Tutorial 2: Train MoCo on CIFAR-10

In this tutorial, we will train a model based on the MoCo Paper Momentum Contrast for Unsupervised Visual Representation Learning.

When training self-supervised models using contrastive loss we usually face one big problem. To get good results, we need many negative examples for the contrastive loss to work. Therefore, we need a large batch size. However, not everyone has access to a cluster full of GPUs or TPUs. To solve this problem, alternative approaches have been developed. Some of them use a memory bank to store old negative examples we can query to compensate for the smaller batch size. MoCo takes this approach one step further by including a momentum encoder.

We use the CIFAR-10 dataset for this tutorial.

In this tutorial you will learn:

  • How to use lightly to load a dataset and train a model

  • How to create a MoCo model with a memory bank

  • How to use the pre-trained model after self-supervised learning for a transfer learning task

Imports

Import the Python frameworks we need for this tutorial. Make sure you have lightly installed.

pip install lightly
import copy

import pytorch_lightning as pl
import torch
import torch.nn as nn
import torchvision

from lightly.data import LightlyDataset
from lightly.loss import NTXentLoss
from lightly.models import ResNetGenerator
from lightly.models.modules.heads import MoCoProjectionHead
from lightly.models.utils import (
    batch_shuffle,
    batch_unshuffle,
    deactivate_requires_grad,
    update_momentum,
)
from lightly.transforms import MoCoV2Transform, utils

Configuration

We set some configuration parameters for our experiment. Feel free to change them and analyze the effect.

The default configuration uses a batch size of 512. This requires around 6.4GB of GPU memory. When training for 100 epochs you should achieve around 73% test set accuracy. When training for 200 epochs accuracy increases to about 80%.

num_workers = 8
batch_size = 512
memory_bank_size = 4096
seed = 1
max_epochs = 100

Replace the path with the location of your CIFAR-10 dataset. We assume we have a train folder with subfolders for each class and .png images inside.

You can download CIFAR-10 in folders from Kaggle.

# The dataset structure should be like this:
# cifar10/train/
#  L airplane/
#    L 10008_airplane.png
#    L ...
#  L automobile/
#  L bird/
#  L cat/
#  L deer/
#  L dog/
#  L frog/
#  L horse/
#  L ship/
#  L truck/
path_to_train = "/datasets/cifar10/train/"
path_to_test = "/datasets/cifar10/test/"

Let’s set the seed to ensure reproducibility of the experiments

pl.seed_everything(seed)
1

Setup data augmentations and loaders

We start with our data preprocessing pipeline. We can implement augmentations from the MoCo paper using the transforms provided by lightly. Images from the CIFAR-10 dataset have a resolution of 32x32 pixels. Let’s use this resolution to train our model.

Note

We could use a higher input resolution to train our model. However, since the original resolution of CIFAR-10 images is low there is no real value in increasing the resolution. A higher resolution results in higher memory consumption and to compensate for that we would need to reduce the batch size.

# disable blur because we're working with tiny images
transform = MoCoV2Transform(
    input_size=32,
    gaussian_blur=0.0,
)

We don’t want any augmentation for our test data. Therefore, we create custom, torchvision based data transformations. Let’s ensure the size is correct and we normalize the data in the same way as we do with the training data.

# Augmentations typically used to train on cifar-10
train_classifier_transforms = torchvision.transforms.Compose(
    [
        torchvision.transforms.RandomCrop(32, padding=4),
        torchvision.transforms.RandomHorizontalFlip(),
        torchvision.transforms.ToTensor(),
        torchvision.transforms.Normalize(
            mean=utils.IMAGENET_NORMALIZE["mean"],
            std=utils.IMAGENET_NORMALIZE["std"],
        ),
    ]
)

# No additional augmentations for the test set
test_transforms = torchvision.transforms.Compose(
    [
        torchvision.transforms.Resize((32, 32)),
        torchvision.transforms.ToTensor(),
        torchvision.transforms.Normalize(
            mean=utils.IMAGENET_NORMALIZE["mean"],
            std=utils.IMAGENET_NORMALIZE["std"],
        ),
    ]
)

# We use the moco augmentations for training moco
dataset_train_moco = LightlyDataset(input_dir=path_to_train, transform=transform)

# Since we also train a linear classifier on the pre-trained moco model we
# reuse the test augmentations here (MoCo augmentations are very strong and
# usually reduce accuracy of models which are not used for contrastive learning.
# Our linear layer will be trained using cross entropy loss and labels provided
# by the dataset. Therefore we chose light augmentations.)
dataset_train_classifier = LightlyDataset(
    input_dir=path_to_train, transform=train_classifier_transforms
)

dataset_test = LightlyDataset(input_dir=path_to_test, transform=test_transforms)

Create the dataloaders to load and preprocess the data in the background.

dataloader_train_moco = torch.utils.data.DataLoader(
    dataset_train_moco,
    batch_size=batch_size,
    shuffle=True,
    drop_last=True,
    num_workers=num_workers,
)

dataloader_train_classifier = torch.utils.data.DataLoader(
    dataset_train_classifier,
    batch_size=batch_size,
    shuffle=True,
    drop_last=True,
    num_workers=num_workers,
)

dataloader_test = torch.utils.data.DataLoader(
    dataset_test,
    batch_size=batch_size,
    shuffle=False,
    drop_last=False,
    num_workers=num_workers,
)

Create the MoCo Lightning Module

Now we create our MoCo model. We use PyTorch Lightning to train our model. We follow the specification of the lightning module. In this example we set the number of features for the hidden dimension to 512. The momentum for the Momentum Encoder is set to 0.99 (default is 0.999) since other reports show that this works better for Cifar-10.

For the backbone we use the lightly variant of a resnet-18. You can use another model following our playground to use custom backbones.

Note

We use a split batch norm to simulate multi-gpu behaviour. Combined with the use of batch shuffling, this prevents the model from communicating through the batch norm layers.

class MocoModel(pl.LightningModule):
    def __init__(self):
        super().__init__()

        # create a ResNet backbone and remove the classification head
        resnet = ResNetGenerator("resnet-18", 1, num_splits=8)
        self.backbone = nn.Sequential(
            *list(resnet.children())[:-1],
            nn.AdaptiveAvgPool2d(1),
        )

        # create a moco model based on ResNet
        self.projection_head = MoCoProjectionHead(512, 512, 128)
        self.backbone_momentum = copy.deepcopy(self.backbone)
        self.projection_head_momentum = copy.deepcopy(self.projection_head)
        deactivate_requires_grad(self.backbone_momentum)
        deactivate_requires_grad(self.projection_head_momentum)

        # create our loss with the optional memory bank
        self.criterion = NTXentLoss(
            temperature=0.1, memory_bank_size=(memory_bank_size, 128)
        )

    def training_step(self, batch, batch_idx):
        (x_q, x_k), _, _ = batch

        # update momentum
        update_momentum(self.backbone, self.backbone_momentum, 0.99)
        update_momentum(self.projection_head, self.projection_head_momentum, 0.99)

        # get queries
        q = self.backbone(x_q).flatten(start_dim=1)
        q = self.projection_head(q)

        # get keys
        k, shuffle = batch_shuffle(x_k)
        k = self.backbone_momentum(k).flatten(start_dim=1)
        k = self.projection_head_momentum(k)
        k = batch_unshuffle(k, shuffle)

        loss = self.criterion(q, k)
        self.log("train_loss_ssl", loss)
        return loss

    def on_train_epoch_end(self):
        self.custom_histogram_weights()

    # We provide a helper method to log weights in tensorboard
    # which is useful for debugging.
    def custom_histogram_weights(self):
        for name, params in self.named_parameters():
            self.logger.experiment.add_histogram(name, params, self.current_epoch)

    def configure_optimizers(self):
        optim = torch.optim.SGD(
            self.parameters(),
            lr=6e-2,
            momentum=0.9,
            weight_decay=5e-4,
        )
        scheduler = torch.optim.lr_scheduler.CosineAnnealingLR(optim, max_epochs)
        return [optim], [scheduler]

Create the Classifier Lightning Module

We create a linear classifier using the features we extract using MoCo and train it on the dataset

class Classifier(pl.LightningModule):
    def __init__(self, backbone):
        super().__init__()
        # use the pretrained ResNet backbone
        self.backbone = backbone

        # freeze the backbone
        deactivate_requires_grad(backbone)

        # create a linear layer for our downstream classification model
        self.fc = nn.Linear(512, 10)

        self.criterion = nn.CrossEntropyLoss()
        self.validation_step_outputs = []

    def forward(self, x):
        y_hat = self.backbone(x).flatten(start_dim=1)
        y_hat = self.fc(y_hat)
        return y_hat

    def training_step(self, batch, batch_idx):
        x, y, _ = batch
        y_hat = self.forward(x)
        loss = self.criterion(y_hat, y)
        self.log("train_loss_fc", loss)
        return loss

    def on_train_epoch_end(self):
        self.custom_histogram_weights()

    # We provide a helper method to log weights in tensorboard
    # which is useful for debugging.
    def custom_histogram_weights(self):
        for name, params in self.named_parameters():
            self.logger.experiment.add_histogram(name, params, self.current_epoch)

    def validation_step(self, batch, batch_idx):
        x, y, _ = batch
        y_hat = self.forward(x)
        y_hat = torch.nn.functional.softmax(y_hat, dim=1)

        # calculate number of correct predictions
        _, predicted = torch.max(y_hat, 1)
        num = predicted.shape[0]
        correct = (predicted == y).float().sum()
        self.validation_step_outputs.append((num, correct))
        return num, correct

    def on_validation_epoch_end(self):
        # calculate and log top1 accuracy
        if self.validation_step_outputs:
            total_num = 0
            total_correct = 0
            for num, correct in self.validation_step_outputs:
                total_num += num
                total_correct += correct
            acc = total_correct / total_num
            self.log("val_acc", acc, on_epoch=True, prog_bar=True)
            self.validation_step_outputs.clear()

    def configure_optimizers(self):
        optim = torch.optim.SGD(self.fc.parameters(), lr=30.0)
        scheduler = torch.optim.lr_scheduler.CosineAnnealingLR(optim, max_epochs)
        return [optim], [scheduler]

Train the MoCo model

We can instantiate the model and train it using the lightning trainer.

model = MocoModel()
trainer = pl.Trainer(max_epochs=max_epochs, devices=1, accelerator="gpu")
trainer.fit(model, dataloader_train_moco)
/datasets/actions-runner/core_gpu_runner_01/hostedtoolcache/Python/3.10.13/x64/lib/python3.10/site-packages/lightning_fabric/plugins/environments/slurm.py:165: PossibleUserWarning: The `srun` command is available on your system but is not used. HINT: If your intention is to run Lightning on SLURM, prepend your python command with `srun` like so: srun python /datasets/actions-runner/core_gpu_runner_01/_work/li ...
  rank_zero_warn(

Training: 0it [00:00, ?it/s]
Training:   0%|          | 0/97 [00:00<?, ?it/s]
Epoch 0:   0%|          | 0/97 [00:00<?, ?it/s]
Epoch 0:   1%|          | 1/97 [00:02<04:23,  2.74s/it]
Epoch 0:   1%|          | 1/97 [00:02<04:28,  2.80s/it, loss=0.6, v_num=0]
Epoch 0:   2%|▏         | 2/97 [00:02<02:16,  1.44s/it, loss=0.6, v_num=0]
Epoch 0:   2%|▏         | 2/97 [00:02<02:19,  1.47s/it, loss=3.43, v_num=0]
Epoch 0:   3%|▎         | 3/97 [00:03<01:34,  1.00s/it, loss=3.43, v_num=0]
Epoch 0:   3%|▎         | 3/97 [00:03<01:35,  1.02s/it, loss=4.6, v_num=0]
Epoch 0:   4%|▍         | 4/97 [00:03<01:12,  1.27it/s, loss=4.6, v_num=0]
Epoch 0:   4%|▍         | 4/97 [00:03<01:14,  1.25it/s, loss=5.26, v_num=0]
Epoch 0:   5%|▌         | 5/97 [00:03<01:00,  1.53it/s, loss=5.26, v_num=0]
Epoch 0:   5%|▌         | 5/97 [00:03<01:01,  1.50it/s, loss=5.7, v_num=0]
Epoch 0:   6%|▌         | 6/97 [00:03<00:51,  1.76it/s, loss=5.7, v_num=0]
Epoch 0:   6%|▌         | 6/97 [00:03<00:52,  1.73it/s, loss=6.02, v_num=0]
Epoch 0:   7%|▋         | 7/97 [00:03<00:45,  1.98it/s, loss=6.02, v_num=0]
Epoch 0:   7%|▋         | 7/97 [00:03<00:46,  1.95it/s, loss=6.27, v_num=0]
Epoch 0:   8%|▊         | 8/97 [00:03<00:40,  2.18it/s, loss=6.27, v_num=0]
Epoch 0:   8%|▊         | 8/97 [00:03<00:41,  2.15it/s, loss=6.47, v_num=0]
Epoch 0:   9%|▉         | 9/97 [00:03<00:37,  2.36it/s, loss=6.47, v_num=0]
Epoch 0:   9%|▉         | 9/97 [00:03<00:37,  2.32it/s, loss=6.62, v_num=0]
Epoch 0:  10%|█         | 10/97 [00:03<00:34,  2.53it/s, loss=6.62, v_num=0]
Epoch 0:  10%|█         | 10/97 [00:04<00:34,  2.50it/s, loss=6.73, v_num=0]
Epoch 0:  11%|█▏        | 11/97 [00:04<00:31,  2.69it/s, loss=6.73, v_num=0]
Epoch 0:  11%|█▏        | 11/97 [00:04<00:32,  2.66it/s, loss=6.82, v_num=0]
Epoch 0:  12%|█▏        | 12/97 [00:04<00:29,  2.85it/s, loss=6.82, v_num=0]
Epoch 0:  12%|█▏        | 12/97 [00:04<00:30,  2.81it/s, loss=6.88, v_num=0]
Epoch 0:  13%|█▎        | 13/97 [00:04<00:28,  2.99it/s, loss=6.88, v_num=0]
Epoch 0:  13%|█▎        | 13/97 [00:04<00:28,  2.95it/s, loss=6.93, v_num=0]
Epoch 0:  14%|█▍        | 14/97 [00:04<00:26,  3.12it/s, loss=6.93, v_num=0]
Epoch 0:  14%|█▍        | 14/97 [00:04<00:26,  3.08it/s, loss=6.96, v_num=0]
Epoch 0:  15%|█▌        | 15/97 [00:04<00:25,  3.25it/s, loss=6.96, v_num=0]
Epoch 0:  15%|█▌        | 15/97 [00:04<00:25,  3.21it/s, loss=6.98, v_num=0]
Epoch 0:  16%|█▋        | 16/97 [00:04<00:24,  3.37it/s, loss=6.98, v_num=0]
Epoch 0:  16%|█▋        | 16/97 [00:04<00:24,  3.33it/s, loss=6.99, v_num=0]
Epoch 0:  18%|█▊        | 17/97 [00:04<00:22,  3.48it/s, loss=6.99, v_num=0]
Epoch 0:  18%|█▊        | 17/97 [00:04<00:23,  3.44it/s, loss=6.99, v_num=0]
Epoch 0:  19%|█▊        | 18/97 [00:05<00:22,  3.58it/s, loss=6.99, v_num=0]
Epoch 0:  19%|█▊        | 18/97 [00:05<00:22,  3.55it/s, loss=6.99, v_num=0]
Epoch 0:  20%|█▉        | 19/97 [00:05<00:21,  3.69it/s, loss=6.99, v_num=0]
Epoch 0:  20%|█▉        | 19/97 [00:05<00:21,  3.65it/s, loss=6.99, v_num=0]
Epoch 0:  21%|██        | 20/97 [00:05<00:20,  3.78it/s, loss=6.99, v_num=0]
Epoch 0:  21%|██        | 20/97 [00:05<00:20,  3.74it/s, loss=6.98, v_num=0]
Epoch 0:  22%|██▏       | 21/97 [00:05<00:19,  3.87it/s, loss=6.98, v_num=0]
Epoch 0:  22%|██▏       | 21/97 [00:05<00:19,  3.83it/s, loss=7.3, v_num=0]
Epoch 0:  23%|██▎       | 22/97 [00:05<00:18,  3.96it/s, loss=7.3, v_num=0]
Epoch 0:  23%|██▎       | 22/97 [00:05<00:19,  3.92it/s, loss=7.32, v_num=0]
Epoch 0:  24%|██▎       | 23/97 [00:05<00:18,  4.04it/s, loss=7.32, v_num=0]
Epoch 0:  24%|██▎       | 23/97 [00:05<00:18,  4.00it/s, loss=7.31, v_num=0]
Epoch 0:  25%|██▍       | 24/97 [00:05<00:17,  4.12it/s, loss=7.31, v_num=0]
Epoch 0:  25%|██▍       | 24/97 [00:05<00:17,  4.08it/s, loss=7.29, v_num=0]
Epoch 0:  26%|██▌       | 25/97 [00:05<00:17,  4.19it/s, loss=7.29, v_num=0]
Epoch 0:  26%|██▌       | 25/97 [00:06<00:17,  4.15it/s, loss=7.25, v_num=0]
Epoch 0:  27%|██▋       | 26/97 [00:06<00:16,  4.25it/s, loss=7.25, v_num=0]
Epoch 0:  27%|██▋       | 26/97 [00:06<00:16,  4.22it/s, loss=7.21, v_num=0]
Epoch 0:  28%|██▊       | 27/97 [00:06<00:16,  4.32it/s, loss=7.21, v_num=0]
Epoch 0:  28%|██▊       | 27/97 [00:06<00:16,  4.29it/s, loss=7.15, v_num=0]
Epoch 0:  29%|██▉       | 28/97 [00:06<00:15,  4.39it/s, loss=7.15, v_num=0]
Epoch 0:  29%|██▉       | 28/97 [00:06<00:15,  4.35it/s, loss=7.1, v_num=0]
Epoch 0:  30%|██▉       | 29/97 [00:06<00:15,  4.45it/s, loss=7.1, v_num=0]
Epoch 0:  30%|██▉       | 29/97 [00:06<00:15,  4.41it/s, loss=7.04, v_num=0]
Epoch 0:  31%|███       | 30/97 [00:06<00:14,  4.51it/s, loss=7.04, v_num=0]
Epoch 0:  31%|███       | 30/97 [00:06<00:14,  4.47it/s, loss=6.99, v_num=0]
Epoch 0:  32%|███▏      | 31/97 [00:06<00:14,  4.57it/s, loss=6.99, v_num=0]
Epoch 0:  32%|███▏      | 31/97 [00:06<00:14,  4.53it/s, loss=6.94, v_num=0]
Epoch 0:  33%|███▎      | 32/97 [00:06<00:14,  4.62it/s, loss=6.94, v_num=0]
Epoch 0:  33%|███▎      | 32/97 [00:06<00:14,  4.59it/s, loss=6.9, v_num=0]
Epoch 0:  34%|███▍      | 33/97 [00:07<00:13,  4.68it/s, loss=6.9, v_num=0]
Epoch 0:  34%|███▍      | 33/97 [00:07<00:13,  4.64it/s, loss=6.86, v_num=0]
Epoch 0:  35%|███▌      | 34/97 [00:07<00:13,  4.73it/s, loss=6.86, v_num=0]
Epoch 0:  35%|███▌      | 34/97 [00:07<00:13,  4.69it/s, loss=6.83, v_num=0]
Epoch 0:  36%|███▌      | 35/97 [00:07<00:12,  4.77it/s, loss=6.83, v_num=0]
Epoch 0:  36%|███▌      | 35/97 [00:07<00:13,  4.74it/s, loss=6.81, v_num=0]
Epoch 0:  37%|███▋      | 36/97 [00:07<00:12,  4.82it/s, loss=6.81, v_num=0]
Epoch 0:  37%|███▋      | 36/97 [00:07<00:12,  4.79it/s, loss=6.79, v_num=0]
Epoch 0:  38%|███▊      | 37/97 [00:07<00:12,  4.87it/s, loss=6.79, v_num=0]
Epoch 0:  38%|███▊      | 37/97 [00:07<00:12,  4.83it/s, loss=6.77, v_num=0]
Epoch 0:  39%|███▉      | 38/97 [00:07<00:12,  4.91it/s, loss=6.77, v_num=0]
Epoch 0:  39%|███▉      | 38/97 [00:07<00:12,  4.88it/s, loss=6.76, v_num=0]
Epoch 0:  40%|████      | 39/97 [00:07<00:11,  4.96it/s, loss=6.76, v_num=0]
Epoch 0:  40%|████      | 39/97 [00:07<00:11,  4.92it/s, loss=6.75, v_num=0]
Epoch 0:  41%|████      | 40/97 [00:08<00:11,  5.00it/s, loss=6.75, v_num=0]
Epoch 0:  41%|████      | 40/97 [00:08<00:11,  4.96it/s, loss=6.75, v_num=0]
Epoch 0:  42%|████▏     | 41/97 [00:08<00:11,  5.04it/s, loss=6.75, v_num=0]
Epoch 0:  42%|████▏     | 41/97 [00:08<00:11,  5.00it/s, loss=6.75, v_num=0]
Epoch 0:  43%|████▎     | 42/97 [00:08<00:10,  5.08it/s, loss=6.75, v_num=0]
Epoch 0:  43%|████▎     | 42/97 [00:08<00:10,  5.04it/s, loss=6.75, v_num=0]
Epoch 0:  44%|████▍     | 43/97 [00:08<00:10,  5.12it/s, loss=6.75, v_num=0]
Epoch 0:  44%|████▍     | 43/97 [00:08<00:10,  5.08it/s, loss=6.76, v_num=0]
Epoch 0:  45%|████▌     | 44/97 [00:08<00:10,  5.15it/s, loss=6.76, v_num=0]
Epoch 0:  45%|████▌     | 44/97 [00:08<00:10,  5.12it/s, loss=6.77, v_num=0]
Epoch 0:  46%|████▋     | 45/97 [00:08<00:10,  5.19it/s, loss=6.77, v_num=0]
Epoch 0:  46%|████▋     | 45/97 [00:08<00:10,  5.15it/s, loss=6.78, v_num=0]
Epoch 0:  47%|████▋     | 46/97 [00:08<00:09,  5.22it/s, loss=6.78, v_num=0]
Epoch 0:  47%|████▋     | 46/97 [00:08<00:09,  5.19it/s, loss=6.78, v_num=0]
Epoch 0:  48%|████▊     | 47/97 [00:08<00:09,  5.26it/s, loss=6.78, v_num=0]
Epoch 0:  48%|████▊     | 47/97 [00:09<00:09,  5.22it/s, loss=6.79, v_num=0]
Epoch 0:  49%|████▉     | 48/97 [00:09<00:09,  5.29it/s, loss=6.79, v_num=0]
Epoch 0:  49%|████▉     | 48/97 [00:09<00:09,  5.25it/s, loss=6.8, v_num=0]
Epoch 0:  51%|█████     | 49/97 [00:09<00:09,  5.32it/s, loss=6.8, v_num=0]
Epoch 0:  51%|█████     | 49/97 [00:09<00:09,  5.29it/s, loss=6.81, v_num=0]
Epoch 0:  52%|█████▏    | 50/97 [00:09<00:08,  5.35it/s, loss=6.81, v_num=0]
Epoch 0:  52%|█████▏    | 50/97 [00:09<00:08,  5.32it/s, loss=6.82, v_num=0]
Epoch 0:  53%|█████▎    | 51/97 [00:09<00:08,  5.38it/s, loss=6.82, v_num=0]
Epoch 0:  53%|█████▎    | 51/97 [00:09<00:08,  5.35it/s, loss=6.83, v_num=0]
Epoch 0:  54%|█████▎    | 52/97 [00:09<00:08,  5.41it/s, loss=6.83, v_num=0]
Epoch 0:  54%|█████▎    | 52/97 [00:09<00:08,  5.37it/s, loss=6.84, v_num=0]
Epoch 0:  55%|█████▍    | 53/97 [00:09<00:08,  5.43it/s, loss=6.84, v_num=0]
Epoch 0:  55%|█████▍    | 53/97 [00:09<00:08,  5.40it/s, loss=6.85, v_num=0]
Epoch 0:  56%|█████▌    | 54/97 [00:09<00:07,  5.46it/s, loss=6.85, v_num=0]
Epoch 0:  56%|█████▌    | 54/97 [00:09<00:07,  5.43it/s, loss=6.86, v_num=0]
Epoch 0:  57%|█████▋    | 55/97 [00:10<00:07,  5.49it/s, loss=6.86, v_num=0]
Epoch 0:  57%|█████▋    | 55/97 [00:10<00:07,  5.46it/s, loss=6.87, v_num=0]
Epoch 0:  58%|█████▊    | 56/97 [00:10<00:07,  5.51it/s, loss=6.87, v_num=0]
Epoch 0:  58%|█████▊    | 56/97 [00:10<00:07,  5.48it/s, loss=6.88, v_num=0]
Epoch 0:  59%|█████▉    | 57/97 [00:10<00:07,  5.54it/s, loss=6.88, v_num=0]
Epoch 0:  59%|█████▉    | 57/97 [00:10<00:07,  5.51it/s, loss=6.88, v_num=0]
Epoch 0:  60%|█████▉    | 58/97 [00:10<00:07,  5.56it/s, loss=6.88, v_num=0]
Epoch 0:  60%|█████▉    | 58/97 [00:10<00:07,  5.53it/s, loss=6.89, v_num=0]
Epoch 0:  61%|██████    | 59/97 [00:10<00:06,  5.58it/s, loss=6.89, v_num=0]
Epoch 0:  61%|██████    | 59/97 [00:10<00:06,  5.56it/s, loss=6.9, v_num=0]
Epoch 0:  62%|██████▏   | 60/97 [00:10<00:06,  5.61it/s, loss=6.9, v_num=0]
Epoch 0:  62%|██████▏   | 60/97 [00:10<00:06,  5.58it/s, loss=6.9, v_num=0]
Epoch 0:  63%|██████▎   | 61/97 [00:10<00:06,  5.63it/s, loss=6.9, v_num=0]
Epoch 0:  63%|██████▎   | 61/97 [00:10<00:06,  5.60it/s, loss=6.91, v_num=0]
Epoch 0:  64%|██████▍   | 62/97 [00:10<00:06,  5.65it/s, loss=6.91, v_num=0]
Epoch 0:  64%|██████▍   | 62/97 [00:11<00:06,  5.63it/s, loss=6.91, v_num=0]
Epoch 0:  65%|██████▍   | 63/97 [00:11<00:05,  5.68it/s, loss=6.91, v_num=0]
Epoch 0:  65%|██████▍   | 63/97 [00:11<00:06,  5.65it/s, loss=6.91, v_num=0]
Epoch 0:  66%|██████▌   | 64/97 [00:11<00:05,  5.70it/s, loss=6.91, v_num=0]
Epoch 0:  66%|██████▌   | 64/97 [00:11<00:05,  5.67it/s, loss=6.92, v_num=0]
Epoch 0:  67%|██████▋   | 65/97 [00:11<00:05,  5.72it/s, loss=6.92, v_num=0]
Epoch 0:  67%|██████▋   | 65/97 [00:11<00:05,  5.69it/s, loss=6.92, v_num=0]
Epoch 0:  68%|██████▊   | 66/97 [00:11<00:05,  5.74it/s, loss=6.92, v_num=0]
Epoch 0:  68%|██████▊   | 66/97 [00:11<00:05,  5.71it/s, loss=6.92, v_num=0]
Epoch 0:  69%|██████▉   | 67/97 [00:11<00:05,  5.76it/s, loss=6.92, v_num=0]
Epoch 0:  69%|██████▉   | 67/97 [00:11<00:05,  5.73it/s, loss=6.92, v_num=0]
Epoch 0:  70%|███████   | 68/97 [00:11<00:05,  5.77it/s, loss=6.92, v_num=0]
Epoch 0:  70%|███████   | 68/97 [00:11<00:05,  5.75it/s, loss=6.91, v_num=0]
Epoch 0:  71%|███████   | 69/97 [00:11<00:04,  5.80it/s, loss=6.91, v_num=0]
Epoch 0:  71%|███████   | 69/97 [00:11<00:04,  5.77it/s, loss=6.92, v_num=0]
Epoch 0:  72%|███████▏  | 70/97 [00:12<00:04,  5.81it/s, loss=6.92, v_num=0]
Epoch 0:  72%|███████▏  | 70/97 [00:12<00:04,  5.79it/s, loss=6.91, v_num=0]
Epoch 0:  73%|███████▎  | 71/97 [00:12<00:04,  5.83it/s, loss=6.91, v_num=0]
Epoch 0:  73%|███████▎  | 71/97 [00:12<00:04,  5.80it/s, loss=6.91, v_num=0]
Epoch 0:  74%|███████▍  | 72/97 [00:12<00:04,  5.85it/s, loss=6.91, v_num=0]
Epoch 0:  74%|███████▍  | 72/97 [00:12<00:04,  5.82it/s, loss=6.91, v_num=0]
Epoch 0:  75%|███████▌  | 73/97 [00:12<00:04,  5.87it/s, loss=6.91, v_num=0]
Epoch 0:  75%|███████▌  | 73/97 [00:12<00:04,  5.84it/s, loss=6.91, v_num=0]
Epoch 0:  76%|███████▋  | 74/97 [00:12<00:03,  5.88it/s, loss=6.91, v_num=0]
Epoch 0:  76%|███████▋  | 74/97 [00:12<00:03,  5.86it/s, loss=6.91, v_num=0]
Epoch 0:  77%|███████▋  | 75/97 [00:12<00:03,  5.90it/s, loss=6.91, v_num=0]
Epoch 0:  77%|███████▋  | 75/97 [00:12<00:03,  5.87it/s, loss=6.91, v_num=0]
Epoch 0:  78%|███████▊  | 76/97 [00:12<00:03,  5.92it/s, loss=6.91, v_num=0]
Epoch 0:  78%|███████▊  | 76/97 [00:12<00:03,  5.89it/s, loss=6.91, v_num=0]
Epoch 0:  79%|███████▉  | 77/97 [00:12<00:03,  5.93it/s, loss=6.91, v_num=0]
Epoch 0:  79%|███████▉  | 77/97 [00:13<00:03,  5.91it/s, loss=6.91, v_num=0]
Epoch 0:  80%|████████  | 78/97 [00:13<00:03,  5.95it/s, loss=6.91, v_num=0]
Epoch 0:  80%|████████  | 78/97 [00:13<00:03,  5.92it/s, loss=6.91, v_num=0]
Epoch 0:  81%|████████▏ | 79/97 [00:13<00:03,  5.96it/s, loss=6.91, v_num=0]
Epoch 0:  81%|████████▏ | 79/97 [00:13<00:03,  5.94it/s, loss=6.91, v_num=0]
Epoch 0:  82%|████████▏ | 80/97 [00:13<00:02,  5.98it/s, loss=6.91, v_num=0]
Epoch 0:  82%|████████▏ | 80/97 [00:13<00:02,  5.95it/s, loss=6.91, v_num=0]
Epoch 0:  84%|████████▎ | 81/97 [00:13<00:02,  5.99it/s, loss=6.91, v_num=0]
Epoch 0:  84%|████████▎ | 81/97 [00:13<00:02,  5.97it/s, loss=6.91, v_num=0]
Epoch 0:  85%|████████▍ | 82/97 [00:13<00:02,  6.00it/s, loss=6.91, v_num=0]
Epoch 0:  85%|████████▍ | 82/97 [00:13<00:02,  5.98it/s, loss=6.9, v_num=0]
Epoch 0:  86%|████████▌ | 83/97 [00:13<00:02,  6.02it/s, loss=6.9, v_num=0]
Epoch 0:  86%|████████▌ | 83/97 [00:13<00:02,  6.00it/s, loss=6.91, v_num=0]
Epoch 0:  87%|████████▋ | 84/97 [00:13<00:02,  6.03it/s, loss=6.91, v_num=0]
Epoch 0:  87%|████████▋ | 84/97 [00:13<00:02,  6.01it/s, loss=6.9, v_num=0]
Epoch 0:  88%|████████▊ | 85/97 [00:14<00:01,  6.05it/s, loss=6.9, v_num=0]
Epoch 0:  88%|████████▊ | 85/97 [00:14<00:01,  6.02it/s, loss=6.9, v_num=0]
Epoch 0:  89%|████████▊ | 86/97 [00:14<00:01,  6.06it/s, loss=6.9, v_num=0]
Epoch 0:  89%|████████▊ | 86/97 [00:14<00:01,  6.04it/s, loss=6.9, v_num=0]
Epoch 0:  90%|████████▉ | 87/97 [00:14<00:01,  6.08it/s, loss=6.9, v_num=0]
Epoch 0:  90%|████████▉ | 87/97 [00:14<00:01,  6.05it/s, loss=6.91, v_num=0]
Epoch 0:  91%|█████████ | 88/97 [00:14<00:01,  6.09it/s, loss=6.91, v_num=0]
Epoch 0:  91%|█████████ | 88/97 [00:14<00:01,  6.06it/s, loss=6.91, v_num=0]
Epoch 0:  92%|█████████▏| 89/97 [00:14<00:01,  6.10it/s, loss=6.91, v_num=0]
Epoch 0:  92%|█████████▏| 89/97 [00:14<00:01,  6.08it/s, loss=6.91, v_num=0]
Epoch 0:  93%|█████████▎| 90/97 [00:14<00:01,  6.11it/s, loss=6.91, v_num=0]
Epoch 0:  93%|█████████▎| 90/97 [00:14<00:01,  6.09it/s, loss=6.91, v_num=0]
Epoch 0:  94%|█████████▍| 91/97 [00:14<00:00,  6.12it/s, loss=6.91, v_num=0]
Epoch 0:  94%|█████████▍| 91/97 [00:14<00:00,  6.10it/s, loss=6.91, v_num=0]
Epoch 0:  95%|█████████▍| 92/97 [00:14<00:00,  6.14it/s, loss=6.91, v_num=0]
Epoch 0:  95%|█████████▍| 92/97 [00:15<00:00,  6.11it/s, loss=6.91, v_num=0]
Epoch 0:  96%|█████████▌| 93/97 [00:15<00:00,  6.15it/s, loss=6.91, v_num=0]
Epoch 0:  96%|█████████▌| 93/97 [00:15<00:00,  6.13it/s, loss=6.92, v_num=0]
Epoch 0:  97%|█████████▋| 94/97 [00:15<00:00,  6.16it/s, loss=6.92, v_num=0]
Epoch 0:  97%|█████████▋| 94/97 [00:15<00:00,  6.14it/s, loss=6.91, v_num=0]
Epoch 0:  98%|█████████▊| 95/97 [00:15<00:00,  6.17it/s, loss=6.91, v_num=0]
Epoch 0:  98%|█████████▊| 95/97 [00:15<00:00,  6.15it/s, loss=6.91, v_num=0]
Epoch 0:  99%|█████████▉| 96/97 [00:15<00:00,  6.18it/s, loss=6.91, v_num=0]
Epoch 0:  99%|█████████▉| 96/97 [00:15<00:00,  6.16it/s, loss=6.91, v_num=0]
Epoch 0: 100%|██████████| 97/97 [00:15<00:00,  6.19it/s, loss=6.91, v_num=0]
Epoch 0: 100%|██████████| 97/97 [00:15<00:00,  6.17it/s, loss=6.91, v_num=0]
Epoch 0: 100%|██████████| 97/97 [00:15<00:00,  6.17it/s, loss=6.91, v_num=0]
Epoch 0:   0%|          | 0/97 [00:00<?, ?it/s, loss=6.91, v_num=0]
Epoch 1:   0%|          | 0/97 [00:00<?, ?it/s, loss=6.91, v_num=0]
Epoch 1:   1%|          | 1/97 [00:01<02:06,  1.32s/it, loss=6.91, v_num=0]
Epoch 1:   1%|          | 1/97 [00:01<02:11,  1.37s/it, loss=6.92, v_num=0]
Epoch 1:   2%|▏         | 2/97 [00:01<01:08,  1.38it/s, loss=6.92, v_num=0]
Epoch 1:   2%|▏         | 2/97 [00:01<01:11,  1.33it/s, loss=6.92, v_num=0]
Epoch 1:   3%|▎         | 3/97 [00:01<00:49,  1.89it/s, loss=6.92, v_num=0]
Epoch 1:   3%|▎         | 3/97 [00:01<00:51,  1.83it/s, loss=6.93, v_num=0]
Epoch 1:   4%|▍         | 4/97 [00:01<00:40,  2.32it/s, loss=6.93, v_num=0]
Epoch 1:   4%|▍         | 4/97 [00:01<00:41,  2.25it/s, loss=6.93, v_num=0]
Epoch 1:   5%|▌         | 5/97 [00:01<00:34,  2.69it/s, loss=6.93, v_num=0]
Epoch 1:   5%|▌         | 5/97 [00:01<00:35,  2.62it/s, loss=6.94, v_num=0]
Epoch 1:   6%|▌         | 6/97 [00:01<00:30,  3.02it/s, loss=6.94, v_num=0]
Epoch 1:   6%|▌         | 6/97 [00:02<00:30,  2.94it/s, loss=6.94, v_num=0]
Epoch 1:   7%|▋         | 7/97 [00:02<00:27,  3.30it/s, loss=6.94, v_num=0]
Epoch 1:   7%|▋         | 7/97 [00:02<00:27,  3.22it/s, loss=6.94, v_num=0]
Epoch 1:   8%|▊         | 8/97 [00:02<00:25,  3.55it/s, loss=6.94, v_num=0]
Epoch 1:   8%|▊         | 8/97 [00:02<00:25,  3.46it/s, loss=6.95, v_num=0]
Epoch 1:   9%|▉         | 9/97 [00:02<00:23,  3.76it/s, loss=6.95, v_num=0]
Epoch 1:   9%|▉         | 9/97 [00:02<00:23,  3.67it/s, loss=6.95, v_num=0]
Epoch 1:  10%|█         | 10/97 [00:02<00:22,  3.95it/s, loss=6.95, v_num=0]
Epoch 1:  10%|█         | 10/97 [00:02<00:22,  3.87it/s, loss=6.95, v_num=0]
Epoch 1:  11%|█▏        | 11/97 [00:02<00:20,  4.13it/s, loss=6.95, v_num=0]
Epoch 1:  11%|█▏        | 11/97 [00:02<00:21,  4.04it/s, loss=6.96, v_num=0]
Epoch 1:  12%|█▏        | 12/97 [00:02<00:19,  4.29it/s, loss=6.96, v_num=0]
Epoch 1:  12%|█▏        | 12/97 [00:02<00:20,  4.21it/s, loss=6.96, v_num=0]
Epoch 1:  13%|█▎        | 13/97 [00:02<00:18,  4.44it/s, loss=6.96, v_num=0]
Epoch 1:  13%|█▎        | 13/97 [00:02<00:19,  4.35it/s, loss=6.97, v_num=0]
Epoch 1:  14%|█▍        | 14/97 [00:03<00:18,  4.57it/s, loss=6.97, v_num=0]
Epoch 1:  14%|█▍        | 14/97 [00:03<00:18,  4.49it/s, loss=6.98, v_num=0]
Epoch 1:  15%|█▌        | 15/97 [00:03<00:17,  4.70it/s, loss=6.98, v_num=0]
Epoch 1:  15%|█▌        | 15/97 [00:03<00:17,  4.61it/s, loss=6.98, v_num=0]
Epoch 1:  16%|█▋        | 16/97 [00:03<00:16,  4.80it/s, loss=6.98, v_num=0]
Epoch 1:  16%|█▋        | 16/97 [00:03<00:17,  4.73it/s, loss=6.99, v_num=0]
Epoch 1:  18%|█▊        | 17/97 [00:03<00:16,  4.90it/s, loss=6.99, v_num=0]
Epoch 1:  18%|█▊        | 17/97 [00:03<00:16,  4.82it/s, loss=6.99, v_num=0]
Epoch 1:  19%|█▊        | 18/97 [00:03<00:15,  5.00it/s, loss=6.99, v_num=0]
Epoch 1:  19%|█▊        | 18/97 [00:03<00:16,  4.92it/s, loss=7, v_num=0]
Epoch 1:  20%|█▉        | 19/97 [00:03<00:15,  5.08it/s, loss=7, v_num=0]
Epoch 1:  20%|█▉        | 19/97 [00:03<00:15,  5.01it/s, loss=7.01, v_num=0]
Epoch 1:  21%|██        | 20/97 [00:03<00:14,  5.17it/s, loss=7.01, v_num=0]
Epoch 1:  21%|██        | 20/97 [00:03<00:15,  5.10it/s, loss=7.01, v_num=0]
Epoch 1:  22%|██▏       | 21/97 [00:04<00:14,  5.24it/s, loss=7.01, v_num=0]
Epoch 1:  22%|██▏       | 21/97 [00:04<00:14,  5.17it/s, loss=7.02, v_num=0]
Epoch 1:  23%|██▎       | 22/97 [00:04<00:14,  5.32it/s, loss=7.02, v_num=0]
Epoch 1:  23%|██▎       | 22/97 [00:04<00:14,  5.25it/s, loss=7.03, v_num=0]
Epoch 1:  24%|██▎       | 23/97 [00:04<00:13,  5.38it/s, loss=7.03, v_num=0]
Epoch 1:  24%|██▎       | 23/97 [00:04<00:13,  5.31it/s, loss=7.04, v_num=0]
Epoch 1:  25%|██▍       | 24/97 [00:04<00:13,  5.45it/s, loss=7.04, v_num=0]
Epoch 1:  25%|██▍       | 24/97 [00:04<00:13,  5.38it/s, loss=7.04, v_num=0]
Epoch 1:  26%|██▌       | 25/97 [00:04<00:13,  5.51it/s, loss=7.04, v_num=0]
Epoch 1:  26%|██▌       | 25/97 [00:04<00:13,  5.44it/s, loss=7.04, v_num=0]
Epoch 1:  27%|██▋       | 26/97 [00:04<00:12,  5.56it/s, loss=7.04, v_num=0]
Epoch 1:  27%|██▋       | 26/97 [00:04<00:12,  5.50it/s, loss=7.05, v_num=0]
Epoch 1:  28%|██▊       | 27/97 [00:04<00:12,  5.61it/s, loss=7.05, v_num=0]
Epoch 1:  28%|██▊       | 27/97 [00:04<00:12,  5.55it/s, loss=7.06, v_num=0]
Epoch 1:  29%|██▉       | 28/97 [00:04<00:12,  5.66it/s, loss=7.06, v_num=0]
Epoch 1:  29%|██▉       | 28/97 [00:04<00:12,  5.60it/s, loss=7.06, v_num=0]
Epoch 1:  30%|██▉       | 29/97 [00:05<00:11,  5.71it/s, loss=7.06, v_num=0]
Epoch 1:  30%|██▉       | 29/97 [00:05<00:12,  5.65it/s, loss=7.07, v_num=0]
Epoch 1:  31%|███       | 30/97 [00:05<00:11,  5.75it/s, loss=7.07, v_num=0]
Epoch 1:  31%|███       | 30/97 [00:05<00:11,  5.69it/s, loss=7.08, v_num=0]
Epoch 1:  32%|███▏      | 31/97 [00:05<00:11,  5.80it/s, loss=7.08, v_num=0]
Epoch 1:  32%|███▏      | 31/97 [00:05<00:11,  5.74it/s, loss=7.08, v_num=0]
Epoch 1:  33%|███▎      | 32/97 [00:05<00:11,  5.84it/s, loss=7.08, v_num=0]
Epoch 1:  33%|███▎      | 32/97 [00:05<00:11,  5.78it/s, loss=7.09, v_num=0]
Epoch 1:  34%|███▍      | 33/97 [00:05<00:10,  5.88it/s, loss=7.09, v_num=0]
Epoch 1:  34%|███▍      | 33/97 [00:05<00:10,  5.82it/s, loss=7.1, v_num=0]
Epoch 1:  35%|███▌      | 34/97 [00:05<00:10,  5.91it/s, loss=7.1, v_num=0]
Epoch 1:  35%|███▌      | 34/97 [00:05<00:10,  5.86it/s, loss=7.11, v_num=0]
Epoch 1:  36%|███▌      | 35/97 [00:05<00:10,  5.95it/s, loss=7.11, v_num=0]
Epoch 1:  36%|███▌      | 35/97 [00:05<00:10,  5.89it/s, loss=7.12, v_num=0]
Epoch 1:  37%|███▋      | 36/97 [00:06<00:10,  5.98it/s, loss=7.12, v_num=0]
Epoch 1:  37%|███▋      | 36/97 [00:06<00:10,  5.93it/s, loss=7.13, v_num=0]
Epoch 1:  38%|███▊      | 37/97 [00:06<00:09,  6.01it/s, loss=7.13, v_num=0]
Epoch 1:  38%|███▊      | 37/97 [00:06<00:10,  5.96it/s, loss=7.14, v_num=0]
Epoch 1:  39%|███▉      | 38/97 [00:06<00:09,  6.04it/s, loss=7.14, v_num=0]
Epoch 1:  39%|███▉      | 38/97 [00:06<00:09,  5.99it/s, loss=7.14, v_num=0]
Epoch 1:  40%|████      | 39/97 [00:06<00:09,  6.07it/s, loss=7.14, v_num=0]
Epoch 1:  40%|████      | 39/97 [00:06<00:09,  6.02it/s, loss=7.16, v_num=0]
Epoch 1:  41%|████      | 40/97 [00:06<00:09,  6.10it/s, loss=7.16, v_num=0]
Epoch 1:  41%|████      | 40/97 [00:06<00:09,  6.05it/s, loss=7.16, v_num=0]
Epoch 1:  42%|████▏     | 41/97 [00:06<00:09,  6.12it/s, loss=7.16, v_num=0]
Epoch 1:  42%|████▏     | 41/97 [00:06<00:09,  6.08it/s, loss=7.17, v_num=0]
Epoch 1:  43%|████▎     | 42/97 [00:06<00:08,  6.15it/s, loss=7.17, v_num=0]
Epoch 1:  43%|████▎     | 42/97 [00:06<00:09,  6.10it/s, loss=7.17, v_num=0]
Epoch 1:  44%|████▍     | 43/97 [00:06<00:08,  6.18it/s, loss=7.17, v_num=0]
Epoch 1:  44%|████▍     | 43/97 [00:07<00:08,  6.13it/s, loss=7.18, v_num=0]
Epoch 1:  45%|████▌     | 44/97 [00:07<00:08,  6.20it/s, loss=7.18, v_num=0]
Epoch 1:  45%|████▌     | 44/97 [00:07<00:08,  6.15it/s, loss=7.19, v_num=0]
Epoch 1:  46%|████▋     | 45/97 [00:07<00:08,  6.22it/s, loss=7.19, v_num=0]
Epoch 1:  46%|████▋     | 45/97 [00:07<00:08,  6.17it/s, loss=7.2, v_num=0]
Epoch 1:  47%|████▋     | 46/97 [00:07<00:08,  6.24it/s, loss=7.2, v_num=0]
Epoch 1:  47%|████▋     | 46/97 [00:07<00:08,  6.20it/s, loss=7.21, v_num=0]
Epoch 1:  48%|████▊     | 47/97 [00:07<00:07,  6.26it/s, loss=7.21, v_num=0]
Epoch 1:  48%|████▊     | 47/97 [00:07<00:08,  6.22it/s, loss=7.21, v_num=0]
Epoch 1:  49%|████▉     | 48/97 [00:07<00:07,  6.28it/s, loss=7.21, v_num=0]
Epoch 1:  49%|████▉     | 48/97 [00:07<00:07,  6.24it/s, loss=7.21, v_num=0]
Epoch 1:  51%|█████     | 49/97 [00:07<00:07,  6.31it/s, loss=7.21, v_num=0]
Epoch 1:  51%|█████     | 49/97 [00:07<00:07,  6.26it/s, loss=7.22, v_num=0]
Epoch 1:  52%|█████▏    | 50/97 [00:07<00:07,  6.32it/s, loss=7.22, v_num=0]
Epoch 1:  52%|█████▏    | 50/97 [00:07<00:07,  6.28it/s, loss=7.22, v_num=0]
Epoch 1:  53%|█████▎    | 51/97 [00:08<00:07,  6.35it/s, loss=7.22, v_num=0]
Epoch 1:  53%|█████▎    | 51/97 [00:08<00:07,  6.30it/s, loss=7.23, v_num=0]
Epoch 1:  54%|█████▎    | 52/97 [00:08<00:07,  6.36it/s, loss=7.23, v_num=0]
Epoch 1:  54%|█████▎    | 52/97 [00:08<00:07,  6.32it/s, loss=7.23, v_num=0]
Epoch 1:  55%|█████▍    | 53/97 [00:08<00:06,  6.38it/s, loss=7.23, v_num=0]
Epoch 1:  55%|█████▍    | 53/97 [00:08<00:06,  6.34it/s, loss=7.24, v_num=0]
Epoch 1:  56%|█████▌    | 54/97 [00:08<00:06,  6.39it/s, loss=7.24, v_num=0]
Epoch 1:  56%|█████▌    | 54/97 [00:08<00:06,  6.35it/s, loss=7.24, v_num=0]
Epoch 1:  57%|█████▋    | 55/97 [00:08<00:06,  6.41it/s, loss=7.24, v_num=0]
Epoch 1:  57%|█████▋    | 55/97 [00:08<00:06,  6.37it/s, loss=7.24, v_num=0]
Epoch 1:  58%|█████▊    | 56/97 [00:08<00:06,  6.42it/s, loss=7.24, v_num=0]
Epoch 1:  58%|█████▊    | 56/97 [00:08<00:06,  6.38it/s, loss=7.24, v_num=0]
Epoch 1:  59%|█████▉    | 57/97 [00:08<00:06,  6.44it/s, loss=7.24, v_num=0]
Epoch 1:  59%|█████▉    | 57/97 [00:08<00:06,  6.40it/s, loss=7.25, v_num=0]
Epoch 1:  60%|█████▉    | 58/97 [00:08<00:06,  6.45it/s, loss=7.25, v_num=0]
Epoch 1:  60%|█████▉    | 58/97 [00:09<00:06,  6.42it/s, loss=7.25, v_num=0]
Epoch 1:  61%|██████    | 59/97 [00:09<00:05,  6.47it/s, loss=7.25, v_num=0]
Epoch 1:  61%|██████    | 59/97 [00:09<00:05,  6.43it/s, loss=7.25, v_num=0]
Epoch 1:  62%|██████▏   | 60/97 [00:09<00:05,  6.48it/s, loss=7.25, v_num=0]
Epoch 1:  62%|██████▏   | 60/97 [00:09<00:05,  6.44it/s, loss=7.25, v_num=0]
Epoch 1:  63%|██████▎   | 61/97 [00:09<00:05,  6.50it/s, loss=7.25, v_num=0]
Epoch 1:  63%|██████▎   | 61/97 [00:09<00:05,  6.46it/s, loss=7.26, v_num=0]
Epoch 1:  64%|██████▍   | 62/97 [00:09<00:05,  6.51it/s, loss=7.26, v_num=0]
Epoch 1:  64%|██████▍   | 62/97 [00:09<00:05,  6.47it/s, loss=7.26, v_num=0]
Epoch 1:  65%|██████▍   | 63/97 [00:09<00:05,  6.52it/s, loss=7.26, v_num=0]
Epoch 1:  65%|██████▍   | 63/97 [00:09<00:05,  6.48it/s, loss=7.25, v_num=0]
Epoch 1:  66%|██████▌   | 64/97 [00:09<00:05,  6.53it/s, loss=7.25, v_num=0]
Epoch 1:  66%|██████▌   | 64/97 [00:09<00:05,  6.50it/s, loss=7.25, v_num=0]
Epoch 1:  67%|██████▋   | 65/97 [00:09<00:04,  6.55it/s, loss=7.25, v_num=0]
Epoch 1:  67%|██████▋   | 65/97 [00:09<00:04,  6.51it/s, loss=7.25, v_num=0]
Epoch 1:  68%|██████▊   | 66/97 [00:10<00:04,  6.56it/s, loss=7.25, v_num=0]
Epoch 1:  68%|██████▊   | 66/97 [00:10<00:04,  6.52it/s, loss=7.24, v_num=0]
Epoch 1:  69%|██████▉   | 67/97 [00:10<00:04,  6.57it/s, loss=7.24, v_num=0]
Epoch 1:  69%|██████▉   | 67/97 [00:10<00:04,  6.53it/s, loss=7.24, v_num=0]
Epoch 1:  70%|███████   | 68/97 [00:10<00:04,  6.58it/s, loss=7.24, v_num=0]
Epoch 1:  70%|███████   | 68/97 [00:10<00:04,  6.55it/s, loss=7.24, v_num=0]
Epoch 1:  71%|███████   | 69/97 [00:10<00:04,  6.59it/s, loss=7.24, v_num=0]
Epoch 1:  71%|███████   | 69/97 [00:10<00:04,  6.56it/s, loss=7.24, v_num=0]
Epoch 1:  72%|███████▏  | 70/97 [00:10<00:04,  6.60it/s, loss=7.24, v_num=0]
Epoch 1:  72%|███████▏  | 70/97 [00:10<00:04,  6.57it/s, loss=7.24, v_num=0]
Epoch 1:  73%|███████▎  | 71/97 [00:10<00:03,  6.61it/s, loss=7.24, v_num=0]
Epoch 1:  73%|███████▎  | 71/97 [00:10<00:03,  6.58it/s, loss=7.24, v_num=0]
Epoch 1:  74%|███████▍  | 72/97 [00:10<00:03,  6.62it/s, loss=7.24, v_num=0]
Epoch 1:  74%|███████▍  | 72/97 [00:10<00:03,  6.59it/s, loss=7.24, v_num=0]
Epoch 1:  75%|███████▌  | 73/97 [00:11<00:03,  6.63it/s, loss=7.24, v_num=0]
Epoch 1:  75%|███████▌  | 73/97 [00:11<00:03,  6.60it/s, loss=7.24, v_num=0]
Epoch 1:  76%|███████▋  | 74/97 [00:11<00:03,  6.64it/s, loss=7.24, v_num=0]
Epoch 1:  76%|███████▋  | 74/97 [00:11<00:03,  6.61it/s, loss=7.24, v_num=0]
Epoch 1:  77%|███████▋  | 75/97 [00:11<00:03,  6.65it/s, loss=7.24, v_num=0]
Epoch 1:  77%|███████▋  | 75/97 [00:11<00:03,  6.62it/s, loss=7.23, v_num=0]
Epoch 1:  78%|███████▊  | 76/97 [00:11<00:03,  6.66it/s, loss=7.23, v_num=0]
Epoch 1:  78%|███████▊  | 76/97 [00:11<00:03,  6.63it/s, loss=7.22, v_num=0]
Epoch 1:  79%|███████▉  | 77/97 [00:11<00:02,  6.67it/s, loss=7.22, v_num=0]
Epoch 1:  79%|███████▉  | 77/97 [00:11<00:03,  6.63it/s, loss=7.21, v_num=0]
Epoch 1:  80%|████████  | 78/97 [00:11<00:02,  6.68it/s, loss=7.21, v_num=0]
Epoch 1:  80%|████████  | 78/97 [00:11<00:02,  6.64it/s, loss=7.2, v_num=0]
Epoch 1:  81%|████████▏ | 79/97 [00:11<00:02,  6.68it/s, loss=7.2, v_num=0]
Epoch 1:  81%|████████▏ | 79/97 [00:11<00:02,  6.65it/s, loss=7.19, v_num=0]
Epoch 1:  82%|████████▏ | 80/97 [00:11<00:02,  6.69it/s, loss=7.19, v_num=0]
Epoch 1:  82%|████████▏ | 80/97 [00:12<00:02,  6.66it/s, loss=7.19, v_num=0]
Epoch 1:  84%|████████▎ | 81/97 [00:12<00:02,  6.70it/s, loss=7.19, v_num=0]
Epoch 1:  84%|████████▎ | 81/97 [00:12<00:02,  6.67it/s, loss=7.18, v_num=0]
Epoch 1:  85%|████████▍ | 82/97 [00:12<00:02,  6.71it/s, loss=7.18, v_num=0]
Epoch 1:  85%|████████▍ | 82/97 [00:12<00:02,  6.68it/s, loss=7.17, v_num=0]
Epoch 1:  86%|████████▌ | 83/97 [00:12<00:02,  6.72it/s, loss=7.17, v_num=0]
Epoch 1:  86%|████████▌ | 83/97 [00:12<00:02,  6.69it/s, loss=7.17, v_num=0]
Epoch 1:  87%|████████▋ | 84/97 [00:12<00:01,  6.72it/s, loss=7.17, v_num=0]
Epoch 1:  87%|████████▋ | 84/97 [00:12<00:01,  6.70it/s, loss=7.17, v_num=0]
Epoch 1:  88%|████████▊ | 85/97 [00:12<00:01,  6.73it/s, loss=7.17, v_num=0]
Epoch 1:  88%|████████▊ | 85/97 [00:12<00:01,  6.70it/s, loss=7.16, v_num=0]
Epoch 1:  89%|████████▊ | 86/97 [00:12<00:01,  6.74it/s, loss=7.16, v_num=0]
Epoch 1:  89%|████████▊ | 86/97 [00:12<00:01,  6.71it/s, loss=7.15, v_num=0]
Epoch 1:  90%|████████▉ | 87/97 [00:12<00:01,  6.75it/s, loss=7.15, v_num=0]
Epoch 1:  90%|████████▉ | 87/97 [00:12<00:01,  6.72it/s, loss=7.15, v_num=0]
Epoch 1:  91%|█████████ | 88/97 [00:13<00:01,  6.75it/s, loss=7.15, v_num=0]
Epoch 1:  91%|█████████ | 88/97 [00:13<00:01,  6.72it/s, loss=7.14, v_num=0]
Epoch 1:  92%|█████████▏| 89/97 [00:13<00:01,  6.76it/s, loss=7.14, v_num=0]
Epoch 1:  92%|█████████▏| 89/97 [00:13<00:01,  6.73it/s, loss=7.13, v_num=0]
Epoch 1:  93%|█████████▎| 90/97 [00:13<00:01,  6.77it/s, loss=7.13, v_num=0]
Epoch 1:  93%|█████████▎| 90/97 [00:13<00:01,  6.74it/s, loss=7.11, v_num=0]
Epoch 1:  94%|█████████▍| 91/97 [00:13<00:00,  6.78it/s, loss=7.11, v_num=0]
Epoch 1:  94%|█████████▍| 91/97 [00:13<00:00,  6.75it/s, loss=7.1, v_num=0]
Epoch 1:  95%|█████████▍| 92/97 [00:13<00:00,  6.78it/s, loss=7.1, v_num=0]
Epoch 1:  95%|█████████▍| 92/97 [00:13<00:00,  6.75it/s, loss=7.1, v_num=0]
Epoch 1:  96%|█████████▌| 93/97 [00:13<00:00,  6.79it/s, loss=7.1, v_num=0]
Epoch 1:  96%|█████████▌| 93/97 [00:13<00:00,  6.76it/s, loss=7.09, v_num=0]
Epoch 1:  97%|█████████▋| 94/97 [00:13<00:00,  6.79it/s, loss=7.09, v_num=0]
Epoch 1:  97%|█████████▋| 94/97 [00:13<00:00,  6.77it/s, loss=7.08, v_num=0]
Epoch 1:  98%|█████████▊| 95/97 [00:13<00:00,  6.80it/s, loss=7.08, v_num=0]
Epoch 1:  98%|█████████▊| 95/97 [00:14<00:00,  6.77it/s, loss=7.07, v_num=0]
Epoch 1:  99%|█████████▉| 96/97 [00:14<00:00,  6.80it/s, loss=7.07, v_num=0]
Epoch 1:  99%|█████████▉| 96/97 [00:14<00:00,  6.78it/s, loss=7.07, v_num=0]
Epoch 1: 100%|██████████| 97/97 [00:14<00:00,  6.81it/s, loss=7.07, v_num=0]
Epoch 1: 100%|██████████| 97/97 [00:14<00:00,  6.79it/s, loss=7.06, v_num=0]
Epoch 1: 100%|██████████| 97/97 [00:14<00:00,  6.78it/s, loss=7.06, v_num=0]
Epoch 1:   0%|          | 0/97 [00:00<?, ?it/s, loss=7.06, v_num=0]
Epoch 2:   0%|          | 0/97 [00:00<?, ?it/s, loss=7.06, v_num=0]
Epoch 2:   1%|          | 1/97 [00:01<02:13,  1.39s/it, loss=7.06, v_num=0]
Epoch 2:   1%|          | 1/97 [00:01<02:18,  1.44s/it, loss=7.06, v_num=0]
Epoch 2:   2%|▏         | 2/97 [00:01<01:12,  1.31it/s, loss=7.06, v_num=0]
Epoch 2:   2%|▏         | 2/97 [00:01<01:15,  1.26it/s, loss=7.06, v_num=0]
Epoch 2:   3%|▎         | 3/97 [00:01<00:52,  1.81it/s, loss=7.06, v_num=0]
Epoch 2:   3%|▎         | 3/97 [00:01<00:53,  1.75it/s, loss=7.06, v_num=0]
Epoch 2:   4%|▍         | 4/97 [00:01<00:41,  2.23it/s, loss=7.06, v_num=0]
Epoch 2:   4%|▍         | 4/97 [00:01<00:43,  2.16it/s, loss=7.05, v_num=0]
Epoch 2:   5%|▌         | 5/97 [00:01<00:35,  2.59it/s, loss=7.05, v_num=0]
Epoch 2:   5%|▌         | 5/97 [00:01<00:36,  2.52it/s, loss=7.05, v_num=0]
Epoch 2:   6%|▌         | 6/97 [00:02<00:31,  2.91it/s, loss=7.05, v_num=0]
Epoch 2:   6%|▌         | 6/97 [00:02<00:32,  2.83it/s, loss=7.04, v_num=0]
Epoch 2:   7%|▋         | 7/97 [00:02<00:28,  3.18it/s, loss=7.04, v_num=0]
Epoch 2:   7%|▋         | 7/97 [00:02<00:29,  3.10it/s, loss=7.03, v_num=0]
Epoch 2:   8%|▊         | 8/97 [00:02<00:25,  3.43it/s, loss=7.03, v_num=0]
Epoch 2:   8%|▊         | 8/97 [00:02<00:26,  3.35it/s, loss=7.03, v_num=0]
Epoch 2:   9%|▉         | 9/97 [00:02<00:24,  3.64it/s, loss=7.03, v_num=0]
Epoch 2:   9%|▉         | 9/97 [00:02<00:24,  3.56it/s, loss=7.03, v_num=0]
Epoch 2:  10%|█         | 10/97 [00:02<00:22,  3.82it/s, loss=7.03, v_num=0]
Epoch 2:  10%|█         | 10/97 [00:02<00:23,  3.75it/s, loss=7.02, v_num=0]
Epoch 2:  11%|█▏        | 11/97 [00:02<00:21,  4.01it/s, loss=7.02, v_num=0]
Epoch 2:  11%|█▏        | 11/97 [00:02<00:21,  3.93it/s, loss=7.02, v_num=0]
Epoch 2:  12%|█▏        | 12/97 [00:02<00:20,  4.17it/s, loss=7.02, v_num=0]
Epoch 2:  12%|█▏        | 12/97 [00:02<00:20,  4.09it/s, loss=7.02, v_num=0]
Epoch 2:  13%|█▎        | 13/97 [00:03<00:19,  4.31it/s, loss=7.02, v_num=0]
Epoch 2:  13%|█▎        | 13/97 [00:03<00:19,  4.24it/s, loss=7.02, v_num=0]
Epoch 2:  14%|█▍        | 14/97 [00:03<00:18,  4.44it/s, loss=7.02, v_num=0]
Epoch 2:  14%|█▍        | 14/97 [00:03<00:18,  4.37it/s, loss=7.01, v_num=0]
Epoch 2:  15%|█▌        | 15/97 [00:03<00:17,  4.56it/s, loss=7.01, v_num=0]
Epoch 2:  15%|█▌        | 15/97 [00:03<00:18,  4.49it/s, loss=7.01, v_num=0]
Epoch 2:  16%|█▋        | 16/97 [00:03<00:17,  4.68it/s, loss=7.01, v_num=0]
Epoch 2:  16%|█▋        | 16/97 [00:03<00:17,  4.60it/s, loss=7.01, v_num=0]
Epoch 2:  18%|█▊        | 17/97 [00:03<00:16,  4.78it/s, loss=7.01, v_num=0]
Epoch 2:  18%|█▊        | 17/97 [00:03<00:16,  4.71it/s, loss=7.01, v_num=0]
Epoch 2:  19%|█▊        | 18/97 [00:03<00:16,  4.87it/s, loss=7.01, v_num=0]
Epoch 2:  19%|█▊        | 18/97 [00:03<00:16,  4.80it/s, loss=7.01, v_num=0]
Epoch 2:  20%|█▉        | 19/97 [00:03<00:15,  4.97it/s, loss=7.01, v_num=0]
Epoch 2:  20%|█▉        | 19/97 [00:03<00:15,  4.90it/s, loss=7.01, v_num=0]
Epoch 2:  21%|██        | 20/97 [00:03<00:15,  5.05it/s, loss=7.01, v_num=0]
Epoch 2:  21%|██        | 20/97 [00:04<00:15,  4.98it/s, loss=7.01, v_num=0]
Epoch 2:  22%|██▏       | 21/97 [00:04<00:14,  5.13it/s, loss=7.01, v_num=0]
Epoch 2:  22%|██▏       | 21/97 [00:04<00:15,  5.06it/s, loss=7, v_num=0]
Epoch 2:  23%|██▎       | 22/97 [00:04<00:14,  5.20it/s, loss=7, v_num=0]
Epoch 2:  23%|██▎       | 22/97 [00:04<00:14,  5.14it/s, loss=7, v_num=0]
Epoch 2:  24%|██▎       | 23/97 [00:04<00:14,  5.27it/s, loss=7, v_num=0]
Epoch 2:  24%|██▎       | 23/97 [00:04<00:14,  5.20it/s, loss=7, v_num=0]
Epoch 2:  25%|██▍       | 24/97 [00:04<00:13,  5.33it/s, loss=7, v_num=0]
Epoch 2:  25%|██▍       | 24/97 [00:04<00:13,  5.27it/s, loss=7, v_num=0]
Epoch 2:  26%|██▌       | 25/97 [00:04<00:13,  5.39it/s, loss=7, v_num=0]
Epoch 2:  26%|██▌       | 25/97 [00:04<00:13,  5.33it/s, loss=7, v_num=0]
Epoch 2:  27%|██▋       | 26/97 [00:04<00:13,  5.45it/s, loss=7, v_num=0]
Epoch 2:  27%|██▋       | 26/97 [00:04<00:13,  5.39it/s, loss=6.99, v_num=0]
Epoch 2:  28%|██▊       | 27/97 [00:04<00:12,  5.51it/s, loss=6.99, v_num=0]
Epoch 2:  28%|██▊       | 27/97 [00:04<00:12,  5.45it/s, loss=7, v_num=0]
Epoch 2:  29%|██▉       | 28/97 [00:05<00:12,  5.56it/s, loss=7, v_num=0]
Epoch 2:  29%|██▉       | 28/97 [00:05<00:12,  5.50it/s, loss=6.99, v_num=0]
Epoch 2:  30%|██▉       | 29/97 [00:05<00:12,  5.61it/s, loss=6.99, v_num=0]
Epoch 2:  30%|██▉       | 29/97 [00:05<00:12,  5.55it/s, loss=6.99, v_num=0]
Epoch 2:  31%|███       | 30/97 [00:05<00:11,  5.65it/s, loss=6.99, v_num=0]
Epoch 2:  31%|███       | 30/97 [00:05<00:11,  5.60it/s, loss=7, v_num=0]
Epoch 2:  32%|███▏      | 31/97 [00:05<00:11,  5.70it/s, loss=7, v_num=0]
Epoch 2:  32%|███▏      | 31/97 [00:05<00:11,  5.64it/s, loss=7, v_num=0]
Epoch 2:  33%|███▎      | 32/97 [00:05<00:11,  5.74it/s, loss=7, v_num=0]
Epoch 2:  33%|███▎      | 32/97 [00:05<00:11,  5.68it/s, loss=6.99, v_num=0]
Epoch 2:  34%|███▍      | 33/97 [00:05<00:11,  5.78it/s, loss=6.99, v_num=0]
Epoch 2:  34%|███▍      | 33/97 [00:05<00:11,  5.72it/s, loss=6.99, v_num=0]
Epoch 2:  35%|███▌      | 34/97 [00:05<00:10,  5.82it/s, loss=6.99, v_num=0]
Epoch 2:  35%|███▌      | 34/97 [00:05<00:10,  5.76it/s, loss=6.99, v_num=0]
Epoch 2:  36%|███▌      | 35/97 [00:05<00:10,  5.85it/s, loss=6.99, v_num=0]
Epoch 2:  36%|███▌      | 35/97 [00:06<00:10,  5.80it/s, loss=6.99, v_num=0]
Epoch 2:  37%|███▋      | 36/97 [00:06<00:10,  5.89it/s, loss=6.99, v_num=0]
Epoch 2:  37%|███▋      | 36/97 [00:06<00:10,  5.84it/s, loss=7, v_num=0]
Epoch 2:  38%|███▊      | 37/97 [00:06<00:10,  5.92it/s, loss=7, v_num=0]
Epoch 2:  38%|███▊      | 37/97 [00:06<00:10,  5.87it/s, loss=6.99, v_num=0]
Epoch 2:  39%|███▉      | 38/97 [00:06<00:09,  5.95it/s, loss=6.99, v_num=0]
Epoch 2:  39%|███▉      | 38/97 [00:06<00:09,  5.90it/s, loss=6.99, v_num=0]
Epoch 2:  40%|████      | 39/97 [00:06<00:09,  5.98it/s, loss=6.99, v_num=0]
Epoch 2:  40%|████      | 39/97 [00:06<00:09,  5.93it/s, loss=6.99, v_num=0]
Epoch 2:  41%|████      | 40/97 [00:06<00:09,  6.01it/s, loss=6.99, v_num=0]
Epoch 2:  41%|████      | 40/97 [00:06<00:09,  5.96it/s, loss=6.99, v_num=0]
Epoch 2:  42%|████▏     | 41/97 [00:06<00:09,  6.04it/s, loss=6.99, v_num=0]
Epoch 2:  42%|████▏     | 41/97 [00:06<00:09,  5.99it/s, loss=6.99, v_num=0]
Epoch 2:  43%|████▎     | 42/97 [00:06<00:09,  6.06it/s, loss=6.99, v_num=0]
Epoch 2:  43%|████▎     | 42/97 [00:06<00:09,  6.02it/s, loss=6.99, v_num=0]
Epoch 2:  44%|████▍     | 43/97 [00:07<00:08,  6.09it/s, loss=6.99, v_num=0]
Epoch 2:  44%|████▍     | 43/97 [00:07<00:08,  6.04it/s, loss=6.99, v_num=0]
Epoch 2:  45%|████▌     | 44/97 [00:07<00:08,  6.11it/s, loss=6.99, v_num=0]
Epoch 2:  45%|████▌     | 44/97 [00:07<00:08,  6.07it/s, loss=6.99, v_num=0]
Epoch 2:  46%|████▋     | 45/97 [00:07<00:08,  6.14it/s, loss=6.99, v_num=0]
Epoch 2:  46%|████▋     | 45/97 [00:07<00:08,  6.09it/s, loss=6.98, v_num=0]
Epoch 2:  47%|████▋     | 46/97 [00:07<00:08,  6.16it/s, loss=6.98, v_num=0]
Epoch 2:  47%|████▋     | 46/97 [00:07<00:08,  6.12it/s, loss=6.99, v_num=0]
Epoch 2:  48%|████▊     | 47/97 [00:07<00:08,  6.18it/s, loss=6.99, v_num=0]
Epoch 2:  48%|████▊     | 47/97 [00:07<00:08,  6.14it/s, loss=6.99, v_num=0]
Epoch 2:  49%|████▉     | 48/97 [00:07<00:07,  6.20it/s, loss=6.99, v_num=0]
Epoch 2:  49%|████▉     | 48/97 [00:07<00:07,  6.16it/s, loss=6.98, v_num=0]
Epoch 2:  51%|█████     | 49/97 [00:07<00:07,  6.22it/s, loss=6.98, v_num=0]
Epoch 2:  51%|█████     | 49/97 [00:07<00:07,  6.18it/s, loss=6.99, v_num=0]
Epoch 2:  52%|█████▏    | 50/97 [00:08<00:07,  6.24it/s, loss=6.99, v_num=0]
Epoch 2:  52%|█████▏    | 50/97 [00:08<00:07,  6.20it/s, loss=6.98, v_num=0]
Epoch 2:  53%|█████▎    | 51/97 [00:08<00:07,  6.26it/s, loss=6.98, v_num=0]
Epoch 2:  53%|█████▎    | 51/97 [00:08<00:07,  6.22it/s, loss=6.99, v_num=0]
Epoch 2:  54%|█████▎    | 52/97 [00:08<00:07,  6.28it/s, loss=6.99, v_num=0]
Epoch 2:  54%|█████▎    | 52/97 [00:08<00:07,  6.24it/s, loss=6.99, v_num=0]
Epoch 2:  55%|█████▍    | 53/97 [00:08<00:06,  6.29it/s, loss=6.99, v_num=0]
Epoch 2:  55%|█████▍    | 53/97 [00:08<00:07,  6.25it/s, loss=6.99, v_num=0]
Epoch 2:  56%|█████▌    | 54/97 [00:08<00:06,  6.31it/s, loss=6.99, v_num=0]
Epoch 2:  56%|█████▌    | 54/97 [00:08<00:06,  6.27it/s, loss=7, v_num=0]
Epoch 2:  57%|█████▋    | 55/97 [00:08<00:06,  6.33it/s, loss=7, v_num=0]
Epoch 2:  57%|█████▋    | 55/97 [00:08<00:06,  6.29it/s, loss=7, v_num=0]
Epoch 2:  58%|█████▊    | 56/97 [00:08<00:06,  6.34it/s, loss=7, v_num=0]
Epoch 2:  58%|█████▊    | 56/97 [00:08<00:06,  6.30it/s, loss=7, v_num=0]
Epoch 2:  59%|█████▉    | 57/97 [00:08<00:06,  6.36it/s, loss=7, v_num=0]
Epoch 2:  59%|█████▉    | 57/97 [00:09<00:06,  6.32it/s, loss=7.01, v_num=0]
Epoch 2:  60%|█████▉    | 58/97 [00:09<00:06,  6.37it/s, loss=7.01, v_num=0]
Epoch 2:  60%|█████▉    | 58/97 [00:09<00:06,  6.34it/s, loss=7.01, v_num=0]
Epoch 2:  61%|██████    | 59/97 [00:09<00:05,  6.39it/s, loss=7.01, v_num=0]
Epoch 2:  61%|██████    | 59/97 [00:09<00:05,  6.35it/s, loss=7.01, v_num=0]
Epoch 2:  62%|██████▏   | 60/97 [00:09<00:05,  6.40it/s, loss=7.01, v_num=0]
Epoch 2:  62%|██████▏   | 60/97 [00:09<00:05,  6.37it/s, loss=7, v_num=0]
Epoch 2:  63%|██████▎   | 61/97 [00:09<00:05,  6.41it/s, loss=7, v_num=0]
Epoch 2:  63%|██████▎   | 61/97 [00:09<00:05,  6.38it/s, loss=7, v_num=0]
Epoch 2:  64%|██████▍   | 62/97 [00:09<00:05,  6.43it/s, loss=7, v_num=0]
Epoch 2:  64%|██████▍   | 62/97 [00:09<00:05,  6.39it/s, loss=7, v_num=0]
Epoch 2:  65%|██████▍   | 63/97 [00:09<00:05,  6.44it/s, loss=7, v_num=0]
Epoch 2:  65%|██████▍   | 63/97 [00:09<00:05,  6.41it/s, loss=7, v_num=0]
Epoch 2:  66%|██████▌   | 64/97 [00:09<00:05,  6.46it/s, loss=7, v_num=0]
Epoch 2:  66%|██████▌   | 64/97 [00:09<00:05,  6.42it/s, loss=7, v_num=0]
Epoch 2:  67%|██████▋   | 65/97 [00:10<00:04,  6.47it/s, loss=7, v_num=0]
Epoch 2:  67%|██████▋   | 65/97 [00:10<00:04,  6.43it/s, loss=7, v_num=0]
Epoch 2:  68%|██████▊   | 66/97 [00:10<00:04,  6.48it/s, loss=7, v_num=0]
Epoch 2:  68%|██████▊   | 66/97 [00:10<00:04,  6.45it/s, loss=6.99, v_num=0]
Epoch 2:  69%|██████▉   | 67/97 [00:10<00:04,  6.49it/s, loss=6.99, v_num=0]
Epoch 2:  69%|██████▉   | 67/97 [00:10<00:04,  6.46it/s, loss=6.99, v_num=0]
Epoch 2:  70%|███████   | 68/97 [00:10<00:04,  6.51it/s, loss=6.99, v_num=0]
Epoch 2:  70%|███████   | 68/97 [00:10<00:04,  6.47it/s, loss=6.99, v_num=0]
Epoch 2:  71%|███████   | 69/97 [00:10<00:04,  6.52it/s, loss=6.99, v_num=0]
Epoch 2:  71%|███████   | 69/97 [00:10<00:04,  6.48it/s, loss=6.99, v_num=0]
Epoch 2:  72%|███████▏  | 70/97 [00:10<00:04,  6.53it/s, loss=6.99, v_num=0]
Epoch 2:  72%|███████▏  | 70/97 [00:10<00:04,  6.49it/s, loss=7, v_num=0]
Epoch 2:  73%|███████▎  | 71/97 [00:10<00:03,  6.54it/s, loss=7, v_num=0]
Epoch 2:  73%|███████▎  | 71/97 [00:10<00:03,  6.51it/s, loss=6.99, v_num=0]
Epoch 2:  74%|███████▍  | 72/97 [00:10<00:03,  6.55it/s, loss=6.99, v_num=0]
Epoch 2:  74%|███████▍  | 72/97 [00:11<00:03,  6.52it/s, loss=6.99, v_num=0]
Epoch 2:  75%|███████▌  | 73/97 [00:11<00:03,  6.56it/s, loss=6.99, v_num=0]
Epoch 2:  75%|███████▌  | 73/97 [00:11<00:03,  6.53it/s, loss=6.98, v_num=0]
Epoch 2:  76%|███████▋  | 74/97 [00:11<00:03,  6.57it/s, loss=6.98, v_num=0]
Epoch 2:  76%|███████▋  | 74/97 [00:11<00:03,  6.54it/s, loss=6.98, v_num=0]
Epoch 2:  77%|███████▋  | 75/97 [00:11<00:03,  6.58it/s, loss=6.98, v_num=0]
Epoch 2:  77%|███████▋  | 75/97 [00:11<00:03,  6.55it/s, loss=6.97, v_num=0]
Epoch 2:  78%|███████▊  | 76/97 [00:11<00:03,  6.59it/s, loss=6.97, v_num=0]
Epoch 2:  78%|███████▊  | 76/97 [00:11<00:03,  6.56it/s, loss=6.97, v_num=0]
Epoch 2:  79%|███████▉  | 77/97 [00:11<00:03,  6.60it/s, loss=6.97, v_num=0]
Epoch 2:  79%|███████▉  | 77/97 [00:11<00:03,  6.57it/s, loss=6.96, v_num=0]
Epoch 2:  80%|████████  | 78/97 [00:11<00:02,  6.61it/s, loss=6.96, v_num=0]
Epoch 2:  80%|████████  | 78/97 [00:11<00:02,  6.58it/s, loss=6.96, v_num=0]
Epoch 2:  81%|████████▏ | 79/97 [00:11<00:02,  6.62it/s, loss=6.96, v_num=0]
Epoch 2:  81%|████████▏ | 79/97 [00:11<00:02,  6.59it/s, loss=6.96, v_num=0]
Epoch 2:  82%|████████▏ | 80/97 [00:12<00:02,  6.63it/s, loss=6.96, v_num=0]
Epoch 2:  82%|████████▏ | 80/97 [00:12<00:02,  6.59it/s, loss=6.96, v_num=0]
Epoch 2:  84%|████████▎ | 81/97 [00:12<00:02,  6.63it/s, loss=6.96, v_num=0]
Epoch 2:  84%|████████▎ | 81/97 [00:12<00:02,  6.60it/s, loss=6.95, v_num=0]
Epoch 2:  85%|████████▍ | 82/97 [00:12<00:02,  6.64it/s, loss=6.95, v_num=0]
Epoch 2:  85%|████████▍ | 82/97 [00:12<00:02,  6.61it/s, loss=6.96, v_num=0]
Epoch 2:  86%|████████▌ | 83/97 [00:12<00:02,  6.64it/s, loss=6.96, v_num=0]
Epoch 2:  86%|████████▌ | 83/97 [00:12<00:02,  6.62it/s, loss=6.95, v_num=0]
Epoch 2:  87%|████████▋ | 84/97 [00:12<00:01,  6.65it/s, loss=6.95, v_num=0]
Epoch 2:  87%|████████▋ | 84/97 [00:12<00:01,  6.62it/s, loss=6.95, v_num=0]
Epoch 2:  88%|████████▊ | 85/97 [00:12<00:01,  6.66it/s, loss=6.95, v_num=0]
Epoch 2:  88%|████████▊ | 85/97 [00:12<00:01,  6.63it/s, loss=6.95, v_num=0]
Epoch 2:  89%|████████▊ | 86/97 [00:12<00:01,  6.67it/s, loss=6.95, v_num=0]
Epoch 2:  89%|████████▊ | 86/97 [00:12<00:01,  6.64it/s, loss=6.95, v_num=0]
Epoch 2:  90%|████████▉ | 87/97 [00:13<00:01,  6.68it/s, loss=6.95, v_num=0]
Epoch 2:  90%|████████▉ | 87/97 [00:13<00:01,  6.65it/s, loss=6.94, v_num=0]
Epoch 2:  91%|█████████ | 88/97 [00:13<00:01,  6.68it/s, loss=6.94, v_num=0]
Epoch 2:  91%|█████████ | 88/97 [00:13<00:01,  6.66it/s, loss=6.95, v_num=0]
Epoch 2:  92%|█████████▏| 89/97 [00:13<00:01,  6.69it/s, loss=6.95, v_num=0]
Epoch 2:  92%|█████████▏| 89/97 [00:13<00:01,  6.66it/s, loss=6.94, v_num=0]
Epoch 2:  93%|█████████▎| 90/97 [00:13<00:01,  6.70it/s, loss=6.94, v_num=0]
Epoch 2:  93%|█████████▎| 90/97 [00:13<00:01,  6.67it/s, loss=6.93, v_num=0]
Epoch 2:  94%|█████████▍| 91/97 [00:13<00:00,  6.71it/s, loss=6.93, v_num=0]
Epoch 2:  94%|█████████▍| 91/97 [00:13<00:00,  6.68it/s, loss=6.93, v_num=0]
Epoch 2:  95%|█████████▍| 92/97 [00:13<00:00,  6.71it/s, loss=6.93, v_num=0]
Epoch 2:  95%|█████████▍| 92/97 [00:13<00:00,  6.69it/s, loss=6.93, v_num=0]
Epoch 2:  96%|█████████▌| 93/97 [00:13<00:00,  6.72it/s, loss=6.93, v_num=0]
Epoch 2:  96%|█████████▌| 93/97 [00:13<00:00,  6.69it/s, loss=6.92, v_num=0]
Epoch 2:  97%|█████████▋| 94/97 [00:13<00:00,  6.73it/s, loss=6.92, v_num=0]
Epoch 2:  97%|█████████▋| 94/97 [00:14<00:00,  6.70it/s, loss=6.92, v_num=0]
Epoch 2:  98%|█████████▊| 95/97 [00:14<00:00,  6.73it/s, loss=6.92, v_num=0]
Epoch 2:  98%|█████████▊| 95/97 [00:14<00:00,  6.71it/s, loss=6.92, v_num=0]
Epoch 2:  99%|█████████▉| 96/97 [00:14<00:00,  6.74it/s, loss=6.92, v_num=0]
Epoch 2:  99%|█████████▉| 96/97 [00:14<00:00,  6.71it/s, loss=6.91, v_num=0]
Epoch 2: 100%|██████████| 97/97 [00:14<00:00,  6.74it/s, loss=6.91, v_num=0]
Epoch 2: 100%|██████████| 97/97 [00:14<00:00,  6.72it/s, loss=6.91, v_num=0]
Epoch 2: 100%|██████████| 97/97 [00:14<00:00,  6.72it/s, loss=6.91, v_num=0]
Epoch 2:   0%|          | 0/97 [00:00<?, ?it/s, loss=6.91, v_num=0]
Epoch 3:   0%|          | 0/97 [00:00<?, ?it/s, loss=6.91, v_num=0]
Epoch 3:   1%|          | 1/97 [00:01<02:13,  1.39s/it, loss=6.91, v_num=0]
Epoch 3:   1%|          | 1/97 [00:01<02:19,  1.45s/it, loss=6.91, v_num=0]
Epoch 3:   2%|▏         | 2/97 [00:01<01:12,  1.31it/s, loss=6.91, v_num=0]
Epoch 3:   2%|▏         | 2/97 [00:01<01:15,  1.26it/s, loss=6.9, v_num=0]
Epoch 3:   3%|▎         | 3/97 [00:01<00:52,  1.81it/s, loss=6.9, v_num=0]
Epoch 3:   3%|▎         | 3/97 [00:01<00:53,  1.75it/s, loss=6.9, v_num=0]
Epoch 3:   4%|▍         | 4/97 [00:01<00:41,  2.23it/s, loss=6.9, v_num=0]
Epoch 3:   4%|▍         | 4/97 [00:01<00:43,  2.16it/s, loss=6.91, v_num=0]
Epoch 3:   5%|▌         | 5/97 [00:01<00:35,  2.59it/s, loss=6.91, v_num=0]
Epoch 3:   5%|▌         | 5/97 [00:01<00:36,  2.52it/s, loss=6.9, v_num=0]
Epoch 3:   6%|▌         | 6/97 [00:02<00:31,  2.90it/s, loss=6.9, v_num=0]
Epoch 3:   6%|▌         | 6/97 [00:02<00:32,  2.83it/s, loss=6.9, v_num=0]
Epoch 3:   7%|▋         | 7/97 [00:02<00:28,  3.18it/s, loss=6.9, v_num=0]
Epoch 3:   7%|▋         | 7/97 [00:02<00:28,  3.11it/s, loss=6.9, v_num=0]
Epoch 3:   8%|▊         | 8/97 [00:02<00:25,  3.43it/s, loss=6.9, v_num=0]
Epoch 3:   8%|▊         | 8/97 [00:02<00:26,  3.35it/s, loss=6.89, v_num=0]
Epoch 3:   9%|▉         | 9/97 [00:02<00:24,  3.64it/s, loss=6.89, v_num=0]
Epoch 3:   9%|▉         | 9/97 [00:02<00:24,  3.56it/s, loss=6.89, v_num=0]
Epoch 3:  10%|█         | 10/97 [00:02<00:22,  3.83it/s, loss=6.89, v_num=0]
Epoch 3:  10%|█         | 10/97 [00:02<00:23,  3.75it/s, loss=6.89, v_num=0]
Epoch 3:  11%|█▏        | 11/97 [00:02<00:21,  4.01it/s, loss=6.89, v_num=0]
Epoch 3:  11%|█▏        | 11/97 [00:02<00:21,  3.93it/s, loss=6.88, v_num=0]
Epoch 3:  12%|█▏        | 12/97 [00:02<00:20,  4.16it/s, loss=6.88, v_num=0]
Epoch 3:  12%|█▏        | 12/97 [00:02<00:20,  4.09it/s, loss=6.88, v_num=0]
Epoch 3:  13%|█▎        | 13/97 [00:03<00:19,  4.32it/s, loss=6.88, v_num=0]
Epoch 3:  13%|█▎        | 13/97 [00:03<00:19,  4.23it/s, loss=6.88, v_num=0]
Epoch 3:  14%|█▍        | 14/97 [00:03<00:18,  4.44it/s, loss=6.88, v_num=0]
Epoch 3:  14%|█▍        | 14/97 [00:03<00:19,  4.36it/s, loss=6.87, v_num=0]
Epoch 3:  15%|█▌        | 15/97 [00:03<00:17,  4.56it/s, loss=6.87, v_num=0]
Epoch 3:  15%|█▌        | 15/97 [00:03<00:18,  4.49it/s, loss=6.87, v_num=0]
Epoch 3:  16%|█▋        | 16/97 [00:03<00:17,  4.68it/s, loss=6.87, v_num=0]
Epoch 3:  16%|█▋        | 16/97 [00:03<00:17,  4.60it/s, loss=6.87, v_num=0]
Epoch 3:  18%|█▊        | 17/97 [00:03<00:16,  4.78it/s, loss=6.87, v_num=0]
Epoch 3:  18%|█▊        | 17/97 [00:03<00:17,  4.70it/s, loss=6.87, v_num=0]
Epoch 3:  19%|█▊        | 18/97 [00:03<00:16,  4.87it/s, loss=6.87, v_num=0]
Epoch 3:  19%|█▊        | 18/97 [00:03<00:16,  4.80it/s, loss=6.87, v_num=0]
Epoch 3:  20%|█▉        | 19/97 [00:03<00:15,  4.97it/s, loss=6.87, v_num=0]
Epoch 3:  20%|█▉        | 19/97 [00:03<00:15,  4.89it/s, loss=6.87, v_num=0]
Epoch 3:  21%|██        | 20/97 [00:03<00:15,  5.04it/s, loss=6.87, v_num=0]
Epoch 3:  21%|██        | 20/97 [00:04<00:15,  4.98it/s, loss=6.88, v_num=0]
Epoch 3:  22%|██▏       | 21/97 [00:04<00:14,  5.12it/s, loss=6.88, v_num=0]
Epoch 3:  22%|██▏       | 21/97 [00:04<00:15,  5.06it/s, loss=6.87, v_num=0]
Epoch 3:  23%|██▎       | 22/97 [00:04<00:14,  5.20it/s, loss=6.87, v_num=0]
Epoch 3:  23%|██▎       | 22/97 [00:04<00:14,  5.13it/s, loss=6.87, v_num=0]
Epoch 3:  24%|██▎       | 23/97 [00:04<00:14,  5.27it/s, loss=6.87, v_num=0]
Epoch 3:  24%|██▎       | 23/97 [00:04<00:14,  5.20it/s, loss=6.87, v_num=0]
Epoch 3:  25%|██▍       | 24/97 [00:04<00:13,  5.33it/s, loss=6.87, v_num=0]
Epoch 3:  25%|██▍       | 24/97 [00:04<00:13,  5.27it/s, loss=6.87, v_num=0]
Epoch 3:  26%|██▌       | 25/97 [00:04<00:13,  5.39it/s, loss=6.87, v_num=0]
Epoch 3:  26%|██▌       | 25/97 [00:04<00:13,  5.33it/s, loss=6.87, v_num=0]
Epoch 3:  27%|██▋       | 26/97 [00:04<00:13,  5.45it/s, loss=6.87, v_num=0]
Epoch 3:  27%|██▋       | 26/97 [00:04<00:13,  5.39it/s, loss=6.87, v_num=0]
Epoch 3:  28%|██▊       | 27/97 [00:04<00:12,  5.50it/s, loss=6.87, v_num=0]
Epoch 3:  28%|██▊       | 27/97 [00:04<00:12,  5.44it/s, loss=6.87, v_num=0]
Epoch 3:  29%|██▉       | 28/97 [00:05<00:12,  5.56it/s, loss=6.87, v_num=0]
Epoch 3:  29%|██▉       | 28/97 [00:05<00:12,  5.49it/s, loss=6.86, v_num=0]
Epoch 3:  30%|██▉       | 29/97 [00:05<00:12,  5.60it/s, loss=6.86, v_num=0]
Epoch 3:  30%|██▉       | 29/97 [00:05<00:12,  5.54it/s, loss=6.86, v_num=0]
Epoch 3:  31%|███       | 30/97 [00:05<00:11,  5.65it/s, loss=6.86, v_num=0]
Epoch 3:  31%|███       | 30/97 [00:05<00:11,  5.59it/s, loss=6.86, v_num=0]
Epoch 3:  32%|███▏      | 31/97 [00:05<00:11,  5.69it/s, loss=6.86, v_num=0]
Epoch 3:  32%|███▏      | 31/97 [00:05<00:11,  5.63it/s, loss=6.86, v_num=0]
Epoch 3:  33%|███▎      | 32/97 [00:05<00:11,  5.73it/s, loss=6.86, v_num=0]
Epoch 3:  33%|███▎      | 32/97 [00:05<00:11,  5.68it/s, loss=6.86, v_num=0]
Epoch 3:  34%|███▍      | 33/97 [00:05<00:11,  5.77it/s, loss=6.86, v_num=0]
Epoch 3:  34%|███▍      | 33/97 [00:05<00:11,  5.72it/s, loss=6.85, v_num=0]
Epoch 3:  35%|███▌      | 34/97 [00:05<00:10,  5.81it/s, loss=6.85, v_num=0]
Epoch 3:  35%|███▌      | 34/97 [00:05<00:10,  5.75it/s, loss=6.85, v_num=0]
Epoch 3:  36%|███▌      | 35/97 [00:05<00:10,  5.84it/s, loss=6.85, v_num=0]
Epoch 3:  36%|███▌      | 35/97 [00:06<00:10,  5.79it/s, loss=6.85, v_num=0]
Epoch 3:  37%|███▋      | 36/97 [00:06<00:10,  5.88it/s, loss=6.85, v_num=0]
Epoch 3:  37%|███▋      | 36/97 [00:06<00:10,  5.83it/s, loss=6.85, v_num=0]
Epoch 3:  38%|███▊      | 37/97 [00:06<00:10,  5.91it/s, loss=6.85, v_num=0]
Epoch 3:  38%|███▊      | 37/97 [00:06<00:10,  5.86it/s, loss=6.84, v_num=0]
Epoch 3:  39%|███▉      | 38/97 [00:06<00:09,  5.94it/s, loss=6.84, v_num=0]
Epoch 3:  39%|███▉      | 38/97 [00:06<00:10,  5.89it/s, loss=6.84, v_num=0]
Epoch 3:  40%|████      | 39/97 [00:06<00:09,  5.97it/s, loss=6.84, v_num=0]
Epoch 3:  40%|████      | 39/97 [00:06<00:09,  5.92it/s, loss=6.84, v_num=0]
Epoch 3:  41%|████      | 40/97 [00:06<00:09,  6.00it/s, loss=6.84, v_num=0]
Epoch 3:  41%|████      | 40/97 [00:06<00:09,  5.95it/s, loss=6.83, v_num=0]
Epoch 3:  42%|████▏     | 41/97 [00:06<00:09,  6.03it/s, loss=6.83, v_num=0]
Epoch 3:  42%|████▏     | 41/97 [00:06<00:09,  5.98it/s, loss=6.83, v_num=0]
Epoch 3:  43%|████▎     | 42/97 [00:06<00:09,  6.05it/s, loss=6.83, v_num=0]
Epoch 3:  43%|████▎     | 42/97 [00:06<00:09,  6.01it/s, loss=6.83, v_num=0]
Epoch 3:  44%|████▍     | 43/97 [00:07<00:08,  6.08it/s, loss=6.83, v_num=0]
Epoch 3:  44%|████▍     | 43/97 [00:07<00:08,  6.03it/s, loss=6.83, v_num=0]
Epoch 3:  45%|████▌     | 44/97 [00:07<00:08,  6.10it/s, loss=6.83, v_num=0]
Epoch 3:  45%|████▌     | 44/97 [00:07<00:08,  6.06it/s, loss=6.83, v_num=0]
Epoch 3:  46%|████▋     | 45/97 [00:07<00:08,  6.12it/s, loss=6.83, v_num=0]
Epoch 3:  46%|████▋     | 45/97 [00:07<00:08,  6.08it/s, loss=6.83, v_num=0]
Epoch 3:  47%|████▋     | 46/97 [00:07<00:08,  6.15it/s, loss=6.83, v_num=0]
Epoch 3:  47%|████▋     | 46/97 [00:07<00:08,  6.10it/s, loss=6.82, v_num=0]
Epoch 3:  48%|████▊     | 47/97 [00:07<00:08,  6.17it/s, loss=6.82, v_num=0]
Epoch 3:  48%|████▊     | 47/97 [00:07<00:08,  6.13it/s, loss=6.81, v_num=0]
Epoch 3:  49%|████▉     | 48/97 [00:07<00:07,  6.19it/s, loss=6.81, v_num=0]
Epoch 3:  49%|████▉     | 48/97 [00:07<00:07,  6.15it/s, loss=6.81, v_num=0]
Epoch 3:  51%|█████     | 49/97 [00:07<00:07,  6.21it/s, loss=6.81, v_num=0]
Epoch 3:  51%|█████     | 49/97 [00:07<00:07,  6.16it/s, loss=6.81, v_num=0]
Epoch 3:  52%|█████▏    | 50/97 [00:08<00:07,  6.23it/s, loss=6.81, v_num=0]
Epoch 3:  52%|█████▏    | 50/97 [00:08<00:07,  6.18it/s, loss=6.81, v_num=0]
Epoch 3:  53%|█████▎    | 51/97 [00:08<00:07,  6.25it/s, loss=6.81, v_num=0]
Epoch 3:  53%|█████▎    | 51/97 [00:08<00:07,  6.20it/s, loss=6.81, v_num=0]
Epoch 3:  54%|█████▎    | 52/97 [00:08<00:07,  6.26it/s, loss=6.81, v_num=0]
Epoch 3:  54%|█████▎    | 52/97 [00:08<00:07,  6.22it/s, loss=6.81, v_num=0]
Epoch 3:  55%|█████▍    | 53/97 [00:08<00:07,  6.28it/s, loss=6.81, v_num=0]
Epoch 3:  55%|█████▍    | 53/97 [00:08<00:07,  6.24it/s, loss=6.8, v_num=0]
Epoch 3:  56%|█████▌    | 54/97 [00:08<00:06,  6.30it/s, loss=6.8, v_num=0]
Epoch 3:  56%|█████▌    | 54/97 [00:08<00:06,  6.26it/s, loss=6.8, v_num=0]
Epoch 3:  57%|█████▋    | 55/97 [00:08<00:06,  6.32it/s, loss=6.8, v_num=0]
Epoch 3:  57%|█████▋    | 55/97 [00:08<00:06,  6.27it/s, loss=6.79, v_num=0]
Epoch 3:  58%|█████▊    | 56/97 [00:08<00:06,  6.33it/s, loss=6.79, v_num=0]
Epoch 3:  58%|█████▊    | 56/97 [00:08<00:06,  6.29it/s, loss=6.78, v_num=0]
Epoch 3:  59%|█████▉    | 57/97 [00:08<00:06,  6.34it/s, loss=6.78, v_num=0]
Epoch 3:  59%|█████▉    | 57/97 [00:09<00:06,  6.30it/s, loss=6.78, v_num=0]
Epoch 3:  60%|█████▉    | 58/97 [00:09<00:06,  6.35it/s, loss=6.78, v_num=0]
Epoch 3:  60%|█████▉    | 58/97 [00:09<00:06,  6.32it/s, loss=6.78, v_num=0]
Epoch 3:  61%|██████    | 59/97 [00:09<00:05,  6.37it/s, loss=6.78, v_num=0]
Epoch 3:  61%|██████    | 59/97 [00:09<00:06,  6.33it/s, loss=6.77, v_num=0]
Epoch 3:  62%|██████▏   | 60/97 [00:09<00:05,  6.38it/s, loss=6.77, v_num=0]
Epoch 3:  62%|██████▏   | 60/97 [00:09<00:05,  6.35it/s, loss=6.76, v_num=0]
Epoch 3:  63%|██████▎   | 61/97 [00:09<00:05,  6.39it/s, loss=6.76, v_num=0]
Epoch 3:  63%|██████▎   | 61/97 [00:09<00:05,  6.36it/s, loss=6.76, v_num=0]
Epoch 3:  64%|██████▍   | 62/97 [00:09<00:05,  6.41it/s, loss=6.76, v_num=0]
Epoch 3:  64%|██████▍   | 62/97 [00:09<00:05,  6.37it/s, loss=6.75, v_num=0]
Epoch 3:  65%|██████▍   | 63/97 [00:09<00:05,  6.42it/s, loss=6.75, v_num=0]
Epoch 3:  65%|██████▍   | 63/97 [00:09<00:05,  6.38it/s, loss=6.74, v_num=0]
Epoch 3:  66%|██████▌   | 64/97 [00:09<00:05,  6.43it/s, loss=6.74, v_num=0]
Epoch 3:  66%|██████▌   | 64/97 [00:10<00:05,  6.39it/s, loss=6.73, v_num=0]
Epoch 3:  67%|██████▋   | 65/97 [00:10<00:04,  6.44it/s, loss=6.73, v_num=0]
Epoch 3:  67%|██████▋   | 65/97 [00:10<00:04,  6.41it/s, loss=6.73, v_num=0]
Epoch 3:  68%|██████▊   | 66/97 [00:10<00:04,  6.45it/s, loss=6.73, v_num=0]
Epoch 3:  68%|██████▊   | 66/97 [00:10<00:04,  6.42it/s, loss=6.72, v_num=0]
Epoch 3:  69%|██████▉   | 67/97 [00:10<00:04,  6.46it/s, loss=6.72, v_num=0]
Epoch 3:  69%|██████▉   | 67/97 [00:10<00:04,  6.43it/s, loss=6.72, v_num=0]
Epoch 3:  70%|███████   | 68/97 [00:10<00:04,  6.47it/s, loss=6.72, v_num=0]
Epoch 3:  70%|███████   | 68/97 [00:10<00:04,  6.44it/s, loss=6.72, v_num=0]
Epoch 3:  71%|███████   | 69/97 [00:10<00:04,  6.49it/s, loss=6.72, v_num=0]
Epoch 3:  71%|███████   | 69/97 [00:10<00:04,  6.45it/s, loss=6.72, v_num=0]
Epoch 3:  72%|███████▏  | 70/97 [00:10<00:04,  6.50it/s, loss=6.72, v_num=0]
Epoch 3:  72%|███████▏  | 70/97 [00:10<00:04,  6.46it/s, loss=6.71, v_num=0]
Epoch 3:  73%|███████▎  | 71/97 [00:10<00:03,  6.51it/s, loss=6.71, v_num=0]
Epoch 3:  73%|███████▎  | 71/97 [00:10<00:04,  6.48it/s, loss=6.71, v_num=0]
Epoch 3:  74%|███████▍  | 72/97 [00:11<00:03,  6.52it/s, loss=6.71, v_num=0]
Epoch 3:  74%|███████▍  | 72/97 [00:11<00:03,  6.49it/s, loss=6.71, v_num=0]
Epoch 3:  75%|███████▌  | 73/97 [00:11<00:03,  6.53it/s, loss=6.71, v_num=0]
Epoch 3:  75%|███████▌  | 73/97 [00:11<00:03,  6.50it/s, loss=6.71, v_num=0]
Epoch 3:  76%|███████▋  | 74/97 [00:11<00:03,  6.54it/s, loss=6.71, v_num=0]
Epoch 3:  76%|███████▋  | 74/97 [00:11<00:03,  6.51it/s, loss=6.7, v_num=0]
Epoch 3:  77%|███████▋  | 75/97 [00:11<00:03,  6.55it/s, loss=6.7, v_num=0]
Epoch 3:  77%|███████▋  | 75/97 [00:11<00:03,  6.52it/s, loss=6.71, v_num=0]
Epoch 3:  78%|███████▊  | 76/97 [00:11<00:03,  6.56it/s, loss=6.71, v_num=0]
Epoch 3:  78%|███████▊  | 76/97 [00:11<00:03,  6.53it/s, loss=6.71, v_num=0]
Epoch 3:  79%|███████▉  | 77/97 [00:11<00:03,  6.57it/s, loss=6.71, v_num=0]
Epoch 3:  79%|███████▉  | 77/97 [00:11<00:03,  6.54it/s, loss=6.71, v_num=0]
Epoch 3:  80%|████████  | 78/97 [00:11<00:02,  6.58it/s, loss=6.71, v_num=0]
Epoch 3:  80%|████████  | 78/97 [00:11<00:02,  6.55it/s, loss=6.7, v_num=0]
Epoch 3:  81%|████████▏ | 79/97 [00:11<00:02,  6.59it/s, loss=6.7, v_num=0]
Epoch 3:  81%|████████▏ | 79/97 [00:12<00:02,  6.56it/s, loss=6.69, v_num=0]
Epoch 3:  82%|████████▏ | 80/97 [00:12<00:02,  6.59it/s, loss=6.69, v_num=0]
Epoch 3:  82%|████████▏ | 80/97 [00:12<00:02,  6.56it/s, loss=6.69, v_num=0]
Epoch 3:  84%|████████▎ | 81/97 [00:12<00:02,  6.60it/s, loss=6.69, v_num=0]
Epoch 3:  84%|████████▎ | 81/97 [00:12<00:02,  6.57it/s, loss=6.68, v_num=0]
Epoch 3:  85%|████████▍ | 82/97 [00:12<00:02,  6.61it/s, loss=6.68, v_num=0]
Epoch 3:  85%|████████▍ | 82/97 [00:12<00:02,  6.58it/s, loss=6.68, v_num=0]
Epoch 3:  86%|████████▌ | 83/97 [00:12<00:02,  6.62it/s, loss=6.68, v_num=0]
Epoch 3:  86%|████████▌ | 83/97 [00:12<00:02,  6.59it/s, loss=6.68, v_num=0]
Epoch 3:  87%|████████▋ | 84/97 [00:12<00:01,  6.63it/s, loss=6.68, v_num=0]
Epoch 3:  87%|████████▋ | 84/97 [00:12<00:01,  6.60it/s, loss=6.68, v_num=0]
Epoch 3:  88%|████████▊ | 85/97 [00:12<00:01,  6.63it/s, loss=6.68, v_num=0]
Epoch 3:  88%|████████▊ | 85/97 [00:12<00:01,  6.60it/s, loss=6.68, v_num=0]
Epoch 3:  89%|████████▊ | 86/97 [00:12<00:01,  6.64it/s, loss=6.68, v_num=0]
Epoch 3:  89%|████████▊ | 86/97 [00:13<00:01,  6.61it/s, loss=6.68, v_num=0]
Epoch 3:  90%|████████▉ | 87/97 [00:13<00:01,  6.65it/s, loss=6.68, v_num=0]
Epoch 3:  90%|████████▉ | 87/97 [00:13<00:01,  6.62it/s, loss=6.67, v_num=0]
Epoch 3:  91%|█████████ | 88/97 [00:13<00:01,  6.66it/s, loss=6.67, v_num=0]
Epoch 3:  91%|█████████ | 88/97 [00:13<00:01,  6.63it/s, loss=6.66, v_num=0]
Epoch 3:  92%|█████████▏| 89/97 [00:13<00:01,  6.67it/s, loss=6.66, v_num=0]
Epoch 3:  92%|█████████▏| 89/97 [00:13<00:01,  6.64it/s, loss=6.66, v_num=0]
Epoch 3:  93%|█████████▎| 90/97 [00:13<00:01,  6.67it/s, loss=6.66, v_num=0]
Epoch 3:  93%|█████████▎| 90/97 [00:13<00:01,  6.64it/s, loss=6.65, v_num=0]
Epoch 3:  94%|█████████▍| 91/97 [00:13<00:00,  6.68it/s, loss=6.65, v_num=0]
Epoch 3:  94%|█████████▍| 91/97 [00:13<00:00,  6.65it/s, loss=6.65, v_num=0]
Epoch 3:  95%|█████████▍| 92/97 [00:13<00:00,  6.69it/s, loss=6.65, v_num=0]
Epoch 3:  95%|█████████▍| 92/97 [00:13<00:00,  6.66it/s, loss=6.63, v_num=0]
Epoch 3:  96%|█████████▌| 93/97 [00:13<00:00,  6.69it/s, loss=6.63, v_num=0]
Epoch 3:  96%|█████████▌| 93/97 [00:13<00:00,  6.66it/s, loss=6.63, v_num=0]
Epoch 3:  97%|█████████▋| 94/97 [00:14<00:00,  6.70it/s, loss=6.63, v_num=0]
Epoch 3:  97%|█████████▋| 94/97 [00:14<00:00,  6.67it/s, loss=6.62, v_num=0]
Epoch 3:  98%|█████████▊| 95/97 [00:14<00:00,  6.70it/s, loss=6.62, v_num=0]
Epoch 3:  98%|█████████▊| 95/97 [00:14<00:00,  6.68it/s, loss=6.61, v_num=0]
Epoch 3:  99%|█████████▉| 96/97 [00:14<00:00,  6.71it/s, loss=6.61, v_num=0]
Epoch 3:  99%|█████████▉| 96/97 [00:14<00:00,  6.68it/s, loss=6.6, v_num=0]
Epoch 3: 100%|██████████| 97/97 [00:14<00:00,  6.72it/s, loss=6.6, v_num=0]
Epoch 3: 100%|██████████| 97/97 [00:14<00:00,  6.69it/s, loss=6.6, v_num=0]
Epoch 3: 100%|██████████| 97/97 [00:14<00:00,  6.69it/s, loss=6.6, v_num=0]
Epoch 3:   0%|          | 0/97 [00:00<?, ?it/s, loss=6.6, v_num=0]
Epoch 4:   0%|          | 0/97 [00:00<?, ?it/s, loss=6.6, v_num=0]
Epoch 4:   1%|          | 1/97 [00:01<02:11,  1.37s/it, loss=6.6, v_num=0]
Epoch 4:   1%|          | 1/97 [00:01<02:16,  1.43s/it, loss=6.6, v_num=0]
Epoch 4:   2%|▏         | 2/97 [00:01<01:11,  1.32it/s, loss=6.6, v_num=0]
Epoch 4:   2%|▏         | 2/97 [00:01<01:14,  1.28it/s, loss=6.6, v_num=0]
Epoch 4:   3%|▎         | 3/97 [00:01<00:51,  1.82it/s, loss=6.6, v_num=0]
Epoch 4:   3%|▎         | 3/97 [00:01<00:53,  1.76it/s, loss=6.59, v_num=0]
Epoch 4:   4%|▍         | 4/97 [00:01<00:41,  2.24it/s, loss=6.59, v_num=0]
Epoch 4:   4%|▍         | 4/97 [00:01<00:42,  2.18it/s, loss=6.59, v_num=0]
Epoch 4:   5%|▌         | 5/97 [00:01<00:35,  2.61it/s, loss=6.59, v_num=0]
Epoch 4:   5%|▌         | 5/97 [00:01<00:36,  2.54it/s, loss=6.58, v_num=0]
Epoch 4:   6%|▌         | 6/97 [00:02<00:31,  2.92it/s, loss=6.58, v_num=0]
Epoch 4:   6%|▌         | 6/97 [00:02<00:31,  2.85it/s, loss=6.57, v_num=0]
Epoch 4:   7%|▋         | 7/97 [00:02<00:28,  3.20it/s, loss=6.57, v_num=0]
Epoch 4:   7%|▋         | 7/97 [00:02<00:28,  3.12it/s, loss=6.56, v_num=0]
Epoch 4:   8%|▊         | 8/97 [00:02<00:25,  3.44it/s, loss=6.56, v_num=0]
Epoch 4:   8%|▊         | 8/97 [00:02<00:26,  3.36it/s, loss=6.55, v_num=0]
Epoch 4:   9%|▉         | 9/97 [00:02<00:24,  3.64it/s, loss=6.55, v_num=0]
Epoch 4:   9%|▉         | 9/97 [00:02<00:24,  3.57it/s, loss=6.54, v_num=0]
Epoch 4:  10%|█         | 10/97 [00:02<00:22,  3.84it/s, loss=6.54, v_num=0]
Epoch 4:  10%|█         | 10/97 [00:02<00:23,  3.76it/s, loss=6.54, v_num=0]
Epoch 4:  11%|█▏        | 11/97 [00:02<00:21,  4.02it/s, loss=6.54, v_num=0]
Epoch 4:  11%|█▏        | 11/97 [00:02<00:21,  3.94it/s, loss=6.54, v_num=0]
Epoch 4:  12%|█▏        | 12/97 [00:02<00:20,  4.18it/s, loss=6.54, v_num=0]
Epoch 4:  12%|█▏        | 12/97 [00:02<00:20,  4.10it/s, loss=6.52, v_num=0]
Epoch 4:  13%|█▎        | 13/97 [00:03<00:19,  4.32it/s, loss=6.52, v_num=0]
Epoch 4:  13%|█▎        | 13/97 [00:03<00:19,  4.24it/s, loss=6.52, v_num=0]
Epoch 4:  14%|█▍        | 14/97 [00:03<00:18,  4.45it/s, loss=6.52, v_num=0]
Epoch 4:  14%|█▍        | 14/97 [00:03<00:18,  4.38it/s, loss=6.52, v_num=0]
Epoch 4:  15%|█▌        | 15/97 [00:03<00:17,  4.58it/s, loss=6.52, v_num=0]
Epoch 4:  15%|█▌        | 15/97 [00:03<00:18,  4.50it/s, loss=6.52, v_num=0]
Epoch 4:  16%|█▋        | 16/97 [00:03<00:17,  4.69it/s, loss=6.52, v_num=0]
Epoch 4:  16%|█▋        | 16/97 [00:03<00:17,  4.61it/s, loss=6.51, v_num=0]
Epoch 4:  18%|█▊        | 17/97 [00:03<00:16,  4.78it/s, loss=6.51, v_num=0]
Epoch 4:  18%|█▊        | 17/97 [00:03<00:16,  4.71it/s, loss=6.51, v_num=0]
Epoch 4:  19%|█▊        | 18/97 [00:03<00:16,  4.88it/s, loss=6.51, v_num=0]
Epoch 4:  19%|█▊        | 18/97 [00:03<00:16,  4.81it/s, loss=6.51, v_num=0]
Epoch 4:  20%|█▉        | 19/97 [00:03<00:15,  4.98it/s, loss=6.51, v_num=0]
Epoch 4:  20%|█▉        | 19/97 [00:03<00:15,  4.90it/s, loss=6.51, v_num=0]
Epoch 4:  21%|██        | 20/97 [00:03<00:15,  5.06it/s, loss=6.51, v_num=0]
Epoch 4:  21%|██        | 20/97 [00:04<00:15,  4.98it/s, loss=6.51, v_num=0]
Epoch 4:  22%|██▏       | 21/97 [00:04<00:14,  5.13it/s, loss=6.51, v_num=0]
Epoch 4:  22%|██▏       | 21/97 [00:04<00:15,  5.06it/s, loss=6.5, v_num=0]
Epoch 4:  23%|██▎       | 22/97 [00:04<00:14,  5.20it/s, loss=6.5, v_num=0]
Epoch 4:  23%|██▎       | 22/97 [00:04<00:14,  5.13it/s, loss=6.5, v_num=0]
Epoch 4:  24%|██▎       | 23/97 [00:04<00:14,  5.27it/s, loss=6.5, v_num=0]
Epoch 4:  24%|██▎       | 23/97 [00:04<00:14,  5.20it/s, loss=6.51, v_num=0]
Epoch 4:  25%|██▍       | 24/97 [00:04<00:13,  5.32it/s, loss=6.51, v_num=0]
Epoch 4:  25%|██▍       | 24/97 [00:04<00:13,  5.26it/s, loss=6.5, v_num=0]
Epoch 4:  26%|██▌       | 25/97 [00:04<00:13,  5.38it/s, loss=6.5, v_num=0]
Epoch 4:  26%|██▌       | 25/97 [00:04<00:13,  5.31it/s, loss=6.5, v_num=0]
Epoch 4:  27%|██▋       | 26/97 [00:04<00:13,  5.43it/s, loss=6.5, v_num=0]
Epoch 4:  27%|██▋       | 26/97 [00:04<00:13,  5.37it/s, loss=6.5, v_num=0]
Epoch 4:  28%|██▊       | 27/97 [00:04<00:12,  5.49it/s, loss=6.5, v_num=0]
Epoch 4:  28%|██▊       | 27/97 [00:04<00:12,  5.43it/s, loss=6.5, v_num=0]
Epoch 4:  29%|██▉       | 28/97 [00:05<00:12,  5.54it/s, loss=6.5, v_num=0]
Epoch 4:  29%|██▉       | 28/97 [00:05<00:12,  5.48it/s, loss=6.5, v_num=0]
Epoch 4:  30%|██▉       | 29/97 [00:05<00:12,  5.59it/s, loss=6.5, v_num=0]
Epoch 4:  30%|██▉       | 29/97 [00:05<00:12,  5.53it/s, loss=6.51, v_num=0]
Epoch 4:  31%|███       | 30/97 [00:05<00:11,  5.63it/s, loss=6.51, v_num=0]
Epoch 4:  31%|███       | 30/97 [00:05<00:12,  5.58it/s, loss=6.51, v_num=0]
Epoch 4:  32%|███▏      | 31/97 [00:05<00:11,  5.68it/s, loss=6.51, v_num=0]
Epoch 4:  32%|███▏      | 31/97 [00:05<00:11,  5.62it/s, loss=6.5, v_num=0]
Epoch 4:  33%|███▎      | 32/97 [00:05<00:11,  5.72it/s, loss=6.5, v_num=0]
Epoch 4:  33%|███▎      | 32/97 [00:05<00:11,  5.66it/s, loss=6.51, v_num=0]
Epoch 4:  34%|███▍      | 33/97 [00:05<00:11,  5.75it/s, loss=6.51, v_num=0]
Epoch 4:  34%|███▍      | 33/97 [00:05<00:11,  5.70it/s, loss=6.5, v_num=0]
Epoch 4:  35%|███▌      | 34/97 [00:05<00:10,  5.79it/s, loss=6.5, v_num=0]
Epoch 4:  35%|███▌      | 34/97 [00:05<00:10,  5.74it/s, loss=6.49, v_num=0]
Epoch 4:  36%|███▌      | 35/97 [00:06<00:10,  5.83it/s, loss=6.49, v_num=0]
Epoch 4:  36%|███▌      | 35/97 [00:06<00:10,  5.77it/s, loss=6.49, v_num=0]
Epoch 4:  37%|███▋      | 36/97 [00:06<00:10,  5.86it/s, loss=6.49, v_num=0]
Epoch 4:  37%|███▋      | 36/97 [00:06<00:10,  5.81it/s, loss=6.49, v_num=0]
Epoch 4:  38%|███▊      | 37/97 [00:06<00:10,  5.89it/s, loss=6.49, v_num=0]
Epoch 4:  38%|███▊      | 37/97 [00:06<00:10,  5.84it/s, loss=6.48, v_num=0]
Epoch 4:  39%|███▉      | 38/97 [00:06<00:09,  5.93it/s, loss=6.48, v_num=0]
Epoch 4:  39%|███▉      | 38/97 [00:06<00:10,  5.87it/s, loss=6.48, v_num=0]
Epoch 4:  40%|████      | 39/97 [00:06<00:09,  5.95it/s, loss=6.48, v_num=0]
Epoch 4:  40%|████      | 39/97 [00:06<00:09,  5.90it/s, loss=6.47, v_num=0]
Epoch 4:  41%|████      | 40/97 [00:06<00:09,  5.98it/s, loss=6.47, v_num=0]
Epoch 4:  41%|████      | 40/97 [00:06<00:09,  5.93it/s, loss=6.47, v_num=0]
Epoch 4:  42%|████▏     | 41/97 [00:06<00:09,  6.01it/s, loss=6.47, v_num=0]
Epoch 4:  42%|████▏     | 41/97 [00:06<00:09,  5.96it/s, loss=6.47, v_num=0]
Epoch 4:  43%|████▎     | 42/97 [00:06<00:09,  6.04it/s, loss=6.47, v_num=0]
Epoch 4:  43%|████▎     | 42/97 [00:07<00:09,  5.99it/s, loss=6.46, v_num=0]
Epoch 4:  44%|████▍     | 43/97 [00:07<00:08,  6.07it/s, loss=6.46, v_num=0]
Epoch 4:  44%|████▍     | 43/97 [00:07<00:08,  6.02it/s, loss=6.46, v_num=0]
Epoch 4:  45%|████▌     | 44/97 [00:07<00:08,  6.09it/s, loss=6.46, v_num=0]
Epoch 4:  45%|████▌     | 44/97 [00:07<00:08,  6.04it/s, loss=6.45, v_num=0]
Epoch 4:  46%|████▋     | 45/97 [00:07<00:08,  6.11it/s, loss=6.45, v_num=0]
Epoch 4:  46%|████▋     | 45/97 [00:07<00:08,  6.07it/s, loss=6.45, v_num=0]
Epoch 4:  47%|████▋     | 46/97 [00:07<00:08,  6.14it/s, loss=6.45, v_num=0]
Epoch 4:  47%|████▋     | 46/97 [00:07<00:08,  6.09it/s, loss=6.46, v_num=0]
Epoch 4:  48%|████▊     | 47/97 [00:07<00:08,  6.16it/s, loss=6.46, v_num=0]
Epoch 4:  48%|████▊     | 47/97 [00:07<00:08,  6.11it/s, loss=6.45, v_num=0]
Epoch 4:  49%|████▉     | 48/97 [00:07<00:07,  6.18it/s, loss=6.45, v_num=0]
Epoch 4:  49%|████▉     | 48/97 [00:07<00:07,  6.13it/s, loss=6.44, v_num=0]
Epoch 4:  51%|█████     | 49/97 [00:07<00:07,  6.20it/s, loss=6.44, v_num=0]
Epoch 4:  51%|█████     | 49/97 [00:07<00:07,  6.16it/s, loss=6.44, v_num=0]
Epoch 4:  52%|█████▏    | 50/97 [00:08<00:07,  6.22it/s, loss=6.44, v_num=0]
Epoch 4:  52%|█████▏    | 50/97 [00:08<00:07,  6.18it/s, loss=6.43, v_num=0]
Epoch 4:  53%|█████▎    | 51/97 [00:08<00:07,  6.24it/s, loss=6.43, v_num=0]
Epoch 4:  53%|█████▎    | 51/97 [00:08<00:07,  6.20it/s, loss=6.43, v_num=0]
Epoch 4:  54%|█████▎    | 52/97 [00:08<00:07,  6.26it/s, loss=6.43, v_num=0]
Epoch 4:  54%|█████▎    | 52/97 [00:08<00:07,  6.22it/s, loss=6.43, v_num=0]
Epoch 4:  55%|█████▍    | 53/97 [00:08<00:07,  6.27it/s, loss=6.43, v_num=0]
Epoch 4:  55%|█████▍    | 53/97 [00:08<00:07,  6.23it/s, loss=6.42, v_num=0]
Epoch 4:  56%|█████▌    | 54/97 [00:08<00:06,  6.29it/s, loss=6.42, v_num=0]
Epoch 4:  56%|█████▌    | 54/97 [00:08<00:06,  6.25it/s, loss=6.42, v_num=0]
Epoch 4:  57%|█████▋    | 55/97 [00:08<00:06,  6.31it/s, loss=6.42, v_num=0]
Epoch 4:  57%|█████▋    | 55/97 [00:08<00:06,  6.26it/s, loss=6.42, v_num=0]
Epoch 4:  58%|█████▊    | 56/97 [00:08<00:06,  6.32it/s, loss=6.42, v_num=0]
Epoch 4:  58%|█████▊    | 56/97 [00:08<00:06,  6.28it/s, loss=6.42, v_num=0]
Epoch 4:  59%|█████▉    | 57/97 [00:08<00:06,  6.34it/s, loss=6.42, v_num=0]
Epoch 4:  59%|█████▉    | 57/97 [00:09<00:06,  6.30it/s, loss=6.41, v_num=0]
Epoch 4:  60%|█████▉    | 58/97 [00:09<00:06,  6.35it/s, loss=6.41, v_num=0]
Epoch 4:  60%|█████▉    | 58/97 [00:09<00:06,  6.31it/s, loss=6.41, v_num=0]
Epoch 4:  61%|██████    | 59/97 [00:09<00:05,  6.37it/s, loss=6.41, v_num=0]
Epoch 4:  61%|██████    | 59/97 [00:09<00:06,  6.33it/s, loss=6.41, v_num=0]
Epoch 4:  62%|██████▏   | 60/97 [00:09<00:05,  6.38it/s, loss=6.41, v_num=0]
Epoch 4:  62%|██████▏   | 60/97 [00:09<00:05,  6.34it/s, loss=6.4, v_num=0]
Epoch 4:  63%|██████▎   | 61/97 [00:09<00:05,  6.39it/s, loss=6.4, v_num=0]
Epoch 4:  63%|██████▎   | 61/97 [00:09<00:05,  6.35it/s, loss=6.4, v_num=0]
Epoch 4:  64%|██████▍   | 62/97 [00:09<00:05,  6.40it/s, loss=6.4, v_num=0]
Epoch 4:  64%|██████▍   | 62/97 [00:09<00:05,  6.37it/s, loss=6.4, v_num=0]
Epoch 4:  65%|██████▍   | 63/97 [00:09<00:05,  6.42it/s, loss=6.4, v_num=0]
Epoch 4:  65%|██████▍   | 63/97 [00:09<00:05,  6.38it/s, loss=6.39, v_num=0]
Epoch 4:  66%|██████▌   | 64/97 [00:09<00:05,  6.43it/s, loss=6.39, v_num=0]
Epoch 4:  66%|██████▌   | 64/97 [00:10<00:05,  6.39it/s, loss=6.39, v_num=0]
Epoch 4:  67%|██████▋   | 65/97 [00:10<00:04,  6.44it/s, loss=6.39, v_num=0]
Epoch 4:  67%|██████▋   | 65/97 [00:10<00:04,  6.41it/s, loss=6.39, v_num=0]
Epoch 4:  68%|██████▊   | 66/97 [00:10<00:04,  6.46it/s, loss=6.39, v_num=0]
Epoch 4:  68%|██████▊   | 66/97 [00:10<00:04,  6.42it/s, loss=6.38, v_num=0]
Epoch 4:  69%|██████▉   | 67/97 [00:10<00:04,  6.47it/s, loss=6.38, v_num=0]
Epoch 4:  69%|██████▉   | 67/97 [00:10<00:04,  6.43it/s, loss=6.38, v_num=0]
Epoch 4:  70%|███████   | 68/97 [00:10<00:04,  6.48it/s, loss=6.38, v_num=0]
Epoch 4:  70%|███████   | 68/97 [00:10<00:04,  6.44it/s, loss=6.38, v_num=0]
Epoch 4:  71%|███████   | 69/97 [00:10<00:04,  6.49it/s, loss=6.38, v_num=0]
Epoch 4:  71%|███████   | 69/97 [00:10<00:04,  6.46it/s, loss=6.37, v_num=0]
Epoch 4:  72%|███████▏  | 70/97 [00:10<00:04,  6.50it/s, loss=6.37, v_num=0]
Epoch 4:  72%|███████▏  | 70/97 [00:10<00:04,  6.47it/s, loss=6.35, v_num=0]
Epoch 4:  73%|███████▎  | 71/97 [00:10<00:03,  6.51it/s, loss=6.35, v_num=0]
Epoch 4:  73%|███████▎  | 71/97 [00:10<00:04,  6.48it/s, loss=6.35, v_num=0]
Epoch 4:  74%|███████▍  | 72/97 [00:11<00:03,  6.52it/s, loss=6.35, v_num=0]
Epoch 4:  74%|███████▍  | 72/97 [00:11<00:03,  6.49it/s, loss=6.34, v_num=0]
Epoch 4:  75%|███████▌  | 73/97 [00:11<00:03,  6.53it/s, loss=6.34, v_num=0]
Epoch 4:  75%|███████▌  | 73/97 [00:11<00:03,  6.50it/s, loss=6.35, v_num=0]
Epoch 4:  76%|███████▋  | 74/97 [00:11<00:03,  6.54it/s, loss=6.35, v_num=0]
Epoch 4:  76%|███████▋  | 74/97 [00:11<00:03,  6.51it/s, loss=6.34, v_num=0]
Epoch 4:  77%|███████▋  | 75/97 [00:11<00:03,  6.55it/s, loss=6.34, v_num=0]
Epoch 4:  77%|███████▋  | 75/97 [00:11<00:03,  6.51it/s, loss=6.33, v_num=0]
Epoch 4:  78%|███████▊  | 76/97 [00:11<00:03,  6.56it/s, loss=6.33, v_num=0]
Epoch 4:  78%|███████▊  | 76/97 [00:11<00:03,  6.52it/s, loss=6.33, v_num=0]
Epoch 4:  79%|███████▉  | 77/97 [00:11<00:03,  6.56it/s, loss=6.33, v_num=0]
Epoch 4:  79%|███████▉  | 77/97 [00:11<00:03,  6.53it/s, loss=6.33, v_num=0]
Epoch 4:  80%|████████  | 78/97 [00:11<00:02,  6.57it/s, loss=6.33, v_num=0]
Epoch 4:  80%|████████  | 78/97 [00:11<00:02,  6.54it/s, loss=6.33, v_num=0]
Epoch 4:  81%|████████▏ | 79/97 [00:12<00:02,  6.58it/s, loss=6.33, v_num=0]
Epoch 4:  81%|████████▏ | 79/97 [00:12<00:02,  6.55it/s, loss=6.32, v_num=0]
Epoch 4:  82%|████████▏ | 80/97 [00:12<00:02,  6.59it/s, loss=6.32, v_num=0]
Epoch 4:  82%|████████▏ | 80/97 [00:12<00:02,  6.56it/s, loss=6.32, v_num=0]
Epoch 4:  84%|████████▎ | 81/97 [00:12<00:02,  6.60it/s, loss=6.32, v_num=0]
Epoch 4:  84%|████████▎ | 81/97 [00:12<00:02,  6.57it/s, loss=6.31, v_num=0]
Epoch 4:  85%|████████▍ | 82/97 [00:12<00:02,  6.61it/s, loss=6.31, v_num=0]
Epoch 4:  85%|████████▍ | 82/97 [00:12<00:02,  6.58it/s, loss=6.31, v_num=0]
Epoch 4:  86%|████████▌ | 83/97 [00:12<00:02,  6.61it/s, loss=6.31, v_num=0]
Epoch 4:  86%|████████▌ | 83/97 [00:12<00:02,  6.59it/s, loss=6.31, v_num=0]
Epoch 4:  87%|████████▋ | 84/97 [00:12<00:01,  6.62it/s, loss=6.31, v_num=0]
Epoch 4:  87%|████████▋ | 84/97 [00:12<00:01,  6.59it/s, loss=6.3, v_num=0]
Epoch 4:  88%|████████▊ | 85/97 [00:12<00:01,  6.63it/s, loss=6.3, v_num=0]
Epoch 4:  88%|████████▊ | 85/97 [00:12<00:01,  6.60it/s, loss=6.3, v_num=0]
Epoch 4:  89%|████████▊ | 86/97 [00:12<00:01,  6.64it/s, loss=6.3, v_num=0]
Epoch 4:  89%|████████▊ | 86/97 [00:13<00:01,  6.61it/s, loss=6.3, v_num=0]
Epoch 4:  90%|████████▉ | 87/97 [00:13<00:01,  6.65it/s, loss=6.3, v_num=0]
Epoch 4:  90%|████████▉ | 87/97 [00:13<00:01,  6.62it/s, loss=6.29, v_num=0]
Epoch 4:  91%|█████████ | 88/97 [00:13<00:01,  6.66it/s, loss=6.29, v_num=0]
Epoch 4:  91%|█████████ | 88/97 [00:13<00:01,  6.63it/s, loss=6.29, v_num=0]
Epoch 4:  92%|█████████▏| 89/97 [00:13<00:01,  6.66it/s, loss=6.29, v_num=0]
Epoch 4:  92%|█████████▏| 89/97 [00:13<00:01,  6.63it/s, loss=6.29, v_num=0]
Epoch 4:  93%|█████████▎| 90/97 [00:13<00:01,  6.67it/s, loss=6.29, v_num=0]
Epoch 4:  93%|█████████▎| 90/97 [00:13<00:01,  6.64it/s, loss=6.29, v_num=0]
Epoch 4:  94%|█████████▍| 91/97 [00:13<00:00,  6.67it/s, loss=6.29, v_num=0]
Epoch 4:  94%|█████████▍| 91/97 [00:13<00:00,  6.65it/s, loss=6.29, v_num=0]
Epoch 4:  95%|█████████▍| 92/97 [00:13<00:00,  6.68it/s, loss=6.29, v_num=0]
Epoch 4:  95%|█████████▍| 92/97 [00:13<00:00,  6.65it/s, loss=6.3, v_num=0]
Epoch 4:  96%|█████████▌| 93/97 [00:13<00:00,  6.69it/s, loss=6.3, v_num=0]
Epoch 4:  96%|█████████▌| 93/97 [00:13<00:00,  6.66it/s, loss=6.28, v_num=0]
Epoch 4:  97%|█████████▋| 94/97 [00:14<00:00,  6.69it/s, loss=6.28, v_num=0]
Epoch 4:  97%|█████████▋| 94/97 [00:14<00:00,  6.67it/s, loss=6.28, v_num=0]
Epoch 4:  98%|█████████▊| 95/97 [00:14<00:00,  6.70it/s, loss=6.28, v_num=0]
Epoch 4:  98%|█████████▊| 95/97 [00:14<00:00,  6.67it/s, loss=6.28, v_num=0]
Epoch 4:  99%|█████████▉| 96/97 [00:14<00:00,  6.71it/s, loss=6.28, v_num=0]
Epoch 4:  99%|█████████▉| 96/97 [00:14<00:00,  6.68it/s, loss=6.28, v_num=0]
Epoch 4: 100%|██████████| 97/97 [00:14<00:00,  6.71it/s, loss=6.28, v_num=0]
Epoch 4: 100%|██████████| 97/97 [00:14<00:00,  6.69it/s, loss=6.27, v_num=0]
Epoch 4: 100%|██████████| 97/97 [00:14<00:00,  6.69it/s, loss=6.27, v_num=0]
Epoch 4:   0%|          | 0/97 [00:00<?, ?it/s, loss=6.27, v_num=0]
Epoch 5:   0%|          | 0/97 [00:00<?, ?it/s, loss=6.27, v_num=0]
Epoch 5:   1%|          | 1/97 [00:01<02:11,  1.37s/it, loss=6.27, v_num=0]
Epoch 5:   1%|          | 1/97 [00:01<02:16,  1.42s/it, loss=6.26, v_num=0]
Epoch 5:   2%|▏         | 2/97 [00:01<01:11,  1.33it/s, loss=6.26, v_num=0]
Epoch 5:   2%|▏         | 2/97 [00:01<01:14,  1.28it/s, loss=6.26, v_num=0]
Epoch 5:   3%|▎         | 3/97 [00:01<00:51,  1.81it/s, loss=6.26, v_num=0]
Epoch 5:   3%|▎         | 3/97 [00:01<00:53,  1.76it/s, loss=6.25, v_num=0]
Epoch 5:   4%|▍         | 4/97 [00:01<00:41,  2.24it/s, loss=6.25, v_num=0]
Epoch 5:   4%|▍         | 4/97 [00:01<00:42,  2.17it/s, loss=6.25, v_num=0]
Epoch 5:   5%|▌         | 5/97 [00:01<00:35,  2.60it/s, loss=6.25, v_num=0]
Epoch 5:   5%|▌         | 5/97 [00:01<00:36,  2.53it/s, loss=6.24, v_num=0]
Epoch 5:   6%|▌         | 6/97 [00:02<00:31,  2.92it/s, loss=6.24, v_num=0]
Epoch 5:   6%|▌         | 6/97 [00:02<00:32,  2.84it/s, loss=6.24, v_num=0]
Epoch 5:   7%|▋         | 7/97 [00:02<00:28,  3.19it/s, loss=6.24, v_num=0]
Epoch 5:   7%|▋         | 7/97 [00:02<00:28,  3.11it/s, loss=6.23, v_num=0]
Epoch 5:   8%|▊         | 8/97 [00:02<00:25,  3.44it/s, loss=6.23, v_num=0]
Epoch 5:   8%|▊         | 8/97 [00:02<00:26,  3.36it/s, loss=6.22, v_num=0]
Epoch 5:   9%|▉         | 9/97 [00:02<00:24,  3.64it/s, loss=6.22, v_num=0]
Epoch 5:   9%|▉         | 9/97 [00:02<00:24,  3.56it/s, loss=6.21, v_num=0]
Epoch 5:  10%|█         | 10/97 [00:02<00:22,  3.83it/s, loss=6.21, v_num=0]
Epoch 5:  10%|█         | 10/97 [00:02<00:23,  3.76it/s, loss=6.21, v_num=0]
Epoch 5:  11%|█▏        | 11/97 [00:02<00:21,  4.00it/s, loss=6.21, v_num=0]
Epoch 5:  11%|█▏        | 11/97 [00:02<00:21,  3.92it/s, loss=6.22, v_num=0]
Epoch 5:  12%|█▏        | 12/97 [00:02<00:20,  4.16it/s, loss=6.22, v_num=0]
Epoch 5:  12%|█▏        | 12/97 [00:02<00:20,  4.08it/s, loss=6.22, v_num=0]
Epoch 5:  13%|█▎        | 13/97 [00:03<00:19,  4.30it/s, loss=6.22, v_num=0]
Epoch 5:  13%|█▎        | 13/97 [00:03<00:19,  4.23it/s, loss=6.21, v_num=0]
Epoch 5:  14%|█▍        | 14/97 [00:03<00:18,  4.44it/s, loss=6.21, v_num=0]
Epoch 5:  14%|█▍        | 14/97 [00:03<00:19,  4.36it/s, loss=6.2, v_num=0]
Epoch 5:  15%|█▌        | 15/97 [00:03<00:17,  4.56it/s, loss=6.2, v_num=0]
Epoch 5:  15%|█▌        | 15/97 [00:03<00:18,  4.48it/s, loss=6.19, v_num=0]
Epoch 5:  16%|█▋        | 16/97 [00:03<00:17,  4.68it/s, loss=6.19, v_num=0]
Epoch 5:  16%|█▋        | 16/97 [00:03<00:17,  4.60it/s, loss=6.19, v_num=0]
Epoch 5:  18%|█▊        | 17/97 [00:03<00:16,  4.77it/s, loss=6.19, v_num=0]
Epoch 5:  18%|█▊        | 17/97 [00:03<00:17,  4.70it/s, loss=6.18, v_num=0]
Epoch 5:  19%|█▊        | 18/97 [00:03<00:16,  4.87it/s, loss=6.18, v_num=0]
Epoch 5:  19%|█▊        | 18/97 [00:03<00:16,  4.79it/s, loss=6.17, v_num=0]
Epoch 5:  20%|█▉        | 19/97 [00:03<00:15,  4.95it/s, loss=6.17, v_num=0]
Epoch 5:  20%|█▉        | 19/97 [00:03<00:15,  4.88it/s, loss=6.18, v_num=0]
Epoch 5:  21%|██        | 20/97 [00:03<00:15,  5.04it/s, loss=6.18, v_num=0]
Epoch 5:  21%|██        | 20/97 [00:04<00:15,  4.97it/s, loss=6.18, v_num=0]
Epoch 5:  22%|██▏       | 21/97 [00:04<00:14,  5.12it/s, loss=6.18, v_num=0]
Epoch 5:  22%|██▏       | 21/97 [00:04<00:15,  5.05it/s, loss=6.18, v_num=0]
Epoch 5:  23%|██▎       | 22/97 [00:04<00:14,  5.19it/s, loss=6.18, v_num=0]
Epoch 5:  23%|██▎       | 22/97 [00:04<00:14,  5.12it/s, loss=6.18, v_num=0]
Epoch 5:  24%|██▎       | 23/97 [00:04<00:14,  5.25it/s, loss=6.18, v_num=0]
Epoch 5:  24%|██▎       | 23/97 [00:04<00:14,  5.19it/s, loss=6.17, v_num=0]
Epoch 5:  25%|██▍       | 24/97 [00:04<00:13,  5.32it/s, loss=6.17, v_num=0]
Epoch 5:  25%|██▍       | 24/97 [00:04<00:13,  5.25it/s, loss=6.17, v_num=0]
Epoch 5:  26%|██▌       | 25/97 [00:04<00:13,  5.38it/s, loss=6.17, v_num=0]
Epoch 5:  26%|██▌       | 25/97 [00:04<00:13,  5.31it/s, loss=6.16, v_num=0]
Epoch 5:  27%|██▋       | 26/97 [00:04<00:13,  5.43it/s, loss=6.16, v_num=0]
Epoch 5:  27%|██▋       | 26/97 [00:04<00:13,  5.37it/s, loss=6.16, v_num=0]
Epoch 5:  28%|██▊       | 27/97 [00:04<00:12,  5.48it/s, loss=6.16, v_num=0]
Epoch 5:  28%|██▊       | 27/97 [00:04<00:12,  5.42it/s, loss=6.16, v_num=0]
Epoch 5:  29%|██▉       | 28/97 [00:05<00:12,  5.54it/s, loss=6.16, v_num=0]
Epoch 5:  29%|██▉       | 28/97 [00:05<00:12,  5.48it/s, loss=6.15, v_num=0]
Epoch 5:  30%|██▉       | 29/97 [00:05<00:12,  5.58it/s, loss=6.15, v_num=0]
Epoch 5:  30%|██▉       | 29/97 [00:05<00:12,  5.53it/s, loss=6.14, v_num=0]
Epoch 5:  31%|███       | 30/97 [00:05<00:11,  5.63it/s, loss=6.14, v_num=0]
Epoch 5:  31%|███       | 30/97 [00:05<00:12,  5.57it/s, loss=6.14, v_num=0]
Epoch 5:  32%|███▏      | 31/97 [00:05<00:11,  5.67it/s, loss=6.14, v_num=0]
Epoch 5:  32%|███▏      | 31/97 [00:05<00:11,  5.62it/s, loss=6.13, v_num=0]
Epoch 5:  33%|███▎      | 32/97 [00:05<00:11,  5.71it/s, loss=6.13, v_num=0]
Epoch 5:  33%|███▎      | 32/97 [00:05<00:11,  5.66it/s, loss=6.12, v_num=0]
Epoch 5:  34%|███▍      | 33/97 [00:05<00:11,  5.75it/s, loss=6.12, v_num=0]
Epoch 5:  34%|███▍      | 33/97 [00:05<00:11,  5.70it/s, loss=6.12, v_num=0]
Epoch 5:  35%|███▌      | 34/97 [00:05<00:10,  5.78it/s, loss=6.12, v_num=0]
Epoch 5:  35%|███▌      | 34/97 [00:05<00:10,  5.73it/s, loss=6.12, v_num=0]
Epoch 5:  36%|███▌      | 35/97 [00:06<00:10,  5.82it/s, loss=6.12, v_num=0]
Epoch 5:  36%|███▌      | 35/97 [00:06<00:10,  5.77it/s, loss=6.11, v_num=0]
Epoch 5:  37%|███▋      | 36/97 [00:06<00:10,  5.85it/s, loss=6.11, v_num=0]
Epoch 5:  37%|███▋      | 36/97 [00:06<00:10,  5.80it/s, loss=6.1, v_num=0]
Epoch 5:  38%|███▊      | 37/97 [00:06<00:10,  5.89it/s, loss=6.1, v_num=0]
Epoch 5:  38%|███▊      | 37/97 [00:06<00:10,  5.84it/s, loss=6.11, v_num=0]
Epoch 5:  39%|███▉      | 38/97 [00:06<00:09,  5.92it/s, loss=6.11, v_num=0]
Epoch 5:  39%|███▉      | 38/97 [00:06<00:10,  5.87it/s, loss=6.11, v_num=0]
Epoch 5:  40%|████      | 39/97 [00:06<00:09,  5.95it/s, loss=6.11, v_num=0]
Epoch 5:  40%|████      | 39/97 [00:06<00:09,  5.90it/s, loss=6.1, v_num=0]
Epoch 5:  41%|████      | 40/97 [00:06<00:09,  5.98it/s, loss=6.1, v_num=0]
Epoch 5:  41%|████      | 40/97 [00:06<00:09,  5.93it/s, loss=6.09, v_num=0]
Epoch 5:  42%|████▏     | 41/97 [00:06<00:09,  6.00it/s, loss=6.09, v_num=0]
Epoch 5:  42%|████▏     | 41/97 [00:06<00:09,  5.96it/s, loss=6.1, v_num=0]
Epoch 5:  43%|████▎     | 42/97 [00:06<00:09,  6.03it/s, loss=6.1, v_num=0]
Epoch 5:  43%|████▎     | 42/97 [00:07<00:09,  5.98it/s, loss=6.09, v_num=0]
Epoch 5:  44%|████▍     | 43/97 [00:07<00:08,  6.06it/s, loss=6.09, v_num=0]
Epoch 5:  44%|████▍     | 43/97 [00:07<00:08,  6.01it/s, loss=6.1, v_num=0]
Epoch 5:  45%|████▌     | 44/97 [00:07<00:08,  6.08it/s, loss=6.1, v_num=0]
Epoch 5:  45%|████▌     | 44/97 [00:07<00:08,  6.03it/s, loss=6.09, v_num=0]
Epoch 5:  46%|████▋     | 45/97 [00:07<00:08,  6.10it/s, loss=6.09, v_num=0]
Epoch 5:  46%|████▋     | 45/97 [00:07<00:08,  6.06it/s, loss=6.09, v_num=0]
Epoch 5:  47%|████▋     | 46/97 [00:07<00:08,  6.13it/s, loss=6.09, v_num=0]
Epoch 5:  47%|████▋     | 46/97 [00:07<00:08,  6.08it/s, loss=6.09, v_num=0]
Epoch 5:  48%|████▊     | 47/97 [00:07<00:08,  6.14it/s, loss=6.09, v_num=0]
Epoch 5:  48%|████▊     | 47/97 [00:07<00:08,  6.10it/s, loss=6.09, v_num=0]
Epoch 5:  49%|████▉     | 48/97 [00:07<00:07,  6.16it/s, loss=6.09, v_num=0]
Epoch 5:  49%|████▉     | 48/97 [00:07<00:08,  6.12it/s, loss=6.09, v_num=0]
Epoch 5:  51%|█████     | 49/97 [00:07<00:07,  6.18it/s, loss=6.09, v_num=0]
Epoch 5:  51%|█████     | 49/97 [00:07<00:07,  6.14it/s, loss=6.09, v_num=0]
Epoch 5:  52%|█████▏    | 50/97 [00:08<00:07,  6.20it/s, loss=6.09, v_num=0]
Epoch 5:  52%|█████▏    | 50/97 [00:08<00:07,  6.16it/s, loss=6.08, v_num=0]
Epoch 5:  53%|█████▎    | 51/97 [00:08<00:07,  6.22it/s, loss=6.08, v_num=0]
Epoch 5:  53%|█████▎    | 51/97 [00:08<00:07,  6.18it/s, loss=6.08, v_num=0]
Epoch 5:  54%|█████▎    | 52/97 [00:08<00:07,  6.24it/s, loss=6.08, v_num=0]
Epoch 5:  54%|█████▎    | 52/97 [00:08<00:07,  6.20it/s, loss=6.07, v_num=0]
Epoch 5:  55%|█████▍    | 53/97 [00:08<00:07,  6.26it/s, loss=6.07, v_num=0]
Epoch 5:  55%|█████▍    | 53/97 [00:08<00:07,  6.22it/s, loss=6.07, v_num=0]
Epoch 5:  56%|█████▌    | 54/97 [00:08<00:06,  6.27it/s, loss=6.07, v_num=0]
Epoch 5:  56%|█████▌    | 54/97 [00:08<00:06,  6.23it/s, loss=6.06, v_num=0]
Epoch 5:  57%|█████▋    | 55/97 [00:08<00:06,  6.29it/s, loss=6.06, v_num=0]
Epoch 5:  57%|█████▋    | 55/97 [00:08<00:06,  6.25it/s, loss=6.06, v_num=0]
Epoch 5:  58%|█████▊    | 56/97 [00:08<00:06,  6.30it/s, loss=6.06, v_num=0]
Epoch 5:  58%|█████▊    | 56/97 [00:08<00:06,  6.26it/s, loss=6.06, v_num=0]
Epoch 5:  59%|█████▉    | 57/97 [00:09<00:06,  6.32it/s, loss=6.06, v_num=0]
Epoch 5:  59%|█████▉    | 57/97 [00:09<00:06,  6.28it/s, loss=6.05, v_num=0]
Epoch 5:  60%|█████▉    | 58/97 [00:09<00:06,  6.33it/s, loss=6.05, v_num=0]
Epoch 5:  60%|█████▉    | 58/97 [00:09<00:06,  6.29it/s, loss=6.06, v_num=0]
Epoch 5:  61%|██████    | 59/97 [00:09<00:05,  6.35it/s, loss=6.06, v_num=0]
Epoch 5:  61%|██████    | 59/97 [00:09<00:06,  6.31it/s, loss=6.05, v_num=0]
Epoch 5:  62%|██████▏   | 60/97 [00:09<00:05,  6.36it/s, loss=6.05, v_num=0]
Epoch 5:  62%|██████▏   | 60/97 [00:09<00:05,  6.32it/s, loss=6.05, v_num=0]
Epoch 5:  63%|██████▎   | 61/97 [00:09<00:05,  6.37it/s, loss=6.05, v_num=0]
Epoch 5:  63%|██████▎   | 61/97 [00:09<00:05,  6.34it/s, loss=6.04, v_num=0]
Epoch 5:  64%|██████▍   | 62/97 [00:09<00:05,  6.39it/s, loss=6.04, v_num=0]
Epoch 5:  64%|██████▍   | 62/97 [00:09<00:05,  6.35it/s, loss=6.04, v_num=0]
Epoch 5:  65%|██████▍   | 63/97 [00:09<00:05,  6.40it/s, loss=6.04, v_num=0]
Epoch 5:  65%|██████▍   | 63/97 [00:09<00:05,  6.36it/s, loss=6.02, v_num=0]
Epoch 5:  66%|██████▌   | 64/97 [00:09<00:05,  6.41it/s, loss=6.02, v_num=0]
Epoch 5:  66%|██████▌   | 64/97 [00:10<00:05,  6.38it/s, loss=6.02, v_num=0]
Epoch 5:  67%|██████▋   | 65/97 [00:10<00:04,  6.42it/s, loss=6.02, v_num=0]
Epoch 5:  67%|██████▋   | 65/97 [00:10<00:05,  6.39it/s, loss=6.01, v_num=0]
Epoch 5:  68%|██████▊   | 66/97 [00:10<00:04,  6.44it/s, loss=6.01, v_num=0]
Epoch 5:  68%|██████▊   | 66/97 [00:10<00:04,  6.40it/s, loss=6.01, v_num=0]
Epoch 5:  69%|██████▉   | 67/97 [00:10<00:04,  6.45it/s, loss=6.01, v_num=0]
Epoch 5:  69%|██████▉   | 67/97 [00:10<00:04,  6.41it/s, loss=6.02, v_num=0]
Epoch 5:  70%|███████   | 68/97 [00:10<00:04,  6.46it/s, loss=6.02, v_num=0]
Epoch 5:  70%|███████   | 68/97 [00:10<00:04,  6.43it/s, loss=6.02, v_num=0]
Epoch 5:  71%|███████   | 69/97 [00:10<00:04,  6.47it/s, loss=6.02, v_num=0]
Epoch 5:  71%|███████   | 69/97 [00:10<00:04,  6.44it/s, loss=6.02, v_num=0]
Epoch 5:  72%|███████▏  | 70/97 [00:10<00:04,  6.48it/s, loss=6.02, v_num=0]
Epoch 5:  72%|███████▏  | 70/97 [00:10<00:04,  6.45it/s, loss=6.02, v_num=0]
Epoch 5:  73%|███████▎  | 71/97 [00:10<00:04,  6.49it/s, loss=6.02, v_num=0]
Epoch 5:  73%|███████▎  | 71/97 [00:10<00:04,  6.46it/s, loss=6.02, v_num=0]
Epoch 5:  74%|███████▍  | 72/97 [00:11<00:03,  6.50it/s, loss=6.02, v_num=0]
Epoch 5:  74%|███████▍  | 72/97 [00:11<00:03,  6.47it/s, loss=6.01, v_num=0]
Epoch 5:  75%|███████▌  | 73/97 [00:11<00:03,  6.51it/s, loss=6.01, v_num=0]
Epoch 5:  75%|███████▌  | 73/97 [00:11<00:03,  6.48it/s, loss=6.01, v_num=0]
Epoch 5:  76%|███████▋  | 74/97 [00:11<00:03,  6.52it/s, loss=6.01, v_num=0]
Epoch 5:  76%|███████▋  | 74/97 [00:11<00:03,  6.49it/s, loss=6.01, v_num=0]
Epoch 5:  77%|███████▋  | 75/97 [00:11<00:03,  6.53it/s, loss=6.01, v_num=0]
Epoch 5:  77%|███████▋  | 75/97 [00:11<00:03,  6.50it/s, loss=6, v_num=0]
Epoch 5:  78%|███████▊  | 76/97 [00:11<00:03,  6.54it/s, loss=6, v_num=0]
Epoch 5:  78%|███████▊  | 76/97 [00:11<00:03,  6.51it/s, loss=6, v_num=0]
Epoch 5:  79%|███████▉  | 77/97 [00:11<00:03,  6.55it/s, loss=6, v_num=0]
Epoch 5:  79%|███████▉  | 77/97 [00:11<00:03,  6.52it/s, loss=6, v_num=0]
Epoch 5:  80%|████████  | 78/97 [00:11<00:02,  6.56it/s, loss=6, v_num=0]
Epoch 5:  80%|████████  | 78/97 [00:11<00:02,  6.53it/s, loss=5.99, v_num=0]
Epoch 5:  81%|████████▏ | 79/97 [00:12<00:02,  6.57it/s, loss=5.99, v_num=0]
Epoch 5:  81%|████████▏ | 79/97 [00:12<00:02,  6.54it/s, loss=5.99, v_num=0]
Epoch 5:  82%|████████▏ | 80/97 [00:12<00:02,  6.58it/s, loss=5.99, v_num=0]
Epoch 5:  82%|████████▏ | 80/97 [00:12<00:02,  6.55it/s, loss=5.99, v_num=0]
Epoch 5:  84%|████████▎ | 81/97 [00:12<00:02,  6.59it/s, loss=5.99, v_num=0]
Epoch 5:  84%|████████▎ | 81/97 [00:12<00:02,  6.56it/s, loss=6, v_num=0]
Epoch 5:  85%|████████▍ | 82/97 [00:12<00:02,  6.59it/s, loss=6, v_num=0]
Epoch 5:  85%|████████▍ | 82/97 [00:12<00:02,  6.56it/s, loss=6, v_num=0]
Epoch 5:  86%|████████▌ | 83/97 [00:12<00:02,  6.60it/s, loss=6, v_num=0]
Epoch 5:  86%|████████▌ | 83/97 [00:12<00:02,  6.57it/s, loss=6, v_num=0]
Epoch 5:  87%|████████▋ | 84/97 [00:12<00:01,  6.61it/s, loss=6, v_num=0]
Epoch 5:  87%|████████▋ | 84/97 [00:12<00:01,  6.58it/s, loss=5.99, v_num=0]
Epoch 5:  88%|████████▊ | 85/97 [00:12<00:01,  6.62it/s, loss=5.99, v_num=0]
Epoch 5:  88%|████████▊ | 85/97 [00:12<00:01,  6.59it/s, loss=5.99, v_num=0]
Epoch 5:  89%|████████▊ | 86/97 [00:12<00:01,  6.63it/s, loss=5.99, v_num=0]
Epoch 5:  89%|████████▊ | 86/97 [00:13<00:01,  6.60it/s, loss=5.98, v_num=0]
Epoch 5:  90%|████████▉ | 87/97 [00:13<00:01,  6.63it/s, loss=5.98, v_num=0]
Epoch 5:  90%|████████▉ | 87/97 [00:13<00:01,  6.61it/s, loss=5.96, v_num=0]
Epoch 5:  91%|█████████ | 88/97 [00:13<00:01,  6.64it/s, loss=5.96, v_num=0]
Epoch 5:  91%|█████████ | 88/97 [00:13<00:01,  6.61it/s, loss=5.97, v_num=0]
Epoch 5:  92%|█████████▏| 89/97 [00:13<00:01,  6.65it/s, loss=5.97, v_num=0]
Epoch 5:  92%|█████████▏| 89/97 [00:13<00:01,  6.62it/s, loss=5.96, v_num=0]
Epoch 5:  93%|█████████▎| 90/97 [00:13<00:01,  6.66it/s, loss=5.96, v_num=0]
Epoch 5:  93%|█████████▎| 90/97 [00:13<00:01,  6.63it/s, loss=5.96, v_num=0]
Epoch 5:  94%|█████████▍| 91/97 [00:13<00:00,  6.66it/s, loss=5.96, v_num=0]
Epoch 5:  94%|█████████▍| 91/97 [00:13<00:00,  6.64it/s, loss=5.96, v_num=0]
Epoch 5:  95%|█████████▍| 92/97 [00:13<00:00,  6.67it/s, loss=5.96, v_num=0]
Epoch 5:  95%|█████████▍| 92/97 [00:13<00:00,  6.64it/s, loss=5.95, v_num=0]
Epoch 5:  96%|█████████▌| 93/97 [00:13<00:00,  6.68it/s, loss=5.95, v_num=0]
Epoch 5:  96%|█████████▌| 93/97 [00:13<00:00,  6.65it/s, loss=5.95, v_num=0]
Epoch 5:  97%|█████████▋| 94/97 [00:14<00:00,  6.68it/s, loss=5.95, v_num=0]
Epoch 5:  97%|█████████▋| 94/97 [00:14<00:00,  6.66it/s, loss=5.95, v_num=0]
Epoch 5:  98%|█████████▊| 95/97 [00:14<00:00,  6.69it/s, loss=5.95, v_num=0]
Epoch 5:  98%|█████████▊| 95/97 [00:14<00:00,  6.66it/s, loss=5.95, v_num=0]
Epoch 5:  99%|█████████▉| 96/97 [00:14<00:00,  6.69it/s, loss=5.95, v_num=0]
Epoch 5:  99%|█████████▉| 96/97 [00:14<00:00,  6.67it/s, loss=5.95, v_num=0]
Epoch 5: 100%|██████████| 97/97 [00:14<00:00,  6.70it/s, loss=5.95, v_num=0]
Epoch 5: 100%|██████████| 97/97 [00:14<00:00,  6.68it/s, loss=5.95, v_num=0]
Epoch 5: 100%|██████████| 97/97 [00:14<00:00,  6.67it/s, loss=5.95, v_num=0]
Epoch 5:   0%|          | 0/97 [00:00<?, ?it/s, loss=5.95, v_num=0]
Epoch 6:   0%|          | 0/97 [00:00<?, ?it/s, loss=5.95, v_num=0]
Epoch 6:   1%|          | 1/97 [00:01<02:15,  1.41s/it, loss=5.95, v_num=0]
Epoch 6:   1%|          | 1/97 [00:01<02:20,  1.46s/it, loss=5.94, v_num=0]
Epoch 6:   2%|▏         | 2/97 [00:01<01:13,  1.29it/s, loss=5.94, v_num=0]
Epoch 6:   2%|▏         | 2/97 [00:01<01:16,  1.25it/s, loss=5.94, v_num=0]
Epoch 6:   3%|▎         | 3/97 [00:01<00:52,  1.79it/s, loss=5.94, v_num=0]
Epoch 6:   3%|▎         | 3/97 [00:01<00:54,  1.73it/s, loss=5.93, v_num=0]
Epoch 6:   4%|▍         | 4/97 [00:01<00:42,  2.21it/s, loss=5.93, v_num=0]
Epoch 6:   4%|▍         | 4/97 [00:01<00:43,  2.14it/s, loss=5.93, v_num=0]
Epoch 6:   5%|▌         | 5/97 [00:01<00:35,  2.56it/s, loss=5.93, v_num=0]
Epoch 6:   5%|▌         | 5/97 [00:02<00:36,  2.49it/s, loss=5.91, v_num=0]
Epoch 6:   6%|▌         | 6/97 [00:02<00:31,  2.88it/s, loss=5.91, v_num=0]
Epoch 6:   6%|▌         | 6/97 [00:02<00:32,  2.80it/s, loss=5.91, v_num=0]
Epoch 6:   7%|▋         | 7/97 [00:02<00:28,  3.15it/s, loss=5.91, v_num=0]
Epoch 6:   7%|▋         | 7/97 [00:02<00:29,  3.07it/s, loss=5.92, v_num=0]
Epoch 6:   8%|▊         | 8/97 [00:02<00:26,  3.40it/s, loss=5.92, v_num=0]
Epoch 6:   8%|▊         | 8/97 [00:02<00:26,  3.32it/s, loss=5.92, v_num=0]
Epoch 6:   9%|▉         | 9/97 [00:02<00:24,  3.59it/s, loss=5.92, v_num=0]
Epoch 6:   9%|▉         | 9/97 [00:02<00:25,  3.52it/s, loss=5.92, v_num=0]
Epoch 6:  10%|█         | 10/97 [00:02<00:22,  3.80it/s, loss=5.92, v_num=0]
Epoch 6:  10%|█         | 10/97 [00:02<00:23,  3.71it/s, loss=5.91, v_num=0]
Epoch 6:  11%|█▏        | 11/97 [00:02<00:21,  3.97it/s, loss=5.91, v_num=0]
Epoch 6:  11%|█▏        | 11/97 [00:02<00:22,  3.89it/s, loss=5.9, v_num=0]
Epoch 6:  12%|█▏        | 12/97 [00:02<00:20,  4.13it/s, loss=5.9, v_num=0]
Epoch 6:  12%|█▏        | 12/97 [00:02<00:20,  4.05it/s, loss=5.9, v_num=0]
Epoch 6:  13%|█▎        | 13/97 [00:03<00:19,  4.28it/s, loss=5.9, v_num=0]
Epoch 6:  13%|█▎        | 13/97 [00:03<00:20,  4.20it/s, loss=5.89, v_num=0]
Epoch 6:  14%|█▍        | 14/97 [00:03<00:18,  4.41it/s, loss=5.89, v_num=0]
Epoch 6:  14%|█▍        | 14/97 [00:03<00:19,  4.33it/s, loss=5.89, v_num=0]
Epoch 6:  15%|█▌        | 15/97 [00:03<00:18,  4.53it/s, loss=5.89, v_num=0]
Epoch 6:  15%|█▌        | 15/97 [00:03<00:18,  4.45it/s, loss=5.9, v_num=0]
Epoch 6:  16%|█▋        | 16/97 [00:03<00:17,  4.65it/s, loss=5.9, v_num=0]
Epoch 6:  16%|█▋        | 16/97 [00:03<00:17,  4.57it/s, loss=5.89, v_num=0]
Epoch 6:  18%|█▊        | 17/97 [00:03<00:16,  4.74it/s, loss=5.89, v_num=0]
Epoch 6:  18%|█▊        | 17/97 [00:03<00:17,  4.67it/s, loss=5.89, v_num=0]
Epoch 6:  19%|█▊        | 18/97 [00:03<00:16,  4.84it/s, loss=5.89, v_num=0]
Epoch 6:  19%|█▊        | 18/97 [00:03<00:16,  4.77it/s, loss=5.9, v_num=0]
Epoch 6:  20%|█▉        | 19/97 [00:03<00:15,  4.92it/s, loss=5.9, v_num=0]
Epoch 6:  20%|█▉        | 19/97 [00:03<00:16,  4.86it/s, loss=5.89, v_num=0]
Epoch 6:  21%|██        | 20/97 [00:03<00:15,  5.01it/s, loss=5.89, v_num=0]
Epoch 6:  21%|██        | 20/97 [00:04<00:15,  4.94it/s, loss=5.88, v_num=0]
Epoch 6:  22%|██▏       | 21/97 [00:04<00:14,  5.09it/s, loss=5.88, v_num=0]
Epoch 6:  22%|██▏       | 21/97 [00:04<00:15,  5.02it/s, loss=5.88, v_num=0]
Epoch 6:  23%|██▎       | 22/97 [00:04<00:14,  5.16it/s, loss=5.88, v_num=0]
Epoch 6:  23%|██▎       | 22/97 [00:04<00:14,  5.09it/s, loss=5.87, v_num=0]
Epoch 6:  24%|██▎       | 23/97 [00:04<00:14,  5.23it/s, loss=5.87, v_num=0]
Epoch 6:  24%|██▎       | 23/97 [00:04<00:14,  5.16it/s, loss=5.87, v_num=0]
Epoch 6:  25%|██▍       | 24/97 [00:04<00:13,  5.30it/s, loss=5.87, v_num=0]
Epoch 6:  25%|██▍       | 24/97 [00:04<00:13,  5.23it/s, loss=5.86, v_num=0]
Epoch 6:  26%|██▌       | 25/97 [00:04<00:13,  5.35it/s, loss=5.86, v_num=0]
Epoch 6:  26%|██▌       | 25/97 [00:04<00:13,  5.29it/s, loss=5.86, v_num=0]
Epoch 6:  27%|██▋       | 26/97 [00:04<00:13,  5.41it/s, loss=5.86, v_num=0]
Epoch 6:  27%|██▋       | 26/97 [00:04<00:13,  5.35it/s, loss=5.86, v_num=0]
Epoch 6:  28%|██▊       | 27/97 [00:04<00:12,  5.46it/s, loss=5.86, v_num=0]
Epoch 6:  28%|██▊       | 27/97 [00:04<00:12,  5.40it/s, loss=5.86, v_num=0]
Epoch 6:  29%|██▉       | 28/97 [00:05<00:12,  5.51it/s, loss=5.86, v_num=0]
Epoch 6:  29%|██▉       | 28/97 [00:05<00:12,  5.45it/s, loss=5.86, v_num=0]
Epoch 6:  30%|██▉       | 29/97 [00:05<00:12,  5.56it/s, loss=5.86, v_num=0]
Epoch 6:  30%|██▉       | 29/97 [00:05<00:12,  5.50it/s, loss=5.85, v_num=0]
Epoch 6:  31%|███       | 30/97 [00:05<00:11,  5.60it/s, loss=5.85, v_num=0]
Epoch 6:  31%|███       | 30/97 [00:05<00:12,  5.55it/s, loss=5.86, v_num=0]
Epoch 6:  32%|███▏      | 31/97 [00:05<00:11,  5.65it/s, loss=5.86, v_num=0]
Epoch 6:  32%|███▏      | 31/97 [00:05<00:11,  5.59it/s, loss=5.86, v_num=0]
Epoch 6:  33%|███▎      | 32/97 [00:05<00:11,  5.69it/s, loss=5.86, v_num=0]
Epoch 6:  33%|███▎      | 32/97 [00:05<00:11,  5.64it/s, loss=5.86, v_num=0]
Epoch 6:  34%|███▍      | 33/97 [00:05<00:11,  5.73it/s, loss=5.86, v_num=0]
Epoch 6:  34%|███▍      | 33/97 [00:05<00:11,  5.67it/s, loss=5.86, v_num=0]
Epoch 6:  35%|███▌      | 34/97 [00:05<00:10,  5.77it/s, loss=5.86, v_num=0]
Epoch 6:  35%|███▌      | 34/97 [00:05<00:11,  5.71it/s, loss=5.86, v_num=0]
Epoch 6:  36%|███▌      | 35/97 [00:06<00:10,  5.81it/s, loss=5.86, v_num=0]
Epoch 6:  36%|███▌      | 35/97 [00:06<00:10,  5.75it/s, loss=5.85, v_num=0]
Epoch 6:  37%|███▋      | 36/97 [00:06<00:10,  5.84it/s, loss=5.85, v_num=0]
Epoch 6:  37%|███▋      | 36/97 [00:06<00:10,  5.79it/s, loss=5.86, v_num=0]
Epoch 6:  38%|███▊      | 37/97 [00:06<00:10,  5.87it/s, loss=5.86, v_num=0]
Epoch 6:  38%|███▊      | 37/97 [00:06<00:10,  5.82it/s, loss=5.85, v_num=0]
Epoch 6:  39%|███▉      | 38/97 [00:06<00:09,  5.90it/s, loss=5.85, v_num=0]
Epoch 6:  39%|███▉      | 38/97 [00:06<00:10,  5.85it/s, loss=5.85, v_num=0]
Epoch 6:  40%|████      | 39/97 [00:06<00:09,  5.93it/s, loss=5.85, v_num=0]
Epoch 6:  40%|████      | 39/97 [00:06<00:09,  5.88it/s, loss=5.85, v_num=0]
Epoch 6:  41%|████      | 40/97 [00:06<00:09,  5.96it/s, loss=5.85, v_num=0]
Epoch 6:  41%|████      | 40/97 [00:06<00:09,  5.91it/s, loss=5.85, v_num=0]
Epoch 6:  42%|████▏     | 41/97 [00:06<00:09,  5.98it/s, loss=5.85, v_num=0]
Epoch 6:  42%|████▏     | 41/97 [00:06<00:09,  5.94it/s, loss=5.85, v_num=0]
Epoch 6:  43%|████▎     | 42/97 [00:06<00:09,  6.01it/s, loss=5.85, v_num=0]
Epoch 6:  43%|████▎     | 42/97 [00:07<00:09,  5.96it/s, loss=5.85, v_num=0]
Epoch 6:  44%|████▍     | 43/97 [00:07<00:08,  6.03it/s, loss=5.85, v_num=0]
Epoch 6:  44%|████▍     | 43/97 [00:07<00:09,  5.99it/s, loss=5.85, v_num=0]
Epoch 6:  45%|████▌     | 44/97 [00:07<00:08,  6.06it/s, loss=5.85, v_num=0]
Epoch 6:  45%|████▌     | 44/97 [00:07<00:08,  6.01it/s, loss=5.84, v_num=0]
Epoch 6:  46%|████▋     | 45/97 [00:07<00:08,  6.08it/s, loss=5.84, v_num=0]
Epoch 6:  46%|████▋     | 45/97 [00:07<00:08,  6.03it/s, loss=5.85, v_num=0]
Epoch 6:  47%|████▋     | 46/97 [00:07<00:08,  6.10it/s, loss=5.85, v_num=0]
Epoch 6:  47%|████▋     | 46/97 [00:07<00:08,  6.06it/s, loss=5.84, v_num=0]
Epoch 6:  48%|████▊     | 47/97 [00:07<00:08,  6.12it/s, loss=5.84, v_num=0]
Epoch 6:  48%|████▊     | 47/97 [00:07<00:08,  6.08it/s, loss=5.83, v_num=0]
Epoch 6:  49%|████▉     | 48/97 [00:07<00:07,  6.15it/s, loss=5.83, v_num=0]
Epoch 6:  49%|████▉     | 48/97 [00:07<00:08,  6.10it/s, loss=5.83, v_num=0]
Epoch 6:  51%|█████     | 49/97 [00:07<00:07,  6.17it/s, loss=5.83, v_num=0]
Epoch 6:  51%|█████     | 49/97 [00:08<00:07,  6.12it/s, loss=5.84, v_num=0]
Epoch 6:  52%|█████▏    | 50/97 [00:08<00:07,  6.19it/s, loss=5.84, v_num=0]
Epoch 6:  52%|█████▏    | 50/97 [00:08<00:07,  6.14it/s, loss=5.83, v_num=0]
Epoch 6:  53%|█████▎    | 51/97 [00:08<00:07,  6.21it/s, loss=5.83, v_num=0]
Epoch 6:  53%|█████▎    | 51/97 [00:08<00:07,  6.16it/s, loss=5.83, v_num=0]
Epoch 6:  54%|█████▎    | 52/97 [00:08<00:07,  6.22it/s, loss=5.83, v_num=0]
Epoch 6:  54%|█████▎    | 52/97 [00:08<00:07,  6.18it/s, loss=5.83, v_num=0]
Epoch 6:  55%|█████▍    | 53/97 [00:08<00:07,  6.24it/s, loss=5.83, v_num=0]
Epoch 6:  55%|█████▍    | 53/97 [00:08<00:07,  6.20it/s, loss=5.82, v_num=0]
Epoch 6:  56%|█████▌    | 54/97 [00:08<00:06,  6.26it/s, loss=5.82, v_num=0]
Epoch 6:  56%|█████▌    | 54/97 [00:08<00:06,  6.22it/s, loss=5.82, v_num=0]
Epoch 6:  57%|█████▋    | 55/97 [00:08<00:06,  6.27it/s, loss=5.82, v_num=0]
Epoch 6:  57%|█████▋    | 55/97 [00:08<00:06,  6.23it/s, loss=5.81, v_num=0]
Epoch 6:  58%|█████▊    | 56/97 [00:08<00:06,  6.29it/s, loss=5.81, v_num=0]
Epoch 6:  58%|█████▊    | 56/97 [00:08<00:06,  6.25it/s, loss=5.81, v_num=0]
Epoch 6:  59%|█████▉    | 57/97 [00:09<00:06,  6.30it/s, loss=5.81, v_num=0]
Epoch 6:  59%|█████▉    | 57/97 [00:09<00:06,  6.27it/s, loss=5.8, v_num=0]
Epoch 6:  60%|█████▉    | 58/97 [00:09<00:06,  6.32it/s, loss=5.8, v_num=0]
Epoch 6:  60%|█████▉    | 58/97 [00:09<00:06,  6.28it/s, loss=5.79, v_num=0]
Epoch 6:  61%|██████    | 59/97 [00:09<00:06,  6.33it/s, loss=5.79, v_num=0]
Epoch 6:  61%|██████    | 59/97 [00:09<00:06,  6.29it/s, loss=5.79, v_num=0]
Epoch 6:  62%|██████▏   | 60/97 [00:09<00:05,  6.34it/s, loss=5.79, v_num=0]
Epoch 6:  62%|██████▏   | 60/97 [00:09<00:05,  6.31it/s, loss=5.79, v_num=0]
Epoch 6:  63%|██████▎   | 61/97 [00:09<00:05,  6.36it/s, loss=5.79, v_num=0]
Epoch 6:  63%|██████▎   | 61/97 [00:09<00:05,  6.32it/s, loss=5.79, v_num=0]
Epoch 6:  64%|██████▍   | 62/97 [00:09<00:05,  6.37it/s, loss=5.79, v_num=0]
Epoch 6:  64%|██████▍   | 62/97 [00:09<00:05,  6.34it/s, loss=5.79, v_num=0]
Epoch 6:  65%|██████▍   | 63/97 [00:09<00:05,  6.38it/s, loss=5.79, v_num=0]
Epoch 6:  65%|██████▍   | 63/97 [00:09<00:05,  6.35it/s, loss=5.8, v_num=0]
Epoch 6:  66%|██████▌   | 64/97 [00:10<00:05,  6.40it/s, loss=5.8, v_num=0]
Epoch 6:  66%|██████▌   | 64/97 [00:10<00:05,  6.36it/s, loss=5.8, v_num=0]
Epoch 6:  67%|██████▋   | 65/97 [00:10<00:04,  6.41it/s, loss=5.8, v_num=0]
Epoch 6:  67%|██████▋   | 65/97 [00:10<00:05,  6.38it/s, loss=5.79, v_num=0]
Epoch 6:  68%|██████▊   | 66/97 [00:10<00:04,  6.42it/s, loss=5.79, v_num=0]
Epoch 6:  68%|██████▊   | 66/97 [00:10<00:04,  6.39it/s, loss=5.79, v_num=0]
Epoch 6:  69%|██████▉   | 67/97 [00:10<00:04,  6.44it/s, loss=5.79, v_num=0]
Epoch 6:  69%|██████▉   | 67/97 [00:10<00:04,  6.40it/s, loss=5.79, v_num=0]
Epoch 6:  70%|███████   | 68/97 [00:10<00:04,  6.45it/s, loss=5.79, v_num=0]
Epoch 6:  70%|███████   | 68/97 [00:10<00:04,  6.41it/s, loss=5.77, v_num=0]
Epoch 6:  71%|███████   | 69/97 [00:10<00:04,  6.46it/s, loss=5.77, v_num=0]
Epoch 6:  71%|███████   | 69/97 [00:10<00:04,  6.42it/s, loss=5.77, v_num=0]
Epoch 6:  72%|███████▏  | 70/97 [00:10<00:04,  6.47it/s, loss=5.77, v_num=0]
Epoch 6:  72%|███████▏  | 70/97 [00:10<00:04,  6.44it/s, loss=5.77, v_num=0]
Epoch 6:  73%|███████▎  | 71/97 [00:10<00:04,  6.48it/s, loss=5.77, v_num=0]
Epoch 6:  73%|███████▎  | 71/97 [00:11<00:04,  6.45it/s, loss=5.77, v_num=0]
Epoch 6:  74%|███████▍  | 72/97 [00:11<00:03,  6.49it/s, loss=5.77, v_num=0]
Epoch 6:  74%|███████▍  | 72/97 [00:11<00:03,  6.46it/s, loss=5.76, v_num=0]
Epoch 6:  75%|███████▌  | 73/97 [00:11<00:03,  6.50it/s, loss=5.76, v_num=0]
Epoch 6:  75%|███████▌  | 73/97 [00:11<00:03,  6.47it/s, loss=5.75, v_num=0]
Epoch 6:  76%|███████▋  | 74/97 [00:11<00:03,  6.51it/s, loss=5.75, v_num=0]
Epoch 6:  76%|███████▋  | 74/97 [00:11<00:03,  6.48it/s, loss=5.75, v_num=0]
Epoch 6:  77%|███████▋  | 75/97 [00:11<00:03,  6.52it/s, loss=5.75, v_num=0]
Epoch 6:  77%|███████▋  | 75/97 [00:11<00:03,  6.49it/s, loss=5.74, v_num=0]
Epoch 6:  78%|███████▊  | 76/97 [00:11<00:03,  6.52it/s, loss=5.74, v_num=0]
Epoch 6:  78%|███████▊  | 76/97 [00:11<00:03,  6.49it/s, loss=5.74, v_num=0]
Epoch 6:  79%|███████▉  | 77/97 [00:11<00:03,  6.53it/s, loss=5.74, v_num=0]
Epoch 6:  79%|███████▉  | 77/97 [00:11<00:03,  6.50it/s, loss=5.74, v_num=0]
Epoch 6:  80%|████████  | 78/97 [00:11<00:02,  6.54it/s, loss=5.74, v_num=0]
Epoch 6:  80%|████████  | 78/97 [00:11<00:02,  6.51it/s, loss=5.73, v_num=0]
Epoch 6:  81%|████████▏ | 79/97 [00:12<00:02,  6.55it/s, loss=5.73, v_num=0]
Epoch 6:  81%|████████▏ | 79/97 [00:12<00:02,  6.52it/s, loss=5.73, v_num=0]
Epoch 6:  82%|████████▏ | 80/97 [00:12<00:02,  6.56it/s, loss=5.73, v_num=0]
Epoch 6:  82%|████████▏ | 80/97 [00:12<00:02,  6.53it/s, loss=5.73, v_num=0]
Epoch 6:  84%|████████▎ | 81/97 [00:12<00:02,  6.57it/s, loss=5.73, v_num=0]
Epoch 6:  84%|████████▎ | 81/97 [00:12<00:02,  6.54it/s, loss=5.72, v_num=0]
Epoch 6:  85%|████████▍ | 82/97 [00:12<00:02,  6.58it/s, loss=5.72, v_num=0]
Epoch 6:  85%|████████▍ | 82/97 [00:12<00:02,  6.55it/s, loss=5.72, v_num=0]
Epoch 6:  86%|████████▌ | 83/97 [00:12<00:02,  6.58it/s, loss=5.72, v_num=0]
Epoch 6:  86%|████████▌ | 83/97 [00:12<00:02,  6.55it/s, loss=5.72, v_num=0]
Epoch 6:  87%|████████▋ | 84/97 [00:12<00:01,  6.59it/s, loss=5.72, v_num=0]
Epoch 6:  87%|████████▋ | 84/97 [00:12<00:01,  6.56it/s, loss=5.72, v_num=0]
Epoch 6:  88%|████████▊ | 85/97 [00:12<00:01,  6.60it/s, loss=5.72, v_num=0]
Epoch 6:  88%|████████▊ | 85/97 [00:12<00:01,  6.57it/s, loss=5.72, v_num=0]
Epoch 6:  89%|████████▊ | 86/97 [00:13<00:01,  6.61it/s, loss=5.72, v_num=0]
Epoch 6:  89%|████████▊ | 86/97 [00:13<00:01,  6.58it/s, loss=5.72, v_num=0]
Epoch 6:  90%|████████▉ | 87/97 [00:13<00:01,  6.61it/s, loss=5.72, v_num=0]
Epoch 6:  90%|████████▉ | 87/97 [00:13<00:01,  6.59it/s, loss=5.71, v_num=0]
Epoch 6:  91%|█████████ | 88/97 [00:13<00:01,  6.62it/s, loss=5.71, v_num=0]
Epoch 6:  91%|█████████ | 88/97 [00:13<00:01,  6.59it/s, loss=5.72, v_num=0]
Epoch 6:  92%|█████████▏| 89/97 [00:13<00:01,  6.63it/s, loss=5.72, v_num=0]
Epoch 6:  92%|█████████▏| 89/97 [00:13<00:01,  6.60it/s, loss=5.72, v_num=0]
Epoch 6:  93%|█████████▎| 90/97 [00:13<00:01,  6.64it/s, loss=5.72, v_num=0]
Epoch 6:  93%|█████████▎| 90/97 [00:13<00:01,  6.61it/s, loss=5.72, v_num=0]
Epoch 6:  94%|█████████▍| 91/97 [00:13<00:00,  6.64it/s, loss=5.72, v_num=0]
Epoch 6:  94%|█████████▍| 91/97 [00:13<00:00,  6.62it/s, loss=5.71, v_num=0]
Epoch 6:  95%|█████████▍| 92/97 [00:13<00:00,  6.65it/s, loss=5.71, v_num=0]
Epoch 6:  95%|█████████▍| 92/97 [00:13<00:00,  6.62it/s, loss=5.71, v_num=0]
Epoch 6:  96%|█████████▌| 93/97 [00:13<00:00,  6.65it/s, loss=5.71, v_num=0]
Epoch 6:  96%|█████████▌| 93/97 [00:14<00:00,  6.63it/s, loss=5.71, v_num=0]
Epoch 6:  97%|█████████▋| 94/97 [00:14<00:00,  6.66it/s, loss=5.71, v_num=0]
Epoch 6:  97%|█████████▋| 94/97 [00:14<00:00,  6.63it/s, loss=5.71, v_num=0]
Epoch 6:  98%|█████████▊| 95/97 [00:14<00:00,  6.67it/s, loss=5.71, v_num=0]
Epoch 6:  98%|█████████▊| 95/97 [00:14<00:00,  6.64it/s, loss=5.72, v_num=0]
Epoch 6:  99%|█████████▉| 96/97 [00:14<00:00,  6.67it/s, loss=5.72, v_num=0]
Epoch 6:  99%|█████████▉| 96/97 [00:14<00:00,  6.65it/s, loss=5.72, v_num=0]
Epoch 6: 100%|██████████| 97/97 [00:14<00:00,  6.68it/s, loss=5.72, v_num=0]
Epoch 6: 100%|██████████| 97/97 [00:14<00:00,  6.65it/s, loss=5.71, v_num=0]
Epoch 6: 100%|██████████| 97/97 [00:14<00:00,  6.65it/s, loss=5.71, v_num=0]
Epoch 6:   0%|          | 0/97 [00:00<?, ?it/s, loss=5.71, v_num=0]
Epoch 7:   0%|          | 0/97 [00:00<?, ?it/s, loss=5.71, v_num=0]
Epoch 7:   1%|          | 1/97 [00:01<02:13,  1.39s/it, loss=5.71, v_num=0]
Epoch 7:   1%|          | 1/97 [00:01<02:18,  1.44s/it, loss=5.71, v_num=0]
Epoch 7:   2%|▏         | 2/97 [00:01<01:12,  1.31it/s, loss=5.71, v_num=0]
Epoch 7:   2%|▏         | 2/97 [00:01<01:15,  1.27it/s, loss=5.7, v_num=0]
Epoch 7:   3%|▎         | 3/97 [00:01<00:51,  1.81it/s, loss=5.7, v_num=0]
Epoch 7:   3%|▎         | 3/97 [00:01<00:53,  1.75it/s, loss=5.7, v_num=0]
Epoch 7:   4%|▍         | 4/97 [00:01<00:41,  2.23it/s, loss=5.7, v_num=0]
Epoch 7:   4%|▍         | 4/97 [00:01<00:43,  2.16it/s, loss=5.7, v_num=0]
Epoch 7:   5%|▌         | 5/97 [00:01<00:35,  2.59it/s, loss=5.7, v_num=0]
Epoch 7:   5%|▌         | 5/97 [00:01<00:36,  2.51it/s, loss=5.7, v_num=0]
Epoch 7:   6%|▌         | 6/97 [00:02<00:31,  2.89it/s, loss=5.7, v_num=0]
Epoch 7:   6%|▌         | 6/97 [00:02<00:32,  2.82it/s, loss=5.69, v_num=0]
Epoch 7:   7%|▋         | 7/97 [00:02<00:28,  3.17it/s, loss=5.69, v_num=0]
Epoch 7:   7%|▋         | 7/97 [00:02<00:29,  3.09it/s, loss=5.69, v_num=0]
Epoch 7:   8%|▊         | 8/97 [00:02<00:26,  3.42it/s, loss=5.69, v_num=0]
Epoch 7:   8%|▊         | 8/97 [00:02<00:26,  3.34it/s, loss=5.68, v_num=0]
Epoch 7:   9%|▉         | 9/97 [00:02<00:24,  3.62it/s, loss=5.68, v_num=0]
Epoch 7:   9%|▉         | 9/97 [00:02<00:24,  3.54it/s, loss=5.68, v_num=0]
Epoch 7:  10%|█         | 10/97 [00:02<00:22,  3.81it/s, loss=5.68, v_num=0]
Epoch 7:  10%|█         | 10/97 [00:02<00:23,  3.73it/s, loss=5.69, v_num=0]
Epoch 7:  11%|█▏        | 11/97 [00:02<00:21,  3.99it/s, loss=5.69, v_num=0]
Epoch 7:  11%|█▏        | 11/97 [00:02<00:22,  3.91it/s, loss=5.69, v_num=0]
Epoch 7:  12%|█▏        | 12/97 [00:02<00:20,  4.15it/s, loss=5.69, v_num=0]
Epoch 7:  12%|█▏        | 12/97 [00:02<00:20,  4.07it/s, loss=5.68, v_num=0]
Epoch 7:  13%|█▎        | 13/97 [00:03<00:19,  4.28it/s, loss=5.68, v_num=0]
Epoch 7:  13%|█▎        | 13/97 [00:03<00:19,  4.20it/s, loss=5.67, v_num=0]
Epoch 7:  14%|█▍        | 14/97 [00:03<00:18,  4.42it/s, loss=5.67, v_num=0]
Epoch 7:  14%|█▍        | 14/97 [00:03<00:19,  4.34it/s, loss=5.68, v_num=0]
Epoch 7:  15%|█▌        | 15/97 [00:03<00:18,  4.53it/s, loss=5.68, v_num=0]
Epoch 7:  15%|█▌        | 15/97 [00:03<00:18,  4.46it/s, loss=5.67, v_num=0]
Epoch 7:  16%|█▋        | 16/97 [00:03<00:17,  4.65it/s, loss=5.67, v_num=0]
Epoch 7:  16%|█▋        | 16/97 [00:03<00:17,  4.57it/s, loss=5.67, v_num=0]
Epoch 7:  18%|█▊        | 17/97 [00:03<00:16,  4.75it/s, loss=5.67, v_num=0]
Epoch 7:  18%|█▊        | 17/97 [00:03<00:17,  4.68it/s, loss=5.67, v_num=0]
Epoch 7:  19%|█▊        | 18/97 [00:03<00:16,  4.84it/s, loss=5.67, v_num=0]
Epoch 7:  19%|█▊        | 18/97 [00:03<00:16,  4.77it/s, loss=5.66, v_num=0]
Epoch 7:  20%|█▉        | 19/97 [00:03<00:15,  4.93it/s, loss=5.66, v_num=0]
Epoch 7:  20%|█▉        | 19/97 [00:03<00:16,  4.86it/s, loss=5.65, v_num=0]
Epoch 7:  21%|██        | 20/97 [00:03<00:15,  5.01it/s, loss=5.65, v_num=0]
Epoch 7:  21%|██        | 20/97 [00:04<00:15,  4.94it/s, loss=5.66, v_num=0]
Epoch 7:  22%|██▏       | 21/97 [00:04<00:14,  5.08it/s, loss=5.66, v_num=0]
Epoch 7:  22%|██▏       | 21/97 [00:04<00:15,  5.02it/s, loss=5.65, v_num=0]
Epoch 7:  23%|██▎       | 22/97 [00:04<00:14,  5.16it/s, loss=5.65, v_num=0]
Epoch 7:  23%|██▎       | 22/97 [00:04<00:14,  5.09it/s, loss=5.65, v_num=0]
Epoch 7:  24%|██▎       | 23/97 [00:04<00:14,  5.23it/s, loss=5.65, v_num=0]
Epoch 7:  24%|██▎       | 23/97 [00:04<00:14,  5.16it/s, loss=5.65, v_num=0]
Epoch 7:  25%|██▍       | 24/97 [00:04<00:13,  5.29it/s, loss=5.65, v_num=0]
Epoch 7:  25%|██▍       | 24/97 [00:04<00:13,  5.22it/s, loss=5.64, v_num=0]
Epoch 7:  26%|██▌       | 25/97 [00:04<00:13,  5.34it/s, loss=5.64, v_num=0]
Epoch 7:  26%|██▌       | 25/97 [00:04<00:13,  5.28it/s, loss=5.63, v_num=0]
Epoch 7:  27%|██▋       | 26/97 [00:04<00:13,  5.40it/s, loss=5.63, v_num=0]
Epoch 7:  27%|██▋       | 26/97 [00:04<00:13,  5.33it/s, loss=5.63, v_num=0]
Epoch 7:  28%|██▊       | 27/97 [00:04<00:12,  5.45it/s, loss=5.63, v_num=0]
Epoch 7:  28%|██▊       | 27/97 [00:05<00:12,  5.39it/s, loss=5.62, v_num=0]
Epoch 7:  29%|██▉       | 28/97 [00:05<00:12,  5.51it/s, loss=5.62, v_num=0]
Epoch 7:  29%|██▉       | 28/97 [00:05<00:12,  5.44it/s, loss=5.61, v_num=0]
Epoch 7:  30%|██▉       | 29/97 [00:05<00:12,  5.54it/s, loss=5.61, v_num=0]
Epoch 7:  30%|██▉       | 29/97 [00:05<00:12,  5.49it/s, loss=5.6, v_num=0]
Epoch 7:  31%|███       | 30/97 [00:05<00:11,  5.59it/s, loss=5.6, v_num=0]
Epoch 7:  31%|███       | 30/97 [00:05<00:12,  5.53it/s, loss=5.6, v_num=0]
Epoch 7:  32%|███▏      | 31/97 [00:05<00:11,  5.63it/s, loss=5.6, v_num=0]
Epoch 7:  32%|███▏      | 31/97 [00:05<00:11,  5.58it/s, loss=5.6, v_num=0]
Epoch 7:  33%|███▎      | 32/97 [00:05<00:11,  5.67it/s, loss=5.6, v_num=0]
Epoch 7:  33%|███▎      | 32/97 [00:05<00:11,  5.62it/s, loss=5.6, v_num=0]
Epoch 7:  34%|███▍      | 33/97 [00:05<00:11,  5.72it/s, loss=5.6, v_num=0]
Epoch 7:  34%|███▍      | 33/97 [00:05<00:11,  5.66it/s, loss=5.6, v_num=0]
Epoch 7:  35%|███▌      | 34/97 [00:05<00:10,  5.75it/s, loss=5.6, v_num=0]
Epoch 7:  35%|███▌      | 34/97 [00:05<00:11,  5.70it/s, loss=5.59, v_num=0]
Epoch 7:  36%|███▌      | 35/97 [00:06<00:10,  5.79it/s, loss=5.59, v_num=0]
Epoch 7:  36%|███▌      | 35/97 [00:06<00:10,  5.73it/s, loss=5.58, v_num=0]
Epoch 7:  37%|███▋      | 36/97 [00:06<00:10,  5.82it/s, loss=5.58, v_num=0]
Epoch 7:  37%|███▋      | 36/97 [00:06<00:10,  5.77it/s, loss=5.59, v_num=0]
Epoch 7:  38%|███▊      | 37/97 [00:06<00:10,  5.85it/s, loss=5.59, v_num=0]
Epoch 7:  38%|███▊      | 37/97 [00:06<00:10,  5.80it/s, loss=5.59, v_num=0]
Epoch 7:  39%|███▉      | 38/97 [00:06<00:10,  5.89it/s, loss=5.59, v_num=0]
Epoch 7:  39%|███▉      | 38/97 [00:06<00:10,  5.83it/s, loss=5.59, v_num=0]
Epoch 7:  40%|████      | 39/97 [00:06<00:09,  5.91it/s, loss=5.59, v_num=0]
Epoch 7:  40%|████      | 39/97 [00:06<00:09,  5.86it/s, loss=5.59, v_num=0]
Epoch 7:  41%|████      | 40/97 [00:06<00:09,  5.95it/s, loss=5.59, v_num=0]
Epoch 7:  41%|████      | 40/97 [00:06<00:09,  5.89it/s, loss=5.58, v_num=0]
Epoch 7:  42%|████▏     | 41/97 [00:06<00:09,  5.97it/s, loss=5.58, v_num=0]
Epoch 7:  42%|████▏     | 41/97 [00:06<00:09,  5.92it/s, loss=5.58, v_num=0]
Epoch 7:  43%|████▎     | 42/97 [00:07<00:09,  5.99it/s, loss=5.58, v_num=0]
Epoch 7:  43%|████▎     | 42/97 [00:07<00:09,  5.94it/s, loss=5.58, v_num=0]
Epoch 7:  44%|████▍     | 43/97 [00:07<00:08,  6.02it/s, loss=5.58, v_num=0]
Epoch 7:  44%|████▍     | 43/97 [00:07<00:09,  5.97it/s, loss=5.58, v_num=0]
Epoch 7:  45%|████▌     | 44/97 [00:07<00:08,  6.04it/s, loss=5.58, v_num=0]
Epoch 7:  45%|████▌     | 44/97 [00:07<00:08,  6.00it/s, loss=5.57, v_num=0]
Epoch 7:  46%|████▋     | 45/97 [00:07<00:08,  6.07it/s, loss=5.57, v_num=0]
Epoch 7:  46%|████▋     | 45/97 [00:07<00:08,  6.02it/s, loss=5.57, v_num=0]
Epoch 7:  47%|████▋     | 46/97 [00:07<00:08,  6.09it/s, loss=5.57, v_num=0]
Epoch 7:  47%|████▋     | 46/97 [00:07<00:08,  6.04it/s, loss=5.57, v_num=0]
Epoch 7:  48%|████▊     | 47/97 [00:07<00:08,  6.11it/s, loss=5.57, v_num=0]
Epoch 7:  48%|████▊     | 47/97 [00:07<00:08,  6.07it/s, loss=5.57, v_num=0]
Epoch 7:  49%|████▉     | 48/97 [00:07<00:07,  6.13it/s, loss=5.57, v_num=0]
Epoch 7:  49%|████▉     | 48/97 [00:07<00:08,  6.09it/s, loss=5.58, v_num=0]
Epoch 7:  51%|█████     | 49/97 [00:07<00:07,  6.15it/s, loss=5.58, v_num=0]
Epoch 7:  51%|█████     | 49/97 [00:08<00:07,  6.11it/s, loss=5.57, v_num=0]
Epoch 7:  52%|█████▏    | 50/97 [00:08<00:07,  6.17it/s, loss=5.57, v_num=0]
Epoch 7:  52%|█████▏    | 50/97 [00:08<00:07,  6.13it/s, loss=5.57, v_num=0]
Epoch 7:  53%|█████▎    | 51/97 [00:08<00:07,  6.18it/s, loss=5.57, v_num=0]
Epoch 7:  53%|█████▎    | 51/97 [00:08<00:07,  6.14it/s, loss=5.56, v_num=0]
Epoch 7:  54%|█████▎    | 52/97 [00:08<00:07,  6.20it/s, loss=5.56, v_num=0]
Epoch 7:  54%|█████▎    | 52/97 [00:08<00:07,  6.16it/s, loss=5.55, v_num=0]
Epoch 7:  55%|█████▍    | 53/97 [00:08<00:07,  6.22it/s, loss=5.55, v_num=0]
Epoch 7:  55%|█████▍    | 53/97 [00:08<00:07,  6.18it/s, loss=5.54, v_num=0]
Epoch 7:  56%|█████▌    | 54/97 [00:08<00:06,  6.23it/s, loss=5.54, v_num=0]
Epoch 7:  56%|█████▌    | 54/97 [00:08<00:06,  6.19it/s, loss=5.55, v_num=0]
Epoch 7:  57%|█████▋    | 55/97 [00:08<00:06,  6.25it/s, loss=5.55, v_num=0]
Epoch 7:  57%|█████▋    | 55/97 [00:08<00:06,  6.21it/s, loss=5.55, v_num=0]
Epoch 7:  58%|█████▊    | 56/97 [00:08<00:06,  6.26it/s, loss=5.55, v_num=0]
Epoch 7:  58%|█████▊    | 56/97 [00:08<00:06,  6.23it/s, loss=5.55, v_num=0]
Epoch 7:  59%|█████▉    | 57/97 [00:09<00:06,  6.28it/s, loss=5.55, v_num=0]
Epoch 7:  59%|█████▉    | 57/97 [00:09<00:06,  6.24it/s, loss=5.55, v_num=0]
Epoch 7:  60%|█████▉    | 58/97 [00:09<00:06,  6.30it/s, loss=5.55, v_num=0]
Epoch 7:  60%|█████▉    | 58/97 [00:09<00:06,  6.26it/s, loss=5.55, v_num=0]
Epoch 7:  61%|██████    | 59/97 [00:09<00:06,  6.31it/s, loss=5.55, v_num=0]
Epoch 7:  61%|██████    | 59/97 [00:09<00:06,  6.27it/s, loss=5.55, v_num=0]
Epoch 7:  62%|██████▏   | 60/97 [00:09<00:05,  6.33it/s, loss=5.55, v_num=0]
Epoch 7:  62%|██████▏   | 60/97 [00:09<00:05,  6.29it/s, loss=5.55, v_num=0]
Epoch 7:  63%|██████▎   | 61/97 [00:09<00:05,  6.34it/s, loss=5.55, v_num=0]
Epoch 7:  63%|██████▎   | 61/97 [00:09<00:05,  6.30it/s, loss=5.55, v_num=0]
Epoch 7:  64%|██████▍   | 62/97 [00:09<00:05,  6.36it/s, loss=5.55, v_num=0]
Epoch 7:  64%|██████▍   | 62/97 [00:09<00:05,  6.32it/s, loss=5.54, v_num=0]
Epoch 7:  65%|██████▍   | 63/97 [00:09<00:05,  6.37it/s, loss=5.54, v_num=0]
Epoch 7:  65%|██████▍   | 63/97 [00:09<00:05,  6.33it/s, loss=5.55, v_num=0]
Epoch 7:  66%|██████▌   | 64/97 [00:10<00:05,  6.38it/s, loss=5.55, v_num=0]
Epoch 7:  66%|██████▌   | 64/97 [00:10<00:05,  6.34it/s, loss=5.55, v_num=0]
Epoch 7:  67%|██████▋   | 65/97 [00:10<00:05,  6.39it/s, loss=5.55, v_num=0]
Epoch 7:  67%|██████▋   | 65/97 [00:10<00:05,  6.36it/s, loss=5.55, v_num=0]
Epoch 7:  68%|██████▊   | 66/97 [00:10<00:04,  6.40it/s, loss=5.55, v_num=0]
Epoch 7:  68%|██████▊   | 66/97 [00:10<00:04,  6.37it/s, loss=5.54, v_num=0]
Epoch 7:  69%|██████▉   | 67/97 [00:10<00:04,  6.42it/s, loss=5.54, v_num=0]
Epoch 7:  69%|██████▉   | 67/97 [00:10<00:04,  6.38it/s, loss=5.54, v_num=0]
Epoch 7:  70%|███████   | 68/97 [00:10<00:04,  6.43it/s, loss=5.54, v_num=0]
Epoch 7:  70%|███████   | 68/97 [00:10<00:04,  6.39it/s, loss=5.54, v_num=0]
Epoch 7:  71%|███████   | 69/97 [00:10<00:04,  6.44it/s, loss=5.54, v_num=0]
Epoch 7:  71%|███████   | 69/97 [00:10<00:04,  6.41it/s, loss=5.53, v_num=0]
Epoch 7:  72%|███████▏  | 70/97 [00:10<00:04,  6.45it/s, loss=5.53, v_num=0]
Epoch 7:  72%|███████▏  | 70/97 [00:10<00:04,  6.42it/s, loss=5.52, v_num=0]
Epoch 7:  73%|███████▎  | 71/97 [00:10<00:04,  6.46it/s, loss=5.52, v_num=0]
Epoch 7:  73%|███████▎  | 71/97 [00:11<00:04,  6.43it/s, loss=5.53, v_num=0]
Epoch 7:  74%|███████▍  | 72/97 [00:11<00:03,  6.47it/s, loss=5.53, v_num=0]
Epoch 7:  74%|███████▍  | 72/97 [00:11<00:03,  6.44it/s, loss=5.53, v_num=0]
Epoch 7:  75%|███████▌  | 73/97 [00:11<00:03,  6.48it/s, loss=5.53, v_num=0]
Epoch 7:  75%|███████▌  | 73/97 [00:11<00:03,  6.45it/s, loss=5.53, v_num=0]
Epoch 7:  76%|███████▋  | 74/97 [00:11<00:03,  6.49it/s, loss=5.53, v_num=0]
Epoch 7:  76%|███████▋  | 74/97 [00:11<00:03,  6.46it/s, loss=5.52, v_num=0]
Epoch 7:  77%|███████▋  | 75/97 [00:11<00:03,  6.50it/s, loss=5.52, v_num=0]
Epoch 7:  77%|███████▋  | 75/97 [00:11<00:03,  6.47it/s, loss=5.53, v_num=0]
Epoch 7:  78%|███████▊  | 76/97 [00:11<00:03,  6.51it/s, loss=5.53, v_num=0]
Epoch 7:  78%|███████▊  | 76/97 [00:11<00:03,  6.48it/s, loss=5.52, v_num=0]
Epoch 7:  79%|███████▉  | 77/97 [00:11<00:03,  6.52it/s, loss=5.52, v_num=0]
Epoch 7:  79%|███████▉  | 77/97 [00:11<00:03,  6.49it/s, loss=5.51, v_num=0]
Epoch 7:  80%|████████  | 78/97 [00:11<00:02,  6.53it/s, loss=5.51, v_num=0]
Epoch 7:  80%|████████  | 78/97 [00:11<00:02,  6.50it/s, loss=5.5, v_num=0]
Epoch 7:  81%|████████▏ | 79/97 [00:12<00:02,  6.54it/s, loss=5.5, v_num=0]
Epoch 7:  81%|████████▏ | 79/97 [00:12<00:02,  6.51it/s, loss=5.5, v_num=0]
Epoch 7:  82%|████████▏ | 80/97 [00:12<00:02,  6.55it/s, loss=5.5, v_num=0]
Epoch 7:  82%|████████▏ | 80/97 [00:12<00:02,  6.52it/s, loss=5.5, v_num=0]
Epoch 7:  84%|████████▎ | 81/97 [00:12<00:02,  6.56it/s, loss=5.5, v_num=0]
Epoch 7:  84%|████████▎ | 81/97 [00:12<00:02,  6.53it/s, loss=5.51, v_num=0]
Epoch 7:  85%|████████▍ | 82/97 [00:12<00:02,  6.57it/s, loss=5.51, v_num=0]
Epoch 7:  85%|████████▍ | 82/97 [00:12<00:02,  6.54it/s, loss=5.5, v_num=0]
Epoch 7:  86%|████████▌ | 83/97 [00:12<00:02,  6.58it/s, loss=5.5, v_num=0]
Epoch 7:  86%|████████▌ | 83/97 [00:12<00:02,  6.54it/s, loss=5.5, v_num=0]
Epoch 7:  87%|████████▋ | 84/97 [00:12<00:01,  6.58it/s, loss=5.5, v_num=0]
Epoch 7:  87%|████████▋ | 84/97 [00:12<00:01,  6.55it/s, loss=5.5, v_num=0]
Epoch 7:  88%|████████▊ | 85/97 [00:12<00:01,  6.59it/s, loss=5.5, v_num=0]
Epoch 7:  88%|████████▊ | 85/97 [00:12<00:01,  6.56it/s, loss=5.5, v_num=0]
Epoch 7:  89%|████████▊ | 86/97 [00:13<00:01,  6.60it/s, loss=5.5, v_num=0]
Epoch 7:  89%|████████▊ | 86/97 [00:13<00:01,  6.57it/s, loss=5.5, v_num=0]
Epoch 7:  90%|████████▉ | 87/97 [00:13<00:01,  6.61it/s, loss=5.5, v_num=0]
Epoch 7:  90%|████████▉ | 87/97 [00:13<00:01,  6.58it/s, loss=5.49, v_num=0]
Epoch 7:  91%|█████████ | 88/97 [00:13<00:01,  6.62it/s, loss=5.49, v_num=0]
Epoch 7:  91%|█████████ | 88/97 [00:13<00:01,  6.59it/s, loss=5.48, v_num=0]
Epoch 7:  92%|█████████▏| 89/97 [00:13<00:01,  6.62it/s, loss=5.48, v_num=0]
Epoch 7:  92%|█████████▏| 89/97 [00:13<00:01,  6.59it/s, loss=5.48, v_num=0]
Epoch 7:  93%|█████████▎| 90/97 [00:13<00:01,  6.63it/s, loss=5.48, v_num=0]
Epoch 7:  93%|█████████▎| 90/97 [00:13<00:01,  6.60it/s, loss=5.49, v_num=0]
Epoch 7:  94%|█████████▍| 91/97 [00:13<00:00,  6.64it/s, loss=5.49, v_num=0]
Epoch 7:  94%|█████████▍| 91/97 [00:13<00:00,  6.61it/s, loss=5.48, v_num=0]
Epoch 7:  95%|█████████▍| 92/97 [00:13<00:00,  6.64it/s, loss=5.48, v_num=0]
Epoch 7:  95%|█████████▍| 92/97 [00:13<00:00,  6.62it/s, loss=5.48, v_num=0]
Epoch 7:  96%|█████████▌| 93/97 [00:13<00:00,  6.65it/s, loss=5.48, v_num=0]
Epoch 7:  96%|█████████▌| 93/97 [00:14<00:00,  6.62it/s, loss=5.47, v_num=0]
Epoch 7:  97%|█████████▋| 94/97 [00:14<00:00,  6.65it/s, loss=5.47, v_num=0]
Epoch 7:  97%|█████████▋| 94/97 [00:14<00:00,  6.63it/s, loss=5.47, v_num=0]
Epoch 7:  98%|█████████▊| 95/97 [00:14<00:00,  6.66it/s, loss=5.47, v_num=0]
Epoch 7:  98%|█████████▊| 95/97 [00:14<00:00,  6.64it/s, loss=5.47, v_num=0]
Epoch 7:  99%|█████████▉| 96/97 [00:14<00:00,  6.67it/s, loss=5.47, v_num=0]
Epoch 7:  99%|█████████▉| 96/97 [00:14<00:00,  6.64it/s, loss=5.47, v_num=0]
Epoch 7: 100%|██████████| 97/97 [00:14<00:00,  6.67it/s, loss=5.47, v_num=0]
Epoch 7: 100%|██████████| 97/97 [00:14<00:00,  6.65it/s, loss=5.47, v_num=0]
Epoch 7: 100%|██████████| 97/97 [00:14<00:00,  6.65it/s, loss=5.47, v_num=0]
Epoch 7:   0%|          | 0/97 [00:00<?, ?it/s, loss=5.47, v_num=0]
Epoch 8:   0%|          | 0/97 [00:00<?, ?it/s, loss=5.47, v_num=0]
Epoch 8:   1%|          | 1/97 [00:01<02:13,  1.39s/it, loss=5.47, v_num=0]
Epoch 8:   1%|          | 1/97 [00:01<02:19,  1.45s/it, loss=5.45, v_num=0]
Epoch 8:   2%|▏         | 2/97 [00:01<01:12,  1.31it/s, loss=5.45, v_num=0]
Epoch 8:   2%|▏         | 2/97 [00:01<01:15,  1.26it/s, loss=5.45, v_num=0]
Epoch 8:   3%|▎         | 3/97 [00:01<00:52,  1.80it/s, loss=5.45, v_num=0]
Epoch 8:   3%|▎         | 3/97 [00:01<00:53,  1.74it/s, loss=5.44, v_num=0]
Epoch 8:   4%|▍         | 4/97 [00:01<00:41,  2.22it/s, loss=5.44, v_num=0]
Epoch 8:   4%|▍         | 4/97 [00:01<00:43,  2.15it/s, loss=5.44, v_num=0]
Epoch 8:   5%|▌         | 5/97 [00:01<00:35,  2.58it/s, loss=5.44, v_num=0]
Epoch 8:   5%|▌         | 5/97 [00:01<00:36,  2.51it/s, loss=5.43, v_num=0]
Epoch 8:   6%|▌         | 6/97 [00:02<00:31,  2.89it/s, loss=5.43, v_num=0]
Epoch 8:   6%|▌         | 6/97 [00:02<00:32,  2.82it/s, loss=5.43, v_num=0]
Epoch 8:   7%|▋         | 7/97 [00:02<00:28,  3.17it/s, loss=5.43, v_num=0]
Epoch 8:   7%|▋         | 7/97 [00:02<00:29,  3.09it/s, loss=5.43, v_num=0]
Epoch 8:   8%|▊         | 8/97 [00:02<00:26,  3.40it/s, loss=5.43, v_num=0]
Epoch 8:   8%|▊         | 8/97 [00:02<00:26,  3.33it/s, loss=5.43, v_num=0]
Epoch 8:   9%|▉         | 9/97 [00:02<00:24,  3.62it/s, loss=5.43, v_num=0]
Epoch 8:   9%|▉         | 9/97 [00:02<00:24,  3.54it/s, loss=5.43, v_num=0]
Epoch 8:  10%|█         | 10/97 [00:02<00:22,  3.81it/s, loss=5.43, v_num=0]
Epoch 8:  10%|█         | 10/97 [00:02<00:23,  3.73it/s, loss=5.43, v_num=0]
Epoch 8:  11%|█▏        | 11/97 [00:02<00:21,  3.98it/s, loss=5.43, v_num=0]
Epoch 8:  11%|█▏        | 11/97 [00:02<00:22,  3.90it/s, loss=5.44, v_num=0]
Epoch 8:  12%|█▏        | 12/97 [00:02<00:20,  4.14it/s, loss=5.44, v_num=0]
Epoch 8:  12%|█▏        | 12/97 [00:02<00:20,  4.06it/s, loss=5.44, v_num=0]
Epoch 8:  13%|█▎        | 13/97 [00:03<00:19,  4.28it/s, loss=5.44, v_num=0]
Epoch 8:  13%|█▎        | 13/97 [00:03<00:19,  4.21it/s, loss=5.44, v_num=0]
Epoch 8:  14%|█▍        | 14/97 [00:03<00:18,  4.42it/s, loss=5.44, v_num=0]
Epoch 8:  14%|█▍        | 14/97 [00:03<00:19,  4.34it/s, loss=5.44, v_num=0]
Epoch 8:  15%|█▌        | 15/97 [00:03<00:18,  4.54it/s, loss=5.44, v_num=0]
Epoch 8:  15%|█▌        | 15/97 [00:03<00:18,  4.46it/s, loss=5.44, v_num=0]
Epoch 8:  16%|█▋        | 16/97 [00:03<00:17,  4.65it/s, loss=5.44, v_num=0]
Epoch 8:  16%|█▋        | 16/97 [00:03<00:17,  4.58it/s, loss=5.44, v_num=0]
Epoch 8:  18%|█▊        | 17/97 [00:03<00:16,  4.75it/s, loss=5.44, v_num=0]
Epoch 8:  18%|█▊        | 17/97 [00:03<00:17,  4.68it/s, loss=5.44, v_num=0]
Epoch 8:  19%|█▊        | 18/97 [00:03<00:16,  4.85it/s, loss=5.44, v_num=0]
Epoch 8:  19%|█▊        | 18/97 [00:03<00:16,  4.78it/s, loss=5.44, v_num=0]
Epoch 8:  20%|█▉        | 19/97 [00:03<00:15,  4.93it/s, loss=5.44, v_num=0]
Epoch 8:  20%|█▉        | 19/97 [00:03<00:16,  4.86it/s, loss=5.43, v_num=0]
Epoch 8:  21%|██        | 20/97 [00:03<00:15,  5.02it/s, loss=5.43, v_num=0]
Epoch 8:  21%|██        | 20/97 [00:04<00:15,  4.95it/s, loss=5.43, v_num=0]
Epoch 8:  22%|██▏       | 21/97 [00:04<00:14,  5.10it/s, loss=5.43, v_num=0]
Epoch 8:  22%|██▏       | 21/97 [00:04<00:15,  5.03it/s, loss=5.45, v_num=0]
Epoch 8:  23%|██▎       | 22/97 [00:04<00:14,  5.17it/s, loss=5.45, v_num=0]
Epoch 8:  23%|██▎       | 22/97 [00:04<00:14,  5.10it/s, loss=5.44, v_num=0]
Epoch 8:  24%|██▎       | 23/97 [00:04<00:14,  5.23it/s, loss=5.44, v_num=0]
Epoch 8:  24%|██▎       | 23/97 [00:04<00:14,  5.17it/s, loss=5.44, v_num=0]
Epoch 8:  25%|██▍       | 24/97 [00:04<00:13,  5.30it/s, loss=5.44, v_num=0]
Epoch 8:  25%|██▍       | 24/97 [00:04<00:13,  5.23it/s, loss=5.45, v_num=0]
Epoch 8:  26%|██▌       | 25/97 [00:04<00:13,  5.36it/s, loss=5.45, v_num=0]
Epoch 8:  26%|██▌       | 25/97 [00:04<00:13,  5.29it/s, loss=5.44, v_num=0]
Epoch 8:  27%|██▋       | 26/97 [00:04<00:13,  5.41it/s, loss=5.44, v_num=0]
Epoch 8:  27%|██▋       | 26/97 [00:04<00:13,  5.35it/s, loss=5.43, v_num=0]
Epoch 8:  28%|██▊       | 27/97 [00:04<00:12,  5.47it/s, loss=5.43, v_num=0]
Epoch 8:  28%|██▊       | 27/97 [00:04<00:12,  5.41it/s, loss=5.44, v_num=0]
Epoch 8:  29%|██▉       | 28/97 [00:05<00:12,  5.52it/s, loss=5.44, v_num=0]
Epoch 8:  29%|██▉       | 28/97 [00:05<00:12,  5.46it/s, loss=5.43, v_num=0]
Epoch 8:  30%|██▉       | 29/97 [00:05<00:12,  5.56it/s, loss=5.43, v_num=0]
Epoch 8:  30%|██▉       | 29/97 [00:05<00:12,  5.51it/s, loss=5.43, v_num=0]
Epoch 8:  31%|███       | 30/97 [00:05<00:11,  5.61it/s, loss=5.43, v_num=0]
Epoch 8:  31%|███       | 30/97 [00:05<00:12,  5.55it/s, loss=5.42, v_num=0]
Epoch 8:  32%|███▏      | 31/97 [00:05<00:11,  5.65it/s, loss=5.42, v_num=0]
Epoch 8:  32%|███▏      | 31/97 [00:05<00:11,  5.60it/s, loss=5.41, v_num=0]
Epoch 8:  33%|███▎      | 32/97 [00:05<00:11,  5.69it/s, loss=5.41, v_num=0]
Epoch 8:  33%|███▎      | 32/97 [00:05<00:11,  5.64it/s, loss=5.41, v_num=0]
Epoch 8:  34%|███▍      | 33/97 [00:05<00:11,  5.73it/s, loss=5.41, v_num=0]
Epoch 8:  34%|███▍      | 33/97 [00:05<00:11,  5.68it/s, loss=5.4, v_num=0]
Epoch 8:  35%|███▌      | 34/97 [00:05<00:10,  5.77it/s, loss=5.4, v_num=0]
Epoch 8:  35%|███▌      | 34/97 [00:05<00:11,  5.72it/s, loss=5.4, v_num=0]
Epoch 8:  36%|███▌      | 35/97 [00:06<00:10,  5.80it/s, loss=5.4, v_num=0]
Epoch 8:  36%|███▌      | 35/97 [00:06<00:10,  5.75it/s, loss=5.4, v_num=0]
Epoch 8:  37%|███▋      | 36/97 [00:06<00:10,  5.84it/s, loss=5.4, v_num=0]
Epoch 8:  37%|███▋      | 36/97 [00:06<00:10,  5.79it/s, loss=5.39, v_num=0]
Epoch 8:  38%|███▊      | 37/97 [00:06<00:10,  5.87it/s, loss=5.39, v_num=0]
Epoch 8:  38%|███▊      | 37/97 [00:06<00:10,  5.82it/s, loss=5.39, v_num=0]
Epoch 8:  39%|███▉      | 38/97 [00:06<00:09,  5.90it/s, loss=5.39, v_num=0]
Epoch 8:  39%|███▉      | 38/97 [00:06<00:10,  5.85it/s, loss=5.38, v_num=0]
Epoch 8:  40%|████      | 39/97 [00:06<00:09,  5.94it/s, loss=5.38, v_num=0]
Epoch 8:  40%|████      | 39/97 [00:06<00:09,  5.88it/s, loss=5.38, v_num=0]
Epoch 8:  41%|████      | 40/97 [00:06<00:09,  5.96it/s, loss=5.38, v_num=0]
Epoch 8:  41%|████      | 40/97 [00:06<00:09,  5.91it/s, loss=5.38, v_num=0]
Epoch 8:  42%|████▏     | 41/97 [00:06<00:09,  5.99it/s, loss=5.38, v_num=0]
Epoch 8:  42%|████▏     | 41/97 [00:06<00:09,  5.94it/s, loss=5.37, v_num=0]
Epoch 8:  43%|████▎     | 42/97 [00:06<00:09,  6.01it/s, loss=5.37, v_num=0]
Epoch 8:  43%|████▎     | 42/97 [00:07<00:09,  5.97it/s, loss=5.36, v_num=0]
Epoch 8:  44%|████▍     | 43/97 [00:07<00:08,  6.04it/s, loss=5.36, v_num=0]
Epoch 8:  44%|████▍     | 43/97 [00:07<00:09,  5.99it/s, loss=5.36, v_num=0]
Epoch 8:  45%|████▌     | 44/97 [00:07<00:08,  6.06it/s, loss=5.36, v_num=0]
Epoch 8:  45%|████▌     | 44/97 [00:07<00:08,  6.02it/s, loss=5.36, v_num=0]
Epoch 8:  46%|████▋     | 45/97 [00:07<00:08,  6.09it/s, loss=5.36, v_num=0]
Epoch 8:  46%|████▋     | 45/97 [00:07<00:08,  6.04it/s, loss=5.36, v_num=0]
Epoch 8:  47%|████▋     | 46/97 [00:07<00:08,  6.11it/s, loss=5.36, v_num=0]
Epoch 8:  47%|████▋     | 46/97 [00:07<00:08,  6.07it/s, loss=5.37, v_num=0]
Epoch 8:  48%|████▊     | 47/97 [00:07<00:08,  6.13it/s, loss=5.37, v_num=0]
Epoch 8:  48%|████▊     | 47/97 [00:07<00:08,  6.09it/s, loss=5.36, v_num=0]
Epoch 8:  49%|████▉     | 48/97 [00:07<00:07,  6.15it/s, loss=5.36, v_num=0]
Epoch 8:  49%|████▉     | 48/97 [00:07<00:08,  6.11it/s, loss=5.35, v_num=0]
Epoch 8:  51%|█████     | 49/97 [00:07<00:07,  6.17it/s, loss=5.35, v_num=0]
Epoch 8:  51%|█████     | 49/97 [00:07<00:07,  6.13it/s, loss=5.36, v_num=0]
Epoch 8:  52%|█████▏    | 50/97 [00:08<00:07,  6.19it/s, loss=5.36, v_num=0]
Epoch 8:  52%|█████▏    | 50/97 [00:08<00:07,  6.15it/s, loss=5.36, v_num=0]
Epoch 8:  53%|█████▎    | 51/97 [00:08<00:07,  6.21it/s, loss=5.36, v_num=0]
Epoch 8:  53%|█████▎    | 51/97 [00:08<00:07,  6.16it/s, loss=5.35, v_num=0]
Epoch 8:  54%|█████▎    | 52/97 [00:08<00:07,  6.22it/s, loss=5.35, v_num=0]
Epoch 8:  54%|█████▎    | 52/97 [00:08<00:07,  6.18it/s, loss=5.34, v_num=0]
Epoch 8:  55%|█████▍    | 53/97 [00:08<00:07,  6.24it/s, loss=5.34, v_num=0]
Epoch 8:  55%|█████▍    | 53/97 [00:08<00:07,  6.20it/s, loss=5.34, v_num=0]
Epoch 8:  56%|█████▌    | 54/97 [00:08<00:06,  6.26it/s, loss=5.34, v_num=0]
Epoch 8:  56%|█████▌    | 54/97 [00:08<00:06,  6.22it/s, loss=5.34, v_num=0]
Epoch 8:  57%|█████▋    | 55/97 [00:08<00:06,  6.28it/s, loss=5.34, v_num=0]
Epoch 8:  57%|█████▋    | 55/97 [00:08<00:06,  6.23it/s, loss=5.34, v_num=0]
Epoch 8:  58%|█████▊    | 56/97 [00:08<00:06,  6.29it/s, loss=5.34, v_num=0]
Epoch 8:  58%|█████▊    | 56/97 [00:08<00:06,  6.25it/s, loss=5.33, v_num=0]
Epoch 8:  59%|█████▉    | 57/97 [00:09<00:06,  6.31it/s, loss=5.33, v_num=0]
Epoch 8:  59%|█████▉    | 57/97 [00:09<00:06,  6.27it/s, loss=5.33, v_num=0]
Epoch 8:  60%|█████▉    | 58/97 [00:09<00:06,  6.32it/s, loss=5.33, v_num=0]
Epoch 8:  60%|█████▉    | 58/97 [00:09<00:06,  6.28it/s, loss=5.33, v_num=0]
Epoch 8:  61%|██████    | 59/97 [00:09<00:05,  6.34it/s, loss=5.33, v_num=0]
Epoch 8:  61%|██████    | 59/97 [00:09<00:06,  6.30it/s, loss=5.33, v_num=0]
Epoch 8:  62%|██████▏   | 60/97 [00:09<00:05,  6.35it/s, loss=5.33, v_num=0]
Epoch 8:  62%|██████▏   | 60/97 [00:09<00:05,  6.31it/s, loss=5.33, v_num=0]
Epoch 8:  63%|██████▎   | 61/97 [00:09<00:05,  6.36it/s, loss=5.33, v_num=0]
Epoch 8:  63%|██████▎   | 61/97 [00:09<00:05,  6.33it/s, loss=5.33, v_num=0]
Epoch 8:  64%|██████▍   | 62/97 [00:09<00:05,  6.38it/s, loss=5.33, v_num=0]
Epoch 8:  64%|██████▍   | 62/97 [00:09<00:05,  6.34it/s, loss=5.32, v_num=0]
Epoch 8:  65%|██████▍   | 63/97 [00:09<00:05,  6.39it/s, loss=5.32, v_num=0]
Epoch 8:  65%|██████▍   | 63/97 [00:09<00:05,  6.35it/s, loss=5.32, v_num=0]
Epoch 8:  66%|██████▌   | 64/97 [00:09<00:05,  6.40it/s, loss=5.32, v_num=0]
Epoch 8:  66%|██████▌   | 64/97 [00:10<00:05,  6.37it/s, loss=5.31, v_num=0]
Epoch 8:  67%|██████▋   | 65/97 [00:10<00:04,  6.41it/s, loss=5.31, v_num=0]
Epoch 8:  67%|██████▋   | 65/97 [00:10<00:05,  6.38it/s, loss=5.32, v_num=0]
Epoch 8:  68%|██████▊   | 66/97 [00:10<00:04,  6.43it/s, loss=5.32, v_num=0]
Epoch 8:  68%|██████▊   | 66/97 [00:10<00:04,  6.39it/s, loss=5.3, v_num=0]
Epoch 8:  69%|██████▉   | 67/97 [00:10<00:04,  6.44it/s, loss=5.3, v_num=0]
Epoch 8:  69%|██████▉   | 67/97 [00:10<00:04,  6.40it/s, loss=5.31, v_num=0]
Epoch 8:  70%|███████   | 68/97 [00:10<00:04,  6.45it/s, loss=5.31, v_num=0]
Epoch 8:  70%|███████   | 68/97 [00:10<00:04,  6.42it/s, loss=5.31, v_num=0]
Epoch 8:  71%|███████   | 69/97 [00:10<00:04,  6.46it/s, loss=5.31, v_num=0]
Epoch 8:  71%|███████   | 69/97 [00:10<00:04,  6.43it/s, loss=5.31, v_num=0]
Epoch 8:  72%|███████▏  | 70/97 [00:10<00:04,  6.47it/s, loss=5.31, v_num=0]
Epoch 8:  72%|███████▏  | 70/97 [00:10<00:04,  6.44it/s, loss=5.3, v_num=0]
Epoch 8:  73%|███████▎  | 71/97 [00:10<00:04,  6.48it/s, loss=5.3, v_num=0]
Epoch 8:  73%|███████▎  | 71/97 [00:11<00:04,  6.45it/s, loss=5.3, v_num=0]
Epoch 8:  74%|███████▍  | 72/97 [00:11<00:03,  6.50it/s, loss=5.3, v_num=0]
Epoch 8:  74%|███████▍  | 72/97 [00:11<00:03,  6.46it/s, loss=5.28, v_num=0]
Epoch 8:  75%|███████▌  | 73/97 [00:11<00:03,  6.50it/s, loss=5.28, v_num=0]
Epoch 8:  75%|███████▌  | 73/97 [00:11<00:03,  6.47it/s, loss=5.28, v_num=0]
Epoch 8:  76%|███████▋  | 74/97 [00:11<00:03,  6.52it/s, loss=5.28, v_num=0]
Epoch 8:  76%|███████▋  | 74/97 [00:11<00:03,  6.48it/s, loss=5.28, v_num=0]
Epoch 8:  77%|███████▋  | 75/97 [00:11<00:03,  6.52it/s, loss=5.28, v_num=0]
Epoch 8:  77%|███████▋  | 75/97 [00:11<00:03,  6.49it/s, loss=5.28, v_num=0]
Epoch 8:  78%|███████▊  | 76/97 [00:11<00:03,  6.53it/s, loss=5.28, v_num=0]
Epoch 8:  78%|███████▊  | 76/97 [00:11<00:03,  6.50it/s, loss=5.27, v_num=0]
Epoch 8:  79%|███████▉  | 77/97 [00:11<00:03,  6.54it/s, loss=5.27, v_num=0]
Epoch 8:  79%|███████▉  | 77/97 [00:11<00:03,  6.51it/s, loss=5.28, v_num=0]
Epoch 8:  80%|████████  | 78/97 [00:11<00:02,  6.55it/s, loss=5.28, v_num=0]
Epoch 8:  80%|████████  | 78/97 [00:11<00:02,  6.52it/s, loss=5.28, v_num=0]
Epoch 8:  81%|████████▏ | 79/97 [00:12<00:02,  6.56it/s, loss=5.28, v_num=0]
Epoch 8:  81%|████████▏ | 79/97 [00:12<00:02,  6.53it/s, loss=5.28, v_num=0]
Epoch 8:  82%|████████▏ | 80/97 [00:12<00:02,  6.57it/s, loss=5.28, v_num=0]
Epoch 8:  82%|████████▏ | 80/97 [00:12<00:02,  6.54it/s, loss=5.27, v_num=0]
Epoch 8:  84%|████████▎ | 81/97 [00:12<00:02,  6.58it/s, loss=5.27, v_num=0]
Epoch 8:  84%|████████▎ | 81/97 [00:12<00:02,  6.55it/s, loss=5.27, v_num=0]
Epoch 8:  85%|████████▍ | 82/97 [00:12<00:02,  6.58it/s, loss=5.27, v_num=0]
Epoch 8:  85%|████████▍ | 82/97 [00:12<00:02,  6.55it/s, loss=5.27, v_num=0]
Epoch 8:  86%|████████▌ | 83/97 [00:12<00:02,  6.59it/s, loss=5.27, v_num=0]
Epoch 8:  86%|████████▌ | 83/97 [00:12<00:02,  6.56it/s, loss=5.28, v_num=0]
Epoch 8:  87%|████████▋ | 84/97 [00:12<00:01,  6.60it/s, loss=5.28, v_num=0]
Epoch 8:  87%|████████▋ | 84/97 [00:12<00:01,  6.57it/s, loss=5.26, v_num=0]
Epoch 8:  88%|████████▊ | 85/97 [00:12<00:01,  6.61it/s, loss=5.26, v_num=0]
Epoch 8:  88%|████████▊ | 85/97 [00:12<00:01,  6.58it/s, loss=5.26, v_num=0]
Epoch 8:  89%|████████▊ | 86/97 [00:12<00:01,  6.62it/s, loss=5.26, v_num=0]
Epoch 8:  89%|████████▊ | 86/97 [00:13<00:01,  6.59it/s, loss=5.26, v_num=0]
Epoch 8:  90%|████████▉ | 87/97 [00:13<00:01,  6.63it/s, loss=5.26, v_num=0]
Epoch 8:  90%|████████▉ | 87/97 [00:13<00:01,  6.60it/s, loss=5.25, v_num=0]
Epoch 8:  91%|█████████ | 88/97 [00:13<00:01,  6.63it/s, loss=5.25, v_num=0]
Epoch 8:  91%|█████████ | 88/97 [00:13<00:01,  6.60it/s, loss=5.25, v_num=0]
Epoch 8:  92%|█████████▏| 89/97 [00:13<00:01,  6.64it/s, loss=5.25, v_num=0]
Epoch 8:  92%|█████████▏| 89/97 [00:13<00:01,  6.61it/s, loss=5.25, v_num=0]
Epoch 8:  93%|█████████▎| 90/97 [00:13<00:01,  6.65it/s, loss=5.25, v_num=0]
Epoch 8:  93%|█████████▎| 90/97 [00:13<00:01,  6.62it/s, loss=5.26, v_num=0]
Epoch 8:  94%|█████████▍| 91/97 [00:13<00:00,  6.65it/s, loss=5.26, v_num=0]
Epoch 8:  94%|█████████▍| 91/97 [00:13<00:00,  6.63it/s, loss=5.26, v_num=0]
Epoch 8:  95%|█████████▍| 92/97 [00:13<00:00,  6.66it/s, loss=5.26, v_num=0]
Epoch 8:  95%|█████████▍| 92/97 [00:13<00:00,  6.63it/s, loss=5.27, v_num=0]
Epoch 8:  96%|█████████▌| 93/97 [00:13<00:00,  6.67it/s, loss=5.27, v_num=0]
Epoch 8:  96%|█████████▌| 93/97 [00:14<00:00,  6.64it/s, loss=5.27, v_num=0]
Epoch 8:  97%|█████████▋| 94/97 [00:14<00:00,  6.67it/s, loss=5.27, v_num=0]
Epoch 8:  97%|█████████▋| 94/97 [00:14<00:00,  6.65it/s, loss=5.27, v_num=0]
Epoch 8:  98%|█████████▊| 95/97 [00:14<00:00,  6.68it/s, loss=5.27, v_num=0]
Epoch 8:  98%|█████████▊| 95/97 [00:14<00:00,  6.65it/s, loss=5.27, v_num=0]
Epoch 8:  99%|█████████▉| 96/97 [00:14<00:00,  6.68it/s, loss=5.27, v_num=0]
Epoch 8:  99%|█████████▉| 96/97 [00:14<00:00,  6.66it/s, loss=5.27, v_num=0]
Epoch 8: 100%|██████████| 97/97 [00:14<00:00,  6.69it/s, loss=5.27, v_num=0]
Epoch 8: 100%|██████████| 97/97 [00:14<00:00,  6.67it/s, loss=5.26, v_num=0]
Epoch 8: 100%|██████████| 97/97 [00:14<00:00,  6.67it/s, loss=5.26, v_num=0]
Epoch 8:   0%|          | 0/97 [00:00<?, ?it/s, loss=5.26, v_num=0]
Epoch 9:   0%|          | 0/97 [00:00<?, ?it/s, loss=5.26, v_num=0]
Epoch 9:   1%|          | 1/97 [00:01<02:13,  1.39s/it, loss=5.26, v_num=0]
Epoch 9:   1%|          | 1/97 [00:01<02:18,  1.44s/it, loss=5.26, v_num=0]
Epoch 9:   2%|▏         | 2/97 [00:01<01:12,  1.31it/s, loss=5.26, v_num=0]
Epoch 9:   2%|▏         | 2/97 [00:01<01:15,  1.27it/s, loss=5.26, v_num=0]
Epoch 9:   3%|▎         | 3/97 [00:01<00:51,  1.81it/s, loss=5.26, v_num=0]
Epoch 9:   3%|▎         | 3/97 [00:01<00:53,  1.75it/s, loss=5.26, v_num=0]
Epoch 9:   4%|▍         | 4/97 [00:01<00:41,  2.23it/s, loss=5.26, v_num=0]
Epoch 9:   4%|▍         | 4/97 [00:01<00:43,  2.16it/s, loss=5.26, v_num=0]
Epoch 9:   5%|▌         | 5/97 [00:01<00:35,  2.59it/s, loss=5.26, v_num=0]
Epoch 9:   5%|▌         | 5/97 [00:01<00:36,  2.51it/s, loss=5.25, v_num=0]
Epoch 9:   6%|▌         | 6/97 [00:02<00:31,  2.89it/s, loss=5.25, v_num=0]
Epoch 9:   6%|▌         | 6/97 [00:02<00:32,  2.82it/s, loss=5.24, v_num=0]
Epoch 9:   7%|▋         | 7/97 [00:02<00:28,  3.17it/s, loss=5.24, v_num=0]
Epoch 9:   7%|▋         | 7/97 [00:02<00:29,  3.09it/s, loss=5.25, v_num=0]
Epoch 9:   8%|▊         | 8/97 [00:02<00:26,  3.42it/s, loss=5.25, v_num=0]
Epoch 9:   8%|▊         | 8/97 [00:02<00:26,  3.33it/s, loss=5.24, v_num=0]
Epoch 9:   9%|▉         | 9/97 [00:02<00:24,  3.62it/s, loss=5.24, v_num=0]
Epoch 9:   9%|▉         | 9/97 [00:02<00:24,  3.55it/s, loss=5.24, v_num=0]
Epoch 9:  10%|█         | 10/97 [00:02<00:22,  3.82it/s, loss=5.24, v_num=0]
Epoch 9:  10%|█         | 10/97 [00:02<00:23,  3.74it/s, loss=5.24, v_num=0]
Epoch 9:  11%|█▏        | 11/97 [00:02<00:21,  3.99it/s, loss=5.24, v_num=0]
Epoch 9:  11%|█▏        | 11/97 [00:02<00:22,  3.90it/s, loss=5.24, v_num=0]
Epoch 9:  12%|█▏        | 12/97 [00:02<00:20,  4.14it/s, loss=5.24, v_num=0]
Epoch 9:  12%|█▏        | 12/97 [00:02<00:20,  4.06it/s, loss=5.24, v_num=0]
Epoch 9:  13%|█▎        | 13/97 [00:03<00:19,  4.29it/s, loss=5.24, v_num=0]
Epoch 9:  13%|█▎        | 13/97 [00:03<00:19,  4.21it/s, loss=5.23, v_num=0]
Epoch 9:  14%|█▍        | 14/97 [00:03<00:18,  4.41it/s, loss=5.23, v_num=0]
Epoch 9:  14%|█▍        | 14/97 [00:03<00:19,  4.34it/s, loss=5.22, v_num=0]
Epoch 9:  15%|█▌        | 15/97 [00:03<00:18,  4.54it/s, loss=5.22, v_num=0]
Epoch 9:  15%|█▌        | 15/97 [00:03<00:18,  4.46it/s, loss=5.21, v_num=0]
Epoch 9:  16%|█▋        | 16/97 [00:03<00:17,  4.64it/s, loss=5.21, v_num=0]
Epoch 9:  16%|█▋        | 16/97 [00:03<00:17,  4.57it/s, loss=5.21, v_num=0]
Epoch 9:  18%|█▊        | 17/97 [00:03<00:16,  4.74it/s, loss=5.21, v_num=0]
Epoch 9:  18%|█▊        | 17/97 [00:03<00:17,  4.67it/s, loss=5.21, v_num=0]
Epoch 9:  19%|█▊        | 18/97 [00:03<00:16,  4.84it/s, loss=5.21, v_num=0]
Epoch 9:  19%|█▊        | 18/97 [00:03<00:16,  4.77it/s, loss=5.21, v_num=0]
Epoch 9:  20%|█▉        | 19/97 [00:03<00:15,  4.93it/s, loss=5.21, v_num=0]
Epoch 9:  20%|█▉        | 19/97 [00:03<00:16,  4.86it/s, loss=5.2, v_num=0]
Epoch 9:  21%|██        | 20/97 [00:03<00:15,  5.01it/s, loss=5.2, v_num=0]
Epoch 9:  21%|██        | 20/97 [00:04<00:15,  4.94it/s, loss=5.2, v_num=0]
Epoch 9:  22%|██▏       | 21/97 [00:04<00:14,  5.08it/s, loss=5.2, v_num=0]
Epoch 9:  22%|██▏       | 21/97 [00:04<00:15,  5.02it/s, loss=5.2, v_num=0]
Epoch 9:  23%|██▎       | 22/97 [00:04<00:14,  5.16it/s, loss=5.2, v_num=0]
Epoch 9:  23%|██▎       | 22/97 [00:04<00:14,  5.09it/s, loss=5.19, v_num=0]
Epoch 9:  24%|██▎       | 23/97 [00:04<00:14,  5.22it/s, loss=5.19, v_num=0]
Epoch 9:  24%|██▎       | 23/97 [00:04<00:14,  5.16it/s, loss=5.19, v_num=0]
Epoch 9:  25%|██▍       | 24/97 [00:04<00:13,  5.28it/s, loss=5.19, v_num=0]
Epoch 9:  25%|██▍       | 24/97 [00:04<00:13,  5.22it/s, loss=5.19, v_num=0]
Epoch 9:  26%|██▌       | 25/97 [00:04<00:13,  5.34it/s, loss=5.19, v_num=0]
Epoch 9:  26%|██▌       | 25/97 [00:04<00:13,  5.28it/s, loss=5.18, v_num=0]
Epoch 9:  27%|██▋       | 26/97 [00:04<00:13,  5.40it/s, loss=5.18, v_num=0]
Epoch 9:  27%|██▋       | 26/97 [00:04<00:13,  5.34it/s, loss=5.18, v_num=0]
Epoch 9:  28%|██▊       | 27/97 [00:04<00:12,  5.45it/s, loss=5.18, v_num=0]
Epoch 9:  28%|██▊       | 27/97 [00:05<00:12,  5.39it/s, loss=5.18, v_num=0]
Epoch 9:  29%|██▉       | 28/97 [00:05<00:12,  5.50it/s, loss=5.18, v_num=0]
Epoch 9:  29%|██▉       | 28/97 [00:05<00:12,  5.44it/s, loss=5.19, v_num=0]
Epoch 9:  30%|██▉       | 29/97 [00:05<00:12,  5.55it/s, loss=5.19, v_num=0]
Epoch 9:  30%|██▉       | 29/97 [00:05<00:12,  5.49it/s, loss=5.18, v_num=0]
Epoch 9:  31%|███       | 30/97 [00:05<00:11,  5.60it/s, loss=5.18, v_num=0]
Epoch 9:  31%|███       | 30/97 [00:05<00:12,  5.54it/s, loss=5.18, v_num=0]
Epoch 9:  32%|███▏      | 31/97 [00:05<00:11,  5.64it/s, loss=5.18, v_num=0]
Epoch 9:  32%|███▏      | 31/97 [00:05<00:11,  5.58it/s, loss=5.18, v_num=0]
Epoch 9:  33%|███▎      | 32/97 [00:05<00:11,  5.68it/s, loss=5.18, v_num=0]
Epoch 9:  33%|███▎      | 32/97 [00:05<00:11,  5.62it/s, loss=5.18, v_num=0]
Epoch 9:  34%|███▍      | 33/97 [00:05<00:11,  5.72it/s, loss=5.18, v_num=0]
Epoch 9:  34%|███▍      | 33/97 [00:05<00:11,  5.66it/s, loss=5.18, v_num=0]
Epoch 9:  35%|███▌      | 34/97 [00:05<00:10,  5.75it/s, loss=5.18, v_num=0]
Epoch 9:  35%|███▌      | 34/97 [00:05<00:11,  5.70it/s, loss=5.17, v_num=0]
Epoch 9:  36%|███▌      | 35/97 [00:06<00:10,  5.79it/s, loss=5.17, v_num=0]
Epoch 9:  36%|███▌      | 35/97 [00:06<00:10,  5.74it/s, loss=5.18, v_num=0]
Epoch 9:  37%|███▋      | 36/97 [00:06<00:10,  5.82it/s, loss=5.18, v_num=0]
Epoch 9:  37%|███▋      | 36/97 [00:06<00:10,  5.77it/s, loss=5.2, v_num=0]
Epoch 9:  38%|███▊      | 37/97 [00:06<00:10,  5.86it/s, loss=5.2, v_num=0]
Epoch 9:  38%|███▊      | 37/97 [00:06<00:10,  5.81it/s, loss=5.19, v_num=0]
Epoch 9:  39%|███▉      | 38/97 [00:06<00:10,  5.89it/s, loss=5.19, v_num=0]
Epoch 9:  39%|███▉      | 38/97 [00:06<00:10,  5.84it/s, loss=5.19, v_num=0]
Epoch 9:  40%|████      | 39/97 [00:06<00:09,  5.92it/s, loss=5.19, v_num=0]
Epoch 9:  40%|████      | 39/97 [00:06<00:09,  5.87it/s, loss=5.19, v_num=0]
Epoch 9:  41%|████      | 40/97 [00:06<00:09,  5.95it/s, loss=5.19, v_num=0]
Epoch 9:  41%|████      | 40/97 [00:06<00:09,  5.90it/s, loss=5.18, v_num=0]
Epoch 9:  42%|████▏     | 41/97 [00:06<00:09,  5.97it/s, loss=5.18, v_num=0]
Epoch 9:  42%|████▏     | 41/97 [00:06<00:09,  5.93it/s, loss=5.17, v_num=0]
Epoch 9:  43%|████▎     | 42/97 [00:06<00:09,  6.00it/s, loss=5.17, v_num=0]
Epoch 9:  43%|████▎     | 42/97 [00:07<00:09,  5.95it/s, loss=5.17, v_num=0]
Epoch 9:  44%|████▍     | 43/97 [00:07<00:08,  6.03it/s, loss=5.17, v_num=0]
Epoch 9:  44%|████▍     | 43/97 [00:07<00:09,  5.98it/s, loss=5.17, v_num=0]
Epoch 9:  45%|████▌     | 44/97 [00:07<00:08,  6.04it/s, loss=5.17, v_num=0]
Epoch 9:  45%|████▌     | 44/97 [00:07<00:08,  6.00it/s, loss=5.17, v_num=0]
Epoch 9:  46%|████▋     | 45/97 [00:07<00:08,  6.06it/s, loss=5.17, v_num=0]
Epoch 9:  46%|████▋     | 45/97 [00:07<00:08,  6.02it/s, loss=5.18, v_num=0]
Epoch 9:  47%|████▋     | 46/97 [00:07<00:08,  6.09it/s, loss=5.18, v_num=0]
Epoch 9:  47%|████▋     | 46/97 [00:07<00:08,  6.04it/s, loss=5.17, v_num=0]
Epoch 9:  48%|████▊     | 47/97 [00:07<00:08,  6.11it/s, loss=5.17, v_num=0]
Epoch 9:  48%|████▊     | 47/97 [00:07<00:08,  6.07it/s, loss=5.17, v_num=0]
Epoch 9:  49%|████▉     | 48/97 [00:07<00:07,  6.13it/s, loss=5.17, v_num=0]
Epoch 9:  49%|████▉     | 48/97 [00:07<00:08,  6.09it/s, loss=5.16, v_num=0]
Epoch 9:  51%|█████     | 49/97 [00:07<00:07,  6.15it/s, loss=5.16, v_num=0]
Epoch 9:  51%|█████     | 49/97 [00:08<00:07,  6.11it/s, loss=5.16, v_num=0]
Epoch 9:  52%|█████▏    | 50/97 [00:08<00:07,  6.17it/s, loss=5.16, v_num=0]
Epoch 9:  52%|█████▏    | 50/97 [00:08<00:07,  6.13it/s, loss=5.17, v_num=0]
Epoch 9:  53%|█████▎    | 51/97 [00:08<00:07,  6.19it/s, loss=5.17, v_num=0]
Epoch 9:  53%|█████▎    | 51/97 [00:08<00:07,  6.15it/s, loss=5.15, v_num=0]
Epoch 9:  54%|█████▎    | 52/97 [00:08<00:07,  6.21it/s, loss=5.15, v_num=0]
Epoch 9:  54%|█████▎    | 52/97 [00:08<00:07,  6.17it/s, loss=5.15, v_num=0]
Epoch 9:  55%|█████▍    | 53/97 [00:08<00:07,  6.23it/s, loss=5.15, v_num=0]
Epoch 9:  55%|█████▍    | 53/97 [00:08<00:07,  6.19it/s, loss=5.14, v_num=0]
Epoch 9:  56%|█████▌    | 54/97 [00:08<00:06,  6.24it/s, loss=5.14, v_num=0]
Epoch 9:  56%|█████▌    | 54/97 [00:08<00:06,  6.20it/s, loss=5.14, v_num=0]
Epoch 9:  57%|█████▋    | 55/97 [00:08<00:06,  6.26it/s, loss=5.14, v_num=0]
Epoch 9:  57%|█████▋    | 55/97 [00:08<00:06,  6.22it/s, loss=5.13, v_num=0]
Epoch 9:  58%|█████▊    | 56/97 [00:08<00:06,  6.28it/s, loss=5.13, v_num=0]
Epoch 9:  58%|█████▊    | 56/97 [00:08<00:06,  6.24it/s, loss=5.12, v_num=0]
Epoch 9:  59%|█████▉    | 57/97 [00:09<00:06,  6.29it/s, loss=5.12, v_num=0]
Epoch 9:  59%|█████▉    | 57/97 [00:09<00:06,  6.25it/s, loss=5.12, v_num=0]
Epoch 9:  60%|█████▉    | 58/97 [00:09<00:06,  6.31it/s, loss=5.12, v_num=0]
Epoch 9:  60%|█████▉    | 58/97 [00:09<00:06,  6.27it/s, loss=5.12, v_num=0]
Epoch 9:  61%|██████    | 59/97 [00:09<00:06,  6.32it/s, loss=5.12, v_num=0]
Epoch 9:  61%|██████    | 59/97 [00:09<00:06,  6.29it/s, loss=5.12, v_num=0]
Epoch 9:  62%|██████▏   | 60/97 [00:09<00:05,  6.34it/s, loss=5.12, v_num=0]
Epoch 9:  62%|██████▏   | 60/97 [00:09<00:05,  6.30it/s, loss=5.13, v_num=0]
Epoch 9:  63%|██████▎   | 61/97 [00:09<00:05,  6.35it/s, loss=5.13, v_num=0]
Epoch 9:  63%|██████▎   | 61/97 [00:09<00:05,  6.32it/s, loss=5.13, v_num=0]
Epoch 9:  64%|██████▍   | 62/97 [00:09<00:05,  6.36it/s, loss=5.13, v_num=0]
Epoch 9:  64%|██████▍   | 62/97 [00:09<00:05,  6.33it/s, loss=5.12, v_num=0]
Epoch 9:  65%|██████▍   | 63/97 [00:09<00:05,  6.38it/s, loss=5.12, v_num=0]
Epoch 9:  65%|██████▍   | 63/97 [00:09<00:05,  6.34it/s, loss=5.12, v_num=0]
Epoch 9:  66%|██████▌   | 64/97 [00:10<00:05,  6.39it/s, loss=5.12, v_num=0]
Epoch 9:  66%|██████▌   | 64/97 [00:10<00:05,  6.36it/s, loss=5.11, v_num=0]
Epoch 9:  67%|██████▋   | 65/97 [00:10<00:04,  6.40it/s, loss=5.11, v_num=0]
Epoch 9:  67%|██████▋   | 65/97 [00:10<00:05,  6.37it/s, loss=5.11, v_num=0]
Epoch 9:  68%|██████▊   | 66/97 [00:10<00:04,  6.42it/s, loss=5.11, v_num=0]
Epoch 9:  68%|██████▊   | 66/97 [00:10<00:04,  6.38it/s, loss=5.11, v_num=0]
Epoch 9:  69%|██████▉   | 67/97 [00:10<00:04,  6.43it/s, loss=5.11, v_num=0]
Epoch 9:  69%|██████▉   | 67/97 [00:10<00:04,  6.39it/s, loss=5.11, v_num=0]
Epoch 9:  70%|███████   | 68/97 [00:10<00:04,  6.44it/s, loss=5.11, v_num=0]
Epoch 9:  70%|███████   | 68/97 [00:10<00:04,  6.41it/s, loss=5.1, v_num=0]
Epoch 9:  71%|███████   | 69/97 [00:10<00:04,  6.45it/s, loss=5.1, v_num=0]
Epoch 9:  71%|███████   | 69/97 [00:10<00:04,  6.42it/s, loss=5.1, v_num=0]
Epoch 9:  72%|███████▏  | 70/97 [00:10<00:04,  6.46it/s, loss=5.1, v_num=0]
Epoch 9:  72%|███████▏  | 70/97 [00:10<00:04,  6.43it/s, loss=5.09, v_num=0]
Epoch 9:  73%|███████▎  | 71/97 [00:10<00:04,  6.47it/s, loss=5.09, v_num=0]
Epoch 9:  73%|███████▎  | 71/97 [00:11<00:04,  6.44it/s, loss=5.09, v_num=0]
Epoch 9:  74%|███████▍  | 72/97 [00:11<00:03,  6.48it/s, loss=5.09, v_num=0]
Epoch 9:  74%|███████▍  | 72/97 [00:11<00:03,  6.45it/s, loss=5.09, v_num=0]
Epoch 9:  75%|███████▌  | 73/97 [00:11<00:03,  6.49it/s, loss=5.09, v_num=0]
Epoch 9:  75%|███████▌  | 73/97 [00:11<00:03,  6.46it/s, loss=5.09, v_num=0]
Epoch 9:  76%|███████▋  | 74/97 [00:11<00:03,  6.50it/s, loss=5.09, v_num=0]
Epoch 9:  76%|███████▋  | 74/97 [00:11<00:03,  6.47it/s, loss=5.09, v_num=0]
Epoch 9:  77%|███████▋  | 75/97 [00:11<00:03,  6.51it/s, loss=5.09, v_num=0]
Epoch 9:  77%|███████▋  | 75/97 [00:11<00:03,  6.48it/s, loss=5.09, v_num=0]
Epoch 9:  78%|███████▊  | 76/97 [00:11<00:03,  6.52it/s, loss=5.09, v_num=0]
Epoch 9:  78%|███████▊  | 76/97 [00:11<00:03,  6.49it/s, loss=5.08, v_num=0]
Epoch 9:  79%|███████▉  | 77/97 [00:11<00:03,  6.53it/s, loss=5.08, v_num=0]
Epoch 9:  79%|███████▉  | 77/97 [00:11<00:03,  6.50it/s, loss=5.07, v_num=0]
Epoch 9:  80%|████████  | 78/97 [00:11<00:02,  6.54it/s, loss=5.07, v_num=0]
Epoch 9:  80%|████████  | 78/97 [00:11<00:02,  6.51it/s, loss=5.06, v_num=0]
Epoch 9:  81%|████████▏ | 79/97 [00:12<00:02,  6.55it/s, loss=5.06, v_num=0]
Epoch 9:  81%|████████▏ | 79/97 [00:12<00:02,  6.52it/s, loss=5.06, v_num=0]
Epoch 9:  82%|████████▏ | 80/97 [00:12<00:02,  6.56it/s, loss=5.06, v_num=0]
Epoch 9:  82%|████████▏ | 80/97 [00:12<00:02,  6.53it/s, loss=5.06, v_num=0]
Epoch 9:  84%|████████▎ | 81/97 [00:12<00:02,  6.57it/s, loss=5.06, v_num=0]
Epoch 9:  84%|████████▎ | 81/97 [00:12<00:02,  6.54it/s, loss=5.07, v_num=0]
Epoch 9:  85%|████████▍ | 82/97 [00:12<00:02,  6.57it/s, loss=5.07, v_num=0]
Epoch 9:  85%|████████▍ | 82/97 [00:12<00:02,  6.54it/s, loss=5.07, v_num=0]
Epoch 9:  86%|████████▌ | 83/97 [00:12<00:02,  6.58it/s, loss=5.07, v_num=0]
Epoch 9:  86%|████████▌ | 83/97 [00:12<00:02,  6.55it/s, loss=5.07, v_num=0]
Epoch 9:  87%|████████▋ | 84/97 [00:12<00:01,  6.59it/s, loss=5.07, v_num=0]
Epoch 9:  87%|████████▋ | 84/97 [00:12<00:01,  6.56it/s, loss=5.07, v_num=0]
Epoch 9:  88%|████████▊ | 85/97 [00:12<00:01,  6.59it/s, loss=5.07, v_num=0]
Epoch 9:  88%|████████▊ | 85/97 [00:12<00:01,  6.57it/s, loss=5.05, v_num=0]
Epoch 9:  89%|████████▊ | 86/97 [00:13<00:01,  6.60it/s, loss=5.05, v_num=0]
Epoch 9:  89%|████████▊ | 86/97 [00:13<00:01,  6.57it/s, loss=5.05, v_num=0]
Epoch 9:  90%|████████▉ | 87/97 [00:13<00:01,  6.61it/s, loss=5.05, v_num=0]
Epoch 9:  90%|████████▉ | 87/97 [00:13<00:01,  6.58it/s, loss=5.05, v_num=0]
Epoch 9:  91%|█████████ | 88/97 [00:13<00:01,  6.62it/s, loss=5.05, v_num=0]
Epoch 9:  91%|█████████ | 88/97 [00:13<00:01,  6.59it/s, loss=5.05, v_num=0]
Epoch 9:  92%|█████████▏| 89/97 [00:13<00:01,  6.63it/s, loss=5.05, v_num=0]
Epoch 9:  92%|█████████▏| 89/97 [00:13<00:01,  6.60it/s, loss=5.05, v_num=0]
Epoch 9:  93%|█████████▎| 90/97 [00:13<00:01,  6.63it/s, loss=5.05, v_num=0]
Epoch 9:  93%|█████████▎| 90/97 [00:13<00:01,  6.61it/s, loss=5.04, v_num=0]
Epoch 9:  94%|█████████▍| 91/97 [00:13<00:00,  6.64it/s, loss=5.04, v_num=0]
Epoch 9:  94%|█████████▍| 91/97 [00:13<00:00,  6.61it/s, loss=5.04, v_num=0]
Epoch 9:  95%|█████████▍| 92/97 [00:13<00:00,  6.65it/s, loss=5.04, v_num=0]
Epoch 9:  95%|█████████▍| 92/97 [00:13<00:00,  6.62it/s, loss=5.04, v_num=0]
Epoch 9:  96%|█████████▌| 93/97 [00:13<00:00,  6.65it/s, loss=5.04, v_num=0]
Epoch 9:  96%|█████████▌| 93/97 [00:14<00:00,  6.63it/s, loss=5.04, v_num=0]
Epoch 9:  97%|█████████▋| 94/97 [00:14<00:00,  6.66it/s, loss=5.04, v_num=0]
Epoch 9:  97%|█████████▋| 94/97 [00:14<00:00,  6.63it/s, loss=5.04, v_num=0]
Epoch 9:  98%|█████████▊| 95/97 [00:14<00:00,  6.67it/s, loss=5.04, v_num=0]
Epoch 9:  98%|█████████▊| 95/97 [00:14<00:00,  6.64it/s, loss=5.04, v_num=0]
Epoch 9:  99%|█████████▉| 96/97 [00:14<00:00,  6.67it/s, loss=5.04, v_num=0]
Epoch 9:  99%|█████████▉| 96/97 [00:14<00:00,  6.65it/s, loss=5.04, v_num=0]
Epoch 9: 100%|██████████| 97/97 [00:14<00:00,  6.68it/s, loss=5.04, v_num=0]
Epoch 9: 100%|██████████| 97/97 [00:14<00:00,  6.65it/s, loss=5.04, v_num=0]
Epoch 9: 100%|██████████| 97/97 [00:14<00:00,  6.65it/s, loss=5.04, v_num=0]
Epoch 9:   0%|          | 0/97 [00:00<?, ?it/s, loss=5.04, v_num=0]
Epoch 10:   0%|          | 0/97 [00:00<?, ?it/s, loss=5.04, v_num=0]
Epoch 10:   1%|          | 1/97 [00:01<02:11,  1.37s/it, loss=5.04, v_num=0]
Epoch 10:   1%|          | 1/97 [00:01<02:16,  1.43s/it, loss=5.04, v_num=0]
Epoch 10:   2%|▏         | 2/97 [00:01<01:11,  1.33it/s, loss=5.04, v_num=0]
Epoch 10:   2%|▏         | 2/97 [00:01<01:14,  1.28it/s, loss=5.03, v_num=0]
Epoch 10:   3%|▎         | 3/97 [00:01<00:51,  1.82it/s, loss=5.03, v_num=0]
Epoch 10:   3%|▎         | 3/97 [00:01<00:53,  1.76it/s, loss=5.03, v_num=0]
Epoch 10:   4%|▍         | 4/97 [00:01<00:41,  2.25it/s, loss=5.03, v_num=0]
Epoch 10:   4%|▍         | 4/97 [00:01<00:42,  2.18it/s, loss=5.03, v_num=0]
Epoch 10:   5%|▌         | 5/97 [00:01<00:35,  2.60it/s, loss=5.03, v_num=0]
Epoch 10:   5%|▌         | 5/97 [00:01<00:36,  2.53it/s, loss=5.03, v_num=0]
Epoch 10:   6%|▌         | 6/97 [00:02<00:31,  2.91it/s, loss=5.03, v_num=0]
Epoch 10:   6%|▌         | 6/97 [00:02<00:32,  2.84it/s, loss=5.04, v_num=0]
Epoch 10:   7%|▋         | 7/97 [00:02<00:28,  3.20it/s, loss=5.04, v_num=0]
Epoch 10:   7%|▋         | 7/97 [00:02<00:28,  3.11it/s, loss=5.04, v_num=0]
Epoch 10:   8%|▊         | 8/97 [00:02<00:25,  3.44it/s, loss=5.04, v_num=0]
Epoch 10:   8%|▊         | 8/97 [00:02<00:26,  3.36it/s, loss=5.03, v_num=0]
Epoch 10:   9%|▉         | 9/97 [00:02<00:24,  3.65it/s, loss=5.03, v_num=0]
Epoch 10:   9%|▉         | 9/97 [00:02<00:24,  3.57it/s, loss=5.03, v_num=0]
Epoch 10:  10%|█         | 10/97 [00:02<00:22,  3.84it/s, loss=5.03, v_num=0]
Epoch 10:  10%|█         | 10/97 [00:02<00:23,  3.76it/s, loss=5.02, v_num=0]
Epoch 10:  11%|█▏        | 11/97 [00:02<00:21,  4.01it/s, loss=5.02, v_num=0]
Epoch 10:  11%|█▏        | 11/97 [00:02<00:21,  3.93it/s, loss=5.02, v_num=0]
Epoch 10:  12%|█▏        | 12/97 [00:02<00:20,  4.17it/s, loss=5.02, v_num=0]
Epoch 10:  12%|█▏        | 12/97 [00:02<00:20,  4.09it/s, loss=5.02, v_num=0]
Epoch 10:  13%|█▎        | 13/97 [00:03<00:19,  4.32it/s, loss=5.02, v_num=0]
Epoch 10:  13%|█▎        | 13/97 [00:03<00:19,  4.24it/s, loss=5.01, v_num=0]
Epoch 10:  14%|█▍        | 14/97 [00:03<00:18,  4.45it/s, loss=5.01, v_num=0]
Epoch 10:  14%|█▍        | 14/97 [00:03<00:18,  4.37it/s, loss=5.01, v_num=0]
Epoch 10:  15%|█▌        | 15/97 [00:03<00:17,  4.57it/s, loss=5.01, v_num=0]
Epoch 10:  15%|█▌        | 15/97 [00:03<00:18,  4.49it/s, loss=5.01, v_num=0]
Epoch 10:  16%|█▋        | 16/97 [00:03<00:17,  4.68it/s, loss=5.01, v_num=0]
Epoch 10:  16%|█▋        | 16/97 [00:03<00:17,  4.60it/s, loss=5.01, v_num=0]
Epoch 10:  18%|█▊        | 17/97 [00:03<00:16,  4.77it/s, loss=5.01, v_num=0]
Epoch 10:  18%|█▊        | 17/97 [00:03<00:17,  4.70it/s, loss=5.01, v_num=0]
Epoch 10:  19%|█▊        | 18/97 [00:03<00:16,  4.87it/s, loss=5.01, v_num=0]
Epoch 10:  19%|█▊        | 18/97 [00:03<00:16,  4.80it/s, loss=5, v_num=0]
Epoch 10:  20%|█▉        | 19/97 [00:03<00:15,  4.96it/s, loss=5, v_num=0]
Epoch 10:  20%|█▉        | 19/97 [00:03<00:15,  4.89it/s, loss=5.01, v_num=0]
Epoch 10:  21%|██        | 20/97 [00:03<00:15,  5.04it/s, loss=5.01, v_num=0]
Epoch 10:  21%|██        | 20/97 [00:04<00:15,  4.97it/s, loss=5, v_num=0]
Epoch 10:  22%|██▏       | 21/97 [00:04<00:14,  5.12it/s, loss=5, v_num=0]
Epoch 10:  22%|██▏       | 21/97 [00:04<00:15,  5.05it/s, loss=4.99, v_num=0]
Epoch 10:  23%|██▎       | 22/97 [00:04<00:14,  5.19it/s, loss=4.99, v_num=0]
Epoch 10:  23%|██▎       | 22/97 [00:04<00:14,  5.12it/s, loss=4.99, v_num=0]
Epoch 10:  24%|██▎       | 23/97 [00:04<00:14,  5.26it/s, loss=4.99, v_num=0]
Epoch 10:  24%|██▎       | 23/97 [00:04<00:14,  5.19it/s, loss=4.99, v_num=0]
Epoch 10:  25%|██▍       | 24/97 [00:04<00:13,  5.32it/s, loss=4.99, v_num=0]
Epoch 10:  25%|██▍       | 24/97 [00:04<00:13,  5.25it/s, loss=4.99, v_num=0]
Epoch 10:  26%|██▌       | 25/97 [00:04<00:13,  5.37it/s, loss=4.99, v_num=0]
Epoch 10:  26%|██▌       | 25/97 [00:04<00:13,  5.31it/s, loss=4.99, v_num=0]
Epoch 10:  27%|██▋       | 26/97 [00:04<00:13,  5.43it/s, loss=4.99, v_num=0]
Epoch 10:  27%|██▋       | 26/97 [00:04<00:13,  5.37it/s, loss=4.97, v_num=0]
Epoch 10:  28%|██▊       | 27/97 [00:04<00:12,  5.49it/s, loss=4.97, v_num=0]
Epoch 10:  28%|██▊       | 27/97 [00:04<00:12,  5.42it/s, loss=4.96, v_num=0]
Epoch 10:  29%|██▉       | 28/97 [00:05<00:12,  5.54it/s, loss=4.96, v_num=0]
Epoch 10:  29%|██▉       | 28/97 [00:05<00:12,  5.48it/s, loss=4.96, v_num=0]
Epoch 10:  30%|██▉       | 29/97 [00:05<00:12,  5.58it/s, loss=4.96, v_num=0]
Epoch 10:  30%|██▉       | 29/97 [00:05<00:12,  5.52it/s, loss=4.96, v_num=0]
Epoch 10:  31%|███       | 30/97 [00:05<00:11,  5.63it/s, loss=4.96, v_num=0]
Epoch 10:  31%|███       | 30/97 [00:05<00:12,  5.57it/s, loss=4.96, v_num=0]
Epoch 10:  32%|███▏      | 31/97 [00:05<00:11,  5.67it/s, loss=4.96, v_num=0]
Epoch 10:  32%|███▏      | 31/97 [00:05<00:11,  5.61it/s, loss=4.95, v_num=0]
Epoch 10:  33%|███▎      | 32/97 [00:05<00:11,  5.71it/s, loss=4.95, v_num=0]
Epoch 10:  33%|███▎      | 32/97 [00:05<00:11,  5.65it/s, loss=4.94, v_num=0]
Epoch 10:  34%|███▍      | 33/97 [00:05<00:11,  5.75it/s, loss=4.94, v_num=0]
Epoch 10:  34%|███▍      | 33/97 [00:05<00:11,  5.69it/s, loss=4.95, v_num=0]
Epoch 10:  35%|███▌      | 34/97 [00:05<00:10,  5.78it/s, loss=4.95, v_num=0]
Epoch 10:  35%|███▌      | 34/97 [00:05<00:10,  5.73it/s, loss=4.94, v_num=0]
Epoch 10:  36%|███▌      | 35/97 [00:06<00:10,  5.82it/s, loss=4.94, v_num=0]
Epoch 10:  36%|███▌      | 35/97 [00:06<00:10,  5.77it/s, loss=4.94, v_num=0]
Epoch 10:  37%|███▋      | 36/97 [00:06<00:10,  5.85it/s, loss=4.94, v_num=0]
Epoch 10:  37%|███▋      | 36/97 [00:06<00:10,  5.80it/s, loss=4.94, v_num=0]
Epoch 10:  38%|███▊      | 37/97 [00:06<00:10,  5.89it/s, loss=4.94, v_num=0]
Epoch 10:  38%|███▊      | 37/97 [00:06<00:10,  5.83it/s, loss=4.94, v_num=0]
Epoch 10:  39%|███▉      | 38/97 [00:06<00:09,  5.91it/s, loss=4.94, v_num=0]
Epoch 10:  39%|███▉      | 38/97 [00:06<00:10,  5.86it/s, loss=4.94, v_num=0]
Epoch 10:  40%|████      | 39/97 [00:06<00:09,  5.95it/s, loss=4.94, v_num=0]
Epoch 10:  40%|████      | 39/97 [00:06<00:09,  5.89it/s, loss=4.93, v_num=0]
Epoch 10:  41%|████      | 40/97 [00:06<00:09,  5.97it/s, loss=4.93, v_num=0]
Epoch 10:  41%|████      | 40/97 [00:06<00:09,  5.92it/s, loss=4.94, v_num=0]
Epoch 10:  42%|████▏     | 41/97 [00:06<00:09,  6.00it/s, loss=4.94, v_num=0]
Epoch 10:  42%|████▏     | 41/97 [00:06<00:09,  5.95it/s, loss=4.94, v_num=0]
Epoch 10:  43%|████▎     | 42/97 [00:06<00:09,  6.02it/s, loss=4.94, v_num=0]
Epoch 10:  43%|████▎     | 42/97 [00:07<00:09,  5.98it/s, loss=4.94, v_num=0]
Epoch 10:  44%|████▍     | 43/97 [00:07<00:08,  6.05it/s, loss=4.94, v_num=0]
Epoch 10:  44%|████▍     | 43/97 [00:07<00:08,  6.00it/s, loss=4.93, v_num=0]
Epoch 10:  45%|████▌     | 44/97 [00:07<00:08,  6.07it/s, loss=4.93, v_num=0]
Epoch 10:  45%|████▌     | 44/97 [00:07<00:08,  6.03it/s, loss=4.93, v_num=0]
Epoch 10:  46%|████▋     | 45/97 [00:07<00:08,  6.09it/s, loss=4.93, v_num=0]
Epoch 10:  46%|████▋     | 45/97 [00:07<00:08,  6.05it/s, loss=4.92, v_num=0]
Epoch 10:  47%|████▋     | 46/97 [00:07<00:08,  6.12it/s, loss=4.92, v_num=0]
Epoch 10:  47%|████▋     | 46/97 [00:07<00:08,  6.07it/s, loss=4.92, v_num=0]
Epoch 10:  48%|████▊     | 47/97 [00:07<00:08,  6.13it/s, loss=4.92, v_num=0]
Epoch 10:  48%|████▊     | 47/97 [00:07<00:08,  6.09it/s, loss=4.94, v_num=0]
Epoch 10:  49%|████▉     | 48/97 [00:07<00:07,  6.15it/s, loss=4.94, v_num=0]
Epoch 10:  49%|████▉     | 48/97 [00:07<00:08,  6.11it/s, loss=4.94, v_num=0]
Epoch 10:  51%|█████     | 49/97 [00:07<00:07,  6.17it/s, loss=4.94, v_num=0]
Epoch 10:  51%|█████     | 49/97 [00:07<00:07,  6.13it/s, loss=4.94, v_num=0]
Epoch 10:  52%|█████▏    | 50/97 [00:08<00:07,  6.19it/s, loss=4.94, v_num=0]
Epoch 10:  52%|█████▏    | 50/97 [00:08<00:07,  6.15it/s, loss=4.95, v_num=0]
Epoch 10:  53%|█████▎    | 51/97 [00:08<00:07,  6.21it/s, loss=4.95, v_num=0]
Epoch 10:  53%|█████▎    | 51/97 [00:08<00:07,  6.16it/s, loss=4.94, v_num=0]
Epoch 10:  54%|█████▎    | 52/97 [00:08<00:07,  6.22it/s, loss=4.94, v_num=0]
Epoch 10:  54%|█████▎    | 52/97 [00:08<00:07,  6.18it/s, loss=4.94, v_num=0]
Epoch 10:  55%|█████▍    | 53/97 [00:08<00:07,  6.24it/s, loss=4.94, v_num=0]
Epoch 10:  55%|█████▍    | 53/97 [00:08<00:07,  6.20it/s, loss=4.94, v_num=0]
Epoch 10:  56%|█████▌    | 54/97 [00:08<00:06,  6.25it/s, loss=4.94, v_num=0]
Epoch 10:  56%|█████▌    | 54/97 [00:08<00:06,  6.22it/s, loss=4.94, v_num=0]
Epoch 10:  57%|█████▋    | 55/97 [00:08<00:06,  6.27it/s, loss=4.94, v_num=0]
Epoch 10:  57%|█████▋    | 55/97 [00:08<00:06,  6.23it/s, loss=4.93, v_num=0]
Epoch 10:  58%|█████▊    | 56/97 [00:08<00:06,  6.29it/s, loss=4.93, v_num=0]
Epoch 10:  58%|█████▊    | 56/97 [00:08<00:06,  6.25it/s, loss=4.93, v_num=0]
Epoch 10:  59%|█████▉    | 57/97 [00:09<00:06,  6.31it/s, loss=4.93, v_num=0]
Epoch 10:  59%|█████▉    | 57/97 [00:09<00:06,  6.27it/s, loss=4.93, v_num=0]
Epoch 10:  60%|█████▉    | 58/97 [00:09<00:06,  6.32it/s, loss=4.93, v_num=0]
Epoch 10:  60%|█████▉    | 58/97 [00:09<00:06,  6.28it/s, loss=4.93, v_num=0]
Epoch 10:  61%|██████    | 59/97 [00:09<00:05,  6.34it/s, loss=4.93, v_num=0]
Epoch 10:  61%|██████    | 59/97 [00:09<00:06,  6.30it/s, loss=4.93, v_num=0]
Epoch 10:  62%|██████▏   | 60/97 [00:09<00:05,  6.35it/s, loss=4.93, v_num=0]
Epoch 10:  62%|██████▏   | 60/97 [00:09<00:05,  6.31it/s, loss=4.93, v_num=0]
Epoch 10:  63%|██████▎   | 61/97 [00:09<00:05,  6.36it/s, loss=4.93, v_num=0]
Epoch 10:  63%|██████▎   | 61/97 [00:09<00:05,  6.32it/s, loss=4.92, v_num=0]
Epoch 10:  64%|██████▍   | 62/97 [00:09<00:05,  6.37it/s, loss=4.92, v_num=0]
Epoch 10:  64%|██████▍   | 62/97 [00:09<00:05,  6.33it/s, loss=4.91, v_num=0]
Epoch 10:  65%|██████▍   | 63/97 [00:09<00:05,  6.38it/s, loss=4.91, v_num=0]
Epoch 10:  65%|██████▍   | 63/97 [00:09<00:05,  6.35it/s, loss=4.91, v_num=0]
Epoch 10:  66%|██████▌   | 64/97 [00:10<00:05,  6.40it/s, loss=4.91, v_num=0]
Epoch 10:  66%|██████▌   | 64/97 [00:10<00:05,  6.36it/s, loss=4.91, v_num=0]
Epoch 10:  67%|██████▋   | 65/97 [00:10<00:04,  6.41it/s, loss=4.91, v_num=0]
Epoch 10:  67%|██████▋   | 65/97 [00:10<00:05,  6.37it/s, loss=4.9, v_num=0]
Epoch 10:  68%|██████▊   | 66/97 [00:10<00:04,  6.42it/s, loss=4.9, v_num=0]
Epoch 10:  68%|██████▊   | 66/97 [00:10<00:04,  6.39it/s, loss=4.9, v_num=0]
Epoch 10:  69%|██████▉   | 67/97 [00:10<00:04,  6.43it/s, loss=4.9, v_num=0]
Epoch 10:  69%|██████▉   | 67/97 [00:10<00:04,  6.40it/s, loss=4.9, v_num=0]
Epoch 10:  70%|███████   | 68/97 [00:10<00:04,  6.45it/s, loss=4.9, v_num=0]
Epoch 10:  70%|███████   | 68/97 [00:10<00:04,  6.41it/s, loss=4.89, v_num=0]
Epoch 10:  71%|███████   | 69/97 [00:10<00:04,  6.45it/s, loss=4.89, v_num=0]
Epoch 10:  71%|███████   | 69/97 [00:10<00:04,  6.42it/s, loss=4.88, v_num=0]
Epoch 10:  72%|███████▏  | 70/97 [00:10<00:04,  6.47it/s, loss=4.88, v_num=0]
Epoch 10:  72%|███████▏  | 70/97 [00:10<00:04,  6.43it/s, loss=4.87, v_num=0]
Epoch 10:  73%|███████▎  | 71/97 [00:10<00:04,  6.48it/s, loss=4.87, v_num=0]
Epoch 10:  73%|███████▎  | 71/97 [00:11<00:04,  6.44it/s, loss=4.87, v_num=0]
Epoch 10:  74%|███████▍  | 72/97 [00:11<00:03,  6.49it/s, loss=4.87, v_num=0]
Epoch 10:  74%|███████▍  | 72/97 [00:11<00:03,  6.45it/s, loss=4.87, v_num=0]
Epoch 10:  75%|███████▌  | 73/97 [00:11<00:03,  6.50it/s, loss=4.87, v_num=0]
Epoch 10:  75%|███████▌  | 73/97 [00:11<00:03,  6.46it/s, loss=4.88, v_num=0]
Epoch 10:  76%|███████▋  | 74/97 [00:11<00:03,  6.51it/s, loss=4.88, v_num=0]
Epoch 10:  76%|███████▋  | 74/97 [00:11<00:03,  6.47it/s, loss=4.87, v_num=0]
Epoch 10:  77%|███████▋  | 75/97 [00:11<00:03,  6.51it/s, loss=4.87, v_num=0]
Epoch 10:  77%|███████▋  | 75/97 [00:11<00:03,  6.48it/s, loss=4.88, v_num=0]
Epoch 10:  78%|███████▊  | 76/97 [00:11<00:03,  6.52it/s, loss=4.88, v_num=0]
Epoch 10:  78%|███████▊  | 76/97 [00:11<00:03,  6.49it/s, loss=4.87, v_num=0]
Epoch 10:  79%|███████▉  | 77/97 [00:11<00:03,  6.53it/s, loss=4.87, v_num=0]
Epoch 10:  79%|███████▉  | 77/97 [00:11<00:03,  6.50it/s, loss=4.87, v_num=0]
Epoch 10:  80%|████████  | 78/97 [00:11<00:02,  6.54it/s, loss=4.87, v_num=0]
Epoch 10:  80%|████████  | 78/97 [00:11<00:02,  6.51it/s, loss=4.87, v_num=0]
Epoch 10:  81%|████████▏ | 79/97 [00:12<00:02,  6.55it/s, loss=4.87, v_num=0]
Epoch 10:  81%|████████▏ | 79/97 [00:12<00:02,  6.52it/s, loss=4.86, v_num=0]
Epoch 10:  82%|████████▏ | 80/97 [00:12<00:02,  6.56it/s, loss=4.86, v_num=0]
Epoch 10:  82%|████████▏ | 80/97 [00:12<00:02,  6.53it/s, loss=4.86, v_num=0]
Epoch 10:  84%|████████▎ | 81/97 [00:12<00:02,  6.57it/s, loss=4.86, v_num=0]
Epoch 10:  84%|████████▎ | 81/97 [00:12<00:02,  6.54it/s, loss=4.87, v_num=0]
Epoch 10:  85%|████████▍ | 82/97 [00:12<00:02,  6.57it/s, loss=4.87, v_num=0]
Epoch 10:  85%|████████▍ | 82/97 [00:12<00:02,  6.55it/s, loss=4.87, v_num=0]
Epoch 10:  86%|████████▌ | 83/97 [00:12<00:02,  6.58it/s, loss=4.87, v_num=0]
Epoch 10:  86%|████████▌ | 83/97 [00:12<00:02,  6.55it/s, loss=4.89, v_num=0]
Epoch 10:  87%|████████▋ | 84/97 [00:12<00:01,  6.59it/s, loss=4.89, v_num=0]
Epoch 10:  87%|████████▋ | 84/97 [00:12<00:01,  6.56it/s, loss=4.88, v_num=0]
Epoch 10:  88%|████████▊ | 85/97 [00:12<00:01,  6.60it/s, loss=4.88, v_num=0]
Epoch 10:  88%|████████▊ | 85/97 [00:12<00:01,  6.57it/s, loss=4.88, v_num=0]
Epoch 10:  89%|████████▊ | 86/97 [00:13<00:01,  6.61it/s, loss=4.88, v_num=0]
Epoch 10:  89%|████████▊ | 86/97 [00:13<00:01,  6.58it/s, loss=4.88, v_num=0]
Epoch 10:  90%|████████▉ | 87/97 [00:13<00:01,  6.61it/s, loss=4.88, v_num=0]
Epoch 10:  90%|████████▉ | 87/97 [00:13<00:01,  6.59it/s, loss=4.88, v_num=0]
Epoch 10:  91%|█████████ | 88/97 [00:13<00:01,  6.62it/s, loss=4.88, v_num=0]
Epoch 10:  91%|█████████ | 88/97 [00:13<00:01,  6.59it/s, loss=4.88, v_num=0]
Epoch 10:  92%|█████████▏| 89/97 [00:13<00:01,  6.63it/s, loss=4.88, v_num=0]
Epoch 10:  92%|█████████▏| 89/97 [00:13<00:01,  6.60it/s, loss=4.88, v_num=0]
Epoch 10:  93%|█████████▎| 90/97 [00:13<00:01,  6.64it/s, loss=4.88, v_num=0]
Epoch 10:  93%|█████████▎| 90/97 [00:13<00:01,  6.61it/s, loss=4.87, v_num=0]
Epoch 10:  94%|█████████▍| 91/97 [00:13<00:00,  6.64it/s, loss=4.87, v_num=0]
Epoch 10:  94%|█████████▍| 91/97 [00:13<00:00,  6.62it/s, loss=4.88, v_num=0]
Epoch 10:  95%|█████████▍| 92/97 [00:13<00:00,  6.65it/s, loss=4.88, v_num=0]
Epoch 10:  95%|█████████▍| 92/97 [00:13<00:00,  6.62it/s, loss=4.89, v_num=0]
Epoch 10:  96%|█████████▌| 93/97 [00:13<00:00,  6.66it/s, loss=4.89, v_num=0]
Epoch 10:  96%|█████████▌| 93/97 [00:14<00:00,  6.63it/s, loss=4.88, v_num=0]
Epoch 10:  97%|█████████▋| 94/97 [00:14<00:00,  6.66it/s, loss=4.88, v_num=0]
Epoch 10:  97%|█████████▋| 94/97 [00:14<00:00,  6.64it/s, loss=4.88, v_num=0]
Epoch 10:  98%|█████████▊| 95/97 [00:14<00:00,  6.67it/s, loss=4.88, v_num=0]
Epoch 10:  98%|█████████▊| 95/97 [00:14<00:00,  6.64it/s, loss=4.87, v_num=0]
Epoch 10:  99%|█████████▉| 96/97 [00:14<00:00,  6.68it/s, loss=4.87, v_num=0]
Epoch 10:  99%|█████████▉| 96/97 [00:14<00:00,  6.65it/s, loss=4.88, v_num=0]
Epoch 10: 100%|██████████| 97/97 [00:14<00:00,  6.68it/s, loss=4.88, v_num=0]
Epoch 10: 100%|██████████| 97/97 [00:14<00:00,  6.66it/s, loss=4.88, v_num=0]
Epoch 10: 100%|██████████| 97/97 [00:14<00:00,  6.66it/s, loss=4.88, v_num=0]
Epoch 10:   0%|          | 0/97 [00:00<?, ?it/s, loss=4.88, v_num=0]
Epoch 11:   0%|          | 0/97 [00:00<?, ?it/s, loss=4.88, v_num=0]
Epoch 11:   1%|          | 1/97 [00:01<02:14,  1.40s/it, loss=4.88, v_num=0]
Epoch 11:   1%|          | 1/97 [00:01<02:19,  1.46s/it, loss=4.88, v_num=0]
Epoch 11:   2%|▏         | 2/97 [00:01<01:12,  1.30it/s, loss=4.88, v_num=0]
Epoch 11:   2%|▏         | 2/97 [00:01<01:15,  1.26it/s, loss=4.88, v_num=0]
Epoch 11:   3%|▎         | 3/97 [00:01<00:52,  1.79it/s, loss=4.88, v_num=0]
Epoch 11:   3%|▎         | 3/97 [00:01<00:54,  1.73it/s, loss=4.88, v_num=0]
Epoch 11:   4%|▍         | 4/97 [00:01<00:42,  2.21it/s, loss=4.88, v_num=0]
Epoch 11:   4%|▍         | 4/97 [00:01<00:43,  2.14it/s, loss=4.87, v_num=0]
Epoch 11:   5%|▌         | 5/97 [00:01<00:35,  2.56it/s, loss=4.87, v_num=0]
Epoch 11:   5%|▌         | 5/97 [00:02<00:36,  2.49it/s, loss=4.86, v_num=0]
Epoch 11:   6%|▌         | 6/97 [00:02<00:31,  2.87it/s, loss=4.86, v_num=0]
Epoch 11:   6%|▌         | 6/97 [00:02<00:32,  2.80it/s, loss=4.85, v_num=0]
Epoch 11:   7%|▋         | 7/97 [00:02<00:28,  3.15it/s, loss=4.85, v_num=0]
Epoch 11:   7%|▋         | 7/97 [00:02<00:29,  3.07it/s, loss=4.86, v_num=0]
Epoch 11:   8%|▊         | 8/97 [00:02<00:26,  3.40it/s, loss=4.86, v_num=0]
Epoch 11:   8%|▊         | 8/97 [00:02<00:26,  3.31it/s, loss=4.86, v_num=0]
Epoch 11:   9%|▉         | 9/97 [00:02<00:24,  3.60it/s, loss=4.86, v_num=0]
Epoch 11:   9%|▉         | 9/97 [00:02<00:25,  3.52it/s, loss=4.86, v_num=0]
Epoch 11:  10%|█         | 10/97 [00:02<00:22,  3.79it/s, loss=4.86, v_num=0]
Epoch 11:  10%|█         | 10/97 [00:02<00:23,  3.71it/s, loss=4.85, v_num=0]
Epoch 11:  11%|█▏        | 11/97 [00:02<00:21,  3.97it/s, loss=4.85, v_num=0]
Epoch 11:  11%|█▏        | 11/97 [00:02<00:22,  3.89it/s, loss=4.84, v_num=0]
Epoch 11:  12%|█▏        | 12/97 [00:02<00:20,  4.12it/s, loss=4.84, v_num=0]
Epoch 11:  12%|█▏        | 12/97 [00:02<00:21,  4.05it/s, loss=4.85, v_num=0]
Epoch 11:  13%|█▎        | 13/97 [00:03<00:19,  4.27it/s, loss=4.85, v_num=0]
Epoch 11:  13%|█▎        | 13/97 [00:03<00:20,  4.19it/s, loss=4.85, v_num=0]
Epoch 11:  14%|█▍        | 14/97 [00:03<00:18,  4.40it/s, loss=4.85, v_num=0]
Epoch 11:  14%|█▍        | 14/97 [00:03<00:19,  4.32it/s, loss=4.84, v_num=0]
Epoch 11:  15%|█▌        | 15/97 [00:03<00:18,  4.53it/s, loss=4.84, v_num=0]
Epoch 11:  15%|█▌        | 15/97 [00:03<00:18,  4.45it/s, loss=4.83, v_num=0]
Epoch 11:  16%|█▋        | 16/97 [00:03<00:17,  4.64it/s, loss=4.83, v_num=0]
Epoch 11:  16%|█▋        | 16/97 [00:03<00:17,  4.56it/s, loss=4.83, v_num=0]
Epoch 11:  18%|█▊        | 17/97 [00:03<00:16,  4.73it/s, loss=4.83, v_num=0]
Epoch 11:  18%|█▊        | 17/97 [00:03<00:17,  4.66it/s, loss=4.82, v_num=0]
Epoch 11:  19%|█▊        | 18/97 [00:03<00:16,  4.83it/s, loss=4.82, v_num=0]
Epoch 11:  19%|█▊        | 18/97 [00:03<00:16,  4.76it/s, loss=4.82, v_num=0]
Epoch 11:  20%|█▉        | 19/97 [00:03<00:15,  4.92it/s, loss=4.82, v_num=0]
Epoch 11:  20%|█▉        | 19/97 [00:03<00:16,  4.85it/s, loss=4.81, v_num=0]
Epoch 11:  21%|██        | 20/97 [00:03<00:15,  5.00it/s, loss=4.81, v_num=0]
Epoch 11:  21%|██        | 20/97 [00:04<00:15,  4.93it/s, loss=4.81, v_num=0]
Epoch 11:  22%|██▏       | 21/97 [00:04<00:14,  5.08it/s, loss=4.81, v_num=0]
Epoch 11:  22%|██▏       | 21/97 [00:04<00:15,  5.01it/s, loss=4.8, v_num=0]
Epoch 11:  23%|██▎       | 22/97 [00:04<00:14,  5.15it/s, loss=4.8, v_num=0]
Epoch 11:  23%|██▎       | 22/97 [00:04<00:14,  5.08it/s, loss=4.79, v_num=0]
Epoch 11:  24%|██▎       | 23/97 [00:04<00:14,  5.20it/s, loss=4.79, v_num=0]
Epoch 11:  24%|██▎       | 23/97 [00:04<00:14,  5.14it/s, loss=4.79, v_num=0]
Epoch 11:  25%|██▍       | 24/97 [00:04<00:13,  5.27it/s, loss=4.79, v_num=0]
Epoch 11:  25%|██▍       | 24/97 [00:04<00:14,  5.21it/s, loss=4.78, v_num=0]
Epoch 11:  26%|██▌       | 25/97 [00:04<00:13,  5.33it/s, loss=4.78, v_num=0]
Epoch 11:  26%|██▌       | 25/97 [00:04<00:13,  5.27it/s, loss=4.77, v_num=0]
Epoch 11:  27%|██▋       | 26/97 [00:04<00:13,  5.39it/s, loss=4.77, v_num=0]
Epoch 11:  27%|██▋       | 26/97 [00:04<00:13,  5.33it/s, loss=4.78, v_num=0]
Epoch 11:  28%|██▊       | 27/97 [00:04<00:12,  5.44it/s, loss=4.78, v_num=0]
Epoch 11:  28%|██▊       | 27/97 [00:05<00:13,  5.38it/s, loss=4.77, v_num=0]
Epoch 11:  29%|██▉       | 28/97 [00:05<00:12,  5.50it/s, loss=4.77, v_num=0]
Epoch 11:  29%|██▉       | 28/97 [00:05<00:12,  5.43it/s, loss=4.76, v_num=0]
Epoch 11:  30%|██▉       | 29/97 [00:05<00:12,  5.54it/s, loss=4.76, v_num=0]
Epoch 11:  30%|██▉       | 29/97 [00:05<00:12,  5.48it/s, loss=4.75, v_num=0]
Epoch 11:  31%|███       | 30/97 [00:05<00:11,  5.59it/s, loss=4.75, v_num=0]
Epoch 11:  31%|███       | 30/97 [00:05<00:12,  5.53it/s, loss=4.75, v_num=0]
Epoch 11:  32%|███▏      | 31/97 [00:05<00:11,  5.63it/s, loss=4.75, v_num=0]
Epoch 11:  32%|███▏      | 31/97 [00:05<00:11,  5.57it/s, loss=4.75, v_num=0]
Epoch 11:  33%|███▎      | 32/97 [00:05<00:11,  5.67it/s, loss=4.75, v_num=0]
Epoch 11:  33%|███▎      | 32/97 [00:05<00:11,  5.61it/s, loss=4.75, v_num=0]
Epoch 11:  34%|███▍      | 33/97 [00:05<00:11,  5.71it/s, loss=4.75, v_num=0]
Epoch 11:  34%|███▍      | 33/97 [00:05<00:11,  5.66it/s, loss=4.74, v_num=0]
Epoch 11:  35%|███▌      | 34/97 [00:05<00:10,  5.75it/s, loss=4.74, v_num=0]
Epoch 11:  35%|███▌      | 34/97 [00:05<00:11,  5.69it/s, loss=4.75, v_num=0]
Epoch 11:  36%|███▌      | 35/97 [00:06<00:10,  5.78it/s, loss=4.75, v_num=0]
Epoch 11:  36%|███▌      | 35/97 [00:06<00:10,  5.73it/s, loss=4.74, v_num=0]
Epoch 11:  37%|███▋      | 36/97 [00:06<00:10,  5.82it/s, loss=4.74, v_num=0]
Epoch 11:  37%|███▋      | 36/97 [00:06<00:10,  5.77it/s, loss=4.74, v_num=0]
Epoch 11:  38%|███▊      | 37/97 [00:06<00:10,  5.85it/s, loss=4.74, v_num=0]
Epoch 11:  38%|███▊      | 37/97 [00:06<00:10,  5.80it/s, loss=4.75, v_num=0]
Epoch 11:  39%|███▉      | 38/97 [00:06<00:10,  5.88it/s, loss=4.75, v_num=0]
Epoch 11:  39%|███▉      | 38/97 [00:06<00:10,  5.83it/s, loss=4.75, v_num=0]
Epoch 11:  40%|████      | 39/97 [00:06<00:09,  5.91it/s, loss=4.75, v_num=0]
Epoch 11:  40%|████      | 39/97 [00:06<00:09,  5.86it/s, loss=4.75, v_num=0]
Epoch 11:  41%|████      | 40/97 [00:06<00:09,  5.94it/s, loss=4.75, v_num=0]
Epoch 11:  41%|████      | 40/97 [00:06<00:09,  5.89it/s, loss=4.75, v_num=0]
Epoch 11:  42%|████▏     | 41/97 [00:06<00:09,  5.97it/s, loss=4.75, v_num=0]
Epoch 11:  42%|████▏     | 41/97 [00:06<00:09,  5.92it/s, loss=4.74, v_num=0]
Epoch 11:  43%|████▎     | 42/97 [00:07<00:09,  5.99it/s, loss=4.74, v_num=0]
Epoch 11:  43%|████▎     | 42/97 [00:07<00:09,  5.95it/s, loss=4.75, v_num=0]
Epoch 11:  44%|████▍     | 43/97 [00:07<00:08,  6.02it/s, loss=4.75, v_num=0]
Epoch 11:  44%|████▍     | 43/97 [00:07<00:09,  5.97it/s, loss=4.75, v_num=0]
Epoch 11:  45%|████▌     | 44/97 [00:07<00:08,  6.05it/s, loss=4.75, v_num=0]
Epoch 11:  45%|████▌     | 44/97 [00:07<00:08,  6.00it/s, loss=4.75, v_num=0]
Epoch 11:  46%|████▋     | 45/97 [00:07<00:08,  6.07it/s, loss=4.75, v_num=0]
Epoch 11:  46%|████▋     | 45/97 [00:07<00:08,  6.02it/s, loss=4.75, v_num=0]
Epoch 11:  47%|████▋     | 46/97 [00:07<00:08,  6.09it/s, loss=4.75, v_num=0]
Epoch 11:  47%|████▋     | 46/97 [00:07<00:08,  6.05it/s, loss=4.75, v_num=0]
Epoch 11:  48%|████▊     | 47/97 [00:07<00:08,  6.11it/s, loss=4.75, v_num=0]
Epoch 11:  48%|████▊     | 47/97 [00:07<00:08,  6.07it/s, loss=4.75, v_num=0]
Epoch 11:  49%|████▉     | 48/97 [00:07<00:07,  6.13it/s, loss=4.75, v_num=0]
Epoch 11:  49%|████▉     | 48/97 [00:07<00:08,  6.09it/s, loss=4.76, v_num=0]
Epoch 11:  51%|█████     | 49/97 [00:07<00:07,  6.16it/s, loss=4.76, v_num=0]
Epoch 11:  51%|█████     | 49/97 [00:08<00:07,  6.11it/s, loss=4.75, v_num=0]
Epoch 11:  52%|█████▏    | 50/97 [00:08<00:07,  6.17it/s, loss=4.75, v_num=0]
Epoch 11:  52%|█████▏    | 50/97 [00:08<00:07,  6.13it/s, loss=4.74, v_num=0]
Epoch 11:  53%|█████▎    | 51/97 [00:08<00:07,  6.19it/s, loss=4.74, v_num=0]
Epoch 11:  53%|█████▎    | 51/97 [00:08<00:07,  6.15it/s, loss=4.75, v_num=0]
Epoch 11:  54%|█████▎    | 52/97 [00:08<00:07,  6.21it/s, loss=4.75, v_num=0]
Epoch 11:  54%|█████▎    | 52/97 [00:08<00:07,  6.17it/s, loss=4.74, v_num=0]
Epoch 11:  55%|█████▍    | 53/97 [00:08<00:07,  6.23it/s, loss=4.74, v_num=0]
Epoch 11:  55%|█████▍    | 53/97 [00:08<00:07,  6.19it/s, loss=4.76, v_num=0]
Epoch 11:  56%|█████▌    | 54/97 [00:08<00:06,  6.24it/s, loss=4.76, v_num=0]
Epoch 11:  56%|█████▌    | 54/97 [00:08<00:06,  6.20it/s, loss=4.75, v_num=0]
Epoch 11:  57%|█████▋    | 55/97 [00:08<00:06,  6.26it/s, loss=4.75, v_num=0]
Epoch 11:  57%|█████▋    | 55/97 [00:08<00:06,  6.22it/s, loss=4.75, v_num=0]
Epoch 11:  58%|█████▊    | 56/97 [00:08<00:06,  6.28it/s, loss=4.75, v_num=0]
Epoch 11:  58%|█████▊    | 56/97 [00:08<00:06,  6.24it/s, loss=4.75, v_num=0]
Epoch 11:  59%|█████▉    | 57/97 [00:09<00:06,  6.29it/s, loss=4.75, v_num=0]
Epoch 11:  59%|█████▉    | 57/97 [00:09<00:06,  6.25it/s, loss=4.74, v_num=0]
Epoch 11:  60%|█████▉    | 58/97 [00:09<00:06,  6.30it/s, loss=4.74, v_num=0]
Epoch 11:  60%|█████▉    | 58/97 [00:09<00:06,  6.27it/s, loss=4.73, v_num=0]
Epoch 11:  61%|██████    | 59/97 [00:09<00:06,  6.32it/s, loss=4.73, v_num=0]
Epoch 11:  61%|██████    | 59/97 [00:09<00:06,  6.28it/s, loss=4.72, v_num=0]
Epoch 11:  62%|██████▏   | 60/97 [00:09<00:05,  6.34it/s, loss=4.72, v_num=0]
Epoch 11:  62%|██████▏   | 60/97 [00:09<00:05,  6.30it/s, loss=4.72, v_num=0]
Epoch 11:  63%|██████▎   | 61/97 [00:09<00:05,  6.34it/s, loss=4.72, v_num=0]
Epoch 11:  63%|██████▎   | 61/97 [00:09<00:05,  6.31it/s, loss=4.72, v_num=0]
Epoch 11:  64%|██████▍   | 62/97 [00:09<00:05,  6.36it/s, loss=4.72, v_num=0]
Epoch 11:  64%|██████▍   | 62/97 [00:09<00:05,  6.32it/s, loss=4.71, v_num=0]
Epoch 11:  65%|██████▍   | 63/97 [00:09<00:05,  6.37it/s, loss=4.71, v_num=0]
Epoch 11:  65%|██████▍   | 63/97 [00:09<00:05,  6.34it/s, loss=4.71, v_num=0]
Epoch 11:  66%|██████▌   | 64/97 [00:10<00:05,  6.38it/s, loss=4.71, v_num=0]
Epoch 11:  66%|██████▌   | 64/97 [00:10<00:05,  6.35it/s, loss=4.72, v_num=0]
Epoch 11:  67%|██████▋   | 65/97 [00:10<00:05,  6.40it/s, loss=4.72, v_num=0]
Epoch 11:  67%|██████▋   | 65/97 [00:10<00:05,  6.36it/s, loss=4.72, v_num=0]
Epoch 11:  68%|██████▊   | 66/97 [00:10<00:04,  6.41it/s, loss=4.72, v_num=0]
Epoch 11:  68%|██████▊   | 66/97 [00:10<00:04,  6.37it/s, loss=4.71, v_num=0]
Epoch 11:  69%|██████▉   | 67/97 [00:10<00:04,  6.42it/s, loss=4.71, v_num=0]
Epoch 11:  69%|██████▉   | 67/97 [00:10<00:04,  6.39it/s, loss=4.71, v_num=0]
Epoch 11:  70%|███████   | 68/97 [00:10<00:04,  6.43it/s, loss=4.71, v_num=0]
Epoch 11:  70%|███████   | 68/97 [00:10<00:04,  6.40it/s, loss=4.71, v_num=0]
Epoch 11:  71%|███████   | 69/97 [00:10<00:04,  6.44it/s, loss=4.71, v_num=0]
Epoch 11:  71%|███████   | 69/97 [00:10<00:04,  6.41it/s, loss=4.71, v_num=0]
Epoch 11:  72%|███████▏  | 70/97 [00:10<00:04,  6.45it/s, loss=4.71, v_num=0]
Epoch 11:  72%|███████▏  | 70/97 [00:10<00:04,  6.42it/s, loss=4.71, v_num=0]
Epoch 11:  73%|███████▎  | 71/97 [00:10<00:04,  6.46it/s, loss=4.71, v_num=0]
Epoch 11:  73%|███████▎  | 71/97 [00:11<00:04,  6.43it/s, loss=4.71, v_num=0]
Epoch 11:  74%|███████▍  | 72/97 [00:11<00:03,  6.48it/s, loss=4.71, v_num=0]
Epoch 11:  74%|███████▍  | 72/97 [00:11<00:03,  6.44it/s, loss=4.71, v_num=0]
Epoch 11:  75%|███████▌  | 73/97 [00:11<00:03,  6.49it/s, loss=4.71, v_num=0]
Epoch 11:  75%|███████▌  | 73/97 [00:11<00:03,  6.45it/s, loss=4.7, v_num=0]
Epoch 11:  76%|███████▋  | 74/97 [00:11<00:03,  6.50it/s, loss=4.7, v_num=0]
Epoch 11:  76%|███████▋  | 74/97 [00:11<00:03,  6.46it/s, loss=4.69, v_num=0]
Epoch 11:  77%|███████▋  | 75/97 [00:11<00:03,  6.50it/s, loss=4.69, v_num=0]
Epoch 11:  77%|███████▋  | 75/97 [00:11<00:03,  6.47it/s, loss=4.69, v_num=0]
Epoch 11:  78%|███████▊  | 76/97 [00:11<00:03,  6.51it/s, loss=4.69, v_num=0]
Epoch 11:  78%|███████▊  | 76/97 [00:11<00:03,  6.48it/s, loss=4.69, v_num=0]
Epoch 11:  79%|███████▉  | 77/97 [00:11<00:03,  6.53it/s, loss=4.69, v_num=0]
Epoch 11:  79%|███████▉  | 77/97 [00:11<00:03,  6.49it/s, loss=4.68, v_num=0]
Epoch 11:  80%|████████  | 78/97 [00:11<00:02,  6.53it/s, loss=4.68, v_num=0]
Epoch 11:  80%|████████  | 78/97 [00:11<00:02,  6.50it/s, loss=4.69, v_num=0]
Epoch 11:  81%|████████▏ | 79/97 [00:12<00:02,  6.54it/s, loss=4.69, v_num=0]
Epoch 11:  81%|████████▏ | 79/97 [00:12<00:02,  6.51it/s, loss=4.69, v_num=0]
Epoch 11:  82%|████████▏ | 80/97 [00:12<00:02,  6.55it/s, loss=4.69, v_num=0]
Epoch 11:  82%|████████▏ | 80/97 [00:12<00:02,  6.52it/s, loss=4.69, v_num=0]
Epoch 11:  84%|████████▎ | 81/97 [00:12<00:02,  6.56it/s, loss=4.69, v_num=0]
Epoch 11:  84%|████████▎ | 81/97 [00:12<00:02,  6.53it/s, loss=4.69, v_num=0]
Epoch 11:  85%|████████▍ | 82/97 [00:12<00:02,  6.56it/s, loss=4.69, v_num=0]
Epoch 11:  85%|████████▍ | 82/97 [00:12<00:02,  6.54it/s, loss=4.69, v_num=0]
Epoch 11:  86%|████████▌ | 83/97 [00:12<00:02,  6.57it/s, loss=4.69, v_num=0]
Epoch 11:  86%|████████▌ | 83/97 [00:12<00:02,  6.54it/s, loss=4.67, v_num=0]
Epoch 11:  87%|████████▋ | 84/97 [00:12<00:01,  6.58it/s, loss=4.67, v_num=0]
Epoch 11:  87%|████████▋ | 84/97 [00:12<00:01,  6.55it/s, loss=4.67, v_num=0]
Epoch 11:  88%|████████▊ | 85/97 [00:12<00:01,  6.59it/s, loss=4.67, v_num=0]
Epoch 11:  88%|████████▊ | 85/97 [00:12<00:01,  6.56it/s, loss=4.66, v_num=0]
Epoch 11:  89%|████████▊ | 86/97 [00:13<00:01,  6.60it/s, loss=4.66, v_num=0]
Epoch 11:  89%|████████▊ | 86/97 [00:13<00:01,  6.57it/s, loss=4.65, v_num=0]
Epoch 11:  90%|████████▉ | 87/97 [00:13<00:01,  6.60it/s, loss=4.65, v_num=0]
Epoch 11:  90%|████████▉ | 87/97 [00:13<00:01,  6.57it/s, loss=4.65, v_num=0]
Epoch 11:  91%|█████████ | 88/97 [00:13<00:01,  6.61it/s, loss=4.65, v_num=0]
Epoch 11:  91%|█████████ | 88/97 [00:13<00:01,  6.58it/s, loss=4.64, v_num=0]
Epoch 11:  92%|█████████▏| 89/97 [00:13<00:01,  6.62it/s, loss=4.64, v_num=0]
Epoch 11:  92%|█████████▏| 89/97 [00:13<00:01,  6.59it/s, loss=4.65, v_num=0]
Epoch 11:  93%|█████████▎| 90/97 [00:13<00:01,  6.62it/s, loss=4.65, v_num=0]
Epoch 11:  93%|█████████▎| 90/97 [00:13<00:01,  6.60it/s, loss=4.64, v_num=0]
Epoch 11:  94%|█████████▍| 91/97 [00:13<00:00,  6.63it/s, loss=4.64, v_num=0]
Epoch 11:  94%|█████████▍| 91/97 [00:13<00:00,  6.60it/s, loss=4.64, v_num=0]
Epoch 11:  95%|█████████▍| 92/97 [00:13<00:00,  6.64it/s, loss=4.64, v_num=0]
Epoch 11:  95%|█████████▍| 92/97 [00:13<00:00,  6.61it/s, loss=4.64, v_num=0]
Epoch 11:  96%|█████████▌| 93/97 [00:13<00:00,  6.64it/s, loss=4.64, v_num=0]
Epoch 11:  96%|█████████▌| 93/97 [00:14<00:00,  6.62it/s, loss=4.63, v_num=0]
Epoch 11:  97%|█████████▋| 94/97 [00:14<00:00,  6.65it/s, loss=4.63, v_num=0]
Epoch 11:  97%|█████████▋| 94/97 [00:14<00:00,  6.63it/s, loss=4.62, v_num=0]
Epoch 11:  98%|█████████▊| 95/97 [00:14<00:00,  6.66it/s, loss=4.62, v_num=0]
Epoch 11:  98%|█████████▊| 95/97 [00:14<00:00,  6.63it/s, loss=4.62, v_num=0]
Epoch 11:  99%|█████████▉| 96/97 [00:14<00:00,  6.66it/s, loss=4.62, v_num=0]
Epoch 11:  99%|█████████▉| 96/97 [00:14<00:00,  6.64it/s, loss=4.63, v_num=0]
Epoch 11: 100%|██████████| 97/97 [00:14<00:00,  6.67it/s, loss=4.63, v_num=0]
Epoch 11: 100%|██████████| 97/97 [00:14<00:00,  6.65it/s, loss=4.64, v_num=0]
Epoch 11: 100%|██████████| 97/97 [00:14<00:00,  6.65it/s, loss=4.64, v_num=0]
Epoch 11:   0%|          | 0/97 [00:00<?, ?it/s, loss=4.64, v_num=0]
Epoch 12:   0%|          | 0/97 [00:00<?, ?it/s, loss=4.64, v_num=0]
Epoch 12:   1%|          | 1/97 [00:01<02:17,  1.44s/it, loss=4.64, v_num=0]
Epoch 12:   1%|          | 1/97 [00:01<02:23,  1.49s/it, loss=4.63, v_num=0]
Epoch 12:   2%|▏         | 2/97 [00:01<01:14,  1.27it/s, loss=4.63, v_num=0]
Epoch 12:   2%|▏         | 2/97 [00:01<01:17,  1.23it/s, loss=4.64, v_num=0]
Epoch 12:   3%|▎         | 3/97 [00:01<00:53,  1.76it/s, loss=4.64, v_num=0]
Epoch 12:   3%|▎         | 3/97 [00:01<00:55,  1.70it/s, loss=4.64, v_num=0]
Epoch 12:   4%|▍         | 4/97 [00:01<00:42,  2.17it/s, loss=4.64, v_num=0]
Epoch 12:   4%|▍         | 4/97 [00:01<00:44,  2.11it/s, loss=4.64, v_num=0]
Epoch 12:   5%|▌         | 5/97 [00:01<00:36,  2.52it/s, loss=4.64, v_num=0]
Epoch 12:   5%|▌         | 5/97 [00:02<00:37,  2.46it/s, loss=4.63, v_num=0]
Epoch 12:   6%|▌         | 6/97 [00:02<00:32,  2.83it/s, loss=4.63, v_num=0]
Epoch 12:   6%|▌         | 6/97 [00:02<00:32,  2.76it/s, loss=4.64, v_num=0]
Epoch 12:   7%|▋         | 7/97 [00:02<00:28,  3.11it/s, loss=4.64, v_num=0]
Epoch 12:   7%|▋         | 7/97 [00:02<00:29,  3.03it/s, loss=4.63, v_num=0]
Epoch 12:   8%|▊         | 8/97 [00:02<00:26,  3.35it/s, loss=4.63, v_num=0]
Epoch 12:   8%|▊         | 8/97 [00:02<00:27,  3.27it/s, loss=4.63, v_num=0]
Epoch 12:   9%|▉         | 9/97 [00:02<00:24,  3.55it/s, loss=4.63, v_num=0]
Epoch 12:   9%|▉         | 9/97 [00:02<00:25,  3.47it/s, loss=4.62, v_num=0]
Epoch 12:  10%|█         | 10/97 [00:02<00:23,  3.74it/s, loss=4.62, v_num=0]
Epoch 12:  10%|█         | 10/97 [00:02<00:23,  3.67it/s, loss=4.62, v_num=0]
Epoch 12:  11%|█▏        | 11/97 [00:02<00:21,  3.92it/s, loss=4.62, v_num=0]
Epoch 12:  11%|█▏        | 11/97 [00:02<00:22,  3.84it/s, loss=4.61, v_num=0]
Epoch 12:  12%|█▏        | 12/97 [00:02<00:20,  4.07it/s, loss=4.61, v_num=0]
Epoch 12:  12%|█▏        | 12/97 [00:03<00:21,  4.00it/s, loss=4.6, v_num=0]
Epoch 12:  13%|█▎        | 13/97 [00:03<00:19,  4.22it/s, loss=4.6, v_num=0]
Epoch 12:  13%|█▎        | 13/97 [00:03<00:20,  4.14it/s, loss=4.61, v_num=0]
Epoch 12:  14%|█▍        | 14/97 [00:03<00:19,  4.35it/s, loss=4.61, v_num=0]
Epoch 12:  14%|█▍        | 14/97 [00:03<00:19,  4.28it/s, loss=4.6, v_num=0]
Epoch 12:  15%|█▌        | 15/97 [00:03<00:18,  4.47it/s, loss=4.6, v_num=0]
Epoch 12:  15%|█▌        | 15/97 [00:03<00:18,  4.40it/s, loss=4.61, v_num=0]
Epoch 12:  16%|█▋        | 16/97 [00:03<00:17,  4.58it/s, loss=4.61, v_num=0]
Epoch 12:  16%|█▋        | 16/97 [00:03<00:17,  4.51it/s, loss=4.61, v_num=0]
Epoch 12:  18%|█▊        | 17/97 [00:03<00:17,  4.68it/s, loss=4.61, v_num=0]
Epoch 12:  18%|█▊        | 17/97 [00:03<00:17,  4.61it/s, loss=4.62, v_num=0]
Epoch 12:  19%|█▊        | 18/97 [00:03<00:16,  4.78it/s, loss=4.62, v_num=0]
Epoch 12:  19%|█▊        | 18/97 [00:03<00:16,  4.71it/s, loss=4.62, v_num=0]
Epoch 12:  20%|█▉        | 19/97 [00:03<00:16,  4.87it/s, loss=4.62, v_num=0]
Epoch 12:  20%|█▉        | 19/97 [00:03<00:16,  4.80it/s, loss=4.61, v_num=0]
Epoch 12:  21%|██        | 20/97 [00:04<00:15,  4.95it/s, loss=4.61, v_num=0]
Epoch 12:  21%|██        | 20/97 [00:04<00:15,  4.88it/s, loss=4.6, v_num=0]
Epoch 12:  22%|██▏       | 21/97 [00:04<00:15,  5.03it/s, loss=4.6, v_num=0]
Epoch 12:  22%|██▏       | 21/97 [00:04<00:15,  4.96it/s, loss=4.6, v_num=0]
Epoch 12:  23%|██▎       | 22/97 [00:04<00:14,  5.10it/s, loss=4.6, v_num=0]
Epoch 12:  23%|██▎       | 22/97 [00:04<00:14,  5.03it/s, loss=4.59, v_num=0]
Epoch 12:  24%|██▎       | 23/97 [00:04<00:14,  5.17it/s, loss=4.59, v_num=0]
Epoch 12:  24%|██▎       | 23/97 [00:04<00:14,  5.10it/s, loss=4.58, v_num=0]
Epoch 12:  25%|██▍       | 24/97 [00:04<00:13,  5.23it/s, loss=4.58, v_num=0]
Epoch 12:  25%|██▍       | 24/97 [00:04<00:14,  5.16it/s, loss=4.58, v_num=0]
Epoch 12:  26%|██▌       | 25/97 [00:04<00:13,  5.28it/s, loss=4.58, v_num=0]
Epoch 12:  26%|██▌       | 25/97 [00:04<00:13,  5.22it/s, loss=4.59, v_num=0]
Epoch 12:  27%|██▋       | 26/97 [00:04<00:13,  5.34it/s, loss=4.59, v_num=0]
Epoch 12:  27%|██▋       | 26/97 [00:04<00:13,  5.28it/s, loss=4.59, v_num=0]
Epoch 12:  28%|██▊       | 27/97 [00:05<00:12,  5.40it/s, loss=4.59, v_num=0]
Epoch 12:  28%|██▊       | 27/97 [00:05<00:13,  5.34it/s, loss=4.6, v_num=0]
Epoch 12:  29%|██▉       | 28/97 [00:05<00:12,  5.45it/s, loss=4.6, v_num=0]
Epoch 12:  29%|██▉       | 28/97 [00:05<00:12,  5.39it/s, loss=4.6, v_num=0]
Epoch 12:  30%|██▉       | 29/97 [00:05<00:12,  5.49it/s, loss=4.6, v_num=0]
Epoch 12:  30%|██▉       | 29/97 [00:05<00:12,  5.44it/s, loss=4.6, v_num=0]
Epoch 12:  31%|███       | 30/97 [00:05<00:12,  5.53it/s, loss=4.6, v_num=0]
Epoch 12:  31%|███       | 30/97 [00:05<00:12,  5.48it/s, loss=4.6, v_num=0]
Epoch 12:  32%|███▏      | 31/97 [00:05<00:11,  5.58it/s, loss=4.6, v_num=0]
Epoch 12:  32%|███▏      | 31/97 [00:05<00:11,  5.52it/s, loss=4.6, v_num=0]
Epoch 12:  33%|███▎      | 32/97 [00:05<00:11,  5.62it/s, loss=4.6, v_num=0]
Epoch 12:  33%|███▎      | 32/97 [00:05<00:11,  5.56it/s, loss=4.6, v_num=0]
Epoch 12:  34%|███▍      | 33/97 [00:05<00:11,  5.65it/s, loss=4.6, v_num=0]
Epoch 12:  34%|███▍      | 33/97 [00:05<00:11,  5.60it/s, loss=4.59, v_num=0]
Epoch 12:  35%|███▌      | 34/97 [00:05<00:11,  5.69it/s, loss=4.59, v_num=0]
Epoch 12:  35%|███▌      | 34/97 [00:06<00:11,  5.64it/s, loss=4.6, v_num=0]
Epoch 12:  36%|███▌      | 35/97 [00:06<00:10,  5.73it/s, loss=4.6, v_num=0]
Epoch 12:  36%|███▌      | 35/97 [00:06<00:10,  5.68it/s, loss=4.59, v_num=0]
Epoch 12:  37%|███▋      | 36/97 [00:06<00:10,  5.76it/s, loss=4.59, v_num=0]
Epoch 12:  37%|███▋      | 36/97 [00:06<00:10,  5.71it/s, loss=4.6, v_num=0]
Epoch 12:  38%|███▊      | 37/97 [00:06<00:10,  5.79it/s, loss=4.6, v_num=0]
Epoch 12:  38%|███▊      | 37/97 [00:06<00:10,  5.74it/s, loss=4.6, v_num=0]
Epoch 12:  39%|███▉      | 38/97 [00:06<00:10,  5.82it/s, loss=4.6, v_num=0]
Epoch 12:  39%|███▉      | 38/97 [00:06<00:10,  5.78it/s, loss=4.59, v_num=0]
Epoch 12:  40%|████      | 39/97 [00:06<00:09,  5.85it/s, loss=4.59, v_num=0]
Epoch 12:  40%|████      | 39/97 [00:06<00:09,  5.81it/s, loss=4.59, v_num=0]
Epoch 12:  41%|████      | 40/97 [00:06<00:09,  5.88it/s, loss=4.59, v_num=0]
Epoch 12:  41%|████      | 40/97 [00:06<00:09,  5.84it/s, loss=4.58, v_num=0]
Epoch 12:  42%|████▏     | 41/97 [00:06<00:09,  5.92it/s, loss=4.58, v_num=0]
Epoch 12:  42%|████▏     | 41/97 [00:06<00:09,  5.87it/s, loss=4.57, v_num=0]
Epoch 12:  43%|████▎     | 42/97 [00:07<00:09,  5.94it/s, loss=4.57, v_num=0]
Epoch 12:  43%|████▎     | 42/97 [00:07<00:09,  5.90it/s, loss=4.58, v_num=0]
Epoch 12:  44%|████▍     | 43/97 [00:07<00:09,  5.97it/s, loss=4.58, v_num=0]
Epoch 12:  44%|████▍     | 43/97 [00:07<00:09,  5.92it/s, loss=4.58, v_num=0]
Epoch 12:  45%|████▌     | 44/97 [00:07<00:08,  5.99it/s, loss=4.58, v_num=0]
Epoch 12:  45%|████▌     | 44/97 [00:07<00:08,  5.95it/s, loss=4.57, v_num=0]
Epoch 12:  46%|████▋     | 45/97 [00:07<00:08,  6.02it/s, loss=4.57, v_num=0]
Epoch 12:  46%|████▋     | 45/97 [00:07<00:08,  5.97it/s, loss=4.57, v_num=0]
Epoch 12:  47%|████▋     | 46/97 [00:07<00:08,  6.04it/s, loss=4.57, v_num=0]
Epoch 12:  47%|████▋     | 46/97 [00:07<00:08,  6.00it/s, loss=4.57, v_num=0]
Epoch 12:  48%|████▊     | 47/97 [00:07<00:08,  6.07it/s, loss=4.57, v_num=0]
Epoch 12:  48%|████▊     | 47/97 [00:07<00:08,  6.02it/s, loss=4.57, v_num=0]
Epoch 12:  49%|████▉     | 48/97 [00:07<00:08,  6.08it/s, loss=4.57, v_num=0]
Epoch 12:  49%|████▉     | 48/97 [00:07<00:08,  6.04it/s, loss=4.57, v_num=0]
Epoch 12:  51%|█████     | 49/97 [00:08<00:07,  6.11it/s, loss=4.57, v_num=0]
Epoch 12:  51%|█████     | 49/97 [00:08<00:07,  6.07it/s, loss=4.56, v_num=0]
Epoch 12:  52%|█████▏    | 50/97 [00:08<00:07,  6.13it/s, loss=4.56, v_num=0]
Epoch 12:  52%|█████▏    | 50/97 [00:08<00:07,  6.09it/s, loss=4.56, v_num=0]
Epoch 12:  53%|█████▎    | 51/97 [00:08<00:07,  6.15it/s, loss=4.56, v_num=0]
Epoch 12:  53%|█████▎    | 51/97 [00:08<00:07,  6.11it/s, loss=4.55, v_num=0]
Epoch 12:  54%|█████▎    | 52/97 [00:08<00:07,  6.17it/s, loss=4.55, v_num=0]
Epoch 12:  54%|█████▎    | 52/97 [00:08<00:07,  6.13it/s, loss=4.56, v_num=0]
Epoch 12:  55%|█████▍    | 53/97 [00:08<00:07,  6.19it/s, loss=4.56, v_num=0]
Epoch 12:  55%|█████▍    | 53/97 [00:08<00:07,  6.15it/s, loss=4.56, v_num=0]
Epoch 12:  56%|█████▌    | 54/97 [00:08<00:06,  6.21it/s, loss=4.56, v_num=0]
Epoch 12:  56%|█████▌    | 54/97 [00:08<00:06,  6.16it/s, loss=4.56, v_num=0]
Epoch 12:  57%|█████▋    | 55/97 [00:08<00:06,  6.22it/s, loss=4.56, v_num=0]
Epoch 12:  57%|█████▋    | 55/97 [00:08<00:06,  6.18it/s, loss=4.56, v_num=0]
Epoch 12:  58%|█████▊    | 56/97 [00:08<00:06,  6.24it/s, loss=4.56, v_num=0]
Epoch 12:  58%|█████▊    | 56/97 [00:09<00:06,  6.20it/s, loss=4.55, v_num=0]
Epoch 12:  59%|█████▉    | 57/97 [00:09<00:06,  6.25it/s, loss=4.55, v_num=0]
Epoch 12:  59%|█████▉    | 57/97 [00:09<00:06,  6.22it/s, loss=4.54, v_num=0]
Epoch 12:  60%|█████▉    | 58/97 [00:09<00:06,  6.27it/s, loss=4.54, v_num=0]
Epoch 12:  60%|█████▉    | 58/97 [00:09<00:06,  6.23it/s, loss=4.54, v_num=0]
Epoch 12:  61%|██████    | 59/97 [00:09<00:06,  6.29it/s, loss=4.54, v_num=0]
Epoch 12:  61%|██████    | 59/97 [00:09<00:06,  6.25it/s, loss=4.55, v_num=0]
Epoch 12:  62%|██████▏   | 60/97 [00:09<00:05,  6.30it/s, loss=4.55, v_num=0]
Epoch 12:  62%|██████▏   | 60/97 [00:09<00:05,  6.26it/s, loss=4.55, v_num=0]
Epoch 12:  63%|██████▎   | 61/97 [00:09<00:05,  6.32it/s, loss=4.55, v_num=0]
Epoch 12:  63%|██████▎   | 61/97 [00:09<00:05,  6.28it/s, loss=4.56, v_num=0]
Epoch 12:  64%|██████▍   | 62/97 [00:09<00:05,  6.33it/s, loss=4.56, v_num=0]
Epoch 12:  64%|██████▍   | 62/97 [00:09<00:05,  6.29it/s, loss=4.55, v_num=0]
Epoch 12:  65%|██████▍   | 63/97 [00:09<00:05,  6.34it/s, loss=4.55, v_num=0]
Epoch 12:  65%|██████▍   | 63/97 [00:09<00:05,  6.31it/s, loss=4.55, v_num=0]
Epoch 12:  66%|██████▌   | 64/97 [00:10<00:05,  6.36it/s, loss=4.55, v_num=0]
Epoch 12:  66%|██████▌   | 64/97 [00:10<00:05,  6.32it/s, loss=4.55, v_num=0]
Epoch 12:  67%|██████▋   | 65/97 [00:10<00:05,  6.37it/s, loss=4.55, v_num=0]
Epoch 12:  67%|██████▋   | 65/97 [00:10<00:05,  6.33it/s, loss=4.55, v_num=0]
Epoch 12:  68%|██████▊   | 66/97 [00:10<00:04,  6.38it/s, loss=4.55, v_num=0]
Epoch 12:  68%|██████▊   | 66/97 [00:10<00:04,  6.35it/s, loss=4.55, v_num=0]
Epoch 12:  69%|██████▉   | 67/97 [00:10<00:04,  6.39it/s, loss=4.55, v_num=0]
Epoch 12:  69%|██████▉   | 67/97 [00:10<00:04,  6.36it/s, loss=4.56, v_num=0]
Epoch 12:  70%|███████   | 68/97 [00:10<00:04,  6.41it/s, loss=4.56, v_num=0]
Epoch 12:  70%|███████   | 68/97 [00:10<00:04,  6.37it/s, loss=4.57, v_num=0]
Epoch 12:  71%|███████   | 69/97 [00:10<00:04,  6.42it/s, loss=4.57, v_num=0]
Epoch 12:  71%|███████   | 69/97 [00:10<00:04,  6.38it/s, loss=4.57, v_num=0]
Epoch 12:  72%|███████▏  | 70/97 [00:10<00:04,  6.43it/s, loss=4.57, v_num=0]
Epoch 12:  72%|███████▏  | 70/97 [00:10<00:04,  6.40it/s, loss=4.58, v_num=0]
Epoch 12:  73%|███████▎  | 71/97 [00:11<00:04,  6.44it/s, loss=4.58, v_num=0]
Epoch 12:  73%|███████▎  | 71/97 [00:11<00:04,  6.41it/s, loss=4.59, v_num=0]
Epoch 12:  74%|███████▍  | 72/97 [00:11<00:03,  6.45it/s, loss=4.59, v_num=0]
Epoch 12:  74%|███████▍  | 72/97 [00:11<00:03,  6.42it/s, loss=4.58, v_num=0]
Epoch 12:  75%|███████▌  | 73/97 [00:11<00:03,  6.46it/s, loss=4.58, v_num=0]
Epoch 12:  75%|███████▌  | 73/97 [00:11<00:03,  6.43it/s, loss=4.58, v_num=0]
Epoch 12:  76%|███████▋  | 74/97 [00:11<00:03,  6.47it/s, loss=4.58, v_num=0]
Epoch 12:  76%|███████▋  | 74/97 [00:11<00:03,  6.44it/s, loss=4.58, v_num=0]
Epoch 12:  77%|███████▋  | 75/97 [00:11<00:03,  6.48it/s, loss=4.58, v_num=0]
Epoch 12:  77%|███████▋  | 75/97 [00:11<00:03,  6.45it/s, loss=4.58, v_num=0]
Epoch 12:  78%|███████▊  | 76/97 [00:11<00:03,  6.49it/s, loss=4.58, v_num=0]
Epoch 12:  78%|███████▊  | 76/97 [00:11<00:03,  6.46it/s, loss=4.58, v_num=0]
Epoch 12:  79%|███████▉  | 77/97 [00:11<00:03,  6.50it/s, loss=4.58, v_num=0]
Epoch 12:  79%|███████▉  | 77/97 [00:11<00:03,  6.47it/s, loss=4.58, v_num=0]
Epoch 12:  80%|████████  | 78/97 [00:11<00:02,  6.51it/s, loss=4.58, v_num=0]
Epoch 12:  80%|████████  | 78/97 [00:12<00:02,  6.48it/s, loss=4.58, v_num=0]
Epoch 12:  81%|████████▏ | 79/97 [00:12<00:02,  6.52it/s, loss=4.58, v_num=0]
Epoch 12:  81%|████████▏ | 79/97 [00:12<00:02,  6.49it/s, loss=4.58, v_num=0]
Epoch 12:  82%|████████▏ | 80/97 [00:12<00:02,  6.53it/s, loss=4.58, v_num=0]
Epoch 12:  82%|████████▏ | 80/97 [00:12<00:02,  6.50it/s, loss=4.58, v_num=0]
Epoch 12:  84%|████████▎ | 81/97 [00:12<00:02,  6.54it/s, loss=4.58, v_num=0]
Epoch 12:  84%|████████▎ | 81/97 [00:12<00:02,  6.51it/s, loss=4.57, v_num=0]
Epoch 12:  85%|████████▍ | 82/97 [00:12<00:02,  6.54it/s, loss=4.57, v_num=0]
Epoch 12:  85%|████████▍ | 82/97 [00:12<00:02,  6.52it/s, loss=4.57, v_num=0]
Epoch 12:  86%|████████▌ | 83/97 [00:12<00:02,  6.55it/s, loss=4.57, v_num=0]
Epoch 12:  86%|████████▌ | 83/97 [00:12<00:02,  6.52it/s, loss=4.57, v_num=0]
Epoch 12:  87%|████████▋ | 84/97 [00:12<00:01,  6.56it/s, loss=4.57, v_num=0]
Epoch 12:  87%|████████▋ | 84/97 [00:12<00:01,  6.53it/s, loss=4.57, v_num=0]
Epoch 12:  88%|████████▊ | 85/97 [00:12<00:01,  6.57it/s, loss=4.57, v_num=0]
Epoch 12:  88%|████████▊ | 85/97 [00:12<00:01,  6.54it/s, loss=4.56, v_num=0]
Epoch 12:  89%|████████▊ | 86/97 [00:13<00:01,  6.58it/s, loss=4.56, v_num=0]
Epoch 12:  89%|████████▊ | 86/97 [00:13<00:01,  6.55it/s, loss=4.56, v_num=0]
Epoch 12:  90%|████████▉ | 87/97 [00:13<00:01,  6.58it/s, loss=4.56, v_num=0]
Epoch 12:  90%|████████▉ | 87/97 [00:13<00:01,  6.56it/s, loss=4.55, v_num=0]
Epoch 12:  91%|█████████ | 88/97 [00:13<00:01,  6.59it/s, loss=4.55, v_num=0]
Epoch 12:  91%|█████████ | 88/97 [00:13<00:01,  6.57it/s, loss=4.54, v_num=0]
Epoch 12:  92%|█████████▏| 89/97 [00:13<00:01,  6.60it/s, loss=4.54, v_num=0]
Epoch 12:  92%|█████████▏| 89/97 [00:13<00:01,  6.57it/s, loss=4.53, v_num=0]
Epoch 12:  93%|█████████▎| 90/97 [00:13<00:01,  6.61it/s, loss=4.53, v_num=0]
Epoch 12:  93%|█████████▎| 90/97 [00:13<00:01,  6.58it/s, loss=4.53, v_num=0]
Epoch 12:  94%|█████████▍| 91/97 [00:13<00:00,  6.61it/s, loss=4.53, v_num=0]
Epoch 12:  94%|█████████▍| 91/97 [00:13<00:00,  6.59it/s, loss=4.53, v_num=0]
Epoch 12:  95%|█████████▍| 92/97 [00:13<00:00,  6.62it/s, loss=4.53, v_num=0]
Epoch 12:  95%|█████████▍| 92/97 [00:13<00:00,  6.60it/s, loss=4.52, v_num=0]
Epoch 12:  96%|█████████▌| 93/97 [00:14<00:00,  6.63it/s, loss=4.52, v_num=0]
Epoch 12:  96%|█████████▌| 93/97 [00:14<00:00,  6.60it/s, loss=4.52, v_num=0]
Epoch 12:  97%|█████████▋| 94/97 [00:14<00:00,  6.63it/s, loss=4.52, v_num=0]
Epoch 12:  97%|█████████▋| 94/97 [00:14<00:00,  6.61it/s, loss=4.51, v_num=0]
Epoch 12:  98%|█████████▊| 95/97 [00:14<00:00,  6.64it/s, loss=4.51, v_num=0]
Epoch 12:  98%|█████████▊| 95/97 [00:14<00:00,  6.62it/s, loss=4.5, v_num=0]
Epoch 12:  99%|█████████▉| 96/97 [00:14<00:00,  6.65it/s, loss=4.5, v_num=0]
Epoch 12:  99%|█████████▉| 96/97 [00:14<00:00,  6.62it/s, loss=4.51, v_num=0]
Epoch 12: 100%|██████████| 97/97 [00:14<00:00,  6.65it/s, loss=4.51, v_num=0]
Epoch 12: 100%|██████████| 97/97 [00:14<00:00,  6.63it/s, loss=4.5, v_num=0]
Epoch 12: 100%|██████████| 97/97 [00:14<00:00,  6.63it/s, loss=4.5, v_num=0]
Epoch 12:   0%|          | 0/97 [00:00<?, ?it/s, loss=4.5, v_num=0]
Epoch 13:   0%|          | 0/97 [00:00<?, ?it/s, loss=4.5, v_num=0]
Epoch 13:   1%|          | 1/97 [00:01<02:12,  1.38s/it, loss=4.5, v_num=0]
Epoch 13:   1%|          | 1/97 [00:01<02:17,  1.43s/it, loss=4.49, v_num=0]
Epoch 13:   2%|▏         | 2/97 [00:01<01:12,  1.32it/s, loss=4.49, v_num=0]
Epoch 13:   2%|▏         | 2/97 [00:01<01:15,  1.27it/s, loss=4.49, v_num=0]
Epoch 13:   3%|▎         | 3/97 [00:01<00:51,  1.81it/s, loss=4.49, v_num=0]
Epoch 13:   3%|▎         | 3/97 [00:01<00:53,  1.75it/s, loss=4.49, v_num=0]
Epoch 13:   4%|▍         | 4/97 [00:01<00:41,  2.23it/s, loss=4.49, v_num=0]
Epoch 13:   4%|▍         | 4/97 [00:01<00:43,  2.16it/s, loss=4.48, v_num=0]
Epoch 13:   5%|▌         | 5/97 [00:01<00:35,  2.59it/s, loss=4.48, v_num=0]
Epoch 13:   5%|▌         | 5/97 [00:01<00:36,  2.51it/s, loss=4.49, v_num=0]
Epoch 13:   6%|▌         | 6/97 [00:02<00:31,  2.90it/s, loss=4.49, v_num=0]
Epoch 13:   6%|▌         | 6/97 [00:02<00:32,  2.82it/s, loss=4.49, v_num=0]
Epoch 13:   7%|▋         | 7/97 [00:02<00:28,  3.17it/s, loss=4.49, v_num=0]
Epoch 13:   7%|▋         | 7/97 [00:02<00:29,  3.09it/s, loss=4.49, v_num=0]
Epoch 13:   8%|▊         | 8/97 [00:02<00:26,  3.41it/s, loss=4.49, v_num=0]
Epoch 13:   8%|▊         | 8/97 [00:02<00:26,  3.33it/s, loss=4.49, v_num=0]
Epoch 13:   9%|▉         | 9/97 [00:02<00:24,  3.62it/s, loss=4.49, v_num=0]
Epoch 13:   9%|▉         | 9/97 [00:02<00:24,  3.54it/s, loss=4.49, v_num=0]
Epoch 13:  10%|█         | 10/97 [00:02<00:22,  3.80it/s, loss=4.49, v_num=0]
Epoch 13:  10%|█         | 10/97 [00:02<00:23,  3.72it/s, loss=4.48, v_num=0]
Epoch 13:  11%|█▏        | 11/97 [00:02<00:21,  3.98it/s, loss=4.48, v_num=0]
Epoch 13:  11%|█▏        | 11/97 [00:02<00:22,  3.90it/s, loss=4.47, v_num=0]
Epoch 13:  12%|█▏        | 12/97 [00:02<00:20,  4.13it/s, loss=4.47, v_num=0]
Epoch 13:  12%|█▏        | 12/97 [00:02<00:20,  4.06it/s, loss=4.47, v_num=0]
Epoch 13:  13%|█▎        | 13/97 [00:03<00:19,  4.28it/s, loss=4.47, v_num=0]
Epoch 13:  13%|█▎        | 13/97 [00:03<00:19,  4.20it/s, loss=4.46, v_num=0]
Epoch 13:  14%|█▍        | 14/97 [00:03<00:18,  4.41it/s, loss=4.46, v_num=0]
Epoch 13:  14%|█▍        | 14/97 [00:03<00:19,  4.34it/s, loss=4.46, v_num=0]
Epoch 13:  15%|█▌        | 15/97 [00:03<00:18,  4.54it/s, loss=4.46, v_num=0]
Epoch 13:  15%|█▌        | 15/97 [00:03<00:18,  4.46it/s, loss=4.45, v_num=0]
Epoch 13:  16%|█▋        | 16/97 [00:03<00:17,  4.65it/s, loss=4.45, v_num=0]
Epoch 13:  16%|█▋        | 16/97 [00:03<00:17,  4.57it/s, loss=4.45, v_num=0]
Epoch 13:  18%|█▊        | 17/97 [00:03<00:16,  4.75it/s, loss=4.45, v_num=0]
Epoch 13:  18%|█▊        | 17/97 [00:03<00:17,  4.67it/s, loss=4.45, v_num=0]
Epoch 13:  19%|█▊        | 18/97 [00:03<00:16,  4.84it/s, loss=4.45, v_num=0]
Epoch 13:  19%|█▊        | 18/97 [00:03<00:16,  4.77it/s, loss=4.46, v_num=0]
Epoch 13:  20%|█▉        | 19/97 [00:03<00:15,  4.93it/s, loss=4.46, v_num=0]
Epoch 13:  20%|█▉        | 19/97 [00:03<00:16,  4.86it/s, loss=4.46, v_num=0]
Epoch 13:  21%|██        | 20/97 [00:03<00:15,  5.01it/s, loss=4.46, v_num=0]
Epoch 13:  21%|██        | 20/97 [00:04<00:15,  4.94it/s, loss=4.46, v_num=0]
Epoch 13:  22%|██▏       | 21/97 [00:04<00:14,  5.09it/s, loss=4.46, v_num=0]
Epoch 13:  22%|██▏       | 21/97 [00:04<00:15,  5.02it/s, loss=4.47, v_num=0]
Epoch 13:  23%|██▎       | 22/97 [00:04<00:14,  5.16it/s, loss=4.47, v_num=0]
Epoch 13:  23%|██▎       | 22/97 [00:04<00:14,  5.09it/s, loss=4.47, v_num=0]
Epoch 13:  24%|██▎       | 23/97 [00:04<00:14,  5.23it/s, loss=4.47, v_num=0]
Epoch 13:  24%|██▎       | 23/97 [00:04<00:14,  5.16it/s, loss=4.47, v_num=0]
Epoch 13:  25%|██▍       | 24/97 [00:04<00:13,  5.28it/s, loss=4.47, v_num=0]
Epoch 13:  25%|██▍       | 24/97 [00:04<00:13,  5.22it/s, loss=4.47, v_num=0]
Epoch 13:  26%|██▌       | 25/97 [00:04<00:13,  5.34it/s, loss=4.47, v_num=0]
Epoch 13:  26%|██▌       | 25/97 [00:04<00:13,  5.28it/s, loss=4.46, v_num=0]
Epoch 13:  27%|██▋       | 26/97 [00:04<00:13,  5.40it/s, loss=4.46, v_num=0]
Epoch 13:  27%|██▋       | 26/97 [00:04<00:13,  5.33it/s, loss=4.47, v_num=0]
Epoch 13:  28%|██▊       | 27/97 [00:04<00:12,  5.45it/s, loss=4.47, v_num=0]
Epoch 13:  28%|██▊       | 27/97 [00:05<00:12,  5.39it/s, loss=4.46, v_num=0]
Epoch 13:  29%|██▉       | 28/97 [00:05<00:12,  5.50it/s, loss=4.46, v_num=0]
Epoch 13:  29%|██▉       | 28/97 [00:05<00:12,  5.44it/s, loss=4.47, v_num=0]
Epoch 13:  30%|██▉       | 29/97 [00:05<00:12,  5.55it/s, loss=4.47, v_num=0]
Epoch 13:  30%|██▉       | 29/97 [00:05<00:12,  5.49it/s, loss=4.46, v_num=0]
Epoch 13:  31%|███       | 30/97 [00:05<00:11,  5.59it/s, loss=4.46, v_num=0]
Epoch 13:  31%|███       | 30/97 [00:05<00:12,  5.53it/s, loss=4.46, v_num=0]
Epoch 13:  32%|███▏      | 31/97 [00:05<00:11,  5.63it/s, loss=4.46, v_num=0]
Epoch 13:  32%|███▏      | 31/97 [00:05<00:11,  5.57it/s, loss=4.47, v_num=0]
Epoch 13:  33%|███▎      | 32/97 [00:05<00:11,  5.67it/s, loss=4.47, v_num=0]
Epoch 13:  33%|███▎      | 32/97 [00:05<00:11,  5.62it/s, loss=4.47, v_num=0]
Epoch 13:  34%|███▍      | 33/97 [00:05<00:11,  5.71it/s, loss=4.47, v_num=0]
Epoch 13:  34%|███▍      | 33/97 [00:05<00:11,  5.65it/s, loss=4.47, v_num=0]
Epoch 13:  35%|███▌      | 34/97 [00:05<00:10,  5.74it/s, loss=4.47, v_num=0]
Epoch 13:  35%|███▌      | 34/97 [00:05<00:11,  5.69it/s, loss=4.47, v_num=0]
Epoch 13:  36%|███▌      | 35/97 [00:06<00:10,  5.78it/s, loss=4.47, v_num=0]
Epoch 13:  36%|███▌      | 35/97 [00:06<00:10,  5.73it/s, loss=4.48, v_num=0]
Epoch 13:  37%|███▋      | 36/97 [00:06<00:10,  5.81it/s, loss=4.48, v_num=0]
Epoch 13:  37%|███▋      | 36/97 [00:06<00:10,  5.76it/s, loss=4.49, v_num=0]
Epoch 13:  38%|███▊      | 37/97 [00:06<00:10,  5.84it/s, loss=4.49, v_num=0]
Epoch 13:  38%|███▊      | 37/97 [00:06<00:10,  5.79it/s, loss=4.5, v_num=0]
Epoch 13:  39%|███▉      | 38/97 [00:06<00:10,  5.87it/s, loss=4.5, v_num=0]
Epoch 13:  39%|███▉      | 38/97 [00:06<00:10,  5.82it/s, loss=4.49, v_num=0]
Epoch 13:  40%|████      | 39/97 [00:06<00:09,  5.90it/s, loss=4.49, v_num=0]
Epoch 13:  40%|████      | 39/97 [00:06<00:09,  5.85it/s, loss=4.48, v_num=0]
Epoch 13:  41%|████      | 40/97 [00:06<00:09,  5.92it/s, loss=4.48, v_num=0]
Epoch 13:  41%|████      | 40/97 [00:06<00:09,  5.88it/s, loss=4.48, v_num=0]
Epoch 13:  42%|████▏     | 41/97 [00:06<00:09,  5.96it/s, loss=4.48, v_num=0]
Epoch 13:  42%|████▏     | 41/97 [00:06<00:09,  5.90it/s, loss=4.48, v_num=0]
Epoch 13:  43%|████▎     | 42/97 [00:07<00:09,  5.98it/s, loss=4.48, v_num=0]
Epoch 13:  43%|████▎     | 42/97 [00:07<00:09,  5.93it/s, loss=4.47, v_num=0]
Epoch 13:  44%|████▍     | 43/97 [00:07<00:08,  6.00it/s, loss=4.47, v_num=0]
Epoch 13:  44%|████▍     | 43/97 [00:07<00:09,  5.95it/s, loss=4.48, v_num=0]
Epoch 13:  45%|████▌     | 44/97 [00:07<00:08,  6.02it/s, loss=4.48, v_num=0]
Epoch 13:  45%|████▌     | 44/97 [00:07<00:08,  5.98it/s, loss=4.48, v_num=0]
Epoch 13:  46%|████▋     | 45/97 [00:07<00:08,  6.05it/s, loss=4.48, v_num=0]
Epoch 13:  46%|████▋     | 45/97 [00:07<00:08,  6.00it/s, loss=4.48, v_num=0]
Epoch 13:  47%|████▋     | 46/97 [00:07<00:08,  6.07it/s, loss=4.48, v_num=0]
Epoch 13:  47%|████▋     | 46/97 [00:07<00:08,  6.03it/s, loss=4.47, v_num=0]
Epoch 13:  48%|████▊     | 47/97 [00:07<00:08,  6.10it/s, loss=4.47, v_num=0]
Epoch 13:  48%|████▊     | 47/97 [00:07<00:08,  6.05it/s, loss=4.47, v_num=0]
Epoch 13:  49%|████▉     | 48/97 [00:07<00:08,  6.11it/s, loss=4.47, v_num=0]
Epoch 13:  49%|████▉     | 48/97 [00:07<00:08,  6.07it/s, loss=4.47, v_num=0]
Epoch 13:  51%|█████     | 49/97 [00:07<00:07,  6.13it/s, loss=4.47, v_num=0]
Epoch 13:  51%|█████     | 49/97 [00:08<00:07,  6.09it/s, loss=4.47, v_num=0]
Epoch 13:  52%|█████▏    | 50/97 [00:08<00:07,  6.15it/s, loss=4.47, v_num=0]
Epoch 13:  52%|█████▏    | 50/97 [00:08<00:07,  6.11it/s, loss=4.48, v_num=0]
Epoch 13:  53%|█████▎    | 51/97 [00:08<00:07,  6.17it/s, loss=4.48, v_num=0]
Epoch 13:  53%|█████▎    | 51/97 [00:08<00:07,  6.13it/s, loss=4.47, v_num=0]
Epoch 13:  54%|█████▎    | 52/97 [00:08<00:07,  6.19it/s, loss=4.47, v_num=0]
Epoch 13:  54%|█████▎    | 52/97 [00:08<00:07,  6.15it/s, loss=4.47, v_num=0]
Epoch 13:  55%|█████▍    | 53/97 [00:08<00:07,  6.20it/s, loss=4.47, v_num=0]
Epoch 13:  55%|█████▍    | 53/97 [00:08<00:07,  6.16it/s, loss=4.46, v_num=0]
Epoch 13:  56%|█████▌    | 54/97 [00:08<00:06,  6.22it/s, loss=4.46, v_num=0]
Epoch 13:  56%|█████▌    | 54/97 [00:08<00:06,  6.18it/s, loss=4.45, v_num=0]
Epoch 13:  57%|█████▋    | 55/97 [00:08<00:06,  6.24it/s, loss=4.45, v_num=0]
Epoch 13:  57%|█████▋    | 55/97 [00:08<00:06,  6.20it/s, loss=4.45, v_num=0]
Epoch 13:  58%|█████▊    | 56/97 [00:08<00:06,  6.25it/s, loss=4.45, v_num=0]
Epoch 13:  58%|█████▊    | 56/97 [00:09<00:06,  6.21it/s, loss=4.45, v_num=0]
Epoch 13:  59%|█████▉    | 57/97 [00:09<00:06,  6.27it/s, loss=4.45, v_num=0]
Epoch 13:  59%|█████▉    | 57/97 [00:09<00:06,  6.23it/s, loss=4.44, v_num=0]
Epoch 13:  60%|█████▉    | 58/97 [00:09<00:06,  6.28it/s, loss=4.44, v_num=0]
Epoch 13:  60%|█████▉    | 58/97 [00:09<00:06,  6.24it/s, loss=4.44, v_num=0]
Epoch 13:  61%|██████    | 59/97 [00:09<00:06,  6.30it/s, loss=4.44, v_num=0]
Epoch 13:  61%|██████    | 59/97 [00:09<00:06,  6.26it/s, loss=4.44, v_num=0]
Epoch 13:  62%|██████▏   | 60/97 [00:09<00:05,  6.31it/s, loss=4.44, v_num=0]
Epoch 13:  62%|██████▏   | 60/97 [00:09<00:05,  6.27it/s, loss=4.44, v_num=0]
Epoch 13:  63%|██████▎   | 61/97 [00:09<00:05,  6.33it/s, loss=4.44, v_num=0]
Epoch 13:  63%|██████▎   | 61/97 [00:09<00:05,  6.29it/s, loss=4.44, v_num=0]
Epoch 13:  64%|██████▍   | 62/97 [00:09<00:05,  6.34it/s, loss=4.44, v_num=0]
Epoch 13:  64%|██████▍   | 62/97 [00:09<00:05,  6.30it/s, loss=4.43, v_num=0]
Epoch 13:  65%|██████▍   | 63/97 [00:09<00:05,  6.35it/s, loss=4.43, v_num=0]
Epoch 13:  65%|██████▍   | 63/97 [00:09<00:05,  6.32it/s, loss=4.42, v_num=0]
Epoch 13:  66%|██████▌   | 64/97 [00:10<00:05,  6.36it/s, loss=4.42, v_num=0]
Epoch 13:  66%|██████▌   | 64/97 [00:10<00:05,  6.33it/s, loss=4.42, v_num=0]
Epoch 13:  67%|██████▋   | 65/97 [00:10<00:05,  6.38it/s, loss=4.42, v_num=0]
Epoch 13:  67%|██████▋   | 65/97 [00:10<00:05,  6.34it/s, loss=4.42, v_num=0]
Epoch 13:  68%|██████▊   | 66/97 [00:10<00:04,  6.39it/s, loss=4.42, v_num=0]
Epoch 13:  68%|██████▊   | 66/97 [00:10<00:04,  6.36it/s, loss=4.43, v_num=0]
Epoch 13:  69%|██████▉   | 67/97 [00:10<00:04,  6.40it/s, loss=4.43, v_num=0]
Epoch 13:  69%|██████▉   | 67/97 [00:10<00:04,  6.37it/s, loss=4.43, v_num=0]
Epoch 13:  70%|███████   | 68/97 [00:10<00:04,  6.41it/s, loss=4.43, v_num=0]
Epoch 13:  70%|███████   | 68/97 [00:10<00:04,  6.38it/s, loss=4.43, v_num=0]
Epoch 13:  71%|███████   | 69/97 [00:10<00:04,  6.42it/s, loss=4.43, v_num=0]
Epoch 13:  71%|███████   | 69/97 [00:10<00:04,  6.39it/s, loss=4.43, v_num=0]
Epoch 13:  72%|███████▏  | 70/97 [00:10<00:04,  6.43it/s, loss=4.43, v_num=0]
Epoch 13:  72%|███████▏  | 70/97 [00:10<00:04,  6.40it/s, loss=4.42, v_num=0]
Epoch 13:  73%|███████▎  | 71/97 [00:11<00:04,  6.45it/s, loss=4.42, v_num=0]
Epoch 13:  73%|███████▎  | 71/97 [00:11<00:04,  6.41it/s, loss=4.42, v_num=0]
Epoch 13:  74%|███████▍  | 72/97 [00:11<00:03,  6.45it/s, loss=4.42, v_num=0]
Epoch 13:  74%|███████▍  | 72/97 [00:11<00:03,  6.42it/s, loss=4.42, v_num=0]
Epoch 13:  75%|███████▌  | 73/97 [00:11<00:03,  6.47it/s, loss=4.42, v_num=0]
Epoch 13:  75%|███████▌  | 73/97 [00:11<00:03,  6.43it/s, loss=4.42, v_num=0]
Epoch 13:  76%|███████▋  | 74/97 [00:11<00:03,  6.47it/s, loss=4.42, v_num=0]
Epoch 13:  76%|███████▋  | 74/97 [00:11<00:03,  6.44it/s, loss=4.42, v_num=0]
Epoch 13:  77%|███████▋  | 75/97 [00:11<00:03,  6.48it/s, loss=4.42, v_num=0]
Epoch 13:  77%|███████▋  | 75/97 [00:11<00:03,  6.45it/s, loss=4.43, v_num=0]
Epoch 13:  78%|███████▊  | 76/97 [00:11<00:03,  6.49it/s, loss=4.43, v_num=0]
Epoch 13:  78%|███████▊  | 76/97 [00:11<00:03,  6.46it/s, loss=4.42, v_num=0]
Epoch 13:  79%|███████▉  | 77/97 [00:11<00:03,  6.50it/s, loss=4.42, v_num=0]
Epoch 13:  79%|███████▉  | 77/97 [00:11<00:03,  6.47it/s, loss=4.42, v_num=0]
Epoch 13:  80%|████████  | 78/97 [00:11<00:02,  6.51it/s, loss=4.42, v_num=0]
Epoch 13:  80%|████████  | 78/97 [00:12<00:02,  6.48it/s, loss=4.42, v_num=0]
Epoch 13:  81%|████████▏ | 79/97 [00:12<00:02,  6.52it/s, loss=4.42, v_num=0]
Epoch 13:  81%|████████▏ | 79/97 [00:12<00:02,  6.49it/s, loss=4.42, v_num=0]
Epoch 13:  82%|████████▏ | 80/97 [00:12<00:02,  6.53it/s, loss=4.42, v_num=0]
Epoch 13:  82%|████████▏ | 80/97 [00:12<00:02,  6.50it/s, loss=4.42, v_num=0]
Epoch 13:  84%|████████▎ | 81/97 [00:12<00:02,  6.54it/s, loss=4.42, v_num=0]
Epoch 13:  84%|████████▎ | 81/97 [00:12<00:02,  6.51it/s, loss=4.41, v_num=0]
Epoch 13:  85%|████████▍ | 82/97 [00:12<00:02,  6.54it/s, loss=4.41, v_num=0]
Epoch 13:  85%|████████▍ | 82/97 [00:12<00:02,  6.51it/s, loss=4.42, v_num=0]
Epoch 13:  86%|████████▌ | 83/97 [00:12<00:02,  6.55it/s, loss=4.42, v_num=0]
Epoch 13:  86%|████████▌ | 83/97 [00:12<00:02,  6.52it/s, loss=4.43, v_num=0]
Epoch 13:  87%|████████▋ | 84/97 [00:12<00:01,  6.56it/s, loss=4.43, v_num=0]
Epoch 13:  87%|████████▋ | 84/97 [00:12<00:01,  6.53it/s, loss=4.43, v_num=0]
Epoch 13:  88%|████████▊ | 85/97 [00:12<00:01,  6.57it/s, loss=4.43, v_num=0]
Epoch 13:  88%|████████▊ | 85/97 [00:13<00:01,  6.54it/s, loss=4.42, v_num=0]
Epoch 13:  89%|████████▊ | 86/97 [00:13<00:01,  6.57it/s, loss=4.42, v_num=0]
Epoch 13:  89%|████████▊ | 86/97 [00:13<00:01,  6.55it/s, loss=4.41, v_num=0]
Epoch 13:  90%|████████▉ | 87/97 [00:13<00:01,  6.58it/s, loss=4.41, v_num=0]
Epoch 13:  90%|████████▉ | 87/97 [00:13<00:01,  6.55it/s, loss=4.41, v_num=0]
Epoch 13:  91%|█████████ | 88/97 [00:13<00:01,  6.59it/s, loss=4.41, v_num=0]
Epoch 13:  91%|█████████ | 88/97 [00:13<00:01,  6.56it/s, loss=4.41, v_num=0]
Epoch 13:  92%|█████████▏| 89/97 [00:13<00:01,  6.60it/s, loss=4.41, v_num=0]
Epoch 13:  92%|█████████▏| 89/97 [00:13<00:01,  6.57it/s, loss=4.41, v_num=0]
Epoch 13:  93%|█████████▎| 90/97 [00:13<00:01,  6.60it/s, loss=4.41, v_num=0]
Epoch 13:  93%|█████████▎| 90/97 [00:13<00:01,  6.58it/s, loss=4.41, v_num=0]
Epoch 13:  94%|█████████▍| 91/97 [00:13<00:00,  6.61it/s, loss=4.41, v_num=0]
Epoch 13:  94%|█████████▍| 91/97 [00:13<00:00,  6.58it/s, loss=4.41, v_num=0]
Epoch 13:  95%|█████████▍| 92/97 [00:13<00:00,  6.62it/s, loss=4.41, v_num=0]
Epoch 13:  95%|█████████▍| 92/97 [00:13<00:00,  6.59it/s, loss=4.42, v_num=0]
Epoch 13:  96%|█████████▌| 93/97 [00:14<00:00,  6.62it/s, loss=4.42, v_num=0]
Epoch 13:  96%|█████████▌| 93/97 [00:14<00:00,  6.60it/s, loss=4.43, v_num=0]
Epoch 13:  97%|█████████▋| 94/97 [00:14<00:00,  6.63it/s, loss=4.43, v_num=0]
Epoch 13:  97%|█████████▋| 94/97 [00:14<00:00,  6.60it/s, loss=4.43, v_num=0]
Epoch 13:  98%|█████████▊| 95/97 [00:14<00:00,  6.64it/s, loss=4.43, v_num=0]
Epoch 13:  98%|█████████▊| 95/97 [00:14<00:00,  6.61it/s, loss=4.43, v_num=0]
Epoch 13:  99%|█████████▉| 96/97 [00:14<00:00,  6.64it/s, loss=4.43, v_num=0]
Epoch 13:  99%|█████████▉| 96/97 [00:14<00:00,  6.62it/s, loss=4.43, v_num=0]
Epoch 13: 100%|██████████| 97/97 [00:14<00:00,  6.65it/s, loss=4.43, v_num=0]
Epoch 13: 100%|██████████| 97/97 [00:14<00:00,  6.62it/s, loss=4.43, v_num=0]
Epoch 13: 100%|██████████| 97/97 [00:14<00:00,  6.62it/s, loss=4.43, v_num=0]
Epoch 13:   0%|          | 0/97 [00:00<?, ?it/s, loss=4.43, v_num=0]
Epoch 14:   0%|          | 0/97 [00:00<?, ?it/s, loss=4.43, v_num=0]
Epoch 14:   1%|          | 1/97 [00:01<02:12,  1.38s/it, loss=4.43, v_num=0]
Epoch 14:   1%|          | 1/97 [00:01<02:17,  1.44s/it, loss=4.43, v_num=0]
Epoch 14:   2%|▏         | 2/97 [00:01<01:12,  1.31it/s, loss=4.43, v_num=0]
Epoch 14:   2%|▏         | 2/97 [00:01<01:15,  1.27it/s, loss=4.43, v_num=0]
Epoch 14:   3%|▎         | 3/97 [00:01<00:52,  1.81it/s, loss=4.43, v_num=0]
Epoch 14:   3%|▎         | 3/97 [00:01<00:53,  1.75it/s, loss=4.42, v_num=0]
Epoch 14:   4%|▍         | 4/97 [00:01<00:41,  2.22it/s, loss=4.42, v_num=0]
Epoch 14:   4%|▍         | 4/97 [00:01<00:43,  2.16it/s, loss=4.42, v_num=0]
Epoch 14:   5%|▌         | 5/97 [00:01<00:35,  2.58it/s, loss=4.42, v_num=0]
Epoch 14:   5%|▌         | 5/97 [00:01<00:36,  2.51it/s, loss=4.41, v_num=0]
Epoch 14:   6%|▌         | 6/97 [00:02<00:31,  2.89it/s, loss=4.41, v_num=0]
Epoch 14:   6%|▌         | 6/97 [00:02<00:32,  2.82it/s, loss=4.4, v_num=0]
Epoch 14:   7%|▋         | 7/97 [00:02<00:28,  3.17it/s, loss=4.4, v_num=0]
Epoch 14:   7%|▋         | 7/97 [00:02<00:29,  3.09it/s, loss=4.4, v_num=0]
Epoch 14:   8%|▊         | 8/97 [00:02<00:26,  3.41it/s, loss=4.4, v_num=0]
Epoch 14:   8%|▊         | 8/97 [00:02<00:26,  3.33it/s, loss=4.4, v_num=0]
Epoch 14:   9%|▉         | 9/97 [00:02<00:24,  3.61it/s, loss=4.4, v_num=0]
Epoch 14:   9%|▉         | 9/97 [00:02<00:24,  3.54it/s, loss=4.4, v_num=0]
Epoch 14:  10%|█         | 10/97 [00:02<00:22,  3.81it/s, loss=4.4, v_num=0]
Epoch 14:  10%|█         | 10/97 [00:02<00:23,  3.73it/s, loss=4.41, v_num=0]
Epoch 14:  11%|█▏        | 11/97 [00:02<00:21,  3.98it/s, loss=4.41, v_num=0]
Epoch 14:  11%|█▏        | 11/97 [00:02<00:22,  3.90it/s, loss=4.41, v_num=0]
Epoch 14:  12%|█▏        | 12/97 [00:02<00:20,  4.14it/s, loss=4.41, v_num=0]
Epoch 14:  12%|█▏        | 12/97 [00:02<00:20,  4.06it/s, loss=4.41, v_num=0]
Epoch 14:  13%|█▎        | 13/97 [00:03<00:19,  4.28it/s, loss=4.41, v_num=0]
Epoch 14:  13%|█▎        | 13/97 [00:03<00:19,  4.20it/s, loss=4.4, v_num=0]
Epoch 14:  14%|█▍        | 14/97 [00:03<00:18,  4.42it/s, loss=4.4, v_num=0]
Epoch 14:  14%|█▍        | 14/97 [00:03<00:19,  4.34it/s, loss=4.4, v_num=0]
Epoch 14:  15%|█▌        | 15/97 [00:03<00:18,  4.53it/s, loss=4.4, v_num=0]
Epoch 14:  15%|█▌        | 15/97 [00:03<00:18,  4.46it/s, loss=4.39, v_num=0]
Epoch 14:  16%|█▋        | 16/97 [00:03<00:17,  4.65it/s, loss=4.39, v_num=0]
Epoch 14:  16%|█▋        | 16/97 [00:03<00:17,  4.57it/s, loss=4.38, v_num=0]
Epoch 14:  18%|█▊        | 17/97 [00:03<00:16,  4.74it/s, loss=4.38, v_num=0]
Epoch 14:  18%|█▊        | 17/97 [00:03<00:17,  4.67it/s, loss=4.38, v_num=0]
Epoch 14:  19%|█▊        | 18/97 [00:03<00:16,  4.84it/s, loss=4.38, v_num=0]
Epoch 14:  19%|█▊        | 18/97 [00:03<00:16,  4.77it/s, loss=4.37, v_num=0]
Epoch 14:  20%|█▉        | 19/97 [00:03<00:15,  4.93it/s, loss=4.37, v_num=0]
Epoch 14:  20%|█▉        | 19/97 [00:03<00:16,  4.86it/s, loss=4.37, v_num=0]
Epoch 14:  21%|██        | 20/97 [00:03<00:15,  5.01it/s, loss=4.37, v_num=0]
Epoch 14:  21%|██        | 20/97 [00:04<00:15,  4.94it/s, loss=4.36, v_num=0]
Epoch 14:  22%|██▏       | 21/97 [00:04<00:14,  5.08it/s, loss=4.36, v_num=0]
Epoch 14:  22%|██▏       | 21/97 [00:04<00:15,  5.01it/s, loss=4.36, v_num=0]
Epoch 14:  23%|██▎       | 22/97 [00:04<00:14,  5.16it/s, loss=4.36, v_num=0]
Epoch 14:  23%|██▎       | 22/97 [00:04<00:14,  5.09it/s, loss=4.36, v_num=0]
Epoch 14:  24%|██▎       | 23/97 [00:04<00:14,  5.22it/s, loss=4.36, v_num=0]
Epoch 14:  24%|██▎       | 23/97 [00:04<00:14,  5.15it/s, loss=4.36, v_num=0]
Epoch 14:  25%|██▍       | 24/97 [00:04<00:13,  5.28it/s, loss=4.36, v_num=0]
Epoch 14:  25%|██▍       | 24/97 [00:04<00:13,  5.22it/s, loss=4.36, v_num=0]
Epoch 14:  26%|██▌       | 25/97 [00:04<00:13,  5.34it/s, loss=4.36, v_num=0]
Epoch 14:  26%|██▌       | 25/97 [00:04<00:13,  5.28it/s, loss=4.36, v_num=0]
Epoch 14:  27%|██▋       | 26/97 [00:04<00:13,  5.39it/s, loss=4.36, v_num=0]
Epoch 14:  27%|██▋       | 26/97 [00:04<00:13,  5.34it/s, loss=4.36, v_num=0]
Epoch 14:  28%|██▊       | 27/97 [00:04<00:12,  5.45it/s, loss=4.36, v_num=0]
Epoch 14:  28%|██▊       | 27/97 [00:05<00:12,  5.39it/s, loss=4.35, v_num=0]
Epoch 14:  29%|██▉       | 28/97 [00:05<00:12,  5.50it/s, loss=4.35, v_num=0]
Epoch 14:  29%|██▉       | 28/97 [00:05<00:12,  5.44it/s, loss=4.34, v_num=0]
Epoch 14:  30%|██▉       | 29/97 [00:05<00:12,  5.55it/s, loss=4.34, v_num=0]
Epoch 14:  30%|██▉       | 29/97 [00:05<00:12,  5.49it/s, loss=4.34, v_num=0]
Epoch 14:  31%|███       | 30/97 [00:05<00:11,  5.59it/s, loss=4.34, v_num=0]
Epoch 14:  31%|███       | 30/97 [00:05<00:12,  5.53it/s, loss=4.33, v_num=0]
Epoch 14:  32%|███▏      | 31/97 [00:05<00:11,  5.63it/s, loss=4.33, v_num=0]
Epoch 14:  32%|███▏      | 31/97 [00:05<00:11,  5.58it/s, loss=4.33, v_num=0]
Epoch 14:  33%|███▎      | 32/97 [00:05<00:11,  5.68it/s, loss=4.33, v_num=0]
Epoch 14:  33%|███▎      | 32/97 [00:05<00:11,  5.62it/s, loss=4.33, v_num=0]
Epoch 14:  34%|███▍      | 33/97 [00:05<00:11,  5.71it/s, loss=4.33, v_num=0]
Epoch 14:  34%|███▍      | 33/97 [00:05<00:11,  5.66it/s, loss=4.34, v_num=0]
Epoch 14:  35%|███▌      | 34/97 [00:05<00:10,  5.75it/s, loss=4.34, v_num=0]
Epoch 14:  35%|███▌      | 34/97 [00:05<00:11,  5.70it/s, loss=4.34, v_num=0]
Epoch 14:  36%|███▌      | 35/97 [00:06<00:10,  5.78it/s, loss=4.34, v_num=0]
Epoch 14:  36%|███▌      | 35/97 [00:06<00:10,  5.73it/s, loss=4.34, v_num=0]
Epoch 14:  37%|███▋      | 36/97 [00:06<00:10,  5.82it/s, loss=4.34, v_num=0]
Epoch 14:  37%|███▋      | 36/97 [00:06<00:10,  5.77it/s, loss=4.35, v_num=0]
Epoch 14:  38%|███▊      | 37/97 [00:06<00:10,  5.85it/s, loss=4.35, v_num=0]
Epoch 14:  38%|███▊      | 37/97 [00:06<00:10,  5.80it/s, loss=4.35, v_num=0]
Epoch 14:  39%|███▉      | 38/97 [00:06<00:10,  5.88it/s, loss=4.35, v_num=0]
Epoch 14:  39%|███▉      | 38/97 [00:06<00:10,  5.83it/s, loss=4.35, v_num=0]
Epoch 14:  40%|████      | 39/97 [00:06<00:09,  5.91it/s, loss=4.35, v_num=0]
Epoch 14:  40%|████      | 39/97 [00:06<00:09,  5.86it/s, loss=4.34, v_num=0]
Epoch 14:  41%|████      | 40/97 [00:06<00:09,  5.94it/s, loss=4.34, v_num=0]
Epoch 14:  41%|████      | 40/97 [00:06<00:09,  5.89it/s, loss=4.34, v_num=0]
Epoch 14:  42%|████▏     | 41/97 [00:06<00:09,  5.97it/s, loss=4.34, v_num=0]
Epoch 14:  42%|████▏     | 41/97 [00:06<00:09,  5.92it/s, loss=4.34, v_num=0]
Epoch 14:  43%|████▎     | 42/97 [00:07<00:09,  5.99it/s, loss=4.34, v_num=0]
Epoch 14:  43%|████▎     | 42/97 [00:07<00:09,  5.95it/s, loss=4.34, v_num=0]
Epoch 14:  44%|████▍     | 43/97 [00:07<00:08,  6.01it/s, loss=4.34, v_num=0]
Epoch 14:  44%|████▍     | 43/97 [00:07<00:09,  5.97it/s, loss=4.34, v_num=0]
Epoch 14:  45%|████▌     | 44/97 [00:07<00:08,  6.04it/s, loss=4.34, v_num=0]
Epoch 14:  45%|████▌     | 44/97 [00:07<00:08,  5.99it/s, loss=4.34, v_num=0]
Epoch 14:  46%|████▋     | 45/97 [00:07<00:08,  6.06it/s, loss=4.34, v_num=0]
Epoch 14:  46%|████▋     | 45/97 [00:07<00:08,  6.02it/s, loss=4.34, v_num=0]
Epoch 14:  47%|████▋     | 46/97 [00:07<00:08,  6.08it/s, loss=4.34, v_num=0]
Epoch 14:  47%|████▋     | 46/97 [00:07<00:08,  6.04it/s, loss=4.34, v_num=0]
Epoch 14:  48%|████▊     | 47/97 [00:07<00:08,  6.11it/s, loss=4.34, v_num=0]
Epoch 14:  48%|████▊     | 47/97 [00:07<00:08,  6.06it/s, loss=4.34, v_num=0]
Epoch 14:  49%|████▉     | 48/97 [00:07<00:07,  6.13it/s, loss=4.34, v_num=0]
Epoch 14:  49%|████▉     | 48/97 [00:07<00:08,  6.08it/s, loss=4.34, v_num=0]
Epoch 14:  51%|█████     | 49/97 [00:07<00:07,  6.15it/s, loss=4.34, v_num=0]
Epoch 14:  51%|█████     | 49/97 [00:08<00:07,  6.10it/s, loss=4.35, v_num=0]
Epoch 14:  52%|█████▏    | 50/97 [00:08<00:07,  6.17it/s, loss=4.35, v_num=0]
Epoch 14:  52%|█████▏    | 50/97 [00:08<00:07,  6.12it/s, loss=4.34, v_num=0]
Epoch 14:  53%|█████▎    | 51/97 [00:08<00:07,  6.18it/s, loss=4.34, v_num=0]
Epoch 14:  53%|█████▎    | 51/97 [00:08<00:07,  6.14it/s, loss=4.34, v_num=0]
Epoch 14:  54%|█████▎    | 52/97 [00:08<00:07,  6.20it/s, loss=4.34, v_num=0]
Epoch 14:  54%|█████▎    | 52/97 [00:08<00:07,  6.16it/s, loss=4.34, v_num=0]
Epoch 14:  55%|█████▍    | 53/97 [00:08<00:07,  6.22it/s, loss=4.34, v_num=0]
Epoch 14:  55%|█████▍    | 53/97 [00:08<00:07,  6.18it/s, loss=4.34, v_num=0]
Epoch 14:  56%|█████▌    | 54/97 [00:08<00:06,  6.24it/s, loss=4.34, v_num=0]
Epoch 14:  56%|█████▌    | 54/97 [00:08<00:06,  6.20it/s, loss=4.33, v_num=0]
Epoch 14:  57%|█████▋    | 55/97 [00:08<00:06,  6.25it/s, loss=4.33, v_num=0]
Epoch 14:  57%|█████▋    | 55/97 [00:08<00:06,  6.21it/s, loss=4.34, v_num=0]
Epoch 14:  58%|█████▊    | 56/97 [00:08<00:06,  6.27it/s, loss=4.34, v_num=0]
Epoch 14:  58%|█████▊    | 56/97 [00:08<00:06,  6.23it/s, loss=4.32, v_num=0]
Epoch 14:  59%|█████▉    | 57/97 [00:09<00:06,  6.28it/s, loss=4.32, v_num=0]
Epoch 14:  59%|█████▉    | 57/97 [00:09<00:06,  6.25it/s, loss=4.32, v_num=0]
Epoch 14:  60%|█████▉    | 58/97 [00:09<00:06,  6.30it/s, loss=4.32, v_num=0]
Epoch 14:  60%|█████▉    | 58/97 [00:09<00:06,  6.26it/s, loss=4.32, v_num=0]
Epoch 14:  61%|██████    | 59/97 [00:09<00:06,  6.32it/s, loss=4.32, v_num=0]
Epoch 14:  61%|██████    | 59/97 [00:09<00:06,  6.28it/s, loss=4.33, v_num=0]
Epoch 14:  62%|██████▏   | 60/97 [00:09<00:05,  6.33it/s, loss=4.33, v_num=0]
Epoch 14:  62%|██████▏   | 60/97 [00:09<00:05,  6.29it/s, loss=4.33, v_num=0]
Epoch 14:  63%|██████▎   | 61/97 [00:09<00:05,  6.34it/s, loss=4.33, v_num=0]
Epoch 14:  63%|██████▎   | 61/97 [00:09<00:05,  6.30it/s, loss=4.32, v_num=0]
Epoch 14:  64%|██████▍   | 62/97 [00:09<00:05,  6.35it/s, loss=4.32, v_num=0]
Epoch 14:  64%|██████▍   | 62/97 [00:09<00:05,  6.32it/s, loss=4.32, v_num=0]
Epoch 14:  65%|██████▍   | 63/97 [00:09<00:05,  6.36it/s, loss=4.32, v_num=0]
Epoch 14:  65%|██████▍   | 63/97 [00:09<00:05,  6.33it/s, loss=4.32, v_num=0]
Epoch 14:  66%|██████▌   | 64/97 [00:10<00:05,  6.38it/s, loss=4.32, v_num=0]
Epoch 14:  66%|██████▌   | 64/97 [00:10<00:05,  6.34it/s, loss=4.32, v_num=0]
Epoch 14:  67%|██████▋   | 65/97 [00:10<00:05,  6.39it/s, loss=4.32, v_num=0]
Epoch 14:  67%|██████▋   | 65/97 [00:10<00:05,  6.35it/s, loss=4.32, v_num=0]
Epoch 14:  68%|██████▊   | 66/97 [00:10<00:04,  6.40it/s, loss=4.32, v_num=0]
Epoch 14:  68%|██████▊   | 66/97 [00:10<00:04,  6.36it/s, loss=4.31, v_num=0]
Epoch 14:  69%|██████▉   | 67/97 [00:10<00:04,  6.41it/s, loss=4.31, v_num=0]
Epoch 14:  69%|██████▉   | 67/97 [00:10<00:04,  6.38it/s, loss=4.31, v_num=0]
Epoch 14:  70%|███████   | 68/97 [00:10<00:04,  6.42it/s, loss=4.31, v_num=0]
Epoch 14:  70%|███████   | 68/97 [00:10<00:04,  6.39it/s, loss=4.31, v_num=0]
Epoch 14:  71%|███████   | 69/97 [00:10<00:04,  6.43it/s, loss=4.31, v_num=0]
Epoch 14:  71%|███████   | 69/97 [00:10<00:04,  6.40it/s, loss=4.31, v_num=0]
Epoch 14:  72%|███████▏  | 70/97 [00:10<00:04,  6.44it/s, loss=4.31, v_num=0]
Epoch 14:  72%|███████▏  | 70/97 [00:10<00:04,  6.41it/s, loss=4.3, v_num=0]
Epoch 14:  73%|███████▎  | 71/97 [00:10<00:04,  6.45it/s, loss=4.3, v_num=0]
Epoch 14:  73%|███████▎  | 71/97 [00:11<00:04,  6.42it/s, loss=4.3, v_num=0]
Epoch 14:  74%|███████▍  | 72/97 [00:11<00:03,  6.47it/s, loss=4.3, v_num=0]
Epoch 14:  74%|███████▍  | 72/97 [00:11<00:03,  6.43it/s, loss=4.29, v_num=0]
Epoch 14:  75%|███████▌  | 73/97 [00:11<00:03,  6.48it/s, loss=4.29, v_num=0]
Epoch 14:  75%|███████▌  | 73/97 [00:11<00:03,  6.44it/s, loss=4.29, v_num=0]
Epoch 14:  76%|███████▋  | 74/97 [00:11<00:03,  6.49it/s, loss=4.29, v_num=0]
Epoch 14:  76%|███████▋  | 74/97 [00:11<00:03,  6.45it/s, loss=4.29, v_num=0]
Epoch 14:  77%|███████▋  | 75/97 [00:11<00:03,  6.50it/s, loss=4.29, v_num=0]
Epoch 14:  77%|███████▋  | 75/97 [00:11<00:03,  6.46it/s, loss=4.29, v_num=0]
Epoch 14:  78%|███████▊  | 76/97 [00:11<00:03,  6.51it/s, loss=4.29, v_num=0]
Epoch 14:  78%|███████▊  | 76/97 [00:11<00:03,  6.47it/s, loss=4.29, v_num=0]
Epoch 14:  79%|███████▉  | 77/97 [00:11<00:03,  6.51it/s, loss=4.29, v_num=0]
Epoch 14:  79%|███████▉  | 77/97 [00:11<00:03,  6.48it/s, loss=4.28, v_num=0]
Epoch 14:  80%|████████  | 78/97 [00:11<00:02,  6.52it/s, loss=4.28, v_num=0]
Epoch 14:  80%|████████  | 78/97 [00:12<00:02,  6.49it/s, loss=4.28, v_num=0]
Epoch 14:  81%|████████▏ | 79/97 [00:12<00:02,  6.53it/s, loss=4.28, v_num=0]
Epoch 14:  81%|████████▏ | 79/97 [00:12<00:02,  6.50it/s, loss=4.28, v_num=0]
Epoch 14:  82%|████████▏ | 80/97 [00:12<00:02,  6.54it/s, loss=4.28, v_num=0]
Epoch 14:  82%|████████▏ | 80/97 [00:12<00:02,  6.51it/s, loss=4.28, v_num=0]
Epoch 14:  84%|████████▎ | 81/97 [00:12<00:02,  6.55it/s, loss=4.28, v_num=0]
Epoch 14:  84%|████████▎ | 81/97 [00:12<00:02,  6.52it/s, loss=4.28, v_num=0]
Epoch 14:  85%|████████▍ | 82/97 [00:12<00:02,  6.56it/s, loss=4.28, v_num=0]
Epoch 14:  85%|████████▍ | 82/97 [00:12<00:02,  6.53it/s, loss=4.28, v_num=0]
Epoch 14:  86%|████████▌ | 83/97 [00:12<00:02,  6.56it/s, loss=4.28, v_num=0]
Epoch 14:  86%|████████▌ | 83/97 [00:12<00:02,  6.54it/s, loss=4.28, v_num=0]
Epoch 14:  87%|████████▋ | 84/97 [00:12<00:01,  6.57it/s, loss=4.28, v_num=0]
Epoch 14:  87%|████████▋ | 84/97 [00:12<00:01,  6.55it/s, loss=4.28, v_num=0]
Epoch 14:  88%|████████▊ | 85/97 [00:12<00:01,  6.58it/s, loss=4.28, v_num=0]
Epoch 14:  88%|████████▊ | 85/97 [00:12<00:01,  6.55it/s, loss=4.28, v_num=0]
Epoch 14:  89%|████████▊ | 86/97 [00:13<00:01,  6.59it/s, loss=4.28, v_num=0]
Epoch 14:  89%|████████▊ | 86/97 [00:13<00:01,  6.56it/s, loss=4.28, v_num=0]
Epoch 14:  90%|████████▉ | 87/97 [00:13<00:01,  6.60it/s, loss=4.28, v_num=0]
Epoch 14:  90%|████████▉ | 87/97 [00:13<00:01,  6.57it/s, loss=4.28, v_num=0]
Epoch 14:  91%|█████████ | 88/97 [00:13<00:01,  6.60it/s, loss=4.28, v_num=0]
Epoch 14:  91%|█████████ | 88/97 [00:13<00:01,  6.58it/s, loss=4.29, v_num=0]
Epoch 14:  92%|█████████▏| 89/97 [00:13<00:01,  6.61it/s, loss=4.29, v_num=0]
Epoch 14:  92%|█████████▏| 89/97 [00:13<00:01,  6.58it/s, loss=4.28, v_num=0]
Epoch 14:  93%|█████████▎| 90/97 [00:13<00:01,  6.62it/s, loss=4.28, v_num=0]
Epoch 14:  93%|█████████▎| 90/97 [00:13<00:01,  6.59it/s, loss=4.28, v_num=0]
Epoch 14:  94%|█████████▍| 91/97 [00:13<00:00,  6.63it/s, loss=4.28, v_num=0]
Epoch 14:  94%|█████████▍| 91/97 [00:13<00:00,  6.60it/s, loss=4.27, v_num=0]
Epoch 14:  95%|█████████▍| 92/97 [00:13<00:00,  6.63it/s, loss=4.27, v_num=0]
Epoch 14:  95%|█████████▍| 92/97 [00:13<00:00,  6.61it/s, loss=4.29, v_num=0]
Epoch 14:  96%|█████████▌| 93/97 [00:14<00:00,  6.64it/s, loss=4.29, v_num=0]
Epoch 14:  96%|█████████▌| 93/97 [00:14<00:00,  6.61it/s, loss=4.29, v_num=0]
Epoch 14:  97%|█████████▋| 94/97 [00:14<00:00,  6.65it/s, loss=4.29, v_num=0]
Epoch 14:  97%|█████████▋| 94/97 [00:14<00:00,  6.62it/s, loss=4.29, v_num=0]
Epoch 14:  98%|█████████▊| 95/97 [00:14<00:00,  6.65it/s, loss=4.29, v_num=0]
Epoch 14:  98%|█████████▊| 95/97 [00:14<00:00,  6.63it/s, loss=4.28, v_num=0]
Epoch 14:  99%|█████████▉| 96/97 [00:14<00:00,  6.66it/s, loss=4.28, v_num=0]
Epoch 14:  99%|█████████▉| 96/97 [00:14<00:00,  6.63it/s, loss=4.29, v_num=0]
Epoch 14: 100%|██████████| 97/97 [00:14<00:00,  6.66it/s, loss=4.29, v_num=0]
Epoch 14: 100%|██████████| 97/97 [00:14<00:00,  6.64it/s, loss=4.29, v_num=0]
Epoch 14: 100%|██████████| 97/97 [00:14<00:00,  6.64it/s, loss=4.29, v_num=0]
Epoch 14:   0%|          | 0/97 [00:00<?, ?it/s, loss=4.29, v_num=0]
Epoch 15:   0%|          | 0/97 [00:00<?, ?it/s, loss=4.29, v_num=0]
Epoch 15:   1%|          | 1/97 [00:01<02:12,  1.38s/it, loss=4.29, v_num=0]
Epoch 15:   1%|          | 1/97 [00:01<02:17,  1.43s/it, loss=4.29, v_num=0]
Epoch 15:   2%|▏         | 2/97 [00:01<01:12,  1.31it/s, loss=4.29, v_num=0]
Epoch 15:   2%|▏         | 2/97 [00:01<01:15,  1.26it/s, loss=4.29, v_num=0]
Epoch 15:   3%|▎         | 3/97 [00:01<00:52,  1.81it/s, loss=4.29, v_num=0]
Epoch 15:   3%|▎         | 3/97 [00:01<00:53,  1.75it/s, loss=4.28, v_num=0]
Epoch 15:   4%|▍         | 4/97 [00:01<00:41,  2.23it/s, loss=4.28, v_num=0]
Epoch 15:   4%|▍         | 4/97 [00:01<00:43,  2.16it/s, loss=4.27, v_num=0]
Epoch 15:   5%|▌         | 5/97 [00:01<00:35,  2.59it/s, loss=4.27, v_num=0]
Epoch 15:   5%|▌         | 5/97 [00:01<00:36,  2.51it/s, loss=4.27, v_num=0]
Epoch 15:   6%|▌         | 6/97 [00:02<00:31,  2.90it/s, loss=4.27, v_num=0]
Epoch 15:   6%|▌         | 6/97 [00:02<00:32,  2.82it/s, loss=4.26, v_num=0]
Epoch 15:   7%|▋         | 7/97 [00:02<00:28,  3.17it/s, loss=4.26, v_num=0]
Epoch 15:   7%|▋         | 7/97 [00:02<00:29,  3.09it/s, loss=4.26, v_num=0]
Epoch 15:   8%|▊         | 8/97 [00:02<00:26,  3.41it/s, loss=4.26, v_num=0]
Epoch 15:   8%|▊         | 8/97 [00:02<00:26,  3.33it/s, loss=4.26, v_num=0]
Epoch 15:   9%|▉         | 9/97 [00:02<00:24,  3.61it/s, loss=4.26, v_num=0]
Epoch 15:   9%|▉         | 9/97 [00:02<00:24,  3.53it/s, loss=4.27, v_num=0]
Epoch 15:  10%|█         | 10/97 [00:02<00:22,  3.80it/s, loss=4.27, v_num=0]
Epoch 15:  10%|█         | 10/97 [00:02<00:23,  3.72it/s, loss=4.27, v_num=0]
Epoch 15:  11%|█▏        | 11/97 [00:02<00:21,  3.97it/s, loss=4.27, v_num=0]
Epoch 15:  11%|█▏        | 11/97 [00:02<00:22,  3.89it/s, loss=4.26, v_num=0]
Epoch 15:  12%|█▏        | 12/97 [00:02<00:20,  4.13it/s, loss=4.26, v_num=0]
Epoch 15:  12%|█▏        | 12/97 [00:02<00:20,  4.05it/s, loss=4.26, v_num=0]
Epoch 15:  13%|█▎        | 13/97 [00:03<00:19,  4.27it/s, loss=4.26, v_num=0]
Epoch 15:  13%|█▎        | 13/97 [00:03<00:20,  4.20it/s, loss=4.26, v_num=0]
Epoch 15:  14%|█▍        | 14/97 [00:03<00:18,  4.40it/s, loss=4.26, v_num=0]
Epoch 15:  14%|█▍        | 14/97 [00:03<00:19,  4.32it/s, loss=4.26, v_num=0]
Epoch 15:  15%|█▌        | 15/97 [00:03<00:18,  4.52it/s, loss=4.26, v_num=0]
Epoch 15:  15%|█▌        | 15/97 [00:03<00:18,  4.45it/s, loss=4.25, v_num=0]
Epoch 15:  16%|█▋        | 16/97 [00:03<00:17,  4.64it/s, loss=4.25, v_num=0]
Epoch 15:  16%|█▋        | 16/97 [00:03<00:17,  4.56it/s, loss=4.24, v_num=0]
Epoch 15:  18%|█▊        | 17/97 [00:03<00:16,  4.73it/s, loss=4.24, v_num=0]
Epoch 15:  18%|█▊        | 17/97 [00:03<00:17,  4.66it/s, loss=4.24, v_num=0]
Epoch 15:  19%|█▊        | 18/97 [00:03<00:16,  4.82it/s, loss=4.24, v_num=0]
Epoch 15:  19%|█▊        | 18/97 [00:03<00:16,  4.75it/s, loss=4.24, v_num=0]
Epoch 15:  20%|█▉        | 19/97 [00:03<00:15,  4.91it/s, loss=4.24, v_num=0]
Epoch 15:  20%|█▉        | 19/97 [00:03<00:16,  4.84it/s, loss=4.23, v_num=0]
Epoch 15:  21%|██        | 20/97 [00:04<00:15,  5.00it/s, loss=4.23, v_num=0]
Epoch 15:  21%|██        | 20/97 [00:04<00:15,  4.93it/s, loss=4.23, v_num=0]
Epoch 15:  22%|██▏       | 21/97 [00:04<00:14,  5.08it/s, loss=4.23, v_num=0]
Epoch 15:  22%|██▏       | 21/97 [00:04<00:15,  5.01it/s, loss=4.22, v_num=0]
Epoch 15:  23%|██▎       | 22/97 [00:04<00:14,  5.15it/s, loss=4.22, v_num=0]
Epoch 15:  23%|██▎       | 22/97 [00:04<00:14,  5.08it/s, loss=4.23, v_num=0]
Epoch 15:  24%|██▎       | 23/97 [00:04<00:14,  5.22it/s, loss=4.23, v_num=0]
Epoch 15:  24%|██▎       | 23/97 [00:04<00:14,  5.15it/s, loss=4.22, v_num=0]
Epoch 15:  25%|██▍       | 24/97 [00:04<00:13,  5.28it/s, loss=4.22, v_num=0]
Epoch 15:  25%|██▍       | 24/97 [00:04<00:14,  5.21it/s, loss=4.22, v_num=0]
Epoch 15:  26%|██▌       | 25/97 [00:04<00:13,  5.34it/s, loss=4.22, v_num=0]
Epoch 15:  26%|██▌       | 25/97 [00:04<00:13,  5.27it/s, loss=4.22, v_num=0]
Epoch 15:  27%|██▋       | 26/97 [00:04<00:13,  5.40it/s, loss=4.22, v_num=0]
Epoch 15:  27%|██▋       | 26/97 [00:04<00:13,  5.33it/s, loss=4.22, v_num=0]
Epoch 15:  28%|██▊       | 27/97 [00:04<00:12,  5.45it/s, loss=4.22, v_num=0]
Epoch 15:  28%|██▊       | 27/97 [00:05<00:12,  5.39it/s, loss=4.21, v_num=0]
Epoch 15:  29%|██▉       | 28/97 [00:05<00:12,  5.50it/s, loss=4.21, v_num=0]
Epoch 15:  29%|██▉       | 28/97 [00:05<00:12,  5.44it/s, loss=4.22, v_num=0]
Epoch 15:  30%|██▉       | 29/97 [00:05<00:12,  5.55it/s, loss=4.22, v_num=0]
Epoch 15:  30%|██▉       | 29/97 [00:05<00:12,  5.49it/s, loss=4.21, v_num=0]
Epoch 15:  31%|███       | 30/97 [00:05<00:11,  5.59it/s, loss=4.21, v_num=0]
Epoch 15:  31%|███       | 30/97 [00:05<00:12,  5.53it/s, loss=4.21, v_num=0]
Epoch 15:  32%|███▏      | 31/97 [00:05<00:11,  5.63it/s, loss=4.21, v_num=0]
Epoch 15:  32%|███▏      | 31/97 [00:05<00:11,  5.58it/s, loss=4.21, v_num=0]
Epoch 15:  33%|███▎      | 32/97 [00:05<00:11,  5.66it/s, loss=4.21, v_num=0]
Epoch 15:  33%|███▎      | 32/97 [00:05<00:11,  5.61it/s, loss=4.21, v_num=0]
Epoch 15:  34%|███▍      | 33/97 [00:05<00:11,  5.70it/s, loss=4.21, v_num=0]
Epoch 15:  34%|███▍      | 33/97 [00:05<00:11,  5.65it/s, loss=4.21, v_num=0]
Epoch 15:  35%|███▌      | 34/97 [00:05<00:10,  5.74it/s, loss=4.21, v_num=0]
Epoch 15:  35%|███▌      | 34/97 [00:05<00:11,  5.69it/s, loss=4.21, v_num=0]
Epoch 15:  36%|███▌      | 35/97 [00:06<00:10,  5.78it/s, loss=4.21, v_num=0]
Epoch 15:  36%|███▌      | 35/97 [00:06<00:10,  5.72it/s, loss=4.21, v_num=0]
Epoch 15:  37%|███▋      | 36/97 [00:06<00:10,  5.81it/s, loss=4.21, v_num=0]
Epoch 15:  37%|███▋      | 36/97 [00:06<00:10,  5.76it/s, loss=4.21, v_num=0]
Epoch 15:  38%|███▊      | 37/97 [00:06<00:10,  5.84it/s, loss=4.21, v_num=0]
Epoch 15:  38%|███▊      | 37/97 [00:06<00:10,  5.79it/s, loss=4.2, v_num=0]
Epoch 15:  39%|███▉      | 38/97 [00:06<00:10,  5.87it/s, loss=4.2, v_num=0]
Epoch 15:  39%|███▉      | 38/97 [00:06<00:10,  5.82it/s, loss=4.21, v_num=0]
Epoch 15:  40%|████      | 39/97 [00:06<00:09,  5.90it/s, loss=4.21, v_num=0]
Epoch 15:  40%|████      | 39/97 [00:06<00:09,  5.85it/s, loss=4.21, v_num=0]
Epoch 15:  41%|████      | 40/97 [00:06<00:09,  5.93it/s, loss=4.21, v_num=0]
Epoch 15:  41%|████      | 40/97 [00:06<00:09,  5.88it/s, loss=4.21, v_num=0]
Epoch 15:  42%|████▏     | 41/97 [00:06<00:09,  5.96it/s, loss=4.21, v_num=0]
Epoch 15:  42%|████▏     | 41/97 [00:06<00:09,  5.91it/s, loss=4.21, v_num=0]
Epoch 15:  43%|████▎     | 42/97 [00:07<00:09,  5.98it/s, loss=4.21, v_num=0]
Epoch 15:  43%|████▎     | 42/97 [00:07<00:09,  5.94it/s, loss=4.2, v_num=0]
Epoch 15:  44%|████▍     | 43/97 [00:07<00:08,  6.00it/s, loss=4.2, v_num=0]
Epoch 15:  44%|████▍     | 43/97 [00:07<00:09,  5.96it/s, loss=4.21, v_num=0]
Epoch 15:  45%|████▌     | 44/97 [00:07<00:08,  6.03it/s, loss=4.21, v_num=0]
Epoch 15:  45%|████▌     | 44/97 [00:07<00:08,  5.98it/s, loss=4.2, v_num=0]
Epoch 15:  46%|████▋     | 45/97 [00:07<00:08,  6.05it/s, loss=4.2, v_num=0]
Epoch 15:  46%|████▋     | 45/97 [00:07<00:08,  6.01it/s, loss=4.21, v_num=0]
Epoch 15:  47%|████▋     | 46/97 [00:07<00:08,  6.08it/s, loss=4.21, v_num=0]
Epoch 15:  47%|████▋     | 46/97 [00:07<00:08,  6.03it/s, loss=4.21, v_num=0]
Epoch 15:  48%|████▊     | 47/97 [00:07<00:08,  6.09it/s, loss=4.21, v_num=0]
Epoch 15:  48%|████▊     | 47/97 [00:07<00:08,  6.05it/s, loss=4.21, v_num=0]
Epoch 15:  49%|████▉     | 48/97 [00:07<00:08,  6.12it/s, loss=4.21, v_num=0]
Epoch 15:  49%|████▉     | 48/97 [00:07<00:08,  6.07it/s, loss=4.2, v_num=0]
Epoch 15:  51%|█████     | 49/97 [00:07<00:07,  6.13it/s, loss=4.2, v_num=0]
Epoch 15:  51%|█████     | 49/97 [00:08<00:07,  6.09it/s, loss=4.21, v_num=0]
Epoch 15:  52%|█████▏    | 50/97 [00:08<00:07,  6.15it/s, loss=4.21, v_num=0]
Epoch 15:  52%|█████▏    | 50/97 [00:08<00:07,  6.11it/s, loss=4.21, v_num=0]
Epoch 15:  53%|█████▎    | 51/97 [00:08<00:07,  6.17it/s, loss=4.21, v_num=0]
Epoch 15:  53%|█████▎    | 51/97 [00:08<00:07,  6.13it/s, loss=4.21, v_num=0]
Epoch 15:  54%|█████▎    | 52/97 [00:08<00:07,  6.19it/s, loss=4.21, v_num=0]
Epoch 15:  54%|█████▎    | 52/97 [00:08<00:07,  6.15it/s, loss=4.21, v_num=0]
Epoch 15:  55%|█████▍    | 53/97 [00:08<00:07,  6.21it/s, loss=4.21, v_num=0]
Epoch 15:  55%|█████▍    | 53/97 [00:08<00:07,  6.17it/s, loss=4.21, v_num=0]
Epoch 15:  56%|█████▌    | 54/97 [00:08<00:06,  6.22it/s, loss=4.21, v_num=0]
Epoch 15:  56%|█████▌    | 54/97 [00:08<00:06,  6.18it/s, loss=4.2, v_num=0]
Epoch 15:  57%|█████▋    | 55/97 [00:08<00:06,  6.24it/s, loss=4.2, v_num=0]
Epoch 15:  57%|█████▋    | 55/97 [00:08<00:06,  6.20it/s, loss=4.2, v_num=0]
Epoch 15:  58%|█████▊    | 56/97 [00:08<00:06,  6.25it/s, loss=4.2, v_num=0]
Epoch 15:  58%|█████▊    | 56/97 [00:09<00:06,  6.22it/s, loss=4.19, v_num=0]
Epoch 15:  59%|█████▉    | 57/97 [00:09<00:06,  6.27it/s, loss=4.19, v_num=0]
Epoch 15:  59%|█████▉    | 57/97 [00:09<00:06,  6.23it/s, loss=4.19, v_num=0]
Epoch 15:  60%|█████▉    | 58/97 [00:09<00:06,  6.28it/s, loss=4.19, v_num=0]
Epoch 15:  60%|█████▉    | 58/97 [00:09<00:06,  6.25it/s, loss=4.18, v_num=0]
Epoch 15:  61%|██████    | 59/97 [00:09<00:06,  6.30it/s, loss=4.18, v_num=0]
Epoch 15:  61%|██████    | 59/97 [00:09<00:06,  6.26it/s, loss=4.18, v_num=0]
Epoch 15:  62%|██████▏   | 60/97 [00:09<00:05,  6.31it/s, loss=4.18, v_num=0]
Epoch 15:  62%|██████▏   | 60/97 [00:09<00:05,  6.28it/s, loss=4.19, v_num=0]
Epoch 15:  63%|██████▎   | 61/97 [00:09<00:05,  6.32it/s, loss=4.19, v_num=0]
Epoch 15:  63%|██████▎   | 61/97 [00:09<00:05,  6.29it/s, loss=4.19, v_num=0]
Epoch 15:  64%|██████▍   | 62/97 [00:09<00:05,  6.34it/s, loss=4.19, v_num=0]
Epoch 15:  64%|██████▍   | 62/97 [00:09<00:05,  6.30it/s, loss=4.19, v_num=0]
Epoch 15:  65%|██████▍   | 63/97 [00:09<00:05,  6.35it/s, loss=4.19, v_num=0]
Epoch 15:  65%|██████▍   | 63/97 [00:09<00:05,  6.31it/s, loss=4.17, v_num=0]
Epoch 15:  66%|██████▌   | 64/97 [00:10<00:05,  6.36it/s, loss=4.17, v_num=0]
Epoch 15:  66%|██████▌   | 64/97 [00:10<00:05,  6.33it/s, loss=4.17, v_num=0]
Epoch 15:  67%|██████▋   | 65/97 [00:10<00:05,  6.38it/s, loss=4.17, v_num=0]
Epoch 15:  67%|██████▋   | 65/97 [00:10<00:05,  6.34it/s, loss=4.17, v_num=0]
Epoch 15:  68%|██████▊   | 66/97 [00:10<00:04,  6.39it/s, loss=4.17, v_num=0]
Epoch 15:  68%|██████▊   | 66/97 [00:10<00:04,  6.35it/s, loss=4.17, v_num=0]
Epoch 15:  69%|██████▉   | 67/97 [00:10<00:04,  6.40it/s, loss=4.17, v_num=0]
Epoch 15:  69%|██████▉   | 67/97 [00:10<00:04,  6.37it/s, loss=4.17, v_num=0]
Epoch 15:  70%|███████   | 68/97 [00:10<00:04,  6.41it/s, loss=4.17, v_num=0]
Epoch 15:  70%|███████   | 68/97 [00:10<00:04,  6.38it/s, loss=4.17, v_num=0]
Epoch 15:  71%|███████   | 69/97 [00:10<00:04,  6.42it/s, loss=4.17, v_num=0]
Epoch 15:  71%|███████   | 69/97 [00:10<00:04,  6.39it/s, loss=4.17, v_num=0]
Epoch 15:  72%|███████▏  | 70/97 [00:10<00:04,  6.43it/s, loss=4.17, v_num=0]
Epoch 15:  72%|███████▏  | 70/97 [00:10<00:04,  6.40it/s, loss=4.15, v_num=0]
Epoch 15:  73%|███████▎  | 71/97 [00:11<00:04,  6.44it/s, loss=4.15, v_num=0]
Epoch 15:  73%|███████▎  | 71/97 [00:11<00:04,  6.41it/s, loss=4.15, v_num=0]
Epoch 15:  74%|███████▍  | 72/97 [00:11<00:03,  6.46it/s, loss=4.15, v_num=0]
Epoch 15:  74%|███████▍  | 72/97 [00:11<00:03,  6.42it/s, loss=4.14, v_num=0]
Epoch 15:  75%|███████▌  | 73/97 [00:11<00:03,  6.47it/s, loss=4.14, v_num=0]
Epoch 15:  75%|███████▌  | 73/97 [00:11<00:03,  6.43it/s, loss=4.14, v_num=0]
Epoch 15:  76%|███████▋  | 74/97 [00:11<00:03,  6.48it/s, loss=4.14, v_num=0]
Epoch 15:  76%|███████▋  | 74/97 [00:11<00:03,  6.44it/s, loss=4.14, v_num=0]
Epoch 15:  77%|███████▋  | 75/97 [00:11<00:03,  6.48it/s, loss=4.14, v_num=0]
Epoch 15:  77%|███████▋  | 75/97 [00:11<00:03,  6.45it/s, loss=4.14, v_num=0]
Epoch 15:  78%|███████▊  | 76/97 [00:11<00:03,  6.49it/s, loss=4.14, v_num=0]
Epoch 15:  78%|███████▊  | 76/97 [00:11<00:03,  6.46it/s, loss=4.14, v_num=0]
Epoch 15:  79%|███████▉  | 77/97 [00:11<00:03,  6.50it/s, loss=4.14, v_num=0]
Epoch 15:  79%|███████▉  | 77/97 [00:11<00:03,  6.47it/s, loss=4.15, v_num=0]
Epoch 15:  80%|████████  | 78/97 [00:11<00:02,  6.51it/s, loss=4.15, v_num=0]
Epoch 15:  80%|████████  | 78/97 [00:12<00:02,  6.48it/s, loss=4.15, v_num=0]
Epoch 15:  81%|████████▏ | 79/97 [00:12<00:02,  6.52it/s, loss=4.15, v_num=0]
Epoch 15:  81%|████████▏ | 79/97 [00:12<00:02,  6.49it/s, loss=4.14, v_num=0]
Epoch 15:  82%|████████▏ | 80/97 [00:12<00:02,  6.53it/s, loss=4.14, v_num=0]
Epoch 15:  82%|████████▏ | 80/97 [00:12<00:02,  6.50it/s, loss=4.14, v_num=0]
Epoch 15:  84%|████████▎ | 81/97 [00:12<00:02,  6.54it/s, loss=4.14, v_num=0]
Epoch 15:  84%|████████▎ | 81/97 [00:12<00:02,  6.51it/s, loss=4.14, v_num=0]
Epoch 15:  85%|████████▍ | 82/97 [00:12<00:02,  6.54it/s, loss=4.14, v_num=0]
Epoch 15:  85%|████████▍ | 82/97 [00:12<00:02,  6.51it/s, loss=4.13, v_num=0]
Epoch 15:  86%|████████▌ | 83/97 [00:12<00:02,  6.55it/s, loss=4.13, v_num=0]
Epoch 15:  86%|████████▌ | 83/97 [00:12<00:02,  6.52it/s, loss=4.15, v_num=0]
Epoch 15:  87%|████████▋ | 84/97 [00:12<00:01,  6.56it/s, loss=4.15, v_num=0]
Epoch 15:  87%|████████▋ | 84/97 [00:12<00:01,  6.53it/s, loss=4.15, v_num=0]
Epoch 15:  88%|████████▊ | 85/97 [00:12<00:01,  6.56it/s, loss=4.15, v_num=0]
Epoch 15:  88%|████████▊ | 85/97 [00:13<00:01,  6.53it/s, loss=4.14, v_num=0]
Epoch 15:  89%|████████▊ | 86/97 [00:13<00:01,  6.57it/s, loss=4.14, v_num=0]
Epoch 15:  89%|████████▊ | 86/97 [00:13<00:01,  6.54it/s, loss=4.13, v_num=0]
Epoch 15:  90%|████████▉ | 87/97 [00:13<00:01,  6.58it/s, loss=4.13, v_num=0]
Epoch 15:  90%|████████▉ | 87/97 [00:13<00:01,  6.55it/s, loss=4.13, v_num=0]
Epoch 15:  91%|█████████ | 88/97 [00:13<00:01,  6.59it/s, loss=4.13, v_num=0]
Epoch 15:  91%|█████████ | 88/97 [00:13<00:01,  6.56it/s, loss=4.13, v_num=0]
Epoch 15:  92%|█████████▏| 89/97 [00:13<00:01,  6.59it/s, loss=4.13, v_num=0]
Epoch 15:  92%|█████████▏| 89/97 [00:13<00:01,  6.57it/s, loss=4.13, v_num=0]
Epoch 15:  93%|█████████▎| 90/97 [00:13<00:01,  6.60it/s, loss=4.13, v_num=0]
Epoch 15:  93%|█████████▎| 90/97 [00:13<00:01,  6.57it/s, loss=4.14, v_num=0]
Epoch 15:  94%|█████████▍| 91/97 [00:13<00:00,  6.61it/s, loss=4.14, v_num=0]
Epoch 15:  94%|█████████▍| 91/97 [00:13<00:00,  6.58it/s, loss=4.14, v_num=0]
Epoch 15:  95%|█████████▍| 92/97 [00:13<00:00,  6.61it/s, loss=4.14, v_num=0]
Epoch 15:  95%|█████████▍| 92/97 [00:13<00:00,  6.58it/s, loss=4.14, v_num=0]
Epoch 15:  96%|█████████▌| 93/97 [00:14<00:00,  6.62it/s, loss=4.14, v_num=0]
Epoch 15:  96%|█████████▌| 93/97 [00:14<00:00,  6.59it/s, loss=4.15, v_num=0]
Epoch 15:  97%|█████████▋| 94/97 [00:14<00:00,  6.62it/s, loss=4.15, v_num=0]
Epoch 15:  97%|█████████▋| 94/97 [00:14<00:00,  6.60it/s, loss=4.14, v_num=0]
Epoch 15:  98%|█████████▊| 95/97 [00:14<00:00,  6.63it/s, loss=4.14, v_num=0]
Epoch 15:  98%|█████████▊| 95/97 [00:14<00:00,  6.61it/s, loss=4.14, v_num=0]
Epoch 15:  99%|█████████▉| 96/97 [00:14<00:00,  6.64it/s, loss=4.14, v_num=0]
Epoch 15:  99%|█████████▉| 96/97 [00:14<00:00,  6.61it/s, loss=4.14, v_num=0]
Epoch 15: 100%|██████████| 97/97 [00:14<00:00,  6.65it/s, loss=4.14, v_num=0]
Epoch 15: 100%|██████████| 97/97 [00:14<00:00,  6.62it/s, loss=4.14, v_num=0]
Epoch 15: 100%|██████████| 97/97 [00:14<00:00,  6.62it/s, loss=4.14, v_num=0]
Epoch 15:   0%|          | 0/97 [00:00<?, ?it/s, loss=4.14, v_num=0]
Epoch 16:   0%|          | 0/97 [00:00<?, ?it/s, loss=4.14, v_num=0]
Epoch 16:   1%|          | 1/97 [00:01<02:13,  1.39s/it, loss=4.14, v_num=0]
Epoch 16:   1%|          | 1/97 [00:01<02:18,  1.44s/it, loss=4.14, v_num=0]
Epoch 16:   2%|▏         | 2/97 [00:01<01:12,  1.30it/s, loss=4.14, v_num=0]
Epoch 16:   2%|▏         | 2/97 [00:01<01:15,  1.26it/s, loss=4.13, v_num=0]
Epoch 16:   3%|▎         | 3/97 [00:01<00:52,  1.79it/s, loss=4.13, v_num=0]
Epoch 16:   3%|▎         | 3/97 [00:01<00:54,  1.74it/s, loss=4.13, v_num=0]
Epoch 16:   4%|▍         | 4/97 [00:01<00:42,  2.21it/s, loss=4.13, v_num=0]
Epoch 16:   4%|▍         | 4/97 [00:01<00:43,  2.15it/s, loss=4.14, v_num=0]
Epoch 16:   5%|▌         | 5/97 [00:01<00:35,  2.57it/s, loss=4.14, v_num=0]
Epoch 16:   5%|▌         | 5/97 [00:02<00:36,  2.50it/s, loss=4.14, v_num=0]
Epoch 16:   6%|▌         | 6/97 [00:02<00:31,  2.89it/s, loss=4.14, v_num=0]
Epoch 16:   6%|▌         | 6/97 [00:02<00:32,  2.81it/s, loss=4.14, v_num=0]
Epoch 16:   7%|▋         | 7/97 [00:02<00:28,  3.16it/s, loss=4.14, v_num=0]
Epoch 16:   7%|▋         | 7/97 [00:02<00:29,  3.08it/s, loss=4.15, v_num=0]
Epoch 16:   8%|▊         | 8/97 [00:02<00:26,  3.40it/s, loss=4.15, v_num=0]
Epoch 16:   8%|▊         | 8/97 [00:02<00:26,  3.32it/s, loss=4.15, v_num=0]
Epoch 16:   9%|▉         | 9/97 [00:02<00:24,  3.60it/s, loss=4.15, v_num=0]
Epoch 16:   9%|▉         | 9/97 [00:02<00:24,  3.53it/s, loss=4.16, v_num=0]
Epoch 16:  10%|█         | 10/97 [00:02<00:22,  3.80it/s, loss=4.16, v_num=0]
Epoch 16:  10%|█         | 10/97 [00:02<00:23,  3.72it/s, loss=4.15, v_num=0]
Epoch 16:  11%|█▏        | 11/97 [00:02<00:21,  3.97it/s, loss=4.15, v_num=0]
Epoch 16:  11%|█▏        | 11/97 [00:02<00:22,  3.89it/s, loss=4.15, v_num=0]
Epoch 16:  12%|█▏        | 12/97 [00:02<00:20,  4.12it/s, loss=4.15, v_num=0]
Epoch 16:  12%|█▏        | 12/97 [00:02<00:21,  4.05it/s, loss=4.15, v_num=0]
Epoch 16:  13%|█▎        | 13/97 [00:03<00:19,  4.26it/s, loss=4.15, v_num=0]
Epoch 16:  13%|█▎        | 13/97 [00:03<00:20,  4.18it/s, loss=4.15, v_num=0]
Epoch 16:  14%|█▍        | 14/97 [00:03<00:18,  4.40it/s, loss=4.15, v_num=0]
Epoch 16:  14%|█▍        | 14/97 [00:03<00:19,  4.32it/s, loss=4.15, v_num=0]
Epoch 16:  15%|█▌        | 15/97 [00:03<00:18,  4.51it/s, loss=4.15, v_num=0]
Epoch 16:  15%|█▌        | 15/97 [00:03<00:18,  4.44it/s, loss=4.15, v_num=0]
Epoch 16:  16%|█▋        | 16/97 [00:03<00:17,  4.62it/s, loss=4.15, v_num=0]
Epoch 16:  16%|█▋        | 16/97 [00:03<00:17,  4.54it/s, loss=4.14, v_num=0]
Epoch 16:  18%|█▊        | 17/97 [00:03<00:16,  4.71it/s, loss=4.14, v_num=0]
Epoch 16:  18%|█▊        | 17/97 [00:03<00:17,  4.64it/s, loss=4.14, v_num=0]
Epoch 16:  19%|█▊        | 18/97 [00:03<00:16,  4.81it/s, loss=4.14, v_num=0]
Epoch 16:  19%|█▊        | 18/97 [00:03<00:16,  4.74it/s, loss=4.14, v_num=0]
Epoch 16:  20%|█▉        | 19/97 [00:03<00:15,  4.90it/s, loss=4.14, v_num=0]
Epoch 16:  20%|█▉        | 19/97 [00:03<00:16,  4.83it/s, loss=4.14, v_num=0]
Epoch 16:  21%|██        | 20/97 [00:04<00:15,  4.98it/s, loss=4.14, v_num=0]
Epoch 16:  21%|██        | 20/97 [00:04<00:15,  4.92it/s, loss=4.14, v_num=0]
Epoch 16:  22%|██▏       | 21/97 [00:04<00:15,  5.06it/s, loss=4.14, v_num=0]
Epoch 16:  22%|██▏       | 21/97 [00:04<00:15,  5.00it/s, loss=4.14, v_num=0]
Epoch 16:  23%|██▎       | 22/97 [00:04<00:14,  5.14it/s, loss=4.14, v_num=0]
Epoch 16:  23%|██▎       | 22/97 [00:04<00:14,  5.07it/s, loss=4.14, v_num=0]
Epoch 16:  24%|██▎       | 23/97 [00:04<00:14,  5.20it/s, loss=4.14, v_num=0]
Epoch 16:  24%|██▎       | 23/97 [00:04<00:14,  5.13it/s, loss=4.14, v_num=0]
Epoch 16:  25%|██▍       | 24/97 [00:04<00:13,  5.26it/s, loss=4.14, v_num=0]
Epoch 16:  25%|██▍       | 24/97 [00:04<00:14,  5.20it/s, loss=4.14, v_num=0]
Epoch 16:  26%|██▌       | 25/97 [00:04<00:13,  5.32it/s, loss=4.14, v_num=0]
Epoch 16:  26%|██▌       | 25/97 [00:04<00:13,  5.26it/s, loss=4.13, v_num=0]
Epoch 16:  27%|██▋       | 26/97 [00:04<00:13,  5.37it/s, loss=4.13, v_num=0]
Epoch 16:  27%|██▋       | 26/97 [00:04<00:13,  5.31it/s, loss=4.13, v_num=0]
Epoch 16:  28%|██▊       | 27/97 [00:04<00:12,  5.43it/s, loss=4.13, v_num=0]
Epoch 16:  28%|██▊       | 27/97 [00:05<00:13,  5.37it/s, loss=4.12, v_num=0]
Epoch 16:  29%|██▉       | 28/97 [00:05<00:12,  5.49it/s, loss=4.12, v_num=0]
Epoch 16:  29%|██▉       | 28/97 [00:05<00:12,  5.42it/s, loss=4.12, v_num=0]
Epoch 16:  30%|██▉       | 29/97 [00:05<00:12,  5.53it/s, loss=4.12, v_num=0]
Epoch 16:  30%|██▉       | 29/97 [00:05<00:12,  5.47it/s, loss=4.11, v_num=0]
Epoch 16:  31%|███       | 30/97 [00:05<00:12,  5.57it/s, loss=4.11, v_num=0]
Epoch 16:  31%|███       | 30/97 [00:05<00:12,  5.52it/s, loss=4.12, v_num=0]
Epoch 16:  32%|███▏      | 31/97 [00:05<00:11,  5.62it/s, loss=4.12, v_num=0]
Epoch 16:  32%|███▏      | 31/97 [00:05<00:11,  5.56it/s, loss=4.12, v_num=0]
Epoch 16:  33%|███▎      | 32/97 [00:05<00:11,  5.66it/s, loss=4.12, v_num=0]
Epoch 16:  33%|███▎      | 32/97 [00:05<00:11,  5.60it/s, loss=4.1, v_num=0]
Epoch 16:  34%|███▍      | 33/97 [00:05<00:11,  5.69it/s, loss=4.1, v_num=0]
Epoch 16:  34%|███▍      | 33/97 [00:05<00:11,  5.64it/s, loss=4.1, v_num=0]
Epoch 16:  35%|███▌      | 34/97 [00:05<00:10,  5.73it/s, loss=4.1, v_num=0]
Epoch 16:  35%|███▌      | 34/97 [00:05<00:11,  5.68it/s, loss=4.1, v_num=0]
Epoch 16:  36%|███▌      | 35/97 [00:06<00:10,  5.77it/s, loss=4.1, v_num=0]
Epoch 16:  36%|███▌      | 35/97 [00:06<00:10,  5.72it/s, loss=4.1, v_num=0]
Epoch 16:  37%|███▋      | 36/97 [00:06<00:10,  5.80it/s, loss=4.1, v_num=0]
Epoch 16:  37%|███▋      | 36/97 [00:06<00:10,  5.75it/s, loss=4.11, v_num=0]
Epoch 16:  38%|███▊      | 37/97 [00:06<00:10,  5.84it/s, loss=4.11, v_num=0]
Epoch 16:  38%|███▊      | 37/97 [00:06<00:10,  5.79it/s, loss=4.1, v_num=0]
Epoch 16:  39%|███▉      | 38/97 [00:06<00:10,  5.87it/s, loss=4.1, v_num=0]
Epoch 16:  39%|███▉      | 38/97 [00:06<00:10,  5.82it/s, loss=4.1, v_num=0]
Epoch 16:  40%|████      | 39/97 [00:06<00:09,  5.90it/s, loss=4.1, v_num=0]
Epoch 16:  40%|████      | 39/97 [00:06<00:09,  5.85it/s, loss=4.09, v_num=0]
Epoch 16:  41%|████      | 40/97 [00:06<00:09,  5.93it/s, loss=4.09, v_num=0]
Epoch 16:  41%|████      | 40/97 [00:06<00:09,  5.88it/s, loss=4.1, v_num=0]
Epoch 16:  42%|████▏     | 41/97 [00:06<00:09,  5.95it/s, loss=4.1, v_num=0]
Epoch 16:  42%|████▏     | 41/97 [00:06<00:09,  5.91it/s, loss=4.1, v_num=0]
Epoch 16:  43%|████▎     | 42/97 [00:07<00:09,  5.98it/s, loss=4.1, v_num=0]
Epoch 16:  43%|████▎     | 42/97 [00:07<00:09,  5.93it/s, loss=4.1, v_num=0]
Epoch 16:  44%|████▍     | 43/97 [00:07<00:08,  6.01it/s, loss=4.1, v_num=0]
Epoch 16:  44%|████▍     | 43/97 [00:07<00:09,  5.96it/s, loss=4.1, v_num=0]
Epoch 16:  45%|████▌     | 44/97 [00:07<00:08,  6.03it/s, loss=4.1, v_num=0]
Epoch 16:  45%|████▌     | 44/97 [00:07<00:08,  5.98it/s, loss=4.1, v_num=0]
Epoch 16:  46%|████▋     | 45/97 [00:07<00:08,  6.05it/s, loss=4.1, v_num=0]
Epoch 16:  46%|████▋     | 45/97 [00:07<00:08,  6.01it/s, loss=4.1, v_num=0]
Epoch 16:  47%|████▋     | 46/97 [00:07<00:08,  6.08it/s, loss=4.1, v_num=0]
Epoch 16:  47%|████▋     | 46/97 [00:07<00:08,  6.03it/s, loss=4.1, v_num=0]
Epoch 16:  48%|████▊     | 47/97 [00:07<00:08,  6.10it/s, loss=4.1, v_num=0]
Epoch 16:  48%|████▊     | 47/97 [00:07<00:08,  6.05it/s, loss=4.11, v_num=0]
Epoch 16:  49%|████▉     | 48/97 [00:07<00:08,  6.12it/s, loss=4.11, v_num=0]
Epoch 16:  49%|████▉     | 48/97 [00:07<00:08,  6.07it/s, loss=4.11, v_num=0]
Epoch 16:  51%|█████     | 49/97 [00:07<00:07,  6.13it/s, loss=4.11, v_num=0]
Epoch 16:  51%|█████     | 49/97 [00:08<00:07,  6.09it/s, loss=4.12, v_num=0]
Epoch 16:  52%|█████▏    | 50/97 [00:08<00:07,  6.15it/s, loss=4.12, v_num=0]
Epoch 16:  52%|█████▏    | 50/97 [00:08<00:07,  6.11it/s, loss=4.11, v_num=0]
Epoch 16:  53%|█████▎    | 51/97 [00:08<00:07,  6.17it/s, loss=4.11, v_num=0]
Epoch 16:  53%|█████▎    | 51/97 [00:08<00:07,  6.13it/s, loss=4.1, v_num=0]
Epoch 16:  54%|█████▎    | 52/97 [00:08<00:07,  6.19it/s, loss=4.1, v_num=0]
Epoch 16:  54%|█████▎    | 52/97 [00:08<00:07,  6.15it/s, loss=4.11, v_num=0]
Epoch 16:  55%|█████▍    | 53/97 [00:08<00:07,  6.21it/s, loss=4.11, v_num=0]
Epoch 16:  55%|█████▍    | 53/97 [00:08<00:07,  6.17it/s, loss=4.11, v_num=0]
Epoch 16:  56%|█████▌    | 54/97 [00:08<00:06,  6.22it/s, loss=4.11, v_num=0]
Epoch 16:  56%|█████▌    | 54/97 [00:08<00:06,  6.18it/s, loss=4.11, v_num=0]
Epoch 16:  57%|█████▋    | 55/97 [00:08<00:06,  6.24it/s, loss=4.11, v_num=0]
Epoch 16:  57%|█████▋    | 55/97 [00:08<00:06,  6.20it/s, loss=4.11, v_num=0]
Epoch 16:  58%|█████▊    | 56/97 [00:08<00:06,  6.25it/s, loss=4.11, v_num=0]
Epoch 16:  58%|█████▊    | 56/97 [00:09<00:06,  6.21it/s, loss=4.11, v_num=0]
Epoch 16:  59%|█████▉    | 57/97 [00:09<00:06,  6.27it/s, loss=4.11, v_num=0]
Epoch 16:  59%|█████▉    | 57/97 [00:09<00:06,  6.23it/s, loss=4.11, v_num=0]
Epoch 16:  60%|█████▉    | 58/97 [00:09<00:06,  6.29it/s, loss=4.11, v_num=0]
Epoch 16:  60%|█████▉    | 58/97 [00:09<00:06,  6.25it/s, loss=4.11, v_num=0]
Epoch 16:  61%|██████    | 59/97 [00:09<00:06,  6.30it/s, loss=4.11, v_num=0]
Epoch 16:  61%|██████    | 59/97 [00:09<00:06,  6.26it/s, loss=4.11, v_num=0]
Epoch 16:  62%|██████▏   | 60/97 [00:09<00:05,  6.31it/s, loss=4.11, v_num=0]
Epoch 16:  62%|██████▏   | 60/97 [00:09<00:05,  6.27it/s, loss=4.11, v_num=0]
Epoch 16:  63%|██████▎   | 61/97 [00:09<00:05,  6.32it/s, loss=4.11, v_num=0]
Epoch 16:  63%|██████▎   | 61/97 [00:09<00:05,  6.29it/s, loss=4.1, v_num=0]
Epoch 16:  64%|██████▍   | 62/97 [00:09<00:05,  6.33it/s, loss=4.1, v_num=0]
Epoch 16:  64%|██████▍   | 62/97 [00:09<00:05,  6.30it/s, loss=4.1, v_num=0]
Epoch 16:  65%|██████▍   | 63/97 [00:09<00:05,  6.35it/s, loss=4.1, v_num=0]
Epoch 16:  65%|██████▍   | 63/97 [00:09<00:05,  6.31it/s, loss=4.09, v_num=0]
Epoch 16:  66%|██████▌   | 64/97 [00:10<00:05,  6.36it/s, loss=4.09, v_num=0]
Epoch 16:  66%|██████▌   | 64/97 [00:10<00:05,  6.32it/s, loss=4.09, v_num=0]
Epoch 16:  67%|██████▋   | 65/97 [00:10<00:05,  6.37it/s, loss=4.09, v_num=0]
Epoch 16:  67%|██████▋   | 65/97 [00:10<00:05,  6.33it/s, loss=4.09, v_num=0]
Epoch 16:  68%|██████▊   | 66/97 [00:10<00:04,  6.38it/s, loss=4.09, v_num=0]
Epoch 16:  68%|██████▊   | 66/97 [00:10<00:04,  6.35it/s, loss=4.09, v_num=0]
Epoch 16:  69%|██████▉   | 67/97 [00:10<00:04,  6.39it/s, loss=4.09, v_num=0]
Epoch 16:  69%|██████▉   | 67/97 [00:10<00:04,  6.36it/s, loss=4.08, v_num=0]
Epoch 16:  70%|███████   | 68/97 [00:10<00:04,  6.41it/s, loss=4.08, v_num=0]
Epoch 16:  70%|███████   | 68/97 [00:10<00:04,  6.37it/s, loss=4.07, v_num=0]
Epoch 16:  71%|███████   | 69/97 [00:10<00:04,  6.42it/s, loss=4.07, v_num=0]
Epoch 16:  71%|███████   | 69/97 [00:10<00:04,  6.38it/s, loss=4.07, v_num=0]
Epoch 16:  72%|███████▏  | 70/97 [00:10<00:04,  6.43it/s, loss=4.07, v_num=0]
Epoch 16:  72%|███████▏  | 70/97 [00:10<00:04,  6.40it/s, loss=4.08, v_num=0]
Epoch 16:  73%|███████▎  | 71/97 [00:11<00:04,  6.44it/s, loss=4.08, v_num=0]
Epoch 16:  73%|███████▎  | 71/97 [00:11<00:04,  6.41it/s, loss=4.09, v_num=0]
Epoch 16:  74%|███████▍  | 72/97 [00:11<00:03,  6.45it/s, loss=4.09, v_num=0]
Epoch 16:  74%|███████▍  | 72/97 [00:11<00:03,  6.42it/s, loss=4.09, v_num=0]
Epoch 16:  75%|███████▌  | 73/97 [00:11<00:03,  6.46it/s, loss=4.09, v_num=0]
Epoch 16:  75%|███████▌  | 73/97 [00:11<00:03,  6.43it/s, loss=4.09, v_num=0]
Epoch 16:  76%|███████▋  | 74/97 [00:11<00:03,  6.47it/s, loss=4.09, v_num=0]
Epoch 16:  76%|███████▋  | 74/97 [00:11<00:03,  6.44it/s, loss=4.08, v_num=0]
Epoch 16:  77%|███████▋  | 75/97 [00:11<00:03,  6.48it/s, loss=4.08, v_num=0]
Epoch 16:  77%|███████▋  | 75/97 [00:11<00:03,  6.45it/s, loss=4.08, v_num=0]
Epoch 16:  78%|███████▊  | 76/97 [00:11<00:03,  6.49it/s, loss=4.08, v_num=0]
Epoch 16:  78%|███████▊  | 76/97 [00:11<00:03,  6.46it/s, loss=4.07, v_num=0]
Epoch 16:  79%|███████▉  | 77/97 [00:11<00:03,  6.49it/s, loss=4.07, v_num=0]
Epoch 16:  79%|███████▉  | 77/97 [00:11<00:03,  6.47it/s, loss=4.07, v_num=0]
Epoch 16:  80%|████████  | 78/97 [00:11<00:02,  6.50it/s, loss=4.07, v_num=0]
Epoch 16:  80%|████████  | 78/97 [00:12<00:02,  6.47it/s, loss=4.08, v_num=0]
Epoch 16:  81%|████████▏ | 79/97 [00:12<00:02,  6.52it/s, loss=4.08, v_num=0]
Epoch 16:  81%|████████▏ | 79/97 [00:12<00:02,  6.48it/s, loss=4.07, v_num=0]
Epoch 16:  82%|████████▏ | 80/97 [00:12<00:02,  6.52it/s, loss=4.07, v_num=0]
Epoch 16:  82%|████████▏ | 80/97 [00:12<00:02,  6.49it/s, loss=4.07, v_num=0]
Epoch 16:  84%|████████▎ | 81/97 [00:12<00:02,  6.53it/s, loss=4.07, v_num=0]
Epoch 16:  84%|████████▎ | 81/97 [00:12<00:02,  6.50it/s, loss=4.07, v_num=0]
Epoch 16:  85%|████████▍ | 82/97 [00:12<00:02,  6.53it/s, loss=4.07, v_num=0]
Epoch 16:  85%|████████▍ | 82/97 [00:12<00:02,  6.50it/s, loss=4.07, v_num=0]
Epoch 16:  86%|████████▌ | 83/97 [00:12<00:02,  6.54it/s, loss=4.07, v_num=0]
Epoch 16:  86%|████████▌ | 83/97 [00:12<00:02,  6.51it/s, loss=4.07, v_num=0]
Epoch 16:  87%|████████▋ | 84/97 [00:12<00:01,  6.55it/s, loss=4.07, v_num=0]
Epoch 16:  87%|████████▋ | 84/97 [00:12<00:01,  6.52it/s, loss=4.06, v_num=0]
Epoch 16:  88%|████████▊ | 85/97 [00:12<00:01,  6.56it/s, loss=4.06, v_num=0]
Epoch 16:  88%|████████▊ | 85/97 [00:13<00:01,  6.53it/s, loss=4.06, v_num=0]
Epoch 16:  89%|████████▊ | 86/97 [00:13<00:01,  6.57it/s, loss=4.06, v_num=0]
Epoch 16:  89%|████████▊ | 86/97 [00:13<00:01,  6.54it/s, loss=4.06, v_num=0]
Epoch 16:  90%|████████▉ | 87/97 [00:13<00:01,  6.57it/s, loss=4.06, v_num=0]
Epoch 16:  90%|████████▉ | 87/97 [00:13<00:01,  6.55it/s, loss=4.06, v_num=0]
Epoch 16:  91%|█████████ | 88/97 [00:13<00:01,  6.58it/s, loss=4.06, v_num=0]
Epoch 16:  91%|█████████ | 88/97 [00:13<00:01,  6.55it/s, loss=4.06, v_num=0]
Epoch 16:  92%|█████████▏| 89/97 [00:13<00:01,  6.59it/s, loss=4.06, v_num=0]
Epoch 16:  92%|█████████▏| 89/97 [00:13<00:01,  6.56it/s, loss=4.05, v_num=0]
Epoch 16:  93%|█████████▎| 90/97 [00:13<00:01,  6.60it/s, loss=4.05, v_num=0]
Epoch 16:  93%|█████████▎| 90/97 [00:13<00:01,  6.57it/s, loss=4.05, v_num=0]
Epoch 16:  94%|█████████▍| 91/97 [00:13<00:00,  6.60it/s, loss=4.05, v_num=0]
Epoch 16:  94%|█████████▍| 91/97 [00:13<00:00,  6.57it/s, loss=4.04, v_num=0]
Epoch 16:  95%|█████████▍| 92/97 [00:13<00:00,  6.61it/s, loss=4.04, v_num=0]
Epoch 16:  95%|█████████▍| 92/97 [00:13<00:00,  6.58it/s, loss=4.04, v_num=0]
Epoch 16:  96%|█████████▌| 93/97 [00:14<00:00,  6.61it/s, loss=4.04, v_num=0]
Epoch 16:  96%|█████████▌| 93/97 [00:14<00:00,  6.59it/s, loss=4.04, v_num=0]
Epoch 16:  97%|█████████▋| 94/97 [00:14<00:00,  6.62it/s, loss=4.04, v_num=0]
Epoch 16:  97%|█████████▋| 94/97 [00:14<00:00,  6.60it/s, loss=4.05, v_num=0]
Epoch 16:  98%|█████████▊| 95/97 [00:14<00:00,  6.63it/s, loss=4.05, v_num=0]
Epoch 16:  98%|█████████▊| 95/97 [00:14<00:00,  6.60it/s, loss=4.04, v_num=0]
Epoch 16:  99%|█████████▉| 96/97 [00:14<00:00,  6.64it/s, loss=4.04, v_num=0]
Epoch 16:  99%|█████████▉| 96/97 [00:14<00:00,  6.61it/s, loss=4.04, v_num=0]
Epoch 16: 100%|██████████| 97/97 [00:14<00:00,  6.64it/s, loss=4.04, v_num=0]
Epoch 16: 100%|██████████| 97/97 [00:14<00:00,  6.62it/s, loss=4.03, v_num=0]
Epoch 16: 100%|██████████| 97/97 [00:14<00:00,  6.62it/s, loss=4.03, v_num=0]
Epoch 16:   0%|          | 0/97 [00:00<?, ?it/s, loss=4.03, v_num=0]
Epoch 17:   0%|          | 0/97 [00:00<?, ?it/s, loss=4.03, v_num=0]
Epoch 17:   1%|          | 1/97 [00:01<02:10,  1.36s/it, loss=4.03, v_num=0]
Epoch 17:   1%|          | 1/97 [00:01<02:15,  1.42s/it, loss=4.03, v_num=0]
Epoch 17:   2%|▏         | 2/97 [00:01<01:11,  1.34it/s, loss=4.03, v_num=0]
Epoch 17:   2%|▏         | 2/97 [00:01<01:13,  1.29it/s, loss=4.04, v_num=0]
Epoch 17:   3%|▎         | 3/97 [00:01<00:51,  1.82it/s, loss=4.04, v_num=0]
Epoch 17:   3%|▎         | 3/97 [00:01<00:53,  1.76it/s, loss=4.04, v_num=0]
Epoch 17:   4%|▍         | 4/97 [00:01<00:41,  2.24it/s, loss=4.04, v_num=0]
Epoch 17:   4%|▍         | 4/97 [00:01<00:42,  2.18it/s, loss=4.05, v_num=0]
Epoch 17:   5%|▌         | 5/97 [00:01<00:35,  2.60it/s, loss=4.05, v_num=0]
Epoch 17:   5%|▌         | 5/97 [00:01<00:36,  2.53it/s, loss=4.05, v_num=0]
Epoch 17:   6%|▌         | 6/97 [00:02<00:31,  2.92it/s, loss=4.05, v_num=0]
Epoch 17:   6%|▌         | 6/97 [00:02<00:32,  2.84it/s, loss=4.05, v_num=0]
Epoch 17:   7%|▋         | 7/97 [00:02<00:28,  3.19it/s, loss=4.05, v_num=0]
Epoch 17:   7%|▋         | 7/97 [00:02<00:28,  3.11it/s, loss=4.05, v_num=0]
Epoch 17:   8%|▊         | 8/97 [00:02<00:25,  3.44it/s, loss=4.05, v_num=0]
Epoch 17:   8%|▊         | 8/97 [00:02<00:26,  3.36it/s, loss=4.05, v_num=0]
Epoch 17:   9%|▉         | 9/97 [00:02<00:24,  3.65it/s, loss=4.05, v_num=0]
Epoch 17:   9%|▉         | 9/97 [00:02<00:24,  3.57it/s, loss=4.04, v_num=0]
Epoch 17:  10%|█         | 10/97 [00:02<00:22,  3.84it/s, loss=4.04, v_num=0]
Epoch 17:  10%|█         | 10/97 [00:02<00:23,  3.76it/s, loss=4.04, v_num=0]
Epoch 17:  11%|█▏        | 11/97 [00:02<00:21,  4.02it/s, loss=4.04, v_num=0]
Epoch 17:  11%|█▏        | 11/97 [00:02<00:21,  3.93it/s, loss=4.04, v_num=0]
Epoch 17:  12%|█▏        | 12/97 [00:02<00:20,  4.17it/s, loss=4.04, v_num=0]
Epoch 17:  12%|█▏        | 12/97 [00:02<00:20,  4.09it/s, loss=4.04, v_num=0]
Epoch 17:  13%|█▎        | 13/97 [00:03<00:19,  4.32it/s, loss=4.04, v_num=0]
Epoch 17:  13%|█▎        | 13/97 [00:03<00:19,  4.24it/s, loss=4.04, v_num=0]
Epoch 17:  14%|█▍        | 14/97 [00:03<00:18,  4.45it/s, loss=4.04, v_num=0]
Epoch 17:  14%|█▍        | 14/97 [00:03<00:18,  4.37it/s, loss=4.05, v_num=0]
Epoch 17:  15%|█▌        | 15/97 [00:03<00:17,  4.57it/s, loss=4.05, v_num=0]
Epoch 17:  15%|█▌        | 15/97 [00:03<00:18,  4.49it/s, loss=4.04, v_num=0]
Epoch 17:  16%|█▋        | 16/97 [00:03<00:17,  4.68it/s, loss=4.04, v_num=0]
Epoch 17:  16%|█▋        | 16/97 [00:03<00:17,  4.60it/s, loss=4.04, v_num=0]
Epoch 17:  18%|█▊        | 17/97 [00:03<00:16,  4.77it/s, loss=4.04, v_num=0]
Epoch 17:  18%|█▊        | 17/97 [00:03<00:17,  4.70it/s, loss=4.03, v_num=0]
Epoch 17:  19%|█▊        | 18/97 [00:03<00:16,  4.87it/s, loss=4.03, v_num=0]
Epoch 17:  19%|█▊        | 18/97 [00:03<00:16,  4.80it/s, loss=4.04, v_num=0]
Epoch 17:  20%|█▉        | 19/97 [00:03<00:15,  4.96it/s, loss=4.04, v_num=0]
Epoch 17:  20%|█▉        | 19/97 [00:03<00:15,  4.89it/s, loss=4.04, v_num=0]
Epoch 17:  21%|██        | 20/97 [00:03<00:15,  5.04it/s, loss=4.04, v_num=0]
Epoch 17:  21%|██        | 20/97 [00:04<00:15,  4.97it/s, loss=4.04, v_num=0]
Epoch 17:  22%|██▏       | 21/97 [00:04<00:14,  5.11it/s, loss=4.04, v_num=0]
Epoch 17:  22%|██▏       | 21/97 [00:04<00:15,  5.05it/s, loss=4.04, v_num=0]
Epoch 17:  23%|██▎       | 22/97 [00:04<00:14,  5.19it/s, loss=4.04, v_num=0]
Epoch 17:  23%|██▎       | 22/97 [00:04<00:14,  5.12it/s, loss=4.04, v_num=0]
Epoch 17:  24%|██▎       | 23/97 [00:04<00:14,  5.25it/s, loss=4.04, v_num=0]
Epoch 17:  24%|██▎       | 23/97 [00:04<00:14,  5.18it/s, loss=4.04, v_num=0]
Epoch 17:  25%|██▍       | 24/97 [00:04<00:13,  5.32it/s, loss=4.04, v_num=0]
Epoch 17:  25%|██▍       | 24/97 [00:04<00:13,  5.25it/s, loss=4.03, v_num=0]
Epoch 17:  26%|██▌       | 25/97 [00:04<00:13,  5.37it/s, loss=4.03, v_num=0]
Epoch 17:  26%|██▌       | 25/97 [00:04<00:13,  5.30it/s, loss=4.03, v_num=0]
Epoch 17:  27%|██▋       | 26/97 [00:04<00:13,  5.43it/s, loss=4.03, v_num=0]
Epoch 17:  27%|██▋       | 26/97 [00:04<00:13,  5.36it/s, loss=4.03, v_num=0]
Epoch 17:  28%|██▊       | 27/97 [00:04<00:12,  5.48it/s, loss=4.03, v_num=0]
Epoch 17:  28%|██▊       | 27/97 [00:04<00:12,  5.42it/s, loss=4.04, v_num=0]
Epoch 17:  29%|██▉       | 28/97 [00:05<00:12,  5.53it/s, loss=4.04, v_num=0]
Epoch 17:  29%|██▉       | 28/97 [00:05<00:12,  5.46it/s, loss=4.04, v_num=0]
Epoch 17:  30%|██▉       | 29/97 [00:05<00:12,  5.58it/s, loss=4.04, v_num=0]
Epoch 17:  30%|██▉       | 29/97 [00:05<00:12,  5.51it/s, loss=4.05, v_num=0]
Epoch 17:  31%|███       | 30/97 [00:05<00:11,  5.61it/s, loss=4.05, v_num=0]
Epoch 17:  31%|███       | 30/97 [00:05<00:12,  5.56it/s, loss=4.05, v_num=0]
Epoch 17:  32%|███▏      | 31/97 [00:05<00:11,  5.66it/s, loss=4.05, v_num=0]
Epoch 17:  32%|███▏      | 31/97 [00:05<00:11,  5.60it/s, loss=4.06, v_num=0]
Epoch 17:  33%|███▎      | 32/97 [00:05<00:11,  5.70it/s, loss=4.06, v_num=0]
Epoch 17:  33%|███▎      | 32/97 [00:05<00:11,  5.64it/s, loss=4.07, v_num=0]
Epoch 17:  34%|███▍      | 33/97 [00:05<00:11,  5.74it/s, loss=4.07, v_num=0]
Epoch 17:  34%|███▍      | 33/97 [00:05<00:11,  5.68it/s, loss=4.07, v_num=0]
Epoch 17:  35%|███▌      | 34/97 [00:05<00:10,  5.77it/s, loss=4.07, v_num=0]
Epoch 17:  35%|███▌      | 34/97 [00:05<00:11,  5.72it/s, loss=4.07, v_num=0]
Epoch 17:  36%|███▌      | 35/97 [00:06<00:10,  5.81it/s, loss=4.07, v_num=0]
Epoch 17:  36%|███▌      | 35/97 [00:06<00:10,  5.75it/s, loss=4.07, v_num=0]
Epoch 17:  37%|███▋      | 36/97 [00:06<00:10,  5.83it/s, loss=4.07, v_num=0]
Epoch 17:  37%|███▋      | 36/97 [00:06<00:10,  5.78it/s, loss=4.07, v_num=0]
Epoch 17:  38%|███▊      | 37/97 [00:06<00:10,  5.87it/s, loss=4.07, v_num=0]
Epoch 17:  38%|███▊      | 37/97 [00:06<00:10,  5.82it/s, loss=4.08, v_num=0]
Epoch 17:  39%|███▉      | 38/97 [00:06<00:10,  5.89it/s, loss=4.08, v_num=0]
Epoch 17:  39%|███▉      | 38/97 [00:06<00:10,  5.84it/s, loss=4.06, v_num=0]
Epoch 17:  40%|████      | 39/97 [00:06<00:09,  5.92it/s, loss=4.06, v_num=0]
Epoch 17:  40%|████      | 39/97 [00:06<00:09,  5.87it/s, loss=4.06, v_num=0]
Epoch 17:  41%|████      | 40/97 [00:06<00:09,  5.95it/s, loss=4.06, v_num=0]
Epoch 17:  41%|████      | 40/97 [00:06<00:09,  5.90it/s, loss=4.07, v_num=0]
Epoch 17:  42%|████▏     | 41/97 [00:06<00:09,  5.98it/s, loss=4.07, v_num=0]
Epoch 17:  42%|████▏     | 41/97 [00:06<00:09,  5.93it/s, loss=4.07, v_num=0]
Epoch 17:  43%|████▎     | 42/97 [00:06<00:09,  6.01it/s, loss=4.07, v_num=0]
Epoch 17:  43%|████▎     | 42/97 [00:07<00:09,  5.96it/s, loss=4.06, v_num=0]
Epoch 17:  44%|████▍     | 43/97 [00:07<00:08,  6.03it/s, loss=4.06, v_num=0]
Epoch 17:  44%|████▍     | 43/97 [00:07<00:09,  5.98it/s, loss=4.05, v_num=0]
Epoch 17:  45%|████▌     | 44/97 [00:07<00:08,  6.06it/s, loss=4.05, v_num=0]
Epoch 17:  45%|████▌     | 44/97 [00:07<00:08,  6.01it/s, loss=4.05, v_num=0]
Epoch 17:  46%|████▋     | 45/97 [00:07<00:08,  6.08it/s, loss=4.05, v_num=0]
Epoch 17:  46%|████▋     | 45/97 [00:07<00:08,  6.03it/s, loss=4.05, v_num=0]
Epoch 17:  47%|████▋     | 46/97 [00:07<00:08,  6.10it/s, loss=4.05, v_num=0]
Epoch 17:  47%|████▋     | 46/97 [00:07<00:08,  6.05it/s, loss=4.04, v_num=0]
Epoch 17:  48%|████▊     | 47/97 [00:07<00:08,  6.11it/s, loss=4.04, v_num=0]
Epoch 17:  48%|████▊     | 47/97 [00:07<00:08,  6.07it/s, loss=4.04, v_num=0]
Epoch 17:  49%|████▉     | 48/97 [00:07<00:07,  6.14it/s, loss=4.04, v_num=0]
Epoch 17:  49%|████▉     | 48/97 [00:07<00:08,  6.09it/s, loss=4.03, v_num=0]
Epoch 17:  51%|█████     | 49/97 [00:07<00:07,  6.15it/s, loss=4.03, v_num=0]
Epoch 17:  51%|█████     | 49/97 [00:08<00:07,  6.11it/s, loss=4.02, v_num=0]
Epoch 17:  52%|█████▏    | 50/97 [00:08<00:07,  6.17it/s, loss=4.02, v_num=0]
Epoch 17:  52%|█████▏    | 50/97 [00:08<00:07,  6.13it/s, loss=4.02, v_num=0]
Epoch 17:  53%|█████▎    | 51/97 [00:08<00:07,  6.19it/s, loss=4.02, v_num=0]
Epoch 17:  53%|█████▎    | 51/97 [00:08<00:07,  6.15it/s, loss=4.01, v_num=0]
Epoch 17:  54%|█████▎    | 52/97 [00:08<00:07,  6.20it/s, loss=4.01, v_num=0]
Epoch 17:  54%|█████▎    | 52/97 [00:08<00:07,  6.16it/s, loss=4, v_num=0]
Epoch 17:  55%|█████▍    | 53/97 [00:08<00:07,  6.22it/s, loss=4, v_num=0]
Epoch 17:  55%|█████▍    | 53/97 [00:08<00:07,  6.18it/s, loss=4, v_num=0]
Epoch 17:  56%|█████▌    | 54/97 [00:08<00:06,  6.24it/s, loss=4, v_num=0]
Epoch 17:  56%|█████▌    | 54/97 [00:08<00:06,  6.19it/s, loss=4, v_num=0]
Epoch 17:  57%|█████▋    | 55/97 [00:08<00:06,  6.25it/s, loss=4, v_num=0]
Epoch 17:  57%|█████▋    | 55/97 [00:08<00:06,  6.21it/s, loss=3.99, v_num=0]
Epoch 17:  58%|█████▊    | 56/97 [00:08<00:06,  6.26it/s, loss=3.99, v_num=0]
Epoch 17:  58%|█████▊    | 56/97 [00:09<00:06,  6.22it/s, loss=3.99, v_num=0]
Epoch 17:  59%|█████▉    | 57/97 [00:09<00:06,  6.28it/s, loss=3.99, v_num=0]
Epoch 17:  59%|█████▉    | 57/97 [00:09<00:06,  6.24it/s, loss=3.99, v_num=0]
Epoch 17:  60%|█████▉    | 58/97 [00:09<00:06,  6.29it/s, loss=3.99, v_num=0]
Epoch 17:  60%|█████▉    | 58/97 [00:09<00:06,  6.25it/s, loss=4, v_num=0]
Epoch 17:  61%|██████    | 59/97 [00:09<00:06,  6.30it/s, loss=4, v_num=0]
Epoch 17:  61%|██████    | 59/97 [00:09<00:06,  6.27it/s, loss=4, v_num=0]
Epoch 17:  62%|██████▏   | 60/97 [00:09<00:05,  6.32it/s, loss=4, v_num=0]
Epoch 17:  62%|██████▏   | 60/97 [00:09<00:05,  6.28it/s, loss=4, v_num=0]
Epoch 17:  63%|██████▎   | 61/97 [00:09<00:05,  6.33it/s, loss=4, v_num=0]
Epoch 17:  63%|██████▎   | 61/97 [00:09<00:05,  6.29it/s, loss=4.01, v_num=0]
Epoch 17:  64%|██████▍   | 62/97 [00:09<00:05,  6.34it/s, loss=4.01, v_num=0]
Epoch 17:  64%|██████▍   | 62/97 [00:09<00:05,  6.31it/s, loss=4.02, v_num=0]
Epoch 17:  65%|██████▍   | 63/97 [00:09<00:05,  6.35it/s, loss=4.02, v_num=0]
Epoch 17:  65%|██████▍   | 63/97 [00:09<00:05,  6.32it/s, loss=4.02, v_num=0]
Epoch 17:  66%|██████▌   | 64/97 [00:10<00:05,  6.37it/s, loss=4.02, v_num=0]
Epoch 17:  66%|██████▌   | 64/97 [00:10<00:05,  6.33it/s, loss=4.02, v_num=0]
Epoch 17:  67%|██████▋   | 65/97 [00:10<00:05,  6.38it/s, loss=4.02, v_num=0]
Epoch 17:  67%|██████▋   | 65/97 [00:10<00:05,  6.34it/s, loss=4.02, v_num=0]
Epoch 17:  68%|██████▊   | 66/97 [00:10<00:04,  6.39it/s, loss=4.02, v_num=0]
Epoch 17:  68%|██████▊   | 66/97 [00:10<00:04,  6.35it/s, loss=4.03, v_num=0]
Epoch 17:  69%|██████▉   | 67/97 [00:10<00:04,  6.40it/s, loss=4.03, v_num=0]
Epoch 17:  69%|██████▉   | 67/97 [00:10<00:04,  6.37it/s, loss=4.04, v_num=0]
Epoch 17:  70%|███████   | 68/97 [00:10<00:04,  6.41it/s, loss=4.04, v_num=0]
Epoch 17:  70%|███████   | 68/97 [00:10<00:04,  6.38it/s, loss=4.03, v_num=0]
Epoch 17:  71%|███████   | 69/97 [00:10<00:04,  6.42it/s, loss=4.03, v_num=0]
Epoch 17:  71%|███████   | 69/97 [00:10<00:04,  6.39it/s, loss=4.03, v_num=0]
Epoch 17:  72%|███████▏  | 70/97 [00:10<00:04,  6.44it/s, loss=4.03, v_num=0]
Epoch 17:  72%|███████▏  | 70/97 [00:10<00:04,  6.40it/s, loss=4.03, v_num=0]
Epoch 17:  73%|███████▎  | 71/97 [00:11<00:04,  6.44it/s, loss=4.03, v_num=0]
Epoch 17:  73%|███████▎  | 71/97 [00:11<00:04,  6.41it/s, loss=4.03, v_num=0]
Epoch 17:  74%|███████▍  | 72/97 [00:11<00:03,  6.46it/s, loss=4.03, v_num=0]
Epoch 17:  74%|███████▍  | 72/97 [00:11<00:03,  6.42it/s, loss=4.03, v_num=0]
Epoch 17:  75%|███████▌  | 73/97 [00:11<00:03,  6.47it/s, loss=4.03, v_num=0]
Epoch 17:  75%|███████▌  | 73/97 [00:11<00:03,  6.43it/s, loss=4.03, v_num=0]
Epoch 17:  76%|███████▋  | 74/97 [00:11<00:03,  6.48it/s, loss=4.03, v_num=0]
Epoch 17:  76%|███████▋  | 74/97 [00:11<00:03,  6.44it/s, loss=4.03, v_num=0]
Epoch 17:  77%|███████▋  | 75/97 [00:11<00:03,  6.48it/s, loss=4.03, v_num=0]
Epoch 17:  77%|███████▋  | 75/97 [00:11<00:03,  6.45it/s, loss=4.03, v_num=0]
Epoch 17:  78%|███████▊  | 76/97 [00:11<00:03,  6.50it/s, loss=4.03, v_num=0]
Epoch 17:  78%|███████▊  | 76/97 [00:11<00:03,  6.46it/s, loss=4.03, v_num=0]
Epoch 17:  79%|███████▉  | 77/97 [00:11<00:03,  6.50it/s, loss=4.03, v_num=0]
Epoch 17:  79%|███████▉  | 77/97 [00:11<00:03,  6.47it/s, loss=4.03, v_num=0]
Epoch 17:  80%|████████  | 78/97 [00:11<00:02,  6.52it/s, loss=4.03, v_num=0]
Epoch 17:  80%|████████  | 78/97 [00:12<00:02,  6.48it/s, loss=4.02, v_num=0]
Epoch 17:  81%|████████▏ | 79/97 [00:12<00:02,  6.52it/s, loss=4.02, v_num=0]
Epoch 17:  81%|████████▏ | 79/97 [00:12<00:02,  6.49it/s, loss=4.03, v_num=0]
Epoch 17:  82%|████████▏ | 80/97 [00:12<00:02,  6.53it/s, loss=4.03, v_num=0]
Epoch 17:  82%|████████▏ | 80/97 [00:12<00:02,  6.50it/s, loss=4.01, v_num=0]
Epoch 17:  84%|████████▎ | 81/97 [00:12<00:02,  6.54it/s, loss=4.01, v_num=0]
Epoch 17:  84%|████████▎ | 81/97 [00:12<00:02,  6.51it/s, loss=4.01, v_num=0]
Epoch 17:  85%|████████▍ | 82/97 [00:12<00:02,  6.55it/s, loss=4.01, v_num=0]
Epoch 17:  85%|████████▍ | 82/97 [00:12<00:02,  6.52it/s, loss=4.01, v_num=0]
Epoch 17:  86%|████████▌ | 83/97 [00:12<00:02,  6.55it/s, loss=4.01, v_num=0]
Epoch 17:  86%|████████▌ | 83/97 [00:12<00:02,  6.53it/s, loss=4.01, v_num=0]
Epoch 17:  87%|████████▋ | 84/97 [00:12<00:01,  6.56it/s, loss=4.01, v_num=0]
Epoch 17:  87%|████████▋ | 84/97 [00:12<00:01,  6.53it/s, loss=4.01, v_num=0]
Epoch 17:  88%|████████▊ | 85/97 [00:12<00:01,  6.57it/s, loss=4.01, v_num=0]
Epoch 17:  88%|████████▊ | 85/97 [00:12<00:01,  6.54it/s, loss=4.01, v_num=0]
Epoch 17:  89%|████████▊ | 86/97 [00:13<00:01,  6.58it/s, loss=4.01, v_num=0]
Epoch 17:  89%|████████▊ | 86/97 [00:13<00:01,  6.55it/s, loss=4.01, v_num=0]
Epoch 17:  90%|████████▉ | 87/97 [00:13<00:01,  6.59it/s, loss=4.01, v_num=0]
Epoch 17:  90%|████████▉ | 87/97 [00:13<00:01,  6.56it/s, loss=4.01, v_num=0]
Epoch 17:  91%|█████████ | 88/97 [00:13<00:01,  6.59it/s, loss=4.01, v_num=0]
Epoch 17:  91%|█████████ | 88/97 [00:13<00:01,  6.57it/s, loss=4, v_num=0]
Epoch 17:  92%|█████████▏| 89/97 [00:13<00:01,  6.60it/s, loss=4, v_num=0]
Epoch 17:  92%|█████████▏| 89/97 [00:13<00:01,  6.57it/s, loss=4, v_num=0]
Epoch 17:  93%|█████████▎| 90/97 [00:13<00:01,  6.61it/s, loss=4, v_num=0]
Epoch 17:  93%|█████████▎| 90/97 [00:13<00:01,  6.58it/s, loss=4.01, v_num=0]
Epoch 17:  94%|█████████▍| 91/97 [00:13<00:00,  6.62it/s, loss=4.01, v_num=0]
Epoch 17:  94%|█████████▍| 91/97 [00:13<00:00,  6.59it/s, loss=4.01, v_num=0]
Epoch 17:  95%|█████████▍| 92/97 [00:13<00:00,  6.62it/s, loss=4.01, v_num=0]
Epoch 17:  95%|█████████▍| 92/97 [00:13<00:00,  6.60it/s, loss=4.01, v_num=0]
Epoch 17:  96%|█████████▌| 93/97 [00:14<00:00,  6.63it/s, loss=4.01, v_num=0]
Epoch 17:  96%|█████████▌| 93/97 [00:14<00:00,  6.60it/s, loss=4.01, v_num=0]
Epoch 17:  97%|█████████▋| 94/97 [00:14<00:00,  6.64it/s, loss=4.01, v_num=0]
Epoch 17:  97%|█████████▋| 94/97 [00:14<00:00,  6.61it/s, loss=4.01, v_num=0]
Epoch 17:  98%|█████████▊| 95/97 [00:14<00:00,  6.64it/s, loss=4.01, v_num=0]
Epoch 17:  98%|█████████▊| 95/97 [00:14<00:00,  6.62it/s, loss=4.01, v_num=0]
Epoch 17:  99%|█████████▉| 96/97 [00:14<00:00,  6.65it/s, loss=4.01, v_num=0]
Epoch 17:  99%|█████████▉| 96/97 [00:14<00:00,  6.62it/s, loss=4.01, v_num=0]
Epoch 17: 100%|██████████| 97/97 [00:14<00:00,  6.66it/s, loss=4.01, v_num=0]
Epoch 17: 100%|██████████| 97/97 [00:14<00:00,  6.63it/s, loss=4, v_num=0]
Epoch 17: 100%|██████████| 97/97 [00:14<00:00,  6.63it/s, loss=4, v_num=0]
Epoch 17:   0%|          | 0/97 [00:00<?, ?it/s, loss=4, v_num=0]
Epoch 18:   0%|          | 0/97 [00:00<?, ?it/s, loss=4, v_num=0]
Epoch 18:   1%|          | 1/97 [00:01<02:16,  1.42s/it, loss=4, v_num=0]
Epoch 18:   1%|          | 1/97 [00:01<02:21,  1.48s/it, loss=4.01, v_num=0]
Epoch 18:   2%|▏         | 2/97 [00:01<01:14,  1.28it/s, loss=4.01, v_num=0]
Epoch 18:   2%|▏         | 2/97 [00:01<01:16,  1.24it/s, loss=4, v_num=0]
Epoch 18:   3%|▎         | 3/97 [00:01<00:53,  1.77it/s, loss=4, v_num=0]
Epoch 18:   3%|▎         | 3/97 [00:01<00:54,  1.71it/s, loss=4, v_num=0]
Epoch 18:   4%|▍         | 4/97 [00:01<00:42,  2.18it/s, loss=4, v_num=0]
Epoch 18:   4%|▍         | 4/97 [00:01<00:43,  2.12it/s, loss=4, v_num=0]
Epoch 18:   5%|▌         | 5/97 [00:01<00:36,  2.54it/s, loss=4, v_num=0]
Epoch 18:   5%|▌         | 5/97 [00:02<00:37,  2.47it/s, loss=4, v_num=0]
Epoch 18:   6%|▌         | 6/97 [00:02<00:31,  2.85it/s, loss=4, v_num=0]
Epoch 18:   6%|▌         | 6/97 [00:02<00:32,  2.77it/s, loss=4, v_num=0]
Epoch 18:   7%|▋         | 7/97 [00:02<00:28,  3.12it/s, loss=4, v_num=0]
Epoch 18:   7%|▋         | 7/97 [00:02<00:29,  3.04it/s, loss=3.99, v_num=0]
Epoch 18:   8%|▊         | 8/97 [00:02<00:26,  3.36it/s, loss=3.99, v_num=0]
Epoch 18:   8%|▊         | 8/97 [00:02<00:27,  3.28it/s, loss=3.98, v_num=0]
Epoch 18:   9%|▉         | 9/97 [00:02<00:24,  3.57it/s, loss=3.98, v_num=0]
Epoch 18:   9%|▉         | 9/97 [00:02<00:25,  3.49it/s, loss=3.98, v_num=0]
Epoch 18:  10%|█         | 10/97 [00:02<00:23,  3.76it/s, loss=3.98, v_num=0]
Epoch 18:  10%|█         | 10/97 [00:02<00:23,  3.68it/s, loss=3.98, v_num=0]
Epoch 18:  11%|█▏        | 11/97 [00:02<00:21,  3.94it/s, loss=3.98, v_num=0]
Epoch 18:  11%|█▏        | 11/97 [00:02<00:22,  3.86it/s, loss=3.98, v_num=0]
Epoch 18:  12%|█▏        | 12/97 [00:02<00:20,  4.09it/s, loss=3.98, v_num=0]
Epoch 18:  12%|█▏        | 12/97 [00:02<00:21,  4.02it/s, loss=3.97, v_num=0]
Epoch 18:  13%|█▎        | 13/97 [00:03<00:19,  4.24it/s, loss=3.97, v_num=0]
Epoch 18:  13%|█▎        | 13/97 [00:03<00:20,  4.16it/s, loss=3.96, v_num=0]
Epoch 18:  14%|█▍        | 14/97 [00:03<00:19,  4.37it/s, loss=3.96, v_num=0]
Epoch 18:  14%|█▍        | 14/97 [00:03<00:19,  4.29it/s, loss=3.96, v_num=0]
Epoch 18:  15%|█▌        | 15/97 [00:03<00:18,  4.49it/s, loss=3.96, v_num=0]
Epoch 18:  15%|█▌        | 15/97 [00:03<00:18,  4.41it/s, loss=3.96, v_num=0]
Epoch 18:  16%|█▋        | 16/97 [00:03<00:17,  4.60it/s, loss=3.96, v_num=0]
Epoch 18:  16%|█▋        | 16/97 [00:03<00:17,  4.53it/s, loss=3.96, v_num=0]
Epoch 18:  18%|█▊        | 17/97 [00:03<00:17,  4.70it/s, loss=3.96, v_num=0]
Epoch 18:  18%|█▊        | 17/97 [00:03<00:17,  4.63it/s, loss=3.96, v_num=0]
Epoch 18:  19%|█▊        | 18/97 [00:03<00:16,  4.80it/s, loss=3.96, v_num=0]
Epoch 18:  19%|█▊        | 18/97 [00:03<00:16,  4.72it/s, loss=3.94, v_num=0]
Epoch 18:  20%|█▉        | 19/97 [00:03<00:15,  4.88it/s, loss=3.94, v_num=0]
Epoch 18:  20%|█▉        | 19/97 [00:03<00:16,  4.81it/s, loss=3.95, v_num=0]
Epoch 18:  21%|██        | 20/97 [00:04<00:15,  4.97it/s, loss=3.95, v_num=0]
Epoch 18:  21%|██        | 20/97 [00:04<00:15,  4.90it/s, loss=3.94, v_num=0]
Epoch 18:  22%|██▏       | 21/97 [00:04<00:15,  5.04it/s, loss=3.94, v_num=0]
Epoch 18:  22%|██▏       | 21/97 [00:04<00:15,  4.98it/s, loss=3.94, v_num=0]
Epoch 18:  23%|██▎       | 22/97 [00:04<00:14,  5.12it/s, loss=3.94, v_num=0]
Epoch 18:  23%|██▎       | 22/97 [00:04<00:14,  5.05it/s, loss=3.94, v_num=0]
Epoch 18:  24%|██▎       | 23/97 [00:04<00:14,  5.18it/s, loss=3.94, v_num=0]
Epoch 18:  24%|██▎       | 23/97 [00:04<00:14,  5.12it/s, loss=3.94, v_num=0]
Epoch 18:  25%|██▍       | 24/97 [00:04<00:13,  5.24it/s, loss=3.94, v_num=0]
Epoch 18:  25%|██▍       | 24/97 [00:04<00:14,  5.18it/s, loss=3.93, v_num=0]
Epoch 18:  26%|██▌       | 25/97 [00:04<00:13,  5.30it/s, loss=3.93, v_num=0]
Epoch 18:  26%|██▌       | 25/97 [00:04<00:13,  5.24it/s, loss=3.93, v_num=0]
Epoch 18:  27%|██▋       | 26/97 [00:04<00:13,  5.36it/s, loss=3.93, v_num=0]
Epoch 18:  27%|██▋       | 26/97 [00:04<00:13,  5.30it/s, loss=3.92, v_num=0]
Epoch 18:  28%|██▊       | 27/97 [00:04<00:12,  5.42it/s, loss=3.92, v_num=0]
Epoch 18:  28%|██▊       | 27/97 [00:05<00:13,  5.35it/s, loss=3.93, v_num=0]
Epoch 18:  29%|██▉       | 28/97 [00:05<00:12,  5.46it/s, loss=3.93, v_num=0]
Epoch 18:  29%|██▉       | 28/97 [00:05<00:12,  5.41it/s, loss=3.92, v_num=0]
Epoch 18:  30%|██▉       | 29/97 [00:05<00:12,  5.51it/s, loss=3.92, v_num=0]
Epoch 18:  30%|██▉       | 29/97 [00:05<00:12,  5.45it/s, loss=3.92, v_num=0]
Epoch 18:  31%|███       | 30/97 [00:05<00:12,  5.55it/s, loss=3.92, v_num=0]
Epoch 18:  31%|███       | 30/97 [00:05<00:12,  5.50it/s, loss=3.91, v_num=0]
Epoch 18:  32%|███▏      | 31/97 [00:05<00:11,  5.59it/s, loss=3.91, v_num=0]
Epoch 18:  32%|███▏      | 31/97 [00:05<00:11,  5.53it/s, loss=3.91, v_num=0]
Epoch 18:  33%|███▎      | 32/97 [00:05<00:11,  5.64it/s, loss=3.91, v_num=0]
Epoch 18:  33%|███▎      | 32/97 [00:05<00:11,  5.58it/s, loss=3.91, v_num=0]
Epoch 18:  34%|███▍      | 33/97 [00:05<00:11,  5.67it/s, loss=3.91, v_num=0]
Epoch 18:  34%|███▍      | 33/97 [00:05<00:11,  5.62it/s, loss=3.93, v_num=0]
Epoch 18:  35%|███▌      | 34/97 [00:05<00:11,  5.71it/s, loss=3.93, v_num=0]
Epoch 18:  35%|███▌      | 34/97 [00:06<00:11,  5.66it/s, loss=3.93, v_num=0]
Epoch 18:  36%|███▌      | 35/97 [00:06<00:10,  5.75it/s, loss=3.93, v_num=0]
Epoch 18:  36%|███▌      | 35/97 [00:06<00:10,  5.69it/s, loss=3.93, v_num=0]
Epoch 18:  37%|███▋      | 36/97 [00:06<00:10,  5.78it/s, loss=3.93, v_num=0]
Epoch 18:  37%|███▋      | 36/97 [00:06<00:10,  5.73it/s, loss=3.93, v_num=0]
Epoch 18:  38%|███▊      | 37/97 [00:06<00:10,  5.81it/s, loss=3.93, v_num=0]
Epoch 18:  38%|███▊      | 37/97 [00:06<00:10,  5.76it/s, loss=3.93, v_num=0]
Epoch 18:  39%|███▉      | 38/97 [00:06<00:10,  5.85it/s, loss=3.93, v_num=0]
Epoch 18:  39%|███▉      | 38/97 [00:06<00:10,  5.79it/s, loss=3.93, v_num=0]
Epoch 18:  40%|████      | 39/97 [00:06<00:09,  5.87it/s, loss=3.93, v_num=0]
Epoch 18:  40%|████      | 39/97 [00:06<00:09,  5.83it/s, loss=3.91, v_num=0]
Epoch 18:  41%|████      | 40/97 [00:06<00:09,  5.90it/s, loss=3.91, v_num=0]
Epoch 18:  41%|████      | 40/97 [00:06<00:09,  5.86it/s, loss=3.91, v_num=0]
Epoch 18:  42%|████▏     | 41/97 [00:06<00:09,  5.93it/s, loss=3.91, v_num=0]
Epoch 18:  42%|████▏     | 41/97 [00:06<00:09,  5.88it/s, loss=3.91, v_num=0]
Epoch 18:  43%|████▎     | 42/97 [00:07<00:09,  5.96it/s, loss=3.91, v_num=0]
Epoch 18:  43%|████▎     | 42/97 [00:07<00:09,  5.91it/s, loss=3.91, v_num=0]
Epoch 18:  44%|████▍     | 43/97 [00:07<00:09,  5.98it/s, loss=3.91, v_num=0]
Epoch 18:  44%|████▍     | 43/97 [00:07<00:09,  5.94it/s, loss=3.92, v_num=0]
Epoch 18:  45%|████▌     | 44/97 [00:07<00:08,  6.01it/s, loss=3.92, v_num=0]
Epoch 18:  45%|████▌     | 44/97 [00:07<00:08,  5.96it/s, loss=3.92, v_num=0]
Epoch 18:  46%|████▋     | 45/97 [00:07<00:08,  6.03it/s, loss=3.92, v_num=0]
Epoch 18:  46%|████▋     | 45/97 [00:07<00:08,  5.99it/s, loss=3.92, v_num=0]
Epoch 18:  47%|████▋     | 46/97 [00:07<00:08,  6.06it/s, loss=3.92, v_num=0]
Epoch 18:  47%|████▋     | 46/97 [00:07<00:08,  6.01it/s, loss=3.92, v_num=0]
Epoch 18:  48%|████▊     | 47/97 [00:07<00:08,  6.08it/s, loss=3.92, v_num=0]
Epoch 18:  48%|████▊     | 47/97 [00:07<00:08,  6.03it/s, loss=3.93, v_num=0]
Epoch 18:  49%|████▉     | 48/97 [00:07<00:08,  6.09it/s, loss=3.93, v_num=0]
Epoch 18:  49%|████▉     | 48/97 [00:07<00:08,  6.05it/s, loss=3.93, v_num=0]
Epoch 18:  51%|█████     | 49/97 [00:08<00:07,  6.11it/s, loss=3.93, v_num=0]
Epoch 18:  51%|█████     | 49/97 [00:08<00:07,  6.07it/s, loss=3.94, v_num=0]
Epoch 18:  52%|█████▏    | 50/97 [00:08<00:07,  6.13it/s, loss=3.94, v_num=0]
Epoch 18:  52%|█████▏    | 50/97 [00:08<00:07,  6.09it/s, loss=3.93, v_num=0]
Epoch 18:  53%|█████▎    | 51/97 [00:08<00:07,  6.15it/s, loss=3.93, v_num=0]
Epoch 18:  53%|█████▎    | 51/97 [00:08<00:07,  6.11it/s, loss=3.93, v_num=0]
Epoch 18:  54%|█████▎    | 52/97 [00:08<00:07,  6.17it/s, loss=3.93, v_num=0]
Epoch 18:  54%|█████▎    | 52/97 [00:08<00:07,  6.13it/s, loss=3.93, v_num=0]
Epoch 18:  55%|█████▍    | 53/97 [00:08<00:07,  6.18it/s, loss=3.93, v_num=0]
Epoch 18:  55%|█████▍    | 53/97 [00:08<00:07,  6.15it/s, loss=3.92, v_num=0]
Epoch 18:  56%|█████▌    | 54/97 [00:08<00:06,  6.20it/s, loss=3.92, v_num=0]
Epoch 18:  56%|█████▌    | 54/97 [00:08<00:06,  6.16it/s, loss=3.91, v_num=0]
Epoch 18:  57%|█████▋    | 55/97 [00:08<00:06,  6.22it/s, loss=3.91, v_num=0]
Epoch 18:  57%|█████▋    | 55/97 [00:08<00:06,  6.18it/s, loss=3.91, v_num=0]
Epoch 18:  58%|█████▊    | 56/97 [00:08<00:06,  6.23it/s, loss=3.91, v_num=0]
Epoch 18:  58%|█████▊    | 56/97 [00:09<00:06,  6.19it/s, loss=3.91, v_num=0]
Epoch 18:  59%|█████▉    | 57/97 [00:09<00:06,  6.25it/s, loss=3.91, v_num=0]
Epoch 18:  59%|█████▉    | 57/97 [00:09<00:06,  6.21it/s, loss=3.9, v_num=0]
Epoch 18:  60%|█████▉    | 58/97 [00:09<00:06,  6.27it/s, loss=3.9, v_num=0]
Epoch 18:  60%|█████▉    | 58/97 [00:09<00:06,  6.23it/s, loss=3.9, v_num=0]
Epoch 18:  61%|██████    | 59/97 [00:09<00:06,  6.28it/s, loss=3.9, v_num=0]
Epoch 18:  61%|██████    | 59/97 [00:09<00:06,  6.24it/s, loss=3.91, v_num=0]
Epoch 18:  62%|██████▏   | 60/97 [00:09<00:05,  6.29it/s, loss=3.91, v_num=0]
Epoch 18:  62%|██████▏   | 60/97 [00:09<00:05,  6.26it/s, loss=3.92, v_num=0]
Epoch 18:  63%|██████▎   | 61/97 [00:09<00:05,  6.31it/s, loss=3.92, v_num=0]
Epoch 18:  63%|██████▎   | 61/97 [00:09<00:05,  6.27it/s, loss=3.91, v_num=0]
Epoch 18:  64%|██████▍   | 62/97 [00:09<00:05,  6.32it/s, loss=3.91, v_num=0]
Epoch 18:  64%|██████▍   | 62/97 [00:09<00:05,  6.28it/s, loss=3.9, v_num=0]
Epoch 18:  65%|██████▍   | 63/97 [00:09<00:05,  6.33it/s, loss=3.9, v_num=0]
Epoch 18:  65%|██████▍   | 63/97 [00:10<00:05,  6.30it/s, loss=3.9, v_num=0]
Epoch 18:  66%|██████▌   | 64/97 [00:10<00:05,  6.35it/s, loss=3.9, v_num=0]
Epoch 18:  66%|██████▌   | 64/97 [00:10<00:05,  6.31it/s, loss=3.91, v_num=0]
Epoch 18:  67%|██████▋   | 65/97 [00:10<00:05,  6.36it/s, loss=3.91, v_num=0]
Epoch 18:  67%|██████▋   | 65/97 [00:10<00:05,  6.32it/s, loss=3.91, v_num=0]
Epoch 18:  68%|██████▊   | 66/97 [00:10<00:04,  6.37it/s, loss=3.91, v_num=0]
Epoch 18:  68%|██████▊   | 66/97 [00:10<00:04,  6.34it/s, loss=3.91, v_num=0]
Epoch 18:  69%|██████▉   | 67/97 [00:10<00:04,  6.38it/s, loss=3.91, v_num=0]
Epoch 18:  69%|██████▉   | 67/97 [00:10<00:04,  6.35it/s, loss=3.9, v_num=0]
Epoch 18:  70%|███████   | 68/97 [00:10<00:04,  6.39it/s, loss=3.9, v_num=0]
Epoch 18:  70%|███████   | 68/97 [00:10<00:04,  6.36it/s, loss=3.91, v_num=0]
Epoch 18:  71%|███████   | 69/97 [00:10<00:04,  6.41it/s, loss=3.91, v_num=0]
Epoch 18:  71%|███████   | 69/97 [00:10<00:04,  6.37it/s, loss=3.9, v_num=0]
Epoch 18:  72%|███████▏  | 70/97 [00:10<00:04,  6.42it/s, loss=3.9, v_num=0]
Epoch 18:  72%|███████▏  | 70/97 [00:10<00:04,  6.39it/s, loss=3.91, v_num=0]
Epoch 18:  73%|███████▎  | 71/97 [00:11<00:04,  6.43it/s, loss=3.91, v_num=0]
Epoch 18:  73%|███████▎  | 71/97 [00:11<00:04,  6.40it/s, loss=3.91, v_num=0]
Epoch 18:  74%|███████▍  | 72/97 [00:11<00:03,  6.44it/s, loss=3.91, v_num=0]
Epoch 18:  74%|███████▍  | 72/97 [00:11<00:03,  6.41it/s, loss=3.91, v_num=0]
Epoch 18:  75%|███████▌  | 73/97 [00:11<00:03,  6.45it/s, loss=3.91, v_num=0]
Epoch 18:  75%|███████▌  | 73/97 [00:11<00:03,  6.42it/s, loss=3.91, v_num=0]
Epoch 18:  76%|███████▋  | 74/97 [00:11<00:03,  6.46it/s, loss=3.91, v_num=0]
Epoch 18:  76%|███████▋  | 74/97 [00:11<00:03,  6.43it/s, loss=3.93, v_num=0]
Epoch 18:  77%|███████▋  | 75/97 [00:11<00:03,  6.47it/s, loss=3.93, v_num=0]
Epoch 18:  77%|███████▋  | 75/97 [00:11<00:03,  6.44it/s, loss=3.93, v_num=0]
Epoch 18:  78%|███████▊  | 76/97 [00:11<00:03,  6.48it/s, loss=3.93, v_num=0]
Epoch 18:  78%|███████▊  | 76/97 [00:11<00:03,  6.45it/s, loss=3.94, v_num=0]
Epoch 18:  79%|███████▉  | 77/97 [00:11<00:03,  6.49it/s, loss=3.94, v_num=0]
Epoch 18:  79%|███████▉  | 77/97 [00:11<00:03,  6.46it/s, loss=3.94, v_num=0]
Epoch 18:  80%|████████  | 78/97 [00:12<00:02,  6.50it/s, loss=3.94, v_num=0]
Epoch 18:  80%|████████  | 78/97 [00:12<00:02,  6.47it/s, loss=3.96, v_num=0]
Epoch 18:  81%|████████▏ | 79/97 [00:12<00:02,  6.51it/s, loss=3.96, v_num=0]
Epoch 18:  81%|████████▏ | 79/97 [00:12<00:02,  6.48it/s, loss=3.96, v_num=0]
Epoch 18:  82%|████████▏ | 80/97 [00:12<00:02,  6.52it/s, loss=3.96, v_num=0]
Epoch 18:  82%|████████▏ | 80/97 [00:12<00:02,  6.49it/s, loss=3.95, v_num=0]
Epoch 18:  84%|████████▎ | 81/97 [00:12<00:02,  6.52it/s, loss=3.95, v_num=0]
Epoch 18:  84%|████████▎ | 81/97 [00:12<00:02,  6.50it/s, loss=3.95, v_num=0]
Epoch 18:  85%|████████▍ | 82/97 [00:12<00:02,  6.53it/s, loss=3.95, v_num=0]
Epoch 18:  85%|████████▍ | 82/97 [00:12<00:02,  6.50it/s, loss=3.97, v_num=0]
Epoch 18:  86%|████████▌ | 83/97 [00:12<00:02,  6.54it/s, loss=3.97, v_num=0]
Epoch 18:  86%|████████▌ | 83/97 [00:12<00:02,  6.51it/s, loss=3.95, v_num=0]
Epoch 18:  87%|████████▋ | 84/97 [00:12<00:01,  6.55it/s, loss=3.95, v_num=0]
Epoch 18:  87%|████████▋ | 84/97 [00:12<00:01,  6.52it/s, loss=3.95, v_num=0]
Epoch 18:  88%|████████▊ | 85/97 [00:12<00:01,  6.55it/s, loss=3.95, v_num=0]
Epoch 18:  88%|████████▊ | 85/97 [00:13<00:01,  6.53it/s, loss=3.95, v_num=0]
Epoch 18:  89%|████████▊ | 86/97 [00:13<00:01,  6.56it/s, loss=3.95, v_num=0]
Epoch 18:  89%|████████▊ | 86/97 [00:13<00:01,  6.53it/s, loss=3.95, v_num=0]
Epoch 18:  90%|████████▉ | 87/97 [00:13<00:01,  6.57it/s, loss=3.95, v_num=0]
Epoch 18:  90%|████████▉ | 87/97 [00:13<00:01,  6.54it/s, loss=3.95, v_num=0]
Epoch 18:  91%|█████████ | 88/97 [00:13<00:01,  6.58it/s, loss=3.95, v_num=0]
Epoch 18:  91%|█████████ | 88/97 [00:13<00:01,  6.55it/s, loss=3.95, v_num=0]
Epoch 18:  92%|█████████▏| 89/97 [00:13<00:01,  6.59it/s, loss=3.95, v_num=0]
Epoch 18:  92%|█████████▏| 89/97 [00:13<00:01,  6.56it/s, loss=3.94, v_num=0]
Epoch 18:  93%|█████████▎| 90/97 [00:13<00:01,  6.59it/s, loss=3.94, v_num=0]
Epoch 18:  93%|█████████▎| 90/97 [00:13<00:01,  6.57it/s, loss=3.94, v_num=0]
Epoch 18:  94%|█████████▍| 91/97 [00:13<00:00,  6.60it/s, loss=3.94, v_num=0]
Epoch 18:  94%|█████████▍| 91/97 [00:13<00:00,  6.57it/s, loss=3.94, v_num=0]
Epoch 18:  95%|█████████▍| 92/97 [00:13<00:00,  6.61it/s, loss=3.94, v_num=0]
Epoch 18:  95%|█████████▍| 92/97 [00:13<00:00,  6.58it/s, loss=3.94, v_num=0]
Epoch 18:  96%|█████████▌| 93/97 [00:14<00:00,  6.61it/s, loss=3.94, v_num=0]
Epoch 18:  96%|█████████▌| 93/97 [00:14<00:00,  6.58it/s, loss=3.93, v_num=0]
Epoch 18:  97%|█████████▋| 94/97 [00:14<00:00,  6.62it/s, loss=3.93, v_num=0]
Epoch 18:  97%|█████████▋| 94/97 [00:14<00:00,  6.59it/s, loss=3.92, v_num=0]
Epoch 18:  98%|█████████▊| 95/97 [00:14<00:00,  6.62it/s, loss=3.92, v_num=0]
Epoch 18:  98%|█████████▊| 95/97 [00:14<00:00,  6.60it/s, loss=3.91, v_num=0]
Epoch 18:  99%|█████████▉| 96/97 [00:14<00:00,  6.63it/s, loss=3.91, v_num=0]
Epoch 18:  99%|█████████▉| 96/97 [00:14<00:00,  6.60it/s, loss=3.91, v_num=0]
Epoch 18: 100%|██████████| 97/97 [00:14<00:00,  6.64it/s, loss=3.91, v_num=0]
Epoch 18: 100%|██████████| 97/97 [00:14<00:00,  6.61it/s, loss=3.9, v_num=0]
Epoch 18: 100%|██████████| 97/97 [00:14<00:00,  6.61it/s, loss=3.9, v_num=0]
Epoch 18:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.9, v_num=0]
Epoch 19:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.9, v_num=0]
Epoch 19:   1%|          | 1/97 [00:01<02:13,  1.39s/it, loss=3.9, v_num=0]
Epoch 19:   1%|          | 1/97 [00:01<02:18,  1.44s/it, loss=3.89, v_num=0]
Epoch 19:   2%|▏         | 2/97 [00:01<01:12,  1.31it/s, loss=3.89, v_num=0]
Epoch 19:   2%|▏         | 2/97 [00:01<01:14,  1.27it/s, loss=3.89, v_num=0]
Epoch 19:   3%|▎         | 3/97 [00:01<00:51,  1.81it/s, loss=3.89, v_num=0]
Epoch 19:   3%|▎         | 3/97 [00:01<00:53,  1.75it/s, loss=3.89, v_num=0]
Epoch 19:   4%|▍         | 4/97 [00:01<00:41,  2.23it/s, loss=3.89, v_num=0]
Epoch 19:   4%|▍         | 4/97 [00:01<00:43,  2.16it/s, loss=3.89, v_num=0]
Epoch 19:   5%|▌         | 5/97 [00:01<00:35,  2.58it/s, loss=3.89, v_num=0]
Epoch 19:   5%|▌         | 5/97 [00:01<00:36,  2.51it/s, loss=3.88, v_num=0]
Epoch 19:   6%|▌         | 6/97 [00:02<00:31,  2.89it/s, loss=3.88, v_num=0]
Epoch 19:   6%|▌         | 6/97 [00:02<00:32,  2.81it/s, loss=3.88, v_num=0]
Epoch 19:   7%|▋         | 7/97 [00:02<00:28,  3.16it/s, loss=3.88, v_num=0]
Epoch 19:   7%|▋         | 7/97 [00:02<00:29,  3.09it/s, loss=3.88, v_num=0]
Epoch 19:   8%|▊         | 8/97 [00:02<00:26,  3.40it/s, loss=3.88, v_num=0]
Epoch 19:   8%|▊         | 8/97 [00:02<00:26,  3.33it/s, loss=3.88, v_num=0]
Epoch 19:   9%|▉         | 9/97 [00:02<00:24,  3.61it/s, loss=3.88, v_num=0]
Epoch 19:   9%|▉         | 9/97 [00:02<00:24,  3.53it/s, loss=3.88, v_num=0]
Epoch 19:  10%|█         | 10/97 [00:02<00:22,  3.80it/s, loss=3.88, v_num=0]
Epoch 19:  10%|█         | 10/97 [00:02<00:23,  3.73it/s, loss=3.88, v_num=0]
Epoch 19:  11%|█▏        | 11/97 [00:02<00:21,  3.98it/s, loss=3.88, v_num=0]
Epoch 19:  11%|█▏        | 11/97 [00:02<00:22,  3.90it/s, loss=3.87, v_num=0]
Epoch 19:  12%|█▏        | 12/97 [00:02<00:20,  4.14it/s, loss=3.87, v_num=0]
Epoch 19:  12%|█▏        | 12/97 [00:02<00:20,  4.06it/s, loss=3.87, v_num=0]
Epoch 19:  13%|█▎        | 13/97 [00:03<00:19,  4.28it/s, loss=3.87, v_num=0]
Epoch 19:  13%|█▎        | 13/97 [00:03<00:19,  4.20it/s, loss=3.87, v_num=0]
Epoch 19:  14%|█▍        | 14/97 [00:03<00:18,  4.41it/s, loss=3.87, v_num=0]
Epoch 19:  14%|█▍        | 14/97 [00:03<00:19,  4.34it/s, loss=3.87, v_num=0]
Epoch 19:  15%|█▌        | 15/97 [00:03<00:18,  4.53it/s, loss=3.87, v_num=0]
Epoch 19:  15%|█▌        | 15/97 [00:03<00:18,  4.46it/s, loss=3.87, v_num=0]
Epoch 19:  16%|█▋        | 16/97 [00:03<00:17,  4.64it/s, loss=3.87, v_num=0]
Epoch 19:  16%|█▋        | 16/97 [00:03<00:17,  4.57it/s, loss=3.86, v_num=0]
Epoch 19:  18%|█▊        | 17/97 [00:03<00:16,  4.75it/s, loss=3.86, v_num=0]
Epoch 19:  18%|█▊        | 17/97 [00:03<00:17,  4.67it/s, loss=3.86, v_num=0]
Epoch 19:  19%|█▊        | 18/97 [00:03<00:16,  4.84it/s, loss=3.86, v_num=0]
Epoch 19:  19%|█▊        | 18/97 [00:03<00:16,  4.77it/s, loss=3.87, v_num=0]
Epoch 19:  20%|█▉        | 19/97 [00:03<00:15,  4.93it/s, loss=3.87, v_num=0]
Epoch 19:  20%|█▉        | 19/97 [00:03<00:16,  4.86it/s, loss=3.87, v_num=0]
Epoch 19:  21%|██        | 20/97 [00:03<00:15,  5.01it/s, loss=3.87, v_num=0]
Epoch 19:  21%|██        | 20/97 [00:04<00:15,  4.94it/s, loss=3.87, v_num=0]
Epoch 19:  22%|██▏       | 21/97 [00:04<00:14,  5.09it/s, loss=3.87, v_num=0]
Epoch 19:  22%|██▏       | 21/97 [00:04<00:15,  5.02it/s, loss=3.87, v_num=0]
Epoch 19:  23%|██▎       | 22/97 [00:04<00:14,  5.16it/s, loss=3.87, v_num=0]
Epoch 19:  23%|██▎       | 22/97 [00:04<00:14,  5.09it/s, loss=3.86, v_num=0]
Epoch 19:  24%|██▎       | 23/97 [00:04<00:14,  5.22it/s, loss=3.86, v_num=0]
Epoch 19:  24%|██▎       | 23/97 [00:04<00:14,  5.16it/s, loss=3.86, v_num=0]
Epoch 19:  25%|██▍       | 24/97 [00:04<00:13,  5.29it/s, loss=3.86, v_num=0]
Epoch 19:  25%|██▍       | 24/97 [00:04<00:13,  5.22it/s, loss=3.87, v_num=0]
Epoch 19:  26%|██▌       | 25/97 [00:04<00:13,  5.35it/s, loss=3.87, v_num=0]
Epoch 19:  26%|██▌       | 25/97 [00:04<00:13,  5.28it/s, loss=3.88, v_num=0]
Epoch 19:  27%|██▋       | 26/97 [00:04<00:13,  5.40it/s, loss=3.88, v_num=0]
Epoch 19:  27%|██▋       | 26/97 [00:04<00:13,  5.34it/s, loss=3.88, v_num=0]
Epoch 19:  28%|██▊       | 27/97 [00:04<00:12,  5.45it/s, loss=3.88, v_num=0]
Epoch 19:  28%|██▊       | 27/97 [00:05<00:12,  5.39it/s, loss=3.88, v_num=0]
Epoch 19:  29%|██▉       | 28/97 [00:05<00:12,  5.50it/s, loss=3.88, v_num=0]
Epoch 19:  29%|██▉       | 28/97 [00:05<00:12,  5.44it/s, loss=3.87, v_num=0]
Epoch 19:  30%|██▉       | 29/97 [00:05<00:12,  5.55it/s, loss=3.87, v_num=0]
Epoch 19:  30%|██▉       | 29/97 [00:05<00:12,  5.49it/s, loss=3.87, v_num=0]
Epoch 19:  31%|███       | 30/97 [00:05<00:11,  5.59it/s, loss=3.87, v_num=0]
Epoch 19:  31%|███       | 30/97 [00:05<00:12,  5.54it/s, loss=3.87, v_num=0]
Epoch 19:  32%|███▏      | 31/97 [00:05<00:11,  5.63it/s, loss=3.87, v_num=0]
Epoch 19:  32%|███▏      | 31/97 [00:05<00:11,  5.58it/s, loss=3.88, v_num=0]
Epoch 19:  33%|███▎      | 32/97 [00:05<00:11,  5.68it/s, loss=3.88, v_num=0]
Epoch 19:  33%|███▎      | 32/97 [00:05<00:11,  5.62it/s, loss=3.88, v_num=0]
Epoch 19:  34%|███▍      | 33/97 [00:05<00:11,  5.72it/s, loss=3.88, v_num=0]
Epoch 19:  34%|███▍      | 33/97 [00:05<00:11,  5.66it/s, loss=3.88, v_num=0]
Epoch 19:  35%|███▌      | 34/97 [00:05<00:10,  5.75it/s, loss=3.88, v_num=0]
Epoch 19:  35%|███▌      | 34/97 [00:05<00:11,  5.70it/s, loss=3.88, v_num=0]
Epoch 19:  36%|███▌      | 35/97 [00:06<00:10,  5.79it/s, loss=3.88, v_num=0]
Epoch 19:  36%|███▌      | 35/97 [00:06<00:10,  5.73it/s, loss=3.88, v_num=0]
Epoch 19:  37%|███▋      | 36/97 [00:06<00:10,  5.82it/s, loss=3.88, v_num=0]
Epoch 19:  37%|███▋      | 36/97 [00:06<00:10,  5.77it/s, loss=3.88, v_num=0]
Epoch 19:  38%|███▊      | 37/97 [00:06<00:10,  5.85it/s, loss=3.88, v_num=0]
Epoch 19:  38%|███▊      | 37/97 [00:06<00:10,  5.80it/s, loss=3.88, v_num=0]
Epoch 19:  39%|███▉      | 38/97 [00:06<00:10,  5.88it/s, loss=3.88, v_num=0]
Epoch 19:  39%|███▉      | 38/97 [00:06<00:10,  5.83it/s, loss=3.88, v_num=0]
Epoch 19:  40%|████      | 39/97 [00:06<00:09,  5.91it/s, loss=3.88, v_num=0]
Epoch 19:  40%|████      | 39/97 [00:06<00:09,  5.86it/s, loss=3.88, v_num=0]
Epoch 19:  41%|████      | 40/97 [00:06<00:09,  5.94it/s, loss=3.88, v_num=0]
Epoch 19:  41%|████      | 40/97 [00:06<00:09,  5.89it/s, loss=3.89, v_num=0]
Epoch 19:  42%|████▏     | 41/97 [00:06<00:09,  5.97it/s, loss=3.89, v_num=0]
Epoch 19:  42%|████▏     | 41/97 [00:06<00:09,  5.92it/s, loss=3.88, v_num=0]
Epoch 19:  43%|████▎     | 42/97 [00:07<00:09,  5.99it/s, loss=3.88, v_num=0]
Epoch 19:  43%|████▎     | 42/97 [00:07<00:09,  5.95it/s, loss=3.89, v_num=0]
Epoch 19:  44%|████▍     | 43/97 [00:07<00:08,  6.02it/s, loss=3.89, v_num=0]
Epoch 19:  44%|████▍     | 43/97 [00:07<00:09,  5.97it/s, loss=3.88, v_num=0]
Epoch 19:  45%|████▌     | 44/97 [00:07<00:08,  6.04it/s, loss=3.88, v_num=0]
Epoch 19:  45%|████▌     | 44/97 [00:07<00:08,  5.99it/s, loss=3.88, v_num=0]
Epoch 19:  46%|████▋     | 45/97 [00:07<00:08,  6.06it/s, loss=3.88, v_num=0]
Epoch 19:  46%|████▋     | 45/97 [00:07<00:08,  6.01it/s, loss=3.87, v_num=0]
Epoch 19:  47%|████▋     | 46/97 [00:07<00:08,  6.08it/s, loss=3.87, v_num=0]
Epoch 19:  47%|████▋     | 46/97 [00:07<00:08,  6.03it/s, loss=3.86, v_num=0]
Epoch 19:  48%|████▊     | 47/97 [00:07<00:08,  6.10it/s, loss=3.86, v_num=0]
Epoch 19:  48%|████▊     | 47/97 [00:07<00:08,  6.06it/s, loss=3.86, v_num=0]
Epoch 19:  49%|████▉     | 48/97 [00:07<00:08,  6.12it/s, loss=3.86, v_num=0]
Epoch 19:  49%|████▉     | 48/97 [00:07<00:08,  6.08it/s, loss=3.87, v_num=0]
Epoch 19:  51%|█████     | 49/97 [00:07<00:07,  6.14it/s, loss=3.87, v_num=0]
Epoch 19:  51%|█████     | 49/97 [00:08<00:07,  6.10it/s, loss=3.87, v_num=0]
Epoch 19:  52%|█████▏    | 50/97 [00:08<00:07,  6.16it/s, loss=3.87, v_num=0]
Epoch 19:  52%|█████▏    | 50/97 [00:08<00:07,  6.12it/s, loss=3.87, v_num=0]
Epoch 19:  53%|█████▎    | 51/97 [00:08<00:07,  6.18it/s, loss=3.87, v_num=0]
Epoch 19:  53%|█████▎    | 51/97 [00:08<00:07,  6.14it/s, loss=3.86, v_num=0]
Epoch 19:  54%|█████▎    | 52/97 [00:08<00:07,  6.20it/s, loss=3.86, v_num=0]
Epoch 19:  54%|█████▎    | 52/97 [00:08<00:07,  6.16it/s, loss=3.87, v_num=0]
Epoch 19:  55%|█████▍    | 53/97 [00:08<00:07,  6.21it/s, loss=3.87, v_num=0]
Epoch 19:  55%|█████▍    | 53/97 [00:08<00:07,  6.17it/s, loss=3.86, v_num=0]
Epoch 19:  56%|█████▌    | 54/97 [00:08<00:06,  6.23it/s, loss=3.86, v_num=0]
Epoch 19:  56%|█████▌    | 54/97 [00:08<00:06,  6.19it/s, loss=3.87, v_num=0]
Epoch 19:  57%|█████▋    | 55/97 [00:08<00:06,  6.24it/s, loss=3.87, v_num=0]
Epoch 19:  57%|█████▋    | 55/97 [00:08<00:06,  6.20it/s, loss=3.86, v_num=0]
Epoch 19:  58%|█████▊    | 56/97 [00:08<00:06,  6.26it/s, loss=3.86, v_num=0]
Epoch 19:  58%|█████▊    | 56/97 [00:09<00:06,  6.22it/s, loss=3.85, v_num=0]
Epoch 19:  59%|█████▉    | 57/97 [00:09<00:06,  6.27it/s, loss=3.85, v_num=0]
Epoch 19:  59%|█████▉    | 57/97 [00:09<00:06,  6.23it/s, loss=3.86, v_num=0]
Epoch 19:  60%|█████▉    | 58/97 [00:09<00:06,  6.29it/s, loss=3.86, v_num=0]
Epoch 19:  60%|█████▉    | 58/97 [00:09<00:06,  6.25it/s, loss=3.85, v_num=0]
Epoch 19:  61%|██████    | 59/97 [00:09<00:06,  6.30it/s, loss=3.85, v_num=0]
Epoch 19:  61%|██████    | 59/97 [00:09<00:06,  6.26it/s, loss=3.85, v_num=0]
Epoch 19:  62%|██████▏   | 60/97 [00:09<00:05,  6.32it/s, loss=3.85, v_num=0]
Epoch 19:  62%|██████▏   | 60/97 [00:09<00:05,  6.28it/s, loss=3.85, v_num=0]
Epoch 19:  63%|██████▎   | 61/97 [00:09<00:05,  6.33it/s, loss=3.85, v_num=0]
Epoch 19:  63%|██████▎   | 61/97 [00:09<00:05,  6.29it/s, loss=3.84, v_num=0]
Epoch 19:  64%|██████▍   | 62/97 [00:09<00:05,  6.34it/s, loss=3.84, v_num=0]
Epoch 19:  64%|██████▍   | 62/97 [00:09<00:05,  6.31it/s, loss=3.84, v_num=0]
Epoch 19:  65%|██████▍   | 63/97 [00:09<00:05,  6.36it/s, loss=3.84, v_num=0]
Epoch 19:  65%|██████▍   | 63/97 [00:09<00:05,  6.32it/s, loss=3.85, v_num=0]
Epoch 19:  66%|██████▌   | 64/97 [00:10<00:05,  6.37it/s, loss=3.85, v_num=0]
Epoch 19:  66%|██████▌   | 64/97 [00:10<00:05,  6.33it/s, loss=3.85, v_num=0]
Epoch 19:  67%|██████▋   | 65/97 [00:10<00:05,  6.38it/s, loss=3.85, v_num=0]
Epoch 19:  67%|██████▋   | 65/97 [00:10<00:05,  6.35it/s, loss=3.85, v_num=0]
Epoch 19:  68%|██████▊   | 66/97 [00:10<00:04,  6.39it/s, loss=3.85, v_num=0]
Epoch 19:  68%|██████▊   | 66/97 [00:10<00:04,  6.36it/s, loss=3.85, v_num=0]
Epoch 19:  69%|██████▉   | 67/97 [00:10<00:04,  6.40it/s, loss=3.85, v_num=0]
Epoch 19:  69%|██████▉   | 67/97 [00:10<00:04,  6.37it/s, loss=3.86, v_num=0]
Epoch 19:  70%|███████   | 68/97 [00:10<00:04,  6.42it/s, loss=3.86, v_num=0]
Epoch 19:  70%|███████   | 68/97 [00:10<00:04,  6.38it/s, loss=3.84, v_num=0]
Epoch 19:  71%|███████   | 69/97 [00:10<00:04,  6.43it/s, loss=3.84, v_num=0]
Epoch 19:  71%|███████   | 69/97 [00:10<00:04,  6.39it/s, loss=3.84, v_num=0]
Epoch 19:  72%|███████▏  | 70/97 [00:10<00:04,  6.44it/s, loss=3.84, v_num=0]
Epoch 19:  72%|███████▏  | 70/97 [00:10<00:04,  6.41it/s, loss=3.84, v_num=0]
Epoch 19:  73%|███████▎  | 71/97 [00:11<00:04,  6.45it/s, loss=3.84, v_num=0]
Epoch 19:  73%|███████▎  | 71/97 [00:11<00:04,  6.42it/s, loss=3.85, v_num=0]
Epoch 19:  74%|███████▍  | 72/97 [00:11<00:03,  6.46it/s, loss=3.85, v_num=0]
Epoch 19:  74%|███████▍  | 72/97 [00:11<00:03,  6.43it/s, loss=3.85, v_num=0]
Epoch 19:  75%|███████▌  | 73/97 [00:11<00:03,  6.47it/s, loss=3.85, v_num=0]
Epoch 19:  75%|███████▌  | 73/97 [00:11<00:03,  6.44it/s, loss=3.86, v_num=0]
Epoch 19:  76%|███████▋  | 74/97 [00:11<00:03,  6.48it/s, loss=3.86, v_num=0]
Epoch 19:  76%|███████▋  | 74/97 [00:11<00:03,  6.45it/s, loss=3.85, v_num=0]
Epoch 19:  77%|███████▋  | 75/97 [00:11<00:03,  6.49it/s, loss=3.85, v_num=0]
Epoch 19:  77%|███████▋  | 75/97 [00:11<00:03,  6.46it/s, loss=3.86, v_num=0]
Epoch 19:  78%|███████▊  | 76/97 [00:11<00:03,  6.50it/s, loss=3.86, v_num=0]
Epoch 19:  78%|███████▊  | 76/97 [00:11<00:03,  6.47it/s, loss=3.85, v_num=0]
Epoch 19:  79%|███████▉  | 77/97 [00:11<00:03,  6.51it/s, loss=3.85, v_num=0]
Epoch 19:  79%|███████▉  | 77/97 [00:11<00:03,  6.48it/s, loss=3.85, v_num=0]
Epoch 19:  80%|████████  | 78/97 [00:11<00:02,  6.51it/s, loss=3.85, v_num=0]
Epoch 19:  80%|████████  | 78/97 [00:12<00:02,  6.48it/s, loss=3.84, v_num=0]
Epoch 19:  81%|████████▏ | 79/97 [00:12<00:02,  6.52it/s, loss=3.84, v_num=0]
Epoch 19:  81%|████████▏ | 79/97 [00:12<00:02,  6.49it/s, loss=3.84, v_num=0]
Epoch 19:  82%|████████▏ | 80/97 [00:12<00:02,  6.53it/s, loss=3.84, v_num=0]
Epoch 19:  82%|████████▏ | 80/97 [00:12<00:02,  6.50it/s, loss=3.83, v_num=0]
Epoch 19:  84%|████████▎ | 81/97 [00:12<00:02,  6.54it/s, loss=3.83, v_num=0]
Epoch 19:  84%|████████▎ | 81/97 [00:12<00:02,  6.51it/s, loss=3.84, v_num=0]
Epoch 19:  85%|████████▍ | 82/97 [00:12<00:02,  6.55it/s, loss=3.84, v_num=0]
Epoch 19:  85%|████████▍ | 82/97 [00:12<00:02,  6.52it/s, loss=3.84, v_num=0]
Epoch 19:  86%|████████▌ | 83/97 [00:12<00:02,  6.56it/s, loss=3.84, v_num=0]
Epoch 19:  86%|████████▌ | 83/97 [00:12<00:02,  6.53it/s, loss=3.84, v_num=0]
Epoch 19:  87%|████████▋ | 84/97 [00:12<00:01,  6.56it/s, loss=3.84, v_num=0]
Epoch 19:  87%|████████▋ | 84/97 [00:12<00:01,  6.54it/s, loss=3.84, v_num=0]
Epoch 19:  88%|████████▊ | 85/97 [00:12<00:01,  6.57it/s, loss=3.84, v_num=0]
Epoch 19:  88%|████████▊ | 85/97 [00:12<00:01,  6.54it/s, loss=3.84, v_num=0]
Epoch 19:  89%|████████▊ | 86/97 [00:13<00:01,  6.58it/s, loss=3.84, v_num=0]
Epoch 19:  89%|████████▊ | 86/97 [00:13<00:01,  6.55it/s, loss=3.84, v_num=0]
Epoch 19:  90%|████████▉ | 87/97 [00:13<00:01,  6.59it/s, loss=3.84, v_num=0]
Epoch 19:  90%|████████▉ | 87/97 [00:13<00:01,  6.56it/s, loss=3.84, v_num=0]
Epoch 19:  91%|█████████ | 88/97 [00:13<00:01,  6.59it/s, loss=3.84, v_num=0]
Epoch 19:  91%|█████████ | 88/97 [00:13<00:01,  6.57it/s, loss=3.84, v_num=0]
Epoch 19:  92%|█████████▏| 89/97 [00:13<00:01,  6.60it/s, loss=3.84, v_num=0]
Epoch 19:  92%|█████████▏| 89/97 [00:13<00:01,  6.57it/s, loss=3.84, v_num=0]
Epoch 19:  93%|█████████▎| 90/97 [00:13<00:01,  6.61it/s, loss=3.84, v_num=0]
Epoch 19:  93%|█████████▎| 90/97 [00:13<00:01,  6.58it/s, loss=3.84, v_num=0]
Epoch 19:  94%|█████████▍| 91/97 [00:13<00:00,  6.62it/s, loss=3.84, v_num=0]
Epoch 19:  94%|█████████▍| 91/97 [00:13<00:00,  6.59it/s, loss=3.84, v_num=0]
Epoch 19:  95%|█████████▍| 92/97 [00:13<00:00,  6.62it/s, loss=3.84, v_num=0]
Epoch 19:  95%|█████████▍| 92/97 [00:13<00:00,  6.60it/s, loss=3.83, v_num=0]
Epoch 19:  96%|█████████▌| 93/97 [00:14<00:00,  6.63it/s, loss=3.83, v_num=0]
Epoch 19:  96%|█████████▌| 93/97 [00:14<00:00,  6.60it/s, loss=3.83, v_num=0]
Epoch 19:  97%|█████████▋| 94/97 [00:14<00:00,  6.64it/s, loss=3.83, v_num=0]
Epoch 19:  97%|█████████▋| 94/97 [00:14<00:00,  6.61it/s, loss=3.83, v_num=0]
Epoch 19:  98%|█████████▊| 95/97 [00:14<00:00,  6.64it/s, loss=3.83, v_num=0]
Epoch 19:  98%|█████████▊| 95/97 [00:14<00:00,  6.62it/s, loss=3.82, v_num=0]
Epoch 19:  99%|█████████▉| 96/97 [00:14<00:00,  6.65it/s, loss=3.82, v_num=0]
Epoch 19:  99%|█████████▉| 96/97 [00:14<00:00,  6.62it/s, loss=3.83, v_num=0]
Epoch 19: 100%|██████████| 97/97 [00:14<00:00,  6.66it/s, loss=3.83, v_num=0]
Epoch 19: 100%|██████████| 97/97 [00:14<00:00,  6.63it/s, loss=3.83, v_num=0]
Epoch 19: 100%|██████████| 97/97 [00:14<00:00,  6.63it/s, loss=3.83, v_num=0]
Epoch 19:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.83, v_num=0]
Epoch 20:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.83, v_num=0]
Epoch 20:   1%|          | 1/97 [00:01<02:13,  1.39s/it, loss=3.83, v_num=0]
Epoch 20:   1%|          | 1/97 [00:01<02:18,  1.44s/it, loss=3.83, v_num=0]
Epoch 20:   2%|▏         | 2/97 [00:01<01:12,  1.31it/s, loss=3.83, v_num=0]
Epoch 20:   2%|▏         | 2/97 [00:01<01:15,  1.27it/s, loss=3.84, v_num=0]
Epoch 20:   3%|▎         | 3/97 [00:01<00:52,  1.80it/s, loss=3.84, v_num=0]
Epoch 20:   3%|▎         | 3/97 [00:01<00:53,  1.74it/s, loss=3.84, v_num=0]
Epoch 20:   4%|▍         | 4/97 [00:01<00:41,  2.22it/s, loss=3.84, v_num=0]
Epoch 20:   4%|▍         | 4/97 [00:01<00:43,  2.15it/s, loss=3.84, v_num=0]
Epoch 20:   5%|▌         | 5/97 [00:01<00:35,  2.58it/s, loss=3.84, v_num=0]
Epoch 20:   5%|▌         | 5/97 [00:01<00:36,  2.51it/s, loss=3.84, v_num=0]
Epoch 20:   6%|▌         | 6/97 [00:02<00:31,  2.90it/s, loss=3.84, v_num=0]
Epoch 20:   6%|▌         | 6/97 [00:02<00:32,  2.82it/s, loss=3.84, v_num=0]
Epoch 20:   7%|▋         | 7/97 [00:02<00:28,  3.16it/s, loss=3.84, v_num=0]
Epoch 20:   7%|▋         | 7/97 [00:02<00:29,  3.09it/s, loss=3.83, v_num=0]
Epoch 20:   8%|▊         | 8/97 [00:02<00:26,  3.41it/s, loss=3.83, v_num=0]
Epoch 20:   8%|▊         | 8/97 [00:02<00:26,  3.33it/s, loss=3.83, v_num=0]
Epoch 20:   9%|▉         | 9/97 [00:02<00:24,  3.61it/s, loss=3.83, v_num=0]
Epoch 20:   9%|▉         | 9/97 [00:02<00:24,  3.53it/s, loss=3.83, v_num=0]
Epoch 20:  10%|█         | 10/97 [00:02<00:22,  3.81it/s, loss=3.83, v_num=0]
Epoch 20:  10%|█         | 10/97 [00:02<00:23,  3.73it/s, loss=3.82, v_num=0]
Epoch 20:  11%|█▏        | 11/97 [00:02<00:21,  3.98it/s, loss=3.82, v_num=0]
Epoch 20:  11%|█▏        | 11/97 [00:02<00:22,  3.90it/s, loss=3.82, v_num=0]
Epoch 20:  12%|█▏        | 12/97 [00:02<00:20,  4.13it/s, loss=3.82, v_num=0]
Epoch 20:  12%|█▏        | 12/97 [00:02<00:20,  4.06it/s, loss=3.83, v_num=0]
Epoch 20:  13%|█▎        | 13/97 [00:03<00:19,  4.28it/s, loss=3.83, v_num=0]
Epoch 20:  13%|█▎        | 13/97 [00:03<00:19,  4.20it/s, loss=3.83, v_num=0]
Epoch 20:  14%|█▍        | 14/97 [00:03<00:18,  4.41it/s, loss=3.83, v_num=0]
Epoch 20:  14%|█▍        | 14/97 [00:03<00:19,  4.34it/s, loss=3.82, v_num=0]
Epoch 20:  15%|█▌        | 15/97 [00:03<00:18,  4.54it/s, loss=3.82, v_num=0]
Epoch 20:  15%|█▌        | 15/97 [00:03<00:18,  4.46it/s, loss=3.82, v_num=0]
Epoch 20:  16%|█▋        | 16/97 [00:03<00:17,  4.65it/s, loss=3.82, v_num=0]
Epoch 20:  16%|█▋        | 16/97 [00:03<00:17,  4.57it/s, loss=3.83, v_num=0]
Epoch 20:  18%|█▊        | 17/97 [00:03<00:16,  4.74it/s, loss=3.83, v_num=0]
Epoch 20:  18%|█▊        | 17/97 [00:03<00:17,  4.67it/s, loss=3.82, v_num=0]
Epoch 20:  19%|█▊        | 18/97 [00:03<00:16,  4.83it/s, loss=3.82, v_num=0]
Epoch 20:  19%|█▊        | 18/97 [00:03<00:16,  4.76it/s, loss=3.81, v_num=0]
Epoch 20:  20%|█▉        | 19/97 [00:03<00:15,  4.93it/s, loss=3.81, v_num=0]
Epoch 20:  20%|█▉        | 19/97 [00:03<00:16,  4.86it/s, loss=3.81, v_num=0]
Epoch 20:  21%|██        | 20/97 [00:03<00:15,  5.01it/s, loss=3.81, v_num=0]
Epoch 20:  21%|██        | 20/97 [00:04<00:15,  4.94it/s, loss=3.81, v_num=0]
Epoch 20:  22%|██▏       | 21/97 [00:04<00:14,  5.08it/s, loss=3.81, v_num=0]
Epoch 20:  22%|██▏       | 21/97 [00:04<00:15,  5.02it/s, loss=3.81, v_num=0]
Epoch 20:  23%|██▎       | 22/97 [00:04<00:14,  5.15it/s, loss=3.81, v_num=0]
Epoch 20:  23%|██▎       | 22/97 [00:04<00:14,  5.08it/s, loss=3.8, v_num=0]
Epoch 20:  24%|██▎       | 23/97 [00:04<00:14,  5.21it/s, loss=3.8, v_num=0]
Epoch 20:  24%|██▎       | 23/97 [00:04<00:14,  5.15it/s, loss=3.8, v_num=0]
Epoch 20:  25%|██▍       | 24/97 [00:04<00:13,  5.28it/s, loss=3.8, v_num=0]
Epoch 20:  25%|██▍       | 24/97 [00:04<00:13,  5.22it/s, loss=3.79, v_num=0]
Epoch 20:  26%|██▌       | 25/97 [00:04<00:13,  5.34it/s, loss=3.79, v_num=0]
Epoch 20:  26%|██▌       | 25/97 [00:04<00:13,  5.27it/s, loss=3.78, v_num=0]
Epoch 20:  27%|██▋       | 26/97 [00:04<00:13,  5.40it/s, loss=3.78, v_num=0]
Epoch 20:  27%|██▋       | 26/97 [00:04<00:13,  5.33it/s, loss=3.78, v_num=0]
Epoch 20:  28%|██▊       | 27/97 [00:04<00:12,  5.45it/s, loss=3.78, v_num=0]
Epoch 20:  28%|██▊       | 27/97 [00:05<00:12,  5.39it/s, loss=3.78, v_num=0]
Epoch 20:  29%|██▉       | 28/97 [00:05<00:12,  5.50it/s, loss=3.78, v_num=0]
Epoch 20:  29%|██▉       | 28/97 [00:05<00:12,  5.44it/s, loss=3.78, v_num=0]
Epoch 20:  30%|██▉       | 29/97 [00:05<00:12,  5.54it/s, loss=3.78, v_num=0]
Epoch 20:  30%|██▉       | 29/97 [00:05<00:12,  5.48it/s, loss=3.79, v_num=0]
Epoch 20:  31%|███       | 30/97 [00:05<00:12,  5.58it/s, loss=3.79, v_num=0]
Epoch 20:  31%|███       | 30/97 [00:05<00:12,  5.53it/s, loss=3.79, v_num=0]
Epoch 20:  32%|███▏      | 31/97 [00:05<00:11,  5.62it/s, loss=3.79, v_num=0]
Epoch 20:  32%|███▏      | 31/97 [00:05<00:11,  5.57it/s, loss=3.79, v_num=0]
Epoch 20:  33%|███▎      | 32/97 [00:05<00:11,  5.67it/s, loss=3.79, v_num=0]
Epoch 20:  33%|███▎      | 32/97 [00:05<00:11,  5.61it/s, loss=3.79, v_num=0]
Epoch 20:  34%|███▍      | 33/97 [00:05<00:11,  5.70it/s, loss=3.79, v_num=0]
Epoch 20:  34%|███▍      | 33/97 [00:05<00:11,  5.65it/s, loss=3.78, v_num=0]
Epoch 20:  35%|███▌      | 34/97 [00:05<00:10,  5.74it/s, loss=3.78, v_num=0]
Epoch 20:  35%|███▌      | 34/97 [00:05<00:11,  5.69it/s, loss=3.78, v_num=0]
Epoch 20:  36%|███▌      | 35/97 [00:06<00:10,  5.78it/s, loss=3.78, v_num=0]
Epoch 20:  36%|███▌      | 35/97 [00:06<00:10,  5.72it/s, loss=3.78, v_num=0]
Epoch 20:  37%|███▋      | 36/97 [00:06<00:10,  5.81it/s, loss=3.78, v_num=0]
Epoch 20:  37%|███▋      | 36/97 [00:06<00:10,  5.76it/s, loss=3.78, v_num=0]
Epoch 20:  38%|███▊      | 37/97 [00:06<00:10,  5.84it/s, loss=3.78, v_num=0]
Epoch 20:  38%|███▊      | 37/97 [00:06<00:10,  5.79it/s, loss=3.79, v_num=0]
Epoch 20:  39%|███▉      | 38/97 [00:06<00:10,  5.88it/s, loss=3.79, v_num=0]
Epoch 20:  39%|███▉      | 38/97 [00:06<00:10,  5.82it/s, loss=3.79, v_num=0]
Epoch 20:  40%|████      | 39/97 [00:06<00:09,  5.90it/s, loss=3.79, v_num=0]
Epoch 20:  40%|████      | 39/97 [00:06<00:09,  5.85it/s, loss=3.79, v_num=0]
Epoch 20:  41%|████      | 40/97 [00:06<00:09,  5.94it/s, loss=3.79, v_num=0]
Epoch 20:  41%|████      | 40/97 [00:06<00:09,  5.88it/s, loss=3.79, v_num=0]
Epoch 20:  42%|████▏     | 41/97 [00:06<00:09,  5.96it/s, loss=3.79, v_num=0]
Epoch 20:  42%|████▏     | 41/97 [00:06<00:09,  5.91it/s, loss=3.79, v_num=0]
Epoch 20:  43%|████▎     | 42/97 [00:07<00:09,  5.99it/s, loss=3.79, v_num=0]
Epoch 20:  43%|████▎     | 42/97 [00:07<00:09,  5.94it/s, loss=3.8, v_num=0]
Epoch 20:  44%|████▍     | 43/97 [00:07<00:08,  6.01it/s, loss=3.8, v_num=0]
Epoch 20:  44%|████▍     | 43/97 [00:07<00:09,  5.97it/s, loss=3.81, v_num=0]
Epoch 20:  45%|████▌     | 44/97 [00:07<00:08,  6.04it/s, loss=3.81, v_num=0]
Epoch 20:  45%|████▌     | 44/97 [00:07<00:08,  5.99it/s, loss=3.82, v_num=0]
Epoch 20:  46%|████▋     | 45/97 [00:07<00:08,  6.06it/s, loss=3.82, v_num=0]
Epoch 20:  46%|████▋     | 45/97 [00:07<00:08,  6.01it/s, loss=3.82, v_num=0]
Epoch 20:  47%|████▋     | 46/97 [00:07<00:08,  6.08it/s, loss=3.82, v_num=0]
Epoch 20:  47%|████▋     | 46/97 [00:07<00:08,  6.04it/s, loss=3.82, v_num=0]
Epoch 20:  48%|████▊     | 47/97 [00:07<00:08,  6.10it/s, loss=3.82, v_num=0]
Epoch 20:  48%|████▊     | 47/97 [00:07<00:08,  6.06it/s, loss=3.81, v_num=0]
Epoch 20:  49%|████▉     | 48/97 [00:07<00:08,  6.12it/s, loss=3.81, v_num=0]
Epoch 20:  49%|████▉     | 48/97 [00:07<00:08,  6.08it/s, loss=3.81, v_num=0]
Epoch 20:  51%|█████     | 49/97 [00:07<00:07,  6.14it/s, loss=3.81, v_num=0]
Epoch 20:  51%|█████     | 49/97 [00:08<00:07,  6.10it/s, loss=3.8, v_num=0]
Epoch 20:  52%|█████▏    | 50/97 [00:08<00:07,  6.16it/s, loss=3.8, v_num=0]
Epoch 20:  52%|█████▏    | 50/97 [00:08<00:07,  6.12it/s, loss=3.79, v_num=0]
Epoch 20:  53%|█████▎    | 51/97 [00:08<00:07,  6.18it/s, loss=3.79, v_num=0]
Epoch 20:  53%|█████▎    | 51/97 [00:08<00:07,  6.14it/s, loss=3.79, v_num=0]
Epoch 20:  54%|█████▎    | 52/97 [00:08<00:07,  6.20it/s, loss=3.79, v_num=0]
Epoch 20:  54%|█████▎    | 52/97 [00:08<00:07,  6.16it/s, loss=3.78, v_num=0]
Epoch 20:  55%|█████▍    | 53/97 [00:08<00:07,  6.22it/s, loss=3.78, v_num=0]
Epoch 20:  55%|█████▍    | 53/97 [00:08<00:07,  6.18it/s, loss=3.78, v_num=0]
Epoch 20:  56%|█████▌    | 54/97 [00:08<00:06,  6.23it/s, loss=3.78, v_num=0]
Epoch 20:  56%|█████▌    | 54/97 [00:08<00:06,  6.19it/s, loss=3.78, v_num=0]
Epoch 20:  57%|█████▋    | 55/97 [00:08<00:06,  6.25it/s, loss=3.78, v_num=0]
Epoch 20:  57%|█████▋    | 55/97 [00:08<00:06,  6.21it/s, loss=3.78, v_num=0]
Epoch 20:  58%|█████▊    | 56/97 [00:08<00:06,  6.27it/s, loss=3.78, v_num=0]
Epoch 20:  58%|█████▊    | 56/97 [00:08<00:06,  6.23it/s, loss=3.78, v_num=0]
Epoch 20:  59%|█████▉    | 57/97 [00:09<00:06,  6.28it/s, loss=3.78, v_num=0]
Epoch 20:  59%|█████▉    | 57/97 [00:09<00:06,  6.24it/s, loss=3.78, v_num=0]
Epoch 20:  60%|█████▉    | 58/97 [00:09<00:06,  6.30it/s, loss=3.78, v_num=0]
Epoch 20:  60%|█████▉    | 58/97 [00:09<00:06,  6.26it/s, loss=3.79, v_num=0]
Epoch 20:  61%|██████    | 59/97 [00:09<00:06,  6.31it/s, loss=3.79, v_num=0]
Epoch 20:  61%|██████    | 59/97 [00:09<00:06,  6.27it/s, loss=3.78, v_num=0]
Epoch 20:  62%|██████▏   | 60/97 [00:09<00:05,  6.33it/s, loss=3.78, v_num=0]
Epoch 20:  62%|██████▏   | 60/97 [00:09<00:05,  6.29it/s, loss=3.79, v_num=0]
Epoch 20:  63%|██████▎   | 61/97 [00:09<00:05,  6.34it/s, loss=3.79, v_num=0]
Epoch 20:  63%|██████▎   | 61/97 [00:09<00:05,  6.30it/s, loss=3.79, v_num=0]
Epoch 20:  64%|██████▍   | 62/97 [00:09<00:05,  6.36it/s, loss=3.79, v_num=0]
Epoch 20:  64%|██████▍   | 62/97 [00:09<00:05,  6.32it/s, loss=3.78, v_num=0]
Epoch 20:  65%|██████▍   | 63/97 [00:09<00:05,  6.37it/s, loss=3.78, v_num=0]
Epoch 20:  65%|██████▍   | 63/97 [00:09<00:05,  6.33it/s, loss=3.78, v_num=0]
Epoch 20:  66%|██████▌   | 64/97 [00:10<00:05,  6.38it/s, loss=3.78, v_num=0]
Epoch 20:  66%|██████▌   | 64/97 [00:10<00:05,  6.34it/s, loss=3.77, v_num=0]
Epoch 20:  67%|██████▋   | 65/97 [00:10<00:05,  6.39it/s, loss=3.77, v_num=0]
Epoch 20:  67%|██████▋   | 65/97 [00:10<00:05,  6.36it/s, loss=3.77, v_num=0]
Epoch 20:  68%|██████▊   | 66/97 [00:10<00:04,  6.40it/s, loss=3.77, v_num=0]
Epoch 20:  68%|██████▊   | 66/97 [00:10<00:04,  6.37it/s, loss=3.78, v_num=0]
Epoch 20:  69%|██████▉   | 67/97 [00:10<00:04,  6.42it/s, loss=3.78, v_num=0]
Epoch 20:  69%|██████▉   | 67/97 [00:10<00:04,  6.38it/s, loss=3.78, v_num=0]
Epoch 20:  70%|███████   | 68/97 [00:10<00:04,  6.43it/s, loss=3.78, v_num=0]
Epoch 20:  70%|███████   | 68/97 [00:10<00:04,  6.39it/s, loss=3.78, v_num=0]
Epoch 20:  71%|███████   | 69/97 [00:10<00:04,  6.44it/s, loss=3.78, v_num=0]
Epoch 20:  71%|███████   | 69/97 [00:10<00:04,  6.41it/s, loss=3.79, v_num=0]
Epoch 20:  72%|███████▏  | 70/97 [00:10<00:04,  6.45it/s, loss=3.79, v_num=0]
Epoch 20:  72%|███████▏  | 70/97 [00:10<00:04,  6.42it/s, loss=3.8, v_num=0]
Epoch 20:  73%|███████▎  | 71/97 [00:10<00:04,  6.46it/s, loss=3.8, v_num=0]
Epoch 20:  73%|███████▎  | 71/97 [00:11<00:04,  6.43it/s, loss=3.8, v_num=0]
Epoch 20:  74%|███████▍  | 72/97 [00:11<00:03,  6.47it/s, loss=3.8, v_num=0]
Epoch 20:  74%|███████▍  | 72/97 [00:11<00:03,  6.44it/s, loss=3.81, v_num=0]
Epoch 20:  75%|███████▌  | 73/97 [00:11<00:03,  6.48it/s, loss=3.81, v_num=0]
Epoch 20:  75%|███████▌  | 73/97 [00:11<00:03,  6.45it/s, loss=3.81, v_num=0]
Epoch 20:  76%|███████▋  | 74/97 [00:11<00:03,  6.49it/s, loss=3.81, v_num=0]
Epoch 20:  76%|███████▋  | 74/97 [00:11<00:03,  6.46it/s, loss=3.8, v_num=0]
Epoch 20:  77%|███████▋  | 75/97 [00:11<00:03,  6.50it/s, loss=3.8, v_num=0]
Epoch 20:  77%|███████▋  | 75/97 [00:11<00:03,  6.47it/s, loss=3.79, v_num=0]
Epoch 20:  78%|███████▊  | 76/97 [00:11<00:03,  6.51it/s, loss=3.79, v_num=0]
Epoch 20:  78%|███████▊  | 76/97 [00:11<00:03,  6.48it/s, loss=3.79, v_num=0]
Epoch 20:  79%|███████▉  | 77/97 [00:11<00:03,  6.52it/s, loss=3.79, v_num=0]
Epoch 20:  79%|███████▉  | 77/97 [00:11<00:03,  6.49it/s, loss=3.79, v_num=0]
Epoch 20:  80%|████████  | 78/97 [00:11<00:02,  6.53it/s, loss=3.79, v_num=0]
Epoch 20:  80%|████████  | 78/97 [00:12<00:02,  6.50it/s, loss=3.78, v_num=0]
Epoch 20:  81%|████████▏ | 79/97 [00:12<00:02,  6.54it/s, loss=3.78, v_num=0]
Epoch 20:  81%|████████▏ | 79/97 [00:12<00:02,  6.51it/s, loss=3.78, v_num=0]
Epoch 20:  82%|████████▏ | 80/97 [00:12<00:02,  6.54it/s, loss=3.78, v_num=0]
Epoch 20:  82%|████████▏ | 80/97 [00:12<00:02,  6.51it/s, loss=3.77, v_num=0]
Epoch 20:  84%|████████▎ | 81/97 [00:12<00:02,  6.55it/s, loss=3.77, v_num=0]
Epoch 20:  84%|████████▎ | 81/97 [00:12<00:02,  6.52it/s, loss=3.76, v_num=0]
Epoch 20:  85%|████████▍ | 82/97 [00:12<00:02,  6.56it/s, loss=3.76, v_num=0]
Epoch 20:  85%|████████▍ | 82/97 [00:12<00:02,  6.53it/s, loss=3.76, v_num=0]
Epoch 20:  86%|████████▌ | 83/97 [00:12<00:02,  6.57it/s, loss=3.76, v_num=0]
Epoch 20:  86%|████████▌ | 83/97 [00:12<00:02,  6.54it/s, loss=3.75, v_num=0]
Epoch 20:  87%|████████▋ | 84/97 [00:12<00:01,  6.57it/s, loss=3.75, v_num=0]
Epoch 20:  87%|████████▋ | 84/97 [00:12<00:01,  6.54it/s, loss=3.76, v_num=0]
Epoch 20:  88%|████████▊ | 85/97 [00:12<00:01,  6.58it/s, loss=3.76, v_num=0]
Epoch 20:  88%|████████▊ | 85/97 [00:12<00:01,  6.55it/s, loss=3.76, v_num=0]
Epoch 20:  89%|████████▊ | 86/97 [00:13<00:01,  6.59it/s, loss=3.76, v_num=0]
Epoch 20:  89%|████████▊ | 86/97 [00:13<00:01,  6.56it/s, loss=3.75, v_num=0]
Epoch 20:  90%|████████▉ | 87/97 [00:13<00:01,  6.60it/s, loss=3.75, v_num=0]
Epoch 20:  90%|████████▉ | 87/97 [00:13<00:01,  6.57it/s, loss=3.75, v_num=0]
Epoch 20:  91%|█████████ | 88/97 [00:13<00:01,  6.60it/s, loss=3.75, v_num=0]
Epoch 20:  91%|█████████ | 88/97 [00:13<00:01,  6.58it/s, loss=3.75, v_num=0]
Epoch 20:  92%|█████████▏| 89/97 [00:13<00:01,  6.61it/s, loss=3.75, v_num=0]
Epoch 20:  92%|█████████▏| 89/97 [00:13<00:01,  6.58it/s, loss=3.75, v_num=0]
Epoch 20:  93%|█████████▎| 90/97 [00:13<00:01,  6.62it/s, loss=3.75, v_num=0]
Epoch 20:  93%|█████████▎| 90/97 [00:13<00:01,  6.59it/s, loss=3.75, v_num=0]
Epoch 20:  94%|█████████▍| 91/97 [00:13<00:00,  6.62it/s, loss=3.75, v_num=0]
Epoch 20:  94%|█████████▍| 91/97 [00:13<00:00,  6.60it/s, loss=3.75, v_num=0]
Epoch 20:  95%|█████████▍| 92/97 [00:13<00:00,  6.63it/s, loss=3.75, v_num=0]
Epoch 20:  95%|█████████▍| 92/97 [00:13<00:00,  6.60it/s, loss=3.75, v_num=0]
Epoch 20:  96%|█████████▌| 93/97 [00:14<00:00,  6.64it/s, loss=3.75, v_num=0]
Epoch 20:  96%|█████████▌| 93/97 [00:14<00:00,  6.61it/s, loss=3.75, v_num=0]
Epoch 20:  97%|█████████▋| 94/97 [00:14<00:00,  6.64it/s, loss=3.75, v_num=0]
Epoch 20:  97%|█████████▋| 94/97 [00:14<00:00,  6.62it/s, loss=3.75, v_num=0]
Epoch 20:  98%|█████████▊| 95/97 [00:14<00:00,  6.65it/s, loss=3.75, v_num=0]
Epoch 20:  98%|█████████▊| 95/97 [00:14<00:00,  6.62it/s, loss=3.76, v_num=0]
Epoch 20:  99%|█████████▉| 96/97 [00:14<00:00,  6.66it/s, loss=3.76, v_num=0]
Epoch 20:  99%|█████████▉| 96/97 [00:14<00:00,  6.63it/s, loss=3.76, v_num=0]
Epoch 20: 100%|██████████| 97/97 [00:14<00:00,  6.66it/s, loss=3.76, v_num=0]
Epoch 20: 100%|██████████| 97/97 [00:14<00:00,  6.64it/s, loss=3.76, v_num=0]
Epoch 20: 100%|██████████| 97/97 [00:14<00:00,  6.64it/s, loss=3.76, v_num=0]
Epoch 20:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.76, v_num=0]
Epoch 21:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.76, v_num=0]
Epoch 21:   1%|          | 1/97 [00:01<02:12,  1.38s/it, loss=3.76, v_num=0]
Epoch 21:   1%|          | 1/97 [00:01<02:17,  1.43s/it, loss=3.76, v_num=0]
Epoch 21:   2%|▏         | 2/97 [00:01<01:12,  1.31it/s, loss=3.76, v_num=0]
Epoch 21:   2%|▏         | 2/97 [00:01<01:14,  1.27it/s, loss=3.77, v_num=0]
Epoch 21:   3%|▎         | 3/97 [00:01<00:51,  1.81it/s, loss=3.77, v_num=0]
Epoch 21:   3%|▎         | 3/97 [00:01<00:53,  1.75it/s, loss=3.78, v_num=0]
Epoch 21:   4%|▍         | 4/97 [00:01<00:41,  2.23it/s, loss=3.78, v_num=0]
Epoch 21:   4%|▍         | 4/97 [00:01<00:43,  2.16it/s, loss=3.79, v_num=0]
Epoch 21:   5%|▌         | 5/97 [00:01<00:35,  2.59it/s, loss=3.79, v_num=0]
Epoch 21:   5%|▌         | 5/97 [00:01<00:36,  2.52it/s, loss=3.78, v_num=0]
Epoch 21:   6%|▌         | 6/97 [00:02<00:31,  2.89it/s, loss=3.78, v_num=0]
Epoch 21:   6%|▌         | 6/97 [00:02<00:32,  2.82it/s, loss=3.78, v_num=0]
Epoch 21:   7%|▋         | 7/97 [00:02<00:28,  3.17it/s, loss=3.78, v_num=0]
Epoch 21:   7%|▋         | 7/97 [00:02<00:29,  3.09it/s, loss=3.77, v_num=0]
Epoch 21:   8%|▊         | 8/97 [00:02<00:26,  3.41it/s, loss=3.77, v_num=0]
Epoch 21:   8%|▊         | 8/97 [00:02<00:26,  3.33it/s, loss=3.77, v_num=0]
Epoch 21:   9%|▉         | 9/97 [00:02<00:24,  3.62it/s, loss=3.77, v_num=0]
Epoch 21:   9%|▉         | 9/97 [00:02<00:24,  3.54it/s, loss=3.77, v_num=0]
Epoch 21:  10%|█         | 10/97 [00:02<00:22,  3.81it/s, loss=3.77, v_num=0]
Epoch 21:  10%|█         | 10/97 [00:02<00:23,  3.73it/s, loss=3.77, v_num=0]
Epoch 21:  11%|█▏        | 11/97 [00:02<00:21,  3.99it/s, loss=3.77, v_num=0]
Epoch 21:  11%|█▏        | 11/97 [00:02<00:22,  3.91it/s, loss=3.77, v_num=0]
Epoch 21:  12%|█▏        | 12/97 [00:02<00:20,  4.15it/s, loss=3.77, v_num=0]
Epoch 21:  12%|█▏        | 12/97 [00:02<00:20,  4.07it/s, loss=3.77, v_num=0]
Epoch 21:  13%|█▎        | 13/97 [00:03<00:19,  4.29it/s, loss=3.77, v_num=0]
Epoch 21:  13%|█▎        | 13/97 [00:03<00:19,  4.21it/s, loss=3.77, v_num=0]
Epoch 21:  14%|█▍        | 14/97 [00:03<00:18,  4.42it/s, loss=3.77, v_num=0]
Epoch 21:  14%|█▍        | 14/97 [00:03<00:19,  4.34it/s, loss=3.76, v_num=0]
Epoch 21:  15%|█▌        | 15/97 [00:03<00:18,  4.54it/s, loss=3.76, v_num=0]
Epoch 21:  15%|█▌        | 15/97 [00:03<00:18,  4.46it/s, loss=3.76, v_num=0]
Epoch 21:  16%|█▋        | 16/97 [00:03<00:17,  4.65it/s, loss=3.76, v_num=0]
Epoch 21:  16%|█▋        | 16/97 [00:03<00:17,  4.57it/s, loss=3.75, v_num=0]
Epoch 21:  18%|█▊        | 17/97 [00:03<00:16,  4.75it/s, loss=3.75, v_num=0]
Epoch 21:  18%|█▊        | 17/97 [00:03<00:17,  4.67it/s, loss=3.76, v_num=0]
Epoch 21:  19%|█▊        | 18/97 [00:03<00:16,  4.84it/s, loss=3.76, v_num=0]
Epoch 21:  19%|█▊        | 18/97 [00:03<00:16,  4.77it/s, loss=3.76, v_num=0]
Epoch 21:  20%|█▉        | 19/97 [00:03<00:15,  4.92it/s, loss=3.76, v_num=0]
Epoch 21:  20%|█▉        | 19/97 [00:03<00:16,  4.86it/s, loss=3.76, v_num=0]
Epoch 21:  21%|██        | 20/97 [00:03<00:15,  5.01it/s, loss=3.76, v_num=0]
Epoch 21:  21%|██        | 20/97 [00:04<00:15,  4.94it/s, loss=3.76, v_num=0]
Epoch 21:  22%|██▏       | 21/97 [00:04<00:14,  5.08it/s, loss=3.76, v_num=0]
Epoch 21:  22%|██▏       | 21/97 [00:04<00:15,  5.02it/s, loss=3.76, v_num=0]
Epoch 21:  23%|██▎       | 22/97 [00:04<00:14,  5.15it/s, loss=3.76, v_num=0]
Epoch 21:  23%|██▎       | 22/97 [00:04<00:14,  5.09it/s, loss=3.75, v_num=0]
Epoch 21:  24%|██▎       | 23/97 [00:04<00:14,  5.22it/s, loss=3.75, v_num=0]
Epoch 21:  24%|██▎       | 23/97 [00:04<00:14,  5.16it/s, loss=3.75, v_num=0]
Epoch 21:  25%|██▍       | 24/97 [00:04<00:13,  5.28it/s, loss=3.75, v_num=0]
Epoch 21:  25%|██▍       | 24/97 [00:04<00:13,  5.22it/s, loss=3.75, v_num=0]
Epoch 21:  26%|██▌       | 25/97 [00:04<00:13,  5.34it/s, loss=3.75, v_num=0]
Epoch 21:  26%|██▌       | 25/97 [00:04<00:13,  5.28it/s, loss=3.76, v_num=0]
Epoch 21:  27%|██▋       | 26/97 [00:04<00:13,  5.40it/s, loss=3.76, v_num=0]
Epoch 21:  27%|██▋       | 26/97 [00:04<00:13,  5.33it/s, loss=3.76, v_num=0]
Epoch 21:  28%|██▊       | 27/97 [00:04<00:12,  5.45it/s, loss=3.76, v_num=0]
Epoch 21:  28%|██▊       | 27/97 [00:05<00:12,  5.39it/s, loss=3.75, v_num=0]
Epoch 21:  29%|██▉       | 28/97 [00:05<00:12,  5.50it/s, loss=3.75, v_num=0]
Epoch 21:  29%|██▉       | 28/97 [00:05<00:12,  5.44it/s, loss=3.76, v_num=0]
Epoch 21:  30%|██▉       | 29/97 [00:05<00:12,  5.55it/s, loss=3.76, v_num=0]
Epoch 21:  30%|██▉       | 29/97 [00:05<00:12,  5.49it/s, loss=3.75, v_num=0]
Epoch 21:  31%|███       | 30/97 [00:05<00:11,  5.59it/s, loss=3.75, v_num=0]
Epoch 21:  31%|███       | 30/97 [00:05<00:12,  5.54it/s, loss=3.75, v_num=0]
Epoch 21:  32%|███▏      | 31/97 [00:05<00:11,  5.63it/s, loss=3.75, v_num=0]
Epoch 21:  32%|███▏      | 31/97 [00:05<00:11,  5.58it/s, loss=3.74, v_num=0]
Epoch 21:  33%|███▎      | 32/97 [00:05<00:11,  5.67it/s, loss=3.74, v_num=0]
Epoch 21:  33%|███▎      | 32/97 [00:05<00:11,  5.62it/s, loss=3.75, v_num=0]
Epoch 21:  34%|███▍      | 33/97 [00:05<00:11,  5.71it/s, loss=3.75, v_num=0]
Epoch 21:  34%|███▍      | 33/97 [00:05<00:11,  5.66it/s, loss=3.75, v_num=0]
Epoch 21:  35%|███▌      | 34/97 [00:05<00:10,  5.75it/s, loss=3.75, v_num=0]
Epoch 21:  35%|███▌      | 34/97 [00:05<00:11,  5.70it/s, loss=3.75, v_num=0]
Epoch 21:  36%|███▌      | 35/97 [00:06<00:10,  5.79it/s, loss=3.75, v_num=0]
Epoch 21:  36%|███▌      | 35/97 [00:06<00:10,  5.73it/s, loss=3.75, v_num=0]
Epoch 21:  37%|███▋      | 36/97 [00:06<00:10,  5.82it/s, loss=3.75, v_num=0]
Epoch 21:  37%|███▋      | 36/97 [00:06<00:10,  5.77it/s, loss=3.75, v_num=0]
Epoch 21:  38%|███▊      | 37/97 [00:06<00:10,  5.85it/s, loss=3.75, v_num=0]
Epoch 21:  38%|███▊      | 37/97 [00:06<00:10,  5.80it/s, loss=3.75, v_num=0]
Epoch 21:  39%|███▉      | 38/97 [00:06<00:10,  5.88it/s, loss=3.75, v_num=0]
Epoch 21:  39%|███▉      | 38/97 [00:06<00:10,  5.83it/s, loss=3.75, v_num=0]
Epoch 21:  40%|████      | 39/97 [00:06<00:09,  5.91it/s, loss=3.75, v_num=0]
Epoch 21:  40%|████      | 39/97 [00:06<00:09,  5.86it/s, loss=3.75, v_num=0]
Epoch 21:  41%|████      | 40/97 [00:06<00:09,  5.94it/s, loss=3.75, v_num=0]
Epoch 21:  41%|████      | 40/97 [00:06<00:09,  5.89it/s, loss=3.74, v_num=0]
Epoch 21:  42%|████▏     | 41/97 [00:06<00:09,  5.97it/s, loss=3.74, v_num=0]
Epoch 21:  42%|████▏     | 41/97 [00:06<00:09,  5.92it/s, loss=3.74, v_num=0]
Epoch 21:  43%|████▎     | 42/97 [00:07<00:09,  5.99it/s, loss=3.74, v_num=0]
Epoch 21:  43%|████▎     | 42/97 [00:07<00:09,  5.95it/s, loss=3.75, v_num=0]
Epoch 21:  44%|████▍     | 43/97 [00:07<00:08,  6.02it/s, loss=3.75, v_num=0]
Epoch 21:  44%|████▍     | 43/97 [00:07<00:09,  5.97it/s, loss=3.74, v_num=0]
Epoch 21:  45%|████▌     | 44/97 [00:07<00:08,  6.05it/s, loss=3.74, v_num=0]
Epoch 21:  45%|████▌     | 44/97 [00:07<00:08,  6.00it/s, loss=3.74, v_num=0]
Epoch 21:  46%|████▋     | 45/97 [00:07<00:08,  6.07it/s, loss=3.74, v_num=0]
Epoch 21:  46%|████▋     | 45/97 [00:07<00:08,  6.02it/s, loss=3.74, v_num=0]
Epoch 21:  47%|████▋     | 46/97 [00:07<00:08,  6.09it/s, loss=3.74, v_num=0]
Epoch 21:  47%|████▋     | 46/97 [00:07<00:08,  6.04it/s, loss=3.74, v_num=0]
Epoch 21:  48%|████▊     | 47/97 [00:07<00:08,  6.11it/s, loss=3.74, v_num=0]
Epoch 21:  48%|████▊     | 47/97 [00:07<00:08,  6.06it/s, loss=3.74, v_num=0]
Epoch 21:  49%|████▉     | 48/97 [00:07<00:08,  6.12it/s, loss=3.74, v_num=0]
Epoch 21:  49%|████▉     | 48/97 [00:07<00:08,  6.08it/s, loss=3.74, v_num=0]
Epoch 21:  51%|█████     | 49/97 [00:07<00:07,  6.14it/s, loss=3.74, v_num=0]
Epoch 21:  51%|█████     | 49/97 [00:08<00:07,  6.10it/s, loss=3.74, v_num=0]
Epoch 21:  52%|█████▏    | 50/97 [00:08<00:07,  6.16it/s, loss=3.74, v_num=0]
Epoch 21:  52%|█████▏    | 50/97 [00:08<00:07,  6.12it/s, loss=3.74, v_num=0]
Epoch 21:  53%|█████▎    | 51/97 [00:08<00:07,  6.18it/s, loss=3.74, v_num=0]
Epoch 21:  53%|█████▎    | 51/97 [00:08<00:07,  6.14it/s, loss=3.74, v_num=0]
Epoch 21:  54%|█████▎    | 52/97 [00:08<00:07,  6.20it/s, loss=3.74, v_num=0]
Epoch 21:  54%|█████▎    | 52/97 [00:08<00:07,  6.16it/s, loss=3.73, v_num=0]
Epoch 21:  55%|█████▍    | 53/97 [00:08<00:07,  6.22it/s, loss=3.73, v_num=0]
Epoch 21:  55%|█████▍    | 53/97 [00:08<00:07,  6.18it/s, loss=3.73, v_num=0]
Epoch 21:  56%|█████▌    | 54/97 [00:08<00:06,  6.24it/s, loss=3.73, v_num=0]
Epoch 21:  56%|█████▌    | 54/97 [00:08<00:06,  6.20it/s, loss=3.74, v_num=0]
Epoch 21:  57%|█████▋    | 55/97 [00:08<00:06,  6.26it/s, loss=3.74, v_num=0]
Epoch 21:  57%|█████▋    | 55/97 [00:08<00:06,  6.21it/s, loss=3.73, v_num=0]
Epoch 21:  58%|█████▊    | 56/97 [00:08<00:06,  6.27it/s, loss=3.73, v_num=0]
Epoch 21:  58%|█████▊    | 56/97 [00:08<00:06,  6.23it/s, loss=3.73, v_num=0]
Epoch 21:  59%|█████▉    | 57/97 [00:09<00:06,  6.28it/s, loss=3.73, v_num=0]
Epoch 21:  59%|█████▉    | 57/97 [00:09<00:06,  6.25it/s, loss=3.73, v_num=0]
Epoch 21:  60%|█████▉    | 58/97 [00:09<00:06,  6.30it/s, loss=3.73, v_num=0]
Epoch 21:  60%|█████▉    | 58/97 [00:09<00:06,  6.26it/s, loss=3.72, v_num=0]
Epoch 21:  61%|██████    | 59/97 [00:09<00:06,  6.31it/s, loss=3.72, v_num=0]
Epoch 21:  61%|██████    | 59/97 [00:09<00:06,  6.28it/s, loss=3.71, v_num=0]
Epoch 21:  62%|██████▏   | 60/97 [00:09<00:05,  6.33it/s, loss=3.71, v_num=0]
Epoch 21:  62%|██████▏   | 60/97 [00:09<00:05,  6.29it/s, loss=3.71, v_num=0]
Epoch 21:  63%|██████▎   | 61/97 [00:09<00:05,  6.34it/s, loss=3.71, v_num=0]
Epoch 21:  63%|██████▎   | 61/97 [00:09<00:05,  6.30it/s, loss=3.72, v_num=0]
Epoch 21:  64%|██████▍   | 62/97 [00:09<00:05,  6.35it/s, loss=3.72, v_num=0]
Epoch 21:  64%|██████▍   | 62/97 [00:09<00:05,  6.32it/s, loss=3.72, v_num=0]
Epoch 21:  65%|██████▍   | 63/97 [00:09<00:05,  6.37it/s, loss=3.72, v_num=0]
Epoch 21:  65%|██████▍   | 63/97 [00:09<00:05,  6.33it/s, loss=3.72, v_num=0]
Epoch 21:  66%|██████▌   | 64/97 [00:10<00:05,  6.38it/s, loss=3.72, v_num=0]
Epoch 21:  66%|██████▌   | 64/97 [00:10<00:05,  6.34it/s, loss=3.72, v_num=0]
Epoch 21:  67%|██████▋   | 65/97 [00:10<00:05,  6.39it/s, loss=3.72, v_num=0]
Epoch 21:  67%|██████▋   | 65/97 [00:10<00:05,  6.36it/s, loss=3.71, v_num=0]
Epoch 21:  68%|██████▊   | 66/97 [00:10<00:04,  6.40it/s, loss=3.71, v_num=0]
Epoch 21:  68%|██████▊   | 66/97 [00:10<00:04,  6.37it/s, loss=3.72, v_num=0]
Epoch 21:  69%|██████▉   | 67/97 [00:10<00:04,  6.42it/s, loss=3.72, v_num=0]
Epoch 21:  69%|██████▉   | 67/97 [00:10<00:04,  6.38it/s, loss=3.72, v_num=0]
Epoch 21:  70%|███████   | 68/97 [00:10<00:04,  6.43it/s, loss=3.72, v_num=0]
Epoch 21:  70%|███████   | 68/97 [00:10<00:04,  6.39it/s, loss=3.72, v_num=0]
Epoch 21:  71%|███████   | 69/97 [00:10<00:04,  6.44it/s, loss=3.72, v_num=0]
Epoch 21:  71%|███████   | 69/97 [00:10<00:04,  6.41it/s, loss=3.71, v_num=0]
Epoch 21:  72%|███████▏  | 70/97 [00:10<00:04,  6.45it/s, loss=3.71, v_num=0]
Epoch 21:  72%|███████▏  | 70/97 [00:10<00:04,  6.42it/s, loss=3.71, v_num=0]
Epoch 21:  73%|███████▎  | 71/97 [00:10<00:04,  6.46it/s, loss=3.71, v_num=0]
Epoch 21:  73%|███████▎  | 71/97 [00:11<00:04,  6.43it/s, loss=3.72, v_num=0]
Epoch 21:  74%|███████▍  | 72/97 [00:11<00:03,  6.47it/s, loss=3.72, v_num=0]
Epoch 21:  74%|███████▍  | 72/97 [00:11<00:03,  6.44it/s, loss=3.72, v_num=0]
Epoch 21:  75%|███████▌  | 73/97 [00:11<00:03,  6.48it/s, loss=3.72, v_num=0]
Epoch 21:  75%|███████▌  | 73/97 [00:11<00:03,  6.45it/s, loss=3.72, v_num=0]
Epoch 21:  76%|███████▋  | 74/97 [00:11<00:03,  6.49it/s, loss=3.72, v_num=0]
Epoch 21:  76%|███████▋  | 74/97 [00:11<00:03,  6.46it/s, loss=3.72, v_num=0]
Epoch 21:  77%|███████▋  | 75/97 [00:11<00:03,  6.50it/s, loss=3.72, v_num=0]
Epoch 21:  77%|███████▋  | 75/97 [00:11<00:03,  6.47it/s, loss=3.72, v_num=0]
Epoch 21:  78%|███████▊  | 76/97 [00:11<00:03,  6.51it/s, loss=3.72, v_num=0]
Epoch 21:  78%|███████▊  | 76/97 [00:11<00:03,  6.48it/s, loss=3.71, v_num=0]
Epoch 21:  79%|███████▉  | 77/97 [00:11<00:03,  6.52it/s, loss=3.71, v_num=0]
Epoch 21:  79%|███████▉  | 77/97 [00:11<00:03,  6.49it/s, loss=3.71, v_num=0]
Epoch 21:  80%|████████  | 78/97 [00:11<00:02,  6.53it/s, loss=3.71, v_num=0]
Epoch 21:  80%|████████  | 78/97 [00:12<00:02,  6.50it/s, loss=3.72, v_num=0]
Epoch 21:  81%|████████▏ | 79/97 [00:12<00:02,  6.54it/s, loss=3.72, v_num=0]
Epoch 21:  81%|████████▏ | 79/97 [00:12<00:02,  6.51it/s, loss=3.71, v_num=0]
Epoch 21:  82%|████████▏ | 80/97 [00:12<00:02,  6.55it/s, loss=3.71, v_num=0]
Epoch 21:  82%|████████▏ | 80/97 [00:12<00:02,  6.52it/s, loss=3.71, v_num=0]
Epoch 21:  84%|████████▎ | 81/97 [00:12<00:02,  6.55it/s, loss=3.71, v_num=0]
Epoch 21:  84%|████████▎ | 81/97 [00:12<00:02,  6.53it/s, loss=3.7, v_num=0]
Epoch 21:  85%|████████▍ | 82/97 [00:12<00:02,  6.56it/s, loss=3.7, v_num=0]
Epoch 21:  85%|████████▍ | 82/97 [00:12<00:02,  6.53it/s, loss=3.7, v_num=0]
Epoch 21:  86%|████████▌ | 83/97 [00:12<00:02,  6.56it/s, loss=3.7, v_num=0]
Epoch 21:  86%|████████▌ | 83/97 [00:12<00:02,  6.54it/s, loss=3.69, v_num=0]
Epoch 21:  87%|████████▋ | 84/97 [00:12<00:01,  6.57it/s, loss=3.69, v_num=0]
Epoch 21:  87%|████████▋ | 84/97 [00:12<00:01,  6.54it/s, loss=3.69, v_num=0]
Epoch 21:  88%|████████▊ | 85/97 [00:12<00:01,  6.58it/s, loss=3.69, v_num=0]
Epoch 21:  88%|████████▊ | 85/97 [00:12<00:01,  6.55it/s, loss=3.69, v_num=0]
Epoch 21:  89%|████████▊ | 86/97 [00:13<00:01,  6.59it/s, loss=3.69, v_num=0]
Epoch 21:  89%|████████▊ | 86/97 [00:13<00:01,  6.56it/s, loss=3.68, v_num=0]
Epoch 21:  90%|████████▉ | 87/97 [00:13<00:01,  6.60it/s, loss=3.68, v_num=0]
Epoch 21:  90%|████████▉ | 87/97 [00:13<00:01,  6.57it/s, loss=3.68, v_num=0]
Epoch 21:  91%|█████████ | 88/97 [00:13<00:01,  6.61it/s, loss=3.68, v_num=0]
Epoch 21:  91%|█████████ | 88/97 [00:13<00:01,  6.58it/s, loss=3.68, v_num=0]
Epoch 21:  92%|█████████▏| 89/97 [00:13<00:01,  6.61it/s, loss=3.68, v_num=0]
Epoch 21:  92%|█████████▏| 89/97 [00:13<00:01,  6.58it/s, loss=3.69, v_num=0]
Epoch 21:  93%|█████████▎| 90/97 [00:13<00:01,  6.62it/s, loss=3.69, v_num=0]
Epoch 21:  93%|█████████▎| 90/97 [00:13<00:01,  6.59it/s, loss=3.69, v_num=0]
Epoch 21:  94%|█████████▍| 91/97 [00:13<00:00,  6.62it/s, loss=3.69, v_num=0]
Epoch 21:  94%|█████████▍| 91/97 [00:13<00:00,  6.59it/s, loss=3.7, v_num=0]
Epoch 21:  95%|█████████▍| 92/97 [00:13<00:00,  6.63it/s, loss=3.7, v_num=0]
Epoch 21:  95%|█████████▍| 92/97 [00:13<00:00,  6.60it/s, loss=3.69, v_num=0]
Epoch 21:  96%|█████████▌| 93/97 [00:14<00:00,  6.63it/s, loss=3.69, v_num=0]
Epoch 21:  96%|█████████▌| 93/97 [00:14<00:00,  6.61it/s, loss=3.69, v_num=0]
Epoch 21:  97%|█████████▋| 94/97 [00:14<00:00,  6.64it/s, loss=3.69, v_num=0]
Epoch 21:  97%|█████████▋| 94/97 [00:14<00:00,  6.61it/s, loss=3.68, v_num=0]
Epoch 21:  98%|█████████▊| 95/97 [00:14<00:00,  6.65it/s, loss=3.68, v_num=0]
Epoch 21:  98%|█████████▊| 95/97 [00:14<00:00,  6.62it/s, loss=3.68, v_num=0]
Epoch 21:  99%|█████████▉| 96/97 [00:14<00:00,  6.65it/s, loss=3.68, v_num=0]
Epoch 21:  99%|█████████▉| 96/97 [00:14<00:00,  6.63it/s, loss=3.68, v_num=0]
Epoch 21: 100%|██████████| 97/97 [00:14<00:00,  6.66it/s, loss=3.68, v_num=0]
Epoch 21: 100%|██████████| 97/97 [00:14<00:00,  6.63it/s, loss=3.69, v_num=0]
Epoch 21: 100%|██████████| 97/97 [00:14<00:00,  6.63it/s, loss=3.69, v_num=0]
Epoch 21:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.69, v_num=0]
Epoch 22:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.69, v_num=0]
Epoch 22:   1%|          | 1/97 [00:01<02:13,  1.39s/it, loss=3.69, v_num=0]
Epoch 22:   1%|          | 1/97 [00:01<02:18,  1.45s/it, loss=3.69, v_num=0]
Epoch 22:   2%|▏         | 2/97 [00:01<01:12,  1.31it/s, loss=3.69, v_num=0]
Epoch 22:   2%|▏         | 2/97 [00:01<01:15,  1.26it/s, loss=3.7, v_num=0]
Epoch 22:   3%|▎         | 3/97 [00:01<00:52,  1.80it/s, loss=3.7, v_num=0]
Epoch 22:   3%|▎         | 3/97 [00:01<00:53,  1.74it/s, loss=3.7, v_num=0]
Epoch 22:   4%|▍         | 4/97 [00:01<00:41,  2.23it/s, loss=3.7, v_num=0]
Epoch 22:   4%|▍         | 4/97 [00:01<00:43,  2.16it/s, loss=3.7, v_num=0]
Epoch 22:   5%|▌         | 5/97 [00:01<00:35,  2.57it/s, loss=3.7, v_num=0]
Epoch 22:   5%|▌         | 5/97 [00:01<00:36,  2.50it/s, loss=3.69, v_num=0]
Epoch 22:   6%|▌         | 6/97 [00:02<00:31,  2.89it/s, loss=3.69, v_num=0]
Epoch 22:   6%|▌         | 6/97 [00:02<00:32,  2.81it/s, loss=3.7, v_num=0]
Epoch 22:   7%|▋         | 7/97 [00:02<00:28,  3.16it/s, loss=3.7, v_num=0]
Epoch 22:   7%|▋         | 7/97 [00:02<00:29,  3.09it/s, loss=3.69, v_num=0]
Epoch 22:   8%|▊         | 8/97 [00:02<00:26,  3.41it/s, loss=3.69, v_num=0]
Epoch 22:   8%|▊         | 8/97 [00:02<00:26,  3.33it/s, loss=3.69, v_num=0]
Epoch 22:   9%|▉         | 9/97 [00:02<00:24,  3.61it/s, loss=3.69, v_num=0]
Epoch 22:   9%|▉         | 9/97 [00:02<00:24,  3.53it/s, loss=3.69, v_num=0]
Epoch 22:  10%|█         | 10/97 [00:02<00:22,  3.81it/s, loss=3.69, v_num=0]
Epoch 22:  10%|█         | 10/97 [00:02<00:23,  3.73it/s, loss=3.69, v_num=0]
Epoch 22:  11%|█▏        | 11/97 [00:02<00:21,  3.98it/s, loss=3.69, v_num=0]
Epoch 22:  11%|█▏        | 11/97 [00:02<00:22,  3.90it/s, loss=3.69, v_num=0]
Epoch 22:  12%|█▏        | 12/97 [00:02<00:20,  4.13it/s, loss=3.69, v_num=0]
Epoch 22:  12%|█▏        | 12/97 [00:02<00:20,  4.05it/s, loss=3.68, v_num=0]
Epoch 22:  13%|█▎        | 13/97 [00:03<00:19,  4.27it/s, loss=3.68, v_num=0]
Epoch 22:  13%|█▎        | 13/97 [00:03<00:20,  4.19it/s, loss=3.68, v_num=0]
Epoch 22:  14%|█▍        | 14/97 [00:03<00:18,  4.40it/s, loss=3.68, v_num=0]
Epoch 22:  14%|█▍        | 14/97 [00:03<00:19,  4.33it/s, loss=3.67, v_num=0]
Epoch 22:  15%|█▌        | 15/97 [00:03<00:18,  4.52it/s, loss=3.67, v_num=0]
Epoch 22:  15%|█▌        | 15/97 [00:03<00:18,  4.45it/s, loss=3.67, v_num=0]
Epoch 22:  16%|█▋        | 16/97 [00:03<00:17,  4.64it/s, loss=3.67, v_num=0]
Epoch 22:  16%|█▋        | 16/97 [00:03<00:17,  4.56it/s, loss=3.66, v_num=0]
Epoch 22:  18%|█▊        | 17/97 [00:03<00:16,  4.72it/s, loss=3.66, v_num=0]
Epoch 22:  18%|█▊        | 17/97 [00:03<00:17,  4.65it/s, loss=3.66, v_num=0]
Epoch 22:  19%|█▊        | 18/97 [00:03<00:16,  4.82it/s, loss=3.66, v_num=0]
Epoch 22:  19%|█▊        | 18/97 [00:03<00:16,  4.75it/s, loss=3.66, v_num=0]
Epoch 22:  20%|█▉        | 19/97 [00:03<00:15,  4.91it/s, loss=3.66, v_num=0]
Epoch 22:  20%|█▉        | 19/97 [00:03<00:16,  4.84it/s, loss=3.65, v_num=0]
Epoch 22:  21%|██        | 20/97 [00:04<00:15,  4.99it/s, loss=3.65, v_num=0]
Epoch 22:  21%|██        | 20/97 [00:04<00:15,  4.92it/s, loss=3.65, v_num=0]
Epoch 22:  22%|██▏       | 21/97 [00:04<00:14,  5.07it/s, loss=3.65, v_num=0]
Epoch 22:  22%|██▏       | 21/97 [00:04<00:15,  5.00it/s, loss=3.64, v_num=0]
Epoch 22:  23%|██▎       | 22/97 [00:04<00:14,  5.14it/s, loss=3.64, v_num=0]
Epoch 22:  23%|██▎       | 22/97 [00:04<00:14,  5.07it/s, loss=3.64, v_num=0]
Epoch 22:  24%|██▎       | 23/97 [00:04<00:14,  5.21it/s, loss=3.64, v_num=0]
Epoch 22:  24%|██▎       | 23/97 [00:04<00:14,  5.14it/s, loss=3.64, v_num=0]
Epoch 22:  25%|██▍       | 24/97 [00:04<00:13,  5.27it/s, loss=3.64, v_num=0]
Epoch 22:  25%|██▍       | 24/97 [00:04<00:14,  5.21it/s, loss=3.64, v_num=0]
Epoch 22:  26%|██▌       | 25/97 [00:04<00:13,  5.33it/s, loss=3.64, v_num=0]
Epoch 22:  26%|██▌       | 25/97 [00:04<00:13,  5.26it/s, loss=3.65, v_num=0]
Epoch 22:  27%|██▋       | 26/97 [00:04<00:13,  5.38it/s, loss=3.65, v_num=0]
Epoch 22:  27%|██▋       | 26/97 [00:04<00:13,  5.32it/s, loss=3.65, v_num=0]
Epoch 22:  28%|██▊       | 27/97 [00:04<00:12,  5.44it/s, loss=3.65, v_num=0]
Epoch 22:  28%|██▊       | 27/97 [00:05<00:13,  5.38it/s, loss=3.67, v_num=0]
Epoch 22:  29%|██▉       | 28/97 [00:05<00:12,  5.49it/s, loss=3.67, v_num=0]
Epoch 22:  29%|██▉       | 28/97 [00:05<00:12,  5.42it/s, loss=3.68, v_num=0]
Epoch 22:  30%|██▉       | 29/97 [00:05<00:12,  5.53it/s, loss=3.68, v_num=0]
Epoch 22:  30%|██▉       | 29/97 [00:05<00:12,  5.47it/s, loss=3.68, v_num=0]
Epoch 22:  31%|███       | 30/97 [00:05<00:12,  5.58it/s, loss=3.68, v_num=0]
Epoch 22:  31%|███       | 30/97 [00:05<00:12,  5.52it/s, loss=3.68, v_num=0]
Epoch 22:  32%|███▏      | 31/97 [00:05<00:11,  5.61it/s, loss=3.68, v_num=0]
Epoch 22:  32%|███▏      | 31/97 [00:05<00:11,  5.56it/s, loss=3.68, v_num=0]
Epoch 22:  33%|███▎      | 32/97 [00:05<00:11,  5.66it/s, loss=3.68, v_num=0]
Epoch 22:  33%|███▎      | 32/97 [00:05<00:11,  5.60it/s, loss=3.68, v_num=0]
Epoch 22:  34%|███▍      | 33/97 [00:05<00:11,  5.69it/s, loss=3.68, v_num=0]
Epoch 22:  34%|███▍      | 33/97 [00:05<00:11,  5.64it/s, loss=3.7, v_num=0]
Epoch 22:  35%|███▌      | 34/97 [00:05<00:10,  5.73it/s, loss=3.7, v_num=0]
Epoch 22:  35%|███▌      | 34/97 [00:05<00:11,  5.68it/s, loss=3.7, v_num=0]
Epoch 22:  36%|███▌      | 35/97 [00:06<00:10,  5.77it/s, loss=3.7, v_num=0]
Epoch 22:  36%|███▌      | 35/97 [00:06<00:10,  5.71it/s, loss=3.7, v_num=0]
Epoch 22:  37%|███▋      | 36/97 [00:06<00:10,  5.80it/s, loss=3.7, v_num=0]
Epoch 22:  37%|███▋      | 36/97 [00:06<00:10,  5.74it/s, loss=3.71, v_num=0]
Epoch 22:  38%|███▊      | 37/97 [00:06<00:10,  5.83it/s, loss=3.71, v_num=0]
Epoch 22:  38%|███▊      | 37/97 [00:06<00:10,  5.78it/s, loss=3.71, v_num=0]
Epoch 22:  39%|███▉      | 38/97 [00:06<00:10,  5.86it/s, loss=3.71, v_num=0]
Epoch 22:  39%|███▉      | 38/97 [00:06<00:10,  5.81it/s, loss=3.71, v_num=0]
Epoch 22:  40%|████      | 39/97 [00:06<00:09,  5.89it/s, loss=3.71, v_num=0]
Epoch 22:  40%|████      | 39/97 [00:06<00:09,  5.84it/s, loss=3.72, v_num=0]
Epoch 22:  41%|████      | 40/97 [00:06<00:09,  5.91it/s, loss=3.72, v_num=0]
Epoch 22:  41%|████      | 40/97 [00:06<00:09,  5.87it/s, loss=3.72, v_num=0]
Epoch 22:  42%|████▏     | 41/97 [00:06<00:09,  5.94it/s, loss=3.72, v_num=0]
Epoch 22:  42%|████▏     | 41/97 [00:06<00:09,  5.90it/s, loss=3.72, v_num=0]
Epoch 22:  43%|████▎     | 42/97 [00:07<00:09,  5.97it/s, loss=3.72, v_num=0]
Epoch 22:  43%|████▎     | 42/97 [00:07<00:09,  5.92it/s, loss=3.71, v_num=0]
Epoch 22:  44%|████▍     | 43/97 [00:07<00:09,  5.99it/s, loss=3.71, v_num=0]
Epoch 22:  44%|████▍     | 43/97 [00:07<00:09,  5.95it/s, loss=3.71, v_num=0]
Epoch 22:  45%|████▌     | 44/97 [00:07<00:08,  6.02it/s, loss=3.71, v_num=0]
Epoch 22:  45%|████▌     | 44/97 [00:07<00:08,  5.97it/s, loss=3.71, v_num=0]
Epoch 22:  46%|████▋     | 45/97 [00:07<00:08,  6.04it/s, loss=3.71, v_num=0]
Epoch 22:  46%|████▋     | 45/97 [00:07<00:08,  6.00it/s, loss=3.71, v_num=0]
Epoch 22:  47%|████▋     | 46/97 [00:07<00:08,  6.06it/s, loss=3.71, v_num=0]
Epoch 22:  47%|████▋     | 46/97 [00:07<00:08,  6.02it/s, loss=3.71, v_num=0]
Epoch 22:  48%|████▊     | 47/97 [00:07<00:08,  6.08it/s, loss=3.71, v_num=0]
Epoch 22:  48%|████▊     | 47/97 [00:07<00:08,  6.04it/s, loss=3.7, v_num=0]
Epoch 22:  49%|████▉     | 48/97 [00:07<00:08,  6.10it/s, loss=3.7, v_num=0]
Epoch 22:  49%|████▉     | 48/97 [00:07<00:08,  6.06it/s, loss=3.7, v_num=0]
Epoch 22:  51%|█████     | 49/97 [00:08<00:07,  6.12it/s, loss=3.7, v_num=0]
Epoch 22:  51%|█████     | 49/97 [00:08<00:07,  6.08it/s, loss=3.69, v_num=0]
Epoch 22:  52%|█████▏    | 50/97 [00:08<00:07,  6.14it/s, loss=3.69, v_num=0]
Epoch 22:  52%|█████▏    | 50/97 [00:08<00:07,  6.10it/s, loss=3.69, v_num=0]
Epoch 22:  53%|█████▎    | 51/97 [00:08<00:07,  6.16it/s, loss=3.69, v_num=0]
Epoch 22:  53%|█████▎    | 51/97 [00:08<00:07,  6.12it/s, loss=3.68, v_num=0]
Epoch 22:  54%|█████▎    | 52/97 [00:08<00:07,  6.18it/s, loss=3.68, v_num=0]
Epoch 22:  54%|█████▎    | 52/97 [00:08<00:07,  6.14it/s, loss=3.67, v_num=0]
Epoch 22:  55%|█████▍    | 53/97 [00:08<00:07,  6.19it/s, loss=3.67, v_num=0]
Epoch 22:  55%|█████▍    | 53/97 [00:08<00:07,  6.16it/s, loss=3.67, v_num=0]
Epoch 22:  56%|█████▌    | 54/97 [00:08<00:06,  6.21it/s, loss=3.67, v_num=0]
Epoch 22:  56%|█████▌    | 54/97 [00:08<00:06,  6.17it/s, loss=3.66, v_num=0]
Epoch 22:  57%|█████▋    | 55/97 [00:08<00:06,  6.23it/s, loss=3.66, v_num=0]
Epoch 22:  57%|█████▋    | 55/97 [00:08<00:06,  6.19it/s, loss=3.65, v_num=0]
Epoch 22:  58%|█████▊    | 56/97 [00:08<00:06,  6.24it/s, loss=3.65, v_num=0]
Epoch 22:  58%|█████▊    | 56/97 [00:09<00:06,  6.20it/s, loss=3.64, v_num=0]
Epoch 22:  59%|█████▉    | 57/97 [00:09<00:06,  6.26it/s, loss=3.64, v_num=0]
Epoch 22:  59%|█████▉    | 57/97 [00:09<00:06,  6.22it/s, loss=3.64, v_num=0]
Epoch 22:  60%|█████▉    | 58/97 [00:09<00:06,  6.27it/s, loss=3.64, v_num=0]
Epoch 22:  60%|█████▉    | 58/97 [00:09<00:06,  6.24it/s, loss=3.64, v_num=0]
Epoch 22:  61%|██████    | 59/97 [00:09<00:06,  6.29it/s, loss=3.64, v_num=0]
Epoch 22:  61%|██████    | 59/97 [00:09<00:06,  6.25it/s, loss=3.64, v_num=0]
Epoch 22:  62%|██████▏   | 60/97 [00:09<00:05,  6.30it/s, loss=3.64, v_num=0]
Epoch 22:  62%|██████▏   | 60/97 [00:09<00:05,  6.27it/s, loss=3.64, v_num=0]
Epoch 22:  63%|██████▎   | 61/97 [00:09<00:05,  6.32it/s, loss=3.64, v_num=0]
Epoch 22:  63%|██████▎   | 61/97 [00:09<00:05,  6.28it/s, loss=3.65, v_num=0]
Epoch 22:  64%|██████▍   | 62/97 [00:09<00:05,  6.33it/s, loss=3.65, v_num=0]
Epoch 22:  64%|██████▍   | 62/97 [00:09<00:05,  6.30it/s, loss=3.64, v_num=0]
Epoch 22:  65%|██████▍   | 63/97 [00:09<00:05,  6.34it/s, loss=3.64, v_num=0]
Epoch 22:  65%|██████▍   | 63/97 [00:09<00:05,  6.31it/s, loss=3.64, v_num=0]
Epoch 22:  66%|██████▌   | 64/97 [00:10<00:05,  6.36it/s, loss=3.64, v_num=0]
Epoch 22:  66%|██████▌   | 64/97 [00:10<00:05,  6.32it/s, loss=3.64, v_num=0]
Epoch 22:  67%|██████▋   | 65/97 [00:10<00:05,  6.37it/s, loss=3.64, v_num=0]
Epoch 22:  67%|██████▋   | 65/97 [00:10<00:05,  6.33it/s, loss=3.63, v_num=0]
Epoch 22:  68%|██████▊   | 66/97 [00:10<00:04,  6.38it/s, loss=3.63, v_num=0]
Epoch 22:  68%|██████▊   | 66/97 [00:10<00:04,  6.35it/s, loss=3.63, v_num=0]
Epoch 22:  69%|██████▉   | 67/97 [00:10<00:04,  6.39it/s, loss=3.63, v_num=0]
Epoch 22:  69%|██████▉   | 67/97 [00:10<00:04,  6.36it/s, loss=3.63, v_num=0]
Epoch 22:  70%|███████   | 68/97 [00:10<00:04,  6.40it/s, loss=3.63, v_num=0]
Epoch 22:  70%|███████   | 68/97 [00:10<00:04,  6.37it/s, loss=3.63, v_num=0]
Epoch 22:  71%|███████   | 69/97 [00:10<00:04,  6.42it/s, loss=3.63, v_num=0]
Epoch 22:  71%|███████   | 69/97 [00:10<00:04,  6.38it/s, loss=3.62, v_num=0]
Epoch 22:  72%|███████▏  | 70/97 [00:10<00:04,  6.43it/s, loss=3.62, v_num=0]
Epoch 22:  72%|███████▏  | 70/97 [00:10<00:04,  6.39it/s, loss=3.63, v_num=0]
Epoch 22:  73%|███████▎  | 71/97 [00:11<00:04,  6.44it/s, loss=3.63, v_num=0]
Epoch 22:  73%|███████▎  | 71/97 [00:11<00:04,  6.40it/s, loss=3.64, v_num=0]
Epoch 22:  74%|███████▍  | 72/97 [00:11<00:03,  6.45it/s, loss=3.64, v_num=0]
Epoch 22:  74%|███████▍  | 72/97 [00:11<00:03,  6.41it/s, loss=3.63, v_num=0]
Epoch 22:  75%|███████▌  | 73/97 [00:11<00:03,  6.46it/s, loss=3.63, v_num=0]
Epoch 22:  75%|███████▌  | 73/97 [00:11<00:03,  6.42it/s, loss=3.63, v_num=0]
Epoch 22:  76%|███████▋  | 74/97 [00:11<00:03,  6.47it/s, loss=3.63, v_num=0]
Epoch 22:  76%|███████▋  | 74/97 [00:11<00:03,  6.44it/s, loss=3.64, v_num=0]
Epoch 22:  77%|███████▋  | 75/97 [00:11<00:03,  6.48it/s, loss=3.64, v_num=0]
Epoch 22:  77%|███████▋  | 75/97 [00:11<00:03,  6.45it/s, loss=3.64, v_num=0]
Epoch 22:  78%|███████▊  | 76/97 [00:11<00:03,  6.49it/s, loss=3.64, v_num=0]
Epoch 22:  78%|███████▊  | 76/97 [00:11<00:03,  6.45it/s, loss=3.64, v_num=0]
Epoch 22:  79%|███████▉  | 77/97 [00:11<00:03,  6.49it/s, loss=3.64, v_num=0]
Epoch 22:  79%|███████▉  | 77/97 [00:11<00:03,  6.46it/s, loss=3.64, v_num=0]
Epoch 22:  80%|████████  | 78/97 [00:11<00:02,  6.51it/s, loss=3.64, v_num=0]
Epoch 22:  80%|████████  | 78/97 [00:12<00:02,  6.47it/s, loss=3.64, v_num=0]
Epoch 22:  81%|████████▏ | 79/97 [00:12<00:02,  6.51it/s, loss=3.64, v_num=0]
Epoch 22:  81%|████████▏ | 79/97 [00:12<00:02,  6.48it/s, loss=3.63, v_num=0]
Epoch 22:  82%|████████▏ | 80/97 [00:12<00:02,  6.52it/s, loss=3.63, v_num=0]
Epoch 22:  82%|████████▏ | 80/97 [00:12<00:02,  6.49it/s, loss=3.63, v_num=0]
Epoch 22:  84%|████████▎ | 81/97 [00:12<00:02,  6.53it/s, loss=3.63, v_num=0]
Epoch 22:  84%|████████▎ | 81/97 [00:12<00:02,  6.50it/s, loss=3.64, v_num=0]
Epoch 22:  85%|████████▍ | 82/97 [00:12<00:02,  6.54it/s, loss=3.64, v_num=0]
Epoch 22:  85%|████████▍ | 82/97 [00:12<00:02,  6.51it/s, loss=3.63, v_num=0]
Epoch 22:  86%|████████▌ | 83/97 [00:12<00:02,  6.55it/s, loss=3.63, v_num=0]
Epoch 22:  86%|████████▌ | 83/97 [00:12<00:02,  6.52it/s, loss=3.63, v_num=0]
Epoch 22:  87%|████████▋ | 84/97 [00:12<00:01,  6.55it/s, loss=3.63, v_num=0]
Epoch 22:  87%|████████▋ | 84/97 [00:12<00:01,  6.53it/s, loss=3.63, v_num=0]
Epoch 22:  88%|████████▊ | 85/97 [00:12<00:01,  6.56it/s, loss=3.63, v_num=0]
Epoch 22:  88%|████████▊ | 85/97 [00:13<00:01,  6.53it/s, loss=3.62, v_num=0]
Epoch 22:  89%|████████▊ | 86/97 [00:13<00:01,  6.57it/s, loss=3.62, v_num=0]
Epoch 22:  89%|████████▊ | 86/97 [00:13<00:01,  6.54it/s, loss=3.62, v_num=0]
Epoch 22:  90%|████████▉ | 87/97 [00:13<00:01,  6.58it/s, loss=3.62, v_num=0]
Epoch 22:  90%|████████▉ | 87/97 [00:13<00:01,  6.55it/s, loss=3.62, v_num=0]
Epoch 22:  91%|█████████ | 88/97 [00:13<00:01,  6.58it/s, loss=3.62, v_num=0]
Epoch 22:  91%|█████████ | 88/97 [00:13<00:01,  6.56it/s, loss=3.62, v_num=0]
Epoch 22:  92%|█████████▏| 89/97 [00:13<00:01,  6.59it/s, loss=3.62, v_num=0]
Epoch 22:  92%|█████████▏| 89/97 [00:13<00:01,  6.56it/s, loss=3.63, v_num=0]
Epoch 22:  93%|█████████▎| 90/97 [00:13<00:01,  6.60it/s, loss=3.63, v_num=0]
Epoch 22:  93%|█████████▎| 90/97 [00:13<00:01,  6.57it/s, loss=3.62, v_num=0]
Epoch 22:  94%|█████████▍| 91/97 [00:13<00:00,  6.60it/s, loss=3.62, v_num=0]
Epoch 22:  94%|█████████▍| 91/97 [00:13<00:00,  6.58it/s, loss=3.63, v_num=0]
Epoch 22:  95%|█████████▍| 92/97 [00:13<00:00,  6.61it/s, loss=3.63, v_num=0]
Epoch 22:  95%|█████████▍| 92/97 [00:13<00:00,  6.58it/s, loss=3.63, v_num=0]
Epoch 22:  96%|█████████▌| 93/97 [00:14<00:00,  6.62it/s, loss=3.63, v_num=0]
Epoch 22:  96%|█████████▌| 93/97 [00:14<00:00,  6.59it/s, loss=3.63, v_num=0]
Epoch 22:  97%|█████████▋| 94/97 [00:14<00:00,  6.62it/s, loss=3.63, v_num=0]
Epoch 22:  97%|█████████▋| 94/97 [00:14<00:00,  6.60it/s, loss=3.64, v_num=0]
Epoch 22:  98%|█████████▊| 95/97 [00:14<00:00,  6.63it/s, loss=3.64, v_num=0]
Epoch 22:  98%|█████████▊| 95/97 [00:14<00:00,  6.60it/s, loss=3.64, v_num=0]
Epoch 22:  99%|█████████▉| 96/97 [00:14<00:00,  6.64it/s, loss=3.64, v_num=0]
Epoch 22:  99%|█████████▉| 96/97 [00:14<00:00,  6.61it/s, loss=3.64, v_num=0]
Epoch 22: 100%|██████████| 97/97 [00:14<00:00,  6.64it/s, loss=3.64, v_num=0]
Epoch 22: 100%|██████████| 97/97 [00:14<00:00,  6.62it/s, loss=3.63, v_num=0]
Epoch 22: 100%|██████████| 97/97 [00:14<00:00,  6.62it/s, loss=3.63, v_num=0]
Epoch 22:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.63, v_num=0]
Epoch 23:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.63, v_num=0]
Epoch 23:   1%|          | 1/97 [00:01<02:12,  1.38s/it, loss=3.63, v_num=0]
Epoch 23:   1%|          | 1/97 [00:01<02:17,  1.43s/it, loss=3.62, v_num=0]
Epoch 23:   2%|▏         | 2/97 [00:01<01:12,  1.32it/s, loss=3.62, v_num=0]
Epoch 23:   2%|▏         | 2/97 [00:01<01:14,  1.27it/s, loss=3.62, v_num=0]
Epoch 23:   3%|▎         | 3/97 [00:01<00:52,  1.80it/s, loss=3.62, v_num=0]
Epoch 23:   3%|▎         | 3/97 [00:01<00:53,  1.74it/s, loss=3.62, v_num=0]
Epoch 23:   4%|▍         | 4/97 [00:01<00:41,  2.22it/s, loss=3.62, v_num=0]
Epoch 23:   4%|▍         | 4/97 [00:01<00:43,  2.16it/s, loss=3.61, v_num=0]
Epoch 23:   5%|▌         | 5/97 [00:01<00:35,  2.59it/s, loss=3.61, v_num=0]
Epoch 23:   5%|▌         | 5/97 [00:01<00:36,  2.51it/s, loss=3.61, v_num=0]
Epoch 23:   6%|▌         | 6/97 [00:02<00:31,  2.90it/s, loss=3.61, v_num=0]
Epoch 23:   6%|▌         | 6/97 [00:02<00:32,  2.82it/s, loss=3.62, v_num=0]
Epoch 23:   7%|▋         | 7/97 [00:02<00:28,  3.17it/s, loss=3.62, v_num=0]
Epoch 23:   7%|▋         | 7/97 [00:02<00:29,  3.09it/s, loss=3.63, v_num=0]
Epoch 23:   8%|▊         | 8/97 [00:02<00:26,  3.41it/s, loss=3.63, v_num=0]
Epoch 23:   8%|▊         | 8/97 [00:02<00:26,  3.34it/s, loss=3.63, v_num=0]
Epoch 23:   9%|▉         | 9/97 [00:02<00:24,  3.63it/s, loss=3.63, v_num=0]
Epoch 23:   9%|▉         | 9/97 [00:02<00:24,  3.55it/s, loss=3.63, v_num=0]
Epoch 23:  10%|█         | 10/97 [00:02<00:22,  3.82it/s, loss=3.63, v_num=0]
Epoch 23:  10%|█         | 10/97 [00:02<00:23,  3.74it/s, loss=3.63, v_num=0]
Epoch 23:  11%|█▏        | 11/97 [00:02<00:21,  3.99it/s, loss=3.63, v_num=0]
Epoch 23:  11%|█▏        | 11/97 [00:02<00:21,  3.91it/s, loss=3.63, v_num=0]
Epoch 23:  12%|█▏        | 12/97 [00:02<00:20,  4.15it/s, loss=3.63, v_num=0]
Epoch 23:  12%|█▏        | 12/97 [00:02<00:20,  4.07it/s, loss=3.63, v_num=0]
Epoch 23:  13%|█▎        | 13/97 [00:03<00:19,  4.29it/s, loss=3.63, v_num=0]
Epoch 23:  13%|█▎        | 13/97 [00:03<00:19,  4.21it/s, loss=3.63, v_num=0]
Epoch 23:  14%|█▍        | 14/97 [00:03<00:18,  4.42it/s, loss=3.63, v_num=0]
Epoch 23:  14%|█▍        | 14/97 [00:03<00:19,  4.35it/s, loss=3.63, v_num=0]
Epoch 23:  15%|█▌        | 15/97 [00:03<00:18,  4.54it/s, loss=3.63, v_num=0]
Epoch 23:  15%|█▌        | 15/97 [00:03<00:18,  4.47it/s, loss=3.62, v_num=0]
Epoch 23:  16%|█▋        | 16/97 [00:03<00:17,  4.65it/s, loss=3.62, v_num=0]
Epoch 23:  16%|█▋        | 16/97 [00:03<00:17,  4.58it/s, loss=3.61, v_num=0]
Epoch 23:  18%|█▊        | 17/97 [00:03<00:16,  4.76it/s, loss=3.61, v_num=0]
Epoch 23:  18%|█▊        | 17/97 [00:03<00:17,  4.68it/s, loss=3.61, v_num=0]
Epoch 23:  19%|█▊        | 18/97 [00:03<00:16,  4.85it/s, loss=3.61, v_num=0]
Epoch 23:  19%|█▊        | 18/97 [00:03<00:16,  4.78it/s, loss=3.61, v_num=0]
Epoch 23:  20%|█▉        | 19/97 [00:03<00:15,  4.93it/s, loss=3.61, v_num=0]
Epoch 23:  20%|█▉        | 19/97 [00:03<00:16,  4.86it/s, loss=3.62, v_num=0]
Epoch 23:  21%|██        | 20/97 [00:04<00:15,  4.99it/s, loss=3.62, v_num=0]
Epoch 23:  21%|██        | 20/97 [00:04<00:15,  4.93it/s, loss=3.62, v_num=0]
Epoch 23:  22%|██▏       | 21/97 [00:04<00:14,  5.07it/s, loss=3.62, v_num=0]
Epoch 23:  22%|██▏       | 21/97 [00:04<00:15,  5.00it/s, loss=3.62, v_num=0]
Epoch 23:  23%|██▎       | 22/97 [00:04<00:14,  5.14it/s, loss=3.62, v_num=0]
Epoch 23:  23%|██▎       | 22/97 [00:04<00:14,  5.07it/s, loss=3.63, v_num=0]
Epoch 23:  24%|██▎       | 23/97 [00:04<00:14,  5.21it/s, loss=3.63, v_num=0]
Epoch 23:  24%|██▎       | 23/97 [00:04<00:14,  5.14it/s, loss=3.63, v_num=0]
Epoch 23:  25%|██▍       | 24/97 [00:04<00:13,  5.27it/s, loss=3.63, v_num=0]
Epoch 23:  25%|██▍       | 24/97 [00:04<00:14,  5.21it/s, loss=3.63, v_num=0]
Epoch 23:  26%|██▌       | 25/97 [00:04<00:13,  5.34it/s, loss=3.63, v_num=0]
Epoch 23:  26%|██▌       | 25/97 [00:04<00:13,  5.27it/s, loss=3.62, v_num=0]
Epoch 23:  27%|██▋       | 26/97 [00:04<00:13,  5.39it/s, loss=3.62, v_num=0]
Epoch 23:  27%|██▋       | 26/97 [00:04<00:13,  5.33it/s, loss=3.62, v_num=0]
Epoch 23:  28%|██▊       | 27/97 [00:04<00:12,  5.44it/s, loss=3.62, v_num=0]
Epoch 23:  28%|██▊       | 27/97 [00:05<00:13,  5.38it/s, loss=3.61, v_num=0]
Epoch 23:  29%|██▉       | 28/97 [00:05<00:12,  5.49it/s, loss=3.61, v_num=0]
Epoch 23:  29%|██▉       | 28/97 [00:05<00:12,  5.43it/s, loss=3.61, v_num=0]
Epoch 23:  30%|██▉       | 29/97 [00:05<00:12,  5.54it/s, loss=3.61, v_num=0]
Epoch 23:  30%|██▉       | 29/97 [00:05<00:12,  5.48it/s, loss=3.6, v_num=0]
Epoch 23:  31%|███       | 30/97 [00:05<00:12,  5.58it/s, loss=3.6, v_num=0]
Epoch 23:  31%|███       | 30/97 [00:05<00:12,  5.52it/s, loss=3.6, v_num=0]
Epoch 23:  32%|███▏      | 31/97 [00:05<00:11,  5.62it/s, loss=3.6, v_num=0]
Epoch 23:  32%|███▏      | 31/97 [00:05<00:11,  5.56it/s, loss=3.61, v_num=0]
Epoch 23:  33%|███▎      | 32/97 [00:05<00:11,  5.66it/s, loss=3.61, v_num=0]
Epoch 23:  33%|███▎      | 32/97 [00:05<00:11,  5.61it/s, loss=3.61, v_num=0]
Epoch 23:  34%|███▍      | 33/97 [00:05<00:11,  5.70it/s, loss=3.61, v_num=0]
Epoch 23:  34%|███▍      | 33/97 [00:05<00:11,  5.65it/s, loss=3.61, v_num=0]
Epoch 23:  35%|███▌      | 34/97 [00:05<00:10,  5.74it/s, loss=3.61, v_num=0]
Epoch 23:  35%|███▌      | 34/97 [00:05<00:11,  5.68it/s, loss=3.61, v_num=0]
Epoch 23:  36%|███▌      | 35/97 [00:06<00:10,  5.77it/s, loss=3.61, v_num=0]
Epoch 23:  36%|███▌      | 35/97 [00:06<00:10,  5.72it/s, loss=3.61, v_num=0]
Epoch 23:  37%|███▋      | 36/97 [00:06<00:10,  5.81it/s, loss=3.61, v_num=0]
Epoch 23:  37%|███▋      | 36/97 [00:06<00:10,  5.75it/s, loss=3.61, v_num=0]
Epoch 23:  38%|███▊      | 37/97 [00:06<00:10,  5.84it/s, loss=3.61, v_num=0]
Epoch 23:  38%|███▊      | 37/97 [00:06<00:10,  5.79it/s, loss=3.61, v_num=0]
Epoch 23:  39%|███▉      | 38/97 [00:06<00:10,  5.87it/s, loss=3.61, v_num=0]
Epoch 23:  39%|███▉      | 38/97 [00:06<00:10,  5.82it/s, loss=3.62, v_num=0]
Epoch 23:  40%|████      | 39/97 [00:06<00:09,  5.90it/s, loss=3.62, v_num=0]
Epoch 23:  40%|████      | 39/97 [00:06<00:09,  5.85it/s, loss=3.62, v_num=0]
Epoch 23:  41%|████      | 40/97 [00:06<00:09,  5.93it/s, loss=3.62, v_num=0]
Epoch 23:  41%|████      | 40/97 [00:06<00:09,  5.88it/s, loss=3.61, v_num=0]
Epoch 23:  42%|████▏     | 41/97 [00:06<00:09,  5.95it/s, loss=3.61, v_num=0]
Epoch 23:  42%|████▏     | 41/97 [00:06<00:09,  5.91it/s, loss=3.61, v_num=0]
Epoch 23:  43%|████▎     | 42/97 [00:07<00:09,  5.98it/s, loss=3.61, v_num=0]
Epoch 23:  43%|████▎     | 42/97 [00:07<00:09,  5.93it/s, loss=3.61, v_num=0]
Epoch 23:  44%|████▍     | 43/97 [00:07<00:09,  6.00it/s, loss=3.61, v_num=0]
Epoch 23:  44%|████▍     | 43/97 [00:07<00:09,  5.95it/s, loss=3.61, v_num=0]
Epoch 23:  45%|████▌     | 44/97 [00:07<00:08,  6.02it/s, loss=3.61, v_num=0]
Epoch 23:  45%|████▌     | 44/97 [00:07<00:08,  5.98it/s, loss=3.62, v_num=0]
Epoch 23:  46%|████▋     | 45/97 [00:07<00:08,  6.04it/s, loss=3.62, v_num=0]
Epoch 23:  46%|████▋     | 45/97 [00:07<00:08,  6.00it/s, loss=3.62, v_num=0]
Epoch 23:  47%|████▋     | 46/97 [00:07<00:08,  6.06it/s, loss=3.62, v_num=0]
Epoch 23:  47%|████▋     | 46/97 [00:07<00:08,  6.02it/s, loss=3.62, v_num=0]
Epoch 23:  48%|████▊     | 47/97 [00:07<00:08,  6.08it/s, loss=3.62, v_num=0]
Epoch 23:  48%|████▊     | 47/97 [00:07<00:08,  6.04it/s, loss=3.62, v_num=0]
Epoch 23:  49%|████▉     | 48/97 [00:07<00:08,  6.10it/s, loss=3.62, v_num=0]
Epoch 23:  49%|████▉     | 48/97 [00:07<00:08,  6.06it/s, loss=3.63, v_num=0]
Epoch 23:  51%|█████     | 49/97 [00:08<00:07,  6.12it/s, loss=3.63, v_num=0]
Epoch 23:  51%|█████     | 49/97 [00:08<00:07,  6.08it/s, loss=3.63, v_num=0]
Epoch 23:  52%|█████▏    | 50/97 [00:08<00:07,  6.14it/s, loss=3.63, v_num=0]
Epoch 23:  52%|█████▏    | 50/97 [00:08<00:07,  6.10it/s, loss=3.64, v_num=0]
Epoch 23:  53%|█████▎    | 51/97 [00:08<00:07,  6.16it/s, loss=3.64, v_num=0]
Epoch 23:  53%|█████▎    | 51/97 [00:08<00:07,  6.12it/s, loss=3.63, v_num=0]
Epoch 23:  54%|█████▎    | 52/97 [00:08<00:07,  6.18it/s, loss=3.63, v_num=0]
Epoch 23:  54%|█████▎    | 52/97 [00:08<00:07,  6.14it/s, loss=3.62, v_num=0]
Epoch 23:  55%|█████▍    | 53/97 [00:08<00:07,  6.20it/s, loss=3.62, v_num=0]
Epoch 23:  55%|█████▍    | 53/97 [00:08<00:07,  6.16it/s, loss=3.62, v_num=0]
Epoch 23:  56%|█████▌    | 54/97 [00:08<00:06,  6.22it/s, loss=3.62, v_num=0]
Epoch 23:  56%|█████▌    | 54/97 [00:08<00:06,  6.18it/s, loss=3.62, v_num=0]
Epoch 23:  57%|█████▋    | 55/97 [00:08<00:06,  6.23it/s, loss=3.62, v_num=0]
Epoch 23:  57%|█████▋    | 55/97 [00:08<00:06,  6.19it/s, loss=3.63, v_num=0]
Epoch 23:  58%|█████▊    | 56/97 [00:08<00:06,  6.25it/s, loss=3.63, v_num=0]
Epoch 23:  58%|█████▊    | 56/97 [00:09<00:06,  6.21it/s, loss=3.62, v_num=0]
Epoch 23:  59%|█████▉    | 57/97 [00:09<00:06,  6.26it/s, loss=3.62, v_num=0]
Epoch 23:  59%|█████▉    | 57/97 [00:09<00:06,  6.22it/s, loss=3.62, v_num=0]
Epoch 23:  60%|█████▉    | 58/97 [00:09<00:06,  6.28it/s, loss=3.62, v_num=0]
Epoch 23:  60%|█████▉    | 58/97 [00:09<00:06,  6.24it/s, loss=3.62, v_num=0]
Epoch 23:  61%|██████    | 59/97 [00:09<00:06,  6.30it/s, loss=3.62, v_num=0]
Epoch 23:  61%|██████    | 59/97 [00:09<00:06,  6.26it/s, loss=3.62, v_num=0]
Epoch 23:  62%|██████▏   | 60/97 [00:09<00:05,  6.31it/s, loss=3.62, v_num=0]
Epoch 23:  62%|██████▏   | 60/97 [00:09<00:05,  6.27it/s, loss=3.63, v_num=0]
Epoch 23:  63%|██████▎   | 61/97 [00:09<00:05,  6.32it/s, loss=3.63, v_num=0]
Epoch 23:  63%|██████▎   | 61/97 [00:09<00:05,  6.28it/s, loss=3.63, v_num=0]
Epoch 23:  64%|██████▍   | 62/97 [00:09<00:05,  6.34it/s, loss=3.63, v_num=0]
Epoch 23:  64%|██████▍   | 62/97 [00:09<00:05,  6.30it/s, loss=3.62, v_num=0]
Epoch 23:  65%|██████▍   | 63/97 [00:09<00:05,  6.35it/s, loss=3.62, v_num=0]
Epoch 23:  65%|██████▍   | 63/97 [00:09<00:05,  6.31it/s, loss=3.62, v_num=0]
Epoch 23:  66%|██████▌   | 64/97 [00:10<00:05,  6.36it/s, loss=3.62, v_num=0]
Epoch 23:  66%|██████▌   | 64/97 [00:10<00:05,  6.33it/s, loss=3.61, v_num=0]
Epoch 23:  67%|██████▋   | 65/97 [00:10<00:05,  6.37it/s, loss=3.61, v_num=0]
Epoch 23:  67%|██████▋   | 65/97 [00:10<00:05,  6.34it/s, loss=3.61, v_num=0]
Epoch 23:  68%|██████▊   | 66/97 [00:10<00:04,  6.39it/s, loss=3.61, v_num=0]
Epoch 23:  68%|██████▊   | 66/97 [00:10<00:04,  6.35it/s, loss=3.6, v_num=0]
Epoch 23:  69%|██████▉   | 67/97 [00:10<00:04,  6.40it/s, loss=3.6, v_num=0]
Epoch 23:  69%|██████▉   | 67/97 [00:10<00:04,  6.36it/s, loss=3.61, v_num=0]
Epoch 23:  70%|███████   | 68/97 [00:10<00:04,  6.41it/s, loss=3.61, v_num=0]
Epoch 23:  70%|███████   | 68/97 [00:10<00:04,  6.37it/s, loss=3.6, v_num=0]
Epoch 23:  71%|███████   | 69/97 [00:10<00:04,  6.42it/s, loss=3.6, v_num=0]
Epoch 23:  71%|███████   | 69/97 [00:10<00:04,  6.38it/s, loss=3.6, v_num=0]
Epoch 23:  72%|███████▏  | 70/97 [00:10<00:04,  6.43it/s, loss=3.6, v_num=0]
Epoch 23:  72%|███████▏  | 70/97 [00:10<00:04,  6.39it/s, loss=3.59, v_num=0]
Epoch 23:  73%|███████▎  | 71/97 [00:11<00:04,  6.44it/s, loss=3.59, v_num=0]
Epoch 23:  73%|███████▎  | 71/97 [00:11<00:04,  6.40it/s, loss=3.59, v_num=0]
Epoch 23:  74%|███████▍  | 72/97 [00:11<00:03,  6.45it/s, loss=3.59, v_num=0]
Epoch 23:  74%|███████▍  | 72/97 [00:11<00:03,  6.42it/s, loss=3.59, v_num=0]
Epoch 23:  75%|███████▌  | 73/97 [00:11<00:03,  6.46it/s, loss=3.59, v_num=0]
Epoch 23:  75%|███████▌  | 73/97 [00:11<00:03,  6.42it/s, loss=3.57, v_num=0]
Epoch 23:  76%|███████▋  | 74/97 [00:11<00:03,  6.47it/s, loss=3.57, v_num=0]
Epoch 23:  76%|███████▋  | 74/97 [00:11<00:03,  6.43it/s, loss=3.57, v_num=0]
Epoch 23:  77%|███████▋  | 75/97 [00:11<00:03,  6.48it/s, loss=3.57, v_num=0]
Epoch 23:  77%|███████▋  | 75/97 [00:11<00:03,  6.44it/s, loss=3.57, v_num=0]
Epoch 23:  78%|███████▊  | 76/97 [00:11<00:03,  6.48it/s, loss=3.57, v_num=0]
Epoch 23:  78%|███████▊  | 76/97 [00:11<00:03,  6.45it/s, loss=3.58, v_num=0]
Epoch 23:  79%|███████▉  | 77/97 [00:11<00:03,  6.50it/s, loss=3.58, v_num=0]
Epoch 23:  79%|███████▉  | 77/97 [00:11<00:03,  6.46it/s, loss=3.58, v_num=0]
Epoch 23:  80%|████████  | 78/97 [00:11<00:02,  6.50it/s, loss=3.58, v_num=0]
Epoch 23:  80%|████████  | 78/97 [00:12<00:02,  6.47it/s, loss=3.57, v_num=0]
Epoch 23:  81%|████████▏ | 79/97 [00:12<00:02,  6.51it/s, loss=3.57, v_num=0]
Epoch 23:  81%|████████▏ | 79/97 [00:12<00:02,  6.48it/s, loss=3.57, v_num=0]
Epoch 23:  82%|████████▏ | 80/97 [00:12<00:02,  6.52it/s, loss=3.57, v_num=0]
Epoch 23:  82%|████████▏ | 80/97 [00:12<00:02,  6.49it/s, loss=3.57, v_num=0]
Epoch 23:  84%|████████▎ | 81/97 [00:12<00:02,  6.53it/s, loss=3.57, v_num=0]
Epoch 23:  84%|████████▎ | 81/97 [00:12<00:02,  6.50it/s, loss=3.57, v_num=0]
Epoch 23:  85%|████████▍ | 82/97 [00:12<00:02,  6.54it/s, loss=3.57, v_num=0]
Epoch 23:  85%|████████▍ | 82/97 [00:12<00:02,  6.51it/s, loss=3.57, v_num=0]
Epoch 23:  86%|████████▌ | 83/97 [00:12<00:02,  6.54it/s, loss=3.57, v_num=0]
Epoch 23:  86%|████████▌ | 83/97 [00:12<00:02,  6.51it/s, loss=3.57, v_num=0]
Epoch 23:  87%|████████▋ | 84/97 [00:12<00:01,  6.55it/s, loss=3.57, v_num=0]
Epoch 23:  87%|████████▋ | 84/97 [00:12<00:01,  6.52it/s, loss=3.57, v_num=0]
Epoch 23:  88%|████████▊ | 85/97 [00:12<00:01,  6.56it/s, loss=3.57, v_num=0]
Epoch 23:  88%|████████▊ | 85/97 [00:13<00:01,  6.53it/s, loss=3.57, v_num=0]
Epoch 23:  89%|████████▊ | 86/97 [00:13<00:01,  6.57it/s, loss=3.57, v_num=0]
Epoch 23:  89%|████████▊ | 86/97 [00:13<00:01,  6.54it/s, loss=3.58, v_num=0]
Epoch 23:  90%|████████▉ | 87/97 [00:13<00:01,  6.58it/s, loss=3.58, v_num=0]
Epoch 23:  90%|████████▉ | 87/97 [00:13<00:01,  6.55it/s, loss=3.57, v_num=0]
Epoch 23:  91%|█████████ | 88/97 [00:13<00:01,  6.58it/s, loss=3.57, v_num=0]
Epoch 23:  91%|█████████ | 88/97 [00:13<00:01,  6.55it/s, loss=3.57, v_num=0]
Epoch 23:  92%|█████████▏| 89/97 [00:13<00:01,  6.59it/s, loss=3.57, v_num=0]
Epoch 23:  92%|█████████▏| 89/97 [00:13<00:01,  6.56it/s, loss=3.56, v_num=0]
Epoch 23:  93%|█████████▎| 90/97 [00:13<00:01,  6.60it/s, loss=3.56, v_num=0]
Epoch 23:  93%|█████████▎| 90/97 [00:13<00:01,  6.57it/s, loss=3.56, v_num=0]
Epoch 23:  94%|█████████▍| 91/97 [00:13<00:00,  6.60it/s, loss=3.56, v_num=0]
Epoch 23:  94%|█████████▍| 91/97 [00:13<00:00,  6.58it/s, loss=3.56, v_num=0]
Epoch 23:  95%|█████████▍| 92/97 [00:13<00:00,  6.61it/s, loss=3.56, v_num=0]
Epoch 23:  95%|█████████▍| 92/97 [00:13<00:00,  6.58it/s, loss=3.56, v_num=0]
Epoch 23:  96%|█████████▌| 93/97 [00:14<00:00,  6.62it/s, loss=3.56, v_num=0]
Epoch 23:  96%|█████████▌| 93/97 [00:14<00:00,  6.59it/s, loss=3.57, v_num=0]
Epoch 23:  97%|█████████▋| 94/97 [00:14<00:00,  6.62it/s, loss=3.57, v_num=0]
Epoch 23:  97%|█████████▋| 94/97 [00:14<00:00,  6.60it/s, loss=3.56, v_num=0]
Epoch 23:  98%|█████████▊| 95/97 [00:14<00:00,  6.63it/s, loss=3.56, v_num=0]
Epoch 23:  98%|█████████▊| 95/97 [00:14<00:00,  6.61it/s, loss=3.56, v_num=0]
Epoch 23:  99%|█████████▉| 96/97 [00:14<00:00,  6.64it/s, loss=3.56, v_num=0]
Epoch 23:  99%|█████████▉| 96/97 [00:14<00:00,  6.61it/s, loss=3.55, v_num=0]
Epoch 23: 100%|██████████| 97/97 [00:14<00:00,  6.65it/s, loss=3.55, v_num=0]
Epoch 23: 100%|██████████| 97/97 [00:14<00:00,  6.62it/s, loss=3.55, v_num=0]
Epoch 23: 100%|██████████| 97/97 [00:14<00:00,  6.62it/s, loss=3.55, v_num=0]
Epoch 23:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.55, v_num=0]
Epoch 24:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.55, v_num=0]
Epoch 24:   1%|          | 1/97 [00:01<02:13,  1.39s/it, loss=3.55, v_num=0]
Epoch 24:   1%|          | 1/97 [00:01<02:18,  1.45s/it, loss=3.55, v_num=0]
Epoch 24:   2%|▏         | 2/97 [00:01<01:12,  1.30it/s, loss=3.55, v_num=0]
Epoch 24:   2%|▏         | 2/97 [00:01<01:15,  1.26it/s, loss=3.54, v_num=0]
Epoch 24:   3%|▎         | 3/97 [00:01<00:52,  1.80it/s, loss=3.54, v_num=0]
Epoch 24:   3%|▎         | 3/97 [00:01<00:54,  1.74it/s, loss=3.54, v_num=0]
Epoch 24:   4%|▍         | 4/97 [00:01<00:42,  2.21it/s, loss=3.54, v_num=0]
Epoch 24:   4%|▍         | 4/97 [00:01<00:43,  2.15it/s, loss=3.55, v_num=0]
Epoch 24:   5%|▌         | 5/97 [00:01<00:35,  2.57it/s, loss=3.55, v_num=0]
Epoch 24:   5%|▌         | 5/97 [00:01<00:36,  2.50it/s, loss=3.54, v_num=0]
Epoch 24:   6%|▌         | 6/97 [00:02<00:31,  2.89it/s, loss=3.54, v_num=0]
Epoch 24:   6%|▌         | 6/97 [00:02<00:32,  2.81it/s, loss=3.54, v_num=0]
Epoch 24:   7%|▋         | 7/97 [00:02<00:28,  3.16it/s, loss=3.54, v_num=0]
Epoch 24:   7%|▋         | 7/97 [00:02<00:29,  3.08it/s, loss=3.54, v_num=0]
Epoch 24:   8%|▊         | 8/97 [00:02<00:26,  3.41it/s, loss=3.54, v_num=0]
Epoch 24:   8%|▊         | 8/97 [00:02<00:26,  3.32it/s, loss=3.54, v_num=0]
Epoch 24:   9%|▉         | 9/97 [00:02<00:24,  3.60it/s, loss=3.54, v_num=0]
Epoch 24:   9%|▉         | 9/97 [00:02<00:24,  3.53it/s, loss=3.54, v_num=0]
Epoch 24:  10%|█         | 10/97 [00:02<00:22,  3.80it/s, loss=3.54, v_num=0]
Epoch 24:  10%|█         | 10/97 [00:02<00:23,  3.72it/s, loss=3.54, v_num=0]
Epoch 24:  11%|█▏        | 11/97 [00:02<00:21,  3.97it/s, loss=3.54, v_num=0]
Epoch 24:  11%|█▏        | 11/97 [00:02<00:22,  3.90it/s, loss=3.54, v_num=0]
Epoch 24:  12%|█▏        | 12/97 [00:02<00:20,  4.13it/s, loss=3.54, v_num=0]
Epoch 24:  12%|█▏        | 12/97 [00:02<00:20,  4.05it/s, loss=3.55, v_num=0]
Epoch 24:  13%|█▎        | 13/97 [00:03<00:19,  4.27it/s, loss=3.55, v_num=0]
Epoch 24:  13%|█▎        | 13/97 [00:03<00:20,  4.20it/s, loss=3.55, v_num=0]
Epoch 24:  14%|█▍        | 14/97 [00:03<00:18,  4.41it/s, loss=3.55, v_num=0]
Epoch 24:  14%|█▍        | 14/97 [00:03<00:19,  4.33it/s, loss=3.56, v_num=0]
Epoch 24:  15%|█▌        | 15/97 [00:03<00:18,  4.53it/s, loss=3.56, v_num=0]
Epoch 24:  15%|█▌        | 15/97 [00:03<00:18,  4.45it/s, loss=3.56, v_num=0]
Epoch 24:  16%|█▋        | 16/97 [00:03<00:17,  4.65it/s, loss=3.56, v_num=0]
Epoch 24:  16%|█▋        | 16/97 [00:03<00:17,  4.57it/s, loss=3.57, v_num=0]
Epoch 24:  18%|█▊        | 17/97 [00:03<00:16,  4.74it/s, loss=3.57, v_num=0]
Epoch 24:  18%|█▊        | 17/97 [00:03<00:17,  4.67it/s, loss=3.56, v_num=0]
Epoch 24:  19%|█▊        | 18/97 [00:03<00:16,  4.84it/s, loss=3.56, v_num=0]
Epoch 24:  19%|█▊        | 18/97 [00:03<00:16,  4.76it/s, loss=3.57, v_num=0]
Epoch 24:  20%|█▉        | 19/97 [00:03<00:15,  4.92it/s, loss=3.57, v_num=0]
Epoch 24:  20%|█▉        | 19/97 [00:03<00:16,  4.85it/s, loss=3.56, v_num=0]
Epoch 24:  21%|██        | 20/97 [00:03<00:15,  5.01it/s, loss=3.56, v_num=0]
Epoch 24:  21%|██        | 20/97 [00:04<00:15,  4.94it/s, loss=3.57, v_num=0]
Epoch 24:  22%|██▏       | 21/97 [00:04<00:14,  5.08it/s, loss=3.57, v_num=0]
Epoch 24:  22%|██▏       | 21/97 [00:04<00:15,  5.02it/s, loss=3.56, v_num=0]
Epoch 24:  23%|██▎       | 22/97 [00:04<00:14,  5.16it/s, loss=3.56, v_num=0]
Epoch 24:  23%|██▎       | 22/97 [00:04<00:14,  5.09it/s, loss=3.57, v_num=0]
Epoch 24:  24%|██▎       | 23/97 [00:04<00:14,  5.23it/s, loss=3.57, v_num=0]
Epoch 24:  24%|██▎       | 23/97 [00:04<00:14,  5.16it/s, loss=3.56, v_num=0]
Epoch 24:  25%|██▍       | 24/97 [00:04<00:13,  5.28it/s, loss=3.56, v_num=0]
Epoch 24:  25%|██▍       | 24/97 [00:04<00:13,  5.22it/s, loss=3.56, v_num=0]
Epoch 24:  26%|██▌       | 25/97 [00:04<00:13,  5.34it/s, loss=3.56, v_num=0]
Epoch 24:  26%|██▌       | 25/97 [00:04<00:13,  5.28it/s, loss=3.56, v_num=0]
Epoch 24:  27%|██▋       | 26/97 [00:04<00:13,  5.40it/s, loss=3.56, v_num=0]
Epoch 24:  27%|██▋       | 26/97 [00:04<00:13,  5.33it/s, loss=3.55, v_num=0]
Epoch 24:  28%|██▊       | 27/97 [00:04<00:12,  5.45it/s, loss=3.55, v_num=0]
Epoch 24:  28%|██▊       | 27/97 [00:05<00:12,  5.39it/s, loss=3.55, v_num=0]
Epoch 24:  29%|██▉       | 28/97 [00:05<00:12,  5.50it/s, loss=3.55, v_num=0]
Epoch 24:  29%|██▉       | 28/97 [00:05<00:12,  5.44it/s, loss=3.55, v_num=0]
Epoch 24:  30%|██▉       | 29/97 [00:05<00:12,  5.55it/s, loss=3.55, v_num=0]
Epoch 24:  30%|██▉       | 29/97 [00:05<00:12,  5.49it/s, loss=3.55, v_num=0]
Epoch 24:  31%|███       | 30/97 [00:05<00:11,  5.60it/s, loss=3.55, v_num=0]
Epoch 24:  31%|███       | 30/97 [00:05<00:12,  5.53it/s, loss=3.53, v_num=0]
Epoch 24:  32%|███▏      | 31/97 [00:05<00:11,  5.64it/s, loss=3.53, v_num=0]
Epoch 24:  32%|███▏      | 31/97 [00:05<00:11,  5.58it/s, loss=3.53, v_num=0]
Epoch 24:  33%|███▎      | 32/97 [00:05<00:11,  5.68it/s, loss=3.53, v_num=0]
Epoch 24:  33%|███▎      | 32/97 [00:05<00:11,  5.62it/s, loss=3.52, v_num=0]
Epoch 24:  34%|███▍      | 33/97 [00:05<00:11,  5.71it/s, loss=3.52, v_num=0]
Epoch 24:  34%|███▍      | 33/97 [00:05<00:11,  5.66it/s, loss=3.53, v_num=0]
Epoch 24:  35%|███▌      | 34/97 [00:05<00:10,  5.75it/s, loss=3.53, v_num=0]
Epoch 24:  35%|███▌      | 34/97 [00:05<00:11,  5.70it/s, loss=3.52, v_num=0]
Epoch 24:  36%|███▌      | 35/97 [00:06<00:10,  5.79it/s, loss=3.52, v_num=0]
Epoch 24:  36%|███▌      | 35/97 [00:06<00:10,  5.73it/s, loss=3.51, v_num=0]
Epoch 24:  37%|███▋      | 36/97 [00:06<00:10,  5.82it/s, loss=3.51, v_num=0]
Epoch 24:  37%|███▋      | 36/97 [00:06<00:10,  5.77it/s, loss=3.5, v_num=0]
Epoch 24:  38%|███▊      | 37/97 [00:06<00:10,  5.85it/s, loss=3.5, v_num=0]
Epoch 24:  38%|███▊      | 37/97 [00:06<00:10,  5.80it/s, loss=3.51, v_num=0]
Epoch 24:  39%|███▉      | 38/97 [00:06<00:10,  5.89it/s, loss=3.51, v_num=0]
Epoch 24:  39%|███▉      | 38/97 [00:06<00:10,  5.83it/s, loss=3.51, v_num=0]
Epoch 24:  40%|████      | 39/97 [00:06<00:09,  5.91it/s, loss=3.51, v_num=0]
Epoch 24:  40%|████      | 39/97 [00:06<00:09,  5.86it/s, loss=3.51, v_num=0]
Epoch 24:  41%|████      | 40/97 [00:06<00:09,  5.94it/s, loss=3.51, v_num=0]
Epoch 24:  41%|████      | 40/97 [00:06<00:09,  5.89it/s, loss=3.5, v_num=0]
Epoch 24:  42%|████▏     | 41/97 [00:06<00:09,  5.97it/s, loss=3.5, v_num=0]
Epoch 24:  42%|████▏     | 41/97 [00:06<00:09,  5.92it/s, loss=3.49, v_num=0]
Epoch 24:  43%|████▎     | 42/97 [00:07<00:09,  6.00it/s, loss=3.49, v_num=0]
Epoch 24:  43%|████▎     | 42/97 [00:07<00:09,  5.95it/s, loss=3.5, v_num=0]
Epoch 24:  44%|████▍     | 43/97 [00:07<00:08,  6.02it/s, loss=3.5, v_num=0]
Epoch 24:  44%|████▍     | 43/97 [00:07<00:09,  5.97it/s, loss=3.5, v_num=0]
Epoch 24:  45%|████▌     | 44/97 [00:07<00:08,  6.04it/s, loss=3.5, v_num=0]
Epoch 24:  45%|████▌     | 44/97 [00:07<00:08,  6.00it/s, loss=3.5, v_num=0]
Epoch 24:  46%|████▋     | 45/97 [00:07<00:08,  6.06it/s, loss=3.5, v_num=0]
Epoch 24:  46%|████▋     | 45/97 [00:07<00:08,  6.02it/s, loss=3.5, v_num=0]
Epoch 24:  47%|████▋     | 46/97 [00:07<00:08,  6.09it/s, loss=3.5, v_num=0]
Epoch 24:  47%|████▋     | 46/97 [00:07<00:08,  6.04it/s, loss=3.52, v_num=0]
Epoch 24:  48%|████▊     | 47/97 [00:07<00:08,  6.11it/s, loss=3.52, v_num=0]
Epoch 24:  48%|████▊     | 47/97 [00:07<00:08,  6.06it/s, loss=3.52, v_num=0]
Epoch 24:  49%|████▉     | 48/97 [00:07<00:07,  6.13it/s, loss=3.52, v_num=0]
Epoch 24:  49%|████▉     | 48/97 [00:07<00:08,  6.09it/s, loss=3.51, v_num=0]
Epoch 24:  51%|█████     | 49/97 [00:07<00:07,  6.15it/s, loss=3.51, v_num=0]
Epoch 24:  51%|█████     | 49/97 [00:08<00:07,  6.10it/s, loss=3.52, v_num=0]
Epoch 24:  52%|█████▏    | 50/97 [00:08<00:07,  6.16it/s, loss=3.52, v_num=0]
Epoch 24:  52%|█████▏    | 50/97 [00:08<00:07,  6.12it/s, loss=3.52, v_num=0]
Epoch 24:  53%|█████▎    | 51/97 [00:08<00:07,  6.18it/s, loss=3.52, v_num=0]
Epoch 24:  53%|█████▎    | 51/97 [00:08<00:07,  6.14it/s, loss=3.52, v_num=0]
Epoch 24:  54%|█████▎    | 52/97 [00:08<00:07,  6.20it/s, loss=3.52, v_num=0]
Epoch 24:  54%|█████▎    | 52/97 [00:08<00:07,  6.16it/s, loss=3.52, v_num=0]
Epoch 24:  55%|█████▍    | 53/97 [00:08<00:07,  6.22it/s, loss=3.52, v_num=0]
Epoch 24:  55%|█████▍    | 53/97 [00:08<00:07,  6.18it/s, loss=3.51, v_num=0]
Epoch 24:  56%|█████▌    | 54/97 [00:08<00:06,  6.24it/s, loss=3.51, v_num=0]
Epoch 24:  56%|█████▌    | 54/97 [00:08<00:06,  6.19it/s, loss=3.51, v_num=0]
Epoch 24:  57%|█████▋    | 55/97 [00:08<00:06,  6.25it/s, loss=3.51, v_num=0]
Epoch 24:  57%|█████▋    | 55/97 [00:08<00:06,  6.21it/s, loss=3.51, v_num=0]
Epoch 24:  58%|█████▊    | 56/97 [00:08<00:06,  6.27it/s, loss=3.51, v_num=0]
Epoch 24:  58%|█████▊    | 56/97 [00:08<00:06,  6.23it/s, loss=3.51, v_num=0]
Epoch 24:  59%|█████▉    | 57/97 [00:09<00:06,  6.28it/s, loss=3.51, v_num=0]
Epoch 24:  59%|█████▉    | 57/97 [00:09<00:06,  6.24it/s, loss=3.51, v_num=0]
Epoch 24:  60%|█████▉    | 58/97 [00:09<00:06,  6.29it/s, loss=3.51, v_num=0]
Epoch 24:  60%|█████▉    | 58/97 [00:09<00:06,  6.26it/s, loss=3.51, v_num=0]
Epoch 24:  61%|██████    | 59/97 [00:09<00:06,  6.31it/s, loss=3.51, v_num=0]
Epoch 24:  61%|██████    | 59/97 [00:09<00:06,  6.27it/s, loss=3.51, v_num=0]
Epoch 24:  62%|██████▏   | 60/97 [00:09<00:05,  6.32it/s, loss=3.51, v_num=0]
Epoch 24:  62%|██████▏   | 60/97 [00:09<00:05,  6.29it/s, loss=3.51, v_num=0]
Epoch 24:  63%|██████▎   | 61/97 [00:09<00:05,  6.34it/s, loss=3.51, v_num=0]
Epoch 24:  63%|██████▎   | 61/97 [00:09<00:05,  6.30it/s, loss=3.52, v_num=0]
Epoch 24:  64%|██████▍   | 62/97 [00:09<00:05,  6.35it/s, loss=3.52, v_num=0]
Epoch 24:  64%|██████▍   | 62/97 [00:09<00:05,  6.31it/s, loss=3.51, v_num=0]
Epoch 24:  65%|██████▍   | 63/97 [00:09<00:05,  6.36it/s, loss=3.51, v_num=0]
Epoch 24:  65%|██████▍   | 63/97 [00:09<00:05,  6.33it/s, loss=3.5, v_num=0]
Epoch 24:  66%|██████▌   | 64/97 [00:10<00:05,  6.38it/s, loss=3.5, v_num=0]
Epoch 24:  66%|██████▌   | 64/97 [00:10<00:05,  6.34it/s, loss=3.5, v_num=0]
Epoch 24:  67%|██████▋   | 65/97 [00:10<00:05,  6.39it/s, loss=3.5, v_num=0]
Epoch 24:  67%|██████▋   | 65/97 [00:10<00:05,  6.35it/s, loss=3.49, v_num=0]
Epoch 24:  68%|██████▊   | 66/97 [00:10<00:04,  6.40it/s, loss=3.49, v_num=0]
Epoch 24:  68%|██████▊   | 66/97 [00:10<00:04,  6.37it/s, loss=3.48, v_num=0]
Epoch 24:  69%|██████▉   | 67/97 [00:10<00:04,  6.41it/s, loss=3.48, v_num=0]
Epoch 24:  69%|██████▉   | 67/97 [00:10<00:04,  6.38it/s, loss=3.48, v_num=0]
Epoch 24:  70%|███████   | 68/97 [00:10<00:04,  6.42it/s, loss=3.48, v_num=0]
Epoch 24:  70%|███████   | 68/97 [00:10<00:04,  6.39it/s, loss=3.49, v_num=0]
Epoch 24:  71%|███████   | 69/97 [00:10<00:04,  6.44it/s, loss=3.49, v_num=0]
Epoch 24:  71%|███████   | 69/97 [00:10<00:04,  6.40it/s, loss=3.49, v_num=0]
Epoch 24:  72%|███████▏  | 70/97 [00:10<00:04,  6.45it/s, loss=3.49, v_num=0]
Epoch 24:  72%|███████▏  | 70/97 [00:10<00:04,  6.41it/s, loss=3.49, v_num=0]
Epoch 24:  73%|███████▎  | 71/97 [00:10<00:04,  6.46it/s, loss=3.49, v_num=0]
Epoch 24:  73%|███████▎  | 71/97 [00:11<00:04,  6.43it/s, loss=3.49, v_num=0]
Epoch 24:  74%|███████▍  | 72/97 [00:11<00:03,  6.47it/s, loss=3.49, v_num=0]
Epoch 24:  74%|███████▍  | 72/97 [00:11<00:03,  6.44it/s, loss=3.5, v_num=0]
Epoch 24:  75%|███████▌  | 73/97 [00:11<00:03,  6.48it/s, loss=3.5, v_num=0]
Epoch 24:  75%|███████▌  | 73/97 [00:11<00:03,  6.45it/s, loss=3.5, v_num=0]
Epoch 24:  76%|███████▋  | 74/97 [00:11<00:03,  6.49it/s, loss=3.5, v_num=0]
Epoch 24:  76%|███████▋  | 74/97 [00:11<00:03,  6.46it/s, loss=3.5, v_num=0]
Epoch 24:  77%|███████▋  | 75/97 [00:11<00:03,  6.50it/s, loss=3.5, v_num=0]
Epoch 24:  77%|███████▋  | 75/97 [00:11<00:03,  6.47it/s, loss=3.5, v_num=0]
Epoch 24:  78%|███████▊  | 76/97 [00:11<00:03,  6.51it/s, loss=3.5, v_num=0]
Epoch 24:  78%|███████▊  | 76/97 [00:11<00:03,  6.48it/s, loss=3.5, v_num=0]
Epoch 24:  79%|███████▉  | 77/97 [00:11<00:03,  6.52it/s, loss=3.5, v_num=0]
Epoch 24:  79%|███████▉  | 77/97 [00:11<00:03,  6.49it/s, loss=3.49, v_num=0]
Epoch 24:  80%|████████  | 78/97 [00:11<00:02,  6.52it/s, loss=3.49, v_num=0]
Epoch 24:  80%|████████  | 78/97 [00:12<00:02,  6.49it/s, loss=3.48, v_num=0]
Epoch 24:  81%|████████▏ | 79/97 [00:12<00:02,  6.53it/s, loss=3.48, v_num=0]
Epoch 24:  81%|████████▏ | 79/97 [00:12<00:02,  6.50it/s, loss=3.48, v_num=0]
Epoch 24:  82%|████████▏ | 80/97 [00:12<00:02,  6.54it/s, loss=3.48, v_num=0]
Epoch 24:  82%|████████▏ | 80/97 [00:12<00:02,  6.51it/s, loss=3.48, v_num=0]
Epoch 24:  84%|████████▎ | 81/97 [00:12<00:02,  6.55it/s, loss=3.48, v_num=0]
Epoch 24:  84%|████████▎ | 81/97 [00:12<00:02,  6.52it/s, loss=3.49, v_num=0]
Epoch 24:  85%|████████▍ | 82/97 [00:12<00:02,  6.55it/s, loss=3.49, v_num=0]
Epoch 24:  85%|████████▍ | 82/97 [00:12<00:02,  6.53it/s, loss=3.49, v_num=0]
Epoch 24:  86%|████████▌ | 83/97 [00:12<00:02,  6.56it/s, loss=3.49, v_num=0]
Epoch 24:  86%|████████▌ | 83/97 [00:12<00:02,  6.53it/s, loss=3.48, v_num=0]
Epoch 24:  87%|████████▋ | 84/97 [00:12<00:01,  6.57it/s, loss=3.48, v_num=0]
Epoch 24:  87%|████████▋ | 84/97 [00:12<00:01,  6.54it/s, loss=3.49, v_num=0]
Epoch 24:  88%|████████▊ | 85/97 [00:12<00:01,  6.58it/s, loss=3.49, v_num=0]
Epoch 24:  88%|████████▊ | 85/97 [00:12<00:01,  6.55it/s, loss=3.49, v_num=0]
Epoch 24:  89%|████████▊ | 86/97 [00:13<00:01,  6.59it/s, loss=3.49, v_num=0]
Epoch 24:  89%|████████▊ | 86/97 [00:13<00:01,  6.56it/s, loss=3.49, v_num=0]
Epoch 24:  90%|████████▉ | 87/97 [00:13<00:01,  6.59it/s, loss=3.49, v_num=0]
Epoch 24:  90%|████████▉ | 87/97 [00:13<00:01,  6.57it/s, loss=3.48, v_num=0]
Epoch 24:  91%|█████████ | 88/97 [00:13<00:01,  6.60it/s, loss=3.48, v_num=0]
Epoch 24:  91%|█████████ | 88/97 [00:13<00:01,  6.57it/s, loss=3.48, v_num=0]
Epoch 24:  92%|█████████▏| 89/97 [00:13<00:01,  6.61it/s, loss=3.48, v_num=0]
Epoch 24:  92%|█████████▏| 89/97 [00:13<00:01,  6.58it/s, loss=3.48, v_num=0]
Epoch 24:  93%|█████████▎| 90/97 [00:13<00:01,  6.62it/s, loss=3.48, v_num=0]
Epoch 24:  93%|█████████▎| 90/97 [00:13<00:01,  6.59it/s, loss=3.48, v_num=0]
Epoch 24:  94%|█████████▍| 91/97 [00:13<00:00,  6.62it/s, loss=3.48, v_num=0]
Epoch 24:  94%|█████████▍| 91/97 [00:13<00:00,  6.60it/s, loss=3.48, v_num=0]
Epoch 24:  95%|█████████▍| 92/97 [00:13<00:00,  6.63it/s, loss=3.48, v_num=0]
Epoch 24:  95%|█████████▍| 92/97 [00:13<00:00,  6.60it/s, loss=3.48, v_num=0]
Epoch 24:  96%|█████████▌| 93/97 [00:14<00:00,  6.64it/s, loss=3.48, v_num=0]
Epoch 24:  96%|█████████▌| 93/97 [00:14<00:00,  6.61it/s, loss=3.47, v_num=0]
Epoch 24:  97%|█████████▋| 94/97 [00:14<00:00,  6.64it/s, loss=3.47, v_num=0]
Epoch 24:  97%|█████████▋| 94/97 [00:14<00:00,  6.62it/s, loss=3.47, v_num=0]
Epoch 24:  98%|█████████▊| 95/97 [00:14<00:00,  6.65it/s, loss=3.47, v_num=0]
Epoch 24:  98%|█████████▊| 95/97 [00:14<00:00,  6.62it/s, loss=3.47, v_num=0]
Epoch 24:  99%|█████████▉| 96/97 [00:14<00:00,  6.66it/s, loss=3.47, v_num=0]
Epoch 24:  99%|█████████▉| 96/97 [00:14<00:00,  6.63it/s, loss=3.48, v_num=0]
Epoch 24: 100%|██████████| 97/97 [00:14<00:00,  6.66it/s, loss=3.48, v_num=0]
Epoch 24: 100%|██████████| 97/97 [00:14<00:00,  6.64it/s, loss=3.48, v_num=0]
Epoch 24: 100%|██████████| 97/97 [00:14<00:00,  6.64it/s, loss=3.48, v_num=0]
Epoch 24:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.48, v_num=0]
Epoch 25:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.48, v_num=0]
Epoch 25:   1%|          | 1/97 [00:01<02:12,  1.38s/it, loss=3.48, v_num=0]
Epoch 25:   1%|          | 1/97 [00:01<02:17,  1.43s/it, loss=3.48, v_num=0]
Epoch 25:   2%|▏         | 2/97 [00:01<01:11,  1.33it/s, loss=3.48, v_num=0]
Epoch 25:   2%|▏         | 2/97 [00:01<01:14,  1.28it/s, loss=3.48, v_num=0]
Epoch 25:   3%|▎         | 3/97 [00:01<00:51,  1.82it/s, loss=3.48, v_num=0]
Epoch 25:   3%|▎         | 3/97 [00:01<00:53,  1.76it/s, loss=3.48, v_num=0]
Epoch 25:   4%|▍         | 4/97 [00:01<00:41,  2.24it/s, loss=3.48, v_num=0]
Epoch 25:   4%|▍         | 4/97 [00:01<00:42,  2.17it/s, loss=3.47, v_num=0]
Epoch 25:   5%|▌         | 5/97 [00:01<00:35,  2.60it/s, loss=3.47, v_num=0]
Epoch 25:   5%|▌         | 5/97 [00:01<00:36,  2.53it/s, loss=3.47, v_num=0]
Epoch 25:   6%|▌         | 6/97 [00:02<00:31,  2.91it/s, loss=3.47, v_num=0]
Epoch 25:   6%|▌         | 6/97 [00:02<00:32,  2.83it/s, loss=3.48, v_num=0]
Epoch 25:   7%|▋         | 7/97 [00:02<00:28,  3.18it/s, loss=3.48, v_num=0]
Epoch 25:   7%|▋         | 7/97 [00:02<00:28,  3.11it/s, loss=3.47, v_num=0]
Epoch 25:   8%|▊         | 8/97 [00:02<00:25,  3.43it/s, loss=3.47, v_num=0]
Epoch 25:   8%|▊         | 8/97 [00:02<00:26,  3.35it/s, loss=3.47, v_num=0]
Epoch 25:   9%|▉         | 9/97 [00:02<00:24,  3.63it/s, loss=3.47, v_num=0]
Epoch 25:   9%|▉         | 9/97 [00:02<00:24,  3.55it/s, loss=3.47, v_num=0]
Epoch 25:  10%|█         | 10/97 [00:02<00:22,  3.83it/s, loss=3.47, v_num=0]
Epoch 25:  10%|█         | 10/97 [00:02<00:23,  3.74it/s, loss=3.47, v_num=0]
Epoch 25:  11%|█▏        | 11/97 [00:02<00:21,  4.00it/s, loss=3.47, v_num=0]
Epoch 25:  11%|█▏        | 11/97 [00:02<00:21,  3.92it/s, loss=3.47, v_num=0]
Epoch 25:  12%|█▏        | 12/97 [00:02<00:20,  4.16it/s, loss=3.47, v_num=0]
Epoch 25:  12%|█▏        | 12/97 [00:02<00:20,  4.08it/s, loss=3.46, v_num=0]
Epoch 25:  13%|█▎        | 13/97 [00:03<00:19,  4.30it/s, loss=3.46, v_num=0]
Epoch 25:  13%|█▎        | 13/97 [00:03<00:19,  4.22it/s, loss=3.47, v_num=0]
Epoch 25:  14%|█▍        | 14/97 [00:03<00:18,  4.44it/s, loss=3.47, v_num=0]
Epoch 25:  14%|█▍        | 14/97 [00:03<00:19,  4.35it/s, loss=3.47, v_num=0]
Epoch 25:  15%|█▌        | 15/97 [00:03<00:18,  4.55it/s, loss=3.47, v_num=0]
Epoch 25:  15%|█▌        | 15/97 [00:03<00:18,  4.48it/s, loss=3.47, v_num=0]
Epoch 25:  16%|█▋        | 16/97 [00:03<00:17,  4.66it/s, loss=3.47, v_num=0]
Epoch 25:  16%|█▋        | 16/97 [00:03<00:17,  4.59it/s, loss=3.47, v_num=0]
Epoch 25:  18%|█▊        | 17/97 [00:03<00:16,  4.76it/s, loss=3.47, v_num=0]
Epoch 25:  18%|█▊        | 17/97 [00:03<00:17,  4.69it/s, loss=3.47, v_num=0]
Epoch 25:  19%|█▊        | 18/97 [00:03<00:16,  4.86it/s, loss=3.47, v_num=0]
Epoch 25:  19%|█▊        | 18/97 [00:03<00:16,  4.78it/s, loss=3.47, v_num=0]
Epoch 25:  20%|█▉        | 19/97 [00:03<00:15,  4.94it/s, loss=3.47, v_num=0]
Epoch 25:  20%|█▉        | 19/97 [00:03<00:16,  4.87it/s, loss=3.47, v_num=0]
Epoch 25:  21%|██        | 20/97 [00:03<00:15,  5.03it/s, loss=3.47, v_num=0]
Epoch 25:  21%|██        | 20/97 [00:04<00:15,  4.96it/s, loss=3.46, v_num=0]
Epoch 25:  22%|██▏       | 21/97 [00:04<00:14,  5.10it/s, loss=3.46, v_num=0]
Epoch 25:  22%|██▏       | 21/97 [00:04<00:15,  5.04it/s, loss=3.47, v_num=0]
Epoch 25:  23%|██▎       | 22/97 [00:04<00:14,  5.17it/s, loss=3.47, v_num=0]
Epoch 25:  23%|██▎       | 22/97 [00:04<00:14,  5.11it/s, loss=3.47, v_num=0]
Epoch 25:  24%|██▎       | 23/97 [00:04<00:14,  5.24it/s, loss=3.47, v_num=0]
Epoch 25:  24%|██▎       | 23/97 [00:04<00:14,  5.18it/s, loss=3.47, v_num=0]
Epoch 25:  25%|██▍       | 24/97 [00:04<00:13,  5.30it/s, loss=3.47, v_num=0]
Epoch 25:  25%|██▍       | 24/97 [00:04<00:13,  5.24it/s, loss=3.47, v_num=0]
Epoch 25:  26%|██▌       | 25/97 [00:04<00:13,  5.36it/s, loss=3.47, v_num=0]
Epoch 25:  26%|██▌       | 25/97 [00:04<00:13,  5.30it/s, loss=3.46, v_num=0]
Epoch 25:  27%|██▋       | 26/97 [00:04<00:13,  5.42it/s, loss=3.46, v_num=0]
Epoch 25:  27%|██▋       | 26/97 [00:04<00:13,  5.35it/s, loss=3.47, v_num=0]
Epoch 25:  28%|██▊       | 27/97 [00:04<00:12,  5.47it/s, loss=3.47, v_num=0]
Epoch 25:  28%|██▊       | 27/97 [00:04<00:12,  5.41it/s, loss=3.47, v_num=0]
Epoch 25:  29%|██▉       | 28/97 [00:05<00:12,  5.52it/s, loss=3.47, v_num=0]
Epoch 25:  29%|██▉       | 28/97 [00:05<00:12,  5.46it/s, loss=3.47, v_num=0]
Epoch 25:  30%|██▉       | 29/97 [00:05<00:12,  5.56it/s, loss=3.47, v_num=0]
Epoch 25:  30%|██▉       | 29/97 [00:05<00:12,  5.51it/s, loss=3.47, v_num=0]
Epoch 25:  31%|███       | 30/97 [00:05<00:11,  5.61it/s, loss=3.47, v_num=0]
Epoch 25:  31%|███       | 30/97 [00:05<00:12,  5.55it/s, loss=3.46, v_num=0]
Epoch 25:  32%|███▏      | 31/97 [00:05<00:11,  5.65it/s, loss=3.46, v_num=0]
Epoch 25:  32%|███▏      | 31/97 [00:05<00:11,  5.59it/s, loss=3.47, v_num=0]
Epoch 25:  33%|███▎      | 32/97 [00:05<00:11,  5.69it/s, loss=3.47, v_num=0]
Epoch 25:  33%|███▎      | 32/97 [00:05<00:11,  5.63it/s, loss=3.46, v_num=0]
Epoch 25:  34%|███▍      | 33/97 [00:05<00:11,  5.73it/s, loss=3.46, v_num=0]
Epoch 25:  34%|███▍      | 33/97 [00:05<00:11,  5.67it/s, loss=3.46, v_num=0]
Epoch 25:  35%|███▌      | 34/97 [00:05<00:10,  5.76it/s, loss=3.46, v_num=0]
Epoch 25:  35%|███▌      | 34/97 [00:05<00:11,  5.71it/s, loss=3.46, v_num=0]
Epoch 25:  36%|███▌      | 35/97 [00:06<00:10,  5.80it/s, loss=3.46, v_num=0]
Epoch 25:  36%|███▌      | 35/97 [00:06<00:10,  5.75it/s, loss=3.46, v_num=0]
Epoch 25:  37%|███▋      | 36/97 [00:06<00:10,  5.83it/s, loss=3.46, v_num=0]
Epoch 25:  37%|███▋      | 36/97 [00:06<00:10,  5.78it/s, loss=3.47, v_num=0]
Epoch 25:  38%|███▊      | 37/97 [00:06<00:10,  5.86it/s, loss=3.47, v_num=0]
Epoch 25:  38%|███▊      | 37/97 [00:06<00:10,  5.81it/s, loss=3.47, v_num=0]
Epoch 25:  39%|███▉      | 38/97 [00:06<00:10,  5.89it/s, loss=3.47, v_num=0]
Epoch 25:  39%|███▉      | 38/97 [00:06<00:10,  5.84it/s, loss=3.47, v_num=0]
Epoch 25:  40%|████      | 39/97 [00:06<00:09,  5.92it/s, loss=3.47, v_num=0]
Epoch 25:  40%|████      | 39/97 [00:06<00:09,  5.87it/s, loss=3.47, v_num=0]
Epoch 25:  41%|████      | 40/97 [00:06<00:09,  5.95it/s, loss=3.47, v_num=0]
Epoch 25:  41%|████      | 40/97 [00:06<00:09,  5.90it/s, loss=3.47, v_num=0]
Epoch 25:  42%|████▏     | 41/97 [00:06<00:09,  5.98it/s, loss=3.47, v_num=0]
Epoch 25:  42%|████▏     | 41/97 [00:06<00:09,  5.93it/s, loss=3.47, v_num=0]
Epoch 25:  43%|████▎     | 42/97 [00:06<00:09,  6.00it/s, loss=3.47, v_num=0]
Epoch 25:  43%|████▎     | 42/97 [00:07<00:09,  5.96it/s, loss=3.47, v_num=0]
Epoch 25:  44%|████▍     | 43/97 [00:07<00:08,  6.03it/s, loss=3.47, v_num=0]
Epoch 25:  44%|████▍     | 43/97 [00:07<00:09,  5.98it/s, loss=3.47, v_num=0]
Epoch 25:  45%|████▌     | 44/97 [00:07<00:08,  6.05it/s, loss=3.47, v_num=0]
Epoch 25:  45%|████▌     | 44/97 [00:07<00:08,  6.01it/s, loss=3.47, v_num=0]
Epoch 25:  46%|████▋     | 45/97 [00:07<00:08,  6.07it/s, loss=3.47, v_num=0]
Epoch 25:  46%|████▋     | 45/97 [00:07<00:08,  6.03it/s, loss=3.47, v_num=0]
Epoch 25:  47%|████▋     | 46/97 [00:07<00:08,  6.10it/s, loss=3.47, v_num=0]
Epoch 25:  47%|████▋     | 46/97 [00:07<00:08,  6.05it/s, loss=3.47, v_num=0]
Epoch 25:  48%|████▊     | 47/97 [00:07<00:08,  6.12it/s, loss=3.47, v_num=0]
Epoch 25:  48%|████▊     | 47/97 [00:07<00:08,  6.07it/s, loss=3.47, v_num=0]
Epoch 25:  49%|████▉     | 48/97 [00:07<00:07,  6.13it/s, loss=3.47, v_num=0]
Epoch 25:  49%|████▉     | 48/97 [00:07<00:08,  6.09it/s, loss=3.48, v_num=0]
Epoch 25:  51%|█████     | 49/97 [00:07<00:07,  6.15it/s, loss=3.48, v_num=0]
Epoch 25:  51%|█████     | 49/97 [00:08<00:07,  6.11it/s, loss=3.48, v_num=0]
Epoch 25:  52%|█████▏    | 50/97 [00:08<00:07,  6.17it/s, loss=3.48, v_num=0]
Epoch 25:  52%|█████▏    | 50/97 [00:08<00:07,  6.13it/s, loss=3.48, v_num=0]
Epoch 25:  53%|█████▎    | 51/97 [00:08<00:07,  6.19it/s, loss=3.48, v_num=0]
Epoch 25:  53%|█████▎    | 51/97 [00:08<00:07,  6.14it/s, loss=3.48, v_num=0]
Epoch 25:  54%|█████▎    | 52/97 [00:08<00:07,  6.20it/s, loss=3.48, v_num=0]
Epoch 25:  54%|█████▎    | 52/97 [00:08<00:07,  6.16it/s, loss=3.49, v_num=0]
Epoch 25:  55%|█████▍    | 53/97 [00:08<00:07,  6.22it/s, loss=3.49, v_num=0]
Epoch 25:  55%|█████▍    | 53/97 [00:08<00:07,  6.18it/s, loss=3.48, v_num=0]
Epoch 25:  56%|█████▌    | 54/97 [00:08<00:06,  6.24it/s, loss=3.48, v_num=0]
Epoch 25:  56%|█████▌    | 54/97 [00:08<00:06,  6.20it/s, loss=3.48, v_num=0]
Epoch 25:  57%|█████▋    | 55/97 [00:08<00:06,  6.25it/s, loss=3.48, v_num=0]
Epoch 25:  57%|█████▋    | 55/97 [00:08<00:06,  6.22it/s, loss=3.48, v_num=0]
Epoch 25:  58%|█████▊    | 56/97 [00:08<00:06,  6.27it/s, loss=3.48, v_num=0]
Epoch 25:  58%|█████▊    | 56/97 [00:08<00:06,  6.23it/s, loss=3.47, v_num=0]
Epoch 25:  59%|█████▉    | 57/97 [00:09<00:06,  6.28it/s, loss=3.47, v_num=0]
Epoch 25:  59%|█████▉    | 57/97 [00:09<00:06,  6.24it/s, loss=3.48, v_num=0]
Epoch 25:  60%|█████▉    | 58/97 [00:09<00:06,  6.30it/s, loss=3.48, v_num=0]
Epoch 25:  60%|█████▉    | 58/97 [00:09<00:06,  6.26it/s, loss=3.47, v_num=0]
Epoch 25:  61%|██████    | 59/97 [00:09<00:06,  6.31it/s, loss=3.47, v_num=0]
Epoch 25:  61%|██████    | 59/97 [00:09<00:06,  6.27it/s, loss=3.48, v_num=0]
Epoch 25:  62%|██████▏   | 60/97 [00:09<00:05,  6.32it/s, loss=3.48, v_num=0]
Epoch 25:  62%|██████▏   | 60/97 [00:09<00:05,  6.28it/s, loss=3.48, v_num=0]
Epoch 25:  63%|██████▎   | 61/97 [00:09<00:05,  6.33it/s, loss=3.48, v_num=0]
Epoch 25:  63%|██████▎   | 61/97 [00:09<00:05,  6.30it/s, loss=3.48, v_num=0]
Epoch 25:  64%|██████▍   | 62/97 [00:09<00:05,  6.34it/s, loss=3.48, v_num=0]
Epoch 25:  64%|██████▍   | 62/97 [00:09<00:05,  6.31it/s, loss=3.48, v_num=0]
Epoch 25:  65%|██████▍   | 63/97 [00:09<00:05,  6.36it/s, loss=3.48, v_num=0]
Epoch 25:  65%|██████▍   | 63/97 [00:09<00:05,  6.32it/s, loss=3.49, v_num=0]
Epoch 25:  66%|██████▌   | 64/97 [00:10<00:05,  6.37it/s, loss=3.49, v_num=0]
Epoch 25:  66%|██████▌   | 64/97 [00:10<00:05,  6.33it/s, loss=3.49, v_num=0]
Epoch 25:  67%|██████▋   | 65/97 [00:10<00:05,  6.38it/s, loss=3.49, v_num=0]
Epoch 25:  67%|██████▋   | 65/97 [00:10<00:05,  6.35it/s, loss=3.49, v_num=0]
Epoch 25:  68%|██████▊   | 66/97 [00:10<00:04,  6.39it/s, loss=3.49, v_num=0]
Epoch 25:  68%|██████▊   | 66/97 [00:10<00:04,  6.36it/s, loss=3.49, v_num=0]
Epoch 25:  69%|██████▉   | 67/97 [00:10<00:04,  6.41it/s, loss=3.49, v_num=0]
Epoch 25:  69%|██████▉   | 67/97 [00:10<00:04,  6.37it/s, loss=3.48, v_num=0]
Epoch 25:  70%|███████   | 68/97 [00:10<00:04,  6.42it/s, loss=3.48, v_num=0]
Epoch 25:  70%|███████   | 68/97 [00:10<00:04,  6.38it/s, loss=3.48, v_num=0]
Epoch 25:  71%|███████   | 69/97 [00:10<00:04,  6.43it/s, loss=3.48, v_num=0]
Epoch 25:  71%|███████   | 69/97 [00:10<00:04,  6.39it/s, loss=3.48, v_num=0]
Epoch 25:  72%|███████▏  | 70/97 [00:10<00:04,  6.44it/s, loss=3.48, v_num=0]
Epoch 25:  72%|███████▏  | 70/97 [00:10<00:04,  6.41it/s, loss=3.48, v_num=0]
Epoch 25:  73%|███████▎  | 71/97 [00:11<00:04,  6.45it/s, loss=3.48, v_num=0]
Epoch 25:  73%|███████▎  | 71/97 [00:11<00:04,  6.41it/s, loss=3.48, v_num=0]
Epoch 25:  74%|███████▍  | 72/97 [00:11<00:03,  6.46it/s, loss=3.48, v_num=0]
Epoch 25:  74%|███████▍  | 72/97 [00:11<00:03,  6.42it/s, loss=3.48, v_num=0]
Epoch 25:  75%|███████▌  | 73/97 [00:11<00:03,  6.47it/s, loss=3.48, v_num=0]
Epoch 25:  75%|███████▌  | 73/97 [00:11<00:03,  6.43it/s, loss=3.48, v_num=0]
Epoch 25:  76%|███████▋  | 74/97 [00:11<00:03,  6.48it/s, loss=3.48, v_num=0]
Epoch 25:  76%|███████▋  | 74/97 [00:11<00:03,  6.44it/s, loss=3.48, v_num=0]
Epoch 25:  77%|███████▋  | 75/97 [00:11<00:03,  6.48it/s, loss=3.48, v_num=0]
Epoch 25:  77%|███████▋  | 75/97 [00:11<00:03,  6.45it/s, loss=3.48, v_num=0]
Epoch 25:  78%|███████▊  | 76/97 [00:11<00:03,  6.50it/s, loss=3.48, v_num=0]
Epoch 25:  78%|███████▊  | 76/97 [00:11<00:03,  6.46it/s, loss=3.48, v_num=0]
Epoch 25:  79%|███████▉  | 77/97 [00:11<00:03,  6.50it/s, loss=3.48, v_num=0]
Epoch 25:  79%|███████▉  | 77/97 [00:11<00:03,  6.47it/s, loss=3.48, v_num=0]
Epoch 25:  80%|████████  | 78/97 [00:11<00:02,  6.51it/s, loss=3.48, v_num=0]
Epoch 25:  80%|████████  | 78/97 [00:12<00:02,  6.48it/s, loss=3.47, v_num=0]
Epoch 25:  81%|████████▏ | 79/97 [00:12<00:02,  6.52it/s, loss=3.47, v_num=0]
Epoch 25:  81%|████████▏ | 79/97 [00:12<00:02,  6.49it/s, loss=3.46, v_num=0]
Epoch 25:  82%|████████▏ | 80/97 [00:12<00:02,  6.53it/s, loss=3.46, v_num=0]
Epoch 25:  82%|████████▏ | 80/97 [00:12<00:02,  6.50it/s, loss=3.46, v_num=0]
Epoch 25:  84%|████████▎ | 81/97 [00:12<00:02,  6.54it/s, loss=3.46, v_num=0]
Epoch 25:  84%|████████▎ | 81/97 [00:12<00:02,  6.51it/s, loss=3.45, v_num=0]
Epoch 25:  85%|████████▍ | 82/97 [00:12<00:02,  6.54it/s, loss=3.45, v_num=0]
Epoch 25:  85%|████████▍ | 82/97 [00:12<00:02,  6.51it/s, loss=3.45, v_num=0]
Epoch 25:  86%|████████▌ | 83/97 [00:12<00:02,  6.55it/s, loss=3.45, v_num=0]
Epoch 25:  86%|████████▌ | 83/97 [00:12<00:02,  6.52it/s, loss=3.44, v_num=0]
Epoch 25:  87%|████████▋ | 84/97 [00:12<00:01,  6.56it/s, loss=3.44, v_num=0]
Epoch 25:  87%|████████▋ | 84/97 [00:12<00:01,  6.53it/s, loss=3.44, v_num=0]
Epoch 25:  88%|████████▊ | 85/97 [00:12<00:01,  6.56it/s, loss=3.44, v_num=0]
Epoch 25:  88%|████████▊ | 85/97 [00:13<00:01,  6.54it/s, loss=3.44, v_num=0]
Epoch 25:  89%|████████▊ | 86/97 [00:13<00:01,  6.57it/s, loss=3.44, v_num=0]
Epoch 25:  89%|████████▊ | 86/97 [00:13<00:01,  6.54it/s, loss=3.43, v_num=0]
Epoch 25:  90%|████████▉ | 87/97 [00:13<00:01,  6.58it/s, loss=3.43, v_num=0]
Epoch 25:  90%|████████▉ | 87/97 [00:13<00:01,  6.55it/s, loss=3.45, v_num=0]
Epoch 25:  91%|█████████ | 88/97 [00:13<00:01,  6.59it/s, loss=3.45, v_num=0]
Epoch 25:  91%|█████████ | 88/97 [00:13<00:01,  6.56it/s, loss=3.44, v_num=0]
Epoch 25:  92%|█████████▏| 89/97 [00:13<00:01,  6.60it/s, loss=3.44, v_num=0]
Epoch 25:  92%|█████████▏| 89/97 [00:13<00:01,  6.57it/s, loss=3.44, v_num=0]
Epoch 25:  93%|█████████▎| 90/97 [00:13<00:01,  6.60it/s, loss=3.44, v_num=0]
Epoch 25:  93%|█████████▎| 90/97 [00:13<00:01,  6.58it/s, loss=3.44, v_num=0]
Epoch 25:  94%|█████████▍| 91/97 [00:13<00:00,  6.61it/s, loss=3.44, v_num=0]
Epoch 25:  94%|█████████▍| 91/97 [00:13<00:00,  6.58it/s, loss=3.44, v_num=0]
Epoch 25:  95%|█████████▍| 92/97 [00:13<00:00,  6.62it/s, loss=3.44, v_num=0]
Epoch 25:  95%|█████████▍| 92/97 [00:13<00:00,  6.59it/s, loss=3.44, v_num=0]
Epoch 25:  96%|█████████▌| 93/97 [00:14<00:00,  6.63it/s, loss=3.44, v_num=0]
Epoch 25:  96%|█████████▌| 93/97 [00:14<00:00,  6.60it/s, loss=3.43, v_num=0]
Epoch 25:  97%|█████████▋| 94/97 [00:14<00:00,  6.63it/s, loss=3.43, v_num=0]
Epoch 25:  97%|█████████▋| 94/97 [00:14<00:00,  6.60it/s, loss=3.43, v_num=0]
Epoch 25:  98%|█████████▊| 95/97 [00:14<00:00,  6.64it/s, loss=3.43, v_num=0]
Epoch 25:  98%|█████████▊| 95/97 [00:14<00:00,  6.61it/s, loss=3.43, v_num=0]
Epoch 25:  99%|█████████▉| 96/97 [00:14<00:00,  6.64it/s, loss=3.43, v_num=0]
Epoch 25:  99%|█████████▉| 96/97 [00:14<00:00,  6.62it/s, loss=3.43, v_num=0]
Epoch 25: 100%|██████████| 97/97 [00:14<00:00,  6.65it/s, loss=3.43, v_num=0]
Epoch 25: 100%|██████████| 97/97 [00:14<00:00,  6.62it/s, loss=3.43, v_num=0]
Epoch 25: 100%|██████████| 97/97 [00:14<00:00,  6.62it/s, loss=3.43, v_num=0]
Epoch 25:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.43, v_num=0]
Epoch 26:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.43, v_num=0]
Epoch 26:   1%|          | 1/97 [00:01<02:09,  1.35s/it, loss=3.43, v_num=0]
Epoch 26:   1%|          | 1/97 [00:01<02:15,  1.41s/it, loss=3.44, v_num=0]
Epoch 26:   2%|▏         | 2/97 [00:01<01:11,  1.34it/s, loss=3.44, v_num=0]
Epoch 26:   2%|▏         | 2/97 [00:01<01:13,  1.29it/s, loss=3.44, v_num=0]
Epoch 26:   3%|▎         | 3/97 [00:01<00:51,  1.84it/s, loss=3.44, v_num=0]
Epoch 26:   3%|▎         | 3/97 [00:01<00:52,  1.78it/s, loss=3.44, v_num=0]
Epoch 26:   4%|▍         | 4/97 [00:01<00:41,  2.27it/s, loss=3.44, v_num=0]
Epoch 26:   4%|▍         | 4/97 [00:01<00:42,  2.19it/s, loss=3.46, v_num=0]
Epoch 26:   5%|▌         | 5/97 [00:01<00:35,  2.63it/s, loss=3.46, v_num=0]
Epoch 26:   5%|▌         | 5/97 [00:01<00:36,  2.55it/s, loss=3.46, v_num=0]
Epoch 26:   6%|▌         | 6/97 [00:02<00:30,  2.94it/s, loss=3.46, v_num=0]
Epoch 26:   6%|▌         | 6/97 [00:02<00:31,  2.86it/s, loss=3.46, v_num=0]
Epoch 26:   7%|▋         | 7/97 [00:02<00:28,  3.21it/s, loss=3.46, v_num=0]
Epoch 26:   7%|▋         | 7/97 [00:02<00:28,  3.13it/s, loss=3.46, v_num=0]
Epoch 26:   8%|▊         | 8/97 [00:02<00:25,  3.45it/s, loss=3.46, v_num=0]
Epoch 26:   8%|▊         | 8/97 [00:02<00:26,  3.37it/s, loss=3.46, v_num=0]
Epoch 26:   9%|▉         | 9/97 [00:02<00:24,  3.66it/s, loss=3.46, v_num=0]
Epoch 26:   9%|▉         | 9/97 [00:02<00:24,  3.58it/s, loss=3.46, v_num=0]
Epoch 26:  10%|█         | 10/97 [00:02<00:22,  3.85it/s, loss=3.46, v_num=0]
Epoch 26:  10%|█         | 10/97 [00:02<00:23,  3.77it/s, loss=3.45, v_num=0]
Epoch 26:  11%|█▏        | 11/97 [00:02<00:21,  4.03it/s, loss=3.45, v_num=0]
Epoch 26:  11%|█▏        | 11/97 [00:02<00:21,  3.95it/s, loss=3.46, v_num=0]
Epoch 26:  12%|█▏        | 12/97 [00:02<00:20,  4.18it/s, loss=3.46, v_num=0]
Epoch 26:  12%|█▏        | 12/97 [00:02<00:20,  4.10it/s, loss=3.45, v_num=0]
Epoch 26:  13%|█▎        | 13/97 [00:03<00:19,  4.33it/s, loss=3.45, v_num=0]
Epoch 26:  13%|█▎        | 13/97 [00:03<00:19,  4.25it/s, loss=3.45, v_num=0]
Epoch 26:  14%|█▍        | 14/97 [00:03<00:18,  4.46it/s, loss=3.45, v_num=0]
Epoch 26:  14%|█▍        | 14/97 [00:03<00:18,  4.38it/s, loss=3.45, v_num=0]
Epoch 26:  15%|█▌        | 15/97 [00:03<00:17,  4.58it/s, loss=3.45, v_num=0]
Epoch 26:  15%|█▌        | 15/97 [00:03<00:18,  4.50it/s, loss=3.45, v_num=0]
Epoch 26:  16%|█▋        | 16/97 [00:03<00:17,  4.69it/s, loss=3.45, v_num=0]
Epoch 26:  16%|█▋        | 16/97 [00:03<00:17,  4.61it/s, loss=3.45, v_num=0]
Epoch 26:  18%|█▊        | 17/97 [00:03<00:16,  4.79it/s, loss=3.45, v_num=0]
Epoch 26:  18%|█▊        | 17/97 [00:03<00:16,  4.72it/s, loss=3.44, v_num=0]
Epoch 26:  19%|█▊        | 18/97 [00:03<00:16,  4.88it/s, loss=3.44, v_num=0]
Epoch 26:  19%|█▊        | 18/97 [00:03<00:16,  4.80it/s, loss=3.44, v_num=0]
Epoch 26:  20%|█▉        | 19/97 [00:03<00:15,  4.96it/s, loss=3.44, v_num=0]
Epoch 26:  20%|█▉        | 19/97 [00:03<00:15,  4.89it/s, loss=3.44, v_num=0]
Epoch 26:  21%|██        | 20/97 [00:03<00:15,  5.04it/s, loss=3.44, v_num=0]
Epoch 26:  21%|██        | 20/97 [00:04<00:15,  4.97it/s, loss=3.44, v_num=0]
Epoch 26:  22%|██▏       | 21/97 [00:04<00:14,  5.11it/s, loss=3.44, v_num=0]
Epoch 26:  22%|██▏       | 21/97 [00:04<00:15,  5.04it/s, loss=3.44, v_num=0]
Epoch 26:  23%|██▎       | 22/97 [00:04<00:14,  5.18it/s, loss=3.44, v_num=0]
Epoch 26:  23%|██▎       | 22/97 [00:04<00:14,  5.11it/s, loss=3.44, v_num=0]
Epoch 26:  24%|██▎       | 23/97 [00:04<00:14,  5.24it/s, loss=3.44, v_num=0]
Epoch 26:  24%|██▎       | 23/97 [00:04<00:14,  5.18it/s, loss=3.43, v_num=0]
Epoch 26:  25%|██▍       | 24/97 [00:04<00:13,  5.31it/s, loss=3.43, v_num=0]
Epoch 26:  25%|██▍       | 24/97 [00:04<00:13,  5.25it/s, loss=3.43, v_num=0]
Epoch 26:  26%|██▌       | 25/97 [00:04<00:13,  5.37it/s, loss=3.43, v_num=0]
Epoch 26:  26%|██▌       | 25/97 [00:04<00:13,  5.31it/s, loss=3.43, v_num=0]
Epoch 26:  27%|██▋       | 26/97 [00:04<00:13,  5.42it/s, loss=3.43, v_num=0]
Epoch 26:  27%|██▋       | 26/97 [00:04<00:13,  5.36it/s, loss=3.43, v_num=0]
Epoch 26:  28%|██▊       | 27/97 [00:04<00:12,  5.48it/s, loss=3.43, v_num=0]
Epoch 26:  28%|██▊       | 27/97 [00:04<00:12,  5.42it/s, loss=3.42, v_num=0]
Epoch 26:  29%|██▉       | 28/97 [00:05<00:12,  5.53it/s, loss=3.42, v_num=0]
Epoch 26:  29%|██▉       | 28/97 [00:05<00:12,  5.47it/s, loss=3.42, v_num=0]
Epoch 26:  30%|██▉       | 29/97 [00:05<00:12,  5.57it/s, loss=3.42, v_num=0]
Epoch 26:  30%|██▉       | 29/97 [00:05<00:12,  5.51it/s, loss=3.42, v_num=0]
Epoch 26:  31%|███       | 30/97 [00:05<00:11,  5.62it/s, loss=3.42, v_num=0]
Epoch 26:  31%|███       | 30/97 [00:05<00:12,  5.56it/s, loss=3.42, v_num=0]
Epoch 26:  32%|███▏      | 31/97 [00:05<00:11,  5.66it/s, loss=3.42, v_num=0]
Epoch 26:  32%|███▏      | 31/97 [00:05<00:11,  5.60it/s, loss=3.41, v_num=0]
Epoch 26:  33%|███▎      | 32/97 [00:05<00:11,  5.70it/s, loss=3.41, v_num=0]
Epoch 26:  33%|███▎      | 32/97 [00:05<00:11,  5.64it/s, loss=3.41, v_num=0]
Epoch 26:  34%|███▍      | 33/97 [00:05<00:11,  5.74it/s, loss=3.41, v_num=0]
Epoch 26:  34%|███▍      | 33/97 [00:05<00:11,  5.68it/s, loss=3.41, v_num=0]
Epoch 26:  35%|███▌      | 34/97 [00:05<00:10,  5.77it/s, loss=3.41, v_num=0]
Epoch 26:  35%|███▌      | 34/97 [00:05<00:11,  5.72it/s, loss=3.41, v_num=0]
Epoch 26:  36%|███▌      | 35/97 [00:06<00:10,  5.81it/s, loss=3.41, v_num=0]
Epoch 26:  36%|███▌      | 35/97 [00:06<00:10,  5.76it/s, loss=3.42, v_num=0]
Epoch 26:  37%|███▋      | 36/97 [00:06<00:10,  5.85it/s, loss=3.42, v_num=0]
Epoch 26:  37%|███▋      | 36/97 [00:06<00:10,  5.79it/s, loss=3.42, v_num=0]
Epoch 26:  38%|███▊      | 37/97 [00:06<00:10,  5.87it/s, loss=3.42, v_num=0]
Epoch 26:  38%|███▊      | 37/97 [00:06<00:10,  5.82it/s, loss=3.42, v_num=0]
Epoch 26:  39%|███▉      | 38/97 [00:06<00:09,  5.91it/s, loss=3.42, v_num=0]
Epoch 26:  39%|███▉      | 38/97 [00:06<00:10,  5.86it/s, loss=3.41, v_num=0]
Epoch 26:  40%|████      | 39/97 [00:06<00:09,  5.94it/s, loss=3.41, v_num=0]
Epoch 26:  40%|████      | 39/97 [00:06<00:09,  5.89it/s, loss=3.41, v_num=0]
Epoch 26:  41%|████      | 40/97 [00:06<00:09,  5.97it/s, loss=3.41, v_num=0]
Epoch 26:  41%|████      | 40/97 [00:06<00:09,  5.91it/s, loss=3.41, v_num=0]
Epoch 26:  42%|████▏     | 41/97 [00:06<00:09,  5.99it/s, loss=3.41, v_num=0]
Epoch 26:  42%|████▏     | 41/97 [00:06<00:09,  5.94it/s, loss=3.42, v_num=0]
Epoch 26:  43%|████▎     | 42/97 [00:06<00:09,  6.02it/s, loss=3.42, v_num=0]
Epoch 26:  43%|████▎     | 42/97 [00:07<00:09,  5.97it/s, loss=3.42, v_num=0]
Epoch 26:  44%|████▍     | 43/97 [00:07<00:08,  6.04it/s, loss=3.42, v_num=0]
Epoch 26:  44%|████▍     | 43/97 [00:07<00:09,  5.99it/s, loss=3.42, v_num=0]
Epoch 26:  45%|████▌     | 44/97 [00:07<00:08,  6.07it/s, loss=3.42, v_num=0]
Epoch 26:  45%|████▌     | 44/97 [00:07<00:08,  6.02it/s, loss=3.42, v_num=0]
Epoch 26:  46%|████▋     | 45/97 [00:07<00:08,  6.09it/s, loss=3.42, v_num=0]
Epoch 26:  46%|████▋     | 45/97 [00:07<00:08,  6.04it/s, loss=3.42, v_num=0]
Epoch 26:  47%|████▋     | 46/97 [00:07<00:08,  6.11it/s, loss=3.42, v_num=0]
Epoch 26:  47%|████▋     | 46/97 [00:07<00:08,  6.07it/s, loss=3.41, v_num=0]
Epoch 26:  48%|████▊     | 47/97 [00:07<00:08,  6.13it/s, loss=3.41, v_num=0]
Epoch 26:  48%|████▊     | 47/97 [00:07<00:08,  6.09it/s, loss=3.41, v_num=0]
Epoch 26:  49%|████▉     | 48/97 [00:07<00:07,  6.16it/s, loss=3.41, v_num=0]
Epoch 26:  49%|████▉     | 48/97 [00:07<00:08,  6.11it/s, loss=3.41, v_num=0]
Epoch 26:  51%|█████     | 49/97 [00:07<00:07,  6.17it/s, loss=3.41, v_num=0]
Epoch 26:  51%|█████     | 49/97 [00:07<00:07,  6.13it/s, loss=3.42, v_num=0]
Epoch 26:  52%|█████▏    | 50/97 [00:08<00:07,  6.20it/s, loss=3.42, v_num=0]
Epoch 26:  52%|█████▏    | 50/97 [00:08<00:07,  6.15it/s, loss=3.42, v_num=0]
Epoch 26:  53%|█████▎    | 51/97 [00:08<00:07,  6.21it/s, loss=3.42, v_num=0]
Epoch 26:  53%|█████▎    | 51/97 [00:08<00:07,  6.17it/s, loss=3.43, v_num=0]
Epoch 26:  54%|█████▎    | 52/97 [00:08<00:07,  6.23it/s, loss=3.43, v_num=0]
Epoch 26:  54%|█████▎    | 52/97 [00:08<00:07,  6.19it/s, loss=3.43, v_num=0]
Epoch 26:  55%|█████▍    | 53/97 [00:08<00:07,  6.25it/s, loss=3.43, v_num=0]
Epoch 26:  55%|█████▍    | 53/97 [00:08<00:07,  6.21it/s, loss=3.43, v_num=0]
Epoch 26:  56%|█████▌    | 54/97 [00:08<00:06,  6.26it/s, loss=3.43, v_num=0]
Epoch 26:  56%|█████▌    | 54/97 [00:08<00:06,  6.22it/s, loss=3.43, v_num=0]
Epoch 26:  57%|█████▋    | 55/97 [00:08<00:06,  6.28it/s, loss=3.43, v_num=0]
Epoch 26:  57%|█████▋    | 55/97 [00:08<00:06,  6.24it/s, loss=3.43, v_num=0]
Epoch 26:  58%|█████▊    | 56/97 [00:08<00:06,  6.30it/s, loss=3.43, v_num=0]
Epoch 26:  58%|█████▊    | 56/97 [00:08<00:06,  6.26it/s, loss=3.42, v_num=0]
Epoch 26:  59%|█████▉    | 57/97 [00:09<00:06,  6.31it/s, loss=3.42, v_num=0]
Epoch 26:  59%|█████▉    | 57/97 [00:09<00:06,  6.27it/s, loss=3.41, v_num=0]
Epoch 26:  60%|█████▉    | 58/97 [00:09<00:06,  6.33it/s, loss=3.41, v_num=0]
Epoch 26:  60%|█████▉    | 58/97 [00:09<00:06,  6.29it/s, loss=3.42, v_num=0]
Epoch 26:  61%|██████    | 59/97 [00:09<00:05,  6.34it/s, loss=3.42, v_num=0]
Epoch 26:  61%|██████    | 59/97 [00:09<00:06,  6.30it/s, loss=3.42, v_num=0]
Epoch 26:  62%|██████▏   | 60/97 [00:09<00:05,  6.35it/s, loss=3.42, v_num=0]
Epoch 26:  62%|██████▏   | 60/97 [00:09<00:05,  6.31it/s, loss=3.42, v_num=0]
Epoch 26:  63%|██████▎   | 61/97 [00:09<00:05,  6.37it/s, loss=3.42, v_num=0]
Epoch 26:  63%|██████▎   | 61/97 [00:09<00:05,  6.33it/s, loss=3.41, v_num=0]
Epoch 26:  64%|██████▍   | 62/97 [00:09<00:05,  6.38it/s, loss=3.41, v_num=0]
Epoch 26:  64%|██████▍   | 62/97 [00:09<00:05,  6.34it/s, loss=3.4, v_num=0]
Epoch 26:  65%|██████▍   | 63/97 [00:09<00:05,  6.39it/s, loss=3.4, v_num=0]
Epoch 26:  65%|██████▍   | 63/97 [00:09<00:05,  6.36it/s, loss=3.41, v_num=0]
Epoch 26:  66%|██████▌   | 64/97 [00:09<00:05,  6.40it/s, loss=3.41, v_num=0]
Epoch 26:  66%|██████▌   | 64/97 [00:10<00:05,  6.37it/s, loss=3.41, v_num=0]
Epoch 26:  67%|██████▋   | 65/97 [00:10<00:04,  6.42it/s, loss=3.41, v_num=0]
Epoch 26:  67%|██████▋   | 65/97 [00:10<00:05,  6.38it/s, loss=3.42, v_num=0]
Epoch 26:  68%|██████▊   | 66/97 [00:10<00:04,  6.43it/s, loss=3.42, v_num=0]
Epoch 26:  68%|██████▊   | 66/97 [00:10<00:04,  6.39it/s, loss=3.42, v_num=0]
Epoch 26:  69%|██████▉   | 67/97 [00:10<00:04,  6.44it/s, loss=3.42, v_num=0]
Epoch 26:  69%|██████▉   | 67/97 [00:10<00:04,  6.40it/s, loss=3.41, v_num=0]
Epoch 26:  70%|███████   | 68/97 [00:10<00:04,  6.45it/s, loss=3.41, v_num=0]
Epoch 26:  70%|███████   | 68/97 [00:10<00:04,  6.42it/s, loss=3.42, v_num=0]
Epoch 26:  71%|███████   | 69/97 [00:10<00:04,  6.46it/s, loss=3.42, v_num=0]
Epoch 26:  71%|███████   | 69/97 [00:10<00:04,  6.43it/s, loss=3.41, v_num=0]
Epoch 26:  72%|███████▏  | 70/97 [00:10<00:04,  6.47it/s, loss=3.41, v_num=0]
Epoch 26:  72%|███████▏  | 70/97 [00:10<00:04,  6.44it/s, loss=3.41, v_num=0]
Epoch 26:  73%|███████▎  | 71/97 [00:10<00:04,  6.48it/s, loss=3.41, v_num=0]
Epoch 26:  73%|███████▎  | 71/97 [00:11<00:04,  6.45it/s, loss=3.41, v_num=0]
Epoch 26:  74%|███████▍  | 72/97 [00:11<00:03,  6.49it/s, loss=3.41, v_num=0]
Epoch 26:  74%|███████▍  | 72/97 [00:11<00:03,  6.46it/s, loss=3.41, v_num=0]
Epoch 26:  75%|███████▌  | 73/97 [00:11<00:03,  6.50it/s, loss=3.41, v_num=0]
Epoch 26:  75%|███████▌  | 73/97 [00:11<00:03,  6.47it/s, loss=3.41, v_num=0]
Epoch 26:  76%|███████▋  | 74/97 [00:11<00:03,  6.51it/s, loss=3.41, v_num=0]
Epoch 26:  76%|███████▋  | 74/97 [00:11<00:03,  6.48it/s, loss=3.41, v_num=0]
Epoch 26:  77%|███████▋  | 75/97 [00:11<00:03,  6.52it/s, loss=3.41, v_num=0]
Epoch 26:  77%|███████▋  | 75/97 [00:11<00:03,  6.49it/s, loss=3.41, v_num=0]
Epoch 26:  78%|███████▊  | 76/97 [00:11<00:03,  6.53it/s, loss=3.41, v_num=0]
Epoch 26:  78%|███████▊  | 76/97 [00:11<00:03,  6.50it/s, loss=3.42, v_num=0]
Epoch 26:  79%|███████▉  | 77/97 [00:11<00:03,  6.54it/s, loss=3.42, v_num=0]
Epoch 26:  79%|███████▉  | 77/97 [00:11<00:03,  6.51it/s, loss=3.42, v_num=0]
Epoch 26:  80%|████████  | 78/97 [00:11<00:02,  6.55it/s, loss=3.42, v_num=0]
Epoch 26:  80%|████████  | 78/97 [00:11<00:02,  6.52it/s, loss=3.41, v_num=0]
Epoch 26:  81%|████████▏ | 79/97 [00:12<00:02,  6.56it/s, loss=3.41, v_num=0]
Epoch 26:  81%|████████▏ | 79/97 [00:12<00:02,  6.52it/s, loss=3.41, v_num=0]
Epoch 26:  82%|████████▏ | 80/97 [00:12<00:02,  6.56it/s, loss=3.41, v_num=0]
Epoch 26:  82%|████████▏ | 80/97 [00:12<00:02,  6.53it/s, loss=3.41, v_num=0]
Epoch 26:  84%|████████▎ | 81/97 [00:12<00:02,  6.57it/s, loss=3.41, v_num=0]
Epoch 26:  84%|████████▎ | 81/97 [00:12<00:02,  6.54it/s, loss=3.41, v_num=0]
Epoch 26:  85%|████████▍ | 82/97 [00:12<00:02,  6.58it/s, loss=3.41, v_num=0]
Epoch 26:  85%|████████▍ | 82/97 [00:12<00:02,  6.55it/s, loss=3.41, v_num=0]
Epoch 26:  86%|████████▌ | 83/97 [00:12<00:02,  6.59it/s, loss=3.41, v_num=0]
Epoch 26:  86%|████████▌ | 83/97 [00:12<00:02,  6.56it/s, loss=3.4, v_num=0]
Epoch 26:  87%|████████▋ | 84/97 [00:12<00:01,  6.59it/s, loss=3.4, v_num=0]
Epoch 26:  87%|████████▋ | 84/97 [00:12<00:01,  6.57it/s, loss=3.4, v_num=0]
Epoch 26:  88%|████████▊ | 85/97 [00:12<00:01,  6.60it/s, loss=3.4, v_num=0]
Epoch 26:  88%|████████▊ | 85/97 [00:12<00:01,  6.57it/s, loss=3.39, v_num=0]
Epoch 26:  89%|████████▊ | 86/97 [00:13<00:01,  6.61it/s, loss=3.39, v_num=0]
Epoch 26:  89%|████████▊ | 86/97 [00:13<00:01,  6.58it/s, loss=3.39, v_num=0]
Epoch 26:  90%|████████▉ | 87/97 [00:13<00:01,  6.62it/s, loss=3.39, v_num=0]
Epoch 26:  90%|████████▉ | 87/97 [00:13<00:01,  6.59it/s, loss=3.39, v_num=0]
Epoch 26:  91%|█████████ | 88/97 [00:13<00:01,  6.62it/s, loss=3.39, v_num=0]
Epoch 26:  91%|█████████ | 88/97 [00:13<00:01,  6.60it/s, loss=3.39, v_num=0]
Epoch 26:  92%|█████████▏| 89/97 [00:13<00:01,  6.63it/s, loss=3.39, v_num=0]
Epoch 26:  92%|█████████▏| 89/97 [00:13<00:01,  6.60it/s, loss=3.39, v_num=0]
Epoch 26:  93%|█████████▎| 90/97 [00:13<00:01,  6.64it/s, loss=3.39, v_num=0]
Epoch 26:  93%|█████████▎| 90/97 [00:13<00:01,  6.61it/s, loss=3.39, v_num=0]
Epoch 26:  94%|█████████▍| 91/97 [00:13<00:00,  6.65it/s, loss=3.39, v_num=0]
Epoch 26:  94%|█████████▍| 91/97 [00:13<00:00,  6.62it/s, loss=3.38, v_num=0]
Epoch 26:  95%|█████████▍| 92/97 [00:13<00:00,  6.65it/s, loss=3.38, v_num=0]
Epoch 26:  95%|█████████▍| 92/97 [00:13<00:00,  6.62it/s, loss=3.38, v_num=0]
Epoch 26:  96%|█████████▌| 93/97 [00:13<00:00,  6.66it/s, loss=3.38, v_num=0]
Epoch 26:  96%|█████████▌| 93/97 [00:14<00:00,  6.63it/s, loss=3.38, v_num=0]
Epoch 26:  97%|█████████▋| 94/97 [00:14<00:00,  6.66it/s, loss=3.38, v_num=0]
Epoch 26:  97%|█████████▋| 94/97 [00:14<00:00,  6.64it/s, loss=3.38, v_num=0]
Epoch 26:  98%|█████████▊| 95/97 [00:14<00:00,  6.67it/s, loss=3.38, v_num=0]
Epoch 26:  98%|█████████▊| 95/97 [00:14<00:00,  6.64it/s, loss=3.37, v_num=0]
Epoch 26:  99%|█████████▉| 96/97 [00:14<00:00,  6.68it/s, loss=3.37, v_num=0]
Epoch 26:  99%|█████████▉| 96/97 [00:14<00:00,  6.65it/s, loss=3.36, v_num=0]
Epoch 26: 100%|██████████| 97/97 [00:14<00:00,  6.68it/s, loss=3.36, v_num=0]
Epoch 26: 100%|██████████| 97/97 [00:14<00:00,  6.66it/s, loss=3.36, v_num=0]
Epoch 26: 100%|██████████| 97/97 [00:14<00:00,  6.66it/s, loss=3.36, v_num=0]
Epoch 26:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.36, v_num=0]
Epoch 27:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.36, v_num=0]
Epoch 27:   1%|          | 1/97 [00:01<02:13,  1.39s/it, loss=3.36, v_num=0]
Epoch 27:   1%|          | 1/97 [00:01<02:18,  1.44s/it, loss=3.36, v_num=0]
Epoch 27:   2%|▏         | 2/97 [00:01<01:12,  1.32it/s, loss=3.36, v_num=0]
Epoch 27:   2%|▏         | 2/97 [00:01<01:14,  1.27it/s, loss=3.37, v_num=0]
Epoch 27:   3%|▎         | 3/97 [00:01<00:52,  1.81it/s, loss=3.37, v_num=0]
Epoch 27:   3%|▎         | 3/97 [00:01<00:53,  1.75it/s, loss=3.37, v_num=0]
Epoch 27:   4%|▍         | 4/97 [00:01<00:41,  2.22it/s, loss=3.37, v_num=0]
Epoch 27:   4%|▍         | 4/97 [00:01<00:43,  2.15it/s, loss=3.38, v_num=0]
Epoch 27:   5%|▌         | 5/97 [00:01<00:35,  2.58it/s, loss=3.38, v_num=0]
Epoch 27:   5%|▌         | 5/97 [00:01<00:36,  2.50it/s, loss=3.39, v_num=0]
Epoch 27:   6%|▌         | 6/97 [00:02<00:31,  2.89it/s, loss=3.39, v_num=0]
Epoch 27:   6%|▌         | 6/97 [00:02<00:32,  2.81it/s, loss=3.38, v_num=0]
Epoch 27:   7%|▋         | 7/97 [00:02<00:28,  3.16it/s, loss=3.38, v_num=0]
Epoch 27:   7%|▋         | 7/97 [00:02<00:29,  3.09it/s, loss=3.39, v_num=0]
Epoch 27:   8%|▊         | 8/97 [00:02<00:26,  3.41it/s, loss=3.39, v_num=0]
Epoch 27:   8%|▊         | 8/97 [00:02<00:26,  3.33it/s, loss=3.39, v_num=0]
Epoch 27:   9%|▉         | 9/97 [00:02<00:24,  3.62it/s, loss=3.39, v_num=0]
Epoch 27:   9%|▉         | 9/97 [00:02<00:24,  3.54it/s, loss=3.39, v_num=0]
Epoch 27:  10%|█         | 10/97 [00:02<00:22,  3.81it/s, loss=3.39, v_num=0]
Epoch 27:  10%|█         | 10/97 [00:02<00:23,  3.73it/s, loss=3.4, v_num=0]
Epoch 27:  11%|█▏        | 11/97 [00:02<00:21,  3.98it/s, loss=3.4, v_num=0]
Epoch 27:  11%|█▏        | 11/97 [00:02<00:22,  3.91it/s, loss=3.39, v_num=0]
Epoch 27:  12%|█▏        | 12/97 [00:02<00:20,  4.14it/s, loss=3.39, v_num=0]
Epoch 27:  12%|█▏        | 12/97 [00:02<00:20,  4.06it/s, loss=3.39, v_num=0]
Epoch 27:  13%|█▎        | 13/97 [00:03<00:19,  4.28it/s, loss=3.39, v_num=0]
Epoch 27:  13%|█▎        | 13/97 [00:03<00:19,  4.21it/s, loss=3.4, v_num=0]
Epoch 27:  14%|█▍        | 14/97 [00:03<00:18,  4.41it/s, loss=3.4, v_num=0]
Epoch 27:  14%|█▍        | 14/97 [00:03<00:19,  4.34it/s, loss=3.4, v_num=0]
Epoch 27:  15%|█▌        | 15/97 [00:03<00:18,  4.54it/s, loss=3.4, v_num=0]
Epoch 27:  15%|█▌        | 15/97 [00:03<00:18,  4.46it/s, loss=3.39, v_num=0]
Epoch 27:  16%|█▋        | 16/97 [00:03<00:17,  4.65it/s, loss=3.39, v_num=0]
Epoch 27:  16%|█▋        | 16/97 [00:03<00:17,  4.57it/s, loss=3.4, v_num=0]
Epoch 27:  18%|█▊        | 17/97 [00:03<00:16,  4.75it/s, loss=3.4, v_num=0]
Epoch 27:  18%|█▊        | 17/97 [00:03<00:17,  4.67it/s, loss=3.39, v_num=0]
Epoch 27:  19%|█▊        | 18/97 [00:03<00:16,  4.84it/s, loss=3.39, v_num=0]
Epoch 27:  19%|█▊        | 18/97 [00:03<00:16,  4.76it/s, loss=3.39, v_num=0]
Epoch 27:  20%|█▉        | 19/97 [00:03<00:15,  4.93it/s, loss=3.39, v_num=0]
Epoch 27:  20%|█▉        | 19/97 [00:03<00:16,  4.86it/s, loss=3.4, v_num=0]
Epoch 27:  21%|██        | 20/97 [00:03<00:15,  5.01it/s, loss=3.4, v_num=0]
Epoch 27:  21%|██        | 20/97 [00:04<00:15,  4.94it/s, loss=3.4, v_num=0]
Epoch 27:  22%|██▏       | 21/97 [00:04<00:14,  5.09it/s, loss=3.4, v_num=0]
Epoch 27:  22%|██▏       | 21/97 [00:04<00:15,  5.02it/s, loss=3.4, v_num=0]
Epoch 27:  23%|██▎       | 22/97 [00:04<00:14,  5.16it/s, loss=3.4, v_num=0]
Epoch 27:  23%|██▎       | 22/97 [00:04<00:14,  5.09it/s, loss=3.38, v_num=0]
Epoch 27:  24%|██▎       | 23/97 [00:04<00:14,  5.22it/s, loss=3.38, v_num=0]
Epoch 27:  24%|██▎       | 23/97 [00:04<00:14,  5.16it/s, loss=3.38, v_num=0]
Epoch 27:  25%|██▍       | 24/97 [00:04<00:13,  5.29it/s, loss=3.38, v_num=0]
Epoch 27:  25%|██▍       | 24/97 [00:04<00:13,  5.22it/s, loss=3.38, v_num=0]
Epoch 27:  26%|██▌       | 25/97 [00:04<00:13,  5.34it/s, loss=3.38, v_num=0]
Epoch 27:  26%|██▌       | 25/97 [00:04<00:13,  5.28it/s, loss=3.37, v_num=0]
Epoch 27:  27%|██▋       | 26/97 [00:04<00:13,  5.40it/s, loss=3.37, v_num=0]
Epoch 27:  27%|██▋       | 26/97 [00:04<00:13,  5.34it/s, loss=3.38, v_num=0]
Epoch 27:  28%|██▊       | 27/97 [00:04<00:12,  5.45it/s, loss=3.38, v_num=0]
Epoch 27:  28%|██▊       | 27/97 [00:05<00:12,  5.39it/s, loss=3.37, v_num=0]
Epoch 27:  29%|██▉       | 28/97 [00:05<00:12,  5.50it/s, loss=3.37, v_num=0]
Epoch 27:  29%|██▉       | 28/97 [00:05<00:12,  5.44it/s, loss=3.37, v_num=0]
Epoch 27:  30%|██▉       | 29/97 [00:05<00:12,  5.55it/s, loss=3.37, v_num=0]
Epoch 27:  30%|██▉       | 29/97 [00:05<00:12,  5.49it/s, loss=3.37, v_num=0]
Epoch 27:  31%|███       | 30/97 [00:05<00:11,  5.59it/s, loss=3.37, v_num=0]
Epoch 27:  31%|███       | 30/97 [00:05<00:12,  5.54it/s, loss=3.37, v_num=0]
Epoch 27:  32%|███▏      | 31/97 [00:05<00:11,  5.64it/s, loss=3.37, v_num=0]
Epoch 27:  32%|███▏      | 31/97 [00:05<00:11,  5.58it/s, loss=3.37, v_num=0]
Epoch 27:  33%|███▎      | 32/97 [00:05<00:11,  5.68it/s, loss=3.37, v_num=0]
Epoch 27:  33%|███▎      | 32/97 [00:05<00:11,  5.62it/s, loss=3.37, v_num=0]
Epoch 27:  34%|███▍      | 33/97 [00:05<00:11,  5.71it/s, loss=3.37, v_num=0]
Epoch 27:  34%|███▍      | 33/97 [00:05<00:11,  5.66it/s, loss=3.36, v_num=0]
Epoch 27:  35%|███▌      | 34/97 [00:05<00:10,  5.75it/s, loss=3.36, v_num=0]
Epoch 27:  35%|███▌      | 34/97 [00:05<00:11,  5.70it/s, loss=3.37, v_num=0]
Epoch 27:  36%|███▌      | 35/97 [00:06<00:10,  5.78it/s, loss=3.37, v_num=0]
Epoch 27:  36%|███▌      | 35/97 [00:06<00:10,  5.73it/s, loss=3.37, v_num=0]
Epoch 27:  37%|███▋      | 36/97 [00:06<00:10,  5.82it/s, loss=3.37, v_num=0]
Epoch 27:  37%|███▋      | 36/97 [00:06<00:10,  5.77it/s, loss=3.37, v_num=0]
Epoch 27:  38%|███▊      | 37/97 [00:06<00:10,  5.85it/s, loss=3.37, v_num=0]
Epoch 27:  38%|███▊      | 37/97 [00:06<00:10,  5.80it/s, loss=3.38, v_num=0]
Epoch 27:  39%|███▉      | 38/97 [00:06<00:10,  5.88it/s, loss=3.38, v_num=0]
Epoch 27:  39%|███▉      | 38/97 [00:06<00:10,  5.83it/s, loss=3.37, v_num=0]
Epoch 27:  40%|████      | 39/97 [00:06<00:09,  5.91it/s, loss=3.37, v_num=0]
Epoch 27:  40%|████      | 39/97 [00:06<00:09,  5.86it/s, loss=3.37, v_num=0]
Epoch 27:  41%|████      | 40/97 [00:06<00:09,  5.94it/s, loss=3.37, v_num=0]
Epoch 27:  41%|████      | 40/97 [00:06<00:09,  5.89it/s, loss=3.37, v_num=0]
Epoch 27:  42%|████▏     | 41/97 [00:06<00:09,  5.97it/s, loss=3.37, v_num=0]
Epoch 27:  42%|████▏     | 41/97 [00:06<00:09,  5.92it/s, loss=3.37, v_num=0]
Epoch 27:  43%|████▎     | 42/97 [00:07<00:09,  5.99it/s, loss=3.37, v_num=0]
Epoch 27:  43%|████▎     | 42/97 [00:07<00:09,  5.95it/s, loss=3.38, v_num=0]
Epoch 27:  44%|████▍     | 43/97 [00:07<00:08,  6.02it/s, loss=3.38, v_num=0]
Epoch 27:  44%|████▍     | 43/97 [00:07<00:09,  5.97it/s, loss=3.39, v_num=0]
Epoch 27:  45%|████▌     | 44/97 [00:07<00:08,  6.04it/s, loss=3.39, v_num=0]
Epoch 27:  45%|████▌     | 44/97 [00:07<00:08,  5.99it/s, loss=3.38, v_num=0]
Epoch 27:  46%|████▋     | 45/97 [00:07<00:08,  6.06it/s, loss=3.38, v_num=0]
Epoch 27:  46%|████▋     | 45/97 [00:07<00:08,  6.01it/s, loss=3.39, v_num=0]
Epoch 27:  47%|████▋     | 46/97 [00:07<00:08,  6.08it/s, loss=3.39, v_num=0]
Epoch 27:  47%|████▋     | 46/97 [00:07<00:08,  6.04it/s, loss=3.39, v_num=0]
Epoch 27:  48%|████▊     | 47/97 [00:07<00:08,  6.10it/s, loss=3.39, v_num=0]
Epoch 27:  48%|████▊     | 47/97 [00:07<00:08,  6.06it/s, loss=3.39, v_num=0]
Epoch 27:  49%|████▉     | 48/97 [00:07<00:07,  6.13it/s, loss=3.39, v_num=0]
Epoch 27:  49%|████▉     | 48/97 [00:07<00:08,  6.08it/s, loss=3.39, v_num=0]
Epoch 27:  51%|█████     | 49/97 [00:07<00:07,  6.14it/s, loss=3.39, v_num=0]
Epoch 27:  51%|█████     | 49/97 [00:08<00:07,  6.10it/s, loss=3.38, v_num=0]
Epoch 27:  52%|█████▏    | 50/97 [00:08<00:07,  6.17it/s, loss=3.38, v_num=0]
Epoch 27:  52%|█████▏    | 50/97 [00:08<00:07,  6.12it/s, loss=3.37, v_num=0]
Epoch 27:  53%|█████▎    | 51/97 [00:08<00:07,  6.18it/s, loss=3.37, v_num=0]
Epoch 27:  53%|█████▎    | 51/97 [00:08<00:07,  6.14it/s, loss=3.37, v_num=0]
Epoch 27:  54%|█████▎    | 52/97 [00:08<00:07,  6.20it/s, loss=3.37, v_num=0]
Epoch 27:  54%|█████▎    | 52/97 [00:08<00:07,  6.16it/s, loss=3.38, v_num=0]
Epoch 27:  55%|█████▍    | 53/97 [00:08<00:07,  6.22it/s, loss=3.38, v_num=0]
Epoch 27:  55%|█████▍    | 53/97 [00:08<00:07,  6.18it/s, loss=3.38, v_num=0]
Epoch 27:  56%|█████▌    | 54/97 [00:08<00:06,  6.23it/s, loss=3.38, v_num=0]
Epoch 27:  56%|█████▌    | 54/97 [00:08<00:06,  6.19it/s, loss=3.38, v_num=0]
Epoch 27:  57%|█████▋    | 55/97 [00:08<00:06,  6.25it/s, loss=3.38, v_num=0]
Epoch 27:  57%|█████▋    | 55/97 [00:08<00:06,  6.21it/s, loss=3.37, v_num=0]
Epoch 27:  58%|█████▊    | 56/97 [00:08<00:06,  6.26it/s, loss=3.37, v_num=0]
Epoch 27:  58%|█████▊    | 56/97 [00:08<00:06,  6.22it/s, loss=3.37, v_num=0]
Epoch 27:  59%|█████▉    | 57/97 [00:09<00:06,  6.28it/s, loss=3.37, v_num=0]
Epoch 27:  59%|█████▉    | 57/97 [00:09<00:06,  6.24it/s, loss=3.36, v_num=0]
Epoch 27:  60%|█████▉    | 58/97 [00:09<00:06,  6.30it/s, loss=3.36, v_num=0]
Epoch 27:  60%|█████▉    | 58/97 [00:09<00:06,  6.25it/s, loss=3.36, v_num=0]
Epoch 27:  61%|██████    | 59/97 [00:09<00:06,  6.31it/s, loss=3.36, v_num=0]
Epoch 27:  61%|██████    | 59/97 [00:09<00:06,  6.27it/s, loss=3.36, v_num=0]
Epoch 27:  62%|██████▏   | 60/97 [00:09<00:05,  6.32it/s, loss=3.36, v_num=0]
Epoch 27:  62%|██████▏   | 60/97 [00:09<00:05,  6.28it/s, loss=3.36, v_num=0]
Epoch 27:  63%|██████▎   | 61/97 [00:09<00:05,  6.33it/s, loss=3.36, v_num=0]
Epoch 27:  63%|██████▎   | 61/97 [00:09<00:05,  6.30it/s, loss=3.35, v_num=0]
Epoch 27:  64%|██████▍   | 62/97 [00:09<00:05,  6.35it/s, loss=3.35, v_num=0]
Epoch 27:  64%|██████▍   | 62/97 [00:09<00:05,  6.31it/s, loss=3.34, v_num=0]
Epoch 27:  65%|██████▍   | 63/97 [00:09<00:05,  6.36it/s, loss=3.34, v_num=0]
Epoch 27:  65%|██████▍   | 63/97 [00:09<00:05,  6.32it/s, loss=3.34, v_num=0]
Epoch 27:  66%|██████▌   | 64/97 [00:10<00:05,  6.37it/s, loss=3.34, v_num=0]
Epoch 27:  66%|██████▌   | 64/97 [00:10<00:05,  6.34it/s, loss=3.34, v_num=0]
Epoch 27:  67%|██████▋   | 65/97 [00:10<00:05,  6.38it/s, loss=3.34, v_num=0]
Epoch 27:  67%|██████▋   | 65/97 [00:10<00:05,  6.35it/s, loss=3.33, v_num=0]
Epoch 27:  68%|██████▊   | 66/97 [00:10<00:04,  6.40it/s, loss=3.33, v_num=0]
Epoch 27:  68%|██████▊   | 66/97 [00:10<00:04,  6.36it/s, loss=3.33, v_num=0]
Epoch 27:  69%|██████▉   | 67/97 [00:10<00:04,  6.41it/s, loss=3.33, v_num=0]
Epoch 27:  69%|██████▉   | 67/97 [00:10<00:04,  6.37it/s, loss=3.33, v_num=0]
Epoch 27:  70%|███████   | 68/97 [00:10<00:04,  6.42it/s, loss=3.33, v_num=0]
Epoch 27:  70%|███████   | 68/97 [00:10<00:04,  6.39it/s, loss=3.32, v_num=0]
Epoch 27:  71%|███████   | 69/97 [00:10<00:04,  6.43it/s, loss=3.32, v_num=0]
Epoch 27:  71%|███████   | 69/97 [00:10<00:04,  6.40it/s, loss=3.33, v_num=0]
Epoch 27:  72%|███████▏  | 70/97 [00:10<00:04,  6.44it/s, loss=3.33, v_num=0]
Epoch 27:  72%|███████▏  | 70/97 [00:10<00:04,  6.41it/s, loss=3.33, v_num=0]
Epoch 27:  73%|███████▎  | 71/97 [00:11<00:04,  6.45it/s, loss=3.33, v_num=0]
Epoch 27:  73%|███████▎  | 71/97 [00:11<00:04,  6.42it/s, loss=3.33, v_num=0]
Epoch 27:  74%|███████▍  | 72/97 [00:11<00:03,  6.46it/s, loss=3.33, v_num=0]
Epoch 27:  74%|███████▍  | 72/97 [00:11<00:03,  6.43it/s, loss=3.32, v_num=0]
Epoch 27:  75%|███████▌  | 73/97 [00:11<00:03,  6.47it/s, loss=3.32, v_num=0]
Epoch 27:  75%|███████▌  | 73/97 [00:11<00:03,  6.44it/s, loss=3.32, v_num=0]
Epoch 27:  76%|███████▋  | 74/97 [00:11<00:03,  6.48it/s, loss=3.32, v_num=0]
Epoch 27:  76%|███████▋  | 74/97 [00:11<00:03,  6.45it/s, loss=3.32, v_num=0]
Epoch 27:  77%|███████▋  | 75/97 [00:11<00:03,  6.49it/s, loss=3.32, v_num=0]
Epoch 27:  77%|███████▋  | 75/97 [00:11<00:03,  6.46it/s, loss=3.32, v_num=0]
Epoch 27:  78%|███████▊  | 76/97 [00:11<00:03,  6.50it/s, loss=3.32, v_num=0]
Epoch 27:  78%|███████▊  | 76/97 [00:11<00:03,  6.47it/s, loss=3.32, v_num=0]
Epoch 27:  79%|███████▉  | 77/97 [00:11<00:03,  6.51it/s, loss=3.32, v_num=0]
Epoch 27:  79%|███████▉  | 77/97 [00:11<00:03,  6.48it/s, loss=3.32, v_num=0]
Epoch 27:  80%|████████  | 78/97 [00:11<00:02,  6.52it/s, loss=3.32, v_num=0]
Epoch 27:  80%|████████  | 78/97 [00:12<00:02,  6.49it/s, loss=3.31, v_num=0]
Epoch 27:  81%|████████▏ | 79/97 [00:12<00:02,  6.53it/s, loss=3.31, v_num=0]
Epoch 27:  81%|████████▏ | 79/97 [00:12<00:02,  6.50it/s, loss=3.31, v_num=0]
Epoch 27:  82%|████████▏ | 80/97 [00:12<00:02,  6.53it/s, loss=3.31, v_num=0]
Epoch 27:  82%|████████▏ | 80/97 [00:12<00:02,  6.50it/s, loss=3.32, v_num=0]
Epoch 27:  84%|████████▎ | 81/97 [00:12<00:02,  6.54it/s, loss=3.32, v_num=0]
Epoch 27:  84%|████████▎ | 81/97 [00:12<00:02,  6.51it/s, loss=3.32, v_num=0]
Epoch 27:  85%|████████▍ | 82/97 [00:12<00:02,  6.55it/s, loss=3.32, v_num=0]
Epoch 27:  85%|████████▍ | 82/97 [00:12<00:02,  6.52it/s, loss=3.33, v_num=0]
Epoch 27:  86%|████████▌ | 83/97 [00:12<00:02,  6.56it/s, loss=3.33, v_num=0]
Epoch 27:  86%|████████▌ | 83/97 [00:12<00:02,  6.53it/s, loss=3.32, v_num=0]
Epoch 27:  87%|████████▋ | 84/97 [00:12<00:01,  6.56it/s, loss=3.32, v_num=0]
Epoch 27:  87%|████████▋ | 84/97 [00:12<00:01,  6.54it/s, loss=3.33, v_num=0]
Epoch 27:  88%|████████▊ | 85/97 [00:12<00:01,  6.57it/s, loss=3.33, v_num=0]
Epoch 27:  88%|████████▊ | 85/97 [00:12<00:01,  6.54it/s, loss=3.34, v_num=0]
Epoch 27:  89%|████████▊ | 86/97 [00:13<00:01,  6.58it/s, loss=3.34, v_num=0]
Epoch 27:  89%|████████▊ | 86/97 [00:13<00:01,  6.55it/s, loss=3.34, v_num=0]
Epoch 27:  90%|████████▉ | 87/97 [00:13<00:01,  6.59it/s, loss=3.34, v_num=0]
Epoch 27:  90%|████████▉ | 87/97 [00:13<00:01,  6.56it/s, loss=3.34, v_num=0]
Epoch 27:  91%|█████████ | 88/97 [00:13<00:01,  6.59it/s, loss=3.34, v_num=0]
Epoch 27:  91%|█████████ | 88/97 [00:13<00:01,  6.57it/s, loss=3.34, v_num=0]
Epoch 27:  92%|█████████▏| 89/97 [00:13<00:01,  6.60it/s, loss=3.34, v_num=0]
Epoch 27:  92%|█████████▏| 89/97 [00:13<00:01,  6.58it/s, loss=3.34, v_num=0]
Epoch 27:  93%|█████████▎| 90/97 [00:13<00:01,  6.61it/s, loss=3.34, v_num=0]
Epoch 27:  93%|█████████▎| 90/97 [00:13<00:01,  6.58it/s, loss=3.34, v_num=0]
Epoch 27:  94%|█████████▍| 91/97 [00:13<00:00,  6.62it/s, loss=3.34, v_num=0]
Epoch 27:  94%|█████████▍| 91/97 [00:13<00:00,  6.59it/s, loss=3.34, v_num=0]
Epoch 27:  95%|█████████▍| 92/97 [00:13<00:00,  6.62it/s, loss=3.34, v_num=0]
Epoch 27:  95%|█████████▍| 92/97 [00:13<00:00,  6.60it/s, loss=3.35, v_num=0]
Epoch 27:  96%|█████████▌| 93/97 [00:14<00:00,  6.63it/s, loss=3.35, v_num=0]
Epoch 27:  96%|█████████▌| 93/97 [00:14<00:00,  6.60it/s, loss=3.34, v_num=0]
Epoch 27:  97%|█████████▋| 94/97 [00:14<00:00,  6.64it/s, loss=3.34, v_num=0]
Epoch 27:  97%|█████████▋| 94/97 [00:14<00:00,  6.61it/s, loss=3.35, v_num=0]
Epoch 27:  98%|█████████▊| 95/97 [00:14<00:00,  6.65it/s, loss=3.35, v_num=0]
Epoch 27:  98%|█████████▊| 95/97 [00:14<00:00,  6.62it/s, loss=3.34, v_num=0]
Epoch 27:  99%|█████████▉| 96/97 [00:14<00:00,  6.65it/s, loss=3.34, v_num=0]
Epoch 27:  99%|█████████▉| 96/97 [00:14<00:00,  6.62it/s, loss=3.33, v_num=0]
Epoch 27: 100%|██████████| 97/97 [00:14<00:00,  6.66it/s, loss=3.33, v_num=0]
Epoch 27: 100%|██████████| 97/97 [00:14<00:00,  6.63it/s, loss=3.34, v_num=0]
Epoch 27: 100%|██████████| 97/97 [00:14<00:00,  6.63it/s, loss=3.34, v_num=0]
Epoch 27:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.34, v_num=0]
Epoch 28:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.34, v_num=0]
Epoch 28:   1%|          | 1/97 [00:01<02:12,  1.38s/it, loss=3.34, v_num=0]
Epoch 28:   1%|          | 1/97 [00:01<02:17,  1.44s/it, loss=3.35, v_num=0]
Epoch 28:   2%|▏         | 2/97 [00:01<01:12,  1.32it/s, loss=3.35, v_num=0]
Epoch 28:   2%|▏         | 2/97 [00:01<01:14,  1.27it/s, loss=3.35, v_num=0]
Epoch 28:   3%|▎         | 3/97 [00:01<00:51,  1.81it/s, loss=3.35, v_num=0]
Epoch 28:   3%|▎         | 3/97 [00:01<00:53,  1.75it/s, loss=3.35, v_num=0]
Epoch 28:   4%|▍         | 4/97 [00:01<00:41,  2.23it/s, loss=3.35, v_num=0]
Epoch 28:   4%|▍         | 4/97 [00:01<00:42,  2.16it/s, loss=3.34, v_num=0]
Epoch 28:   5%|▌         | 5/97 [00:01<00:35,  2.59it/s, loss=3.34, v_num=0]
Epoch 28:   5%|▌         | 5/97 [00:01<00:36,  2.51it/s, loss=3.33, v_num=0]
Epoch 28:   6%|▌         | 6/97 [00:02<00:31,  2.90it/s, loss=3.33, v_num=0]
Epoch 28:   6%|▌         | 6/97 [00:02<00:32,  2.82it/s, loss=3.34, v_num=0]
Epoch 28:   7%|▋         | 7/97 [00:02<00:28,  3.18it/s, loss=3.34, v_num=0]
Epoch 28:   7%|▋         | 7/97 [00:02<00:29,  3.10it/s, loss=3.33, v_num=0]
Epoch 28:   8%|▊         | 8/97 [00:02<00:26,  3.42it/s, loss=3.33, v_num=0]
Epoch 28:   8%|▊         | 8/97 [00:02<00:26,  3.34it/s, loss=3.33, v_num=0]
Epoch 28:   9%|▉         | 9/97 [00:02<00:24,  3.62it/s, loss=3.33, v_num=0]
Epoch 28:   9%|▉         | 9/97 [00:02<00:24,  3.54it/s, loss=3.32, v_num=0]
Epoch 28:  10%|█         | 10/97 [00:02<00:22,  3.82it/s, loss=3.32, v_num=0]
Epoch 28:  10%|█         | 10/97 [00:02<00:23,  3.73it/s, loss=3.31, v_num=0]
Epoch 28:  11%|█▏        | 11/97 [00:02<00:21,  3.99it/s, loss=3.31, v_num=0]
Epoch 28:  11%|█▏        | 11/97 [00:02<00:22,  3.91it/s, loss=3.32, v_num=0]
Epoch 28:  12%|█▏        | 12/97 [00:02<00:20,  4.15it/s, loss=3.32, v_num=0]
Epoch 28:  12%|█▏        | 12/97 [00:02<00:20,  4.07it/s, loss=3.31, v_num=0]
Epoch 28:  13%|█▎        | 13/97 [00:03<00:19,  4.29it/s, loss=3.31, v_num=0]
Epoch 28:  13%|█▎        | 13/97 [00:03<00:19,  4.21it/s, loss=3.31, v_num=0]
Epoch 28:  14%|█▍        | 14/97 [00:03<00:18,  4.42it/s, loss=3.31, v_num=0]
Epoch 28:  14%|█▍        | 14/97 [00:03<00:19,  4.34it/s, loss=3.31, v_num=0]
Epoch 28:  15%|█▌        | 15/97 [00:03<00:18,  4.54it/s, loss=3.31, v_num=0]
Epoch 28:  15%|█▌        | 15/97 [00:03<00:18,  4.47it/s, loss=3.3, v_num=0]
Epoch 28:  16%|█▋        | 16/97 [00:03<00:17,  4.65it/s, loss=3.3, v_num=0]
Epoch 28:  16%|█▋        | 16/97 [00:03<00:17,  4.58it/s, loss=3.31, v_num=0]
Epoch 28:  18%|█▊        | 17/97 [00:03<00:16,  4.75it/s, loss=3.31, v_num=0]
Epoch 28:  18%|█▊        | 17/97 [00:03<00:17,  4.68it/s, loss=3.3, v_num=0]
Epoch 28:  19%|█▊        | 18/97 [00:03<00:16,  4.85it/s, loss=3.3, v_num=0]
Epoch 28:  19%|█▊        | 18/97 [00:03<00:16,  4.77it/s, loss=3.31, v_num=0]
Epoch 28:  20%|█▉        | 19/97 [00:03<00:15,  4.93it/s, loss=3.31, v_num=0]
Epoch 28:  20%|█▉        | 19/97 [00:03<00:16,  4.86it/s, loss=3.32, v_num=0]
Epoch 28:  21%|██        | 20/97 [00:03<00:15,  5.02it/s, loss=3.32, v_num=0]
Epoch 28:  21%|██        | 20/97 [00:04<00:15,  4.95it/s, loss=3.31, v_num=0]
Epoch 28:  22%|██▏       | 21/97 [00:04<00:14,  5.09it/s, loss=3.31, v_num=0]
Epoch 28:  22%|██▏       | 21/97 [00:04<00:15,  5.02it/s, loss=3.3, v_num=0]
Epoch 28:  23%|██▎       | 22/97 [00:04<00:14,  5.17it/s, loss=3.3, v_num=0]
Epoch 28:  23%|██▎       | 22/97 [00:04<00:14,  5.10it/s, loss=3.3, v_num=0]
Epoch 28:  24%|██▎       | 23/97 [00:04<00:14,  5.23it/s, loss=3.3, v_num=0]
Epoch 28:  24%|██▎       | 23/97 [00:04<00:14,  5.16it/s, loss=3.29, v_num=0]
Epoch 28:  25%|██▍       | 24/97 [00:04<00:13,  5.29it/s, loss=3.29, v_num=0]
Epoch 28:  25%|██▍       | 24/97 [00:04<00:13,  5.22it/s, loss=3.29, v_num=0]
Epoch 28:  26%|██▌       | 25/97 [00:04<00:13,  5.35it/s, loss=3.29, v_num=0]
Epoch 28:  26%|██▌       | 25/97 [00:04<00:13,  5.28it/s, loss=3.3, v_num=0]
Epoch 28:  27%|██▋       | 26/97 [00:04<00:13,  5.41it/s, loss=3.3, v_num=0]
Epoch 28:  27%|██▋       | 26/97 [00:04<00:13,  5.34it/s, loss=3.29, v_num=0]
Epoch 28:  28%|██▊       | 27/97 [00:04<00:12,  5.45it/s, loss=3.29, v_num=0]
Epoch 28:  28%|██▊       | 27/97 [00:05<00:12,  5.39it/s, loss=3.29, v_num=0]
Epoch 28:  29%|██▉       | 28/97 [00:05<00:12,  5.50it/s, loss=3.29, v_num=0]
Epoch 28:  29%|██▉       | 28/97 [00:05<00:12,  5.44it/s, loss=3.29, v_num=0]
Epoch 28:  30%|██▉       | 29/97 [00:05<00:12,  5.55it/s, loss=3.29, v_num=0]
Epoch 28:  30%|██▉       | 29/97 [00:05<00:12,  5.49it/s, loss=3.3, v_num=0]
Epoch 28:  31%|███       | 30/97 [00:05<00:11,  5.59it/s, loss=3.3, v_num=0]
Epoch 28:  31%|███       | 30/97 [00:05<00:12,  5.53it/s, loss=3.29, v_num=0]
Epoch 28:  32%|███▏      | 31/97 [00:05<00:11,  5.63it/s, loss=3.29, v_num=0]
Epoch 28:  32%|███▏      | 31/97 [00:05<00:11,  5.58it/s, loss=3.29, v_num=0]
Epoch 28:  33%|███▎      | 32/97 [00:05<00:11,  5.67it/s, loss=3.29, v_num=0]
Epoch 28:  33%|███▎      | 32/97 [00:05<00:11,  5.62it/s, loss=3.28, v_num=0]
Epoch 28:  34%|███▍      | 33/97 [00:05<00:11,  5.72it/s, loss=3.28, v_num=0]
Epoch 28:  34%|███▍      | 33/97 [00:05<00:11,  5.66it/s, loss=3.3, v_num=0]
Epoch 28:  35%|███▌      | 34/97 [00:05<00:10,  5.75it/s, loss=3.3, v_num=0]
Epoch 28:  35%|███▌      | 34/97 [00:05<00:11,  5.70it/s, loss=3.29, v_num=0]
Epoch 28:  36%|███▌      | 35/97 [00:06<00:10,  5.79it/s, loss=3.29, v_num=0]
Epoch 28:  36%|███▌      | 35/97 [00:06<00:10,  5.73it/s, loss=3.3, v_num=0]
Epoch 28:  37%|███▋      | 36/97 [00:06<00:10,  5.81it/s, loss=3.3, v_num=0]
Epoch 28:  37%|███▋      | 36/97 [00:06<00:10,  5.76it/s, loss=3.3, v_num=0]
Epoch 28:  38%|███▊      | 37/97 [00:06<00:10,  5.84it/s, loss=3.3, v_num=0]
Epoch 28:  38%|███▊      | 37/97 [00:06<00:10,  5.79it/s, loss=3.29, v_num=0]
Epoch 28:  39%|███▉      | 38/97 [00:06<00:10,  5.88it/s, loss=3.29, v_num=0]
Epoch 28:  39%|███▉      | 38/97 [00:06<00:10,  5.83it/s, loss=3.29, v_num=0]
Epoch 28:  40%|████      | 39/97 [00:06<00:09,  5.91it/s, loss=3.29, v_num=0]
Epoch 28:  40%|████      | 39/97 [00:06<00:09,  5.86it/s, loss=3.3, v_num=0]
Epoch 28:  41%|████      | 40/97 [00:06<00:09,  5.93it/s, loss=3.3, v_num=0]
Epoch 28:  41%|████      | 40/97 [00:06<00:09,  5.89it/s, loss=3.31, v_num=0]
Epoch 28:  42%|████▏     | 41/97 [00:06<00:09,  5.97it/s, loss=3.31, v_num=0]
Epoch 28:  42%|████▏     | 41/97 [00:06<00:09,  5.91it/s, loss=3.31, v_num=0]
Epoch 28:  43%|████▎     | 42/97 [00:07<00:09,  5.99it/s, loss=3.31, v_num=0]
Epoch 28:  43%|████▎     | 42/97 [00:07<00:09,  5.94it/s, loss=3.31, v_num=0]
Epoch 28:  44%|████▍     | 43/97 [00:07<00:08,  6.02it/s, loss=3.31, v_num=0]
Epoch 28:  44%|████▍     | 43/97 [00:07<00:09,  5.97it/s, loss=3.32, v_num=0]
Epoch 28:  45%|████▌     | 44/97 [00:07<00:08,  6.04it/s, loss=3.32, v_num=0]
Epoch 28:  45%|████▌     | 44/97 [00:07<00:08,  5.99it/s, loss=3.32, v_num=0]
Epoch 28:  46%|████▋     | 45/97 [00:07<00:08,  6.06it/s, loss=3.32, v_num=0]
Epoch 28:  46%|████▋     | 45/97 [00:07<00:08,  6.02it/s, loss=3.31, v_num=0]
Epoch 28:  47%|████▋     | 46/97 [00:07<00:08,  6.08it/s, loss=3.31, v_num=0]
Epoch 28:  47%|████▋     | 46/97 [00:07<00:08,  6.04it/s, loss=3.32, v_num=0]
Epoch 28:  48%|████▊     | 47/97 [00:07<00:08,  6.11it/s, loss=3.32, v_num=0]
Epoch 28:  48%|████▊     | 47/97 [00:07<00:08,  6.06it/s, loss=3.32, v_num=0]
Epoch 28:  49%|████▉     | 48/97 [00:07<00:07,  6.13it/s, loss=3.32, v_num=0]
Epoch 28:  49%|████▉     | 48/97 [00:07<00:08,  6.08it/s, loss=3.33, v_num=0]
Epoch 28:  51%|█████     | 49/97 [00:07<00:07,  6.15it/s, loss=3.33, v_num=0]
Epoch 28:  51%|█████     | 49/97 [00:08<00:07,  6.10it/s, loss=3.32, v_num=0]
Epoch 28:  52%|█████▏    | 50/97 [00:08<00:07,  6.16it/s, loss=3.32, v_num=0]
Epoch 28:  52%|█████▏    | 50/97 [00:08<00:07,  6.12it/s, loss=3.32, v_num=0]
Epoch 28:  53%|█████▎    | 51/97 [00:08<00:07,  6.18it/s, loss=3.32, v_num=0]
Epoch 28:  53%|█████▎    | 51/97 [00:08<00:07,  6.14it/s, loss=3.33, v_num=0]
Epoch 28:  54%|█████▎    | 52/97 [00:08<00:07,  6.20it/s, loss=3.33, v_num=0]
Epoch 28:  54%|█████▎    | 52/97 [00:08<00:07,  6.16it/s, loss=3.33, v_num=0]
Epoch 28:  55%|█████▍    | 53/97 [00:08<00:07,  6.22it/s, loss=3.33, v_num=0]
Epoch 28:  55%|█████▍    | 53/97 [00:08<00:07,  6.18it/s, loss=3.32, v_num=0]
Epoch 28:  56%|█████▌    | 54/97 [00:08<00:06,  6.24it/s, loss=3.32, v_num=0]
Epoch 28:  56%|█████▌    | 54/97 [00:08<00:06,  6.20it/s, loss=3.31, v_num=0]
Epoch 28:  57%|█████▋    | 55/97 [00:08<00:06,  6.25it/s, loss=3.31, v_num=0]
Epoch 28:  57%|█████▋    | 55/97 [00:08<00:06,  6.21it/s, loss=3.31, v_num=0]
Epoch 28:  58%|█████▊    | 56/97 [00:08<00:06,  6.26it/s, loss=3.31, v_num=0]
Epoch 28:  58%|█████▊    | 56/97 [00:08<00:06,  6.22it/s, loss=3.3, v_num=0]
Epoch 28:  59%|█████▉    | 57/97 [00:09<00:06,  6.28it/s, loss=3.3, v_num=0]
Epoch 28:  59%|█████▉    | 57/97 [00:09<00:06,  6.24it/s, loss=3.31, v_num=0]
Epoch 28:  60%|█████▉    | 58/97 [00:09<00:06,  6.29it/s, loss=3.31, v_num=0]
Epoch 28:  60%|█████▉    | 58/97 [00:09<00:06,  6.25it/s, loss=3.3, v_num=0]
Epoch 28:  61%|██████    | 59/97 [00:09<00:06,  6.31it/s, loss=3.3, v_num=0]
Epoch 28:  61%|██████    | 59/97 [00:09<00:06,  6.27it/s, loss=3.29, v_num=0]
Epoch 28:  62%|██████▏   | 60/97 [00:09<00:05,  6.32it/s, loss=3.29, v_num=0]
Epoch 28:  62%|██████▏   | 60/97 [00:09<00:05,  6.28it/s, loss=3.29, v_num=0]
Epoch 28:  63%|██████▎   | 61/97 [00:09<00:05,  6.33it/s, loss=3.29, v_num=0]
Epoch 28:  63%|██████▎   | 61/97 [00:09<00:05,  6.29it/s, loss=3.29, v_num=0]
Epoch 28:  64%|██████▍   | 62/97 [00:09<00:05,  6.34it/s, loss=3.29, v_num=0]
Epoch 28:  64%|██████▍   | 62/97 [00:09<00:05,  6.31it/s, loss=3.29, v_num=0]
Epoch 28:  65%|██████▍   | 63/97 [00:09<00:05,  6.36it/s, loss=3.29, v_num=0]
Epoch 28:  65%|██████▍   | 63/97 [00:09<00:05,  6.32it/s, loss=3.29, v_num=0]
Epoch 28:  66%|██████▌   | 64/97 [00:10<00:05,  6.37it/s, loss=3.29, v_num=0]
Epoch 28:  66%|██████▌   | 64/97 [00:10<00:05,  6.34it/s, loss=3.29, v_num=0]
Epoch 28:  67%|██████▋   | 65/97 [00:10<00:05,  6.38it/s, loss=3.29, v_num=0]
Epoch 28:  67%|██████▋   | 65/97 [00:10<00:05,  6.35it/s, loss=3.29, v_num=0]
Epoch 28:  68%|██████▊   | 66/97 [00:10<00:04,  6.40it/s, loss=3.29, v_num=0]
Epoch 28:  68%|██████▊   | 66/97 [00:10<00:04,  6.36it/s, loss=3.29, v_num=0]
Epoch 28:  69%|██████▉   | 67/97 [00:10<00:04,  6.41it/s, loss=3.29, v_num=0]
Epoch 28:  69%|██████▉   | 67/97 [00:10<00:04,  6.37it/s, loss=3.29, v_num=0]
Epoch 28:  70%|███████   | 68/97 [00:10<00:04,  6.42it/s, loss=3.29, v_num=0]
Epoch 28:  70%|███████   | 68/97 [00:10<00:04,  6.39it/s, loss=3.28, v_num=0]
Epoch 28:  71%|███████   | 69/97 [00:10<00:04,  6.43it/s, loss=3.28, v_num=0]
Epoch 28:  71%|███████   | 69/97 [00:10<00:04,  6.40it/s, loss=3.28, v_num=0]
Epoch 28:  72%|███████▏  | 70/97 [00:10<00:04,  6.44it/s, loss=3.28, v_num=0]
Epoch 28:  72%|███████▏  | 70/97 [00:10<00:04,  6.41it/s, loss=3.28, v_num=0]
Epoch 28:  73%|███████▎  | 71/97 [00:11<00:04,  6.45it/s, loss=3.28, v_num=0]
Epoch 28:  73%|███████▎  | 71/97 [00:11<00:04,  6.42it/s, loss=3.28, v_num=0]
Epoch 28:  74%|███████▍  | 72/97 [00:11<00:03,  6.46it/s, loss=3.28, v_num=0]
Epoch 28:  74%|███████▍  | 72/97 [00:11<00:03,  6.43it/s, loss=3.28, v_num=0]
Epoch 28:  75%|███████▌  | 73/97 [00:11<00:03,  6.47it/s, loss=3.28, v_num=0]
Epoch 28:  75%|███████▌  | 73/97 [00:11<00:03,  6.44it/s, loss=3.28, v_num=0]
Epoch 28:  76%|███████▋  | 74/97 [00:11<00:03,  6.48it/s, loss=3.28, v_num=0]
Epoch 28:  76%|███████▋  | 74/97 [00:11<00:03,  6.45it/s, loss=3.28, v_num=0]
Epoch 28:  77%|███████▋  | 75/97 [00:11<00:03,  6.49it/s, loss=3.28, v_num=0]
Epoch 28:  77%|███████▋  | 75/97 [00:11<00:03,  6.46it/s, loss=3.28, v_num=0]
Epoch 28:  78%|███████▊  | 76/97 [00:11<00:03,  6.50it/s, loss=3.28, v_num=0]
Epoch 28:  78%|███████▊  | 76/97 [00:11<00:03,  6.47it/s, loss=3.28, v_num=0]
Epoch 28:  79%|███████▉  | 77/97 [00:11<00:03,  6.50it/s, loss=3.28, v_num=0]
Epoch 28:  79%|███████▉  | 77/97 [00:11<00:03,  6.47it/s, loss=3.28, v_num=0]
Epoch 28:  80%|████████  | 78/97 [00:11<00:02,  6.51it/s, loss=3.28, v_num=0]
Epoch 28:  80%|████████  | 78/97 [00:12<00:02,  6.48it/s, loss=3.28, v_num=0]
Epoch 28:  81%|████████▏ | 79/97 [00:12<00:02,  6.52it/s, loss=3.28, v_num=0]
Epoch 28:  81%|████████▏ | 79/97 [00:12<00:02,  6.49it/s, loss=3.29, v_num=0]
Epoch 28:  82%|████████▏ | 80/97 [00:12<00:02,  6.53it/s, loss=3.29, v_num=0]
Epoch 28:  82%|████████▏ | 80/97 [00:12<00:02,  6.50it/s, loss=3.28, v_num=0]
Epoch 28:  84%|████████▎ | 81/97 [00:12<00:02,  6.54it/s, loss=3.28, v_num=0]
Epoch 28:  84%|████████▎ | 81/97 [00:12<00:02,  6.51it/s, loss=3.28, v_num=0]
Epoch 28:  85%|████████▍ | 82/97 [00:12<00:02,  6.54it/s, loss=3.28, v_num=0]
Epoch 28:  85%|████████▍ | 82/97 [00:12<00:02,  6.52it/s, loss=3.29, v_num=0]
Epoch 28:  86%|████████▌ | 83/97 [00:12<00:02,  6.55it/s, loss=3.29, v_num=0]
Epoch 28:  86%|████████▌ | 83/97 [00:12<00:02,  6.52it/s, loss=3.29, v_num=0]
Epoch 28:  87%|████████▋ | 84/97 [00:12<00:01,  6.56it/s, loss=3.29, v_num=0]
Epoch 28:  87%|████████▋ | 84/97 [00:12<00:01,  6.53it/s, loss=3.29, v_num=0]
Epoch 28:  88%|████████▊ | 85/97 [00:12<00:01,  6.57it/s, loss=3.29, v_num=0]
Epoch 28:  88%|████████▊ | 85/97 [00:12<00:01,  6.54it/s, loss=3.29, v_num=0]
Epoch 28:  89%|████████▊ | 86/97 [00:13<00:01,  6.57it/s, loss=3.29, v_num=0]
Epoch 28:  89%|████████▊ | 86/97 [00:13<00:01,  6.55it/s, loss=3.29, v_num=0]
Epoch 28:  90%|████████▉ | 87/97 [00:13<00:01,  6.58it/s, loss=3.29, v_num=0]
Epoch 28:  90%|████████▉ | 87/97 [00:13<00:01,  6.55it/s, loss=3.29, v_num=0]
Epoch 28:  91%|█████████ | 88/97 [00:13<00:01,  6.59it/s, loss=3.29, v_num=0]
Epoch 28:  91%|█████████ | 88/97 [00:13<00:01,  6.56it/s, loss=3.28, v_num=0]
Epoch 28:  92%|█████████▏| 89/97 [00:13<00:01,  6.60it/s, loss=3.28, v_num=0]
Epoch 28:  92%|█████████▏| 89/97 [00:13<00:01,  6.57it/s, loss=3.28, v_num=0]
Epoch 28:  93%|█████████▎| 90/97 [00:13<00:01,  6.60it/s, loss=3.28, v_num=0]
Epoch 28:  93%|█████████▎| 90/97 [00:13<00:01,  6.58it/s, loss=3.29, v_num=0]
Epoch 28:  94%|█████████▍| 91/97 [00:13<00:00,  6.61it/s, loss=3.29, v_num=0]
Epoch 28:  94%|█████████▍| 91/97 [00:13<00:00,  6.59it/s, loss=3.29, v_num=0]
Epoch 28:  95%|█████████▍| 92/97 [00:13<00:00,  6.62it/s, loss=3.29, v_num=0]
Epoch 28:  95%|█████████▍| 92/97 [00:13<00:00,  6.59it/s, loss=3.29, v_num=0]
Epoch 28:  96%|█████████▌| 93/97 [00:14<00:00,  6.63it/s, loss=3.29, v_num=0]
Epoch 28:  96%|█████████▌| 93/97 [00:14<00:00,  6.60it/s, loss=3.29, v_num=0]
Epoch 28:  97%|█████████▋| 94/97 [00:14<00:00,  6.63it/s, loss=3.29, v_num=0]
Epoch 28:  97%|█████████▋| 94/97 [00:14<00:00,  6.61it/s, loss=3.29, v_num=0]
Epoch 28:  98%|█████████▊| 95/97 [00:14<00:00,  6.64it/s, loss=3.29, v_num=0]
Epoch 28:  98%|█████████▊| 95/97 [00:14<00:00,  6.61it/s, loss=3.3, v_num=0]
Epoch 28:  99%|█████████▉| 96/97 [00:14<00:00,  6.65it/s, loss=3.3, v_num=0]
Epoch 28:  99%|█████████▉| 96/97 [00:14<00:00,  6.62it/s, loss=3.3, v_num=0]
Epoch 28: 100%|██████████| 97/97 [00:14<00:00,  6.65it/s, loss=3.3, v_num=0]
Epoch 28: 100%|██████████| 97/97 [00:14<00:00,  6.63it/s, loss=3.29, v_num=0]
Epoch 28: 100%|██████████| 97/97 [00:14<00:00,  6.63it/s, loss=3.29, v_num=0]
Epoch 28:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.29, v_num=0]
Epoch 29:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.29, v_num=0]
Epoch 29:   1%|          | 1/97 [00:01<02:12,  1.38s/it, loss=3.29, v_num=0]
Epoch 29:   1%|          | 1/97 [00:01<02:17,  1.43s/it, loss=3.29, v_num=0]
Epoch 29:   2%|▏         | 2/97 [00:01<01:12,  1.31it/s, loss=3.29, v_num=0]
Epoch 29:   2%|▏         | 2/97 [00:01<01:14,  1.27it/s, loss=3.29, v_num=0]
Epoch 29:   3%|▎         | 3/97 [00:01<00:51,  1.81it/s, loss=3.29, v_num=0]
Epoch 29:   3%|▎         | 3/97 [00:01<00:53,  1.75it/s, loss=3.3, v_num=0]
Epoch 29:   4%|▍         | 4/97 [00:01<00:41,  2.23it/s, loss=3.3, v_num=0]
Epoch 29:   4%|▍         | 4/97 [00:01<00:42,  2.16it/s, loss=3.29, v_num=0]
Epoch 29:   5%|▌         | 5/97 [00:01<00:35,  2.60it/s, loss=3.29, v_num=0]
Epoch 29:   5%|▌         | 5/97 [00:01<00:36,  2.52it/s, loss=3.29, v_num=0]
Epoch 29:   6%|▌         | 6/97 [00:02<00:31,  2.90it/s, loss=3.29, v_num=0]
Epoch 29:   6%|▌         | 6/97 [00:02<00:32,  2.83it/s, loss=3.29, v_num=0]
Epoch 29:   7%|▋         | 7/97 [00:02<00:28,  3.18it/s, loss=3.29, v_num=0]
Epoch 29:   7%|▋         | 7/97 [00:02<00:29,  3.10it/s, loss=3.29, v_num=0]
Epoch 29:   8%|▊         | 8/97 [00:02<00:26,  3.42it/s, loss=3.29, v_num=0]
Epoch 29:   8%|▊         | 8/97 [00:02<00:26,  3.34it/s, loss=3.28, v_num=0]
Epoch 29:   9%|▉         | 9/97 [00:02<00:24,  3.64it/s, loss=3.28, v_num=0]
Epoch 29:   9%|▉         | 9/97 [00:02<00:24,  3.55it/s, loss=3.28, v_num=0]
Epoch 29:  10%|█         | 10/97 [00:02<00:22,  3.81it/s, loss=3.28, v_num=0]
Epoch 29:  10%|█         | 10/97 [00:02<00:23,  3.74it/s, loss=3.28, v_num=0]
Epoch 29:  11%|█▏        | 11/97 [00:02<00:21,  3.99it/s, loss=3.28, v_num=0]
Epoch 29:  11%|█▏        | 11/97 [00:02<00:21,  3.91it/s, loss=3.28, v_num=0]
Epoch 29:  12%|█▏        | 12/97 [00:02<00:20,  4.15it/s, loss=3.28, v_num=0]
Epoch 29:  12%|█▏        | 12/97 [00:02<00:20,  4.07it/s, loss=3.29, v_num=0]
Epoch 29:  13%|█▎        | 13/97 [00:03<00:19,  4.29it/s, loss=3.29, v_num=0]
Epoch 29:  13%|█▎        | 13/97 [00:03<00:19,  4.22it/s, loss=3.28, v_num=0]
Epoch 29:  14%|█▍        | 14/97 [00:03<00:18,  4.43it/s, loss=3.28, v_num=0]
Epoch 29:  14%|█▍        | 14/97 [00:03<00:19,  4.35it/s, loss=3.28, v_num=0]
Epoch 29:  15%|█▌        | 15/97 [00:03<00:18,  4.55it/s, loss=3.28, v_num=0]
Epoch 29:  15%|█▌        | 15/97 [00:03<00:18,  4.47it/s, loss=3.28, v_num=0]
Epoch 29:  16%|█▋        | 16/97 [00:03<00:17,  4.66it/s, loss=3.28, v_num=0]
Epoch 29:  16%|█▋        | 16/97 [00:03<00:17,  4.59it/s, loss=3.28, v_num=0]
Epoch 29:  18%|█▊        | 17/97 [00:03<00:16,  4.76it/s, loss=3.28, v_num=0]
Epoch 29:  18%|█▊        | 17/97 [00:03<00:17,  4.68it/s, loss=3.27, v_num=0]
Epoch 29:  19%|█▊        | 18/97 [00:03<00:16,  4.85it/s, loss=3.27, v_num=0]
Epoch 29:  19%|█▊        | 18/97 [00:03<00:16,  4.78it/s, loss=3.28, v_num=0]
Epoch 29:  20%|█▉        | 19/97 [00:03<00:15,  4.94it/s, loss=3.28, v_num=0]
Epoch 29:  20%|█▉        | 19/97 [00:03<00:16,  4.87it/s, loss=3.27, v_num=0]
Epoch 29:  21%|██        | 20/97 [00:03<00:15,  5.02it/s, loss=3.27, v_num=0]
Epoch 29:  21%|██        | 20/97 [00:04<00:15,  4.95it/s, loss=3.27, v_num=0]
Epoch 29:  22%|██▏       | 21/97 [00:04<00:14,  5.10it/s, loss=3.27, v_num=0]
Epoch 29:  22%|██▏       | 21/97 [00:04<00:15,  5.03it/s, loss=3.27, v_num=0]
Epoch 29:  23%|██▎       | 22/97 [00:04<00:14,  5.17it/s, loss=3.27, v_num=0]
Epoch 29:  23%|██▎       | 22/97 [00:04<00:14,  5.10it/s, loss=3.27, v_num=0]
Epoch 29:  24%|██▎       | 23/97 [00:04<00:14,  5.24it/s, loss=3.27, v_num=0]
Epoch 29:  24%|██▎       | 23/97 [00:04<00:14,  5.17it/s, loss=3.26, v_num=0]
Epoch 29:  25%|██▍       | 24/97 [00:04<00:13,  5.29it/s, loss=3.26, v_num=0]
Epoch 29:  25%|██▍       | 24/97 [00:04<00:13,  5.23it/s, loss=3.27, v_num=0]
Epoch 29:  26%|██▌       | 25/97 [00:04<00:13,  5.35it/s, loss=3.27, v_num=0]
Epoch 29:  26%|██▌       | 25/97 [00:04<00:13,  5.29it/s, loss=3.26, v_num=0]
Epoch 29:  27%|██▋       | 26/97 [00:04<00:13,  5.40it/s, loss=3.26, v_num=0]
Epoch 29:  27%|██▋       | 26/97 [00:04<00:13,  5.34it/s, loss=3.24, v_num=0]
Epoch 29:  28%|██▊       | 27/97 [00:04<00:12,  5.46it/s, loss=3.24, v_num=0]
Epoch 29:  28%|██▊       | 27/97 [00:05<00:12,  5.40it/s, loss=3.25, v_num=0]
Epoch 29:  29%|██▉       | 28/97 [00:05<00:12,  5.51it/s, loss=3.25, v_num=0]
Epoch 29:  29%|██▉       | 28/97 [00:05<00:12,  5.45it/s, loss=3.25, v_num=0]
Epoch 29:  30%|██▉       | 29/97 [00:05<00:12,  5.56it/s, loss=3.25, v_num=0]
Epoch 29:  30%|██▉       | 29/97 [00:05<00:12,  5.50it/s, loss=3.25, v_num=0]
Epoch 29:  31%|███       | 30/97 [00:05<00:11,  5.60it/s, loss=3.25, v_num=0]
Epoch 29:  31%|███       | 30/97 [00:05<00:12,  5.54it/s, loss=3.25, v_num=0]
Epoch 29:  32%|███▏      | 31/97 [00:05<00:11,  5.65it/s, loss=3.25, v_num=0]
Epoch 29:  32%|███▏      | 31/97 [00:05<00:11,  5.59it/s, loss=3.25, v_num=0]
Epoch 29:  33%|███▎      | 32/97 [00:05<00:11,  5.69it/s, loss=3.25, v_num=0]
Epoch 29:  33%|███▎      | 32/97 [00:05<00:11,  5.63it/s, loss=3.25, v_num=0]
Epoch 29:  34%|███▍      | 33/97 [00:05<00:11,  5.73it/s, loss=3.25, v_num=0]
Epoch 29:  34%|███▍      | 33/97 [00:05<00:11,  5.67it/s, loss=3.25, v_num=0]
Epoch 29:  35%|███▌      | 34/97 [00:05<00:10,  5.76it/s, loss=3.25, v_num=0]
Epoch 29:  35%|███▌      | 34/97 [00:05<00:11,  5.71it/s, loss=3.25, v_num=0]
Epoch 29:  36%|███▌      | 35/97 [00:06<00:10,  5.80it/s, loss=3.25, v_num=0]
Epoch 29:  36%|███▌      | 35/97 [00:06<00:10,  5.74it/s, loss=3.24, v_num=0]
Epoch 29:  37%|███▋      | 36/97 [00:06<00:10,  5.83it/s, loss=3.24, v_num=0]
Epoch 29:  37%|███▋      | 36/97 [00:06<00:10,  5.78it/s, loss=3.24, v_num=0]
Epoch 29:  38%|███▊      | 37/97 [00:06<00:10,  5.86it/s, loss=3.24, v_num=0]
Epoch 29:  38%|███▊      | 37/97 [00:06<00:10,  5.81it/s, loss=3.25, v_num=0]
Epoch 29:  39%|███▉      | 38/97 [00:06<00:10,  5.89it/s, loss=3.25, v_num=0]
Epoch 29:  39%|███▉      | 38/97 [00:06<00:10,  5.84it/s, loss=3.25, v_num=0]
Epoch 29:  40%|████      | 39/97 [00:06<00:09,  5.92it/s, loss=3.25, v_num=0]
Epoch 29:  40%|████      | 39/97 [00:06<00:09,  5.87it/s, loss=3.25, v_num=0]
Epoch 29:  41%|████      | 40/97 [00:06<00:09,  5.95it/s, loss=3.25, v_num=0]
Epoch 29:  41%|████      | 40/97 [00:06<00:09,  5.90it/s, loss=3.25, v_num=0]
Epoch 29:  42%|████▏     | 41/97 [00:06<00:09,  5.97it/s, loss=3.25, v_num=0]
Epoch 29:  42%|████▏     | 41/97 [00:06<00:09,  5.93it/s, loss=3.24, v_num=0]
Epoch 29:  43%|████▎     | 42/97 [00:06<00:09,  6.00it/s, loss=3.24, v_num=0]
Epoch 29:  43%|████▎     | 42/97 [00:07<00:09,  5.95it/s, loss=3.24, v_num=0]
Epoch 29:  44%|████▍     | 43/97 [00:07<00:08,  6.03it/s, loss=3.24, v_num=0]
Epoch 29:  44%|████▍     | 43/97 [00:07<00:09,  5.98it/s, loss=3.25, v_num=0]
Epoch 29:  45%|████▌     | 44/97 [00:07<00:08,  6.05it/s, loss=3.25, v_num=0]
Epoch 29:  45%|████▌     | 44/97 [00:07<00:08,  6.00it/s, loss=3.24, v_num=0]
Epoch 29:  46%|████▋     | 45/97 [00:07<00:08,  6.07it/s, loss=3.24, v_num=0]
Epoch 29:  46%|████▋     | 45/97 [00:07<00:08,  6.02it/s, loss=3.24, v_num=0]
Epoch 29:  47%|████▋     | 46/97 [00:07<00:08,  6.09it/s, loss=3.24, v_num=0]
Epoch 29:  47%|████▋     | 46/97 [00:07<00:08,  6.04it/s, loss=3.24, v_num=0]
Epoch 29:  48%|████▊     | 47/97 [00:07<00:08,  6.11it/s, loss=3.24, v_num=0]
Epoch 29:  48%|████▊     | 47/97 [00:07<00:08,  6.07it/s, loss=3.23, v_num=0]
Epoch 29:  49%|████▉     | 48/97 [00:07<00:07,  6.13it/s, loss=3.23, v_num=0]
Epoch 29:  49%|████▉     | 48/97 [00:07<00:08,  6.09it/s, loss=3.24, v_num=0]
Epoch 29:  51%|█████     | 49/97 [00:07<00:07,  6.15it/s, loss=3.24, v_num=0]
Epoch 29:  51%|█████     | 49/97 [00:08<00:07,  6.11it/s, loss=3.24, v_num=0]
Epoch 29:  52%|█████▏    | 50/97 [00:08<00:07,  6.17it/s, loss=3.24, v_num=0]
Epoch 29:  52%|█████▏    | 50/97 [00:08<00:07,  6.13it/s, loss=3.24, v_num=0]
Epoch 29:  53%|█████▎    | 51/97 [00:08<00:07,  6.19it/s, loss=3.24, v_num=0]
Epoch 29:  53%|█████▎    | 51/97 [00:08<00:07,  6.15it/s, loss=3.23, v_num=0]
Epoch 29:  54%|█████▎    | 52/97 [00:08<00:07,  6.21it/s, loss=3.23, v_num=0]
Epoch 29:  54%|█████▎    | 52/97 [00:08<00:07,  6.17it/s, loss=3.23, v_num=0]
Epoch 29:  55%|█████▍    | 53/97 [00:08<00:07,  6.22it/s, loss=3.23, v_num=0]
Epoch 29:  55%|█████▍    | 53/97 [00:08<00:07,  6.18it/s, loss=3.22, v_num=0]
Epoch 29:  56%|█████▌    | 54/97 [00:08<00:06,  6.24it/s, loss=3.22, v_num=0]
Epoch 29:  56%|█████▌    | 54/97 [00:08<00:06,  6.20it/s, loss=3.22, v_num=0]
Epoch 29:  57%|█████▋    | 55/97 [00:08<00:06,  6.25it/s, loss=3.22, v_num=0]
Epoch 29:  57%|█████▋    | 55/97 [00:08<00:06,  6.22it/s, loss=3.22, v_num=0]
Epoch 29:  58%|█████▊    | 56/97 [00:08<00:06,  6.27it/s, loss=3.22, v_num=0]
Epoch 29:  58%|█████▊    | 56/97 [00:08<00:06,  6.23it/s, loss=3.22, v_num=0]
Epoch 29:  59%|█████▉    | 57/97 [00:09<00:06,  6.28it/s, loss=3.22, v_num=0]
Epoch 29:  59%|█████▉    | 57/97 [00:09<00:06,  6.24it/s, loss=3.22, v_num=0]
Epoch 29:  60%|█████▉    | 58/97 [00:09<00:06,  6.30it/s, loss=3.22, v_num=0]
Epoch 29:  60%|█████▉    | 58/97 [00:09<00:06,  6.26it/s, loss=3.21, v_num=0]
Epoch 29:  61%|██████    | 59/97 [00:09<00:06,  6.31it/s, loss=3.21, v_num=0]
Epoch 29:  61%|██████    | 59/97 [00:09<00:06,  6.27it/s, loss=3.22, v_num=0]
Epoch 29:  62%|██████▏   | 60/97 [00:09<00:05,  6.32it/s, loss=3.22, v_num=0]
Epoch 29:  62%|██████▏   | 60/97 [00:09<00:05,  6.29it/s, loss=3.22, v_num=0]
Epoch 29:  63%|██████▎   | 61/97 [00:09<00:05,  6.34it/s, loss=3.22, v_num=0]
Epoch 29:  63%|██████▎   | 61/97 [00:09<00:05,  6.30it/s, loss=3.22, v_num=0]
Epoch 29:  64%|██████▍   | 62/97 [00:09<00:05,  6.35it/s, loss=3.22, v_num=0]
Epoch 29:  64%|██████▍   | 62/97 [00:09<00:05,  6.31it/s, loss=3.21, v_num=0]
Epoch 29:  65%|██████▍   | 63/97 [00:09<00:05,  6.36it/s, loss=3.21, v_num=0]
Epoch 29:  65%|██████▍   | 63/97 [00:09<00:05,  6.33it/s, loss=3.22, v_num=0]
Epoch 29:  66%|██████▌   | 64/97 [00:10<00:05,  6.37it/s, loss=3.22, v_num=0]
Epoch 29:  66%|██████▌   | 64/97 [00:10<00:05,  6.34it/s, loss=3.23, v_num=0]
Epoch 29:  67%|██████▋   | 65/97 [00:10<00:05,  6.39it/s, loss=3.23, v_num=0]
Epoch 29:  67%|██████▋   | 65/97 [00:10<00:05,  6.35it/s, loss=3.23, v_num=0]
Epoch 29:  68%|██████▊   | 66/97 [00:10<00:04,  6.40it/s, loss=3.23, v_num=0]
Epoch 29:  68%|██████▊   | 66/97 [00:10<00:04,  6.36it/s, loss=3.23, v_num=0]
Epoch 29:  69%|██████▉   | 67/97 [00:10<00:04,  6.41it/s, loss=3.23, v_num=0]
Epoch 29:  69%|██████▉   | 67/97 [00:10<00:04,  6.38it/s, loss=3.22, v_num=0]
Epoch 29:  70%|███████   | 68/97 [00:10<00:04,  6.42it/s, loss=3.22, v_num=0]
Epoch 29:  70%|███████   | 68/97 [00:10<00:04,  6.39it/s, loss=3.22, v_num=0]
Epoch 29:  71%|███████   | 69/97 [00:10<00:04,  6.44it/s, loss=3.22, v_num=0]
Epoch 29:  71%|███████   | 69/97 [00:10<00:04,  6.40it/s, loss=3.23, v_num=0]
Epoch 29:  72%|███████▏  | 70/97 [00:10<00:04,  6.44it/s, loss=3.23, v_num=0]
Epoch 29:  72%|███████▏  | 70/97 [00:10<00:04,  6.41it/s, loss=3.24, v_num=0]
Epoch 29:  73%|███████▎  | 71/97 [00:10<00:04,  6.46it/s, loss=3.24, v_num=0]
Epoch 29:  73%|███████▎  | 71/97 [00:11<00:04,  6.42it/s, loss=3.25, v_num=0]
Epoch 29:  74%|███████▍  | 72/97 [00:11<00:03,  6.46it/s, loss=3.25, v_num=0]
Epoch 29:  74%|███████▍  | 72/97 [00:11<00:03,  6.43it/s, loss=3.24, v_num=0]
Epoch 29:  75%|███████▌  | 73/97 [00:11<00:03,  6.48it/s, loss=3.24, v_num=0]
Epoch 29:  75%|███████▌  | 73/97 [00:11<00:03,  6.44it/s, loss=3.25, v_num=0]
Epoch 29:  76%|███████▋  | 74/97 [00:11<00:03,  6.48it/s, loss=3.25, v_num=0]
Epoch 29:  76%|███████▋  | 74/97 [00:11<00:03,  6.45it/s, loss=3.25, v_num=0]
Epoch 29:  77%|███████▋  | 75/97 [00:11<00:03,  6.49it/s, loss=3.25, v_num=0]
Epoch 29:  77%|███████▋  | 75/97 [00:11<00:03,  6.46it/s, loss=3.25, v_num=0]
Epoch 29:  78%|███████▊  | 76/97 [00:11<00:03,  6.50it/s, loss=3.25, v_num=0]
Epoch 29:  78%|███████▊  | 76/97 [00:11<00:03,  6.47it/s, loss=3.25, v_num=0]
Epoch 29:  79%|███████▉  | 77/97 [00:11<00:03,  6.51it/s, loss=3.25, v_num=0]
Epoch 29:  79%|███████▉  | 77/97 [00:11<00:03,  6.48it/s, loss=3.26, v_num=0]
Epoch 29:  80%|████████  | 78/97 [00:11<00:02,  6.52it/s, loss=3.26, v_num=0]
Epoch 29:  80%|████████  | 78/97 [00:12<00:02,  6.49it/s, loss=3.25, v_num=0]
Epoch 29:  81%|████████▏ | 79/97 [00:12<00:02,  6.53it/s, loss=3.25, v_num=0]
Epoch 29:  81%|████████▏ | 79/97 [00:12<00:02,  6.50it/s, loss=3.24, v_num=0]
Epoch 29:  82%|████████▏ | 80/97 [00:12<00:02,  6.54it/s, loss=3.24, v_num=0]
Epoch 29:  82%|████████▏ | 80/97 [00:12<00:02,  6.51it/s, loss=3.24, v_num=0]
Epoch 29:  84%|████████▎ | 81/97 [00:12<00:02,  6.54it/s, loss=3.24, v_num=0]
Epoch 29:  84%|████████▎ | 81/97 [00:12<00:02,  6.51it/s, loss=3.25, v_num=0]
Epoch 29:  85%|████████▍ | 82/97 [00:12<00:02,  6.55it/s, loss=3.25, v_num=0]
Epoch 29:  85%|████████▍ | 82/97 [00:12<00:02,  6.52it/s, loss=3.25, v_num=0]
Epoch 29:  86%|████████▌ | 83/97 [00:12<00:02,  6.56it/s, loss=3.25, v_num=0]
Epoch 29:  86%|████████▌ | 83/97 [00:12<00:02,  6.53it/s, loss=3.25, v_num=0]
Epoch 29:  87%|████████▋ | 84/97 [00:12<00:01,  6.57it/s, loss=3.25, v_num=0]
Epoch 29:  87%|████████▋ | 84/97 [00:12<00:01,  6.54it/s, loss=3.25, v_num=0]
Epoch 29:  88%|████████▊ | 85/97 [00:12<00:01,  6.57it/s, loss=3.25, v_num=0]
Epoch 29:  88%|████████▊ | 85/97 [00:12<00:01,  6.54it/s, loss=3.25, v_num=0]
Epoch 29:  89%|████████▊ | 86/97 [00:13<00:01,  6.58it/s, loss=3.25, v_num=0]
Epoch 29:  89%|████████▊ | 86/97 [00:13<00:01,  6.55it/s, loss=3.25, v_num=0]
Epoch 29:  90%|████████▉ | 87/97 [00:13<00:01,  6.59it/s, loss=3.25, v_num=0]
Epoch 29:  90%|████████▉ | 87/97 [00:13<00:01,  6.56it/s, loss=3.25, v_num=0]
Epoch 29:  91%|█████████ | 88/97 [00:13<00:01,  6.60it/s, loss=3.25, v_num=0]
Epoch 29:  91%|█████████ | 88/97 [00:13<00:01,  6.57it/s, loss=3.26, v_num=0]
Epoch 29:  92%|█████████▏| 89/97 [00:13<00:01,  6.60it/s, loss=3.26, v_num=0]
Epoch 29:  92%|█████████▏| 89/97 [00:13<00:01,  6.58it/s, loss=3.25, v_num=0]
Epoch 29:  93%|█████████▎| 90/97 [00:13<00:01,  6.61it/s, loss=3.25, v_num=0]
Epoch 29:  93%|█████████▎| 90/97 [00:13<00:01,  6.58it/s, loss=3.24, v_num=0]
Epoch 29:  94%|█████████▍| 91/97 [00:13<00:00,  6.62it/s, loss=3.24, v_num=0]
Epoch 29:  94%|█████████▍| 91/97 [00:13<00:00,  6.59it/s, loss=3.24, v_num=0]
Epoch 29:  95%|█████████▍| 92/97 [00:13<00:00,  6.63it/s, loss=3.24, v_num=0]
Epoch 29:  95%|█████████▍| 92/97 [00:13<00:00,  6.60it/s, loss=3.25, v_num=0]
Epoch 29:  96%|█████████▌| 93/97 [00:14<00:00,  6.63it/s, loss=3.25, v_num=0]
Epoch 29:  96%|█████████▌| 93/97 [00:14<00:00,  6.60it/s, loss=3.24, v_num=0]
Epoch 29:  97%|█████████▋| 94/97 [00:14<00:00,  6.64it/s, loss=3.24, v_num=0]
Epoch 29:  97%|█████████▋| 94/97 [00:14<00:00,  6.61it/s, loss=3.24, v_num=0]
Epoch 29:  98%|█████████▊| 95/97 [00:14<00:00,  6.64it/s, loss=3.24, v_num=0]
Epoch 29:  98%|█████████▊| 95/97 [00:14<00:00,  6.62it/s, loss=3.24, v_num=0]
Epoch 29:  99%|█████████▉| 96/97 [00:14<00:00,  6.65it/s, loss=3.24, v_num=0]
Epoch 29:  99%|█████████▉| 96/97 [00:14<00:00,  6.62it/s, loss=3.24, v_num=0]
Epoch 29: 100%|██████████| 97/97 [00:14<00:00,  6.66it/s, loss=3.24, v_num=0]
Epoch 29: 100%|██████████| 97/97 [00:14<00:00,  6.63it/s, loss=3.23, v_num=0]
Epoch 29: 100%|██████████| 97/97 [00:14<00:00,  6.63it/s, loss=3.23, v_num=0]
Epoch 29:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.23, v_num=0]
Epoch 30:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.23, v_num=0]
Epoch 30:   1%|          | 1/97 [00:01<02:14,  1.40s/it, loss=3.23, v_num=0]
Epoch 30:   1%|          | 1/97 [00:01<02:19,  1.45s/it, loss=3.24, v_num=0]
Epoch 30:   2%|▏         | 2/97 [00:01<01:12,  1.31it/s, loss=3.24, v_num=0]
Epoch 30:   2%|▏         | 2/97 [00:01<01:15,  1.26it/s, loss=3.24, v_num=0]
Epoch 30:   3%|▎         | 3/97 [00:01<00:52,  1.79it/s, loss=3.24, v_num=0]
Epoch 30:   3%|▎         | 3/97 [00:01<00:54,  1.74it/s, loss=3.24, v_num=0]
Epoch 30:   4%|▍         | 4/97 [00:01<00:41,  2.22it/s, loss=3.24, v_num=0]
Epoch 30:   4%|▍         | 4/97 [00:01<00:43,  2.15it/s, loss=3.23, v_num=0]
Epoch 30:   5%|▌         | 5/97 [00:01<00:35,  2.57it/s, loss=3.23, v_num=0]
Epoch 30:   5%|▌         | 5/97 [00:01<00:36,  2.50it/s, loss=3.23, v_num=0]
Epoch 30:   6%|▌         | 6/97 [00:02<00:31,  2.89it/s, loss=3.23, v_num=0]
Epoch 30:   6%|▌         | 6/97 [00:02<00:32,  2.81it/s, loss=3.23, v_num=0]
Epoch 30:   7%|▋         | 7/97 [00:02<00:28,  3.16it/s, loss=3.23, v_num=0]
Epoch 30:   7%|▋         | 7/97 [00:02<00:29,  3.08it/s, loss=3.22, v_num=0]
Epoch 30:   8%|▊         | 8/97 [00:02<00:26,  3.41it/s, loss=3.22, v_num=0]
Epoch 30:   8%|▊         | 8/97 [00:02<00:26,  3.32it/s, loss=3.22, v_num=0]
Epoch 30:   9%|▉         | 9/97 [00:02<00:24,  3.60it/s, loss=3.22, v_num=0]
Epoch 30:   9%|▉         | 9/97 [00:02<00:24,  3.52it/s, loss=3.22, v_num=0]
Epoch 30:  10%|█         | 10/97 [00:02<00:22,  3.80it/s, loss=3.22, v_num=0]
Epoch 30:  10%|█         | 10/97 [00:02<00:23,  3.71it/s, loss=3.22, v_num=0]
Epoch 30:  11%|█▏        | 11/97 [00:02<00:21,  3.97it/s, loss=3.22, v_num=0]
Epoch 30:  11%|█▏        | 11/97 [00:02<00:22,  3.89it/s, loss=3.21, v_num=0]
Epoch 30:  12%|█▏        | 12/97 [00:02<00:20,  4.13it/s, loss=3.21, v_num=0]
Epoch 30:  12%|█▏        | 12/97 [00:02<00:20,  4.05it/s, loss=3.2, v_num=0]
Epoch 30:  13%|█▎        | 13/97 [00:03<00:19,  4.27it/s, loss=3.2, v_num=0]
Epoch 30:  13%|█▎        | 13/97 [00:03<00:20,  4.20it/s, loss=3.2, v_num=0]
Epoch 30:  14%|█▍        | 14/97 [00:03<00:18,  4.41it/s, loss=3.2, v_num=0]
Epoch 30:  14%|█▍        | 14/97 [00:03<00:19,  4.33it/s, loss=3.2, v_num=0]
Epoch 30:  15%|█▌        | 15/97 [00:03<00:18,  4.52it/s, loss=3.2, v_num=0]
Epoch 30:  15%|█▌        | 15/97 [00:03<00:18,  4.45it/s, loss=3.2, v_num=0]
Epoch 30:  16%|█▋        | 16/97 [00:03<00:17,  4.64it/s, loss=3.2, v_num=0]
Epoch 30:  16%|█▋        | 16/97 [00:03<00:17,  4.56it/s, loss=3.2, v_num=0]
Epoch 30:  18%|█▊        | 17/97 [00:03<00:16,  4.74it/s, loss=3.2, v_num=0]
Epoch 30:  18%|█▊        | 17/97 [00:03<00:17,  4.66it/s, loss=3.2, v_num=0]
Epoch 30:  19%|█▊        | 18/97 [00:03<00:16,  4.83it/s, loss=3.2, v_num=0]
Epoch 30:  19%|█▊        | 18/97 [00:03<00:16,  4.76it/s, loss=3.2, v_num=0]
Epoch 30:  20%|█▉        | 19/97 [00:03<00:15,  4.92it/s, loss=3.2, v_num=0]
Epoch 30:  20%|█▉        | 19/97 [00:03<00:16,  4.85it/s, loss=3.19, v_num=0]
Epoch 30:  21%|██        | 20/97 [00:03<00:15,  5.01it/s, loss=3.19, v_num=0]
Epoch 30:  21%|██        | 20/97 [00:04<00:15,  4.94it/s, loss=3.18, v_num=0]
Epoch 30:  22%|██▏       | 21/97 [00:04<00:14,  5.08it/s, loss=3.18, v_num=0]
Epoch 30:  22%|██▏       | 21/97 [00:04<00:15,  5.02it/s, loss=3.18, v_num=0]
Epoch 30:  23%|██▎       | 22/97 [00:04<00:14,  5.15it/s, loss=3.18, v_num=0]
Epoch 30:  23%|██▎       | 22/97 [00:04<00:14,  5.08it/s, loss=3.17, v_num=0]
Epoch 30:  24%|██▎       | 23/97 [00:04<00:14,  5.22it/s, loss=3.17, v_num=0]
Epoch 30:  24%|██▎       | 23/97 [00:04<00:14,  5.15it/s, loss=3.17, v_num=0]
Epoch 30:  25%|██▍       | 24/97 [00:04<00:13,  5.28it/s, loss=3.17, v_num=0]
Epoch 30:  25%|██▍       | 24/97 [00:04<00:13,  5.22it/s, loss=3.17, v_num=0]
Epoch 30:  26%|██▌       | 25/97 [00:04<00:13,  5.34it/s, loss=3.17, v_num=0]
Epoch 30:  26%|██▌       | 25/97 [00:04<00:13,  5.28it/s, loss=3.17, v_num=0]
Epoch 30:  27%|██▋       | 26/97 [00:04<00:13,  5.40it/s, loss=3.17, v_num=0]
Epoch 30:  27%|██▋       | 26/97 [00:04<00:13,  5.33it/s, loss=3.16, v_num=0]
Epoch 30:  28%|██▊       | 27/97 [00:04<00:12,  5.45it/s, loss=3.16, v_num=0]
Epoch 30:  28%|██▊       | 27/97 [00:05<00:12,  5.39it/s, loss=3.17, v_num=0]
Epoch 30:  29%|██▉       | 28/97 [00:05<00:12,  5.50it/s, loss=3.17, v_num=0]
Epoch 30:  29%|██▉       | 28/97 [00:05<00:12,  5.44it/s, loss=3.16, v_num=0]
Epoch 30:  30%|██▉       | 29/97 [00:05<00:12,  5.55it/s, loss=3.16, v_num=0]
Epoch 30:  30%|██▉       | 29/97 [00:05<00:12,  5.49it/s, loss=3.17, v_num=0]
Epoch 30:  31%|███       | 30/97 [00:05<00:11,  5.59it/s, loss=3.17, v_num=0]
Epoch 30:  31%|███       | 30/97 [00:05<00:12,  5.54it/s, loss=3.17, v_num=0]
Epoch 30:  32%|███▏      | 31/97 [00:05<00:11,  5.64it/s, loss=3.17, v_num=0]
Epoch 30:  32%|███▏      | 31/97 [00:05<00:11,  5.58it/s, loss=3.17, v_num=0]
Epoch 30:  33%|███▎      | 32/97 [00:05<00:11,  5.68it/s, loss=3.17, v_num=0]
Epoch 30:  33%|███▎      | 32/97 [00:05<00:11,  5.62it/s, loss=3.17, v_num=0]
Epoch 30:  34%|███▍      | 33/97 [00:05<00:11,  5.71it/s, loss=3.17, v_num=0]
Epoch 30:  34%|███▍      | 33/97 [00:05<00:11,  5.66it/s, loss=3.18, v_num=0]
Epoch 30:  35%|███▌      | 34/97 [00:05<00:10,  5.75it/s, loss=3.18, v_num=0]
Epoch 30:  35%|███▌      | 34/97 [00:05<00:11,  5.70it/s, loss=3.17, v_num=0]
Epoch 30:  36%|███▌      | 35/97 [00:06<00:10,  5.79it/s, loss=3.17, v_num=0]
Epoch 30:  36%|███▌      | 35/97 [00:06<00:10,  5.74it/s, loss=3.18, v_num=0]
Epoch 30:  37%|███▋      | 36/97 [00:06<00:10,  5.82it/s, loss=3.18, v_num=0]
Epoch 30:  37%|███▋      | 36/97 [00:06<00:10,  5.77it/s, loss=3.18, v_num=0]
Epoch 30:  38%|███▊      | 37/97 [00:06<00:10,  5.85it/s, loss=3.18, v_num=0]
Epoch 30:  38%|███▊      | 37/97 [00:06<00:10,  5.80it/s, loss=3.18, v_num=0]
Epoch 30:  39%|███▉      | 38/97 [00:06<00:10,  5.89it/s, loss=3.18, v_num=0]
Epoch 30:  39%|███▉      | 38/97 [00:06<00:10,  5.84it/s, loss=3.17, v_num=0]
Epoch 30:  40%|████      | 39/97 [00:06<00:09,  5.92it/s, loss=3.17, v_num=0]
Epoch 30:  40%|████      | 39/97 [00:06<00:09,  5.87it/s, loss=3.18, v_num=0]
Epoch 30:  41%|████      | 40/97 [00:06<00:09,  5.94it/s, loss=3.18, v_num=0]
Epoch 30:  41%|████      | 40/97 [00:06<00:09,  5.90it/s, loss=3.19, v_num=0]
Epoch 30:  42%|████▏     | 41/97 [00:06<00:09,  5.97it/s, loss=3.19, v_num=0]
Epoch 30:  42%|████▏     | 41/97 [00:06<00:09,  5.92it/s, loss=3.2, v_num=0]
Epoch 30:  43%|████▎     | 42/97 [00:07<00:09,  6.00it/s, loss=3.2, v_num=0]
Epoch 30:  43%|████▎     | 42/97 [00:07<00:09,  5.95it/s, loss=3.2, v_num=0]
Epoch 30:  44%|████▍     | 43/97 [00:07<00:08,  6.02it/s, loss=3.2, v_num=0]
Epoch 30:  44%|████▍     | 43/97 [00:07<00:09,  5.98it/s, loss=3.2, v_num=0]
Epoch 30:  45%|████▌     | 44/97 [00:07<00:08,  6.05it/s, loss=3.2, v_num=0]
Epoch 30:  45%|████▌     | 44/97 [00:07<00:08,  6.00it/s, loss=3.2, v_num=0]
Epoch 30:  46%|████▋     | 45/97 [00:07<00:08,  6.07it/s, loss=3.2, v_num=0]
Epoch 30:  46%|████▋     | 45/97 [00:07<00:08,  6.02it/s, loss=3.21, v_num=0]
Epoch 30:  47%|████▋     | 46/97 [00:07<00:08,  6.09it/s, loss=3.21, v_num=0]
Epoch 30:  47%|████▋     | 46/97 [00:07<00:08,  6.05it/s, loss=3.21, v_num=0]
Epoch 30:  48%|████▊     | 47/97 [00:07<00:08,  6.11it/s, loss=3.21, v_num=0]
Epoch 30:  48%|████▊     | 47/97 [00:07<00:08,  6.07it/s, loss=3.22, v_num=0]
Epoch 30:  49%|████▉     | 48/97 [00:07<00:07,  6.14it/s, loss=3.22, v_num=0]
Epoch 30:  49%|████▉     | 48/97 [00:07<00:08,  6.09it/s, loss=3.21, v_num=0]
Epoch 30:  51%|█████     | 49/97 [00:07<00:07,  6.16it/s, loss=3.21, v_num=0]
Epoch 30:  51%|█████     | 49/97 [00:08<00:07,  6.11it/s, loss=3.2, v_num=0]
Epoch 30:  52%|█████▏    | 50/97 [00:08<00:07,  6.18it/s, loss=3.2, v_num=0]
Epoch 30:  52%|█████▏    | 50/97 [00:08<00:07,  6.13it/s, loss=3.21, v_num=0]
Epoch 30:  53%|█████▎    | 51/97 [00:08<00:07,  6.19it/s, loss=3.21, v_num=0]
Epoch 30:  53%|█████▎    | 51/97 [00:08<00:07,  6.15it/s, loss=3.21, v_num=0]
Epoch 30:  54%|█████▎    | 52/97 [00:08<00:07,  6.21it/s, loss=3.21, v_num=0]
Epoch 30:  54%|█████▎    | 52/97 [00:08<00:07,  6.17it/s, loss=3.21, v_num=0]
Epoch 30:  55%|█████▍    | 53/97 [00:08<00:07,  6.23it/s, loss=3.21, v_num=0]
Epoch 30:  55%|█████▍    | 53/97 [00:08<00:07,  6.19it/s, loss=3.2, v_num=0]
Epoch 30:  56%|█████▌    | 54/97 [00:08<00:06,  6.25it/s, loss=3.2, v_num=0]
Epoch 30:  56%|█████▌    | 54/97 [00:08<00:06,  6.21it/s, loss=3.21, v_num=0]
Epoch 30:  57%|█████▋    | 55/97 [00:08<00:06,  6.26it/s, loss=3.21, v_num=0]
Epoch 30:  57%|█████▋    | 55/97 [00:08<00:06,  6.22it/s, loss=3.2, v_num=0]
Epoch 30:  58%|█████▊    | 56/97 [00:08<00:06,  6.28it/s, loss=3.2, v_num=0]
Epoch 30:  58%|█████▊    | 56/97 [00:08<00:06,  6.24it/s, loss=3.21, v_num=0]
Epoch 30:  59%|█████▉    | 57/97 [00:09<00:06,  6.29it/s, loss=3.21, v_num=0]
Epoch 30:  59%|█████▉    | 57/97 [00:09<00:06,  6.25it/s, loss=3.21, v_num=0]
Epoch 30:  60%|█████▉    | 58/97 [00:09<00:06,  6.31it/s, loss=3.21, v_num=0]
Epoch 30:  60%|█████▉    | 58/97 [00:09<00:06,  6.27it/s, loss=3.21, v_num=0]
Epoch 30:  61%|██████    | 59/97 [00:09<00:06,  6.32it/s, loss=3.21, v_num=0]
Epoch 30:  61%|██████    | 59/97 [00:09<00:06,  6.28it/s, loss=3.21, v_num=0]
Epoch 30:  62%|██████▏   | 60/97 [00:09<00:05,  6.33it/s, loss=3.21, v_num=0]
Epoch 30:  62%|██████▏   | 60/97 [00:09<00:05,  6.30it/s, loss=3.2, v_num=0]
Epoch 30:  63%|██████▎   | 61/97 [00:09<00:05,  6.35it/s, loss=3.2, v_num=0]
Epoch 30:  63%|██████▎   | 61/97 [00:09<00:05,  6.31it/s, loss=3.2, v_num=0]
Epoch 30:  64%|██████▍   | 62/97 [00:09<00:05,  6.36it/s, loss=3.2, v_num=0]
Epoch 30:  64%|██████▍   | 62/97 [00:09<00:05,  6.33it/s, loss=3.21, v_num=0]
Epoch 30:  65%|██████▍   | 63/97 [00:09<00:05,  6.37it/s, loss=3.21, v_num=0]
Epoch 30:  65%|██████▍   | 63/97 [00:09<00:05,  6.34it/s, loss=3.21, v_num=0]
Epoch 30:  66%|██████▌   | 64/97 [00:10<00:05,  6.38it/s, loss=3.21, v_num=0]
Epoch 30:  66%|██████▌   | 64/97 [00:10<00:05,  6.35it/s, loss=3.22, v_num=0]
Epoch 30:  67%|██████▋   | 65/97 [00:10<00:05,  6.40it/s, loss=3.22, v_num=0]
Epoch 30:  67%|██████▋   | 65/97 [00:10<00:05,  6.36it/s, loss=3.22, v_num=0]
Epoch 30:  68%|██████▊   | 66/97 [00:10<00:04,  6.41it/s, loss=3.22, v_num=0]
Epoch 30:  68%|██████▊   | 66/97 [00:10<00:04,  6.37it/s, loss=3.22, v_num=0]
Epoch 30:  69%|██████▉   | 67/97 [00:10<00:04,  6.42it/s, loss=3.22, v_num=0]
Epoch 30:  69%|██████▉   | 67/97 [00:10<00:04,  6.39it/s, loss=3.21, v_num=0]
Epoch 30:  70%|███████   | 68/97 [00:10<00:04,  6.43it/s, loss=3.21, v_num=0]
Epoch 30:  70%|███████   | 68/97 [00:10<00:04,  6.40it/s, loss=3.22, v_num=0]
Epoch 30:  71%|███████   | 69/97 [00:10<00:04,  6.44it/s, loss=3.22, v_num=0]
Epoch 30:  71%|███████   | 69/97 [00:10<00:04,  6.41it/s, loss=3.22, v_num=0]
Epoch 30:  72%|███████▏  | 70/97 [00:10<00:04,  6.45it/s, loss=3.22, v_num=0]
Epoch 30:  72%|███████▏  | 70/97 [00:10<00:04,  6.42it/s, loss=3.22, v_num=0]
Epoch 30:  73%|███████▎  | 71/97 [00:10<00:04,  6.46it/s, loss=3.22, v_num=0]
Epoch 30:  73%|███████▎  | 71/97 [00:11<00:04,  6.43it/s, loss=3.23, v_num=0]
Epoch 30:  74%|███████▍  | 72/97 [00:11<00:03,  6.47it/s, loss=3.23, v_num=0]
Epoch 30:  74%|███████▍  | 72/97 [00:11<00:03,  6.44it/s, loss=3.22, v_num=0]
Epoch 30:  75%|███████▌  | 73/97 [00:11<00:03,  6.49it/s, loss=3.22, v_num=0]
Epoch 30:  75%|███████▌  | 73/97 [00:11<00:03,  6.45it/s, loss=3.23, v_num=0]
Epoch 30:  76%|███████▋  | 74/97 [00:11<00:03,  6.49it/s, loss=3.23, v_num=0]
Epoch 30:  76%|███████▋  | 74/97 [00:11<00:03,  6.46it/s, loss=3.23, v_num=0]
Epoch 30:  77%|███████▋  | 75/97 [00:11<00:03,  6.50it/s, loss=3.23, v_num=0]
Epoch 30:  77%|███████▋  | 75/97 [00:11<00:03,  6.47it/s, loss=3.23, v_num=0]
Epoch 30:  78%|███████▊  | 76/97 [00:11<00:03,  6.51it/s, loss=3.23, v_num=0]
Epoch 30:  78%|███████▊  | 76/97 [00:11<00:03,  6.48it/s, loss=3.23, v_num=0]
Epoch 30:  79%|███████▉  | 77/97 [00:11<00:03,  6.52it/s, loss=3.23, v_num=0]
Epoch 30:  79%|███████▉  | 77/97 [00:11<00:03,  6.49it/s, loss=3.22, v_num=0]
Epoch 30:  80%|████████  | 78/97 [00:11<00:02,  6.53it/s, loss=3.22, v_num=0]
Epoch 30:  80%|████████  | 78/97 [00:12<00:02,  6.50it/s, loss=3.23, v_num=0]
Epoch 30:  81%|████████▏ | 79/97 [00:12<00:02,  6.53it/s, loss=3.23, v_num=0]
Epoch 30:  81%|████████▏ | 79/97 [00:12<00:02,  6.51it/s, loss=3.23, v_num=0]
Epoch 30:  82%|████████▏ | 80/97 [00:12<00:02,  6.55it/s, loss=3.23, v_num=0]
Epoch 30:  82%|████████▏ | 80/97 [00:12<00:02,  6.51it/s, loss=3.22, v_num=0]
Epoch 30:  84%|████████▎ | 81/97 [00:12<00:02,  6.55it/s, loss=3.22, v_num=0]
Epoch 30:  84%|████████▎ | 81/97 [00:12<00:02,  6.52it/s, loss=3.22, v_num=0]
Epoch 30:  85%|████████▍ | 82/97 [00:12<00:02,  6.56it/s, loss=3.22, v_num=0]
Epoch 30:  85%|████████▍ | 82/97 [00:12<00:02,  6.53it/s, loss=3.22, v_num=0]
Epoch 30:  86%|████████▌ | 83/97 [00:12<00:02,  6.57it/s, loss=3.22, v_num=0]
Epoch 30:  86%|████████▌ | 83/97 [00:12<00:02,  6.54it/s, loss=3.21, v_num=0]
Epoch 30:  87%|████████▋ | 84/97 [00:12<00:01,  6.57it/s, loss=3.21, v_num=0]
Epoch 30:  87%|████████▋ | 84/97 [00:12<00:01,  6.55it/s, loss=3.21, v_num=0]
Epoch 30:  88%|████████▊ | 85/97 [00:12<00:01,  6.58it/s, loss=3.21, v_num=0]
Epoch 30:  88%|████████▊ | 85/97 [00:12<00:01,  6.55it/s, loss=3.2, v_num=0]
Epoch 30:  89%|████████▊ | 86/97 [00:13<00:01,  6.59it/s, loss=3.2, v_num=0]
Epoch 30:  89%|████████▊ | 86/97 [00:13<00:01,  6.56it/s, loss=3.2, v_num=0]
Epoch 30:  90%|████████▉ | 87/97 [00:13<00:01,  6.60it/s, loss=3.2, v_num=0]
Epoch 30:  90%|████████▉ | 87/97 [00:13<00:01,  6.57it/s, loss=3.2, v_num=0]
Epoch 30:  91%|█████████ | 88/97 [00:13<00:01,  6.60it/s, loss=3.2, v_num=0]
Epoch 30:  91%|█████████ | 88/97 [00:13<00:01,  6.58it/s, loss=3.2, v_num=0]
Epoch 30:  92%|█████████▏| 89/97 [00:13<00:01,  6.61it/s, loss=3.2, v_num=0]
Epoch 30:  92%|█████████▏| 89/97 [00:13<00:01,  6.58it/s, loss=3.2, v_num=0]
Epoch 30:  93%|█████████▎| 90/97 [00:13<00:01,  6.62it/s, loss=3.2, v_num=0]
Epoch 30:  93%|█████████▎| 90/97 [00:13<00:01,  6.59it/s, loss=3.2, v_num=0]
Epoch 30:  94%|█████████▍| 91/97 [00:13<00:00,  6.62it/s, loss=3.2, v_num=0]
Epoch 30:  94%|█████████▍| 91/97 [00:13<00:00,  6.60it/s, loss=3.19, v_num=0]
Epoch 30:  95%|█████████▍| 92/97 [00:13<00:00,  6.63it/s, loss=3.19, v_num=0]
Epoch 30:  95%|█████████▍| 92/97 [00:13<00:00,  6.60it/s, loss=3.19, v_num=0]
Epoch 30:  96%|█████████▌| 93/97 [00:14<00:00,  6.64it/s, loss=3.19, v_num=0]
Epoch 30:  96%|█████████▌| 93/97 [00:14<00:00,  6.61it/s, loss=3.19, v_num=0]
Epoch 30:  97%|█████████▋| 94/97 [00:14<00:00,  6.64it/s, loss=3.19, v_num=0]
Epoch 30:  97%|█████████▋| 94/97 [00:14<00:00,  6.62it/s, loss=3.18, v_num=0]
Epoch 30:  98%|█████████▊| 95/97 [00:14<00:00,  6.65it/s, loss=3.18, v_num=0]
Epoch 30:  98%|█████████▊| 95/97 [00:14<00:00,  6.62it/s, loss=3.18, v_num=0]
Epoch 30:  99%|█████████▉| 96/97 [00:14<00:00,  6.66it/s, loss=3.18, v_num=0]
Epoch 30:  99%|█████████▉| 96/97 [00:14<00:00,  6.63it/s, loss=3.18, v_num=0]
Epoch 30: 100%|██████████| 97/97 [00:14<00:00,  6.66it/s, loss=3.18, v_num=0]
Epoch 30: 100%|██████████| 97/97 [00:14<00:00,  6.64it/s, loss=3.18, v_num=0]
Epoch 30: 100%|██████████| 97/97 [00:14<00:00,  6.64it/s, loss=3.18, v_num=0]
Epoch 30:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.18, v_num=0]
Epoch 31:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.18, v_num=0]
Epoch 31:   1%|          | 1/97 [00:01<02:36,  1.63s/it, loss=3.18, v_num=0]
Epoch 31:   1%|          | 1/97 [00:01<02:41,  1.68s/it, loss=3.17, v_num=0]
Epoch 31:   2%|▏         | 2/97 [00:01<01:24,  1.13it/s, loss=3.17, v_num=0]
Epoch 31:   2%|▏         | 2/97 [00:01<01:26,  1.09it/s, loss=3.18, v_num=0]
Epoch 31:   3%|▎         | 3/97 [00:01<01:00,  1.55it/s, loss=3.18, v_num=0]
Epoch 31:   3%|▎         | 3/97 [00:01<01:02,  1.51it/s, loss=3.18, v_num=0]
Epoch 31:   4%|▍         | 4/97 [00:02<00:48,  1.92it/s, loss=3.18, v_num=0]
Epoch 31:   4%|▍         | 4/97 [00:02<00:49,  1.87it/s, loss=3.18, v_num=0]
Epoch 31:   5%|▌         | 5/97 [00:02<00:40,  2.25it/s, loss=3.18, v_num=0]
Epoch 31:   5%|▌         | 5/97 [00:02<00:41,  2.20it/s, loss=3.18, v_num=0]
Epoch 31:   6%|▌         | 6/97 [00:02<00:35,  2.54it/s, loss=3.18, v_num=0]
Epoch 31:   6%|▌         | 6/97 [00:02<00:36,  2.48it/s, loss=3.19, v_num=0]
Epoch 31:   7%|▋         | 7/97 [00:02<00:32,  2.79it/s, loss=3.19, v_num=0]
Epoch 31:   7%|▋         | 7/97 [00:02<00:32,  2.74it/s, loss=3.19, v_num=0]
Epoch 31:   8%|▊         | 8/97 [00:02<00:29,  3.02it/s, loss=3.19, v_num=0]
Epoch 31:   8%|▊         | 8/97 [00:02<00:30,  2.96it/s, loss=3.19, v_num=0]
Epoch 31:   9%|▉         | 9/97 [00:02<00:27,  3.22it/s, loss=3.19, v_num=0]
Epoch 31:   9%|▉         | 9/97 [00:02<00:27,  3.16it/s, loss=3.19, v_num=0]
Epoch 31:  10%|█         | 10/97 [00:02<00:25,  3.41it/s, loss=3.19, v_num=0]
Epoch 31:  10%|█         | 10/97 [00:02<00:26,  3.34it/s, loss=3.2, v_num=0]
Epoch 31:  11%|█▏        | 11/97 [00:03<00:24,  3.57it/s, loss=3.2, v_num=0]
Epoch 31:  11%|█▏        | 11/97 [00:03<00:24,  3.51it/s, loss=3.2, v_num=0]
Epoch 31:  12%|█▏        | 12/97 [00:03<00:22,  3.72it/s, loss=3.2, v_num=0]
Epoch 31:  12%|█▏        | 12/97 [00:03<00:23,  3.66it/s, loss=3.2, v_num=0]
Epoch 31:  13%|█▎        | 13/97 [00:03<00:21,  3.86it/s, loss=3.2, v_num=0]
Epoch 31:  13%|█▎        | 13/97 [00:03<00:22,  3.80it/s, loss=3.21, v_num=0]
Epoch 31:  14%|█▍        | 14/97 [00:03<00:20,  3.99it/s, loss=3.21, v_num=0]
Epoch 31:  14%|█▍        | 14/97 [00:03<00:21,  3.93it/s, loss=3.22, v_num=0]
Epoch 31:  15%|█▌        | 15/97 [00:03<00:19,  4.12it/s, loss=3.22, v_num=0]
Epoch 31:  15%|█▌        | 15/97 [00:03<00:20,  4.05it/s, loss=3.21, v_num=0]
Epoch 31:  16%|█▋        | 16/97 [00:03<00:19,  4.23it/s, loss=3.21, v_num=0]
Epoch 31:  16%|█▋        | 16/97 [00:03<00:19,  4.17it/s, loss=3.21, v_num=0]
Epoch 31:  18%|█▊        | 17/97 [00:03<00:18,  4.34it/s, loss=3.21, v_num=0]
Epoch 31:  18%|█▊        | 17/97 [00:03<00:18,  4.28it/s, loss=3.22, v_num=0]
Epoch 31:  19%|█▊        | 18/97 [00:04<00:17,  4.43it/s, loss=3.22, v_num=0]
Epoch 31:  19%|█▊        | 18/97 [00:04<00:18,  4.37it/s, loss=3.22, v_num=0]
Epoch 31:  20%|█▉        | 19/97 [00:04<00:17,  4.50it/s, loss=3.22, v_num=0]
Epoch 31:  20%|█▉        | 19/97 [00:04<00:17,  4.45it/s, loss=3.22, v_num=0]
Epoch 31:  21%|██        | 20/97 [00:04<00:16,  4.59it/s, loss=3.22, v_num=0]
Epoch 31:  21%|██        | 20/97 [00:04<00:17,  4.53it/s, loss=3.22, v_num=0]
Epoch 31:  22%|██▏       | 21/97 [00:04<00:16,  4.66it/s, loss=3.22, v_num=0]
Epoch 31:  22%|██▏       | 21/97 [00:04<00:16,  4.61it/s, loss=3.22, v_num=0]
Epoch 31:  23%|██▎       | 22/97 [00:04<00:15,  4.74it/s, loss=3.22, v_num=0]
Epoch 31:  23%|██▎       | 22/97 [00:04<00:16,  4.68it/s, loss=3.22, v_num=0]
Epoch 31:  24%|██▎       | 23/97 [00:04<00:15,  4.81it/s, loss=3.22, v_num=0]
Epoch 31:  24%|██▎       | 23/97 [00:04<00:15,  4.75it/s, loss=3.22, v_num=0]
Epoch 31:  25%|██▍       | 24/97 [00:04<00:14,  4.88it/s, loss=3.22, v_num=0]
Epoch 31:  25%|██▍       | 24/97 [00:04<00:15,  4.82it/s, loss=3.22, v_num=0]
Epoch 31:  26%|██▌       | 25/97 [00:05<00:14,  4.94it/s, loss=3.22, v_num=0]
Epoch 31:  26%|██▌       | 25/97 [00:05<00:14,  4.89it/s, loss=3.23, v_num=0]
Epoch 31:  27%|██▋       | 26/97 [00:05<00:14,  5.00it/s, loss=3.23, v_num=0]
Epoch 31:  27%|██▋       | 26/97 [00:05<00:14,  4.95it/s, loss=3.23, v_num=0]
Epoch 31:  28%|██▊       | 27/97 [00:05<00:13,  5.05it/s, loss=3.23, v_num=0]
Epoch 31:  28%|██▊       | 27/97 [00:05<00:13,  5.01it/s, loss=3.24, v_num=0]
Epoch 31:  29%|██▉       | 28/97 [00:05<00:13,  5.12it/s, loss=3.24, v_num=0]
Epoch 31:  29%|██▉       | 28/97 [00:05<00:13,  5.06it/s, loss=3.23, v_num=0]
Epoch 31:  30%|██▉       | 29/97 [00:05<00:13,  5.17it/s, loss=3.23, v_num=0]
Epoch 31:  30%|██▉       | 29/97 [00:05<00:13,  5.12it/s, loss=3.24, v_num=0]
Epoch 31:  31%|███       | 30/97 [00:05<00:12,  5.21it/s, loss=3.24, v_num=0]
Epoch 31:  31%|███       | 30/97 [00:05<00:12,  5.16it/s, loss=3.23, v_num=0]
Epoch 31:  32%|███▏      | 31/97 [00:05<00:12,  5.26it/s, loss=3.23, v_num=0]
Epoch 31:  32%|███▏      | 31/97 [00:05<00:12,  5.21it/s, loss=3.22, v_num=0]
Epoch 31:  33%|███▎      | 32/97 [00:06<00:12,  5.31it/s, loss=3.22, v_num=0]
Epoch 31:  33%|███▎      | 32/97 [00:06<00:12,  5.26it/s, loss=3.22, v_num=0]
Epoch 31:  34%|███▍      | 33/97 [00:06<00:11,  5.35it/s, loss=3.22, v_num=0]
Epoch 31:  34%|███▍      | 33/97 [00:06<00:12,  5.30it/s, loss=3.21, v_num=0]
Epoch 31:  35%|███▌      | 34/97 [00:06<00:11,  5.39it/s, loss=3.21, v_num=0]
Epoch 31:  35%|███▌      | 34/97 [00:06<00:11,  5.34it/s, loss=3.2, v_num=0]
Epoch 31:  36%|███▌      | 35/97 [00:06<00:11,  5.43it/s, loss=3.2, v_num=0]
Epoch 31:  36%|███▌      | 35/97 [00:06<00:11,  5.38it/s, loss=3.21, v_num=0]
Epoch 31:  37%|███▋      | 36/97 [00:06<00:11,  5.47it/s, loss=3.21, v_num=0]
Epoch 31:  37%|███▋      | 36/97 [00:06<00:11,  5.42it/s, loss=3.21, v_num=0]
Epoch 31:  38%|███▊      | 37/97 [00:06<00:10,  5.50it/s, loss=3.21, v_num=0]
Epoch 31:  38%|███▊      | 37/97 [00:06<00:10,  5.46it/s, loss=3.2, v_num=0]
Epoch 31:  39%|███▉      | 38/97 [00:06<00:10,  5.54it/s, loss=3.2, v_num=0]
Epoch 31:  39%|███▉      | 38/97 [00:06<00:10,  5.49it/s, loss=3.2, v_num=0]
Epoch 31:  40%|████      | 39/97 [00:06<00:10,  5.57it/s, loss=3.2, v_num=0]
Epoch 31:  40%|████      | 39/97 [00:07<00:10,  5.53it/s, loss=3.18, v_num=0]
Epoch 31:  41%|████      | 40/97 [00:07<00:10,  5.60it/s, loss=3.18, v_num=0]
Epoch 31:  41%|████      | 40/97 [00:07<00:10,  5.56it/s, loss=3.2, v_num=0]
Epoch 31:  42%|████▏     | 41/97 [00:07<00:09,  5.63it/s, loss=3.2, v_num=0]
Epoch 31:  42%|████▏     | 41/97 [00:07<00:10,  5.59it/s, loss=3.2, v_num=0]
Epoch 31:  43%|████▎     | 42/97 [00:07<00:09,  5.66it/s, loss=3.2, v_num=0]
Epoch 31:  43%|████▎     | 42/97 [00:07<00:09,  5.62it/s, loss=3.19, v_num=0]
Epoch 31:  44%|████▍     | 43/97 [00:07<00:09,  5.69it/s, loss=3.19, v_num=0]
Epoch 31:  44%|████▍     | 43/97 [00:07<00:09,  5.65it/s, loss=3.18, v_num=0]
Epoch 31:  45%|████▌     | 44/97 [00:07<00:09,  5.72it/s, loss=3.18, v_num=0]
Epoch 31:  45%|████▌     | 44/97 [00:07<00:09,  5.67it/s, loss=3.18, v_num=0]
Epoch 31:  46%|████▋     | 45/97 [00:07<00:09,  5.74it/s, loss=3.18, v_num=0]
Epoch 31:  46%|████▋     | 45/97 [00:07<00:09,  5.70it/s, loss=3.18, v_num=0]
Epoch 31:  47%|████▋     | 46/97 [00:07<00:08,  5.77it/s, loss=3.18, v_num=0]
Epoch 31:  47%|████▋     | 46/97 [00:08<00:08,  5.73it/s, loss=3.18, v_num=0]
Epoch 31:  48%|████▊     | 47/97 [00:08<00:08,  5.79it/s, loss=3.18, v_num=0]
Epoch 31:  48%|████▊     | 47/97 [00:08<00:08,  5.75it/s, loss=3.17, v_num=0]
Epoch 31:  49%|████▉     | 48/97 [00:08<00:08,  5.82it/s, loss=3.17, v_num=0]
Epoch 31:  49%|████▉     | 48/97 [00:08<00:08,  5.78it/s, loss=3.16, v_num=0]
Epoch 31:  51%|█████     | 49/97 [00:08<00:08,  5.84it/s, loss=3.16, v_num=0]
Epoch 31:  51%|█████     | 49/97 [00:08<00:08,  5.80it/s, loss=3.15, v_num=0]
Epoch 31:  52%|█████▏    | 50/97 [00:08<00:08,  5.86it/s, loss=3.15, v_num=0]
Epoch 31:  52%|█████▏    | 50/97 [00:08<00:08,  5.82it/s, loss=3.16, v_num=0]
Epoch 31:  53%|█████▎    | 51/97 [00:08<00:07,  5.88it/s, loss=3.16, v_num=0]
Epoch 31:  53%|█████▎    | 51/97 [00:08<00:07,  5.84it/s, loss=3.17, v_num=0]
Epoch 31:  54%|█████▎    | 52/97 [00:08<00:07,  5.90it/s, loss=3.17, v_num=0]
Epoch 31:  54%|█████▎    | 52/97 [00:08<00:07,  5.87it/s, loss=3.16, v_num=0]
Epoch 31:  55%|█████▍    | 53/97 [00:08<00:07,  5.92it/s, loss=3.16, v_num=0]
Epoch 31:  55%|█████▍    | 53/97 [00:09<00:07,  5.89it/s, loss=3.16, v_num=0]
Epoch 31:  56%|█████▌    | 54/97 [00:09<00:07,  5.94it/s, loss=3.16, v_num=0]
Epoch 31:  56%|█████▌    | 54/97 [00:09<00:07,  5.91it/s, loss=3.16, v_num=0]
Epoch 31:  57%|█████▋    | 55/97 [00:09<00:07,  5.96it/s, loss=3.16, v_num=0]
Epoch 31:  57%|█████▋    | 55/97 [00:09<00:07,  5.92it/s, loss=3.16, v_num=0]
Epoch 31:  58%|█████▊    | 56/97 [00:09<00:06,  5.98it/s, loss=3.16, v_num=0]
Epoch 31:  58%|█████▊    | 56/97 [00:09<00:06,  5.94it/s, loss=3.16, v_num=0]
Epoch 31:  59%|█████▉    | 57/97 [00:09<00:06,  6.00it/s, loss=3.16, v_num=0]
Epoch 31:  59%|█████▉    | 57/97 [00:09<00:06,  5.96it/s, loss=3.16, v_num=0]
Epoch 31:  60%|█████▉    | 58/97 [00:09<00:06,  6.02it/s, loss=3.16, v_num=0]
Epoch 31:  60%|█████▉    | 58/97 [00:09<00:06,  5.98it/s, loss=3.16, v_num=0]
Epoch 31:  61%|██████    | 59/97 [00:09<00:06,  6.03it/s, loss=3.16, v_num=0]
Epoch 31:  61%|██████    | 59/97 [00:09<00:06,  6.00it/s, loss=3.16, v_num=0]
Epoch 31:  62%|██████▏   | 60/97 [00:09<00:06,  6.05it/s, loss=3.16, v_num=0]
Epoch 31:  62%|██████▏   | 60/97 [00:09<00:06,  6.01it/s, loss=3.15, v_num=0]
Epoch 31:  63%|██████▎   | 61/97 [00:10<00:05,  6.06it/s, loss=3.15, v_num=0]
Epoch 31:  63%|██████▎   | 61/97 [00:10<00:05,  6.03it/s, loss=3.16, v_num=0]
Epoch 31:  64%|██████▍   | 62/97 [00:10<00:05,  6.08it/s, loss=3.16, v_num=0]
Epoch 31:  64%|██████▍   | 62/97 [00:10<00:05,  6.04it/s, loss=3.15, v_num=0]
Epoch 31:  65%|██████▍   | 63/97 [00:10<00:05,  6.09it/s, loss=3.15, v_num=0]
Epoch 31:  65%|██████▍   | 63/97 [00:10<00:05,  6.06it/s, loss=3.16, v_num=0]
Epoch 31:  66%|██████▌   | 64/97 [00:10<00:05,  6.11it/s, loss=3.16, v_num=0]
Epoch 31:  66%|██████▌   | 64/97 [00:10<00:05,  6.07it/s, loss=3.15, v_num=0]
Epoch 31:  67%|██████▋   | 65/97 [00:10<00:05,  6.12it/s, loss=3.15, v_num=0]
Epoch 31:  67%|██████▋   | 65/97 [00:10<00:05,  6.09it/s, loss=3.15, v_num=0]
Epoch 31:  68%|██████▊   | 66/97 [00:10<00:05,  6.14it/s, loss=3.15, v_num=0]
Epoch 31:  68%|██████▊   | 66/97 [00:10<00:05,  6.10it/s, loss=3.14, v_num=0]
Epoch 31:  69%|██████▉   | 67/97 [00:10<00:04,  6.15it/s, loss=3.14, v_num=0]
Epoch 31:  69%|██████▉   | 67/97 [00:10<00:04,  6.12it/s, loss=3.14, v_num=0]
Epoch 31:  70%|███████   | 68/97 [00:11<00:04,  6.16it/s, loss=3.14, v_num=0]
Epoch 31:  70%|███████   | 68/97 [00:11<00:04,  6.13it/s, loss=3.15, v_num=0]
Epoch 31:  71%|███████   | 69/97 [00:11<00:04,  6.18it/s, loss=3.15, v_num=0]
Epoch 31:  71%|███████   | 69/97 [00:11<00:04,  6.15it/s, loss=3.15, v_num=0]
Epoch 31:  72%|███████▏  | 70/97 [00:11<00:04,  6.19it/s, loss=3.15, v_num=0]
Epoch 31:  72%|███████▏  | 70/97 [00:11<00:04,  6.16it/s, loss=3.14, v_num=0]
Epoch 31:  73%|███████▎  | 71/97 [00:11<00:04,  6.20it/s, loss=3.14, v_num=0]
Epoch 31:  73%|███████▎  | 71/97 [00:11<00:04,  6.17it/s, loss=3.14, v_num=0]
Epoch 31:  74%|███████▍  | 72/97 [00:11<00:04,  6.22it/s, loss=3.14, v_num=0]
Epoch 31:  74%|███████▍  | 72/97 [00:11<00:04,  6.19it/s, loss=3.14, v_num=0]
Epoch 31:  75%|███████▌  | 73/97 [00:11<00:03,  6.23it/s, loss=3.14, v_num=0]
Epoch 31:  75%|███████▌  | 73/97 [00:11<00:03,  6.20it/s, loss=3.14, v_num=0]
Epoch 31:  76%|███████▋  | 74/97 [00:11<00:03,  6.24it/s, loss=3.14, v_num=0]
Epoch 31:  76%|███████▋  | 74/97 [00:11<00:03,  6.21it/s, loss=3.14, v_num=0]
Epoch 31:  77%|███████▋  | 75/97 [00:11<00:03,  6.25it/s, loss=3.14, v_num=0]
Epoch 31:  77%|███████▋  | 75/97 [00:12<00:03,  6.22it/s, loss=3.14, v_num=0]
Epoch 31:  78%|███████▊  | 76/97 [00:12<00:03,  6.27it/s, loss=3.14, v_num=0]
Epoch 31:  78%|███████▊  | 76/97 [00:12<00:03,  6.24it/s, loss=3.13, v_num=0]
Epoch 31:  79%|███████▉  | 77/97 [00:12<00:03,  6.28it/s, loss=3.13, v_num=0]
Epoch 31:  79%|███████▉  | 77/97 [00:12<00:03,  6.25it/s, loss=3.13, v_num=0]
Epoch 31:  80%|████████  | 78/97 [00:12<00:03,  6.29it/s, loss=3.13, v_num=0]
Epoch 31:  80%|████████  | 78/97 [00:12<00:03,  6.26it/s, loss=3.13, v_num=0]
Epoch 31:  81%|████████▏ | 79/97 [00:12<00:02,  6.30it/s, loss=3.13, v_num=0]
Epoch 31:  81%|████████▏ | 79/97 [00:12<00:02,  6.27it/s, loss=3.14, v_num=0]
Epoch 31:  82%|████████▏ | 80/97 [00:12<00:02,  6.31it/s, loss=3.14, v_num=0]
Epoch 31:  82%|████████▏ | 80/97 [00:12<00:02,  6.28it/s, loss=3.14, v_num=0]
Epoch 31:  84%|████████▎ | 81/97 [00:12<00:02,  6.32it/s, loss=3.14, v_num=0]
Epoch 31:  84%|████████▎ | 81/97 [00:12<00:02,  6.29it/s, loss=3.13, v_num=0]
Epoch 31:  85%|████████▍ | 82/97 [00:12<00:02,  6.32it/s, loss=3.13, v_num=0]
Epoch 31:  85%|████████▍ | 82/97 [00:13<00:02,  6.29it/s, loss=3.13, v_num=0]
Epoch 31:  86%|████████▌ | 83/97 [00:13<00:02,  6.33it/s, loss=3.13, v_num=0]
Epoch 31:  86%|████████▌ | 83/97 [00:13<00:02,  6.30it/s, loss=3.14, v_num=0]
Epoch 31:  87%|████████▋ | 84/97 [00:13<00:02,  6.34it/s, loss=3.14, v_num=0]
Epoch 31:  87%|████████▋ | 84/97 [00:13<00:02,  6.31it/s, loss=3.14, v_num=0]
Epoch 31:  88%|████████▊ | 85/97 [00:13<00:01,  6.35it/s, loss=3.14, v_num=0]
Epoch 31:  88%|████████▊ | 85/97 [00:13<00:01,  6.32it/s, loss=3.14, v_num=0]
Epoch 31:  89%|████████▊ | 86/97 [00:13<00:01,  6.36it/s, loss=3.14, v_num=0]
Epoch 31:  89%|████████▊ | 86/97 [00:13<00:01,  6.33it/s, loss=3.15, v_num=0]
Epoch 31:  90%|████████▉ | 87/97 [00:13<00:01,  6.37it/s, loss=3.15, v_num=0]
Epoch 31:  90%|████████▉ | 87/97 [00:13<00:01,  6.34it/s, loss=3.16, v_num=0]
Epoch 31:  91%|█████████ | 88/97 [00:13<00:01,  6.38it/s, loss=3.16, v_num=0]
Epoch 31:  91%|█████████ | 88/97 [00:13<00:01,  6.35it/s, loss=3.16, v_num=0]
Epoch 31:  92%|█████████▏| 89/97 [00:13<00:01,  6.39it/s, loss=3.16, v_num=0]
Epoch 31:  92%|█████████▏| 89/97 [00:13<00:01,  6.36it/s, loss=3.15, v_num=0]
Epoch 31:  93%|█████████▎| 90/97 [00:14<00:01,  6.40it/s, loss=3.15, v_num=0]
Epoch 31:  93%|█████████▎| 90/97 [00:14<00:01,  6.37it/s, loss=3.16, v_num=0]
Epoch 31:  94%|█████████▍| 91/97 [00:14<00:00,  6.41it/s, loss=3.16, v_num=0]
Epoch 31:  94%|█████████▍| 91/97 [00:14<00:00,  6.38it/s, loss=3.15, v_num=0]
Epoch 31:  95%|█████████▍| 92/97 [00:14<00:00,  6.41it/s, loss=3.15, v_num=0]
Epoch 31:  95%|█████████▍| 92/97 [00:14<00:00,  6.39it/s, loss=3.15, v_num=0]
Epoch 31:  96%|█████████▌| 93/97 [00:14<00:00,  6.42it/s, loss=3.15, v_num=0]
Epoch 31:  96%|█████████▌| 93/97 [00:14<00:00,  6.40it/s, loss=3.15, v_num=0]
Epoch 31:  97%|█████████▋| 94/97 [00:14<00:00,  6.43it/s, loss=3.15, v_num=0]
Epoch 31:  97%|█████████▋| 94/97 [00:14<00:00,  6.41it/s, loss=3.14, v_num=0]
Epoch 31:  98%|█████████▊| 95/97 [00:14<00:00,  6.44it/s, loss=3.14, v_num=0]
Epoch 31:  98%|█████████▊| 95/97 [00:14<00:00,  6.41it/s, loss=3.14, v_num=0]
Epoch 31:  99%|█████████▉| 96/97 [00:14<00:00,  6.45it/s, loss=3.14, v_num=0]
Epoch 31:  99%|█████████▉| 96/97 [00:14<00:00,  6.42it/s, loss=3.14, v_num=0]
Epoch 31: 100%|██████████| 97/97 [00:15<00:00,  6.46it/s, loss=3.14, v_num=0]
Epoch 31: 100%|██████████| 97/97 [00:15<00:00,  6.43it/s, loss=3.14, v_num=0]
Epoch 31: 100%|██████████| 97/97 [00:15<00:00,  6.43it/s, loss=3.14, v_num=0]
Epoch 31:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.14, v_num=0]
Epoch 32:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.14, v_num=0]
Epoch 32:   1%|          | 1/97 [00:01<02:21,  1.47s/it, loss=3.14, v_num=0]
Epoch 32:   1%|          | 1/97 [00:01<02:26,  1.53s/it, loss=3.15, v_num=0]
Epoch 32:   2%|▏         | 2/97 [00:01<01:16,  1.24it/s, loss=3.15, v_num=0]
Epoch 32:   2%|▏         | 2/97 [00:01<01:19,  1.20it/s, loss=3.14, v_num=0]
Epoch 32:   3%|▎         | 3/97 [00:01<00:55,  1.71it/s, loss=3.14, v_num=0]
Epoch 32:   3%|▎         | 3/97 [00:01<00:56,  1.66it/s, loss=3.14, v_num=0]
Epoch 32:   4%|▍         | 4/97 [00:01<00:43,  2.11it/s, loss=3.14, v_num=0]
Epoch 32:   4%|▍         | 4/97 [00:01<00:45,  2.05it/s, loss=3.13, v_num=0]
Epoch 32:   5%|▌         | 5/97 [00:02<00:37,  2.46it/s, loss=3.13, v_num=0]
Epoch 32:   5%|▌         | 5/97 [00:02<00:38,  2.40it/s, loss=3.13, v_num=0]
Epoch 32:   6%|▌         | 6/97 [00:02<00:32,  2.77it/s, loss=3.13, v_num=0]
Epoch 32:   6%|▌         | 6/97 [00:02<00:33,  2.70it/s, loss=3.12, v_num=0]
Epoch 32:   7%|▋         | 7/97 [00:02<00:29,  3.03it/s, loss=3.12, v_num=0]
Epoch 32:   7%|▋         | 7/97 [00:02<00:30,  2.96it/s, loss=3.12, v_num=0]
Epoch 32:   8%|▊         | 8/97 [00:02<00:27,  3.27it/s, loss=3.12, v_num=0]
Epoch 32:   8%|▊         | 8/97 [00:02<00:27,  3.20it/s, loss=3.13, v_num=0]
Epoch 32:   9%|▉         | 9/97 [00:02<00:25,  3.48it/s, loss=3.13, v_num=0]
Epoch 32:   9%|▉         | 9/97 [00:02<00:25,  3.41it/s, loss=3.12, v_num=0]
Epoch 32:  10%|█         | 10/97 [00:02<00:23,  3.66it/s, loss=3.12, v_num=0]
Epoch 32:  10%|█         | 10/97 [00:02<00:24,  3.59it/s, loss=3.11, v_num=0]
Epoch 32:  11%|█▏        | 11/97 [00:02<00:22,  3.82it/s, loss=3.11, v_num=0]
Epoch 32:  11%|█▏        | 11/97 [00:02<00:22,  3.75it/s, loss=3.11, v_num=0]
Epoch 32:  12%|█▏        | 12/97 [00:03<00:21,  3.96it/s, loss=3.11, v_num=0]
Epoch 32:  12%|█▏        | 12/97 [00:03<00:21,  3.89it/s, loss=3.1, v_num=0]
Epoch 32:  13%|█▎        | 13/97 [00:03<00:20,  4.08it/s, loss=3.1, v_num=0]
Epoch 32:  13%|█▎        | 13/97 [00:03<00:20,  4.02it/s, loss=3.11, v_num=0]
Epoch 32:  14%|█▍        | 14/97 [00:03<00:19,  4.20it/s, loss=3.11, v_num=0]
Epoch 32:  14%|█▍        | 14/97 [00:03<00:20,  4.14it/s, loss=3.11, v_num=0]
Epoch 32:  15%|█▌        | 15/97 [00:03<00:19,  4.31it/s, loss=3.11, v_num=0]
Epoch 32:  15%|█▌        | 15/97 [00:03<00:19,  4.25it/s, loss=3.11, v_num=0]
Epoch 32:  16%|█▋        | 16/97 [00:03<00:18,  4.42it/s, loss=3.11, v_num=0]
Epoch 32:  16%|█▋        | 16/97 [00:03<00:18,  4.35it/s, loss=3.11, v_num=0]
Epoch 32:  18%|█▊        | 17/97 [00:03<00:17,  4.51it/s, loss=3.11, v_num=0]
Epoch 32:  18%|█▊        | 17/97 [00:03<00:17,  4.45it/s, loss=3.12, v_num=0]
Epoch 32:  19%|█▊        | 18/97 [00:03<00:17,  4.59it/s, loss=3.12, v_num=0]
Epoch 32:  19%|█▊        | 18/97 [00:03<00:17,  4.53it/s, loss=3.12, v_num=0]
Epoch 32:  20%|█▉        | 19/97 [00:04<00:16,  4.67it/s, loss=3.12, v_num=0]
Epoch 32:  20%|█▉        | 19/97 [00:04<00:16,  4.62it/s, loss=3.11, v_num=0]
Epoch 32:  21%|██        | 20/97 [00:04<00:16,  4.75it/s, loss=3.11, v_num=0]
Epoch 32:  21%|██        | 20/97 [00:04<00:16,  4.70it/s, loss=3.11, v_num=0]
Epoch 32:  22%|██▏       | 21/97 [00:04<00:15,  4.84it/s, loss=3.11, v_num=0]
Epoch 32:  22%|██▏       | 21/97 [00:04<00:15,  4.78it/s, loss=3.1, v_num=0]
Epoch 32:  23%|██▎       | 22/97 [00:04<00:15,  4.91it/s, loss=3.1, v_num=0]
Epoch 32:  23%|██▎       | 22/97 [00:04<00:15,  4.85it/s, loss=3.11, v_num=0]
Epoch 32:  24%|██▎       | 23/97 [00:04<00:14,  4.97it/s, loss=3.11, v_num=0]
Epoch 32:  24%|██▎       | 23/97 [00:04<00:15,  4.91it/s, loss=3.1, v_num=0]
Epoch 32:  25%|██▍       | 24/97 [00:04<00:14,  5.03it/s, loss=3.1, v_num=0]
Epoch 32:  25%|██▍       | 24/97 [00:04<00:14,  4.97it/s, loss=3.11, v_num=0]
Epoch 32:  26%|██▌       | 25/97 [00:04<00:14,  5.09it/s, loss=3.11, v_num=0]
Epoch 32:  26%|██▌       | 25/97 [00:04<00:14,  5.03it/s, loss=3.11, v_num=0]
Epoch 32:  27%|██▋       | 26/97 [00:05<00:13,  5.14it/s, loss=3.11, v_num=0]
Epoch 32:  27%|██▋       | 26/97 [00:05<00:13,  5.08it/s, loss=3.12, v_num=0]
Epoch 32:  28%|██▊       | 27/97 [00:05<00:13,  5.19it/s, loss=3.12, v_num=0]
Epoch 32:  28%|██▊       | 27/97 [00:05<00:13,  5.14it/s, loss=3.12, v_num=0]
Epoch 32:  29%|██▉       | 28/97 [00:05<00:13,  5.25it/s, loss=3.12, v_num=0]
Epoch 32:  29%|██▉       | 28/97 [00:05<00:13,  5.19it/s, loss=3.11, v_num=0]
Epoch 32:  30%|██▉       | 29/97 [00:05<00:12,  5.30it/s, loss=3.11, v_num=0]
Epoch 32:  30%|██▉       | 29/97 [00:05<00:12,  5.24it/s, loss=3.11, v_num=0]
Epoch 32:  31%|███       | 30/97 [00:05<00:12,  5.34it/s, loss=3.11, v_num=0]
Epoch 32:  31%|███       | 30/97 [00:05<00:12,  5.29it/s, loss=3.11, v_num=0]
Epoch 32:  32%|███▏      | 31/97 [00:05<00:12,  5.39it/s, loss=3.11, v_num=0]
Epoch 32:  32%|███▏      | 31/97 [00:05<00:12,  5.34it/s, loss=3.11, v_num=0]
Epoch 32:  33%|███▎      | 32/97 [00:05<00:11,  5.43it/s, loss=3.11, v_num=0]
Epoch 32:  33%|███▎      | 32/97 [00:05<00:12,  5.38it/s, loss=3.11, v_num=0]
Epoch 32:  34%|███▍      | 33/97 [00:06<00:11,  5.47it/s, loss=3.11, v_num=0]
Epoch 32:  34%|███▍      | 33/97 [00:06<00:11,  5.42it/s, loss=3.11, v_num=0]
Epoch 32:  35%|███▌      | 34/97 [00:06<00:11,  5.51it/s, loss=3.11, v_num=0]
Epoch 32:  35%|███▌      | 34/97 [00:06<00:11,  5.46it/s, loss=3.1, v_num=0]
Epoch 32:  36%|███▌      | 35/97 [00:06<00:11,  5.54it/s, loss=3.1, v_num=0]
Epoch 32:  36%|███▌      | 35/97 [00:06<00:11,  5.50it/s, loss=3.1, v_num=0]
Epoch 32:  37%|███▋      | 36/97 [00:06<00:10,  5.57it/s, loss=3.1, v_num=0]
Epoch 32:  37%|███▋      | 36/97 [00:06<00:11,  5.53it/s, loss=3.1, v_num=0]
Epoch 32:  38%|███▊      | 37/97 [00:06<00:10,  5.61it/s, loss=3.1, v_num=0]
Epoch 32:  38%|███▊      | 37/97 [00:06<00:10,  5.56it/s, loss=3.1, v_num=0]
Epoch 32:  39%|███▉      | 38/97 [00:06<00:10,  5.64it/s, loss=3.1, v_num=0]
Epoch 32:  39%|███▉      | 38/97 [00:06<00:10,  5.59it/s, loss=3.1, v_num=0]
Epoch 32:  40%|████      | 39/97 [00:06<00:10,  5.67it/s, loss=3.1, v_num=0]
Epoch 32:  40%|████      | 39/97 [00:06<00:10,  5.62it/s, loss=3.11, v_num=0]
Epoch 32:  41%|████      | 40/97 [00:07<00:09,  5.70it/s, loss=3.11, v_num=0]
Epoch 32:  41%|████      | 40/97 [00:07<00:10,  5.66it/s, loss=3.11, v_num=0]
Epoch 32:  42%|████▏     | 41/97 [00:07<00:09,  5.73it/s, loss=3.11, v_num=0]
Epoch 32:  42%|████▏     | 41/97 [00:07<00:09,  5.69it/s, loss=3.1, v_num=0]
Epoch 32:  43%|████▎     | 42/97 [00:07<00:09,  5.76it/s, loss=3.1, v_num=0]
Epoch 32:  43%|████▎     | 42/97 [00:07<00:09,  5.72it/s, loss=3.1, v_num=0]
Epoch 32:  44%|████▍     | 43/97 [00:07<00:09,  5.79it/s, loss=3.1, v_num=0]
Epoch 32:  44%|████▍     | 43/97 [00:07<00:09,  5.74it/s, loss=3.1, v_num=0]
Epoch 32:  45%|████▌     | 44/97 [00:07<00:09,  5.81it/s, loss=3.1, v_num=0]
Epoch 32:  45%|████▌     | 44/97 [00:07<00:09,  5.77it/s, loss=3.1, v_num=0]
Epoch 32:  46%|████▋     | 45/97 [00:07<00:08,  5.84it/s, loss=3.1, v_num=0]
Epoch 32:  46%|████▋     | 45/97 [00:07<00:08,  5.79it/s, loss=3.11, v_num=0]
Epoch 32:  47%|████▋     | 46/97 [00:07<00:08,  5.86it/s, loss=3.11, v_num=0]
Epoch 32:  47%|████▋     | 46/97 [00:07<00:08,  5.82it/s, loss=3.11, v_num=0]
Epoch 32:  48%|████▊     | 47/97 [00:07<00:08,  5.89it/s, loss=3.11, v_num=0]
Epoch 32:  48%|████▊     | 47/97 [00:08<00:08,  5.84it/s, loss=3.11, v_num=0]
Epoch 32:  49%|████▉     | 48/97 [00:08<00:08,  5.91it/s, loss=3.11, v_num=0]
Epoch 32:  49%|████▉     | 48/97 [00:08<00:08,  5.87it/s, loss=3.12, v_num=0]
Epoch 32:  51%|█████     | 49/97 [00:08<00:08,  5.93it/s, loss=3.12, v_num=0]
Epoch 32:  51%|█████     | 49/97 [00:08<00:08,  5.89it/s, loss=3.12, v_num=0]
Epoch 32:  52%|█████▏    | 50/97 [00:08<00:07,  5.96it/s, loss=3.12, v_num=0]
Epoch 32:  52%|█████▏    | 50/97 [00:08<00:07,  5.91it/s, loss=3.12, v_num=0]
Epoch 32:  53%|█████▎    | 51/97 [00:08<00:07,  5.98it/s, loss=3.12, v_num=0]
Epoch 32:  53%|█████▎    | 51/97 [00:08<00:07,  5.94it/s, loss=3.12, v_num=0]
Epoch 32:  54%|█████▎    | 52/97 [00:08<00:07,  6.00it/s, loss=3.12, v_num=0]
Epoch 32:  54%|█████▎    | 52/97 [00:08<00:07,  5.96it/s, loss=3.12, v_num=0]
Epoch 32:  55%|█████▍    | 53/97 [00:08<00:07,  6.02it/s, loss=3.12, v_num=0]
Epoch 32:  55%|█████▍    | 53/97 [00:08<00:07,  5.98it/s, loss=3.12, v_num=0]
Epoch 32:  56%|█████▌    | 54/97 [00:08<00:07,  6.04it/s, loss=3.12, v_num=0]
Epoch 32:  56%|█████▌    | 54/97 [00:09<00:07,  6.00it/s, loss=3.12, v_num=0]
Epoch 32:  57%|█████▋    | 55/97 [00:09<00:06,  6.06it/s, loss=3.12, v_num=0]
Epoch 32:  57%|█████▋    | 55/97 [00:09<00:06,  6.02it/s, loss=3.12, v_num=0]
Epoch 32:  58%|█████▊    | 56/97 [00:09<00:06,  6.07it/s, loss=3.12, v_num=0]
Epoch 32:  58%|█████▊    | 56/97 [00:09<00:06,  6.04it/s, loss=3.12, v_num=0]
Epoch 32:  59%|█████▉    | 57/97 [00:09<00:06,  6.09it/s, loss=3.12, v_num=0]
Epoch 32:  59%|█████▉    | 57/97 [00:09<00:06,  6.05it/s, loss=3.12, v_num=0]
Epoch 32:  60%|█████▉    | 58/97 [00:09<00:06,  6.10it/s, loss=3.12, v_num=0]
Epoch 32:  60%|█████▉    | 58/97 [00:09<00:06,  6.07it/s, loss=3.12, v_num=0]
Epoch 32:  61%|██████    | 59/97 [00:09<00:06,  6.12it/s, loss=3.12, v_num=0]
Epoch 32:  61%|██████    | 59/97 [00:09<00:06,  6.09it/s, loss=3.12, v_num=0]
Epoch 32:  62%|██████▏   | 60/97 [00:09<00:06,  6.14it/s, loss=3.12, v_num=0]
Epoch 32:  62%|██████▏   | 60/97 [00:09<00:06,  6.10it/s, loss=3.13, v_num=0]
Epoch 32:  63%|██████▎   | 61/97 [00:09<00:05,  6.15it/s, loss=3.13, v_num=0]
Epoch 32:  63%|██████▎   | 61/97 [00:09<00:05,  6.12it/s, loss=3.13, v_num=0]
Epoch 32:  64%|██████▍   | 62/97 [00:10<00:05,  6.17it/s, loss=3.13, v_num=0]
Epoch 32:  64%|██████▍   | 62/97 [00:10<00:05,  6.13it/s, loss=3.13, v_num=0]
Epoch 32:  65%|██████▍   | 63/97 [00:10<00:05,  6.18it/s, loss=3.13, v_num=0]
Epoch 32:  65%|██████▍   | 63/97 [00:10<00:05,  6.15it/s, loss=3.13, v_num=0]
Epoch 32:  66%|██████▌   | 64/97 [00:10<00:05,  6.20it/s, loss=3.13, v_num=0]
Epoch 32:  66%|██████▌   | 64/97 [00:10<00:05,  6.16it/s, loss=3.13, v_num=0]
Epoch 32:  67%|██████▋   | 65/97 [00:10<00:05,  6.21it/s, loss=3.13, v_num=0]
Epoch 32:  67%|██████▋   | 65/97 [00:10<00:05,  6.18it/s, loss=3.12, v_num=0]
Epoch 32:  68%|██████▊   | 66/97 [00:10<00:04,  6.23it/s, loss=3.12, v_num=0]
Epoch 32:  68%|██████▊   | 66/97 [00:10<00:05,  6.19it/s, loss=3.12, v_num=0]
Epoch 32:  69%|██████▉   | 67/97 [00:10<00:04,  6.24it/s, loss=3.12, v_num=0]
Epoch 32:  69%|██████▉   | 67/97 [00:10<00:04,  6.21it/s, loss=3.11, v_num=0]
Epoch 32:  70%|███████   | 68/97 [00:10<00:04,  6.25it/s, loss=3.11, v_num=0]
Epoch 32:  70%|███████   | 68/97 [00:10<00:04,  6.22it/s, loss=3.12, v_num=0]
Epoch 32:  71%|███████   | 69/97 [00:11<00:04,  6.27it/s, loss=3.12, v_num=0]
Epoch 32:  71%|███████   | 69/97 [00:11<00:04,  6.23it/s, loss=3.12, v_num=0]
Epoch 32:  72%|███████▏  | 70/97 [00:11<00:04,  6.28it/s, loss=3.12, v_num=0]
Epoch 32:  72%|███████▏  | 70/97 [00:11<00:04,  6.24it/s, loss=3.12, v_num=0]
Epoch 32:  73%|███████▎  | 71/97 [00:11<00:04,  6.29it/s, loss=3.12, v_num=0]
Epoch 32:  73%|███████▎  | 71/97 [00:11<00:04,  6.26it/s, loss=3.12, v_num=0]
Epoch 32:  74%|███████▍  | 72/97 [00:11<00:03,  6.30it/s, loss=3.12, v_num=0]
Epoch 32:  74%|███████▍  | 72/97 [00:11<00:03,  6.27it/s, loss=3.13, v_num=0]
Epoch 32:  75%|███████▌  | 73/97 [00:11<00:03,  6.31it/s, loss=3.13, v_num=0]
Epoch 32:  75%|███████▌  | 73/97 [00:11<00:03,  6.28it/s, loss=3.15, v_num=0]
Epoch 32:  76%|███████▋  | 74/97 [00:11<00:03,  6.32it/s, loss=3.15, v_num=0]
Epoch 32:  76%|███████▋  | 74/97 [00:11<00:03,  6.29it/s, loss=3.15, v_num=0]
Epoch 32:  77%|███████▋  | 75/97 [00:11<00:03,  6.33it/s, loss=3.15, v_num=0]
Epoch 32:  77%|███████▋  | 75/97 [00:11<00:03,  6.30it/s, loss=3.15, v_num=0]
Epoch 32:  78%|███████▊  | 76/97 [00:11<00:03,  6.34it/s, loss=3.15, v_num=0]
Epoch 32:  78%|███████▊  | 76/97 [00:12<00:03,  6.31it/s, loss=3.15, v_num=0]
Epoch 32:  79%|███████▉  | 77/97 [00:12<00:03,  6.35it/s, loss=3.15, v_num=0]
Epoch 32:  79%|███████▉  | 77/97 [00:12<00:03,  6.32it/s, loss=3.15, v_num=0]
Epoch 32:  80%|████████  | 78/97 [00:12<00:02,  6.36it/s, loss=3.15, v_num=0]
Epoch 32:  80%|████████  | 78/97 [00:12<00:02,  6.34it/s, loss=3.15, v_num=0]
Epoch 32:  81%|████████▏ | 79/97 [00:12<00:02,  6.37it/s, loss=3.15, v_num=0]
Epoch 32:  81%|████████▏ | 79/97 [00:12<00:02,  6.34it/s, loss=3.15, v_num=0]
Epoch 32:  82%|████████▏ | 80/97 [00:12<00:02,  6.38it/s, loss=3.15, v_num=0]
Epoch 32:  82%|████████▏ | 80/97 [00:12<00:02,  6.35it/s, loss=3.14, v_num=0]
Epoch 32:  84%|████████▎ | 81/97 [00:12<00:02,  6.39it/s, loss=3.14, v_num=0]
Epoch 32:  84%|████████▎ | 81/97 [00:12<00:02,  6.36it/s, loss=3.14, v_num=0]
Epoch 32:  85%|████████▍ | 82/97 [00:12<00:02,  6.40it/s, loss=3.14, v_num=0]
Epoch 32:  85%|████████▍ | 82/97 [00:12<00:02,  6.37it/s, loss=3.15, v_num=0]
Epoch 32:  86%|████████▌ | 83/97 [00:12<00:02,  6.41it/s, loss=3.15, v_num=0]
Epoch 32:  86%|████████▌ | 83/97 [00:13<00:02,  6.38it/s, loss=3.15, v_num=0]
Epoch 32:  87%|████████▋ | 84/97 [00:13<00:02,  6.41it/s, loss=3.15, v_num=0]
Epoch 32:  87%|████████▋ | 84/97 [00:13<00:02,  6.39it/s, loss=3.14, v_num=0]
Epoch 32:  88%|████████▊ | 85/97 [00:13<00:01,  6.42it/s, loss=3.14, v_num=0]
Epoch 32:  88%|████████▊ | 85/97 [00:13<00:01,  6.39it/s, loss=3.15, v_num=0]
Epoch 32:  89%|████████▊ | 86/97 [00:13<00:01,  6.43it/s, loss=3.15, v_num=0]
Epoch 32:  89%|████████▊ | 86/97 [00:13<00:01,  6.40it/s, loss=3.15, v_num=0]
Epoch 32:  90%|████████▉ | 87/97 [00:13<00:01,  6.44it/s, loss=3.15, v_num=0]
Epoch 32:  90%|████████▉ | 87/97 [00:13<00:01,  6.41it/s, loss=3.15, v_num=0]
Epoch 32:  91%|█████████ | 88/97 [00:13<00:01,  6.44it/s, loss=3.15, v_num=0]
Epoch 32:  91%|█████████ | 88/97 [00:13<00:01,  6.42it/s, loss=3.15, v_num=0]
Epoch 32:  92%|█████████▏| 89/97 [00:13<00:01,  6.45it/s, loss=3.15, v_num=0]
Epoch 32:  92%|█████████▏| 89/97 [00:13<00:01,  6.42it/s, loss=3.15, v_num=0]
Epoch 32:  93%|█████████▎| 90/97 [00:13<00:01,  6.46it/s, loss=3.15, v_num=0]
Epoch 32:  93%|█████████▎| 90/97 [00:13<00:01,  6.43it/s, loss=3.14, v_num=0]
Epoch 32:  94%|█████████▍| 91/97 [00:14<00:00,  6.47it/s, loss=3.14, v_num=0]
Epoch 32:  94%|█████████▍| 91/97 [00:14<00:00,  6.44it/s, loss=3.15, v_num=0]
Epoch 32:  95%|█████████▍| 92/97 [00:14<00:00,  6.47it/s, loss=3.15, v_num=0]
Epoch 32:  95%|█████████▍| 92/97 [00:14<00:00,  6.45it/s, loss=3.14, v_num=0]
Epoch 32:  96%|█████████▌| 93/97 [00:14<00:00,  6.48it/s, loss=3.14, v_num=0]
Epoch 32:  96%|█████████▌| 93/97 [00:14<00:00,  6.45it/s, loss=3.13, v_num=0]
Epoch 32:  97%|█████████▋| 94/97 [00:14<00:00,  6.49it/s, loss=3.13, v_num=0]
Epoch 32:  97%|█████████▋| 94/97 [00:14<00:00,  6.46it/s, loss=3.12, v_num=0]
Epoch 32:  98%|█████████▊| 95/97 [00:14<00:00,  6.49it/s, loss=3.12, v_num=0]
Epoch 32:  98%|█████████▊| 95/97 [00:14<00:00,  6.47it/s, loss=3.12, v_num=0]
Epoch 32:  99%|█████████▉| 96/97 [00:14<00:00,  6.50it/s, loss=3.12, v_num=0]
Epoch 32:  99%|█████████▉| 96/97 [00:14<00:00,  6.48it/s, loss=3.12, v_num=0]
Epoch 32: 100%|██████████| 97/97 [00:14<00:00,  6.51it/s, loss=3.12, v_num=0]
Epoch 32: 100%|██████████| 97/97 [00:14<00:00,  6.48it/s, loss=3.12, v_num=0]
Epoch 32: 100%|██████████| 97/97 [00:14<00:00,  6.48it/s, loss=3.12, v_num=0]
Epoch 32:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.12, v_num=0]
Epoch 33:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.12, v_num=0]
Epoch 33:   1%|          | 1/97 [00:01<02:27,  1.54s/it, loss=3.12, v_num=0]
Epoch 33:   1%|          | 1/97 [00:01<02:32,  1.59s/it, loss=3.12, v_num=0]
Epoch 33:   2%|▏         | 2/97 [00:01<01:19,  1.19it/s, loss=3.12, v_num=0]
Epoch 33:   2%|▏         | 2/97 [00:01<01:22,  1.15it/s, loss=3.12, v_num=0]
Epoch 33:   3%|▎         | 3/97 [00:01<00:57,  1.65it/s, loss=3.12, v_num=0]
Epoch 33:   3%|▎         | 3/97 [00:01<00:58,  1.60it/s, loss=3.12, v_num=0]
Epoch 33:   4%|▍         | 4/97 [00:01<00:45,  2.04it/s, loss=3.12, v_num=0]
Epoch 33:   4%|▍         | 4/97 [00:02<00:46,  1.98it/s, loss=3.12, v_num=0]
Epoch 33:   5%|▌         | 5/97 [00:02<00:38,  2.38it/s, loss=3.12, v_num=0]
Epoch 33:   5%|▌         | 5/97 [00:02<00:39,  2.32it/s, loss=3.12, v_num=0]
Epoch 33:   6%|▌         | 6/97 [00:02<00:33,  2.68it/s, loss=3.12, v_num=0]
Epoch 33:   6%|▌         | 6/97 [00:02<00:34,  2.61it/s, loss=3.12, v_num=0]
Epoch 33:   7%|▋         | 7/97 [00:02<00:30,  2.93it/s, loss=3.12, v_num=0]
Epoch 33:   7%|▋         | 7/97 [00:02<00:31,  2.86it/s, loss=3.11, v_num=0]
Epoch 33:   8%|▊         | 8/97 [00:02<00:28,  3.15it/s, loss=3.11, v_num=0]
Epoch 33:   8%|▊         | 8/97 [00:02<00:28,  3.09it/s, loss=3.12, v_num=0]
Epoch 33:   9%|▉         | 9/97 [00:02<00:26,  3.35it/s, loss=3.12, v_num=0]
Epoch 33:   9%|▉         | 9/97 [00:02<00:26,  3.28it/s, loss=3.13, v_num=0]
Epoch 33:  10%|█         | 10/97 [00:02<00:24,  3.52it/s, loss=3.13, v_num=0]
Epoch 33:  10%|█         | 10/97 [00:02<00:25,  3.45it/s, loss=3.12, v_num=0]
Epoch 33:  11%|█▏        | 11/97 [00:02<00:23,  3.68it/s, loss=3.12, v_num=0]
Epoch 33:  11%|█▏        | 11/97 [00:03<00:23,  3.61it/s, loss=3.12, v_num=0]
Epoch 33:  12%|█▏        | 12/97 [00:03<00:22,  3.83it/s, loss=3.12, v_num=0]
Epoch 33:  12%|█▏        | 12/97 [00:03<00:22,  3.77it/s, loss=3.12, v_num=0]
Epoch 33:  13%|█▎        | 13/97 [00:03<00:21,  3.97it/s, loss=3.12, v_num=0]
Epoch 33:  13%|█▎        | 13/97 [00:03<00:21,  3.91it/s, loss=3.12, v_num=0]
Epoch 33:  14%|█▍        | 14/97 [00:03<00:20,  4.10it/s, loss=3.12, v_num=0]
Epoch 33:  14%|█▍        | 14/97 [00:03<00:20,  4.03it/s, loss=3.12, v_num=0]
Epoch 33:  15%|█▌        | 15/97 [00:03<00:19,  4.21it/s, loss=3.12, v_num=0]
Epoch 33:  15%|█▌        | 15/97 [00:03<00:19,  4.15it/s, loss=3.12, v_num=0]
Epoch 33:  16%|█▋        | 16/97 [00:03<00:18,  4.32it/s, loss=3.12, v_num=0]
Epoch 33:  16%|█▋        | 16/97 [00:03<00:19,  4.25it/s, loss=3.12, v_num=0]
Epoch 33:  18%|█▊        | 17/97 [00:03<00:18,  4.41it/s, loss=3.12, v_num=0]
Epoch 33:  18%|█▊        | 17/97 [00:03<00:18,  4.35it/s, loss=3.12, v_num=0]
Epoch 33:  19%|█▊        | 18/97 [00:04<00:17,  4.49it/s, loss=3.12, v_num=0]
Epoch 33:  19%|█▊        | 18/97 [00:04<00:17,  4.44it/s, loss=3.12, v_num=0]
Epoch 33:  20%|█▉        | 19/97 [00:04<00:17,  4.59it/s, loss=3.12, v_num=0]
Epoch 33:  20%|█▉        | 19/97 [00:04<00:17,  4.53it/s, loss=3.12, v_num=0]
Epoch 33:  21%|██        | 20/97 [00:04<00:16,  4.66it/s, loss=3.12, v_num=0]
Epoch 33:  21%|██        | 20/97 [00:04<00:16,  4.60it/s, loss=3.12, v_num=0]
Epoch 33:  22%|██▏       | 21/97 [00:04<00:16,  4.73it/s, loss=3.12, v_num=0]
Epoch 33:  22%|██▏       | 21/97 [00:04<00:16,  4.68it/s, loss=3.12, v_num=0]
Epoch 33:  23%|██▎       | 22/97 [00:04<00:15,  4.80it/s, loss=3.12, v_num=0]
Epoch 33:  23%|██▎       | 22/97 [00:04<00:15,  4.75it/s, loss=3.13, v_num=0]
Epoch 33:  24%|██▎       | 23/97 [00:04<00:15,  4.87it/s, loss=3.13, v_num=0]
Epoch 33:  24%|██▎       | 23/97 [00:04<00:15,  4.81it/s, loss=3.11, v_num=0]
Epoch 33:  25%|██▍       | 24/97 [00:04<00:14,  4.93it/s, loss=3.11, v_num=0]
Epoch 33:  25%|██▍       | 24/97 [00:04<00:14,  4.88it/s, loss=3.11, v_num=0]
Epoch 33:  26%|██▌       | 25/97 [00:05<00:14,  4.99it/s, loss=3.11, v_num=0]
Epoch 33:  26%|██▌       | 25/97 [00:05<00:14,  4.94it/s, loss=3.12, v_num=0]
Epoch 33:  27%|██▋       | 26/97 [00:05<00:14,  5.05it/s, loss=3.12, v_num=0]
Epoch 33:  27%|██▋       | 26/97 [00:05<00:14,  5.00it/s, loss=3.11, v_num=0]
Epoch 33:  28%|██▊       | 27/97 [00:05<00:13,  5.10it/s, loss=3.11, v_num=0]
Epoch 33:  28%|██▊       | 27/97 [00:05<00:13,  5.05it/s, loss=3.11, v_num=0]
Epoch 33:  29%|██▉       | 28/97 [00:05<00:13,  5.15it/s, loss=3.11, v_num=0]
Epoch 33:  29%|██▉       | 28/97 [00:05<00:13,  5.10it/s, loss=3.11, v_num=0]
Epoch 33:  30%|██▉       | 29/97 [00:05<00:13,  5.20it/s, loss=3.11, v_num=0]
Epoch 33:  30%|██▉       | 29/97 [00:05<00:13,  5.15it/s, loss=3.11, v_num=0]
Epoch 33:  31%|███       | 30/97 [00:05<00:12,  5.25it/s, loss=3.11, v_num=0]
Epoch 33:  31%|███       | 30/97 [00:05<00:12,  5.20it/s, loss=3.1, v_num=0]
Epoch 33:  32%|███▏      | 31/97 [00:05<00:12,  5.29it/s, loss=3.1, v_num=0]
Epoch 33:  32%|███▏      | 31/97 [00:05<00:12,  5.24it/s, loss=3.1, v_num=0]
Epoch 33:  33%|███▎      | 32/97 [00:05<00:12,  5.33it/s, loss=3.1, v_num=0]
Epoch 33:  33%|███▎      | 32/97 [00:06<00:12,  5.28it/s, loss=3.1, v_num=0]
Epoch 33:  34%|███▍      | 33/97 [00:06<00:11,  5.37it/s, loss=3.1, v_num=0]
Epoch 33:  34%|███▍      | 33/97 [00:06<00:12,  5.33it/s, loss=3.11, v_num=0]
Epoch 33:  35%|███▌      | 34/97 [00:06<00:11,  5.41it/s, loss=3.11, v_num=0]
Epoch 33:  35%|███▌      | 34/97 [00:06<00:11,  5.36it/s, loss=3.1, v_num=0]
Epoch 33:  36%|███▌      | 35/97 [00:06<00:11,  5.44it/s, loss=3.1, v_num=0]
Epoch 33:  36%|███▌      | 35/97 [00:06<00:11,  5.40it/s, loss=3.1, v_num=0]
Epoch 33:  37%|███▋      | 36/97 [00:06<00:11,  5.48it/s, loss=3.1, v_num=0]
Epoch 33:  37%|███▋      | 36/97 [00:06<00:11,  5.44it/s, loss=3.1, v_num=0]
Epoch 33:  38%|███▊      | 37/97 [00:06<00:10,  5.52it/s, loss=3.1, v_num=0]
Epoch 33:  38%|███▊      | 37/97 [00:06<00:10,  5.47it/s, loss=3.1, v_num=0]
Epoch 33:  39%|███▉      | 38/97 [00:06<00:10,  5.55it/s, loss=3.1, v_num=0]
Epoch 33:  39%|███▉      | 38/97 [00:06<00:10,  5.51it/s, loss=3.1, v_num=0]
Epoch 33:  40%|████      | 39/97 [00:06<00:10,  5.58it/s, loss=3.1, v_num=0]
Epoch 33:  40%|████      | 39/97 [00:07<00:10,  5.54it/s, loss=3.1, v_num=0]
Epoch 33:  41%|████      | 40/97 [00:07<00:10,  5.62it/s, loss=3.1, v_num=0]
Epoch 33:  41%|████      | 40/97 [00:07<00:10,  5.57it/s, loss=3.11, v_num=0]
Epoch 33:  42%|████▏     | 41/97 [00:07<00:09,  5.65it/s, loss=3.11, v_num=0]
Epoch 33:  42%|████▏     | 41/97 [00:07<00:09,  5.60it/s, loss=3.1, v_num=0]
Epoch 33:  43%|████▎     | 42/97 [00:07<00:09,  5.67it/s, loss=3.1, v_num=0]
Epoch 33:  43%|████▎     | 42/97 [00:07<00:09,  5.63it/s, loss=3.1, v_num=0]
Epoch 33:  44%|████▍     | 43/97 [00:07<00:09,  5.70it/s, loss=3.1, v_num=0]
Epoch 33:  44%|████▍     | 43/97 [00:07<00:09,  5.65it/s, loss=3.11, v_num=0]
Epoch 33:  45%|████▌     | 44/97 [00:07<00:09,  5.72it/s, loss=3.11, v_num=0]
Epoch 33:  45%|████▌     | 44/97 [00:07<00:09,  5.68it/s, loss=3.11, v_num=0]
Epoch 33:  46%|████▋     | 45/97 [00:07<00:09,  5.75it/s, loss=3.11, v_num=0]
Epoch 33:  46%|████▋     | 45/97 [00:07<00:09,  5.70it/s, loss=3.1, v_num=0]
Epoch 33:  47%|████▋     | 46/97 [00:07<00:08,  5.77it/s, loss=3.1, v_num=0]
Epoch 33:  47%|████▋     | 46/97 [00:08<00:08,  5.73it/s, loss=3.11, v_num=0]
Epoch 33:  48%|████▊     | 47/97 [00:08<00:08,  5.79it/s, loss=3.11, v_num=0]
Epoch 33:  48%|████▊     | 47/97 [00:08<00:08,  5.75it/s, loss=3.11, v_num=0]
Epoch 33:  49%|████▉     | 48/97 [00:08<00:08,  5.81it/s, loss=3.11, v_num=0]
Epoch 33:  49%|████▉     | 48/97 [00:08<00:08,  5.77it/s, loss=3.1, v_num=0]
Epoch 33:  51%|█████     | 49/97 [00:08<00:08,  5.83it/s, loss=3.1, v_num=0]
Epoch 33:  51%|█████     | 49/97 [00:08<00:08,  5.79it/s, loss=3.1, v_num=0]
Epoch 33:  52%|█████▏    | 50/97 [00:08<00:08,  5.85it/s, loss=3.1, v_num=0]
Epoch 33:  52%|█████▏    | 50/97 [00:08<00:08,  5.81it/s, loss=3.09, v_num=0]
Epoch 33:  53%|█████▎    | 51/97 [00:08<00:07,  5.87it/s, loss=3.09, v_num=0]
Epoch 33:  53%|█████▎    | 51/97 [00:08<00:07,  5.83it/s, loss=3.1, v_num=0]
Epoch 33:  54%|█████▎    | 52/97 [00:08<00:07,  5.89it/s, loss=3.1, v_num=0]
Epoch 33:  54%|█████▎    | 52/97 [00:08<00:07,  5.85it/s, loss=3.1, v_num=0]
Epoch 33:  55%|█████▍    | 53/97 [00:08<00:07,  5.91it/s, loss=3.1, v_num=0]
Epoch 33:  55%|█████▍    | 53/97 [00:09<00:07,  5.87it/s, loss=3.09, v_num=0]
Epoch 33:  56%|█████▌    | 54/97 [00:09<00:07,  5.93it/s, loss=3.09, v_num=0]
Epoch 33:  56%|█████▌    | 54/97 [00:09<00:07,  5.89it/s, loss=3.09, v_num=0]
Epoch 33:  57%|█████▋    | 55/97 [00:09<00:07,  5.94it/s, loss=3.09, v_num=0]
Epoch 33:  57%|█████▋    | 55/97 [00:09<00:07,  5.91it/s, loss=3.09, v_num=0]
Epoch 33:  58%|█████▊    | 56/97 [00:09<00:06,  5.96it/s, loss=3.09, v_num=0]
Epoch 33:  58%|█████▊    | 56/97 [00:09<00:06,  5.93it/s, loss=3.1, v_num=0]
Epoch 33:  59%|█████▉    | 57/97 [00:09<00:06,  5.98it/s, loss=3.1, v_num=0]
Epoch 33:  59%|█████▉    | 57/97 [00:09<00:06,  5.95it/s, loss=3.1, v_num=0]
Epoch 33:  60%|█████▉    | 58/97 [00:09<00:06,  6.00it/s, loss=3.1, v_num=0]
Epoch 33:  60%|█████▉    | 58/97 [00:09<00:06,  5.97it/s, loss=3.09, v_num=0]
Epoch 33:  61%|██████    | 59/97 [00:09<00:06,  6.01it/s, loss=3.09, v_num=0]
Epoch 33:  61%|██████    | 59/97 [00:09<00:06,  5.98it/s, loss=3.1, v_num=0]
Epoch 33:  62%|██████▏   | 60/97 [00:09<00:06,  6.03it/s, loss=3.1, v_num=0]
Epoch 33:  62%|██████▏   | 60/97 [00:10<00:06,  6.00it/s, loss=3.1, v_num=0]
Epoch 33:  63%|██████▎   | 61/97 [00:10<00:05,  6.04it/s, loss=3.1, v_num=0]
Epoch 33:  63%|██████▎   | 61/97 [00:10<00:05,  6.01it/s, loss=3.1, v_num=0]
Epoch 33:  64%|██████▍   | 62/97 [00:10<00:05,  6.06it/s, loss=3.1, v_num=0]
Epoch 33:  64%|██████▍   | 62/97 [00:10<00:05,  6.03it/s, loss=3.1, v_num=0]
Epoch 33:  65%|██████▍   | 63/97 [00:10<00:05,  6.08it/s, loss=3.1, v_num=0]
Epoch 33:  65%|██████▍   | 63/97 [00:10<00:05,  6.04it/s, loss=3.1, v_num=0]
Epoch 33:  66%|██████▌   | 64/97 [00:10<00:05,  6.09it/s, loss=3.1, v_num=0]
Epoch 33:  66%|██████▌   | 64/97 [00:10<00:05,  6.06it/s, loss=3.11, v_num=0]
Epoch 33:  67%|██████▋   | 65/97 [00:10<00:05,  6.11it/s, loss=3.11, v_num=0]
Epoch 33:  67%|██████▋   | 65/97 [00:10<00:05,  6.07it/s, loss=3.11, v_num=0]
Epoch 33:  68%|██████▊   | 66/97 [00:10<00:05,  6.12it/s, loss=3.11, v_num=0]
Epoch 33:  68%|██████▊   | 66/97 [00:10<00:05,  6.09it/s, loss=3.11, v_num=0]
Epoch 33:  69%|██████▉   | 67/97 [00:10<00:04,  6.13it/s, loss=3.11, v_num=0]
Epoch 33:  69%|██████▉   | 67/97 [00:10<00:04,  6.10it/s, loss=3.12, v_num=0]
Epoch 33:  70%|███████   | 68/97 [00:11<00:04,  6.15it/s, loss=3.12, v_num=0]
Epoch 33:  70%|███████   | 68/97 [00:11<00:04,  6.12it/s, loss=3.12, v_num=0]
Epoch 33:  71%|███████   | 69/97 [00:11<00:04,  6.16it/s, loss=3.12, v_num=0]
Epoch 33:  71%|███████   | 69/97 [00:11<00:04,  6.13it/s, loss=3.12, v_num=0]
Epoch 33:  72%|███████▏  | 70/97 [00:11<00:04,  6.17it/s, loss=3.12, v_num=0]
Epoch 33:  72%|███████▏  | 70/97 [00:11<00:04,  6.14it/s, loss=3.12, v_num=0]
Epoch 33:  73%|███████▎  | 71/97 [00:11<00:04,  6.19it/s, loss=3.12, v_num=0]
Epoch 33:  73%|███████▎  | 71/97 [00:11<00:04,  6.16it/s, loss=3.12, v_num=0]
Epoch 33:  74%|███████▍  | 72/97 [00:11<00:04,  6.20it/s, loss=3.12, v_num=0]
Epoch 33:  74%|███████▍  | 72/97 [00:11<00:04,  6.17it/s, loss=3.11, v_num=0]
Epoch 33:  75%|███████▌  | 73/97 [00:11<00:03,  6.21it/s, loss=3.11, v_num=0]
Epoch 33:  75%|███████▌  | 73/97 [00:11<00:03,  6.18it/s, loss=3.11, v_num=0]
Epoch 33:  76%|███████▋  | 74/97 [00:11<00:03,  6.22it/s, loss=3.11, v_num=0]
Epoch 33:  76%|███████▋  | 74/97 [00:11<00:03,  6.19it/s, loss=3.11, v_num=0]
Epoch 33:  77%|███████▋  | 75/97 [00:12<00:03,  6.23it/s, loss=3.11, v_num=0]
Epoch 33:  77%|███████▋  | 75/97 [00:12<00:03,  6.20it/s, loss=3.1, v_num=0]
Epoch 33:  78%|███████▊  | 76/97 [00:12<00:03,  6.24it/s, loss=3.1, v_num=0]
Epoch 33:  78%|███████▊  | 76/97 [00:12<00:03,  6.22it/s, loss=3.09, v_num=0]
Epoch 33:  79%|███████▉  | 77/97 [00:12<00:03,  6.26it/s, loss=3.09, v_num=0]
Epoch 33:  79%|███████▉  | 77/97 [00:12<00:03,  6.23it/s, loss=3.1, v_num=0]
Epoch 33:  80%|████████  | 78/97 [00:12<00:03,  6.27it/s, loss=3.1, v_num=0]
Epoch 33:  80%|████████  | 78/97 [00:12<00:03,  6.24it/s, loss=3.11, v_num=0]
Epoch 33:  81%|████████▏ | 79/97 [00:12<00:02,  6.28it/s, loss=3.11, v_num=0]
Epoch 33:  81%|████████▏ | 79/97 [00:12<00:02,  6.25it/s, loss=3.1, v_num=0]
Epoch 33:  82%|████████▏ | 80/97 [00:12<00:02,  6.29it/s, loss=3.1, v_num=0]
Epoch 33:  82%|████████▏ | 80/97 [00:12<00:02,  6.26it/s, loss=3.1, v_num=0]
Epoch 33:  84%|████████▎ | 81/97 [00:12<00:02,  6.30it/s, loss=3.1, v_num=0]
Epoch 33:  84%|████████▎ | 81/97 [00:12<00:02,  6.27it/s, loss=3.09, v_num=0]
Epoch 33:  85%|████████▍ | 82/97 [00:13<00:02,  6.31it/s, loss=3.09, v_num=0]
Epoch 33:  85%|████████▍ | 82/97 [00:13<00:02,  6.28it/s, loss=3.08, v_num=0]
Epoch 33:  86%|████████▌ | 83/97 [00:13<00:02,  6.32it/s, loss=3.08, v_num=0]
Epoch 33:  86%|████████▌ | 83/97 [00:13<00:02,  6.29it/s, loss=3.07, v_num=0]
Epoch 33:  87%|████████▋ | 84/97 [00:13<00:02,  6.32it/s, loss=3.07, v_num=0]
Epoch 33:  87%|████████▋ | 84/97 [00:13<00:02,  6.30it/s, loss=3.07, v_num=0]
Epoch 33:  88%|████████▊ | 85/97 [00:13<00:01,  6.33it/s, loss=3.07, v_num=0]
Epoch 33:  88%|████████▊ | 85/97 [00:13<00:01,  6.31it/s, loss=3.06, v_num=0]
Epoch 33:  89%|████████▊ | 86/97 [00:13<00:01,  6.34it/s, loss=3.06, v_num=0]
Epoch 33:  89%|████████▊ | 86/97 [00:13<00:01,  6.32it/s, loss=3.06, v_num=0]
Epoch 33:  90%|████████▉ | 87/97 [00:13<00:01,  6.35it/s, loss=3.06, v_num=0]
Epoch 33:  90%|████████▉ | 87/97 [00:13<00:01,  6.33it/s, loss=3.05, v_num=0]
Epoch 33:  91%|█████████ | 88/97 [00:13<00:01,  6.36it/s, loss=3.05, v_num=0]
Epoch 33:  91%|█████████ | 88/97 [00:13<00:01,  6.34it/s, loss=3.05, v_num=0]
Epoch 33:  92%|█████████▏| 89/97 [00:13<00:01,  6.37it/s, loss=3.05, v_num=0]
Epoch 33:  92%|█████████▏| 89/97 [00:14<00:01,  6.35it/s, loss=3.05, v_num=0]
Epoch 33:  93%|█████████▎| 90/97 [00:14<00:01,  6.38it/s, loss=3.05, v_num=0]
Epoch 33:  93%|█████████▎| 90/97 [00:14<00:01,  6.36it/s, loss=3.06, v_num=0]
Epoch 33:  94%|█████████▍| 91/97 [00:14<00:00,  6.39it/s, loss=3.06, v_num=0]
Epoch 33:  94%|█████████▍| 91/97 [00:14<00:00,  6.37it/s, loss=3.05, v_num=0]
Epoch 33:  95%|█████████▍| 92/97 [00:14<00:00,  6.40it/s, loss=3.05, v_num=0]
Epoch 33:  95%|█████████▍| 92/97 [00:14<00:00,  6.37it/s, loss=3.05, v_num=0]
Epoch 33:  96%|█████████▌| 93/97 [00:14<00:00,  6.41it/s, loss=3.05, v_num=0]
Epoch 33:  96%|█████████▌| 93/97 [00:14<00:00,  6.38it/s, loss=3.05, v_num=0]
Epoch 33:  97%|█████████▋| 94/97 [00:14<00:00,  6.42it/s, loss=3.05, v_num=0]
Epoch 33:  97%|█████████▋| 94/97 [00:14<00:00,  6.39it/s, loss=3.06, v_num=0]
Epoch 33:  98%|█████████▊| 95/97 [00:14<00:00,  6.43it/s, loss=3.06, v_num=0]
Epoch 33:  98%|█████████▊| 95/97 [00:14<00:00,  6.40it/s, loss=3.06, v_num=0]
Epoch 33:  99%|█████████▉| 96/97 [00:14<00:00,  6.43it/s, loss=3.06, v_num=0]
Epoch 33:  99%|█████████▉| 96/97 [00:14<00:00,  6.41it/s, loss=3.07, v_num=0]
Epoch 33: 100%|██████████| 97/97 [00:15<00:00,  6.44it/s, loss=3.07, v_num=0]
Epoch 33: 100%|██████████| 97/97 [00:15<00:00,  6.42it/s, loss=3.06, v_num=0]
Epoch 33: 100%|██████████| 97/97 [00:15<00:00,  6.42it/s, loss=3.06, v_num=0]
Epoch 33:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.06, v_num=0]
Epoch 34:   0%|          | 0/97 [00:00<?, ?it/s, loss=3.06, v_num=0]
Epoch 34:   1%|          | 1/97 [00:01<02:58,  1.86s/it, loss=3.06, v_num=0]
Epoch 34:   1%|          | 1/97 [00:01<03:03,  1.91s/it, loss=3.06, v_num=0]
Epoch 34:   2%|▏         | 2/97 [00:01<01:34,  1.00it/s, loss=3.06, v_num=0]
Epoch 34:   2%|▏         | 2/97 [00:02<01:37,  1.03s/it, loss=3.06, v_num=0]
Epoch 34:   3%|▎         | 3/97 [00:02<01:07,  1.40it/s, loss=3.06, v_num=0]
Epoch 34:   3%|▎         | 3/97 [00:02<01:08,  1.37it/s, loss=3.06, v_num=0]
Epoch 34:   4%|▍         | 4/97 [00:02<00:53,  1.75it/s, loss=3.06, v_num=0]
Epoch 34:   4%|▍         | 4/97 [00:02<00:54,  1.71it/s, loss=3.06, v_num=0]
Epoch 34:   5%|▌         | 5/97 [00:02<00:44,  2.06it/s, loss=3.06, v_num=0]
Epoch 34:   5%|▌         | 5/97 [00:02<00:45,  2.02it/s, loss=3.06, v_num=0]
Epoch 34:   6%|▌         | 6/97 [00:02<00:38,  2.34it/s, loss=3.06, v_num=0]
Epoch 34:   6%|▌         | 6/97 [00:02<00:39,  2.29it/s, loss=3.06, v_num=0]
Epoch 34:   7%|▋         | 7/97 [00:02<00:34,  2.59it/s, loss=3.06, v_num=0]
Epoch 34:   7%|▋         | 7/97 [00:02<00:35,  2.53it/s, loss=3.06, v_num=0]
Epoch 34:   8%|▊         | 8/97 [00:02<00:31,  2.81it/s, loss=3.06, v_num=0]
Epoch 34:   8%|▊         | 8/97 [00:02<00:32,  2.76it/s, loss=3.07, v_num=0]
Epoch 34:   9%|▉         | 9/97 [00:02<00:29,  3.00it/s, loss=3.07, v_num=0]
Epoch 34:   9%|▉         | 9/97 [00:03<00:29,  2.95it/s, loss=3.07, v_num=0]
Epoch 34:  10%|█         | 10/97 [00:03<00:27,  3.19it/s, loss=3.07, v_num=0]
Epoch 34:  10%|█         | 10/97 [00:03<00:27,  3.13it/s, loss=3.06, v_num=0]
Epoch 34:  11%|█▏        | 11/97 [00:03<00:25,  3.36it/s, loss=3.06, v_num=0]
Epoch 34:  11%|█▏        | 11/97 [00:03<00:26,  3.30it/s, loss=3.07, v_num=0]
Epoch 34:  12%|█▏        | 12/97 [00:03<00:24,  3.51it/s, loss=3.07, v_num=0]
Epoch 34:  12%|█▏        | 12/97 [00:03<00:24,  3.46it/s, loss=3.06, v_num=0]
Epoch 34:  13%|█▎        | 13/97 [00:03<00:22,  3.66it/s, loss=3.06, v_num=0]
Epoch 34:  13%|█▎        | 13/97 [00:03<00:23,  3.60it/s, loss=3.06, v_num=0]
Epoch 34:  14%|█▍        | 14/97 [00:03<00:21,  3.79it/s, loss=3.06, v_num=0]
Epoch 34:  14%|█▍        | 14/97 [00:03<00:22,  3.74it/s, loss=3.06, v_num=0]
Epoch 34:  15%|█▌        | 15/97 [00:03<00:20,  3.92it/s, loss=3.06, v_num=0]
Epoch 34:  15%|█▌        | 15/97 [00:03<00:21,  3.86it/s, loss=3.06, v_num=0]
Epoch 34:  16%|█▋        | 16/97 [00:03<00:20,  4.04it/s, loss=3.06, v_num=0]
Epoch 34:  16%|█▋        | 16/97 [00:04<00:20,  3.98it/s, loss=3.06, v_num=0]
Epoch 34:  18%|█▊        | 17/97 [00:04<00:19,  4.14it/s, loss=3.06, v_num=0]
Epoch 34:  18%|█▊        | 17/97 [00:04<00:19,  4.08it/s, loss=3.06, v_num=0]
Epoch 34:  19%|█▊        | 18/97 [00:04<00:18,  4.23it/s, loss=3.06, v_num=0]
Epoch 34:  19%|█▊        | 18/97 [00:04<00:18,  4.18it/s, loss=3.06, v_num=0]
Epoch 34:  20%|█▉        | 19/97 [00:04<00:18,  4.33it/s, loss=3.06, v_num=0]
Epoch 34:  20%|█▉        | 19/97 [00:04<00:18,  4.27it/s, loss=3.05, v_num=0]
Epoch 34:  21%|██        | 20/97 [00:04<00:17,  4.42it/s, loss=3.05, v_num=0]
Epoch 34:  21%|██        | 20/97 [00:04<00:17,  4.36it/s, loss=3.05, v_num=0]
Epoch 34:  22%|██▏       | 21/97 [00:04<00:16,  4.50it/s, loss=3.05, v_num=0]
Epoch 34:  22%|██▏       | 21/97 [00:04<00:17,  4.45it/s, loss=3.04, v_num=0]
Epoch 34:  23%|██▎       | 22/97 [00:04<00:16,  4.58it/s, loss=3.04, v_num=0]
Epoch 34:  23%|██▎       | 22/97 [00:04<00:16,  4.52it/s, loss=3.04, v_num=0]
Epoch 34:  24%|██▎       | 23/97 [00:04<00:15,  4.65it/s, loss=3.04, v_num=0]
Epoch 34:  24%|██▎       | 23/97 [00:05<00:16,  4.60it/s, loss=3.04, v_num=0]
Epoch 34:  25%|██▍       | 24/97 [00:05<00:15,  4.72it/s, loss=3.04, v_num=0]
Epoch 34:  25%|██▍       | 24/97 [00:05<00:15,  4.67it/s, loss=3.04, v_num=0]
Epoch 34:  26%|██▌       | 25/97 [00:05<00:15,  4.79it/s, loss=3.04, v_num=0]
Epoch 34:  26%|██▌       | 25/97 [00:05<00:15,  4.74it/s, loss=3.04, v_num=0]
Epoch 34:  27%|██▋       | 26/97 [00:05<00:14,  4.85it/s, loss=3.04, v_num=0]
Epoch 34:  27%|██▋       | 26/97 [00:05<00:14,  4.80it/s, loss=3.05, v_num=0]
Epoch 34:  28%|██▊       | 27/97 [00:05<00:14,  4.91it/s, loss=3.05, v_num=0]
Epoch 34:  28%|██▊       | 27/97 [00:05<00:14,  4.86it/s, loss=3.05, v_num=0]
Epoch 34:  29%|██▉       | 28/97 [00:05<00:13,  4.97it/s, loss=3.05, v_num=0]
Epoch 34:  29%|██▉       | 28/97 [00:05<00:14,  4.92it/s, loss=3.05, v_num=0]
Epoch 34:  30%|██▉       | 29/97 [00:05<00:13,  5.02it/s, loss=3.05, v_num=0]
Epoch 34:  30%|██▉       | 29/97 [00:05<00:13,  4.98it/s, loss=3.05, v_num=0]
Epoch 34:  31%|███       | 30/97 [00:05<00:13,  5.07it/s, loss=3.05, v_num=0]
Epoch 34:  31%|███       | 30/97 [00:05<00:13,  5.03it/s, loss=3.05, v_num=0]
Epoch 34:  32%|███▏      | 31/97 [00:06<00:12,  5.13it/s, loss=3.05, v_num=0]
Epoch 34:  32%|███▏      | 31/97 [00:06<00:12,  5.08it/s, loss=3.05, v_num=0]
Epoch 34:  33%|███▎      | 32/97 [00:06<00:12,  5.17it/s, loss=3.05, v_num=0]
Epoch 34:  33%|███▎      | 32/97 [00:06<00:12,  5.13it/s, loss=3.05, v_num=0]
Epoch 34:  34%|███▍      | 33/97 [00:06<00:12,  5.22it/s, loss=3.05, v_num=0]
Epoch 34:  34%|███▍      | 33/97 [00:06<00:12,  5.17it/s, loss=3.06, v_num=0]
Epoch 34:  35%|███▌      | 34/97 [00:06<00:11,  5.26it/s, loss=3.06, v_num=0]
Epoch 34:  35%|███▌      | 34/97 [00:06<00:12,  5.22it/s, loss=3.06, v_num=0]
Epoch 34:  36%|███▌      | 35/97 [00:06<00:11,  5.31it/s, loss=3.06, v_num=0]
Epoch 34:  36%|███▌      | 35/97 [00:06<00:11,  5.26it/s, loss=3.06, v_num=0]
Epoch 34:  37%|███▋      | 36/97 [00:06<00:11,  5.35it/s, loss=3.06, v_num=0]
Epoch 34:  37%|███▋      | 36/97 [00:06<00:11,  5.30it/s, loss=3.06, v_num=0]
Epoch 34:  38%|███▊      | 37/97 [00:06<00:11,  5.38it/s, loss=3.06, v_num=0]
Epoch 34:  38%|███▊      | 37/97 [00:06<00:11,  5.34it/s, loss=3.07, v_num=0]
Epoch 34:  39%|███▉      | 38/97 [00:07<00:10,  5.42it/s, loss=3.07, v_num=0]
Epoch 34:  39%|███▉      | 38/97 [00:07<00:10,  5.38it/s, loss=3.07, v_num=0]
Epoch 34:  40%|████      | 39/97 [00:07<00:10,  5.46it/s, loss=3.07, v_num=0]
Epoch 34:  40%|████      | 39/97 [00:07<00:10,  5.42it/s, loss=3.07, v_num=0]
Epoch 34:  41%|████      | 40/97 [00:07<00:10,  5.49it/s, loss=3.07, v_num=0]
Epoch 34:  41%|████      | 40/97 [00:07<00:10,  5.45it/s, loss=3.07, v_num=0]
Epoch 34:  42%|████▏     | 41/97 [00:07<00:10,  5.52it/s, loss=3.07, v_num=0]
Epoch 34:  42%|████▏     | 41/97 [00:07<00:10,  5.48it/s, loss=3.08, v_num=0]
Epoch 34:  43%|████▎     | 42/97 [00:07<00:09,  5.55it/s, loss=3.08, v_num=0]
Epoch 34:  43%|████▎     | 42/97 [00:07<00:09,  5.51it/s, loss=3.07, v_num=0]
Epoch 34:  44%|████▍     | 43/97 [00:07<00:09,  5.58it/s, loss=3.07, v_num=0]
Epoch 34:  44%|████▍     | 43/97 [00:07<00:09,  5.54it/s, loss=3.07, v_num=0]
Epoch 34:  45%|████▌     | 44/97 [00:07<00:09,  5.62it/s, loss=3.07, v_num=0]
Epoch 34:  45%|████▌     | 44/97 [00:07<00:09,  5.57it/s, loss=3.07, v_num=0]
Epoch 34:  46%|████▋     | 45/97 [00:07<00:09,  5.64it/s, loss=3.07, v_num=0]
Epoch 34:  46%|████▋     | 45/97 [00:08<00:09,  5.60it/s, loss=3.07, v_num=0]
Epoch 34:  47%|████▋     | 46/97 [00:08<00:08,  5.67it/s, loss=3.07, v_num=0]
Epoch 34:  47%|████▋     | 46/97 [00:08<00:09,  5.63it/s, loss=3.07, v_num=0]
Epoch 34:  48%|████▊     | 47/97 [00:08<00:08,  5.70it/s, loss=3.07, v_num=0]
Epoch 34:  48%|████▊     | 47/97 [00:08<00:08,  5.66it/s, loss=3.07, v_num=0]
Epoch 34:  49%|████▉     | 48/97 [00:08<00:08,  5.72it/s, loss=3.07, v_num=0]
Epoch 34:  49%|████▉     | 48/97 [00:08<00:08,  5.68it/s, loss=3.07, v_num=0]
Epoch 34:  51%|█████     | 49/97 [00:08<00:08,  5.74it/s, loss=3.07, v_num=0]
Epoch 34:  51%|█████     | 49/97 [00:08<00:08,  5.71it/s, loss=3.07, v_num=0]
Epoch 34:  52%|█████▏    | 50/97 [00:08<00:08,  5.77it/s, loss=3.07, v_num=0]
Epoch 34:  52%|█████▏    | 50/97 [00:08<00:08,  5.73it/s, loss=3.07, v_num=0]
Epoch 34:  53%|█████▎    | 51/97 [00:08<00:07,  5.79it/s, loss=3.07, v_num=0]
Epoch 34:  53%|█████▎    | 51/97 [00:08<00:07,  5.76it/s, loss=3.08, v_num=0]
Epoch 34:  54%|█████▎    | 52/97 [00:08<00:07,  5.82it/s, loss=3.08, v_num=0]
Epoch 34:  54%|█████▎    | 52/97 [00:08<00:07,  5.78it/s, loss=3.08, v_num=0]
Epoch 34:  55%|█████▍    | 53/97 [00:09<00:07,  5.84it/s, loss=3.08, v_num=0]
Epoch 34:  55%|█████▍    | 53/97 [00:09<00:07,  5.80it/s, loss=3.08, v_num=0]
Epoch 34:  56%|█████▌    | 54/97 [00:09<00:07,  5.86it/s, loss=3.08, v_num=0]
Epoch 34:  56%|█████▌    | 54/97 [00:09<00:07,  5.82it/s, loss=3.09, v_num=0]
Epoch 34:  57%|█████▋    | 55/97 [00:09<00:07,  5.88it/s, loss=3.09, v_num=0]
Epoch 34:  57%|█████▋    | 55/97 [00:09<00:07,  5.85it/s, loss=3.09, v_num=0]
Epoch 34:  58%|█████▊    | 56/97 [00:09<00:06,  5.90it/s, loss=3.09, v_num=0]
Epoch 34:  58%|█████▊    | 56/97 [00:09<00:06,  5.87it/s, loss=3.08, v_num=0]
Epoch 34:  59%|█████▉    | 57/97 [00:09<00:06,  5.92it/s, loss=3.08, v_num=0]
Epoch 34:  59%|█████▉    | 57/97 [00:09<00:06,  5.88it/s, loss=3.08, v_num=0]
Epoch 34:  60%|█████▉    | 58/97 [00:09<00:06,  5.94it/s, loss=3.08, v_num=0]
Epoch 34:  60%|█████▉    | 58/97 [00:09<00:06,  5.90it/s, loss=3.07, v_num=0]
Epoch 34:  61%|██████    | 59/97 [00:09<00:06,  5.95it/s, loss=3.07, v_num=0]
Epoch 34:  61%|██████    | 59/97 [00:09<00:06,  5.92it/s, loss=3.07, v_num=0]
Epoch 34:  62%|██████▏   | 60/97 [00:10<00:06,  5.97it/s, loss=3.07, v_num=0]
Epoch 34:  62%|██████▏   | 60/97 [00:10<00:06,  5.94it/s, loss=3.08, v_num=0]
Epoch 34:  63%|██████▎   | 61/97 [00:10<00:06,  5.99it/s, loss=3.08, v_num=0]
Epoch 34:  63%|██████▎   | 61/97 [00:10<00:06,  5.96it/s, loss=3.07, v_num=0]
Epoch 34:  64%|██████▍   | 62/97 [00:10<00:05,  6.01it/s, loss=3.07, v_num=0]
Epoch 34:  64%|██████▍   | 62/97 [00:10<00:05,  5.98it/s, loss=3.07, v_num=0]
Epoch 34:  65%|██████▍   | 63/97 [00:10<00:05,  6.02it/s, loss=3.07, v_num=0]
Epoch 34:  65%|██████▍   | 63/97 [00:10<00:05,  5.99it/s, loss=3.07, v_num=0]
Epoch 34:  66%|██████▌   | 64/97 [00:10<00:05,  6.04it/s, loss=3.07, v_num=0]
Epoch 34:  66%|██████▌   | 64/97 [00:10<00:05,  6.01it/s, loss=3.07, v_num=0]
Epoch 34:  67%|██████▋   | 65/97 [00:10<00:05,  6.06it/s, loss=3.07, v_num=0]
Epoch 34:  67%|██████▋   | 65/97 [00:10<00:05,  6.03it/s, loss=3.07, v_num=0]
Epoch 34:  68%|██████▊   | 66/97 [00:10<00:05,  6.07it/s, loss=3.07, v_num=0]
Epoch 34:  68%|██████▊   | 66/97 [00:10<00:05,  6.04it/s, loss=3.08, v_num=0]
Epoch 34:  69%|██████▉   | 67/97 [00:11<00:04,  6.09it/s, loss=3.08, v_num=0]
Epoch 34:  69%|██████▉   | 67/97 [00:11<00:04,  6.06it/s, loss=3.08, v_num=0]
Epoch 34:  70%|███████   | 68/97 [00:11<00:04,  6.10it/s, loss=3.08, v_num=0]
Epoch 34:  70%|███████   | 68/97 [00:11<00:04,  6.07it/s, loss=3.07, v_num=0]
Epoch 34:  71%|███████   | 69/97 [00:11<00:04,  6.12it/s, loss=3.07, v_num=0]
Epoch 34:  71%|███████   | 69/97 [00:11<00:04,  6.09it/s, loss=3.07, v_num=0]
Epoch 34:  72%|███████▏  | 70/97 [00:11<00:04,  6.13it/s, loss=3.07, v_num=0]
Epoch 34:  72%|███████▏  | 70/97 [00:11<00:04,  6.10it/s, loss=3.07, v_num=0]
Epoch 34:  73%|███████▎  | 71/97 [00:11<00:04,  6.15it/s, loss=3.07, v_num=0]
Epoch 34:  73%|███████▎  | 71/97 [00:11<00:04,  6.11it/s, loss=3.05, v_num=0]
Epoch 34:  74%|███████▍  | 72/97 [00:11<00:04,  6.16it/s, loss=3.05, v_num=0]
Epoch 34:  74%|███████▍  | 72/97 [00:11<00:04,  6.13it/s, loss=3.04, v_num=0]
Epoch 34:  75%|███████▌  | 73/97 [00:11<00:03,  6.17it/s, loss=3.04, v_num=0]
Epoch 34:  75%|███████▌  | 73/97 [00:11<00:03,  6.14it/s, loss=3.04, v_num=0]
Epoch 34:  76%|███████▋  | 74/97 [00:11<00:03,  6.19it/s, loss=3.04, v_num=0]
Epoch 34:  76%|███████▋  | 74/97 [00:12<00:03,  6.15it/s, loss=3.03, v_num=0]
Epoch 34:  77%|███████▋  | 75/97 [00:12<00:03,  6.20it/s, loss=3.03, v_num=0]
Epoch 34:  77%|███████▋  | 75/97 [00:12<00:03,  6.17it/s, loss=3.03, v_num=0]
Epoch 34:  78%|███████▊  | 76/97 [00:12<00:03,  6.21it/s, loss=3.03, v_num=0]
Epoch 34:  78%|███████▊  | 76/97 [00:12<00:03,  6.18it/s, loss=3.04, v_num=0]
Epoch 34:  79%|███████▉  | 77/97 [00:12<00:03,  6.22it/s, loss=3.04, v_num=0]
Epoch 34:  79%|███████▉  | 77/97 [00:12<00:03,  6.19it/s, loss=3.02, v_num=0]
Epoch 34:  80%|████████  | 78/97 [00:12<00:03,  6.23it/s, loss=3.02, v_num=0]
Epoch 34:  80%|████████  | 78/97 [00:12<00:03,  6.20it/s, loss=3.02, v_num=0]
Epoch 34:  81%|████████▏ | 79/97 [00:12<00:02,  6.24it/s, loss=3.02, v_num=0]
Epoch 34:  81%|████████▏ | 79/97 [00:12<00:02,  6.21it/s, loss=3.03, v_num=0]
Epoch 34:  82%|████████▏ | 80/97 [00:12<00:02,  6.25it/s, loss=3.03, v_num=0]
Epoch 34:  82%|████████▏ | 80/97 [00:12<00:02,  6.23it/s, loss=3.02, v_num=0]
Epoch 34:  84%|████████▎ | 81/97 [00:12<00:02,  6.26it/s, loss=3.02, v_num=0]
Epoch 34:  84%|████████▎ | 81/97 [00:12<00:02,  6.24it/s, loss=3.02, v_num=0]
Epoch 34:  85%|████████▍ | 82/97 [00:13<00:02,  6.27it/s, loss=3.02, v_num=0]
Epoch 34:  85%|████████▍ | 82/97 [00:13<00:02,  6.25it/s, loss=3.03, v_num=0]
Epoch 34:  86%|████████▌ | 83/97 [00:13<00:02,  6.28it/s, loss=3.03, v_num=0]
Epoch 34:  86%|████████▌ | 83/97 [00:13<00:02,  6.26it/s, loss=3.03, v_num=0]
Epoch 34:  87%|████████▋ | 84/97 [00:13<00:02,  6.29it/s, loss=3.03, v_num=0]
Epoch 34:  87%|████████▋ | 84/97 [00:13<00:02,  6.27it/s, loss=3.02, v_num=0]
Epoch 34:  88%|████████▊ | 85/97 [00:13<00:01,  6.31it/s, loss=3.02, v_num=0]
Epoch 34:  88%|████████▊ | 85/97 [00:13<00:01,  6.28it/s, loss=3.01, v_num=0]
Epoch 34:  89%|████████▊ | 86/97 [00:13<00:01,  6.31it/s, loss=3.01, v_num=0]
Epoch 34:  89%|████████▊ | 86/97 [00:13<00:01,  6.29it/s, loss=3, v_num=0]
Epoch 34:  90%|████████▉ | 87/97 [00:13<00:01,  6.32it/s, loss=3, v_num=0]
Epoch 34:  90%|████████▉ | 87/97 [00:13<00:01,  6.30it/s, loss=2.99, v_num=0]
Epoch 34:  91%|█████████ | 88/97 [00:13<00:01,  6.33it/s, loss=2.99, v_num=0]
Epoch 34:  91%|█████████ | 88/97 [00:13<00:01,  6.31it/s, loss=2.99, v_num=0]
Epoch 34:  92%|█████████▏| 89/97 [00:14<00:01,  6.34it/s, loss=2.99, v_num=0]
Epoch 34:  92%|█████████▏| 89/97 [00:14<00:01,  6.32it/s, loss=3, v_num=0]
Epoch 34:  93%|█████████▎| 90/97 [00:14<00:01,  6.35it/s, loss=3, v_num=0]
Epoch 34:  93%|█████████▎| 90/97 [00:14<00:01,  6.33it/s, loss=3, v_num=0]
Epoch 34:  94%|█████████▍| 91/97 [00:14<00:00,  6.36it/s, loss=3, v_num=0]
Epoch 34:  94%|█████████▍| 91/97 [00:14<00:00,  6.34it/s, loss=3, v_num=0]
Epoch 34:  95%|█████████▍| 92/97 [00:14<00:00,  6.37it/s, loss=3, v_num=0]
Epoch 34:  95%|█████████▍| 92/97 [00:14<00:00,  6.35it/s, loss=3.01, v_num=0]
Epoch 34:  96%|█████████▌| 93/97 [00:14<00:00,  6.38it/s, loss=3.01, v_num=0]
Epoch 34:  96%|█████████▌| 93/97 [00:14<00:00,  6.36it/s, loss=3.01, v_num=0]
Epoch 34:  97%|█████████▋| 94/97 [00:14<00:00,  6.39it/s, loss=3.01, v_num=0]
Epoch 34:  97%|█████████▋| 94/97 [00:14<00:00,  6.36it/s, loss=3, v_num=0]
Epoch 34:  98%|█████████▊| 95/97 [00:14<00:00,  6.40it/s, loss=3, v_num=0]
Epoch 34:  98%|█████████▊| 95/97 [00:14<00:00,  6.37it/s, loss=3, v_num=0]
Epoch 34:  99%|█████████▉| 96/97 [00:14<00:00,  6.40it/s, loss=3, v_num=0]
Epoch 34:  99%|█████████▉| 96/97 [00:15<00:00,  6.38it/s, loss=2.99, v_num=0]
Epoch 34: 100%|██████████| 97/97 [00:15<00:00,  6.41it/s, loss=2.99, v_num=0]
Epoch 34: 100%|██████████| 97/97 [00:15<00:00,  6.39it/s, loss=3, v_num=0]
Epoch 34: 100%|██████████| 97/97 [00:15<00:00,  6.39it/s, loss=3, v_num=0]
Epoch 34:   0%|          | 0/97 [00:00<?, ?it/s, loss=3, v_num=0]
Epoch 35:   0%|          | 0/97 [00:00<?, ?it/s, loss=3, v_num=0]
Epoch 35:   1%|          | 1/97 [00:01<02:28,  1.55s/it, loss=3, v_num=0]
Epoch 35:   1%|          | 1/97 [00:01<02:33,  1.60s/it, loss=3, v_num=0]
Epoch 35:   2%|▏         | 2/97 [00:01<01:20,  1.18it/s, loss=3, v_num=0]
Epoch 35:   2%|▏         | 2/97 [00:01<01:22,  1.15it/s, loss=3, v_num=0]
Epoch 35:   3%|▎         | 3/97 [00:01<00:57,  1.64it/s, loss=3, v_num=0]
Epoch 35:   3%|▎         | 3/97 [00:01<00:58,  1.59it/s, loss=3.01, v_num=0]
Epoch 35:   4%|▍         | 4/97 [00:01<00:45,  2.04it/s, loss=3.01, v_num=0]
Epoch 35:   4%|▍         | 4/97 [00:02<00:46,  1.98it/s, loss=3.01, v_num=0]
Epoch 35:   5%|▌         | 5/97 [00:02<00:38,  2.38it/s, loss=3.01, v_num=0]
Epoch 35:   5%|▌         | 5/97 [00:02<00:39,  2.32it/s, loss=3.01, v_num=0]
Epoch 35:   6%|▌         | 6/97 [00:02<00:33,  2.68it/s, loss=3.01, v_num=0]
Epoch 35:   6%|▌         | 6/97 [00:02<00:34,  2.62it/s, loss=3.01, v_num=0]
Epoch 35:   7%|▋         | 7/97 [00:02<00:30,  2.95it/s, loss=3.01, v_num=0]
Epoch 35:   7%|▋         | 7/97 [00:02<00:31,  2.88it/s, loss=3.01, v_num=0]
Epoch 35:   8%|▊         | 8/97 [00:02<00:27,  3.18it/s, loss=3.01, v_num=0]
Epoch 35:   8%|▊         | 8/97 [00:02<00:28,  3.12it/s, loss=3.02, v_num=0]
Epoch 35:   9%|▉         | 9/97 [00:02<00:25,  3.39it/s, loss=3.02, v_num=0]
Epoch 35:   9%|▉         | 9/97 [00:02<00:26,  3.32it/s, loss=3.02, v_num=0]
Epoch 35:  10%|█         | 10/97 [00:02<00:24,  3.57it/s, loss=3.02, v_num=0]
Epoch 35:  10%|█         | 10/97 [00:02<00:24,  3.50it/s, loss=3.03, v_num=0]
Epoch 35:  11%|█▏        | 11/97 [00:02<00:22,  3.75it/s, loss=3.03, v_num=0]
Epoch 35:  11%|█▏        | 11/97 [00:02<00:23,  3.68it/s, loss=3.03, v_num=0]
Epoch 35:  12%|█▏        | 12/97 [00:03<00:21,  3.90it/s, loss=3.03, v_num=0]
Epoch 35:  12%|█▏        | 12/97 [00:03<00:22,  3.84it/s, loss=3.03, v_num=0]
Epoch 35:  13%|█▎        | 13/97 [00:03<00:20,  4.06it/s, loss=3.03, v_num=0]
Epoch 35:  13%|█▎        | 13/97 [00:03<00:21,  3.98it/s, loss=3.02, v_num=0]
Epoch 35:  14%|█▍        | 14/97 [00:03<00:19,  4.18it/s, loss=3.02, v_num=0]
Epoch 35:  14%|█▍        | 14/97 [00:03<00:20,  4.12it/s, loss=3.03, v_num=0]
Epoch 35:  15%|█▌        | 15/97 [00:03<00:19,  4.31it/s, loss=3.03, v_num=0]
Epoch 35:  15%|█▌        | 15/97 [00:03<00:19,  4.24it/s, loss=3.03, v_num=0]
Epoch 35:  16%|█▋        | 16/97 [00:03<00:18,  4.42it/s, loss=3.03, v_num=0]
Epoch 35:  16%|█▋        | 16/97 [00:03<00:18,  4.35it/s, loss=3.03, v_num=0]
Epoch 35:  18%|█▊        | 17/97 [00:03<00:17,  4.52it/s, loss=3.03, v_num=0]
Epoch 35:  18%|█▊        | 17/97 [00:03<00:17,  4.45it/s, loss=3.02, v_num=0]
Epoch 35:  19%|█▊        | 18/97 [00:03<00:17,  4.62it/s, loss=3.02, v_num=0]
Epoch 35:  19%|█▊        | 18/97 [00:03<00:17,  4.55it/s, loss=3.03, v_num=0]
Epoch 35:  20%|█▉        | 19/97 [00:04<00:16,  4.71it/s, loss=3.03, v_num=0]
Epoch 35:  20%|█▉        | 19/97 [00:04<00:16,  4.64it/s, loss=3.02, v_num=0]
Epoch 35:  21%|██        | 20/97 [00:04<00:16,  4.79it/s, loss=3.02, v_num=0]
Epoch 35:  21%|██        | 20/97 [00:04<00:16,  4.72it/s, loss=3.02, v_num=0]
Epoch 35:  22%|██▏       | 21/97 [00:04<00:15,  4.87it/s, loss=3.02, v_num=0]
Epoch 35:  22%|██▏       | 21/97 [00:04<00:15,  4.80it/s, loss=3.03, v_num=0]
Epoch 35:  23%|██▎       | 22/97 [00:04<00:15,  4.93it/s, loss=3.03, v_num=0]
Epoch 35:  23%|██▎       | 22/97 [00:04<00:15,  4.87it/s, loss=3.02, v_num=0]
Epoch 35:  24%|██▎       | 23/97 [00:04<00:14,  5.00it/s, loss=3.02, v_num=0]
Epoch 35:  24%|██▎       | 23/97 [00:04<00:14,  4.94it/s, loss=3.02, v_num=0]
Epoch 35:  25%|██▍       | 24/97 [00:04<00:14,  5.06it/s, loss=3.02, v_num=0]
Epoch 35:  25%|██▍       | 24/97 [00:04<00:14,  5.00it/s, loss=3.02, v_num=0]
Epoch 35:  26%|██▌       | 25/97 [00:04<00:14,  5.12it/s, loss=3.02, v_num=0]
Epoch 35:  26%|██▌       | 25/97 [00:04<00:14,  5.07it/s, loss=3.02, v_num=0]
Epoch 35:  27%|██▋       | 26/97 [00:05<00:13,  5.18it/s, loss=3.02, v_num=0]
Epoch 35:  27%|██▋       | 26/97 [00:05<00:13,  5.12it/s, loss=3.02, v_num=0]
Epoch 35:  28%|██▊       | 27/97 [00:05<00:13,  5.23it/s, loss=3.02, v_num=0]
Epoch 35:  28%|██▊       | 27/97 [00:05<00:13,  5.18it/s, loss=3.03, v_num=0]
Epoch 35:  29%|██▉       | 28/97 [00:05<00:13,  5.29it/s, loss=3.03, v_num=0]
Epoch 35:  29%|██▉       | 28/97 [00:05<00:13,  5.23it/s, loss=3.02, v_num=0]
Epoch 35:  30%|██▉       | 29/97 [00:05<00:12,  5.34it/s, loss=3.02, v_num=0]
Epoch 35:  30%|██▉       | 29/97 [00:05<00:12,  5.28it/s, loss=3.02, v_num=0]
Epoch 35:  31%|███       | 30/97 [00:05<00:12,  5.39it/s, loss=3.02, v_num=0]
Epoch 35:  31%|███       | 30/97 [00:05<00:12,  5.33it/s, loss=3.01, v_num=0]
Epoch 35:  32%|███▏      | 31/97 [00:05<00:12,  5.43it/s, loss=3.01, v_num=0]
Epoch 35:  32%|███▏      | 31/97 [00:05<00:12,  5.38it/s, loss=3.01, v_num=0]
Epoch 35:  33%|███▎      | 32/97 [00:05<00:11,  5.48it/s, loss=3.01, v_num=0]
Epoch 35:  33%|███▎      | 32/97 [00:05<00:11,  5.42it/s, loss=3, v_num=0]
Epoch 35:  34%|███▍      | 33/97 [00:05<00:11,  5.51it/s, loss=3, v_num=0]
Epoch 35:  34%|███▍      | 33/97 [00:06<00:11,  5.46it/s, loss=3.01, v_num=0]
Epoch 35:  35%|███▌      | 34/97 [00:06<00:11,  5.55it/s, loss=3.01, v_num=0]
Epoch 35:  35%|███▌      | 34/97 [00:06<00:11,  5.50it/s, loss=3, v_num=0]
Epoch 35:  36%|███▌      | 35/97 [00:06<00:11,  5.59it/s, loss=3, v_num=0]
Epoch 35:  36%|███▌      | 35/97 [00:06<00:11,  5.54it/s, loss=3, v_num=0]
Epoch 35:  37%|███▋      | 36/97 [00:06<00:10,  5.62it/s, loss=3, v_num=0]
Epoch 35:  37%|███▋      | 36/97 [00:06<00:10,  5.58it/s, loss=3.01, v_num=0]
Epoch 35:  38%|███▊      | 37/97 [00:06<00:10,  5.66it/s, loss=3.01, v_num=0]
Epoch 35:  38%|███▊      | 37/97 [00:06<00:10,  5.61it/s, loss=3.01, v_num=0]
Epoch 35:  39%|███▉      | 38/97 [00:06<00:10,  5.69it/s, loss=3.01, v_num=0]
Epoch 35:  39%|███▉      | 38/97 [00:06<00:10,  5.65it/s, loss=3.01, v_num=0]
Epoch 35:  40%|████      | 39/97 [00:06<00:10,  5.72it/s, loss=3.01, v_num=0]
Epoch 35:  40%|████      | 39/97 [00:06<00:10,  5.68it/s, loss=3.02, v_num=0]
Epoch 35:  41%|████      | 40/97 [00:06<00:09,  5.75it/s, loss=3.02, v_num=0]
Epoch 35:  41%|████      | 40/97 [00:07<00:09,  5.70it/s, loss=3.01, v_num=0]
Epoch 35:  42%|████▏     | 41/97 [00:07<00:09,  5.78it/s, loss=3.01, v_num=0]
Epoch 35:  42%|████▏     | 41/97 [00:07<00:09,  5.73it/s, loss=3.01, v_num=0]
Epoch 35:  43%|████▎     | 42/97 [00:07<00:09,  5.80it/s, loss=3.01, v_num=0]
Epoch 35:  43%|████▎     | 42/97 [00:07<00:09,  5.76it/s, loss=3.02, v_num=0]
Epoch 35:  44%|████▍     | 43/97 [00:07<00:09,  5.83it/s, loss=3.02, v_num=0]
Epoch 35:  44%|████▍     | 43/97 [00:07<00:09,  5.79it/s, loss=3.02, v_num=0]
Epoch 35:  45%|████▌     | 44/97 [00:07<00:09,  5.86it/s, loss=3.02, v_num=0]
Epoch 35:  45%|████▌     | 44/97 [00:07<00:09,  5.82it/s, loss=3.01, v_num=0]
Epoch 35:  46%|████▋     | 45/97 [00:07<00:08,  5.89it/s, loss=3.01, v_num=0]
Epoch 35:  46%|████▋     | 45/97 [00:07<00:08,  5.84it/s, loss=3.01, v_num=0]
Epoch 35:  47%|████▋     | 46/97 [00:07<00:08,  5.91it/s, loss=3.01, v_num=0]
Epoch 35:  47%|████▋     | 46/97 [00:07<00:08,  5.86it/s, loss=3, v_num=0]
Epoch 35:  48%|████▊     | 47/97 [00:07<00:08,  5.93it/s, loss=3, v_num=0]
Epoch 35:  48%|████▊     | 47/97 [00:07<00:08,  5.89it/s, loss=3, v_num=0]
Epoch 35:  49%|████▉     | 48/97 [00:08<00:08,  5.95it/s, loss=3, v_num=0]
Epoch 35:  49%|████▉     | 48/97 [00:08<00:08,  5.91it/s, loss=3, v_num=0]
Epoch 35:  51%|█████     | 49/97 [00:08<00:08,  5.98it/s, loss=3, v_num=0]
Epoch 35:  51%|█████     | 49/97 [00:08<00:08,  5.93it/s, loss=3.01, v_num=0]
Epoch 35:  52%|█████▏    | 50/97 [00:08<00:07,  6.00it/s, loss=3.01, v_num=0]
Epoch 35:  52%|█████▏    | 50/97 [00:08<00:07,  5.96it/s, loss=3.01, v_num=0]
Epoch 35:  53%|█████▎    | 51/97 [00:08<00:07,  6.02it/s, loss=3.01, v_num=0]
Epoch 35:  53%|█████▎    | 51/97 [00:08<00:07,  5.98it/s, loss=3.02, v_num=0]
Epoch 35:  54%|█████▎    | 52/97 [00:08<00:07,  6.04it/s, loss=3.02, v_num=0]
Epoch 35:  54%|█████▎    | 52/97 [00:08<00:07,  6.00it/s, loss=3.03, v_num=0]
Epoch 35:  55%|█████▍    | 53/97 [00:08<00:07,  6.06it/s, loss=3.03, v_num=0]
Epoch 35:  55%|█████▍    | 53/97 [00:08<00:07,  6.02it/s, loss=3.02, v_num=0]
Epoch 35:  56%|█████▌    | 54/97 [00:08<00:07,  6.08it/s, loss=3.02, v_num=0]
Epoch 35:  56%|█████▌    | 54/97 [00:08<00:07,  6.04it/s, loss=3.03, v_num=0]
Epoch 35:  57%|█████▋    | 55/97 [00:09<00:06,  6.10it/s, loss=3.03, v_num=0]
Epoch 35:  57%|█████▋    | 55/97 [00:09<00:06,  6.06it/s, loss=3.02, v_num=0]
Epoch 35:  58%|█████▊    | 56/97 [00:09<00:06,  6.11it/s, loss=3.02, v_num=0]
Epoch 35:  58%|█████▊    | 56/97 [00:09<00:06,  6.07it/s, loss=3.02, v_num=0]
Epoch 35:  59%|█████▉    | 57/97 [00:09<00:06,  6.13it/s, loss=3.02, v_num=0]
Epoch 35:  59%|█████▉    | 57/97 [00:09<00:06,  6.09it/s, loss=3.02, v_num=0]
Epoch 35:  60%|█████▉    | 58/97 [00:09<00:06,  6.15it/s, loss=3.02, v_num=0]
Epoch 35:  60%|█████▉    | 58/97 [00:09<00:06,  6.11it/s, loss=3.02, v_num=0]
Epoch 35:  61%|██████    | 59/97 [00:09<00:06,  6.16it/s, loss=3.02, v_num=0]
Epoch 35:  61%|██████    | 59/97 [00:09<00:06,  6.13it/s, loss=3.02, v_num=0]
Epoch 35:  62%|██████▏   | 60/97 [00:09<00:05,  6.18it/s, loss=3.02, v_num=0]
Epoch 35:  62%|██████▏   | 60/97 [00:09<00:06,  6.14it/s, loss=3.02, v_num=0]
Epoch 35:  63%|██████▎   | 61/97 [00:09<00:05,  6.19it/s, loss=3.02, v_num=0]
Epoch 35:  63%|██████▎   | 61/97 [00:09<00:05,  6.16it/s, loss=3.02, v_num=0]
Epoch 35:  64%|██████▍   | 62/97 [00:09<00:05,  6.21it/s, loss=3.02, v_num=0]
Epoch 35:  64%|██████▍   | 62/97 [00:10<00:05,  6.17it/s, loss=3.02, v_num=0]
Epoch 35:  65%|██████▍   | 63/97 [00:10<00:05,  6.22it/s, loss=3.02, v_num=0]
Epoch 35:  65%|██████▍   | 63/97 [00:10<00:05,  6.19it/s, loss=3.02, v_num=0]
Epoch 35:  66%|██████▌   | 64/97 [00:10<00:05,  6.24it/s, loss=3.02, v_num=0]
Epoch 35:  66%|██████▌   | 64/97 [00:10<00:05,  6.20it/s, loss=3.02, v_num=0]
Epoch 35:  67%|██████▋   | 65/97 [00:10<00:05,  6.25it/s, loss=3.02, v_num=0]
Epoch 35:  67%|██████▋   | 65/97 [00:10<00:05,  6.22it/s, loss=3.02, v_num=0]
Epoch 35:  68%|██████▊   | 66/97 [00:10<00:04,  6.26it/s, loss=3.02, v_num=0]
Epoch 35:  68%|██████▊   | 66/97 [00:10<00:04,  6.23it/s, loss=3.02, v_num=0]
Epoch 35:  69%|██████▉   | 67/97 [00:10<00:04,  6.27it/s, loss=3.02, v_num=0]
Epoch 35:  69%|██████▉   | 67/97 [00:10<00:04,  6.24it/s, loss=3.02, v_num=0]
Epoch 35:  70%|███████   | 68/97 [00:10<00:04,  6.29it/s, loss=3.02, v_num=0]
Epoch 35:  70%|███████   | 68/97 [00:10<00:04,  6.26it/s, loss=3.02, v_num=0]
Epoch 35:  71%|███████   | 69/97 [00:10<00:04,  6.30it/s, loss=3.02, v_num=0]
Epoch 35:  71%|███████   | 69/97 [00:11<00:04,  6.27it/s, loss=3.01, v_num=0]
Epoch 35:  72%|███████▏  | 70/97 [00:11<00:04,  6.31it/s, loss=3.01, v_num=0]
Epoch 35:  72%|███████▏  | 70/97 [00:11<00:04,  6.28it/s, loss=3, v_num=0]
Epoch 35:  73%|███████▎  | 71/97 [00:11<00:04,  6.32it/s, loss=3, v_num=0]
Epoch 35:  73%|███████▎  | 71/97 [00:11<00:04,  6.29it/s, loss=3, v_num=0]
Epoch 35:  74%|███████▍  | 72/97 [00:11<00:03,  6.34it/s, loss=3, v_num=0]
Epoch 35:  74%|███████▍  | 72/97 [00:11<00:03,  6.31it/s, loss=2.99, v_num=0]
Epoch 35:  75%|███████▌  | 73/97 [00:11<00:03,  6.35it/s, loss=2.99, v_num=0]
Epoch 35:  75%|███████▌  | 73/97 [00:11<00:03,  6.32it/s, loss=2.99, v_num=0]
Epoch 35:  76%|███████▋  | 74/97 [00:11<00:03,  6.36it/s, loss=2.99, v_num=0]
Epoch 35:  76%|███████▋  | 74/97 [00:11<00:03,  6.33it/s, loss=2.99, v_num=0]
Epoch 35:  77%|███████▋  | 75/97 [00:11<00:03,  6.37it/s, loss=2.99, v_num=0]
Epoch 35:  77%|███████▋  | 75/97 [00:11<00:03,  6.34it/s, loss=2.99, v_num=0]
Epoch 35:  78%|███████▊  | 76/97 [00:11<00:03,  6.38it/s, loss=2.99, v_num=0]
Epoch 35:  78%|███████▊  | 76/97 [00:11<00:03,  6.35it/s, loss=2.97, v_num=0]
Epoch 35:  79%|███████▉  | 77/97 [00:12<00:03,  6.39it/s, loss=2.97, v_num=0]
Epoch 35:  79%|███████▉  | 77/97 [00:12<00:03,  6.36it/s, loss=2.97, v_num=0]
Epoch 35:  80%|████████  | 78/97 [00:12<00:02,  6.40it/s, loss=2.97, v_num=0]
Epoch 35:  80%|████████  | 78/97 [00:12<00:02,  6.37it/s, loss=2.97, v_num=0]
Epoch 35:  81%|████████▏ | 79/97 [00:12<00:02,  6.41it/s, loss=2.97, v_num=0]
Epoch 35:  81%|████████▏ | 79/97 [00:12<00:02,  6.38it/s, loss=2.98, v_num=0]
Epoch 35:  82%|████████▏ | 80/97 [00:12<00:02,  6.42it/s, loss=2.98, v_num=0]
Epoch 35:  82%|████████▏ | 80/97 [00:12<00:02,  6.39it/s, loss=2.98, v_num=0]
Epoch 35:  84%|████████▎ | 81/97 [00:12<00:02,  6.43it/s, loss=2.98, v_num=0]
Epoch 35:  84%|████████▎ | 81/97 [00:12<00:02,  6.40it/s, loss=2.98, v_num=0]
Epoch 35:  85%|████████▍ | 82/97 [00:12<00:02,  6.44it/s, loss=2.98, v_num=0]
Epoch 35:  85%|████████▍ | 82/97 [00:12<00:02,  6.41it/s, loss=2.97, v_num=0]
Epoch 35:  86%|████████▌ | 83/97 [00:12<00:02,  6.44it/s, loss=2.97, v_num=0]
Epoch 35:  86%|████████▌ | 83/97 [00:12<00:02,  6.41it/s, loss=2.97, v_num=0]
Epoch 35:  87%|████████▋ | 84/97 [00:13<00:02,  6.45it/s, loss=2.97, v_num=0]
Epoch 35:  87%|████████▋ | 84/97 [00:13<00:02,  6.42it/s, loss=2.98, v_num=0]
Epoch 35:  88%|████████▊ | 85/97 [00:13<00:01,  6.46it/s, loss=2.98, v_num=0]
Epoch 35:  88%|████████▊ | 85/97 [00:13<00:01,  6.43it/s, loss=2.98, v_num=0]
Epoch 35:  89%|████████▊ | 86/97 [00:13<00:01,  6.47it/s, loss=2.98, v_num=0]
Epoch 35:  89%|████████▊ | 86/97 [00:13<00:01,  6.44it/s, loss=2.98, v_num=0]
Epoch 35:  90%|████████▉ | 87/97 [00:13<00:01,  6.47it/s, loss=2.98, v_num=0]
Epoch 35:  90%|████████▉ | 87/97 [00:13<00:01,  6.45it/s, loss=2.98, v_num=0]
Epoch 35:  91%|█████████ | 88/97 [00:13<00:01,  6.48it/s, loss=2.98, v_num=0]
Epoch 35:  91%|█████████ | 88/97 [00:13<00:01,  6.46it/s, loss=2.97, v_num=0]
Epoch 35:  92%|█████████▏| 89/97 [00:13<00:01,  6.49it/s, loss=2.97, v_num=0]
Epoch 35:  92%|█████████▏| 89/97 [00:13<00:01,  6.46it/s, loss=2.97, v_num=0]
Epoch 35:  93%|█████████▎| 90/97 [00:13<00:01,  6.50it/s, loss=2.97, v_num=0]
Epoch 35:  93%|█████████▎| 90/97 [00:13<00:01,  6.47it/s, loss=2.97, v_num=0]
Epoch 35:  94%|█████████▍| 91/97 [00:13<00:00,  6.50it/s, loss=2.97, v_num=0]
Epoch 35:  94%|█████████▍| 91/97 [00:14<00:00,  6.48it/s, loss=2.97, v_num=0]
Epoch 35:  95%|█████████▍| 92/97 [00:14<00:00,  6.51it/s, loss=2.97, v_num=0]
Epoch 35:  95%|█████████▍| 92/97 [00:14<00:00,  6.48it/s, loss=2.97, v_num=0]
Epoch 35:  96%|█████████▌| 93/97 [00:14<00:00,  6.52it/s, loss=2.97, v_num=0]
Epoch 35:  96%|█████████▌| 93/97 [00:14<00:00,  6.49it/s, loss=2.99, v_num=0]
Epoch 35:  97%|█████████▋| 94/97 [00:14<00:00,  6.52it/s, loss=2.99, v_num=0]
Epoch 35:  97%|█████████▋| 94/97 [00:14<00:00,  6.50it/s, loss=2.98, v_num=0]
Epoch 35:  98%|█████████▊| 95/97 [00:14<00:00,  6.53it/s, loss=2.98, v_num=0]
Epoch 35:  98%|█████████▊| 95/97 [00:14<00:00,  6.51it/s, loss=2.98, v_num=0]
Epoch 35:  99%|█████████▉| 96/97 [00:14<00:00,  6.54it/s, loss=2.98, v_num=0]
Epoch 35:  99%|█████████▉| 96/97 [00:14<00:00,  6.51it/s, loss=2.99, v_num=0]
Epoch 35: 100%|██████████| 97/97 [00:14<00:00,  6.54it/s, loss=2.99, v_num=0]
Epoch 35: 100%|██████████| 97/97 [00:14<00:00,  6.52it/s, loss=2.99, v_num=0]
Epoch 35: 100%|██████████| 97/97 [00:14<00:00,  6.52it/s, loss=2.99, v_num=0]
Epoch 35:   0%|          | 0/97 [00:00<?, ?it/s, loss=2.99, v_num=0]
Epoch 36:   0%|          | 0/97 [00:00<?, ?it/s, loss=2.99, v_num=0]
Epoch 36:   1%|          | 1/97 [00:01<02:13,  1.39s/it, loss=2.99, v_num=0]
Epoch 36:   1%|          | 1/97 [00:01<02:18,  1.45s/it, loss=2.99, v_num=0]
Epoch 36:   2%|▏         | 2/97 [00:01<01:12,  1.31it/s, loss=2.99, v_num=0]
Epoch 36:   2%|▏         | 2/97 [00:01<01:15,  1.26it/s, loss=3, v_num=0]
Epoch 36:   3%|▎         | 3/97 [00:01<00:52,  1.79it/s, loss=3, v_num=0]
Epoch 36:   3%|▎         | 3/97 [00:01<00:54,  1.74it/s, loss=2.99, v_num=0]
Epoch 36:   4%|▍         | 4/97 [00:01<00:42,  2.21it/s, loss=2.99, v_num=0]
Epoch 36:   4%|▍         | 4/97 [00:01<00:43,  2.14it/s, loss=2.99, v_num=0]
Epoch 36:   5%|▌         | 5/97 [00:01<00:35,  2.57it/s, loss=2.99, v_num=0]
Epoch 36:   5%|▌         | 5/97 [00:02<00:36,  2.50it/s, loss=2.99, v_num=0]
Epoch 36:   6%|▌         | 6/97 [00:02<00:31,  2.88it/s, loss=2.99, v_num=0]
Epoch 36:   6%|▌         | 6/97 [00:02<00:32,  2.81it/s, loss=2.99, v_num=0]
Epoch 36:   7%|▋         | 7/97 [00:02<00:28,  3.15it/s, loss=2.99, v_num=0]
Epoch 36:   7%|▋         | 7/97 [00:02<00:29,  3.08it/s, loss=2.98, v_num=0]
Epoch 36:   8%|▊         | 8/97 [00:02<00:26,  3.40it/s, loss=2.98, v_num=0]
Epoch 36:   8%|▊         | 8/97 [00:02<00:26,  3.32it/s, loss=2.98, v_num=0]
Epoch 36:   9%|▉         | 9/97 [00:02<00:24,  3.61it/s, loss=2.98, v_num=0]
Epoch 36:   9%|▉         | 9/97 [00:02<00:24,  3.53it/s, loss=2.98, v_num=0]
Epoch 36:  10%|█         | 10/97 [00:02<00:22,  3.79it/s, loss=2.98, v_num=0]
Epoch 36:  10%|█         | 10/97 [00:02<00:23,  3.72it/s, loss=2.99, v_num=0]
Epoch 36:  11%|█▏        | 11/97 [00:02<00:21,  3.97it/s, loss=2.99, v_num=0]
Epoch 36:  11%|█▏        | 11/97 [00:02<00:22,  3.89it/s, loss=2.99, v_num=0]
Epoch 36:  12%|█▏        | 12/97 [00:02<00:20,  4.12it/s, loss=2.99, v_num=0]
Epoch 36:  12%|█▏        | 12/97 [00:02<00:21,  4.05it/s, loss=2.99, v_num=0]
Epoch 36:  13%|█▎        | 13/97 [00:03<00:19,  4.27it/s, loss=2.99, v_num=0]
Epoch 36:  13%|█▎        | 13/97 [00:03<00:20,  4.19it/s, loss=2.99, v_num=0]
Epoch 36:  14%|█▍        | 14/97 [00:03<00:18,  4.40it/s, loss=2.99, v_num=0]
Epoch 36:  14%|█▍        | 14/97 [00:03<00:19,  4.33it/s, loss=2.99, v_num=0]
Epoch 36:  15%|█▌        | 15/97 [00:03<00:18,  4.53it/s, loss=2.99, v_num=0]
Epoch 36:  15%|█▌        | 15/97 [00:03<00:18,  4.45it/s, loss=2.99, v_num=0]
Epoch 36:  16%|█▋        | 16/97 [00:03<00:17,  4.64it/s, loss=2.99, v_num=0]
Epoch 36:  16%|█▋        | 16/97 [00:03<00:17,  4.56it/s, loss=2.98, v_num=0]
Epoch 36:  18%|█▊        | 17/97 [00:03<00:16,  4.74it/s, loss=2.98, v_num=0]
Epoch 36:  18%|█▊        | 17/97 [00:03<00:17,  4.66it/s, loss=2.99, v_num=0]
Epoch 36:  19%|█▊        | 18/97 [00:03<00:16,  4.83it/s, loss=2.99, v_num=0]
Epoch 36:  19%|█▊        | 18/97 [00:03<00:16,  4.76it/s, loss=2.98, v_num=0]
Epoch 36:  20%|█▉        | 19/97 [00:03<00:15,  4.93it/s, loss=2.98, v_num=0]
Epoch 36:  20%|█▉        | 19/97 [00:03<00:16,  4.85it/s, loss=2.98, v_num=0]
Epoch 36:  21%|██        | 20/97 [00:04<00:15,  4.99it/s, loss=2.98, v_num=0]
Epoch 36:  21%|██        | 20/97 [00:04<00:15,  4.93it/s, loss=2.98, v_num=0]
Epoch 36:  22%|██▏       | 21/97 [00:04<00:14,  5.07it/s, loss=2.98, v_num=0]
Epoch 36:  22%|██▏       | 21/97 [00:04<00:15,  5.01it/s, loss=2.98, v_num=0]
Epoch 36:  23%|██▎       | 22/97 [00:04<00:14,  5.15it/s, loss=2.98, v_num=0]
Epoch 36:  23%|██▎       | 22/97 [00:04<00:14,  5.08it/s, loss=2.97, v_num=0]
Epoch 36:  24%|██▎       | 23/97 [00:04<00:14,  5.21it/s, loss=2.97, v_num=0]
Epoch 36:  24%|██▎       | 23/97 [00:04<00:14,  5.14it/s, loss=2.98, v_num=0]
Epoch 36:  25%|██▍       | 24/97 [00:04<00:13,  5.28it/s, loss=2.98, v_num=0]
Epoch 36:  25%|██▍       | 24/97 [00:04<00:14,  5.21it/s, loss=2.98, v_num=0]
Epoch 36:  26%|██▌       | 25/97 [00:04<00:13,  5.33it/s, loss=2.98, v_num=0]
Epoch 36:  26%|██▌       | 25/97 [00:04<00:13,  5.27it/s, loss=2.98, v_num=0]
Epoch 36:  27%|██▋       | 26/97 [00:04<00:13,  5.38it/s, loss=2.98, v_num=0]
Epoch 36:  27%|██▋       | 26/97 [00:04<00:13,  5.33it/s, loss=2.97, v_num=0]
Epoch 36:  28%|██▊       | 27/97 [00:04<00:12,  5.44it/s, loss=2.97, v_num=0]
Epoch 36:  28%|██▊       | 27/97 [00:05<00:13,  5.38it/s, loss=2.98, v_num=0]
Epoch 36:  29%|██▉       | 28/97 [00:05<00:12,  5.49it/s, loss=2.98, v_num=0]
Epoch 36:  29%|██▉       | 28/97 [00:05<00:12,  5.43it/s, loss=2.98, v_num=0]
Epoch 36:  30%|██▉       | 29/97 [00:05<00:12,  5.54it/s, loss=2.98, v_num=0]
Epoch 36:  30%|██▉       | 29/97 [00:05<00:12,  5.48it/s, loss=2.98, v_num=0]
Epoch 36:  31%|███       | 30/97 [00:05<00:12,  5.58it/s, loss=2.98, v_num=0]
Epoch 36:  31%|███       | 30/97 [00:05<00:12,  5.52it/s, loss=2.99, v_num=0]
Epoch 36:  32%|███▏      | 31/97 [00:05<00:11,  5.62it/s, loss=2.99, v_num=0]
Epoch 36:  32%|███▏      | 31/97 [00:05<00:11,  5.57it/s, loss=2.99, v_num=0]
Epoch 36:  33%|███▎      | 32/97 [00:05<00:11,  5.66it/s, loss=2.99, v_num=0]
Epoch 36:  33%|███▎      | 32/97 [00:05<00:11,  5.61it/s, loss=2.99, v_num=0]
Epoch 36:  34%|███▍      | 33/97 [00:05<00:11,  5.70it/s, loss=2.99, v_num=0]
Epoch 36:  34%|███▍      | 33/97 [00:05<00:11,  5.65it/s, loss=2.99, v_num=0]
Epoch 36:  35%|███▌      | 34/97 [00:05<00:10,  5.74it/s, loss=2.99, v_num=0]
Epoch 36:  35%|███▌      | 34/97 [00:05<00:11,  5.69it/s, loss=3, v_num=0]
Epoch 36:  36%|███▌      | 35/97 [00:06<00:10,  5.78it/s, loss=3, v_num=0]
Epoch 36:  36%|███▌      | 35/97 [00:06<00:10,  5.72it/s, loss=3.01, v_num=0]
Epoch 36:  37%|███▋      | 36/97 [00:06<00:10,  5.81it/s, loss=3.01, v_num=0]
Epoch 36:  37%|███▋      | 36/97 [00:06<00:10,  5.76it/s, loss=3, v_num=0]
Epoch 36:  38%|███▊      | 37/97 [00:06<00:10,  5.84it/s, loss=3, v_num=0]
Epoch 36:  38%|███▊      | 37/97 [00:06<00:10,  5.79it/s, loss=3, v_num=0]
Epoch 36:  39%|███▉      | 38/97 [00:06<00:10,  5.87it/s, loss=3, v_num=0]
Epoch 36:  39%|███▉      | 38/97 [00:06<00:10,  5.82it/s, loss=3.01, v_num=0]
Epoch 36:  40%|████      | 39/97 [00:06<00:09,  5.90it/s, loss=3.01, v_num=0]
Epoch 36:  40%|████      | 39/97 [00:06<00:09,  5.85it/s, loss=3.01, v_num=0]
Epoch 36:  41%|████      | 40/97 [00:06<00:09,  5.93it/s, loss=3.01, v_num=0]
Epoch 36:  41%|████      | 40/97 [00:06<00:09,  5.88it/s, loss=3.01, v_num=0]
Epoch 36:  42%|████▏     | 41/97 [00:06<00:09,  5.95it/s, loss=3.01, v_num=0]
Epoch 36:  42%|████▏     | 41/97 [00:06<00:09,  5.91it/s, loss=3.01, v_num=0]
Epoch 36:  43%|████▎     | 42/97 [00:07<00:09,  5.98it/s, loss=3.01, v_num=0]
Epoch 36:  43%|████▎     | 42/97 [00:07<00:09,  5.93it/s, loss=3.01, v_num=0]
Epoch 36:  44%|████▍     | 43/97 [00:07<00:08,  6.01it/s, loss=3.01, v_num=0]
Epoch 36:  44%|████▍     | 43/97 [00:07<00:09,  5.96it/s, loss=3, v_num=0]
Epoch 36:  45%|████▌     | 44/97 [00:07<00:08,  6.03it/s, loss=3, v_num=0]
Epoch 36:  45%|████▌     | 44/97 [00:07<00:08,  5.99it/s, loss=3, v_num=0]
Epoch 36:  46%|████▋     | 45/97 [00:07<00:08,  6.05it/s, loss=3, v_num=0]
Epoch 36:  46%|████▋     | 45/97 [00:07<00:08,  6.01it/s, loss=3, v_num=0]
Epoch 36:  47%|████▋     | 46/97 [00:07<00:08,  6.08it/s, loss=3, v_num=0]
Epoch 36:  47%|████▋     | 46/97 [00:07<00:08,  6.03it/s, loss=3.01, v_num=0]
Epoch 36:  48%|████▊     | 47/97 [00:07<00:08,  6.10it/s, loss=3.01, v_num=0]
Epoch 36:  48%|████▊     | 47/97 [00:07<00:08,  6.06it/s, loss=3.01, v_num=0]
Epoch 36:  49%|████▉     | 48/97 [00:07<00:08,  6.12it/s, loss=3.01, v_num=0]
Epoch 36:  49%|████▉     | 48/97 [00:07<00:08,  6.08it/s, loss=3.01, v_num=0]
Epoch 36:  51%|█████     | 49/97 [00:07<00:07,  6.14it/s, loss=3.01, v_num=0]
Epoch 36:  51%|█████     | 49/97 [00:08<00:07,  6.10it/s, loss=3.01, v_num=0]
Epoch 36:  52%|█████▏    | 50/97 [00:08<00:07,  6.16it/s, loss=3.01, v_num=0]
Epoch 36:  52%|█████▏    | 50/97 [00:08<00:07,  6.12it/s, loss=3, v_num=0]
Epoch 36:  53%|█████▎    | 51/97 [00:08<00:07,  6.18it/s, loss=3, v_num=0]
Epoch 36:  53%|█████▎    | 51/97 [00:08<00:07,  6.14it/s, loss=3, v_num=0]
Epoch 36:  54%|█████▎    | 52/97 [00:08<00:07,  6.20it/s, loss=3, v_num=0]
Epoch 36:  54%|█████▎    | 52/97 [00:08<00:07,  6.16it/s, loss=3.01, v_num=0]
Epoch 36:  55%|█████▍    | 53/97 [00:08<00:07,  6.22it/s, loss=3.01, v_num=0]
Epoch 36:  55%|█████▍    | 53/97 [00:08<00:07,  6.18it/s, loss=3, v_num=0]
Epoch 36:  56%|█████▌    | 54/97 [00:08<00:06,  6.23it/s, loss=3, v_num=0]
Epoch 36:  56%|█████▌    | 54/97 [00:08<00:06,  6.19it/s, loss=3, v_num=0]
Epoch 36:  57%|█████▋    | 55/97 [00:08<00:06,  6.25it/s, loss=3, v_num=0]
Epoch 36:  57%|█████▋    | 55/97 [00:08<00:06,  6.21it/s, loss=3, v_num=0]
Epoch 36:  58%|█████▊    | 56/97 [00:08<00:06,  6.27it/s, loss=3, v_num=0]
Epoch 36:  58%|█████▊    | 56/97 [00:08<00:06,  6.23it/s, loss=3, v_num=0]
Epoch 36:  59%|█████▉    | 57/97 [00:09<00:06,  6.28it/s, loss=3, v_num=0]
Epoch 36:  59%|█████▉    | 57/97 [00:09<00:06,  6.24it/s, loss=3, v_num=0]
Epoch 36:  60%|█████▉    | 58/97 [00:09<00:06,  6.30it/s, loss=3, v_num=0]
Epoch 36:  60%|█████▉    | 58/97 [00:09<00:06,  6.26it/s, loss=2.99, v_num=0]
Epoch 36:  61%|██████    | 59/97 [00:09<00:06,  6.31it/s, loss=2.99, v_num=0]
Epoch 36:  61%|██████    | 59/97 [00:09<00:06,  6.28it/s, loss=3, v_num=0]
Epoch 36:  62%|██████▏   | 60/97 [00:09<00:05,  6.33it/s, loss=3, v_num=0]
Epoch 36:  62%|██████▏   | 60/97 [00:09<00:05,  6.29it/s, loss=3, v_num=0]
Epoch 36:  63%|██████▎   | 61/97 [00:09<00:05,  6.34it/s, loss=3, v_num=0]
Epoch 36:  63%|██████▎   | 61/97 [00:09<00:05,  6.31it/s, loss=2.99, v_num=0]
Epoch 36:  64%|██████▍   | 62/97 [00:09<00:05,  6.36it/s, loss=2.99, v_num=0]
Epoch 36:  64%|██████▍   | 62/97 [00:09<00:05,  6.32it/s, loss=2.99, v_num=0]
Epoch 36:  65%|██████▍   | 63/97 [00:09<00:05,  6.37it/s, loss=2.99, v_num=0]
Epoch 36:  65%|██████▍   | 63/97 [00:09<00:05,  6.33it/s, loss=2.99, v_num=0]
Epoch 36:  66%|██████▌   | 64/97 [00:10<00:05,  6.38it/s, loss=2.99, v_num=0]
Epoch 36:  66%|██████▌   | 64/97 [00:10<00:05,  6.35it/s, loss=2.99, v_num=0]
Epoch 36:  67%|██████▋   | 65/97 [00:10<00:05,  6.40it/s, loss=2.99, v_num=0]
Epoch 36:  67%|██████▋   | 65/97 [00:10<00:05,  6.36it/s, loss=2.98, v_num=0]
Epoch 36:  68%|██████▊   | 66/97 [00:10<00:04,  6.41it/s, loss=2.98, v_num=0]
Epoch 36:  68%|██████▊   | 66/97 [00:10<00:04,  6.37it/s, loss=2.98, v_num=0]
Epoch 36:  69%|██████▉   | 67/97 [00:10<00:04,  6.42it/s, loss=2.98, v_num=0]
Epoch 36:  69%|██████▉   | 67/97 [00:10<00:04,  6.39it/s, loss=2.98, v_num=0]
Epoch 36:  70%|███████   | 68/97 [00:10<00:04,  6.43it/s, loss=2.98, v_num=0]
Epoch 36:  70%|███████   | 68/97 [00:10<00:04,  6.40it/s, loss=2.97, v_num=0]
Epoch 36:  71%|███████   | 69/97 [00:10<00:04,  6.44it/s, loss=2.97, v_num=0]
Epoch 36:  71%|███████   | 69/97 [00:10<00:04,  6.41it/s, loss=2.97, v_num=0]
Epoch 36:  72%|███████▏  | 70/97 [00:10<00:04,  6.45it/s, loss=2.97, v_num=0]
Epoch 36:  72%|███████▏  | 70/97 [00:10<00:04,  6.42it/s, loss=2.97, v_num=0]
Epoch 36:  73%|███████▎  | 71/97 [00:10<00:04,  6.47it/s, loss=2.97, v_num=0]
Epoch 36:  73%|███████▎  | 71/97 [00:11<00:04,  6.43it/s, loss=2.97, v_num=0]
Epoch 36:  74%|███████▍  | 72/97 [00:11<00:03,  6.47it/s, loss=2.97, v_num=0]
Epoch 36:  74%|███████▍  | 72/97 [00:11<00:03,  6.44it/s, loss=2.96, v_num=0]
Epoch 36:  75%|███████▌  | 73/97 [00:11<00:03,  6.49it/s, loss=2.96, v_num=0]
Epoch 36:  75%|███████▌  | 73/97 [00:11<00:03,  6.45it/s, loss=2.96, v_num=0]
Epoch 36:  76%|███████▋  | 74/97 [00:11<00:03,  6.49it/s, loss=2.96, v_num=0]
Epoch 36:  76%|███████▋  | 74/97 [00:11<00:03,  6.46it/s, loss=2.96, v_num=0]
Epoch 36:  77%|███████▋  | 75/97 [00:11<00:03,  6.50it/s, loss=2.96, v_num=0]
Epoch 36:  77%|███████▋  | 75/97 [00:11<00:03,  6.47it/s, loss=2.95, v_num=0]
Epoch 36:  78%|███████▊  | 76/97 [00:11<00:03,  6.51it/s, loss=2.95, v_num=0]
Epoch 36:  78%|███████▊  | 76/97 [00:11<00:03,  6.48it/s, loss=2.95, v_num=0]
Epoch 36:  79%|███████▉  | 77/97 [00:11<00:03,  6.52it/s, loss=2.95, v_num=0]
Epoch 36:  79%|███████▉  | 77/97 [00:11<00:03,  6.49it/s, loss=2.96, v_num=0]
Epoch 36:  80%|████████  | 78/97 [00:11<00:02,  6.53it/s, loss=2.96, v_num=0]
Epoch 36:  80%|████████  | 78/97 [00:11<00:02,  6.50it/s, loss=2.97, v_num=0]
Epoch 36:  81%|████████▏ | 79/97 [00:12<00:02,  6.54it/s, loss=2.97, v_num=0]
Epoch 36:  81%|████████▏ | 79/97 [00:12<00:02,  6.51it/s, loss=2.96, v_num=0]
Epoch 36:  82%|████████▏ | 80/97 [00:12<00:02,  6.55it/s, loss=2.96, v_num=0]
Epoch 36:  82%|████████▏ | 80/97 [00:12<00:02,  6.52it/s, loss=2.95, v_num=0]
Epoch 36:  84%|████████▎ | 81/97 [00:12<00:02,  6.56it/s, loss=2.95, v_num=0]
Epoch 36:  84%|████████▎ | 81/97 [00:12<00:02,  6.53it/s, loss=2.96, v_num=0]
Epoch 36:  85%|████████▍ | 82/97 [00:12<00:02,  6.57it/s, loss=2.96, v_num=0]
Epoch 36:  85%|████████▍ | 82/97 [00:12<00:02,  6.54it/s, loss=2.96, v_num=0]
Epoch 36:  86%|████████▌ | 83/97 [00:12<00:02,  6.57it/s, loss=2.96, v_num=0]
Epoch 36:  86%|████████▌ | 83/97 [00:12<00:02,  6.54it/s, loss=2.96, v_num=0]
Epoch 36:  87%|████████▋ | 84/97 [00:12<00:01,  6.58it/s, loss=2.96, v_num=0]
Epoch 36:  87%|████████▋ | 84/97 [00:12<00:01,  6.55it/s, loss=2.95, v_num=0]
Epoch 36:  88%|████████▊ | 85/97 [00:12<00:01,  6.59it/s, loss=2.95, v_num=0]
Epoch 36:  88%|████████▊ | 85/97 [00:12<00:01,  6.56it/s, loss=2.95, v_num=0]
Epoch 36:  89%|████████▊ | 86/97 [00:13<00:01,  6.60it/s, loss=2.95, v_num=0]
Epoch 36:  89%|████████▊ | 86/97 [00:13<00:01,  6.57it/s, loss=2.95, v_num=0]
Epoch 36:  90%|████████▉ | 87/97 [00:13<00:01,  6.61it/s, loss=2.95, v_num=0]
Epoch 36:  90%|████████▉ | 87/97 [00:13<00:01,  6.58it/s, loss=2.95, v_num=0]
Epoch 36:  91%|█████████ | 88/97 [00:13<00:01,  6.61it/s, loss=2.95, v_num=0]
Epoch 36:  91%|█████████ | 88/97 [00:13<00:01,  6.59it/s, loss=2.94, v_num=0]
Epoch 36:  92%|█████████▏| 89/97 [00:13<00:01,  6.62it/s, loss=2.94, v_num=0]
Epoch 36:  92%|█████████▏| 89/97 [00:13<00:01,  6.59it/s, loss=2.95, v_num=0]
Epoch 36:  93%|█████████▎| 90/97 [00:13<00:01,  6.63it/s, loss=2.95, v_num=0]
Epoch 36:  93%|█████████▎| 90/97 [00:13<00:01,  6.60it/s, loss=2.95, v_num=0]
Epoch 36:  94%|█████████▍| 91/97 [00:13<00:00,  6.64it/s, loss=2.95, v_num=0]
Epoch 36:  94%|█████████▍| 91/97 [00:13<00:00,  6.61it/s, loss=2.95, v_num=0]
Epoch 36:  95%|█████████▍| 92/97 [00:13<00:00,  6.64it/s, loss=2.95, v_num=0]
Epoch 36:  95%|█████████▍| 92/97 [00:13<00:00,  6.62it/s, loss=2.95, v_num=0]
Epoch 36:  96%|█████████▌| 93/97 [00:13<00:00,  6.65it/s, loss=2.95, v_num=0]
Epoch 36:  96%|█████████▌| 93/97 [00:14<00:00,  6.62it/s, loss=2.95, v_num=0]
Epoch 36:  97%|█████████▋| 94/97 [00:14<00:00,  6.65it/s, loss=2.95, v_num=0]
Epoch 36:  97%|█████████▋| 94/97 [00:14<00:00,  6.63it/s, loss=2.95, v_num=0]
Epoch 36:  98%|█████████▊| 95/97 [00:14<00:00,  6.66it/s, loss=2.95, v_num=0]
Epoch 36:  98%|█████████▊| 95/97 [00:14<00:00,  6.63it/s, loss=2.95, v_num=0]
Epoch 36:  99%|█████████▉| 96/97 [00:14<00:00,  6.67it/s, loss=2.95, v_num=0]
Epoch 36:  99%|█████████▉| 96/97 [00:14<00:00,  6.64it/s, loss=2.95, v_num=0]
Epoch 36: 100%|██████████| 97/97 [00:14<00:00,  6.67it/s, loss=2.95, v_num=0]
Epoch 36: 100%|██████████| 97/97 [00:14<00:00,  6.65it/s, loss=2.95, v_num=0]
Epoch 36: 100%|██████████| 97/97 [00:14<00:00,  6.65it/s, loss=2.95, v_num=0]
Epoch 36:   0%|          | 0/97 [00:00<?, ?it/s, loss=2.95, v_num=0]
Epoch 37:   0%|          | 0/97 [00:00<?, ?it/s, loss=2.95, v_num=0]
Epoch 37:   1%|          | 1/97 [00:01<02:15,  1.41s/it, loss=2.95, v_num=0]
Epoch 37:   1%|          | 1/97 [00:01<02:20,  1.47s/it, loss=2.94, v_num=0]
Epoch 37:   2%|▏         | 2/97 [00:01<01:13,  1.29it/s, loss=2.94, v_num=0]
Epoch 37:   2%|▏         | 2/97 [00:01<01:16,  1.25it/s, loss=2.95, v_num=0]
Epoch 37:   3%|▎         | 3/97 [00:01<00:52,  1.78it/s, loss=2.95, v_num=0]
Epoch 37:   3%|▎         | 3/97 [00:01<00:54,  1.73it/s, loss=2.95, v_num=0]
Epoch 37:   4%|▍         | 4/97 [00:01<00:42,  2.20it/s, loss=2.95, v_num=0]
Epoch 37:   4%|▍         | 4/97 [00:01<00:43,  2.13it/s, loss=2.94, v_num=0]
Epoch 37:   5%|▌         | 5/97 [00:01<00:36,  2.55it/s, loss=2.94, v_num=0]
Epoch 37:   5%|▌         | 5/97 [00:02<00:37,  2.49it/s, loss=2.95, v_num=0]
Epoch 37:   6%|▌         | 6/97 [00:02<00:31,  2.86it/s, loss=2.95, v_num=0]
Epoch 37:   6%|▌         | 6/97 [00:02<00:32,  2.79it/s, loss=2.96, v_num=0]
Epoch 37:   7%|▋         | 7/97 [00:02<00:28,  3.14it/s, loss=2.96, v_num=0]
Epoch 37:   7%|▋         | 7/97 [00:02<00:29,  3.06it/s, loss=2.96, v_num=0]
Epoch 37:   8%|▊         | 8/97 [00:02<00:26,  3.38it/s, loss=2.96, v_num=0]
Epoch 37:   8%|▊         | 8/97 [00:02<00:26,  3.30it/s, loss=2.96, v_num=0]
Epoch 37:   9%|▉         | 9/97 [00:02<00:24,  3.58it/s, loss=2.96, v_num=0]
Epoch 37:   9%|▉         | 9/97 [00:02<00:25,  3.50it/s, loss=2.96, v_num=0]
Epoch 37:  10%|█         | 10/97 [00:02<00:23,  3.77it/s, loss=2.96, v_num=0]
Epoch 37:  10%|█         | 10/97 [00:02<00:23,  3.69it/s, loss=2.97, v_num=0]
Epoch 37:  11%|█▏        | 11/97 [00:02<00:21,  3.94it/s, loss=2.97, v_num=0]
Epoch 37:  11%|█▏        | 11/97 [00:02<00:22,  3.87it/s, loss=2.97, v_num=0]
Epoch 37:  12%|█▏        | 12/97 [00:02<00:20,  4.10it/s, loss=2.97, v_num=0]
Epoch 37:  12%|█▏        | 12/97 [00:02<00:21,  4.02it/s, loss=2.97, v_num=0]
Epoch 37:  13%|█▎        | 13/97 [00:03<00:19,  4.25it/s, loss=2.97, v_num=0]
Epoch 37:  13%|█▎        | 13/97 [00:03<00:20,  4.17it/s, loss=2.97, v_num=0]
Epoch 37:  14%|█▍        | 14/97 [00:03<00:18,  4.38it/s, loss=2.97, v_num=0]
Epoch 37:  14%|█▍        | 14/97 [00:03<00:19,  4.30it/s, loss=2.97, v_num=0]
Epoch 37:  15%|█▌        | 15/97 [00:03<00:18,  4.50it/s, loss=2.97, v_num=0]
Epoch 37:  15%|█▌        | 15/97 [00:03<00:18,  4.42it/s, loss=2.97, v_num=0]
Epoch 37:  16%|█▋        | 16/97 [00:03<00:17,  4.62it/s, loss=2.97, v_num=0]
Epoch 37:  16%|█▋        | 16/97 [00:03<00:17,  4.54it/s, loss=2.97, v_num=0]
Epoch 37:  18%|█▊        | 17/97 [00:03<00:16,  4.71it/s, loss=2.97, v_num=0]
Epoch 37:  18%|█▊        | 17/97 [00:03<00:17,  4.64it/s, loss=2.96, v_num=0]
Epoch 37:  19%|█▊        | 18/97 [00:03<00:16,  4.80it/s, loss=2.96, v_num=0]
Epoch 37:  19%|█▊        | 18/97 [00:03<00:16,  4.73it/s, loss=2.96, v_num=0]
Epoch 37:  20%|█▉        | 19/97 [00:03<00:15,  4.90it/s, loss=2.96, v_num=0]
Epoch 37:  20%|█▉        | 19/97 [00:03<00:16,  4.83it/s, loss=2.96, v_num=0]
Epoch 37:  21%|██        | 20/97 [00:04<00:15,  4.98it/s, loss=2.96, v_num=0]
Epoch 37:  21%|██        | 20/97 [00:04<00:15,  4.91it/s, loss=2.95, v_num=0]
Epoch 37:  22%|██▏       | 21/97 [00:04<00:15,  5.06it/s, loss=2.95, v_num=0]
Epoch 37:  22%|██▏       | 21/97 [00:04<00:15,  4.99it/s, loss=2.95, v_num=0]
Epoch 37:  23%|██▎       | 22/97 [00:04<00:14,  5.13it/s, loss=2.95, v_num=0]
Epoch 37:  23%|██▎       | 22/97 [00:04<00:14,  5.06it/s, loss=2.96, v_num=0]
Epoch 37:  24%|██▎       | 23/97 [00:04<00:14,  5.20it/s, loss=2.96, v_num=0]
Epoch 37:  24%|██▎       | 23/97 [00:04<00:14,  5.13it/s, loss=2.96, v_num=0]
Epoch 37:  25%|██▍       | 24/97 [00:04<00:13,  5.26it/s, loss=2.96, v_num=0]
Epoch 37:  25%|██▍       | 24/97 [00:04<00:14,  5.19it/s, loss=2.96, v_num=0]
Epoch 37:  26%|██▌       | 25/97 [00:04<00:13,  5.31it/s, loss=2.96, v_num=0]
Epoch 37:  26%|██▌       | 25/97 [00:04<00:13,  5.25it/s, loss=2.95, v_num=0]
Epoch 37:  27%|██▋       | 26/97 [00:04<00:13,  5.37it/s, loss=2.95, v_num=0]
Epoch 37:  27%|██▋       | 26/97 [00:04<00:13,  5.31it/s, loss=2.95, v_num=0]
Epoch 37:  28%|██▊       | 27/97 [00:04<00:12,  5.43it/s, loss=2.95, v_num=0]
Epoch 37:  28%|██▊       | 27/97 [00:05<00:13,  5.37it/s, loss=2.95, v_num=0]
Epoch 37:  29%|██▉       | 28/97 [00:05<00:12,  5.48it/s, loss=2.95, v_num=0]
Epoch 37:  29%|██▉       | 28/97 [00:05<00:12,  5.42it/s, loss=2.95, v_num=0]
Epoch 37:  30%|██▉       | 29/97 [00:05<00:12,  5.52it/s, loss=2.95, v_num=0]
Epoch 37:  30%|██▉       | 29/97 [00:05<00:12,  5.47it/s, loss=2.94, v_num=0]
Epoch 37:  31%|███       | 30/97 [00:05<00:12,  5.57it/s, loss=2.94, v_num=0]
Epoch 37:  31%|███       | 30/97 [00:05<00:12,  5.51it/s, loss=2.94, v_num=0]
Epoch 37:  32%|███▏      | 31/97 [00:05<00:11,  5.61it/s, loss=2.94, v_num=0]
Epoch 37:  32%|███▏      | 31/97 [00:05<00:11,  5.56it/s, loss=2.93, v_num=0]
Epoch 37:  33%|███▎      | 32/97 [00:05<00:11,  5.65it/s, loss=2.93, v_num=0]
Epoch 37:  33%|███▎      | 32/97 [00:05<00:11,  5.60it/s, loss=2.92, v_num=0]
Epoch 37:  34%|███▍      | 33/97 [00:05<00:11,  5.69it/s, loss=2.92, v_num=0]
Epoch 37:  34%|███▍      | 33/97 [00:05<00:11,  5.64it/s, loss=2.92, v_num=0]
Epoch 37:  35%|███▌      | 34/97 [00:05<00:10,  5.73it/s, loss=2.92, v_num=0]
Epoch 37:  35%|███▌      | 34/97 [00:05<00:11,  5.68it/s, loss=2.91, v_num=0]
Epoch 37:  36%|███▌      | 35/97 [00:06<00:10,  5.77it/s, loss=2.91, v_num=0]
Epoch 37:  36%|███▌      | 35/97 [00:06<00:10,  5.72it/s, loss=2.92, v_num=0]
Epoch 37:  37%|███▋      | 36/97 [00:06<00:10,  5.80it/s, loss=2.92, v_num=0]
Epoch 37:  37%|███▋      | 36/97 [00:06<00:10,  5.75it/s, loss=2.92, v_num=0]
Epoch 37:  38%|███▊      | 37/97 [00:06<00:10,  5.83it/s, loss=2.92, v_num=0]
Epoch 37:  38%|███▊      | 37/97 [00:06<00:10,  5.78it/s, loss=2.93, v_num=0]
Epoch 37:  39%|███▉      | 38/97 [00:06<00:10,  5.86it/s, loss=2.93, v_num=0]
Epoch 37:  39%|███▉      | 38/97 [00:06<00:10,  5.82it/s, loss=2.93, v_num=0]
Epoch 37:  40%|████      | 39/97 [00:06<00:09,  5.89it/s, loss=2.93, v_num=0]
Epoch 37:  40%|████      | 39/97 [00:06<00:09,  5.85it/s, loss=2.93, v_num=0]
Epoch 37:  41%|████      | 40/97 [00:06<00:09,  5.92it/s, loss=2.93, v_num=0]
Epoch 37:  41%|████      | 40/97 [00:06<00:09,  5.87it/s, loss=2.93, v_num=0]
Epoch 37:  42%|████▏     | 41/97 [00:06<00:09,  5.95it/s, loss=2.93, v_num=0]
Epoch 37:  42%|████▏     | 41/97 [00:06<00:09,  5.90it/s, loss=2.93, v_num=0]
Epoch 37:  43%|████▎     | 42/97 [00:07<00:09,  5.98it/s, loss=2.93, v_num=0]
Epoch 37:  43%|████▎     | 42/97 [00:07<00:09,  5.93it/s, loss=2.93, v_num=0]
Epoch 37:  44%|████▍     | 43/97 [00:07<00:08,  6.00it/s, loss=2.93, v_num=0]
Epoch 37:  44%|████▍     | 43/97 [00:07<00:09,  5.95it/s, loss=2.93, v_num=0]
Epoch 37:  45%|████▌     | 44/97 [00:07<00:08,  6.02it/s, loss=2.93, v_num=0]
Epoch 37:  45%|████▌     | 44/97 [00:07<00:08,  5.98it/s, loss=2.93, v_num=0]
Epoch 37:  46%|████▋     | 45/97 [00:07<00:08,  6.05it/s, loss=2.93, v_num=0]
Epoch 37:  46%|████▋     | 45/97 [00:07<00:08,  6.00it/s, loss=2.94, v_num=0]
Epoch 37:  47%|████▋     | 46/97 [00:07<00:08,  6.07it/s, loss=2.94, v_num=0]
Epoch 37:  47%|████▋     | 46/97 [00:07<00:08,  6.03it/s, loss=2.94, v_num=0]
Epoch 37:  48%|████▊     | 47/97 [00:07<00:08,  6.09it/s, loss=2.94, v_num=0]
Epoch 37:  48%|████▊     | 47/97 [00:07<00:08,  6.05it/s, loss=2.93, v_num=0]
Epoch 37:  49%|████▉     | 48/97 [00:07<00:08,  6.11it/s, loss=2.93, v_num=0]
Epoch 37:  49%|████▉     | 48/97 [00:07<00:08,  6.07it/s, loss=2.93, v_num=0]
Epoch 37:  51%|█████     | 49/97 [00:07<00:07,  6.13it/s, loss=2.93, v_num=0]
Epoch 37:  51%|█████     | 49/97 [00:08<00:07,  6.09it/s, loss=2.94, v_num=0]
Epoch 37:  52%|█████▏    | 50/97 [00:08<00:07,  6.15it/s, loss=2.94, v_num=0]
Epoch 37:  52%|█████▏    | 50/97 [00:08<00:07,  6.11it/s, loss=2.93, v_num=0]
Epoch 37:  53%|█████▎    | 51/97 [00:08<00:07,  6.17it/s, loss=2.93, v_num=0]
Epoch 37:  53%|█████▎    | 51/97 [00:08<00:07,  6.13it/s, loss=2.94, v_num=0]
Epoch 37:  54%|█████▎    | 52/97 [00:08<00:07,  6.19it/s, loss=2.94, v_num=0]
Epoch 37:  54%|█████▎    | 52/97 [00:08<00:07,  6.15it/s, loss=2.94, v_num=0]
Epoch 37:  55%|█████▍    | 53/97 [00:08<00:07,  6.20it/s, loss=2.94, v_num=0]
Epoch 37:  55%|█████▍    | 53/97 [00:08<00:07,  6.16it/s, loss=2.94, v_num=0]
Epoch 37:  56%|█████▌    | 54/97 [00:08<00:06,  6.22it/s, loss=2.94, v_num=0]
Epoch 37:  56%|█████▌    | 54/97 [00:08<00:06,  6.18it/s, loss=2.94, v_num=0]
Epoch 37:  57%|█████▋    | 55/97 [00:08<00:06,  6.23it/s, loss=2.94, v_num=0]
Epoch 37:  57%|█████▋    | 55/97 [00:08<00:06,  6.19it/s, loss=2.94, v_num=0]
Epoch 37:  58%|█████▊    | 56/97 [00:08<00:06,  6.25it/s, loss=2.94, v_num=0]
Epoch 37:  58%|█████▊    | 56/97 [00:09<00:06,  6.21it/s, loss=2.94, v_num=0]
Epoch 37:  59%|█████▉    | 57/97 [00:09<00:06,  6.26it/s, loss=2.94, v_num=0]
Epoch 37:  59%|█████▉    | 57/97 [00:09<00:06,  6.22it/s, loss=2.94, v_num=0]
Epoch 37:  60%|█████▉    | 58/97 [00:09<00:06,  6.28it/s, loss=2.94, v_num=0]
Epoch 37:  60%|█████▉    | 58/97 [00:09<00:06,  6.24it/s, loss=2.93, v_num=0]
Epoch 37:  61%|██████    | 59/97 [00:09<00:06,  6.29it/s, loss=2.93, v_num=0]
Epoch 37:  61%|██████    | 59/97 [00:09<00:06,  6.26it/s, loss=2.93, v_num=0]
Epoch 37:  62%|██████▏   | 60/97 [00:09<00:05,  6.31it/s, loss=2.93, v_num=0]
Epoch 37:  62%|██████▏   | 60/97 [00:09<00:05,  6.27it/s, loss=2.93, v_num=0]
Epoch 37:  63%|██████▎   | 61/97 [00:09<00:05,  6.32it/s, loss=2.93, v_num=0]
Epoch 37:  63%|██████▎   | 61/97 [00:09<00:05,  6.28it/s, loss=2.93, v_num=0]
Epoch 37:  64%|██████▍   | 62/97 [00:09<00:05,  6.33it/s, loss=2.93, v_num=0]
Epoch 37:  64%|██████▍   | 62/97 [00:09<00:05,  6.30it/s, loss=2.92, v_num=0]
Epoch 37:  65%|██████▍   | 63/97 [00:09<00:05,  6.35it/s, loss=2.92, v_num=0]
Epoch 37:  65%|██████▍   | 63/97 [00:09<00:05,  6.31it/s, loss=2.92, v_num=0]
Epoch 37:  66%|██████▌   | 64/97 [00:10<00:05,  6.36it/s, loss=2.92, v_num=0]
Epoch 37:  66%|██████▌   | 64/97 [00:10<00:05,  6.32it/s, loss=2.91, v_num=0]
Epoch 37:  67%|██████▋   | 65/97 [00:10<00:05,  6.37it/s, loss=2.91, v_num=0]
Epoch 37:  67%|██████▋   | 65/97 [00:10<00:05,  6.34it/s, loss=2.91, v_num=0]
Epoch 37:  68%|██████▊   | 66/97 [00:10<00:04,  6.38it/s, loss=2.91, v_num=0]
Epoch 37:  68%|██████▊   | 66/97 [00:10<00:04,  6.35it/s, loss=2.91, v_num=0]
Epoch 37:  69%|██████▉   | 67/97 [00:10<00:04,  6.40it/s, loss=2.91, v_num=0]
Epoch 37:  69%|██████▉   | 67/97 [00:10<00:04,  6.36it/s, loss=2.91, v_num=0]
Epoch 37:  70%|███████   | 68/97 [00:10<00:04,  6.41it/s, loss=2.91, v_num=0]
Epoch 37:  70%|███████   | 68/97 [00:10<00:04,  6.37it/s, loss=2.91, v_num=0]
Epoch 37:  71%|███████   | 69/97 [00:10<00:04,  6.42it/s, loss=2.91, v_num=0]
Epoch 37:  71%|███████   | 69/97 [00:10<00:04,  6.39it/s, loss=2.9, v_num=0]
Epoch 37:  72%|███████▏  | 70/97 [00:10<00:04,  6.43it/s, loss=2.9, v_num=0]
Epoch 37:  72%|███████▏  | 70/97 [00:10<00:04,  6.40it/s, loss=2.91, v_num=0]
Epoch 37:  73%|███████▎  | 71/97 [00:11<00:04,  6.44it/s, loss=2.91, v_num=0]
Epoch 37:  73%|███████▎  | 71/97 [00:11<00:04,  6.41it/s, loss=2.91, v_num=0]
Epoch 37:  74%|███████▍  | 72/97 [00:11<00:03,  6.45it/s, loss=2.91, v_num=0]
Epoch 37:  74%|███████▍  | 72/97 [00:11<00:03,  6.42it/s, loss=2.9, v_num=0]
Epoch 37:  75%|███████▌  | 73/97 [00:11<00:03,  6.46it/s, loss=2.9, v_num=0]
Epoch 37:  75%|███████▌  | 73/97 [00:11<00:03,  6.43it/s, loss=2.9, v_num=0]
Epoch 37:  76%|███████▋  | 74/97 [00:11<00:03,  6.47it/s, loss=2.9, v_num=0]
Epoch 37:  76%|███████▋  | 74/97 [00:11<00:03,  6.44it/s, loss=2.9, v_num=0]
Epoch 37:  77%|███████▋  | 75/97 [00:11<00:03,  6.48it/s, loss=2.9, v_num=0]
Epoch 37:  77%|███████▋  | 75/97 [00:11<00:03,  6.45it/s, loss=2.9, v_num=0]
Epoch 37:  78%|███████▊  | 76/97 [00:11<00:03,  6.49it/s, loss=2.9, v_num=0]
Epoch 37:  78%|███████▊  | 76/97 [00:11<00:03,  6.46it/s, loss=2.9, v_num=0]
Epoch 37:  79%|███████▉  | 77/97 [00:11<00:03,  6.50it/s, loss=2.9, v_num=0]
Epoch 37:  79%|███████▉  | 77/97 [00:11<00:03,  6.47it/s, loss=2.89, v_num=0]
Epoch 37:  80%|████████  | 78/97 [00:11<00:02,  6.51it/s, loss=2.89, v_num=0]
Epoch 37:  80%|████████  | 78/97 [00:12<00:02,  6.48it/s, loss=2.89, v_num=0]
Epoch 37:  81%|████████▏ | 79/97 [00:12<00:02,  6.52it/s, loss=2.89, v_num=0]
Epoch 37:  81%|████████▏ | 79/97 [00:12<00:02,  6.49it/s, loss=2.9, v_num=0]
Epoch 37:  82%|████████▏ | 80/97 [00:12<00:02,  6.53it/s, loss=2.9, v_num=0]
Epoch 37:  82%|████████▏ | 80/97 [00:12<00:02,  6.50it/s, loss=2.9, v_num=0]
Epoch 37:  84%|████████▎ | 81/97 [00:12<00:02,  6.54it/s, loss=2.9, v_num=0]
Epoch 37:  84%|████████▎ | 81/97 [00:12<00:02,  6.51it/s, loss=2.9, v_num=0]
Epoch 37:  85%|████████▍ | 82/97 [00:12<00:02,  6.54it/s, loss=2.9, v_num=0]
Epoch 37:  85%|████████▍ | 82/97 [00:12<00:02,  6.51it/s, loss=2.9, v_num=0]
Epoch 37:  86%|████████▌ | 83/97 [00:12<00:02,  6.55it/s, loss=2.9, v_num=0]
Epoch 37:  86%|████████▌ | 83/97 [00:12<00:02,  6.52it/s, loss=2.89, v_num=0]
Epoch 37:  87%|████████▋ | 84/97 [00:12<00:01,  6.56it/s, loss=2.89, v_num=0]
Epoch 37:  87%|████████▋ | 84/97 [00:12<00:01,  6.53it/s, loss=2.91, v_num=0]
Epoch 37:  88%|████████▊ | 85/97 [00:12<00:01,  6.57it/s, loss=2.91, v_num=0]
Epoch 37:  88%|████████▊ | 85/97 [00:12<00:01,  6.54it/s, loss=2.91, v_num=0]
Epoch 37:  89%|████████▊ | 86/97 [00:13<00:01,  6.57it/s, loss=2.91, v_num=0]
Epoch 37:  89%|████████▊ | 86/97 [00:13<00:01,  6.55it/s, loss=2.91, v_num=0]
Epoch 37:  90%|████████▉ | 87/97 [00:13<00:01,  6.59it/s, loss=2.91, v_num=0]
Epoch 37:  90%|████████▉ | 87/97 [00:13<00:01,  6.56it/s, loss=2.91, v_num=0]
Epoch 37:  91%|█████████ | 88/97 [00:13<00:01,  6.59it/s, loss=2.91, v_num=0]
Epoch 37:  91%|█████████ | 88/97 [00:13<00:01,  6.56it/s, loss=2.91, v_num=0]
Epoch 37:  92%|█████████▏| 89/97 [00:13<00:01,  6.60it/s, loss=2.91, v_num=0]
Epoch 37:  92%|█████████▏| 89/97 [00:13<00:01,  6.57it/s, loss=2.91, v_num=0]
Epoch 37:  93%|█████████▎| 90/97 [00:13<00:01,  6.61it/s, loss=2.91, v_num=0]
Epoch 37:  93%|█████████▎| 90/97 [00:13<00:01,  6.58it/s, loss=2.9, v_num=0]
Epoch 37:  94%|█████████▍| 91/97 [00:13<00:00,  6.61it/s, loss=2.9, v_num=0]
Epoch 37:  94%|█████████▍| 91/97 [00:13<00:00,  6.59it/s, loss=2.9, v_num=0]
Epoch 37:  95%|█████████▍| 92/97 [00:13<00:00,  6.62it/s, loss=2.9, v_num=0]
Epoch 37:  95%|█████████▍| 92/97 [00:13<00:00,  6.59it/s, loss=2.91, v_num=0]
Epoch 37:  96%|█████████▌| 93/97 [00:14<00:00,  6.63it/s, loss=2.91, v_num=0]
Epoch 37:  96%|█████████▌| 93/97 [00:14<00:00,  6.60it/s, loss=2.92, v_num=0]
Epoch 37:  97%|█████████▋| 94/97 [00:14<00:00,  6.63it/s, loss=2.92, v_num=0]
Epoch 37:  97%|█████████▋| 94/97 [00:14<00:00,  6.61it/s, loss=2.92, v_num=0]
Epoch 37:  98%|█████████▊| 95/97 [00:14<00:00,  6.64it/s, loss=2.92, v_num=0]
Epoch 37:  98%|█████████▊| 95/97 [00:14<00:00,  6.61it/s, loss=2.93, v_num=0]
Epoch 37:  99%|█████████▉| 96/97 [00:14<00:00,  6.64it/s, loss=2.93, v_num=0]
Epoch 37:  99%|█████████▉| 96/97 [00:14<00:00,  6.62it/s, loss=2.93, v_num=0]
Epoch 37: 100%|██████████| 97/97 [00:14<00:00,  6.65it/s, loss=2.93, v_num=0]
Epoch 37: 100%|██████████| 97/97 [00:14<00:00,  6.63it/s, loss=2.93, v_num=0]
Epoch 37: 100%|██████████| 97/97 [00:14<00:00,  6.63it/s, loss=2.93, v_num=0]
Epoch 37:   0%|          | 0/97 [00:00<?, ?it/s, loss=2.93, v_num=0]
Epoch 38:   0%|          | 0/97 [00:00<?, ?it/s, loss=2.93, v_num=0]
Epoch 38:   1%|          | 1/97 [00:01<02:14,  1.41s/it, loss=2.93, v_num=0]
Epoch 38:   1%|          | 1/97 [00:01<02:20,  1.46s/it, loss=2.93, v_num=0]
Epoch 38:   2%|▏         | 2/97 [00:01<01:13,  1.30it/s, loss=2.93, v_num=0]
Epoch 38:   2%|▏         | 2/97 [00:01<01:16,  1.25it/s, loss=2.93, v_num=0]
Epoch 38:   3%|▎         | 3/97 [00:01<00:52,  1.78it/s, loss=2.93, v_num=0]
Epoch 38:   3%|▎         | 3/97 [00:01<00:54,  1.73it/s, loss=2.93, v_num=0]
Epoch 38:   4%|▍         | 4/97 [00:01<00:42,  2.20it/s, loss=2.93, v_num=0]
Epoch 38:   4%|▍         | 4/97 [00:01<00:43,  2.14it/s, loss=2.93, v_num=0]
Epoch 38:   5%|▌         | 5/97 [00:01<00:35,  2.57it/s, loss=2.93, v_num=0]
Epoch 38:   5%|▌         | 5/97 [00:02<00:36,  2.49it/s, loss=2.93, v_num=0]
Epoch 38:   6%|▌         | 6/97 [00:02<00:31,  2.87it/s, loss=2.93, v_num=0]
Epoch 38:   6%|▌         | 6/97 [00:02<00:32,  2.80it/s, loss=2.93, v_num=0]
Epoch 38:   7%|▋         | 7/97 [00:02<00:28,  3.15it/s, loss=2.93, v_num=0]
Epoch 38:   7%|▋         | 7/97 [00:02<00:29,  3.07it/s, loss=2.92, v_num=0]
Epoch 38:   8%|▊         | 8/97 [00:02<00:26,  3.40it/s, loss=2.92, v_num=0]
Epoch 38:   8%|▊         | 8/97 [00:02<00:26,  3.31it/s, loss=2.93, v_num=0]
Epoch 38:   9%|▉         | 9/97 [00:02<00:24,  3.59it/s, loss=2.93, v_num=0]
Epoch 38:   9%|▉         | 9/97 [00:02<00:25,  3.51it/s, loss=2.92, v_num=0]
Epoch 38:  10%|█         | 10/97 [00:02<00:23,  3.78it/s, loss=2.92, v_num=0]
Epoch 38:  10%|█         | 10/97 [00:02<00:23,  3.70it/s, loss=2.92, v_num=0]
Epoch 38:  11%|█▏        | 11/97 [00:02<00:21,  3.96it/s, loss=2.92, v_num=0]
Epoch 38:  11%|█▏        | 11/97 [00:02<00:22,  3.88it/s, loss=2.93, v_num=0]
Epoch 38:  12%|█▏        | 12/97 [00:02<00:20,  4.12it/s, loss=2.93, v_num=0]
Epoch 38:  12%|█▏        | 12/97 [00:02<00:21,  4.04it/s, loss=2.93, v_num=0]
Epoch 38:  13%|█▎        | 13/97 [00:03<00:19,  4.26it/s, loss=2.93, v_num=0]
Epoch 38:  13%|█▎        | 13/97 [00:03<00:20,  4.19it/s, loss=2.93, v_num=0]
Epoch 38:  14%|█▍        | 14/97 [00:03<00:18,  4.40it/s, loss=2.93, v_num=0]
Epoch 38:  14%|█▍        | 14/97 [00:03<00:19,  4.32it/s, loss=2.93, v_num=0]
Epoch 38:  15%|█▌        | 15/97 [00:03<00:18,  4.52it/s, loss=2.93, v_num=0]
Epoch 38:  15%|█▌        | 15/97 [00:03<00:18,  4.44it/s, loss=2.93, v_num=0]
Epoch 38:  16%|█▋        | 16/97 [00:03<00:17,  4.63it/s, loss=2.93, v_num=0]
Epoch 38:  16%|█▋        | 16/97 [00:03<00:17,  4.55it/s, loss=2.92, v_num=0]
Epoch 38:  18%|█▊        | 17/97 [00:03<00:16,  4.72it/s, loss=2.92, v_num=0]
Epoch 38:  18%|█▊        | 17/97 [00:03<00:17,  4.65it/s, loss=2.92, v_num=0]
Epoch 38:  19%|█▊        | 18/97 [00:03<00:16,  4.82it/s, loss=2.92, v_num=0]
Epoch 38:  19%|█▊        | 18/97 [00:03<00:16,  4.75it/s, loss=2.92, v_num=0]
Epoch 38:  20%|█▉        | 19/97 [00:03<00:15,  4.91it/s, loss=2.92, v_num=0]
Epoch 38:  20%|█▉        | 19/97 [00:03<00:16,  4.84it/s, loss=2.91, v_num=0]
Epoch 38:  21%|██        | 20/97 [00:04<00:15,  4.99it/s, loss=2.91, v_num=0]
Epoch 38:  21%|██        | 20/97 [00:04<00:15,  4.92it/s, loss=2.91, v_num=0]
Epoch 38:  22%|██▏       | 21/97 [00:04<00:14,  5.07it/s, loss=2.91, v_num=0]
Epoch 38:  22%|██▏       | 21/97 [00:04<00:15,  5.00it/s, loss=2.91, v_num=0]
Epoch 38:  23%|██▎       | 22/97 [00:04<00:14,  5.13it/s, loss=2.91, v_num=0]
Epoch 38:  23%|██▎       | 22/97 [00:04<00:14,  5.07it/s, loss=2.91, v_num=0]
Epoch 38:  24%|██▎       | 23/97 [00:04<00:14,  5.20it/s, loss=2.91, v_num=0]
Epoch 38:  24%|██▎       | 23/97 [00:04<00:14,  5.14it/s, loss=2.91, v_num=0]
Epoch 38:  25%|██▍       | 24/97 [00:04<00:13,  5.26it/s, loss=2.91, v_num=0]
Epoch 38:  25%|██▍       | 24/97 [00:04<00:14,  5.20it/s, loss=2.91, v_num=0]
Epoch 38:  26%|██▌       | 25/97 [00:04<00:13,  5.32it/s, loss=2.91, v_num=0]
Epoch 38:  26%|██▌       | 25/97 [00:04<00:13,  5.26it/s, loss=2.91, v_num=0]
Epoch 38:  27%|██▋       | 26/97 [00:04<00:13,  5.37it/s, loss=2.91, v_num=0]
Epoch 38:  27%|██▋       | 26/97 [00:04<00:13,  5.31it/s, loss=2.9, v_num=0]
Epoch 38:  28%|██▊       | 27/97 [00:04<00:12,  5.43it/s, loss=2.9, v_num=0]
Epoch 38:  28%|██▊       | 27/97 [00:05<00:13,  5.37it/s, loss=2.9, v_num=0]
Epoch 38:  29%|██▉       | 28/97 [00:05<00:12,  5.48it/s, loss=2.9, v_num=0]
Epoch 38:  29%|██▉       | 28/97 [00:05<00:12,  5.42it/s, loss=2.89, v_num=0]
Epoch 38:  30%|██▉       | 29/97 [00:05<00:12,  5.53it/s, loss=2.89, v_num=0]
Epoch 38:  30%|██▉       | 29/97 [00:05<00:12,  5.47it/s, loss=2.9, v_num=0]
Epoch 38:  31%|███       | 30/97 [00:05<00:12,  5.57it/s, loss=2.9, v_num=0]
Epoch 38:  31%|███       | 30/97 [00:05<00:12,  5.51it/s, loss=2.9, v_num=0]
Epoch 38:  32%|███▏      | 31/97 [00:05<00:11,  5.61it/s, loss=2.9, v_num=0]
Epoch 38:  32%|███▏      | 31/97 [00:05<00:11,  5.56it/s, loss=2.9, v_num=0]
Epoch 38:  33%|███▎      | 32/97 [00:05<00:11,  5.66it/s, loss=2.9, v_num=0]
Epoch 38:  33%|███▎      | 32/97 [00:05<00:11,  5.60it/s, loss=2.9, v_num=0]
Epoch 38:  34%|███▍      | 33/97 [00:05<00:11,  5.69it/s, loss=2.9, v_num=0]
Epoch 38:  34%|███▍      | 33/97 [00:05<00:11,  5.64it/s, loss=2.9, v_num=0]
Epoch 38:  35%|███▌      | 34/97 [00:05<00:10,  5.73it/s, loss=2.9, v_num=0]
Epoch 38:  35%|███▌      | 34/97 [00:05<00:11,  5.68it/s, loss=2.9, v_num=0]
Epoch 38:  36%|███▌      | 35/97 [00:06<00:10,  5.76it/s, loss=2.9, v_num=0]
Epoch 38:  36%|███▌      | 35/97 [00:06<00:10,  5.71it/s, loss=2.9, v_num=0]
Epoch 38:  37%|███▋      | 36/97 [00:06<00:10,  5.80it/s, loss=2.9, v_num=0]
Epoch 38:  37%|███▋      | 36/97 [00:06<00:10,  5.74it/s, loss=2.9, v_num=0]
Epoch 38:  38%|███▊      | 37/97 [00:06<00:10,  5.82it/s, loss=2.9, v_num=0]
Epoch 38:  38%|███▊      | 37/97 [00:06<00:10,  5.77it/s, loss=2.89, v_num=0]
Epoch 38:  39%|███▉      | 38/97 [00:06<00:10,  5.86it/s, loss=2.89, v_num=0]
Epoch 38:  39%|███▉      | 38/97 [00:06<00:10,  5.80it/s, loss=2.9, v_num=0]
Epoch 38:  40%|████      | 39/97 [00:06<00:09,  5.88it/s, loss=2.9, v_num=0]
Epoch 38:  40%|████      | 39/97 [00:06<00:09,  5.84it/s, loss=2.9, v_num=0]
Epoch 38:  41%|████      | 40/97 [00:06<00:09,  5.91it/s, loss=2.9, v_num=0]
Epoch 38:  41%|████      | 40/97 [00:06<00:09,  5.87it/s, loss=2.9, v_num=0]
Epoch 38:  42%|████▏     | 41/97 [00:06<00:09,  5.94it/s, loss=2.9, v_num=0]
Epoch 38:  42%|████▏     | 41/97 [00:06<00:09,  5.89it/s, loss=2.9, v_num=0]
Epoch 38:  43%|████▎     | 42/97 [00:07<00:09,  5.97it/s, loss=2.9, v_num=0]
Epoch 38:  43%|████▎     | 42/97 [00:07<00:09,  5.92it/s, loss=2.9, v_num=0]
Epoch 38:  44%|████▍     | 43/97 [00:07<00:09,  5.99it/s, loss=2.9, v_num=0]
Epoch 38:  44%|████▍     | 43/97 [00:07<00:09,  5.95it/s, loss=2.89, v_num=0]
Epoch 38:  45%|████▌     | 44/97 [00:07<00:08,  6.02it/s, loss=2.89, v_num=0]
Epoch 38:  45%|████▌     | 44/97 [00:07<00:08,  5.97it/s, loss=2.89, v_num=0]
Epoch 38:  46%|████▋     | 45/97 [00:07<00:08,  6.04it/s, loss=2.89, v_num=0]
Epoch 38:  46%|████▋     | 45/97 [00:07<00:08,  6.00it/s, loss=2.9, v_num=0]
Epoch 38:  47%|████▋     | 46/97 [00:07<00:08,  6.06it/s, loss=2.9, v_num=0]
Epoch 38:  47%|████▋     | 46/97 [00:07<00:08,  6.02it/s, loss=2.9, v_num=0]
Epoch 38:  48%|████▊     | 47/97 [00:07<00:08,  6.09it/s, loss=2.9, v_num=0]
Epoch 38:  48%|████▊     | 47/97 [00:07<00:08,  6.04it/s, loss=2.89, v_num=0]
Epoch 38:  49%|████▉     | 48/97 [00:07<00:08,  6.11it/s, loss=2.89, v_num=0]
Epoch 38:  49%|████▉     | 48/97 [00:07<00:08,  6.06it/s, loss=2.88, v_num=0]
Epoch 38:  51%|█████     | 49/97 [00:08<00:07,  6.12it/s, loss=2.88, v_num=0]
Epoch 38:  51%|█████     | 49/97 [00:08<00:07,  6.08it/s, loss=2.88, v_num=0]
Epoch 38:  52%|█████▏    | 50/97 [00:08<00:07,  6.15it/s, loss=2.88, v_num=0]
Epoch 38:  52%|█████▏    | 50/97 [00:08<00:07,  6.10it/s, loss=2.87, v_num=0]
Epoch 38:  53%|█████▎    | 51/97 [00:08<00:07,  6.16it/s, loss=2.87, v_num=0]
Epoch 38:  53%|█████▎    | 51/97 [00:08<00:07,  6.12it/s, loss=2.87, v_num=0]
Epoch 38:  54%|█████▎    | 52/97 [00:08<00:07,  6.18it/s, loss=2.87, v_num=0]
Epoch 38:  54%|█████▎    | 52/97 [00:08<00:07,  6.14it/s, loss=2.87, v_num=0]
Epoch 38:  55%|█████▍    | 53/97 [00:08<00:07,  6.20it/s, loss=2.87, v_num=0]
Epoch 38:  55%|█████▍    | 53/97 [00:08<00:07,  6.16it/s, loss=2.87, v_num=0]
Epoch 38:  56%|█████▌    | 54/97 [00:08<00:06,  6.21it/s, loss=2.87, v_num=0]
Epoch 38:  56%|█████▌    | 54/97 [00:08<00:06,  6.17it/s, loss=2.88, v_num=0]
Epoch 38:  57%|█████▋    | 55/97 [00:08<00:06,  6.23it/s, loss=2.88, v_num=0]
Epoch 38:  57%|█████▋    | 55/97 [00:08<00:06,  6.19it/s, loss=2.88, v_num=0]
Epoch 38:  58%|█████▊    | 56/97 [00:08<00:06,  6.24it/s, loss=2.88, v_num=0]
Epoch 38:  58%|█████▊    | 56/97 [00:09<00:06,  6.20it/s, loss=2.88, v_num=0]
Epoch 38:  59%|█████▉    | 57/97 [00:09<00:06,  6.26it/s, loss=2.88, v_num=0]
Epoch 38:  59%|█████▉    | 57/97 [00:09<00:06,  6.22it/s, loss=2.88, v_num=0]
Epoch 38:  60%|█████▉    | 58/97 [00:09<00:06,  6.27it/s, loss=2.88, v_num=0]
Epoch 38:  60%|█████▉    | 58/97 [00:09<00:06,  6.23it/s, loss=2.88, v_num=0]
Epoch 38:  61%|██████    | 59/97 [00:09<00:06,  6.29it/s, loss=2.88, v_num=0]
Epoch 38:  61%|██████    | 59/97 [00:09<00:06,  6.25it/s, loss=2.88, v_num=0]
Epoch 38:  62%|██████▏   | 60/97 [00:09<00:05,  6.30it/s, loss=2.88, v_num=0]
Epoch 38:  62%|██████▏   | 60/97 [00:09<00:05,  6.26it/s, loss=2.88, v_num=0]
Epoch 38:  63%|██████▎   | 61/97 [00:09<00:05,  6.31it/s, loss=2.88, v_num=0]
Epoch 38:  63%|██████▎   | 61/97 [00:09<00:05,  6.27it/s, loss=2.88, v_num=0]
Epoch 38:  64%|██████▍   | 62/97 [00:09<00:05,  6.32it/s, loss=2.88, v_num=0]
Epoch 38:  64%|██████▍   | 62/97 [00:09<00:05,  6.29it/s, loss=2.87, v_num=0]
Epoch 38:  65%|██████▍   | 63/97 [00:09<00:05,  6.33it/s, loss=2.87, v_num=0]
Epoch 38:  65%|██████▍   | 63/97 [00:10<00:05,  6.30it/s, loss=2.88, v_num=0]
Epoch 38:  66%|██████▌   | 64/97 [00:10<00:05,  6.35it/s, loss=2.88, v_num=0]
Epoch 38:  66%|██████▌   | 64/97 [00:10<00:05,  6.31it/s, loss=2.89, v_num=0]
Epoch 38:  67%|██████▋   | 65/97 [00:10<00:05,  6.36it/s, loss=2.89, v_num=0]
Epoch 38:  67%|██████▋   | 65/97 [00:10<00:05,  6.32it/s, loss=2.88, v_num=0]
Epoch 38:  68%|██████▊   | 66/97 [00:10<00:04,  6.37it/s, loss=2.88, v_num=0]
Epoch 38:  68%|██████▊   | 66/97 [00:10<00:04,  6.34it/s, loss=2.88, v_num=0]
Epoch 38:  69%|██████▉   | 67/97 [00:10<00:04,  6.39it/s, loss=2.88, v_num=0]
Epoch 38:  69%|██████▉   | 67/97 [00:10<00:04,  6.35it/s, loss=2.89, v_num=0]
Epoch 38:  70%|███████   | 68/97 [00:10<00:04,  6.40it/s, loss=2.89, v_num=0]
Epoch 38:  70%|███████   | 68/97 [00:10<00:04,  6.36it/s, loss=2.89, v_num=0]
Epoch 38:  71%|███████   | 69/97 [00:10<00:04,  6.41it/s, loss=2.89, v_num=0]
Epoch 38:  71%|███████   | 69/97 [00:10<00:04,  6.37it/s, loss=2.9, v_num=0]
Epoch 38:  72%|███████▏  | 70/97 [00:10<00:04,  6.42it/s, loss=2.9, v_num=0]
Epoch 38:  72%|███████▏  | 70/97 [00:10<00:04,  6.39it/s, loss=2.89, v_num=0]
Epoch 38:  73%|███████▎  | 71/97 [00:11<00:04,  6.43it/s, loss=2.89, v_num=0]
Epoch 38:  73%|███████▎  | 71/97 [00:11<00:04,  6.40it/s, loss=2.89, v_num=0]
Epoch 38:  74%|███████▍  | 72/97 [00:11<00:03,  6.44it/s, loss=2.89, v_num=0]
Epoch 38:  74%|███████▍  | 72/97 [00:11<00:03,  6.41it/s, loss=2.89, v_num=0]
Epoch 38:  75%|███████▌  | 73/97 [00:11<00:03,  6.45it/s, loss=2.89, v_num=0]
Epoch 38:  75%|███████▌  | 73/97 [00:11<00:03,  6.42it/s, loss=2.89, v_num=0]
Epoch 38:  76%|███████▋  | 74/97 [00:11<00:03,  6.46it/s, loss=2.89, v_num=0]
Epoch 38:  76%|███████▋  | 74/97 [00:11<00:03,  6.43it/s, loss=2.89, v_num=0]
Epoch 38:  77%|███████▋  | 75/97 [00:11<00:03,  6.47it/s, loss=2.89, v_num=0]
Epoch 38:  77%|███████▋  | 75/97 [00:11<00:03,  6.44it/s, loss=2.89, v_num=0]
Epoch 38:  78%|███████▊  | 76/97 [00:11<00:03,  6.48it/s, loss=2.89, v_num=0]
Epoch 38:  78%|███████▊  | 76/97 [00:11<00:03,  6.45it/s, loss=2.89, v_num=0]
Epoch 38:  79%|███████▉  | 77/97 [00:11<00:03,  6.49it/s, loss=2.89, v_num=0]
Epoch 38:  79%|███████▉  | 77/97 [00:11<00:03,  6.46it/s, loss=2.89, v_num=0]
Epoch 38:  80%|████████  | 78/97 [00:12<00:02,  6.50it/s, loss=2.89, v_num=0]
Epoch 38:  80%|████████  | 78/97 [00:12<00:02,  6.47it/s, loss=2.89, v_num=0]
Epoch 38:  81%|████████▏ | 79/97 [00:12<00:02,  6.51it/s, loss=2.89, v_num=0]
Epoch 38:  81%|████████▏ | 79/97 [00:12<00:02,  6.48it/s, loss=2.89, v_num=0]
Epoch 38:  82%|████████▏ | 80/97 [00:12<00:02,  6.52it/s, loss=2.89, v_num=0]
Epoch 38:  82%|████████▏ | 80/97 [00:12<00:02,  6.49it/s, loss=2.89, v_num=0]
Epoch 38:  84%|████████▎ | 81/97 [00:12<00:02,  6.52it/s, loss=2.89, v_num=0]
Epoch 38:  84%|████████▎ | 81/97 [00:12<00:02,  6.50it/s, loss=2.9, v_num=0]
Epoch 38:  85%|████████▍ | 82/97 [00:12<00:02,  6.53it/s, loss=2.9, v_num=0]
Epoch 38:  85%|████████▍ | 82/97 [00:12<00:02,  6.50it/s, loss=2.9, v_num=0]
Epoch 38:  86%|████████▌ | 83/97 [00:12<00:02,  6.54it/s, loss=2.9, v_num=0]
Epoch 38:  86%|████████▌ | 83/97 [00:12<00:02,  6.51it/s, loss=2.9, v_num=0]
Epoch 38:  87%|████████▋ | 84/97 [00:12<00:01,  6.55it/s, loss=2.9, v_num=0]
Epoch 38:  87%|████████▋ | 84/97 [00:12<00:01,  6.52it/s, loss=2.89, v_num=0]
Epoch 38:  88%|████████▊ | 85/97 [00:12<00:01,  6.56it/s, loss=2.89, v_num=0]
Epoch 38:  88%|████████▊ | 85/97 [00:13<00:01,  6.53it/s, loss=2.89, v_num=0]
Epoch 38:  89%|████████▊ | 86/97 [00:13<00:01,  6.56it/s, loss=2.89, v_num=0]
Epoch 38:  89%|████████▊ | 86/97 [00:13<00:01,  6.53it/s, loss=2.9, v_num=0]
Epoch 38:  90%|████████▉ | 87/97 [00:13<00:01,  6.57it/s, loss=2.9, v_num=0]
Epoch 38:  90%|████████▉ | 87/97 [00:13<00:01,  6.54it/s, loss=2.9, v_num=0]
Epoch 38:  91%|█████████ | 88/97 [00:13<00:01,  6.58it/s, loss=2.9, v_num=0]
Epoch 38:  91%|█████████ | 88/97 [00:13<00:01,  6.55it/s, loss=2.89, v_num=0]
Epoch 38:  92%|█████████▏| 89/97 [00:13<00:01,  6.58it/s, loss=2.89, v_num=0]
Epoch 38:  92%|█████████▏| 89/97 [00:13<00:01,  6.56it/s, loss=2.89, v_num=0]
Epoch 38:  93%|█████████▎| 90/97 [00:13<00:01,  6.59it/s, loss=2.89, v_num=0]
Epoch 38:  93%|█████████▎| 90/97 [00:13<00:01,  6.56it/s, loss=2.89, v_num=0]
Epoch 38:  94%|█████████▍| 91/97 [00:13<00:00,  6.60it/s, loss=2.89, v_num=0]
Epoch 38:  94%|█████████▍| 91/97 [00:13<00:00,  6.57it/s, loss=2.9, v_num=0]
Epoch 38:  95%|█████████▍| 92/97 [00:13<00:00,  6.60it/s, loss=2.9, v_num=0]
Epoch 38:  95%|█████████▍| 92/97 [00:13<00:00,  6.58it/s, loss=2.89, v_num=0]
Epoch 38:  96%|█████████▌| 93/97 [00:14<00:00,  6.61it/s, loss=2.89, v_num=0]
Epoch 38:  96%|█████████▌| 93/97 [00:14<00:00,  6.59it/s, loss=2.89, v_num=0]
Epoch 38:  97%|█████████▋| 94/97 [00:14<00:00,  6.62it/s, loss=2.89, v_num=0]
Epoch 38:  97%|█████████▋| 94/97 [00:14<00:00,  6.59it/s, loss=2.89, v_num=0]
Epoch 38:  98%|█████████▊| 95/97 [00:14<00:00,  6.62it/s, loss=2.89, v_num=0]
Epoch 38:  98%|█████████▊| 95/97 [00:14<00:00,  6.60it/s, loss=2.89, v_num=0]
Epoch 38:  99%|█████████▉| 96/97 [00:14<00:00,  6.63it/s, loss=2.89, v_num=0]
Epoch 38:  99%|█████████▉| 96/97 [00:14<00:00,  6.61it/s, loss=2.89, v_num=0]
Epoch 38: 100%|██████████| 97/97 [00:14<00:00,  6.64it/s, loss=2.89, v_num=0]
Epoch 38: 100%|██████████| 97/97 [00:14<00:00,  6.61it/s, loss=2.89, v_num=0]
Epoch 38: 100%|██████████| 97/97 [00:14<00:00,  6.61it/s, loss=2.89, v_num=0]
Epoch 38:   0%|          | 0/97 [00:00<?, ?it/s, loss=2.89, v_num=0]
Epoch 39:   0%|          | 0/97 [00:00<?, ?it/s, loss=2.89, v_num=0]
Epoch 39:   1%|          | 1/97 [00:01<02:13,  1.39s/it, loss=2.89, v_num=0]
Epoch 39:   1%|          | 1/97 [00:01<02:18,  1.44s/it, loss=2.88, v_num=0]
Epoch 39:   2%|▏         | 2/97 [00:01<01:12,  1.31it/s, loss=2.88, v_num=0]
Epoch 39:   2%|▏         | 2/97 [00:01<01:15,  1.26it/s, loss=2.88, v_num=0]
Epoch 39:   3%|▎         | 3/97 [00:01<00:52,  1.80it/s, loss=2.88, v_num=0]
Epoch 39:   3%|▎         | 3/97 [00:01<00:54,  1.74it/s, loss=2.87, v_num=0]
Epoch 39:   4%|▍         | 4/97 [00:01<00:42,  2.21it/s, loss=2.87, v_num=0]
Epoch 39:   4%|▍         | 4/97 [00:01<00:43,  2.15it/s, loss=2.87, v_num=0]
Epoch 39:   5%|▌         | 5/97 [00:01<00:35,  2.57it/s, loss=2.87, v_num=0]
Epoch 39:   5%|▌         | 5/97 [00:01<00:36,  2.50it/s, loss=2.87, v_num=0]
Epoch 39:   6%|▌         | 6/97 [00:02<00:31,  2.88it/s, loss=2.87, v_num=0]
Epoch 39:   6%|▌         | 6/97 [00:02<00:32,  2.81it/s, loss=2.86, v_num=0]
Epoch 39:   7%|▋         | 7/97 [00:02<00:28,  3.16it/s, loss=2.86, v_num=0]
Epoch 39:   7%|▋         | 7/97 [00:02<00:29,  3.08it/s, loss=2.86, v_num=0]
Epoch 39:   8%|▊         | 8/97 [00:02<00:26,  3.41it/s, loss=2.86, v_num=0]
Epoch 39:   8%|▊         | 8/97 [00:02<00:26,  3.32it/s, loss=2.87, v_num=0]
Epoch 39:   9%|▉         | 9/97 [00:02<00:24,  3.61it/s, loss=2.87, v_num=0]
Epoch 39:   9%|▉         | 9/97 [00:02<00:24,  3.53it/s, loss=2.87, v_num=0]
Epoch 39:  10%|█         | 10/97 [00:02<00:22,  3.80it/s, loss=2.87, v_num=0]
Epoch 39:  10%|█         | 10/97 [00:02<00:23,  3.73it/s, loss=2.87, v_num=0]
Epoch 39:  11%|█▏        | 11/97 [00:02<00:21,  3.98it/s, loss=2.87, v_num=0]
Epoch 39:  11%|█▏        | 11/97 [00:02<00:22,  3.90it/s, loss=2.88, v_num=0]
Epoch 39:  12%|█▏        | 12/97 [00:02<00:20,  4.14it/s, loss=2.88, v_num=0]
Epoch 39:  12%|█▏        | 12/97 [00:02<00:20,  4.06it/s, loss=2.88, v_num=0]
Epoch 39:  13%|█▎        | 13/97 [00:03<00:19,  4.29it/s, loss=2.88, v_num=0]
Epoch 39:  13%|█▎        | 13/97 [00:03<00:19,  4.20it/s, loss=2.88, v_num=0]
Epoch 39:  14%|█▍        | 14/97 [00:03<00:18,  4.42it/s, loss=2.88, v_num=0]
Epoch 39:  14%|█▍        | 14/97 [00:03<00:19,  4.34it/s, loss=2.88, v_num=0]
Epoch 39:  15%|█▌        | 15/97 [00:03<00:18,  4.54it/s, loss=2.88, v_num=0]
Epoch 39:  15%|█▌        | 15/97 [00:03<00:18,  4.46it/s, loss=2.88, v_num=0]
Epoch 39:  16%|█▋        | 16/97 [00:03<00:17,  4.64it/s, loss=2.88, v_num=0]
Epoch 39:  16%|█▋        | 16/97 [00:03<00:17,  4.57it/s, loss=2.87, v_num=0]
Epoch 39:  18%|█▊        | 17/97 [00:03<00:16,  4.74it/s, loss=2.87, v_num=0]
Epoch 39:  18%|█▊        | 17/97 [00:03<00:17,  4.67it/s, loss=2.87, v_num=0]
Epoch 39:  19%|█▊        | 18/97 [00:03<00:16,  4.83it/s, loss=2.87, v_num=0]
Epoch 39:  19%|█▊        | 18/97 [00:03<00:16,  4.77it/s, loss=2.87, v_num=0]
Epoch 39:  20%|█▉        | 19/97 [00:03<00:15,  4.92it/s, loss=2.87, v_num=0]
Epoch 39:  20%|█▉        | 19/97 [00:03<00:16,  4.85it/s, loss=2.86, v_num=0]
Epoch 39:  21%|██        | 20/97 [00:03<00:15,  5.01it/s, loss=2.86, v_num=0]
Epoch 39:  21%|██        | 20/97 [00:04<00:15,  4.94it/s, loss=2.86, v_num=0]
Epoch 39:  22%|██▏       | 21/97 [00:04<00:14,  5.08it/s, loss=2.86, v_num=0]
Epoch 39:  22%|██▏       | 21/97 [00:04<00:15,  5.02it/s, loss=2.86, v_num=0]
Epoch 39:  23%|██▎       | 22/97 [00:04<00:14,  5.15it/s, loss=2.86, v_num=0]
Epoch 39:  23%|██▎       | 22/97 [00:04<00:14,  5.09it/s, loss=2.86, v_num=0]
Epoch 39:  24%|██▎       | 23/97 [00:04<00:14,  5.23it/s, loss=2.86, v_num=0]
Epoch 39:  24%|██▎       | 23/97 [00:04<00:14,  5.16it/s, loss=2.87, v_num=0]
Epoch 39:  25%|██▍       | 24/97 [00:04<00:13,  5.28it/s, loss=2.87, v_num=0]
Epoch 39:  25%|██▍       | 24/97 [00:04<00:13,  5.22it/s, loss=2.87, v_num=0]
Epoch 39:  26%|██▌       | 25/97 [00:04<00:13,  5.34it/s, loss=2.87, v_num=0]
Epoch 39:  26%|██▌       | 25/97 [00:04<00:13,  5.28it/s, loss=2.87, v_num=0]
Epoch 39:  27%|██▋       | 26/97 [00:04<00:13,  5.40it/s, loss=2.87, v_num=0]
Epoch 39:  27%|██▋       | 26/97 [00:04<00:13,  5.34it/s, loss=2.88, v_num=0]
Epoch 39:  28%|██▊       | 27/97 [00:04<00:12,  5.45it/s, loss=2.88, v_num=0]
Epoch 39:  28%|██▊       | 27/97 [00:05<00:12,  5.39it/s, loss=2.87, v_num=0]
Epoch 39:  29%|██▉       | 28/97 [00:05<00:12,  5.50it/s, loss=2.87, v_num=0]
Epoch 39:  29%|██▉       | 28/97 [00:05<00:12,  5.44it/s, loss=2.86, v_num=0]
Epoch 39:  30%|██▉       | 29/97 [00:05<00:12,  5.55it/s, loss=2.86, v_num=0]
Epoch 39:  30%|██▉       | 29/97 [00:05<00:12,  5.49it/s, loss=2.86, v_num=0]
Epoch 39:  31%|███       | 30/97 [00:05<00:11,  5.60it/s, loss=2.86, v_num=0]
Epoch 39:  31%|███       | 30/97 [00:05<00:12,  5.54it/s, loss=2.87, v_num=0]
Epoch 39:  32%|███▏      | 31/97 [00:05<00:11,  5.64it/s, loss=2.87, v_num=0]
Epoch 39:  32%|███▏      | 31/97 [00:05<00:11,  5.58it/s, loss=2.87, v_num=0]
Epoch 39:  33%|███▎      | 32/97 [00:05<00:11,  5.68it/s, loss=2.87, v_num=0]
Epoch 39:  33%|███▎      | 32/97 [00:05<00:11,  5.62it/s, loss=2.87, v_num=0]
Epoch 39:  34%|███▍      | 33/97 [00:05<00:11,  5.72it/s, loss=2.87, v_num=0]
Epoch 39:  34%|███▍      | 33/97 [00:05<00:11,  5.66it/s, loss=2.88, v_num=0]
Epoch 39:  35%|███▌      | 34/97 [00:05<00:10,  5.75it/s, loss=2.88, v_num=0]
Epoch 39:  35%|███▌      | 34/97 [00:05<00:11,  5.70it/s, loss=2.88, v_num=0]
Epoch 39:  36%|███▌      | 35/97 [00:06<00:10,  5.79it/s, loss=2.88, v_num=0]
Epoch 39:  36%|███▌      | 35/97 [00:06<00:10,  5.74it/s, loss=2.88, v_num=0]
Epoch 39:  37%|███▋      | 36/97 [00:06<00:10,  5.82it/s, loss=2.88, v_num=0]
Epoch 39:  37%|███▋      | 36/97 [00:06<00:10,  5.77it/s, loss=2.88, v_num=0]
Epoch 39:  38%|███▊      | 37/97 [00:06<00:10,  5.85it/s, loss=2.88, v_num=0]
Epoch 39:  38%|███▊      | 37/97 [00:06<00:10,  5.80it/s, loss=2.87, v_num=0]
Epoch 39:  39%|███▉      | 38/97 [00:06<00:10,  5.88it/s, loss=2.87, v_num=0]
Epoch 39:  39%|███▉      | 38/97 [00:06<00:10,  5.83it/s, loss=2.87, v_num=0]
Epoch 39:  40%|████      | 39/97 [00:06<00:09,  5.91it/s, loss=2.87, v_num=0]
Epoch 39:  40%|████      | 39/97 [00:06<00:09,  5.86it/s, loss=2.88, v_num=0]
Epoch 39:  41%|████      | 40/97 [00:06<00:09,  5.94it/s, loss=2.88, v_num=0]
Epoch 39:  41%|████      | 40/97 [00:06<00:09,  5.89it/s, loss=2.88, v_num=0]
Epoch 39:  42%|████▏     | 41/97 [00:06<00:09,  5.96it/s, loss=2.88, v_num=0]
Epoch 39:  42%|████▏     | 41/97 [00:06<00:09,  5.92it/s, loss=2.88, v_num=0]
Epoch 39:  43%|████▎     | 42/97 [00:07<00:09,  5.99it/s, loss=2.88, v_num=0]
Epoch 39:  43%|████▎     | 42/97 [00:07<00:09,  5.94it/s, loss=2.88, v_num=0]
Epoch 39:  44%|████▍     | 43/97 [00:07<00:08,  6.01it/s, loss=2.88, v_num=0]
Epoch 39:  44%|████▍     | 43/97 [00:07<00:09,  5.97it/s, loss=2.88, v_num=0]
Epoch 39:  45%|████▌     | 44/97 [00:07<00:08,  6.04it/s, loss=2.88, v_num=0]
Epoch 39:  45%|████▌     | 44/97 [00:07<00:08,  5.99it/s, loss=2.88, v_num=0]
Epoch 39:  46%|████▋     | 45/97 [00:07<00:08,  6.06it/s, loss=2.88, v_num=0]
Epoch 39:  46%|████▋     | 45/97 [00:07<00:08,  6.02it/s, loss=2.87, v_num=0]
Epoch 39:  47%|████▋     | 46/97 [00:07<00:08,  6.08it/s, loss=2.87, v_num=0]
Epoch 39:  47%|████▋     | 46/97 [00:07<00:08,  6.04it/s, loss=2.87, v_num=0]
Epoch 39:  48%|████▊     | 47/97 [00:07<00:08,  6.10it/s, loss=2.87, v_num=0]
Epoch 39:  48%|████▊     | 47/97 [00:07<00:08,  6.06it/s, loss=2.88, v_num=0]
Epoch 39:  49%|████▉     | 48/97 [00:07<00:08,  6.12it/s, loss=2.88, v_num=0]
Epoch 39:  49%|████▉     | 48/97 [00:07<00:08,  6.08it/s, loss=2.88, v_num=0]
Epoch 39:  51%|█████     | 49/97 [00:07<00:07,  6.14it/s, loss=2.88, v_num=0]
Epoch 39:  51%|█████     | 49/97 [00:08<00:07,  6.10it/s, loss=2.87, v_num=0]
Epoch 39:  52%|█████▏    | 50/97 [00:08<00:07,  6.16it/s, loss=2.87, v_num=0]
Epoch 39:  52%|█████▏    | 50/97 [00:08<00:07,  6.12it/s, loss=2.87, v_num=0]
Epoch 39:  53%|█████▎    | 51/97 [00:08<00:07,  6.18it/s, loss=2.87, v_num=0]
Epoch 39:  53%|█████▎    | 51/97 [00:08<00:07,  6.14it/s, loss=2.86, v_num=0]
Epoch 39:  54%|█████▎    | 52/97 [00:08<00:07,  6.19it/s, loss=2.86, v_num=0]
Epoch 39:  54%|█████▎    | 52/97 [00:08<00:07,  6.15it/s, loss=2.86, v_num=0]
Epoch 39:  55%|█████▍    | 53/97 [00:08<00:07,  6.21it/s, loss=2.86, v_num=0]
Epoch 39:  55%|█████▍    | 53/97 [00:08<00:07,  6.17it/s, loss=2.86, v_num=0]
Epoch 39:  56%|█████▌    | 54/97 [00:08<00:06,  6.22it/s, loss=2.86, v_num=0]
Epoch 39:  56%|█████▌    | 54/97 [00:08<00:06,  6.19it/s, loss=2.86, v_num=0]
Epoch 39:  57%|█████▋    | 55/97 [00:08<00:06,  6.24it/s, loss=2.86, v_num=0]
Epoch 39:  57%|█████▋    | 55/97 [00:08<00:06,  6.20it/s, loss=2.86, v_num=0]
Epoch 39:  58%|█████▊    | 56/97 [00:08<00:06,  6.26it/s, loss=2.86, v_num=0]
Epoch 39:  58%|█████▊    | 56/97 [00:09<00:06,  6.22it/s, loss=2.86, v_num=0]
Epoch 39:  59%|█████▉    | 57/97 [00:09<00:06,  6.27it/s, loss=2.86, v_num=0]
Epoch 39:  59%|█████▉    | 57/97 [00:09<00:06,  6.23it/s, loss=2.86, v_num=0]
Epoch 39:  60%|█████▉    | 58/97 [00:09<00:06,  6.29it/s, loss=2.86, v_num=0]
Epoch 39:  60%|█████▉    | 58/97 [00:09<00:06,  6.25it/s, loss=2.86, v_num=0]
Epoch 39:  61%|██████    | 59/97 [00:09<00:06,  6.30it/s, loss=2.86, v_num=0]
Epoch 39:  61%|██████    | 59/97 [00:09<00:06,  6.26it/s, loss=2.86, v_num=0]
Epoch 39:  62%|██████▏   | 60/97 [00:09<00:05,  6.31it/s, loss=2.86, v_num=0]
Epoch 39:  62%|██████▏   | 60/97 [00:09<00:05,  6.28it/s, loss=2.86, v_num=0]
Epoch 39:  63%|██████▎   | 61/97 [00:09<00:05,  6.32it/s, loss=2.86, v_num=0]
Epoch 39:  63%|██████▎   | 61/97 [00:09<00:05,  6.29it/s, loss=2.86, v_num=0]
Epoch 39:  64%|██████▍   | 62/97 [00:09<00:05,  6.34it/s, loss=2.86, v_num=0]
Epoch 39:  64%|██████▍   | 62/97 [00:09<00:05,  6.30it/s, loss=2.86, v_num=0]
Epoch 39:  65%|██████▍   | 63/97 [00:09<00:05,  6.35it/s, loss=2.86, v_num=0]
Epoch 39:  65%|██████▍   | 63/97 [00:09<00:05,  6.31it/s, loss=2.85, v_num=0]
Epoch 39:  66%|██████▌   | 64/97 [00:10<00:05,  6.36it/s, loss=2.85, v_num=0]
Epoch 39:  66%|██████▌   | 64/97 [00:10<00:05,  6.33it/s, loss=2.85, v_num=0]
Epoch 39:  67%|██████▋   | 65/97 [00:10<00:05,  6.37it/s, loss=2.85, v_num=0]
Epoch 39:  67%|██████▋   | 65/97 [00:10<00:05,  6.34it/s, loss=2.86, v_num=0]
Epoch 39:  68%|██████▊   | 66/97 [00:10<00:04,  6.38it/s, loss=2.86, v_num=0]
Epoch 39:  68%|██████▊   | 66/97 [00:10<00:04,  6.35it/s, loss=2.86, v_num=0]
Epoch 39:  69%|██████▉   | 67/97 [00:10<00:04,  6.39it/s, loss=2.86, v_num=0]
Epoch 39:  69%|██████▉   | 67/97 [00:10<00:04,  6.36it/s, loss=2.86, v_num=0]
Epoch 39:  70%|███████   | 68/97 [00:10<00:04,  6.40it/s, loss=2.86, v_num=0]
Epoch 39:  70%|███████   | 68/97 [00:10<00:04,  6.37it/s, loss=2.86, v_num=0]
Epoch 39:  71%|███████   | 69/97 [00:10<00:04,  6.41it/s, loss=2.86, v_num=0]
Epoch 39:  71%|███████   | 69/97 [00:10<00:04,  6.38it/s, loss=2.86, v_num=0]
Epoch 39:  72%|███████▏  | 70/97 [00:10<00:04,  6.42it/s, loss=2.86, v_num=0]
Epoch 39:  72%|███████▏  | 70/97 [00:10<00:04,  6.39it/s, loss=2.85, v_num=0]
Epoch 39:  73%|███████▎  | 71/97 [00:11<00:04,  6.44it/s, loss=2.85, v_num=0]
Epoch 39:  73%|███████▎  | 71/97 [00:11<00:04,  6.40it/s, loss=2.86, v_num=0]
Epoch 39:  74%|███████▍  | 72/97 [00:11<00:03,  6.45it/s, loss=2.86, v_num=0]
Epoch 39:  74%|███████▍  | 72/97 [00:11<00:03,  6.41it/s, loss=2.86, v_num=0]
Epoch 39:  75%|███████▌  | 73/97 [00:11<00:03,  6.46it/s, loss=2.86, v_num=0]
Epoch 39:  75%|███████▌  | 73/97 [00:11<00:03,  6.42it/s, loss=2.85, v_num=0]
Epoch 39:  76%|███████▋  | 74/97 [00:11<00:03,  6.47it/s, loss=2.85, v_num=0]
Epoch 39:  76%|███████▋  | 74/97 [00:11<00:03,  6.43it/s, loss=2.86, v_num=0]
Epoch 39:  77%|███████▋  | 75/97 [00:11<00:03,  6.47it/s, loss=2.86, v_num=0]
Epoch 39:  77%|███████▋  | 75/97 [00:11<00:03,  6.44it/s, loss=2.86, v_num=0]
Epoch 39:  78%|███████▊  | 76/97 [00:11<00:03,  6.48it/s, loss=2.86, v_num=0]
Epoch 39:  78%|███████▊  | 76/97 [00:11<00:03,  6.45it/s, loss=2.86, v_num=0]
Epoch 39:  79%|███████▉  | 77/97 [00:11<00:03,  6.49it/s, loss=2.86, v_num=0]
Epoch 39:  79%|███████▉  | 77/97 [00:11<00:03,  6.46it/s, loss=2.86, v_num=0]
Epoch 39:  80%|████████  | 78/97 [00:11<00:02,  6.50it/s, loss=2.86, v_num=0]
Epoch 39:  80%|████████  | 78/97 [00:12<00:02,  6.47it/s, loss=2.86, v_num=0]
Epoch 39:  81%|████████▏ | 79/97 [00:12<00:02,  6.51it/s, loss=2.86, v_num=0]
Epoch 39:  81%|████████▏ | 79/97 [00:12<00:02,  6.48it/s, loss=2.86, v_num=0]
Epoch 39:  82%|████████▏ | 80/97 [00:12<00:02,  6.52it/s, loss=2.86, v_num=0]
Epoch 39:  82%|████████▏ | 80/97 [00:12<00:02,  6.49it/s, loss=2.87, v_num=0]
Epoch 39:  84%|████████▎ | 81/97 [00:12<00:02,  6.53it/s, loss=2.87, v_num=0]
Epoch 39:  84%|████████▎ | 81/97 [00:12<00:02,  6.50it/s, loss=2.87, v_num=0]
Epoch 39:  85%|████████▍ | 82/97 [00:12<00:02,  6.53it/s, loss=2.87, v_num=0]
Epoch 39:  85%|████████▍ | 82/97 [00:12<00:02,  6.51it/s, loss=2.87, v_num=0]
Epoch 39:  86%|████████▌ | 83/97 [00:12<00:02,  6.55it/s, loss=2.87, v_num=0]
Epoch 39:  86%|████████▌ | 83/97 [00:12<00:02,  6.52it/s, loss=2.87, v_num=0]
Epoch 39:  87%|████████▋ | 84/97 [00:12<00:01,  6.55it/s, loss=2.87, v_num=0]
Epoch 39:  87%|████████▋ | 84/97 [00:12<00:01,  6.52it/s, loss=2.87, v_num=0]
Epoch 39:  88%|████████▊ | 85/97 [00:12<00:01,  6.56it/s, loss=2.87, v_num=0]
Epoch 39:  88%|████████▊ | 85/97 [00:13<00:01,  6.53it/s, loss=2.87, v_num=0]
Epoch 39:  89%|████████▊ | 86/97 [00:13<00:01,  6.57it/s, loss=2.87, v_num=0]
Epoch 39:  89%|████████▊ | 86/97 [00:13<00:01,  6.54it/s, loss=2.86, v_num=0]
Epoch 39:  90%|████████▉ | 87/97 [00:13<00:01,  6.58it/s, loss=2.86, v_num=0]
Epoch 39:  90%|████████▉ | 87/97 [00:13<00:01,  6.55it/s, loss=2.86, v_num=0]
Epoch 39:  91%|█████████ | 88/97 [00:13<00:01,  6.58it/s, loss=2.86, v_num=0]
Epoch 39:  91%|█████████ | 88/97 [00:13<00:01,  6.56it/s, loss=2.86, v_num=0]
Epoch 39:  92%|█████████▏| 89/97 [00:13<00:01,  6.59it/s, loss=2.86, v_num=0]
Epoch 39:  92%|█████████▏| 89/97 [00:13<00:01,  6.56it/s, loss=2.85, v_num=0]
Epoch 39:  93%|█████████▎| 90/97 [00:13<00:01,  6.60it/s, loss=2.85, v_num=0]
Epoch 39:  93%|█████████▎| 90/97 [00:13<00:01,  6.57it/s, loss=2.85, v_num=0]
Epoch 39:  94%|█████████▍| 91/97 [00:13<00:00,  6.60it/s, loss=2.85, v_num=0]
Epoch 39:  94%|█████████▍| 91/97 [00:13<00:00,  6.58it/s, loss=2.85, v_num=0]
Epoch 39:  95%|█████████▍| 92/97 [00:13<00:00,  6.61it/s, loss=2.85, v_num=0]
Epoch 39:  95%|█████████▍| 92/97 [00:13<00:00,  6.58it/s, loss=2.84, v_num=0]
Epoch 39:  96%|█████████▌| 93/97 [00:14<00:00,  6.62it/s, loss=2.84, v_num=0]
Epoch 39:  96%|█████████▌| 93/97 [00:14<00:00,  6.59it/s, loss=2.85, v_num=0]
Epoch 39:  97%|█████████▋| 94/97 [00:14<00:00,  6.63it/s, loss=2.85, v_num=0]
Epoch 39:  97%|█████████▋| 94/97 [00:14<00:00,  6.60it/s, loss=2.85, v_num=0]
Epoch 39:  98%|█████████▊| 95/97 [00:14<00:00,  6.63it/s, loss=2.85, v_num=0]
Epoch 39:  98%|█████████▊| 95/97 [00:14<00:00,  6.61it/s, loss=2.85, v_num=0]
Epoch 39:  99%|█████████▉| 96/97 [00:14<00:00,  6.64it/s, loss=2.85, v_num=0]
Epoch 39:  99%|█████████▉| 96/97 [00:14<00:00,  6.61it/s, loss=2.84, v_num=0]
Epoch 39: 100%|██████████| 97/97 [00:14<00:00,  6.64it/s, loss=2.84, v_num=0]
Epoch 39: 100%|██████████| 97/97 [00:14<00:00,  6.62it/s, loss=2.84, v_num=0]
Epoch 39: 100%|██████████| 97/97 [00:14<00:00,  6.62it/s, loss=2.84, v_num=0]
Epoch 39:   0%|          | 0/97 [00:00<?, ?it/s, loss=2.84, v_num=0]
Epoch 40:   0%|          | 0/97 [00:00<?, ?it/s, loss=2.84, v_num=0]
Epoch 40:   1%|          | 1/97 [00:01<02:14,  1.40s/it, loss=2.84, v_num=0]
Epoch 40:   1%|          | 1/97 [00:01<02:19,  1.45s/it, loss=2.84, v_num=0]
Epoch 40:   2%|▏         | 2/97 [00:01<01:13,  1.30it/s, loss=2.84, v_num=0]
Epoch 40:   2%|▏         | 2/97 [00:01<01:15,  1.25it/s, loss=2.84, v_num=0]
Epoch 40:   3%|▎         | 3/97 [00:01<00:52,  1.79it/s, loss=2.84, v_num=0]
Epoch 40:   3%|▎         | 3/97 [00:01<00:54,  1.73it/s, loss=2.83, v_num=0]
Epoch 40:   4%|▍         | 4/97 [00:01<00:42,  2.20it/s, loss=2.83, v_num=0]
Epoch 40:   4%|▍         | 4/97 [00:01<00:43,  2.14it/s, loss=2.82, v_num=0]
Epoch 40:   5%|▌         | 5/97 [00:01<00:35,  2.56it/s, loss=2.82, v_num=0]
Epoch 40:   5%|▌         | 5/97 [00:02<00:36,  2.49it/s, loss=2.83, v_num=0]
Epoch 40:   6%|▌         | 6/97 [00:02<00:31,  2.88it/s, loss=2.83, v_num=0]
Epoch 40:   6%|▌         | 6/97 [00:02<00:32,  2.80it/s, loss=2.82, v_num=0]
Epoch 40:   7%|▋         | 7/97 [00:02<00:28,  3.14it/s, loss=2.82, v_num=0]
Epoch 40:   7%|▋         | 7/97 [00:02<00:29,  3.07it/s, loss=2.83, v_num=0]
Epoch 40:   8%|▊         | 8/97 [00:02<00:26,  3.39it/s, loss=2.83, v_num=0]
Epoch 40:   8%|▊         | 8/97 [00:02<00:26,  3.31it/s, loss=2.84, v_num=0]
Epoch 40:   9%|▉         | 9/97 [00:02<00:24,  3.59it/s, loss=2.84, v_num=0]
Epoch 40:   9%|▉         | 9/97 [00:02<00:25,  3.52it/s, loss=2.82, v_num=0]
Epoch 40:  10%|█         | 10/97 [00:02<00:22,  3.79it/s, loss=2.82, v_num=0]
Epoch 40:  10%|█         | 10/97 [00:02<00:23,  3.71it/s, loss=2.82, v_num=0]
Epoch 40:  11%|█▏        | 11/97 [00:02<00:21,  3.96it/s, loss=2.82, v_num=0]
Epoch 40:  11%|█▏        | 11/97 [00:02<00:22,  3.88it/s, loss=2.83, v_num=0]
Epoch 40:  12%|█▏        | 12/97 [00:02<00:20,  4.12it/s, loss=2.83, v_num=0]
Epoch 40:  12%|█▏        | 12/97 [00:02<00:21,  4.04it/s, loss=2.83, v_num=0]
Epoch 40:  13%|█▎        | 13/97 [00:03<00:19,  4.26it/s, loss=2.83, v_num=0]
Epoch 40:  13%|█▎        | 13/97 [00:03<00:20,  4.19it/s, loss=2.82, v_num=0]
Epoch 40:  14%|█▍        | 14/97 [00:03<00:18,  4.40it/s, loss=2.82, v_num=0]
Epoch 40:  14%|█▍        | 14/97 [00:03<00:19,  4.32it/s, loss=2.83, v_num=0]
Epoch 40:  15%|█▌        | 15/97 [00:03<00:18,  4.51it/s, loss=2.83, v_num=0]
Epoch 40:  15%|█▌        | 15/97 [00:03<00:18,  4.44it/s, loss=2.83, v_num=0]
Epoch 40:  16%|█▋        | 16/97 [00:03<00:17,  4.63it/s, loss=2.83, v_num=0]
Epoch 40:  16%|█▋        | 16/97 [00:03<00:17,  4.55it/s, loss=2.83, v_num=0]
Epoch 40:  18%|█▊        | 17/97 [00:03<00:16,  4.72it/s, loss=2.83, v_num=0]
Epoch 40:  18%|█▊        | 17/97 [00:03<00:17,  4.65it/s, loss=2.83, v_num=0]
Epoch 40:  19%|█▊        | 18/97 [00:03<00:16,  4.82it/s, loss=2.83, v_num=0]
Epoch 40:  19%|█▊        | 18/97 [00:03<00:16,  4.75it/s, loss=2.83, v_num=0]
Epoch 40:  20%|█▉        | 19/97 [00:03<00:15,  4.91it/s, loss=2.83, v_num=0]
Epoch 40:  20%|█▉        | 19/97 [00:03<00:16,  4.84it/s, loss=2.84, v_num=0]
Epoch 40:  21%|██        | 20/97 [00:04<00:15,  4.99it/s, loss=2.84, v_num=0]
Epoch 40:  21%|██        | 20/97 [00:04<00:15,  4.92it/s, loss=2.84, v_num=0]
Epoch 40:  22%|██▏       | 21/97 [00:04<00:14,  5.07it/s, loss=2.84, v_num=0]
Epoch 40:  22%|██▏       | 21/97 [00:04<00:15,  5.00it/s, loss=2.84, v_num=0]
Epoch 40:  23%|██▎       | 22/97 [00:04<00:14,  5.14it/s, loss=2.84, v_num=0]
Epoch 40:  23%|██▎       | 22/97 [00:04<00:14,  5.07it/s, loss=2.84, v_num=0]
Epoch 40:  24%|██▎       | 23/97 [00:04<00:14,  5.20it/s, loss=2.84, v_num=0]
Epoch 40:  24%|██▎       | 23/97 [00:04<00:14,  5.14it/s, loss=2.85, v_num=0]
Epoch 40:  25%|██▍       | 24/97 [00:04<00:13,  5.27it/s, loss=2.85, v_num=0]
Epoch 40:  25%|██▍       | 24/97 [00:04<00:14,  5.20it/s, loss=2.85, v_num=0]
Epoch 40:  26%|██▌       | 25/97 [00:04<00:13,  5.32it/s, loss=2.85, v_num=0]
Epoch 40:  26%|██▌       | 25/97 [00:04<00:13,  5.26it/s, loss=2.85, v_num=0]
Epoch 40:  27%|██▋       | 26/97 [00:04<00:13,  5.38it/s, loss=2.85, v_num=0]
Epoch 40:  27%|██▋       | 26/97 [00:04<00:13,  5.32it/s, loss=2.86, v_num=0]
Epoch 40:  28%|██▊       | 27/97 [00:04<00:12,  5.43it/s, loss=2.86, v_num=0]
Epoch 40:  28%|██▊       | 27/97 [00:05<00:13,  5.37it/s, loss=2.85, v_num=0]
Epoch 40:  29%|██▉       | 28/97 [00:05<00:12,  5.48it/s, loss=2.85, v_num=0]
Epoch 40:  29%|██▉       | 28/97 [00:05<00:12,  5.42it/s, loss=2.84, v_num=0]
Epoch 40:  30%|██▉       | 29/97 [00:05<00:12,  5.53it/s, loss=2.84, v_num=0]
Epoch 40:  30%|██▉       | 29/97 [00:05<00:12,  5.47it/s, loss=2.85, v_num=0]
Epoch 40:  31%|███       | 30/97 [00:05<00:12,  5.58it/s, loss=2.85, v_num=0]
Epoch 40:  31%|███       | 30/97 [00:05<00:12,  5.52it/s, loss=2.85, v_num=0]
Epoch 40:  32%|███▏      | 31/97 [00:05<00:11,  5.62it/s, loss=2.85, v_num=0]
Epoch 40:  32%|███▏      | 31/97 [00:05<00:11,  5.56it/s, loss=2.84, v_num=0]
Epoch 40:  33%|███▎      | 32/97 [00:05<00:11,  5.66it/s, loss=2.84, v_num=0]
Epoch 40:  33%|███▎      | 32/97 [00:05<00:11,  5.61it/s, loss=2.84, v_num=0]
Epoch 40:  34%|███▍      | 33/97 [00:05<00:11,  5.70it/s, loss=2.84, v_num=0]
Epoch 40:  34%|███▍      | 33/97 [00:05<00:11,  5.65it/s, loss=2.84, v_num=0]
Epoch 40:  35%|███▌      | 34/97 [00:05<00:10,  5.74it/s, loss=2.84, v_num=0]
Epoch 40:  35%|███▌      | 34/97 [00:05<00:11,  5.68it/s, loss=2.84, v_num=0]
Epoch 40:  36%|███▌      | 35/97 [00:06<00:10,  5.78it/s, loss=2.84, v_num=0]
Epoch 40:  36%|███▌      | 35/97 [00:06<00:10,  5.72it/s, loss=2.84, v_num=0]
Epoch 40:  37%|███▋      | 36/97 [00:06<00:10,  5.81it/s, loss=2.84, v_num=0]
Epoch 40:  37%|███▋      | 36/97 [00:06<00:10,  5.76it/s, loss=2.83, v_num=0]
Epoch 40:  38%|███▊      | 37/97 [00:06<00:10,  5.84it/s, loss=2.83, v_num=0]
Epoch 40:  38%|███▊      | 37/97 [00:06<00:10,  5.79it/s, loss=2.82, v_num=0]
Epoch 40:  39%|███▉      | 38/97 [00:06<00:10,  5.87it/s, loss=2.82, v_num=0]
Epoch 40:  39%|███▉      | 38/97 [00:06<00:10,  5.82it/s, loss=2.81, v_num=0]
Epoch 40:  40%|████      | 39/97 [00:06<00:09,  5.90it/s, loss=2.81, v_num=0]
Epoch 40:  40%|████      | 39/97 [00:06<00:09,  5.85it/s, loss=2.81, v_num=0]
Epoch 40:  41%|████      | 40/97 [00:06<00:09,  5.93it/s, loss=2.81, v_num=0]
Epoch 40:  41%|████      | 40/97 [00:06<00:09,  5.88it/s, loss=2.8, v_num=0]
Epoch 40:  42%|████▏     | 41/97 [00:06<00:09,  5.96it/s, loss=2.8, v_num=0]
Epoch 40:  42%|████▏     | 41/97 [00:06<00:09,  5.91it/s, loss=2.79, v_num=0]
Epoch 40:  43%|████▎     | 42/97 [00:07<00:09,  5.99it/s, loss=2.79, v_num=0]
Epoch 40:  43%|████▎     | 42/97 [00:07<00:09,  5.94it/s, loss=2.78, v_num=0]
Epoch 40:  44%|████▍     | 43/97 [00:07<00:08,  6.01it/s, loss=2.78, v_num=0]
Epoch 40:  44%|████▍     | 43/97 [00:07<00:09,  5.96it/s, loss=2.77, v_num=0]
Epoch 40:  45%|████▌     | 44/97 [00:07<00:08,  6.03it/s, loss=2.77, v_num=0]
Epoch 40:  45%|████▌     | 44/97 [00:07<00:08,  5.99it/s, loss=2.77, v_num=0]
Epoch 40:  46%|████▋     | 45/97 [00:07<00:08,  6.06it/s, loss=2.77, v_num=0]
Epoch 40:  46%|████▋     | 45/97 [00:07<00:08,  6.01it/s, loss=2.78, v_num=0]
Epoch 40:  47%|████▋     | 46/97 [00:07<00:08,  6.08it/s, loss=2.78, v_num=0]
Epoch 40:  47%|████▋     | 46/97 [00:07<00:08,  6.03it/s, loss=2.77, v_num=0]
Epoch 40:  48%|████▊     | 47/97 [00:07<00:08,  6.10it/s, loss=2.77, v_num=0]
Epoch 40:  48%|████▊     | 47/97 [00:07<00:08,  6.06it/s, loss=2.78, v_num=0]
Epoch 40:  49%|████▉     | 48/97 [00:07<00:08,  6.12it/s, loss=2.78, v_num=0]
Epoch 40:  49%|████▉     | 48/97 [00:07<00:08,  6.08it/s, loss=2.78, v_num=0]
Epoch 40:  51%|█████     | 49/97 [00:07<00:07,  6.15it/s, loss=2.78, v_num=0]
Epoch 40:  51%|█████     | 49/97 [00:08<00:07,  6.10it/s, loss=2.78, v_num=0]
Epoch 40:  52%|█████▏    | 50/97 [00:08<00:07,  6.16it/s, loss=2.78, v_num=0]
Epoch 40:  52%|█████▏    | 50/97 [00:08<00:07,  6.12it/s, loss=2.79, v_num=0]
Epoch 40:  53%|█████▎    | 51/97 [00:08<00:07,  6.18it/s, loss=2.79, v_num=0]
Epoch 40:  53%|█████▎    | 51/97 [00:08<00:07,  6.14it/s, loss=2.79, v_num=0]
Epoch 40:  54%|█████▎    | 52/97 [00:08<00:07,  6.20it/s, loss=2.79, v_num=0]
Epoch 40:  54%|█████▎    | 52/97 [00:08<00:07,  6.16it/s, loss=2.8, v_num=0]
Epoch 40:  55%|█████▍    | 53/97 [00:08<00:07,  6.22it/s, loss=2.8, v_num=0]
Epoch 40:  55%|█████▍    | 53/97 [00:08<00:07,  6.18it/s, loss=2.8, v_num=0]
Epoch 40:  56%|█████▌    | 54/97 [00:08<00:06,  6.23it/s, loss=2.8, v_num=0]
Epoch 40:  56%|█████▌    | 54/97 [00:08<00:06,  6.19it/s, loss=2.8, v_num=0]
Epoch 40:  57%|█████▋    | 55/97 [00:08<00:06,  6.25it/s, loss=2.8, v_num=0]
Epoch 40:  57%|█████▋    | 55/97 [00:08<00:06,  6.21it/s, loss=2.8, v_num=0]
Epoch 40:  58%|█████▊    | 56/97 [00:08<00:06,  6.26it/s, loss=2.8, v_num=0]
Epoch 40:  58%|█████▊    | 56/97 [00:09<00:06,  6.22it/s, loss=2.8, v_num=0]
Epoch 40:  59%|█████▉    | 57/97 [00:09<00:06,  6.27it/s, loss=2.8, v_num=0]
Epoch 40:  59%|█████▉    | 57/97 [00:09<00:06,  6.24it/s, loss=2.8, v_num=0]
Epoch 40:  60%|█████▉    | 58/97 [00:09<00:06,  6.29it/s, loss=2.8, v_num=0]
Epoch 40:  60%|█████▉    | 58/97 [00:09<00:06,  6.25it/s, loss=2.8, v_num=0]
Epoch 40:  61%|██████    | 59/97 [00:09<00:06,  6.30it/s, loss=2.8, v_num=0]
Epoch 40:  61%|██████    | 59/97 [00:09<00:06,  6.27it/s, loss=2.81, v_num=0]
Epoch 40:  62%|██████▏   | 60/97 [00:09<00:05,  6.31it/s, loss=2.81, v_num=0]
Epoch 40:  62%|██████▏   | 60/97 [00:09<00:05,  6.28it/s, loss=2.81, v_num=0]
Epoch 40:  63%|██████▎   | 61/97 [00:09<00:05,  6.33it/s, loss=2.81, v_num=0]
Epoch 40:  63%|██████▎   | 61/97 [00:09<00:05,  6.29it/s, loss=2.82, v_num=0]
Epoch 40:  64%|██████▍   | 62/97 [00:09<00:05,  6.34it/s, loss=2.82, v_num=0]
Epoch 40:  64%|██████▍   | 62/97 [00:09<00:05,  6.31it/s, loss=2.83, v_num=0]
Epoch 40:  65%|██████▍   | 63/97 [00:09<00:05,  6.36it/s, loss=2.83, v_num=0]
Epoch 40:  65%|██████▍   | 63/97 [00:09<00:05,  6.32it/s, loss=2.83, v_num=0]
Epoch 40:  66%|██████▌   | 64/97 [00:10<00:05,  6.37it/s, loss=2.83, v_num=0]
Epoch 40:  66%|██████▌   | 64/97 [00:10<00:05,  6.33it/s, loss=2.83, v_num=0]
Epoch 40:  67%|██████▋   | 65/97 [00:10<00:05,  6.38it/s, loss=2.83, v_num=0]
Epoch 40:  67%|██████▋   | 65/97 [00:10<00:05,  6.35it/s, loss=2.82, v_num=0]
Epoch 40:  68%|██████▊   | 66/97 [00:10<00:04,  6.39it/s, loss=2.82, v_num=0]
Epoch 40:  68%|██████▊   | 66/97 [00:10<00:04,  6.36it/s, loss=2.82, v_num=0]
Epoch 40:  69%|██████▉   | 67/97 [00:10<00:04,  6.40it/s, loss=2.82, v_num=0]
Epoch 40:  69%|██████▉   | 67/97 [00:10<00:04,  6.37it/s, loss=2.83, v_num=0]
Epoch 40:  70%|███████   | 68/97 [00:10<00:04,  6.42it/s, loss=2.83, v_num=0]
Epoch 40:  70%|███████   | 68/97 [00:10<00:04,  6.38it/s, loss=2.83, v_num=0]
Epoch 40:  71%|███████   | 69/97 [00:10<00:04,  6.42it/s, loss=2.83, v_num=0]
Epoch 40:  71%|███████   | 69/97 [00:10<00:04,  6.39it/s, loss=2.82, v_num=0]
Epoch 40:  72%|███████▏  | 70/97 [00:10<00:04,  6.44it/s, loss=2.82, v_num=0]
Epoch 40:  72%|███████▏  | 70/97 [00:10<00:04,  6.40it/s, loss=2.83, v_num=0]
Epoch 40:  73%|███████▎  | 71/97 [00:11<00:04,  6.44it/s, loss=2.83, v_num=0]
Epoch 40:  73%|███████▎  | 71/97 [00:11<00:04,  6.41it/s, loss=2.83, v_num=0]
Epoch 40:  74%|███████▍  | 72/97 [00:11<00:03,  6.45it/s, loss=2.83, v_num=0]
Epoch 40:  74%|███████▍  | 72/97 [00:11<00:03,  6.42it/s, loss=2.82, v_num=0]
Epoch 40:  75%|███████▌  | 73/97 [00:11<00:03,  6.46it/s, loss=2.82, v_num=0]
Epoch 40:  75%|███████▌  | 73/97 [00:11<00:03,  6.43it/s, loss=2.82, v_num=0]
Epoch 40:  76%|███████▋  | 74/97 [00:11<00:03,  6.47it/s, loss=2.82, v_num=0]
Epoch 40:  76%|███████▋  | 74/97 [00:11<00:03,  6.44it/s, loss=2.82, v_num=0]
Epoch 40:  77%|███████▋  | 75/97 [00:11<00:03,  6.48it/s, loss=2.82, v_num=0]
Epoch 40:  77%|███████▋  | 75/97 [00:11<00:03,  6.45it/s, loss=2.83, v_num=0]
Epoch 40:  78%|███████▊  | 76/97 [00:11<00:03,  6.49it/s, loss=2.83, v_num=0]
Epoch 40:  78%|███████▊  | 76/97 [00:11<00:03,  6.46it/s, loss=2.83, v_num=0]
Epoch 40:  79%|███████▉  | 77/97 [00:11<00:03,  6.50it/s, loss=2.83, v_num=0]
Epoch 40:  79%|███████▉  | 77/97 [00:11<00:03,  6.47it/s, loss=2.83, v_num=0]
Epoch 40:  80%|████████  | 78/97 [00:11<00:02,  6.51it/s, loss=2.83, v_num=0]
Epoch 40:  80%|████████  | 78/97 [00:12<00:02,  6.48it/s, loss=2.83, v_num=0]
Epoch 40:  81%|████████▏ | 79/97 [00:12<00:02,  6.52it/s, loss=2.83, v_num=0]
Epoch 40:  81%|████████▏ | 79/97 [00:12<00:02,  6.49it/s, loss=2.83, v_num=0]
Epoch 40:  82%|████████▏ | 80/97 [00:12<00:02,  6.53it/s, loss=2.83, v_num=0]
Epoch 40:  82%|████████▏ | 80/97 [00:12<00:02,  6.50it/s, loss=2.84, v_num=0]
Epoch 40:  84%|████████▎ | 81/97 [00:12<00:02,  6.53it/s, loss=2.84, v_num=0]
Epoch 40:  84%|████████▎ | 81/97 [00:12<00:02,  6.51it/s, loss=2.83, v_num=0]
Epoch 40:  85%|████████▍ | 82/97 [00:12<00:02,  6.54it/s, loss=2.83, v_num=0]
Epoch 40:  85%|████████▍ | 82/97 [00:12<00:02,  6.51it/s, loss=2.83, v_num=0]
Epoch 40:  86%|████████▌ | 83/97 [00:12<00:02,  6.55it/s, loss=2.83, v_num=0]
Epoch 40:  86%|████████▌ | 83/97 [00:12<00:02,  6.52it/s, loss=2.84, v_num=0]
Epoch 40:  87%|████████▋ | 84/97 [00:12<00:01,  6.56it/s, loss=2.84, v_num=0]
Epoch 40:  87%|████████▋ | 84/97 [00:12<00:01,  6.53it/s, loss=2.84, v_num=0]
Epoch 40:  88%|████████▊ | 85/97 [00:12<00:01,  6.56it/s, loss=2.84, v_num=0]
Epoch 40:  88%|████████▊ | 85/97 [00:13<00:01,  6.54it/s, loss=2.84, v_num=0]
Epoch 40:  89%|████████▊ | 86/97 [00:13<00:01,  6.57it/s, loss=2.84, v_num=0]
Epoch 40:  89%|████████▊ | 86/97 [00:13<00:01,  6.54it/s, loss=2.83, v_num=0]
Epoch 40:  90%|████████▉ | 87/97 [00:13<00:01,  6.58it/s, loss=2.83, v_num=0]
Epoch 40:  90%|████████▉ | 87/97 [00:13<00:01,  6.55it/s, loss=2.83, v_num=0]
Epoch 40:  91%|█████████ | 88/97 [00:13<00:01,  6.59it/s, loss=2.83, v_num=0]
Epoch 40:  91%|█████████ | 88/97 [00:13<00:01,  6.56it/s, loss=2.82, v_num=0]
Epoch 40:  92%|█████████▏| 89/97 [00:13<00:01,  6.60it/s, loss=2.82, v_num=0]
Epoch 40:  92%|█████████▏| 89/97 [00:13<00:01,  6.57it/s, loss=2.82, v_num=0]
Epoch 40:  93%|█████████▎| 90/97 [00:13<00:01,  6.60it/s, loss=2.82, v_num=0]
Epoch 40:  93%|█████████▎| 90/97 [00:13<00:01,  6.57it/s, loss=2.81, v_num=0]
Epoch 40:  94%|█████████▍| 91/97 [00:13<00:00,  6.61it/s, loss=2.81, v_num=0]
Epoch 40:  94%|█████████▍| 91/97 [00:13<00:00,  6.58it/s, loss=2.81, v_num=0]
Epoch 40:  95%|█████████▍| 92/97 [00:13<00:00,  6.61it/s, loss=2.81, v_num=0]
Epoch 40:  95%|█████████▍| 92/97 [00:13<00:00,  6.59it/s, loss=2.82, v_num=0]
Epoch 40:  96%|█████████▌| 93/97 [00:14<00:00,  6.62it/s, loss=2.82, v_num=0]
Epoch 40:  96%|█████████▌| 93/97 [00:14<00:00,  6.60it/s, loss=2.82, v_num=0]
Epoch 40:  97%|█████████▋| 94/97 [00:14<00:00,  6.63it/s, loss=2.82, v_num=0]
Epoch 40:  97%|█████████▋| 94/97 [00:14<00:00,  6.60it/s, loss=2.81, v_num=0]
Epoch 40:  98%|█████████▊| 95/97 [00:14<00:00,  6.64it/s, loss=2.81, v_num=0]
Epoch 40:  98%|█████████▊| 95/97 [00:14<00:00,  6.61it/s, loss=2.8, v_num=0]
Epoch 40:  99%|█████████▉| 96/97 [00:14<00:00,  6.64it/s, loss=2.8, v_num=0]
Epoch 40:  99%|█████████▉| 96/97 [00:14<00:00,  6.62it/s, loss=2.8, v_num=0]
Epoch 40: 100%|██████████| 97/97 [00:14<00:00,  6.65it/s, loss=2.8, v_num=0]
Epoch 40: 100%|██████████| 97/97 [00:14<00:00,  6.62it/s, loss=2.81, v_num=0]
Epoch 40: 100%|██████████| 97/97 [00:14<00:00,  6.62it/s, loss=2.81, v_num=0]
Epoch 40:   0%|          | 0/97 [00:00<?, ?it/s, loss=2.81, v_num=0]
Epoch 41:   0%|          | 0/97 [00:00<?, ?it/s, loss=2.81, v_num=0]
Epoch 41:   1%|          | 1/97 [00:01<02:12,  1.38s/it, loss=2.81, v_num=0]
Epoch 41:   1%|          | 1/97 [00:01<02:17,  1.44s/it, loss=2.8, v_num=0]
Epoch 41:   2%|▏         | 2/97 [00:01<01:12,  1.31it/s, loss=2.8, v_num=0]
Epoch 41:   2%|▏         | 2/97 [00:01<01:14,  1.27it/s, loss=2.81, v_num=0]
Epoch 41:   3%|▎         | 3/97 [00:01<00:51,  1.81it/s, loss=2.81, v_num=0]
Epoch 41:   3%|▎         | 3/97 [00:01<00:53,  1.75it/s, loss=2.8, v_num=0]
Epoch 41:   4%|▍         | 4/97 [00:01<00:41,  2.23it/s, loss=2.8, v_num=0]
Epoch 41:   4%|▍         | 4/97 [00:01<00:43,  2.16it/s, loss=2.81, v_num=0]
Epoch 41:   5%|▌         | 5/97 [00:01<00:35,  2.59it/s, loss=2.81, v_num=0]
Epoch 41:   5%|▌         | 5/97 [00:01<00:36,  2.52it/s, loss=2.81, v_num=0]
Epoch 41:   6%|▌         | 6/97 [00:02<00:31,  2.90it/s, loss=2.81, v_num=0]
Epoch 41:   6%|▌         | 6/97 [00:02<00:32,  2.83it/s, loss=2.81, v_num=0]
Epoch 41:   7%|▋         | 7/97 [00:02<00:28,  3.17it/s, loss=2.81, v_num=0]
Epoch 41:   7%|▋         | 7/97 [00:02<00:29,  3.10it/s, loss=2.8, v_num=0]
Epoch 41:   8%|▊         | 8/97 [00:02<00:26,  3.42it/s, loss=2.8, v_num=0]
Epoch 41:   8%|▊         | 8/97 [00:02<00:26,  3.34it/s, loss=2.81, v_num=0]
Epoch 41:   9%|▉         | 9/97 [00:02<00:24,  3.62it/s, loss=2.81, v_num=0]
Epoch 41:   9%|▉         | 9/97 [00:02<00:24,  3.55it/s, loss=2.81, v_num=0]
Epoch 41:  10%|█         | 10/97 [00:02<00:22,  3.81it/s, loss=2.81, v_num=0]
Epoch 41:  10%|█         | 10/97 [00:02<00:23,  3.73it/s, loss=2.81, v_num=0]
Epoch 41:  11%|█▏        | 11/97 [00:02<00:21,  3.99it/s, loss=2.81, v_num=0]
Epoch 41:  11%|█▏        | 11/97 [00:02<00:21,  3.91it/s, loss=2.81, v_num=0]
Epoch 41:  12%|█▏        | 12/97 [00:02<00:20,  4.15it/s, loss=2.81, v_num=0]
Epoch 41:  12%|█▏        | 12/97 [00:02<00:20,  4.07it/s, loss=2.81, v_num=0]
Epoch 41:  13%|█▎        | 13/97 [00:03<00:19,  4.29it/s, loss=2.81, v_num=0]
Epoch 41:  13%|█▎        | 13/97 [00:03<00:19,  4.21it/s, loss=2.81, v_num=0]
Epoch 41:  14%|█▍        | 14/97 [00:03<00:18,  4.42it/s, loss=2.81, v_num=0]
Epoch 41:  14%|█▍        | 14/97 [00:03<00:19,  4.34it/s, loss=2.81, v_num=0]
Epoch 41:  15%|█▌        | 15/97 [00:03<00:18,  4.55it/s, loss=2.81, v_num=0]
Epoch 41:  15%|█▌        | 15/97 [00:03<00:18,  4.47it/s, loss=2.8, v_num=0]
Epoch 41:  16%|█▋        | 16/97 [00:03<00:17,  4.65it/s, loss=2.8, v_num=0]
Epoch 41:  16%|█▋        | 16/97 [00:03<00:17,  4.58it/s, loss=2.81, v_num=0]
Epoch 41:  18%|█▊        | 17/97 [00:03<00:16,  4.75it/s, loss=2.81, v_num=0]
Epoch 41:  18%|█▊        | 17/97 [00:03<00:17,  4.67it/s, loss=2.81, v_num=0]
Epoch 41:  19%|█▊        | 18/97 [00:03<00:16,  4.84it/s, loss=2.81, v_num=0]
Epoch 41:  19%|█▊        | 18/97 [00:03<00:16,  4.77it/s, loss=2.81, v_num=0]
Epoch 41:  20%|█▉        | 19/97 [00:03<00:15,  4.93it/s, loss=2.81, v_num=0]
Epoch 41:  20%|█▉        | 19/97 [00:03<00:16,  4.86it/s, loss=2.81, v_num=0]
Epoch 41:  21%|██        | 20/97 [00:03<00:15,  5.01it/s, loss=2.81, v_num=0]
Epoch 41:  21%|██        | 20/97 [00:04<00:15,  4.94it/s, loss=2.81, v_num=0]
Epoch 41:  22%|██▏       | 21/97 [00:04<00:14,  5.09it/s, loss=2.81, v_num=0]
Epoch 41:  22%|██▏       | 21/97 [00:04<00:15,  5.02it/s, loss=2.81, v_num=0]
Epoch 41:  23%|██▎       | 22/97 [00:04<00:14,  5.16it/s, loss=2.81, v_num=0]
Epoch 41:  23%|██▎       | 22/97 [00:04<00:14,  5.09it/s, loss=2.8, v_num=0]
Epoch 41:  24%|██▎       | 23/97 [00:04<00:14,  5.23it/s, loss=2.8, v_num=0]
Epoch 41:  24%|██▎       | 23/97 [00:04<00:14,  5.16it/s, loss=2.8, v_num=0]
Epoch 41:  25%|██▍       | 24/97 [00:04<00:13,  5.28it/s, loss=2.8, v_num=0]
Epoch 41:  25%|██▍       | 24/97 [00:04<00:14,  5.21it/s, loss=2.79, v_num=0]
Epoch 41:  26%|██▌       | 25/97 [00:04<00:13,  5.33it/s, loss=2.79, v_num=0]
Epoch 41:  26%|██▌       | 25/97 [00:04<00:13,  5.27it/s, loss=2.79, v_num=0]
Epoch 41:  27%|██▋       | 26/97 [00:04<00:13,  5.39it/s, loss=2.79, v_num=0]
Epoch 41:  27%|██▋       | 26/97 [00:04<00:13,  5.33it/s, loss=2.79, v_num=0]
Epoch 41:  28%|██▊       | 27/97 [00:04<00:12,  5.44it/s, loss=2.79, v_num=0]
Epoch 41:  28%|██▊       | 27/97 [00:05<00:13,  5.38it/s, loss=2.8, v_num=0]
Epoch 41:  29%|██▉       | 28/97 [00:05<00:12,  5.49it/s, loss=2.8, v_num=0]
Epoch 41:  29%|██▉       | 28/97 [00:05<00:12,  5.43it/s, loss=2.79, v_num=0]
Epoch 41:  30%|██▉       | 29/97 [00:05<00:12,  5.54it/s, loss=2.79, v_num=0]
Epoch 41:  30%|██▉       | 29/97 [00:05<00:12,  5.48it/s, loss=2.79, v_num=0]
Epoch 41:  31%|███       | 30/97 [00:05<00:11,  5.58it/s, loss=2.79, v_num=0]
Epoch 41:  31%|███       | 30/97 [00:05<00:12,  5.53it/s, loss=2.79, v_num=0]
Epoch 41:  32%|███▏      | 31/97 [00:05<00:11,  5.63it/s, loss=2.79, v_num=0]
Epoch 41:  32%|███▏      | 31/97 [00:05<00:11,  5.57it/s, loss=2.79, v_num=0]
Epoch 41:  33%|███▎      | 32/97 [00:05<00:11,  5.66it/s, loss=2.79, v_num=0]
Epoch 41:  33%|███▎      | 32/97 [00:05<00:11,  5.61it/s, loss=2.79, v_num=0]
Epoch 41:  34%|███▍      | 33/97 [00:05<00:11,  5.70it/s, loss=2.79, v_num=0]
Epoch 41:  34%|███▍      | 33/97 [00:05<00:11,  5.65it/s, loss=2.79, v_num=0]
Epoch 41:  35%|███▌      | 34/97 [00:05<00:10,  5.74it/s, loss=2.79, v_num=0]
Epoch 41:  35%|███▌      | 34/97 [00:05<00:11,  5.69it/s, loss=2.79, v_num=0]
Epoch 41:  36%|███▌      | 35/97 [00:06<00:10,  5.78it/s, loss=2.79, v_num=0]
Epoch 41:  36%|███▌      | 35/97 [00:06<00:10,  5.72it/s, loss=2.79, v_num=0]
Epoch 41:  37%|███▋      | 36/97 [00:06<00:10,  5.81it/s, loss=2.79, v_num=0]
Epoch 41:  37%|███▋      | 36/97 [00:06<00:10,  5.76it/s, loss=2.79, v_num=0]
Epoch 41:  38%|███▊      | 37/97 [00:06<00:10,  5.84it/s, loss=2.79, v_num=0]
Epoch 41:  38%|███▊      | 37/97 [00:06<00:10,  5.79it/s, loss=2.79, v_num=0]
Epoch 41:  39%|███▉      | 38/97 [00:06<00:10,  5.87it/s, loss=2.79, v_num=0]
Epoch 41:  39%|███▉      | 38/97 [00:06<00:10,  5.82it/s, loss=2.79, v_num=0]
Epoch 41:  40%|████      | 39/97 [00:06<00:09,  5.90it/s, loss=2.79, v_num=0]
Epoch 41:  40%|████      | 39/97 [00:06<00:09,  5.85it/s, loss=2.8, v_num=0]
Epoch 41:  41%|████      | 40/97 [00:06<00:09,  5.93it/s, loss=2.8, v_num=0]
Epoch 41:  41%|████      | 40/97 [00:06<00:09,  5.88it/s, loss=2.8, v_num=0]
Epoch 41:  42%|████▏     | 41/97 [00:06<00:09,  5.96it/s, loss=2.8, v_num=0]
Epoch 41:  42%|████▏     | 41/97 [00:06<00:09,  5.91it/s, loss=2.81, v_num=0]
Epoch 41:  43%|████▎     | 42/97 [00:07<00:09,  5.99it/s, loss=2.81, v_num=0]
Epoch 41:  43%|████▎     | 42/97 [00:07<00:09,  5.94it/s, loss=2.81, v_num=0]
Epoch 41:  44%|████▍     | 43/97 [00:07<00:08,  6.01it/s, loss=2.81, v_num=0]
Epoch 41:  44%|████▍     | 43/97 [00:07<00:09,  5.96it/s, loss=2.82, v_num=0]
Epoch 41:  45%|████▌     | 44/97 [00:07<00:08,  6.03it/s, loss=2.82, v_num=0]
Epoch 41:  45%|████▌     | 44/97 [00:07<00:08,  5.99it/s, loss=2.83, v_num=0]
Epoch 41:  46%|████▋     | 45/97 [00:07<00:08,  6.06it/s, loss=2.83, v_num=0]
Epoch 41:  46%|████▋     | 45/97 [00:07<00:08,  6.01it/s, loss=2.82, v_num=0]
Epoch 41:  47%|████▋     | 46/97 [00:07<00:08,  6.08it/s, loss=2.82, v_num=0]
Epoch 41:  47%|████▋     | 46/97 [00:07<00:08,  6.03it/s, loss=2.82, v_num=0]
Epoch 41:  48%|████▊     | 47/97 [00:07<00:08,  6.10it/s, loss=2.82, v_num=0]
Epoch 41:  48%|████▊     | 47/97 [00:07<00:08,  6.05it/s, loss=2.82, v_num=0]
Epoch 41:  49%|████▉     | 48/97 [00:07<00:08,  6.12it/s, loss=2.82, v_num=0]
Epoch 41:  49%|████▉     | 48/97 [00:07<00:08,  6.07it/s, loss=2.82, v_num=0]
Epoch 41:  51%|█████     | 49/97 [00:07<00:07,  6.14it/s, loss=2.82, v_num=0]
Epoch 41:  51%|█████     | 49/97 [00:08<00:07,  6.09it/s, loss=2.82, v_num=0]
Epoch 41:  52%|█████▏    | 50/97 [00:08<00:07,  6.16it/s, loss=2.82, v_num=0]
Epoch 41:  52%|█████▏    | 50/97 [00:08<00:07,  6.12it/s, loss=2.82, v_num=0]
Epoch 41:  53%|█████▎    | 51/97 [00:08<00:07,  6.18it/s, loss=2.82, v_num=0]
Epoch 41:  53%|█████▎    | 51/97 [00:08<00:07,  6.13it/s, loss=2.83, v_num=0]
Epoch 41:  54%|█████▎    | 52/97 [00:08<00:07,  6.20it/s, loss=2.83, v_num=0]
Epoch 41:  54%|█████▎    | 52/97 [00:08<00:07,  6.15it/s, loss=2.83, v_num=0]
Epoch 41:  55%|█████▍    | 53/97 [00:08<00:07,  6.21it/s, loss=2.83, v_num=0]
Epoch 41:  55%|█████▍    | 53/97 [00:08<00:07,  6.17it/s, loss=2.83, v_num=0]
Epoch 41:  56%|█████▌    | 54/97 [00:08<00:06,  6.23it/s, loss=2.83, v_num=0]
Epoch 41:  56%|█████▌    | 54/97 [00:08<00:06,  6.18it/s, loss=2.83, v_num=0]
Epoch 41:  57%|█████▋    | 55/97 [00:08<00:06,  6.24it/s, loss=2.83, v_num=0]
Epoch 41:  57%|█████▋    | 55/97 [00:08<00:06,  6.20it/s, loss=2.83, v_num=0]
Epoch 41:  58%|█████▊    | 56/97 [00:08<00:06,  6.26it/s, loss=2.83, v_num=0]
Epoch 41:  58%|█████▊    | 56/97 [00:09<00:06,  6.22it/s, loss=2.82, v_num=0]
Epoch 41:  59%|█████▉    | 57/97 [00:09<00:06,  6.28it/s, loss=2.82, v_num=0]
Epoch 41:  59%|█████▉    | 57/97 [00:09<00:06,  6.24it/s, loss=2.82, v_num=0]
Epoch 41:  60%|█████▉    | 58/97 [00:09<00:06,  6.29it/s, loss=2.82, v_num=0]
Epoch 41:  60%|█████▉    | 58/97 [00:09<00:06,  6.25it/s, loss=2.82, v_num=0]
Epoch 41:  61%|██████    | 59/97 [00:09<00:06,  6.31it/s, loss=2.82, v_num=0]
Epoch 41:  61%|██████    | 59/97 [00:09<00:06,  6.27it/s, loss=2.81, v_num=0]
Epoch 41:  62%|██████▏   | 60/97 [00:09<00:05,  6.32it/s, loss=2.81, v_num=0]
Epoch 41:  62%|██████▏   | 60/97 [00:09<00:05,  6.28it/s, loss=2.81, v_num=0]
Epoch 41:  63%|██████▎   | 61/97 [00:09<00:05,  6.33it/s, loss=2.81, v_num=0]
Epoch 41:  63%|██████▎   | 61/97 [00:09<00:05,  6.29it/s, loss=2.81, v_num=0]
Epoch 41:  64%|██████▍   | 62/97 [00:09<00:05,  6.34it/s, loss=2.81, v_num=0]
Epoch 41:  64%|██████▍   | 62/97 [00:09<00:05,  6.31it/s, loss=2.8, v_num=0]
Epoch 41:  65%|██████▍   | 63/97 [00:09<00:05,  6.35it/s, loss=2.8, v_num=0]
Epoch 41:  65%|██████▍   | 63/97 [00:09<00:05,  6.32it/s, loss=2.79, v_num=0]
Epoch 41:  66%|██████▌   | 64/97 [00:10<00:05,  6.37it/s, loss=2.79, v_num=0]
Epoch 41:  66%|██████▌   | 64/97 [00:10<00:05,  6.33it/s, loss=2.79, v_num=0]
Epoch 41:  67%|██████▋   | 65/97 [00:10<00:05,  6.38it/s, loss=2.79, v_num=0]
Epoch 41:  67%|██████▋   | 65/97 [00:10<00:05,  6.35it/s, loss=2.8, v_num=0]
Epoch 41:  68%|██████▊   | 66/97 [00:10<00:04,  6.39it/s, loss=2.8, v_num=0]
Epoch 41:  68%|██████▊   | 66/97 [00:10<00:04,  6.36it/s, loss=2.79, v_num=0]
Epoch 41:  69%|██████▉   | 67/97 [00:10<00:04,  6.41it/s, loss=2.79, v_num=0]
Epoch 41:  69%|██████▉   | 67/97 [00:10<00:04,  6.37it/s, loss=2.79, v_num=0]
Epoch 41:  70%|███████   | 68/97 [00:10<00:04,  6.41it/s, loss=2.79, v_num=0]
Epoch 41:  70%|███████   | 68/97 [00:10<00:04,  6.38it/s, loss=2.79, v_num=0]
Epoch 41:  71%|███████   | 69/97 [00:10<00:04,  6.43it/s, loss=2.79, v_num=0]
Epoch 41:  71%|███████   | 69/97 [00:10<00:04,  6.39it/s, loss=2.79, v_num=0]
Epoch 41:  72%|███████▏  | 70/97 [00:10<00:04,  6.44it/s, loss=2.79, v_num=0]
Epoch 41:  72%|███████▏  | 70/97 [00:10<00:04,  6.40it/s, loss=2.79, v_num=0]
Epoch 41:  73%|███████▎  | 71/97 [00:11<00:04,  6.45it/s, loss=2.79, v_num=0]
Epoch 41:  73%|███████▎  | 71/97 [00:11<00:04,  6.41it/s, loss=2.79, v_num=0]
Epoch 41:  74%|███████▍  | 72/97 [00:11<00:03,  6.46it/s, loss=2.79, v_num=0]
Epoch 41:  74%|███████▍  | 72/97 [00:11<00:03,  6.42it/s, loss=2.78, v_num=0]
Epoch 41:  75%|███████▌  | 73/97 [00:11<00:03,  6.47it/s, loss=2.78, v_num=0]
Epoch 41:  75%|███████▌  | 73/97 [00:11<00:03,  6.43it/s, loss=2.77, v_num=0]
Epoch 41:  76%|███████▋  | 74/97 [00:11<00:03,  6.48it/s, loss=2.77, v_num=0]
Epoch 41:  76%|███████▋  | 74/97 [00:11<00:03,  6.44it/s, loss=2.78, v_num=0]
Epoch 41:  77%|███████▋  | 75/97 [00:11<00:03,  6.49it/s, loss=2.78, v_num=0]
Epoch 41:  77%|███████▋  | 75/97 [00:11<00:03,  6.45it/s, loss=2.78, v_num=0]
Epoch 41:  78%|███████▊  | 76/97 [00:11<00:03,  6.50it/s, loss=2.78, v_num=0]
Epoch 41:  78%|███████▊  | 76/97 [00:11<00:03,  6.46it/s, loss=2.78, v_num=0]
Epoch 41:  79%|███████▉  | 77/97 [00:11<00:03,  6.51it/s, loss=2.78, v_num=0]
Epoch 41:  79%|███████▉  | 77/97 [00:11<00:03,  6.47it/s, loss=2.77, v_num=0]
Epoch 41:  80%|████████  | 78/97 [00:11<00:02,  6.51it/s, loss=2.77, v_num=0]
Epoch 41:  80%|████████  | 78/97 [00:12<00:02,  6.48it/s, loss=2.78, v_num=0]
Epoch 41:  81%|████████▏ | 79/97 [00:12<00:02,  6.52it/s, loss=2.78, v_num=0]
Epoch 41:  81%|████████▏ | 79/97 [00:12<00:02,  6.49it/s, loss=2.78, v_num=0]
Epoch 41:  82%|████████▏ | 80/97 [00:12<00:02,  6.53it/s, loss=2.78, v_num=0]
Epoch 41:  82%|████████▏ | 80/97 [00:12<00:02,  6.50it/s, loss=2.77, v_num=0]
Epoch 41:  84%|████████▎ | 81/97 [00:12<00:02,  6.54it/s, loss=2.77, v_num=0]
Epoch 41:  84%|████████▎ | 81/97 [00:12<00:02,  6.51it/s, loss=2.77, v_num=0]
Epoch 41:  85%|████████▍ | 82/97 [00:12<00:02,  6.54it/s, loss=2.77, v_num=0]
Epoch 41:  85%|████████▍ | 82/97 [00:12<00:02,  6.52it/s, loss=2.77, v_num=0]
Epoch 41:  86%|████████▌ | 83/97 [00:12<00:02,  6.55it/s, loss=2.77, v_num=0]
Epoch 41:  86%|████████▌ | 83/97 [00:12<00:02,  6.52it/s, loss=2.77, v_num=0]
Epoch 41:  87%|████████▋ | 84/97 [00:12<00:01,  6.56it/s, loss=2.77, v_num=0]
Epoch 41:  87%|████████▋ | 84/97 [00:12<00:01,  6.53it/s, loss=2.77, v_num=0]
Epoch 41:  88%|████████▊ | 85/97 [00:12<00:01,  6.57it/s, loss=2.77, v_num=0]
Epoch 41:  88%|████████▊ | 85/97 [00:12<00:01,  6.54it/s, loss=2.76, v_num=0]
Epoch 41:  89%|████████▊ | 86/97 [00:13<00:01,  6.58it/s, loss=2.76, v_num=0]
Epoch 41:  89%|████████▊ | 86/97 [00:13<00:01,  6.55it/s, loss=2.78, v_num=0]
Epoch 41:  90%|████████▉ | 87/97 [00:13<00:01,  6.58it/s, loss=2.78, v_num=0]
Epoch 41:  90%|████████▉ | 87/97 [00:13<00:01,  6.56it/s, loss=2.78, v_num=0]
Epoch 41:  91%|█████████ | 88/97 [00:13<00:01,  6.59it/s, loss=2.78, v_num=0]
Epoch 41:  91%|█████████ | 88/97 [00:13<00:01,  6.56it/s, loss=2.78, v_num=0]
Epoch 41:  92%|█████████▏| 89/97 [00:13<00:01,  6.60it/s, loss=2.78, v_num=0]
Epoch 41:  92%|█████████▏| 89/97 [00:13<00:01,  6.57it/s, loss=2.77, v_num=0]
Epoch 41:  93%|█████████▎| 90/97 [00:13<00:01,  6.61it/s, loss=2.77, v_num=0]
Epoch 41:  93%|█████████▎| 90/97 [00:13<00:01,  6.58it/s, loss=2.77, v_num=0]
Epoch 41:  94%|█████████▍| 91/97 [00:13<00:00,  6.61it/s, loss=2.77, v_num=0]
Epoch 41:  94%|█████████▍| 91/97 [00:13<00:00,  6.58it/s, loss=2.78, v_num=0]
Epoch 41:  95%|█████████▍| 92/97 [00:13<00:00,  6.62it/s, loss=2.78, v_num=0]
Epoch 41:  95%|█████████▍| 92/97 [00:13<00:00,  6.59it/s, loss=2.78, v_num=0]
Epoch 41:  96%|█████████▌| 93/97 [00:14<00:00,  6.62it/s, loss=2.78, v_num=0]
Epoch 41:  96%|█████████▌| 93/97 [00:14<00:00,  6.60it/s, loss=2.78, v_num=0]
Epoch 41:  97%|█████████▋| 94/97 [00:14<00:00,  6.63it/s, loss=2.78, v_num=0]
Epoch 41:  97%|█████████▋| 94/97 [00:14<00:00,  6.60it/s, loss=2.78, v_num=0]
Epoch 41:  98%|█████████▊| 95/97 [00:14<00:00,  6.64it/s, loss=2.78, v_num=0]
Epoch 41:  98%|█████████▊| 95/97 [00:14<00:00,  6.61it/s, loss=2.78, v_num=0]
Epoch 41:  99%|█████████▉| 96/97 [00:14<00:00,  6.64it/s, loss=2.78, v_num=0]
Epoch 41:  99%|█████████▉| 96/97 [00:14<00:00,  6.62it/s, loss=2.78, v_num=0]
Epoch 41: 100%|██████████| 97/97 [00:14<00:00,  6.65it/s, loss=2.78, v_num=0]
Epoch 41: 100%|██████████| 97/97 [00:14<00:00,  6.62it/s, loss=2.79, v_num=0]
Epoch 41: 100%|██████████| 97/97 [00:14<00:00,  6.62it/s, loss=2.79, v_num=0]
Epoch 41:   0%|          | 0/97 [00:00<?, ?it/s, loss=2.79, v_num=0]
Epoch 42:   0%|          | 0/97 [00:00<?, ?it/s, loss=2.79, v_num=0]
Epoch 42:   1%|          | 1/97 [00:01<02:15,  1.41s/it, loss=2.79, v_num=0]
Epoch 42:   1%|          | 1/97 [00:01<02:20,  1.46s/it, loss=2.79, v_num=0]
Epoch 42:   2%|▏         | 2/97 [00:01<01:13,  1.30it/s, loss=2.79, v_num=0]
Epoch 42:   2%|▏         | 2/97 [00:01<01:15,  1.25it/s, loss=2.79, v_num=0]
Epoch 42:   3%|▎         | 3/97 [00:01<00:52,  1.79it/s, loss=2.79, v_num=0]
Epoch 42:   3%|▎         | 3/97 [00:01<00:54,  1.73it/s, loss=2.79, v_num=0]
Epoch 42:   4%|▍         | 4/97 [00:01<00:42,  2.21it/s, loss=2.79, v_num=0]
Epoch 42:   4%|▍         | 4/97 [00:01<00:43,  2.14it/s, loss=2.79, v_num=0]
Epoch 42:   5%|▌         | 5/97 [00:01<00:35,  2.57it/s, loss=2.79, v_num=0]
Epoch 42:   5%|▌         | 5/97 [00:02<00:36,  2.49it/s, loss=2.79, v_num=0]
Epoch 42:   6%|▌         | 6/97 [00:02<00:31,  2.88it/s, loss=2.79, v_num=0]
Epoch 42:   6%|▌         | 6/97 [00:02<00:32,  2.80it/s, loss=2.79, v_num=0]
Epoch 42:   7%|▋         | 7/97 [00:02<00:28,  3.15it/s, loss=2.79, v_num=0]
Epoch 42:   7%|▋         | 7/97 [00:02<00:29,  3.07it/s, loss=2.8, v_num=0]
Epoch 42:   8%|▊         | 8/97 [00:02<00:26,  3.40it/s, loss=2.8, v_num=0]
Epoch 42:   8%|▊         | 8/97 [00:02<00:26,  3.31it/s, loss=2.8, v_num=0]
Epoch 42:   9%|▉         | 9/97 [00:02<00:24,  3.61it/s, loss=2.8, v_num=0]
Epoch 42:   9%|▉         | 9/97 [00:02<00:24,  3.52it/s, loss=2.78, v_num=0]
Epoch 42:  10%|█         | 10/97 [00:02<00:22,  3.80it/s, loss=2.78, v_num=0]
Epoch 42:  10%|█         | 10/97 [00:02<00:23,  3.72it/s, loss=2.78, v_num=0]
Epoch 42:  11%|█▏        | 11/97 [00:02<00:21,  3.97it/s, loss=2.78, v_num=0]
Epoch 42:  11%|█▏        | 11/97 [00:02<00:22,  3.88it/s, loss=2.78, v_num=0]
Epoch 42:  12%|█▏        | 12/97 [00:02<00:20,  4.12it/s, loss=2.78, v_num=0]
Epoch 42:  12%|█▏        | 12/97 [00:02<00:21,  4.04it/s, loss=2.79, v_num=0]
Epoch 42:  13%|█▎        | 13/97 [00:03<00:19,  4.27it/s, loss=2.79, v_num=0]
Epoch 42:  13%|█▎        | 13/97 [00:03<00:20,  4.19it/s, loss=2.78, v_num=0]
Epoch 42:  14%|█▍        | 14/97 [00:03<00:18,  4.39it/s, loss=2.78, v_num=0]
Epoch 42:  14%|█▍        | 14/97 [00:03<00:19,  4.32it/s, loss=2.77, v_num=0]
Epoch 42:  15%|█▌        | 15/97 [00:03<00:18,  4.51it/s, loss=2.77, v_num=0]
Epoch 42:  15%|█▌        | 15/97 [00:03<00:18,  4.44it/s, loss=2.78, v_num=0]
Epoch 42:  16%|█▋        | 16/97 [00:03<00:17,  4.63it/s, loss=2.78, v_num=0]
Epoch 42:  16%|█▋        | 16/97 [00:03<00:17,  4.55it/s, loss=2.78, v_num=0]
Epoch 42:  18%|█▊        | 17/97 [00:03<00:16,  4.73it/s, loss=2.78, v_num=0]
Epoch 42:  18%|█▊        | 17/97 [00:03<00:17,  4.66it/s, loss=2.77, v_num=0]
Epoch 42:  19%|█▊        | 18/97 [00:03<00:16,  4.82it/s, loss=2.77, v_num=0]
Epoch 42:  19%|█▊        | 18/97 [00:03<00:16,  4.75it/s, loss=2.77, v_num=0]
Epoch 42:  20%|█▉        | 19/97 [00:03<00:15,  4.91it/s, loss=2.77, v_num=0]
Epoch 42:  20%|█▉        | 19/97 [00:03<00:16,  4.84it/s, loss=2.77, v_num=0]
Epoch 42:  21%|██        | 20/97 [00:04<00:15,  4.99it/s, loss=2.77, v_num=0]
Epoch 42:  21%|██        | 20/97 [00:04<00:15,  4.92it/s, loss=2.77, v_num=0]
Epoch 42:  22%|██▏       | 21/97 [00:04<00:14,  5.07it/s, loss=2.77, v_num=0]
Epoch 42:  22%|██▏       | 21/97 [00:04<00:15,  5.00it/s, loss=2.76, v_num=0]
Epoch 42:  23%|██▎       | 22/97 [00:04<00:14,  5.14it/s, loss=2.76, v_num=0]
Epoch 42:  23%|██▎       | 22/97 [00:04<00:14,  5.08it/s, loss=2.76, v_num=0]
Epoch 42:  24%|██▎       | 23/97 [00:04<00:14,  5.21it/s, loss=2.76, v_num=0]
Epoch 42:  24%|██▎       | 23/97 [00:04<00:14,  5.14it/s, loss=2.76, v_num=0]
Epoch 42:  25%|██▍       | 24/97 [00:04<00:13,  5.27it/s, loss=2.76, v_num=0]
Epoch 42:  25%|██▍       | 24/97 [00:04<00:14,  5.21it/s, loss=2.77, v_num=0]
Epoch 42:  26%|██▌       | 25/97 [00:04<00:13,  5.33it/s, loss=2.77, v_num=0]
Epoch 42:  26%|██▌       | 25/97 [00:04<00:13,  5.27it/s, loss=2.76, v_num=0]
Epoch 42:  27%|██▋       | 26/97 [00:04<00:13,  5.39it/s, loss=2.76, v_num=0]
Epoch 42:  27%|██▋       | 26/97 [00:04<00:13,  5.33it/s, loss=2.77, v_num=0]
Epoch 42:  28%|██▊       | 27/97 [00:04<00:12,  5.44it/s, loss=2.77, v_num=0]
Epoch 42:  28%|██▊       | 27/97 [00:05<00:13,  5.38it/s, loss=2.77, v_num=0]
Epoch 42:  29%|██▉       | 28/97 [00:05<00:12,  5.49it/s, loss=2.77, v_num=0]
Epoch 42:  29%|██▉       | 28/97 [00:05<00:12,  5.43it/s, loss=2.77, v_num=0]
Epoch 42:  30%|██▉       | 29/97 [00:05<00:12,  5.54it/s, loss=2.77, v_num=0]
Epoch 42:  30%|██▉       | 29/97 [00:05<00:12,  5.48it/s, loss=2.76, v_num=0]
Epoch 42:  31%|███       | 30/97 [00:05<00:12,  5.58it/s, loss=2.76, v_num=0]
Epoch 42:  31%|███       | 30/97 [00:05<00:12,  5.53it/s, loss=2.76, v_num=0]
Epoch 42:  32%|███▏      | 31/97 [00:05<00:11,  5.62it/s, loss=2.76, v_num=0]
Epoch 42:  32%|███▏      | 31/97 [00:05<00:11,  5.57it/s, loss=2.76, v_num=0]
Epoch 42:  33%|███▎      | 32/97 [00:05<00:11,  5.67it/s, loss=2.76, v_num=0]
Epoch 42:  33%|███▎      | 32/97 [00:05<00:11,  5.61it/s, loss=2.75, v_num=0]
Epoch 42:  34%|███▍      | 33/97 [00:05<00:11,  5.70it/s, loss=2.75, v_num=0]
Epoch 42:  34%|███▍      | 33/97 [00:05<00:11,  5.65it/s, loss=2.74, v_num=0]
Epoch 42:  35%|███▌      | 34/97 [00:05<00:10,  5.74it/s, loss=2.74, v_num=0]
Epoch 42:  35%|███▌      | 34/97 [00:05<00:11,  5.69it/s, loss=2.75, v_num=0]
Epoch 42:  36%|███▌      | 35/97 [00:06<00:10,  5.78it/s, loss=2.75, v_num=0]
Epoch 42:  36%|███▌      | 35/97 [00:06<00:10,  5.72it/s, loss=2.75, v_num=0]
Epoch 42:  37%|███▋      | 36/97 [00:06<00:10,  5.81it/s, loss=2.75, v_num=0]
Epoch 42:  37%|███▋      | 36/97 [00:06<00:10,  5.76it/s, loss=2.76, v_num=0]
Epoch 42:  38%|███▊      | 37/97 [00:06<00:10,  5.84it/s, loss=2.76, v_num=0]
Epoch 42:  38%|███▊      | 37/97 [00:06<00:10,  5.79it/s, loss=2.77, v_num=0]
Epoch 42:  39%|███▉      | 38/97 [00:06<00:10,  5.87it/s, loss=2.77, v_num=0]
Epoch 42:  39%|███▉      | 38/97 [00:06<00:10,  5.82it/s, loss=2.76, v_num=0]
Epoch 42:  40%|████      | 39/97 [00:06<00:09,  5.90it/s, loss=2.76, v_num=0]
Epoch 42:  40%|████      | 39/97 [00:06<00:09,  5.85it/s, loss=2.77, v_num=0]
Epoch 42:  41%|████      | 40/97 [00:06<00:09,  5.93it/s, loss=2.77, v_num=0]
Epoch 42:  41%|████      | 40/97 [00:06<00:09,  5.88it/s, loss=2.77, v_num=0]
Epoch 42:  42%|████▏     | 41/97 [00:06<00:09,  5.96it/s, loss=2.77, v_num=0]
Epoch 42:  42%|████▏     | 41/97 [00:06<00:09,  5.91it/s, loss=2.77, v_num=0]
Epoch 42:  43%|████▎     | 42/97 [00:07<00:09,  5.98it/s, loss=2.77, v_num=0]
Epoch 42:  43%|████▎     | 42/97 [00:07<00:09,  5.94it/s, loss=2.76, v_num=0]
Epoch 42:  44%|████▍     | 43/97 [00:07<00:08,  6.01it/s, loss=2.76, v_num=0]
Epoch 42:  44%|████▍     | 43/97 [00:07<00:09,  5.96it/s, loss=2.76, v_num=0]
Epoch 42:  45%|████▌     | 44/97 [00:07<00:08,  6.03it/s, loss=2.76, v_num=0]
Epoch 42:  45%|████▌     | 44/97 [00:07<00:08,  5.98it/s, loss=2.75, v_num=0]
Epoch 42:  46%|████▋     | 45/97 [00:07<00:08,  6.05it/s, loss=2.75, v_num=0]
Epoch 42:  46%|████▋     | 45/97 [00:07<00:08,  6.01it/s, loss=2.75, v_num=0]
Epoch 42:  47%|████▋     | 46/97 [00:07<00:08,  6.07it/s, loss=2.75, v_num=0]
Epoch 42:  47%|████▋     | 46/97 [00:07<00:08,  6.03it/s, loss=2.75, v_num=0]
Epoch 42:  48%|████▊     | 47/97 [00:07<00:08,  6.10it/s, loss=2.75, v_num=0]
Epoch 42:  48%|████▊     | 47/97 [00:07<00:08,  6.05it/s, loss=2.75, v_num=0]
Epoch 42:  49%|████▉     | 48/97 [00:07<00:08,  6.12it/s, loss=2.75, v_num=0]
Epoch 42:  49%|█