W0623 05:22:45.563000 2950623 site-packages/torch/distributed/run.py:766] 
W0623 05:22:45.563000 2950623 site-packages/torch/distributed/run.py:766] *****************************************
W0623 05:22:45.563000 2950623 site-packages/torch/distributed/run.py:766] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed. 
W0623 05:22:45.563000 2950623 site-packages/torch/distributed/run.py:766] *****************************************
[INFO] Running in WANDB offline mode
[INFO] Running in WANDB offline mode
[INFO] Running in WANDB offline mode
[INFO] Running in WANDB offline mode
Trainer._get_train_sampler replaced with custom implementation.
warmup_ratio is deprecated and will be removed in v5.2. Use `warmup_steps` instead.
[2026-06-23 05:22:52,028] [INFO] [real_accelerator.py:222:get_accelerator] Setting ds_accelerator to cuda (auto detect)
df: /home/dungpq6/.triton/autotune: No such file or directory
Trainer._get_train_sampler replaced with custom implementation.
warmup_ratio is deprecated and will be removed in v5.2. Use `warmup_steps` instead.
[2026-06-23 05:22:52,189] [INFO] [real_accelerator.py:222:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Trainer._get_train_sampler replaced with custom implementation.
warmup_ratio is deprecated and will be removed in v5.2. Use `warmup_steps` instead.
[2026-06-23 05:22:52,278] [INFO] [real_accelerator.py:222:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Trainer._get_train_sampler replaced with custom implementation.
warmup_ratio is deprecated and will be removed in v5.2. Use `warmup_steps` instead.
[2026-06-23 05:22:52,354] [INFO] [comm.py:658:init_distributed] cdb=None
[2026-06-23 05:22:52,354] [INFO] [comm.py:689:init_distributed] Initializing TorchBackend in DeepSpeed with backend nccl
[2026-06-23 05:22:52,390] [INFO] [real_accelerator.py:222:get_accelerator] Setting ds_accelerator to cuda (auto detect)
[2026-06-23 05:22:52,504] [INFO] [comm.py:658:init_distributed] cdb=None
[2026-06-23 05:22:52,588] [INFO] [comm.py:658:init_distributed] cdb=None
[2026-06-23 05:22:52,733] [INFO] [comm.py:658:init_distributed] cdb=None

Loading weights:   0%|          | 0/723 [00:00<?, ?it/s]
Loading weights: 100%|██████████| 723/723 [00:00<00:00, 12114.86it/s]

Loading weights:   0%|          | 0/723 [00:00<?, ?it/s]
Loading weights:   0%|          | 0/723 [00:00<?, ?it/s]
Loading weights:   0%|          | 0/723 [00:00<?, ?it/s]
Loading weights: 100%|██████████| 723/723 [00:00<00:00, 12156.18it/s]

Loading weights: 100%|██████████| 723/723 [00:00<00:00, 11861.34it/s]

Loading weights: 100%|██████████| 723/723 [00:00<00:00, 11249.71it/s]
[VGGTEncoder] eval KV-cache window: start=8 recent=48 (total=56 frames)
[FeatureFusion] geometry fusion delta scaled by 0.5
[VGGTEncoder] eval KV-cache window: start=8 recent=48 (total=56 frames)
[FeatureFusion] geometry fusion delta scaled by 0.5
[VGGTEncoder] eval KV-cache window: start=8 recent=48 (total=56 frames)
[FeatureFusion] geometry fusion delta scaled by 0.5
[VGGTEncoder] eval KV-cache window: start=8 recent=48 (total=56 frames)
[FeatureFusion] geometry fusion delta scaled by 0.5
Loading weights from local directory
Loading weights from local directory
Loading weights from local directory
Loading weights from local directory
Qwen3_5Config {
  "architectures": [
    "Qwen3_5ForConditionalGeneration"
  ],
  "dtype": "bfloat16",
  "feature_fusion_method": "deepstack_language_add",
  "fusion_num_layers": 1,
  "geometry_encoder_layers": [
    11,
    17,
    23
  ],
  "geometry_encoder_path": "/mnt/data/vmo-ai-task/dungpq6/model-checkpoint/VGGT-1B",
  "geometry_encoder_streaming": true,
  "geometry_encoder_type": "vggt",
  "geometry_fusion_layers": [
    0,
    1,
    2
  ],
  "geometry_fusion_scale": 0.5,
  "geometry_merger_type": "mlp",
  "image_token_id": 248056,
  "include_camera_token": false,
  "model_type": "qwen3_5",
  "pos_encoding_type": "none",
  "reference_frame": "first",
  "text_config": {
    "attention_bias": false,
    "attention_dropout": 0.0,
    "attn_output_gate": true,
    "bos_token_id": null,
    "dtype": "bfloat16",
    "eos_token_id": 248044,
    "full_attention_interval": 4,
    "head_dim": 256,
    "hidden_act": "silu",
    "hidden_size": 2560,
    "initializer_range": 0.02,
    "intermediate_size": 9216,
    "layer_types": [
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention"
    ],
    "linear_conv_kernel_dim": 4,
    "linear_key_head_dim": 128,
    "linear_num_key_heads": 16,
    "linear_num_value_heads": 32,
    "linear_value_head_dim": 128,
    "mamba_ssm_dtype": "float32",
    "max_position_embeddings": 262144,
    "mlp_only_layers": [],
    "model_type": "qwen3_5_text",
    "mtp_num_hidden_layers": 1,
    "mtp_use_dedicated_embeddings": false,
    "num_attention_heads": 16,
    "num_hidden_layers": 32,
    "num_key_value_heads": 4,
    "pad_token_id": null,
    "partial_rotary_factor": 0.25,
    "rms_norm_eps": 1e-06,
    "rope_parameters": {
      "mrope_interleaved": true,
      "mrope_section": [
        11,
        11,
        10
      ],
      "partial_rotary_factor": 0.25,
      "rope_theta": 10000000,
      "rope_type": "default"
    },
    "tie_word_embeddings": true,
    "use_cache": true,
    "vocab_size": 248320
  },
  "tie_word_embeddings": true,
  "transformers_version": "5.3.0",
  "use_cache": false,
  "use_geometry_encoder": true,
  "video_token_id": 248057,
  "vision_config": {
    "deepstack_visual_indexes": [],
    "depth": 24,
    "dtype": "bfloat16",
    "hidden_act": "gelu_pytorch_tanh",
    "hidden_size": 1024,
    "in_channels": 3,
    "initializer_range": 0.02,
    "intermediate_size": 4096,
    "model_type": "qwen3_5",
    "num_heads": 16,
    "num_position_embeddings": 2304,
    "out_hidden_size": 2560,
    "patch_size": 16,
    "spatial_merge_size": 2,
    "temporal_patch_size": 2
  },
  "vision_end_token_id": 248054,
  "vision_language_fusion_layers": null,
  "vision_start_token_id": 248053
}

Loading datasets: [{'annotation_path': '/mnt/data/vmo-ai-task/anhdh35/JanusVLN/train_r2r_rxr.json', 'data_path': '/mnt/data/vmo-ai-task/anhdh35/JanusVLN', 'tag': '3d', 'sampling_rate': 1.0, 'dataset_name': 'train_r2r_rxr'}]
Qwen3_5Config {
  "architectures": [
    "Qwen3_5ForConditionalGeneration"
  ],
  "dtype": "bfloat16",
  "feature_fusion_method": "deepstack_language_add",
  "fusion_num_layers": 1,
  "geometry_encoder_layers": [
    11,
    17,
    23
  ],
  "geometry_encoder_path": "/mnt/data/vmo-ai-task/dungpq6/model-checkpoint/VGGT-1B",
  "geometry_encoder_streaming": true,
  "geometry_encoder_type": "vggt",
  "geometry_fusion_layers": [
    0,
    1,
    2
  ],
  "geometry_fusion_scale": 0.5,
  "geometry_merger_type": "mlp",
  "image_token_id": 248056,
  "include_camera_token": false,
  "model_type": "qwen3_5",
  "pos_encoding_type": "none",
  "reference_frame": "first",
  "text_config": {
    "attention_bias": false,
    "attention_dropout": 0.0,
    "attn_output_gate": true,
    "bos_token_id": null,
    "dtype": "bfloat16",
    "eos_token_id": 248044,
    "full_attention_interval": 4,
    "head_dim": 256,
    "hidden_act": "silu",
    "hidden_size": 2560,
    "initializer_range": 0.02,
    "intermediate_size": 9216,
    "layer_types": [
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention"
    ],
    "linear_conv_kernel_dim": 4,
    "linear_key_head_dim": 128,
    "linear_num_key_heads": 16,
    "linear_num_value_heads": 32,
    "linear_value_head_dim": 128,
    "mamba_ssm_dtype": "float32",
    "max_position_embeddings": 262144,
    "mlp_only_layers": [],
    "model_type": "qwen3_5_text",
    "mtp_num_hidden_layers": 1,
    "mtp_use_dedicated_embeddings": false,
    "num_attention_heads": 16,
    "num_hidden_layers": 32,
    "num_key_value_heads": 4,
    "pad_token_id": null,
    "partial_rotary_factor": 0.25,
    "rms_norm_eps": 1e-06,
    "rope_parameters": {
      "mrope_interleaved": true,
      "mrope_section": [
        11,
        11,
        10
      ],
      "partial_rotary_factor": 0.25,
      "rope_theta": 10000000,
      "rope_type": "default"
    },
    "tie_word_embeddings": true,
    "use_cache": true,
    "vocab_size": 248320
  },
  "tie_word_embeddings": true,
  "transformers_version": "5.3.0",
  "use_cache": false,
  "use_geometry_encoder": true,
  "video_token_id": 248057,
  "vision_config": {
    "deepstack_visual_indexes": [],
    "depth": 24,
    "dtype": "bfloat16",
    "hidden_act": "gelu_pytorch_tanh",
    "hidden_size": 1024,
    "in_channels": 3,
    "initializer_range": 0.02,
    "intermediate_size": 4096,
    "model_type": "qwen3_5",
    "num_heads": 16,
    "num_position_embeddings": 2304,
    "out_hidden_size": 2560,
    "patch_size": 16,
    "spatial_merge_size": 2,
    "temporal_patch_size": 2
  },
  "vision_end_token_id": 248054,
  "vision_language_fusion_layers": null,
  "vision_start_token_id": 248053
}

Loading datasets: [{'annotation_path': '/mnt/data/vmo-ai-task/anhdh35/JanusVLN/train_r2r_rxr.json', 'data_path': '/mnt/data/vmo-ai-task/anhdh35/JanusVLN', 'tag': '3d', 'sampling_rate': 1.0, 'dataset_name': 'train_r2r_rxr'}]
Vision Module - Attention Blocks:
Trainable Block Indices: None
Non-Trainable Block Indices: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23]
Merger Module Trainable: True
LLM Module - Embed Tokens Trainable: True
LLM Module - Trainable Layer Indices: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31]
LLM Module - Non-Trainable Layer Indices: None
Qwen3_5Config {
  "architectures": [
    "Qwen3_5ForConditionalGeneration"
  ],
  "dtype": "bfloat16",
  "feature_fusion_method": "deepstack_language_add",
  "fusion_num_layers": 1,
  "geometry_encoder_layers": [
    11,
    17,
    23
  ],
  "geometry_encoder_path": "/mnt/data/vmo-ai-task/dungpq6/model-checkpoint/VGGT-1B",
  "geometry_encoder_streaming": true,
  "geometry_encoder_type": "vggt",
  "geometry_fusion_layers": [
    0,
    1,
    2
  ],
  "geometry_fusion_scale": 0.5,
  "geometry_merger_type": "mlp",
  "image_token_id": 248056,
  "include_camera_token": false,
  "model_type": "qwen3_5",
  "pos_encoding_type": "none",
  "reference_frame": "first",
  "text_config": {
    "attention_bias": false,
    "attention_dropout": 0.0,
    "attn_output_gate": true,
    "bos_token_id": null,
    "dtype": "bfloat16",
    "eos_token_id": 248044,
    "full_attention_interval": 4,
    "head_dim": 256,
    "hidden_act": "silu",
    "hidden_size": 2560,
    "initializer_range": 0.02,
    "intermediate_size": 9216,
    "layer_types": [
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention"
    ],
    "linear_conv_kernel_dim": 4,
    "linear_key_head_dim": 128,
    "linear_num_key_heads": 16,
    "linear_num_value_heads": 32,
    "linear_value_head_dim": 128,
    "mamba_ssm_dtype": "float32",
    "max_position_embeddings": 262144,
    "mlp_only_layers": [],
    "model_type": "qwen3_5_text",
    "mtp_num_hidden_layers": 1,
    "mtp_use_dedicated_embeddings": false,
    "num_attention_heads": 16,
    "num_hidden_layers": 32,
    "num_key_value_heads": 4,
    "pad_token_id": null,
    "partial_rotary_factor": 0.25,
    "rms_norm_eps": 1e-06,
    "rope_parameters": {
      "mrope_interleaved": true,
      "mrope_section": [
        11,
        11,
        10
      ],
      "partial_rotary_factor": 0.25,
      "rope_theta": 10000000,
      "rope_type": "default"
    },
    "tie_word_embeddings": true,
    "use_cache": true,
    "vocab_size": 248320
  },
  "tie_word_embeddings": true,
  "transformers_version": "5.3.0",
  "use_cache": false,
  "use_geometry_encoder": true,
  "video_token_id": 248057,
  "vision_config": {
    "deepstack_visual_indexes": [],
    "depth": 24,
    "dtype": "bfloat16",
    "hidden_act": "gelu_pytorch_tanh",
    "hidden_size": 1024,
    "in_channels": 3,
    "initializer_range": 0.02,
    "intermediate_size": 4096,
    "model_type": "qwen3_5",
    "num_heads": 16,
    "num_position_embeddings": 2304,
    "out_hidden_size": 2560,
    "patch_size": 16,
    "spatial_merge_size": 2,
    "temporal_patch_size": 2
  },
  "vision_end_token_id": 248054,
  "vision_language_fusion_layers": null,
  "vision_start_token_id": 248053
}

Loading datasets: [{'annotation_path': '/mnt/data/vmo-ai-task/anhdh35/JanusVLN/train_r2r_rxr.json', 'data_path': '/mnt/data/vmo-ai-task/anhdh35/JanusVLN', 'tag': '3d', 'sampling_rate': 1.0, 'dataset_name': 'train_r2r_rxr'}]
Qwen3_5Config {
  "architectures": [
    "Qwen3_5ForConditionalGeneration"
  ],
  "dtype": "bfloat16",
  "feature_fusion_method": "deepstack_language_add",
  "fusion_num_layers": 1,
  "geometry_encoder_layers": [
    11,
    17,
    23
  ],
  "geometry_encoder_path": "/mnt/data/vmo-ai-task/dungpq6/model-checkpoint/VGGT-1B",
  "geometry_encoder_streaming": true,
  "geometry_encoder_type": "vggt",
  "geometry_fusion_layers": [
    0,
    1,
    2
  ],
  "geometry_fusion_scale": 0.5,
  "geometry_merger_type": "mlp",
  "image_token_id": 248056,
  "include_camera_token": false,
  "model_type": "qwen3_5",
  "pos_encoding_type": "none",
  "reference_frame": "first",
  "text_config": {
    "attention_bias": false,
    "attention_dropout": 0.0,
    "attn_output_gate": true,
    "bos_token_id": null,
    "dtype": "bfloat16",
    "eos_token_id": 248044,
    "full_attention_interval": 4,
    "head_dim": 256,
    "hidden_act": "silu",
    "hidden_size": 2560,
    "initializer_range": 0.02,
    "intermediate_size": 9216,
    "layer_types": [
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention",
      "linear_attention",
      "linear_attention",
      "linear_attention",
      "full_attention"
    ],
    "linear_conv_kernel_dim": 4,
    "linear_key_head_dim": 128,
    "linear_num_key_heads": 16,
    "linear_num_value_heads": 32,
    "linear_value_head_dim": 128,
    "mamba_ssm_dtype": "float32",
    "max_position_embeddings": 262144,
    "mlp_only_layers": [],
    "model_type": "qwen3_5_text",
    "mtp_num_hidden_layers": 1,
    "mtp_use_dedicated_embeddings": false,
    "num_attention_heads": 16,
    "num_hidden_layers": 32,
    "num_key_value_heads": 4,
    "pad_token_id": null,
    "partial_rotary_factor": 0.25,
    "rms_norm_eps": 1e-06,
    "rope_parameters": {
      "mrope_interleaved": true,
      "mrope_section": [
        11,
        11,
        10
      ],
      "partial_rotary_factor": 0.25,
      "rope_theta": 10000000,
      "rope_type": "default"
    },
    "tie_word_embeddings": true,
    "use_cache": true,
    "vocab_size": 248320
  },
  "tie_word_embeddings": true,
  "transformers_version": "5.3.0",
  "use_cache": false,
  "use_geometry_encoder": true,
  "video_token_id": 248057,
  "vision_config": {
    "deepstack_visual_indexes": [],
    "depth": 24,
    "dtype": "bfloat16",
    "hidden_act": "gelu_pytorch_tanh",
    "hidden_size": 1024,
    "in_channels": 3,
    "initializer_range": 0.02,
    "intermediate_size": 4096,
    "model_type": "qwen3_5",
    "num_heads": 16,
    "num_position_embeddings": 2304,
    "out_hidden_size": 2560,
    "patch_size": 16,
    "spatial_merge_size": 2,
    "temporal_patch_size": 2
  },
  "vision_end_token_id": 248054,
  "vision_language_fusion_layers": null,
  "vision_start_token_id": 248053
}

Loading datasets: [{'annotation_path': '/mnt/data/vmo-ai-task/anhdh35/JanusVLN/train_r2r_rxr.json', 'data_path': '/mnt/data/vmo-ai-task/anhdh35/JanusVLN', 'tag': '3d', 'sampling_rate': 1.0, 'dataset_name': 'train_r2r_rxr'}]
Total training samples: 631244
Total training samples: 631244
Formatting inputs...Skip in lazy mode
Formatting inputs...Skip in lazy mode
Total training samples: 631244
Formatting inputs...Skip in lazy mode
Total training samples: 631244
The tokenizer has new PAD/BOS/EOS tokens that differ from the model config and generation config. The model config and generation config were aligned accordingly, being updated with the tokenizer's values. Updated tokens: {'eos_token_id': 248046, 'pad_token_id': 248044}.
The tokenizer has new PAD/BOS/EOS tokens that differ from the model config and generation config. The model config and generation config were aligned accordingly, being updated with the tokenizer's values. Updated tokens: {'eos_token_id': 248046, 'pad_token_id': 248044}.
Formatting inputs...Skip in lazy mode
The tokenizer has new PAD/BOS/EOS tokens that differ from the model config and generation config. The model config and generation config were aligned accordingly, being updated with the tokenizer's values. Updated tokens: {'eos_token_id': 248046, 'pad_token_id': 248044}.
The tokenizer has new PAD/BOS/EOS tokens that differ from the model config and generation config. The model config and generation config were aligned accordingly, being updated with the tokenizer's values. Updated tokens: {'eos_token_id': 248046, 'pad_token_id': 248044}.
NCCL version 2.26.2+cuda12.2

  0%|          | 0/9864 [00:00<?, ?it/s]/home/dungpq6/Project/SpatialStack/src/qwen_vl/model/geometry_encoders/vggt_encoder.py:285: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(dtype=dtype):
/home/dungpq6/Project/SpatialStack/src/qwen_vl/model/geometry_encoders/vggt_encoder.py:285: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(dtype=dtype):
/home/dungpq6/Project/SpatialStack/src/qwen_vl/model/geometry_encoders/vggt_encoder.py:285: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(dtype=dtype):
/home/dungpq6/Project/SpatialStack/src/qwen_vl/model/geometry_encoders/vggt_encoder.py:285: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(dtype=dtype):
/home/dungpq6/Project/SpatialStack/src/qwen_vl/model/geometry_encoders/vggt_encoder.py:285: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(dtype=dtype):
/home/dungpq6/Project/SpatialStack/src/qwen_vl/model/geometry_encoders/vggt_encoder.py:285: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(dtype=dtype):
/home/dungpq6/Project/SpatialStack/src/qwen_vl/model/geometry_encoders/vggt_encoder.py:285: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(dtype=dtype):
/home/dungpq6/Project/SpatialStack/src/qwen_vl/model/geometry_encoders/vggt_encoder.py:285: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(dtype=dtype):
/home/dungpq6/Project/SpatialStack/src/qwen_vl/model/geometry_encoders/vggt_encoder.py:285: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(dtype=dtype):
/home/dungpq6/Project/SpatialStack/src/qwen_vl/model/geometry_encoders/vggt_encoder.py:285: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(dtype=dtype):
/home/dungpq6/Project/SpatialStack/src/qwen_vl/model/geometry_encoders/vggt_encoder.py:285: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(dtype=dtype):
/home/dungpq6/Project/SpatialStack/src/qwen_vl/model/geometry_encoders/vggt_encoder.py:285: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(dtype=dtype):
/home/dungpq6/Project/SpatialStack/src/qwen_vl/model/geometry_encoders/vggt_encoder.py:285: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(dtype=dtype):

  0%|          | 1/9864 [02:13<364:38:05, 133.09s/it]
  0%|          | 2/9864 [02:27<173:10:12, 63.21s/it] 
  0%|          | 3/9864 [02:41<112:16:34, 40.99s/it]
  0%|          | 4/9864 [02:55<82:48:07, 30.23s/it] 
  0%|          | 5/9864 [03:09<67:04:06, 24.49s/it]
  0%|          | 6/9864 [03:24<57:28:54, 20.99s/it]
  0%|          | 7/9864 [03:40<53:26:19, 19.52s/it]
  0%|          | 8/9864 [03:54<48:38:01, 17.76s/it]
  0%|          | 9/9864 [04:13<49:38:57, 18.14s/it]
  0%|          | 10/9864 [04:27<45:40:33, 16.69s/it]
                                                    
{'loss': '1.469', 'grad_norm': '50.73', 'learning_rate': '6.081e-07', 'epoch': '0.001014'}

  0%|          | 10/9864 [04:27<45:40:33, 16.69s/it]
  0%|          | 11/9864 [04:42<44:13:39, 16.16s/it]
  0%|          | 12/9864 [04:55<42:23:21, 15.49s/it]
  0%|          | 13/9864 [05:10<41:13:28, 15.07s/it]
  0%|          | 14/9864 [05:25<41:46:17, 15.27s/it]
  0%|          | 15/9864 [05:40<41:08:03, 15.04s/it]
  0%|          | 16/9864 [05:57<43:18:16, 15.83s/it]
  0%|          | 17/9864 [06:13<43:07:08, 15.76s/it]
  0%|          | 18/9864 [06:27<41:54:15, 15.32s/it]
  0%|          | 19/9864 [06:42<40:58:16, 14.98s/it]
  0%|          | 20/9864 [06:55<39:55:57, 14.60s/it]
                                                    
{'loss': '0.2675', 'grad_norm': '19.65', 'learning_rate': '1.284e-06', 'epoch': '0.002028'}

  0%|          | 20/9864 [06:55<39:55:57, 14.60s/it]
  0%|          | 21/9864 [07:09<39:29:19, 14.44s/it]
  0%|          | 22/9864 [07:25<40:37:22, 14.86s/it]
  0%|          | 23/9864 [07:40<40:28:34, 14.81s/it]
  0%|          | 24/9864 [07:55<41:07:22, 15.04s/it]
  0%|          | 25/9864 [08:11<41:54:24, 15.33s/it]
  0%|          | 26/9864 [08:27<41:54:59, 15.34s/it]
  0%|          | 27/9864 [08:41<40:54:23, 14.97s/it]
  0%|          | 28/9864 [08:55<40:15:15, 14.73s/it]
  0%|          | 29/9864 [09:09<39:36:40, 14.50s/it]
  0%|          | 30/9864 [09:25<40:48:05, 14.94s/it]
                                                    
{'loss': '0.2197', 'grad_norm': '14', 'learning_rate': '1.959e-06', 'epoch': '0.003042'}

  0%|          | 30/9864 [09:25<40:48:05, 14.94s/it]
  0%|          | 31/9864 [09:40<41:05:45, 15.05s/it]
  0%|          | 32/9864 [09:56<41:40:09, 15.26s/it]
  0%|          | 33/9864 [10:10<40:46:50, 14.93s/it]
  0%|          | 34/9864 [10:26<41:40:27, 15.26s/it]
  0%|          | 35/9864 [10:40<40:17:38, 14.76s/it]
  0%|          | 36/9864 [10:54<39:44:39, 14.56s/it]
  0%|          | 37/9864 [11:08<39:18:45, 14.40s/it]
  0%|          | 38/9864 [11:24<40:25:27, 14.81s/it]
  0%|          | 39/9864 [11:40<41:19:47, 15.14s/it]
  0%|          | 40/9864 [11:56<41:55:29, 15.36s/it]
                                                    
{'loss': '0.2126', 'grad_norm': '16.08', 'learning_rate': '2.635e-06', 'epoch': '0.004055'}

  0%|          | 40/9864 [11:56<41:55:29, 15.36s/it]
  0%|          | 41/9864 [12:10<40:44:53, 14.93s/it]
  0%|          | 42/9864 [12:25<41:27:46, 15.20s/it]
  0%|          | 43/9864 [12:40<40:48:06, 14.96s/it]
  0%|          | 44/9864 [12:54<39:58:29, 14.65s/it]
  0%|          | 45/9864 [13:10<41:01:05, 15.04s/it]
  0%|          | 46/9864 [13:26<41:53:07, 15.36s/it]
  0%|          | 47/9864 [13:40<40:55:07, 15.01s/it]
  0%|          | 48/9864 [13:56<41:32:54, 15.24s/it]
  0%|          | 49/9864 [14:09<39:59:27, 14.67s/it]
  1%|          | 50/9864 [14:25<41:03:33, 15.06s/it]
                                                    
{'loss': '0.2124', 'grad_norm': '13.42', 'learning_rate': '3.311e-06', 'epoch': '0.005069'}

  1%|          | 50/9864 [14:25<41:03:33, 15.06s/it]
  1%|          | 51/9864 [14:39<40:22:10, 14.81s/it]
  1%|          | 52/9864 [14:53<39:36:14, 14.53s/it]
  1%|          | 53/9864 [15:09<40:48:16, 14.97s/it]
  1%|          | 54/9864 [15:25<41:37:35, 15.28s/it]
  1%|          | 55/9864 [15:41<42:30:41, 15.60s/it]
  1%|          | 56/9864 [15:56<41:34:52, 15.26s/it]
  1%|          | 57/9864 [16:10<40:46:05, 14.97s/it]
  1%|          | 58/9864 [16:26<41:04:14, 15.08s/it]
  1%|          | 59/9864 [16:40<40:29:47, 14.87s/it]
  1%|          | 60/9864 [16:54<39:34:51, 14.53s/it]
                                                    
{'loss': '0.2208', 'grad_norm': '7.979', 'learning_rate': '3.986e-06', 'epoch': '0.006083'}

  1%|          | 60/9864 [16:54<39:34:51, 14.53s/it]
  1%|          | 61/9864 [17:11<42:01:38, 15.43s/it]
  1%|          | 62/9864 [17:25<40:49:34, 14.99s/it]
  1%|          | 63/9864 [17:41<41:36:10, 15.28s/it]
  1%|          | 64/9864 [17:55<40:47:25, 14.98s/it]
  1%|          | 65/9864 [18:11<41:17:17, 15.17s/it]
  1%|          | 66/9864 [18:25<40:28:53, 14.87s/it]
  1%|          | 67/9864 [18:39<39:32:03, 14.53s/it]
  1%|          | 68/9864 [18:55<40:37:00, 14.93s/it]
  1%|          | 69/9864 [19:11<41:18:38, 15.18s/it]
  1%|          | 70/9864 [19:26<41:24:57, 15.22s/it]
                                                    
{'loss': '0.1998', 'grad_norm': '10.19', 'learning_rate': '4.662e-06', 'epoch': '0.007097'}

  1%|          | 70/9864 [19:26<41:24:57, 15.22s/it]
  1%|          | 71/9864 [19:40<40:42:23, 14.96s/it]
  1%|          | 72/9864 [19:55<40:09:12, 14.76s/it]
  1%|          | 73/9864 [20:10<40:44:35, 14.98s/it]
  1%|          | 74/9864 [20:24<39:52:57, 14.67s/it]
  1%|          | 75/9864 [20:39<39:50:10, 14.65s/it]
  1%|          | 76/9864 [20:56<41:54:05, 15.41s/it]
  1%|          | 77/9864 [21:11<41:52:38, 15.40s/it]
  1%|          | 78/9864 [21:25<41:01:03, 15.09s/it]
  1%|          | 79/9864 [21:39<40:03:23, 14.74s/it]
  1%|          | 80/9864 [21:54<39:51:21, 14.66s/it]
                                                    
{'loss': '0.2234', 'grad_norm': '10.56', 'learning_rate': '5.338e-06', 'epoch': '0.008111'}

  1%|          | 80/9864 [21:54<39:51:21, 14.66s/it]
  1%|          | 81/9864 [22:10<41:04:22, 15.11s/it]
  1%|          | 82/9864 [22:24<40:01:26, 14.73s/it]
  1%|          | 83/9864 [22:40<40:47:24, 15.01s/it]
  1%|          | 84/9864 [22:56<41:40:49, 15.34s/it]
  1%|          | 85/9864 [23:11<41:58:10, 15.45s/it]
  1%|          | 86/9864 [23:25<40:50:21, 15.04s/it]
  1%|          | 87/9864 [23:40<40:14:32, 14.82s/it]
  1%|          | 88/9864 [23:55<40:47:55, 15.02s/it]
  1%|          | 89/9864 [24:09<39:56:44, 14.71s/it]
  1%|          | 90/9864 [24:24<39:41:33, 14.62s/it]
                                                    
{'loss': '0.213', 'grad_norm': '8.021', 'learning_rate': '6.014e-06', 'epoch': '0.009125'}

  1%|          | 90/9864 [24:24<39:41:33, 14.62s/it]
  1%|          | 91/9864 [24:41<41:58:55, 15.46s/it]
  1%|          | 92/9864 [24:57<42:06:51, 15.51s/it]
  1%|          | 93/9864 [25:10<40:32:38, 14.94s/it]
  1%|          | 94/9864 [25:24<39:30:37, 14.56s/it]
  1%|          | 95/9864 [25:38<39:18:59, 14.49s/it]
  1%|          | 96/9864 [25:54<40:03:30, 14.76s/it]
  1%|          | 97/9864 [26:08<39:35:37, 14.59s/it]
  1%|          | 98/9864 [26:23<40:12:52, 14.82s/it]
  1%|          | 99/9864 [26:39<41:15:24, 15.21s/it]
  1%|          | 100/9864 [26:55<41:37:51, 15.35s/it]
                                                     
{'loss': '0.1958', 'grad_norm': '6.12', 'learning_rate': '6.689e-06', 'epoch': '0.01014'}

  1%|          | 100/9864 [26:55<41:37:51, 15.35s/it]
  1%|          | 101/9864 [27:09<40:20:01, 14.87s/it]
  1%|          | 102/9864 [27:23<39:32:43, 14.58s/it]
  1%|          | 103/9864 [27:37<39:28:33, 14.56s/it]
  1%|          | 104/9864 [27:51<39:12:34, 14.46s/it]
  1%|          | 105/9864 [28:07<40:02:44, 14.77s/it]
  1%|          | 106/9864 [28:23<41:01:35, 15.14s/it]
  1%|          | 107/9864 [28:40<42:39:21, 15.74s/it]
  1%|          | 108/9864 [28:54<41:24:31, 15.28s/it]
  1%|          | 109/9864 [29:09<40:36:36, 14.99s/it]
  1%|          | 110/9864 [29:23<39:46:09, 14.68s/it]
                                                     
{'loss': '0.224', 'grad_norm': '8.377', 'learning_rate': '7.365e-06', 'epoch': '0.01115'}

  1%|          | 110/9864 [29:23<39:46:09, 14.68s/it]
  1%|          | 111/9864 [29:37<39:43:11, 14.66s/it]
  1%|          | 112/9864 [29:52<39:39:32, 14.64s/it]
  1%|          | 113/9864 [30:08<40:47:47, 15.06s/it]
  1%|          | 114/9864 [30:25<42:55:50, 15.85s/it]
  1%|          | 115/9864 [30:42<43:18:32, 15.99s/it]
  1%|          | 116/9864 [30:57<42:15:38, 15.61s/it]
  1%|          | 117/9864 [31:11<41:15:29, 15.24s/it]
  1%|          | 118/9864 [31:25<40:21:45, 14.91s/it]
  1%|          | 119/9864 [31:39<39:22:25, 14.55s/it]
  1%|          | 120/9864 [31:53<38:49:29, 14.34s/it]
                                                     
{'loss': '0.2036', 'grad_norm': '12.65', 'learning_rate': '8.041e-06', 'epoch': '0.01217'}

  1%|          | 120/9864 [31:53<38:49:29, 14.34s/it]
  1%|          | 121/9864 [32:09<40:10:44, 14.85s/it]
  1%|          | 122/9864 [32:26<42:17:14, 15.63s/it]
  1%|          | 123/9864 [32:42<42:09:33, 15.58s/it]
  1%|▏         | 124/9864 [32:56<40:59:22, 15.15s/it]
  1%|▏         | 125/9864 [33:10<40:20:24, 14.91s/it]
  1%|▏         | 126/9864 [33:24<39:47:17, 14.71s/it]
  1%|▏         | 127/9864 [33:38<39:15:36, 14.52s/it]
  1%|▏         | 128/9864 [33:54<40:18:03, 14.90s/it]
  1%|▏         | 129/9864 [34:10<40:45:41, 15.07s/it]
  1%|▏         | 130/9864 [34:25<41:04:05, 15.19s/it]
                                                     
{'loss': '0.2071', 'grad_norm': '5.785', 'learning_rate': '8.716e-06', 'epoch': '0.01318'}

  1%|▏         | 130/9864 [34:25<41:04:05, 15.19s/it]
  1%|▏         | 131/9864 [34:41<41:24:07, 15.31s/it]
  1%|▏         | 132/9864 [34:55<40:16:02, 14.90s/it]
  1%|▏         | 133/9864 [35:08<39:24:38, 14.58s/it]
  1%|▏         | 134/9864 [35:23<39:20:31, 14.56s/it]
  1%|▏         | 135/9864 [35:37<38:48:10, 14.36s/it]
  1%|▏         | 136/9864 [35:52<39:47:35, 14.73s/it]
  1%|▏         | 137/9864 [36:08<40:24:11, 14.95s/it]
  1%|▏         | 138/9864 [36:26<42:53:06, 15.87s/it]
  1%|▏         | 139/9864 [36:40<41:28:31, 15.35s/it]
  1%|▏         | 140/9864 [36:54<40:15:36, 14.91s/it]
                                                     
{'loss': '0.2127', 'grad_norm': '9.337', 'learning_rate': '9.392e-06', 'epoch': '0.01419'}

  1%|▏         | 140/9864 [36:54<40:15:36, 14.91s/it]
  1%|▏         | 141/9864 [37:08<39:19:32, 14.56s/it]
  1%|▏         | 142/9864 [37:22<39:16:13, 14.54s/it]
  1%|▏         | 143/9864 [37:36<38:26:38, 14.24s/it]
  1%|▏         | 144/9864 [37:52<40:04:26, 14.84s/it]
  1%|▏         | 145/9864 [38:08<40:55:39, 15.16s/it]
  1%|▏         | 146/9864 [38:25<42:29:29, 15.74s/it]
  1%|▏         | 147/9864 [38:39<41:17:28, 15.30s/it]
  2%|▏         | 148/9864 [38:53<40:25:18, 14.98s/it]
  2%|▏         | 149/9864 [39:08<39:53:12, 14.78s/it]
  2%|▏         | 150/9864 [39:22<39:19:29, 14.57s/it]
                                                     
{'loss': '0.1992', 'grad_norm': '13.57', 'learning_rate': '1.007e-05', 'epoch': '0.01521'}

  2%|▏         | 150/9864 [39:22<39:19:29, 14.57s/it]
  2%|▏         | 151/9864 [39:38<40:32:31, 15.03s/it]
  2%|▏         | 152/9864 [39:52<39:22:24, 14.59s/it]
  2%|▏         | 153/9864 [40:07<40:19:52, 14.95s/it]
  2%|▏         | 154/9864 [40:25<42:22:32, 15.71s/it]
  2%|▏         | 155/9864 [40:39<40:55:41, 15.18s/it]
  2%|▏         | 156/9864 [40:53<39:53:54, 14.80s/it]
  2%|▏         | 157/9864 [41:06<38:52:37, 14.42s/it]
  2%|▏         | 158/9864 [41:22<39:48:11, 14.76s/it]
  2%|▏         | 159/9864 [41:36<38:58:57, 14.46s/it]
  2%|▏         | 160/9864 [41:51<39:49:18, 14.77s/it]
                                                     
{'loss': '0.1818', 'grad_norm': '6.873', 'learning_rate': '1.074e-05', 'epoch': '0.01622'}

  2%|▏         | 160/9864 [41:51<39:49:18, 14.77s/it]
  2%|▏         | 161/9864 [42:07<40:47:48, 15.14s/it]
  2%|▏         | 162/9864 [42:21<39:50:21, 14.78s/it]
  2%|▏         | 163/9864 [42:37<40:37:14, 15.07s/it]
  2%|▏         | 164/9864 [42:51<39:44:34, 14.75s/it]
  2%|▏         | 165/9864 [43:07<40:55:00, 15.19s/it]
  2%|▏         | 166/9864 [43:21<39:56:40, 14.83s/it]
  2%|▏         | 167/9864 [43:35<39:05:01, 14.51s/it]
  2%|▏         | 168/9864 [43:50<40:00:24, 14.85s/it]
  2%|▏         | 169/9864 [44:04<39:01:40, 14.49s/it]
  2%|▏         | 170/9864 [44:21<41:01:22, 15.23s/it]
                                                     
{'loss': '0.1837', 'grad_norm': '10.19', 'learning_rate': '1.142e-05', 'epoch': '0.01724'}

  2%|▏         | 170/9864 [44:21<41:01:22, 15.23s/it]
  2%|▏         | 171/9864 [44:35<40:02:33, 14.87s/it]
  2%|▏         | 172/9864 [44:48<38:56:03, 14.46s/it]
  2%|▏         | 173/9864 [45:04<39:31:55, 14.69s/it]
  2%|▏         | 174/9864 [45:18<39:09:48, 14.55s/it]
  2%|▏         | 175/9864 [45:31<38:15:55, 14.22s/it]
  2%|▏         | 176/9864 [45:46<38:59:54, 14.49s/it]
  2%|▏         | 177/9864 [46:02<39:35:49, 14.72s/it]
  2%|▏         | 178/9864 [46:18<40:33:34, 15.07s/it]
  2%|▏         | 179/9864 [46:32<39:43:12, 14.76s/it]
  2%|▏         | 180/9864 [46:46<39:03:17, 14.52s/it]
                                                     
{'loss': '0.1895', 'grad_norm': '12.03', 'learning_rate': '1.209e-05', 'epoch': '0.01825'}

  2%|▏         | 180/9864 [46:46<39:03:17, 14.52s/it]
  2%|▏         | 181/9864 [47:01<39:32:42, 14.70s/it]
  2%|▏         | 182/9864 [47:15<38:54:57, 14.47s/it]
  2%|▏         | 183/9864 [47:28<38:17:00, 14.24s/it]
  2%|▏         | 184/9864 [47:47<41:25:24, 15.41s/it]
  2%|▏         | 185/9864 [48:00<39:59:32, 14.87s/it]
  2%|▏         | 186/9864 [48:16<40:54:03, 15.21s/it]
  2%|▏         | 187/9864 [48:30<39:57:25, 14.86s/it]
  2%|▏         | 188/9864 [48:45<39:35:01, 14.73s/it]
  2%|▏         | 189/9864 [49:00<40:18:31, 15.00s/it]
  2%|▏         | 190/9864 [49:14<39:33:54, 14.72s/it]
                                                     
{'loss': '0.1876', 'grad_norm': '12.34', 'learning_rate': '1.277e-05', 'epoch': '0.01926'}

  2%|▏         | 190/9864 [49:14<39:33:54, 14.72s/it]
  2%|▏         | 191/9864 [49:30<40:03:40, 14.91s/it]
  2%|▏         | 192/9864 [49:45<40:39:22, 15.13s/it]
  2%|▏         | 193/9864 [49:59<39:20:41, 14.65s/it]
  2%|▏         | 194/9864 [50:15<40:32:29, 15.09s/it]
  2%|▏         | 195/9864 [50:29<39:42:08, 14.78s/it]
  2%|▏         | 196/9864 [50:43<39:01:24, 14.53s/it]
  2%|▏         | 197/9864 [50:58<39:44:09, 14.80s/it]
  2%|▏         | 198/9864 [51:12<39:02:03, 14.54s/it]
  2%|▏         | 199/9864 [51:28<40:14:33, 14.99s/it]
  2%|▏         | 200/9864 [51:43<39:41:58, 14.79s/it]
                                                     
{'loss': '0.2158', 'grad_norm': '11.82', 'learning_rate': '1.345e-05', 'epoch': '0.02028'}

  2%|▏         | 200/9864 [51:43<39:41:58, 14.79s/it]
  2%|▏         | 201/9864 [51:57<39:00:10, 14.53s/it]
  2%|▏         | 202/9864 [52:13<40:10:00, 14.97s/it]
  2%|▏         | 203/9864 [52:27<39:30:38, 14.72s/it]
  2%|▏         | 204/9864 [52:41<38:55:34, 14.51s/it]
  2%|▏         | 205/9864 [52:56<39:41:45, 14.80s/it]
  2%|▏         | 206/9864 [53:10<39:03:30, 14.56s/it]
  2%|▏         | 207/9864 [53:26<40:14:56, 15.00s/it]
  2%|▏         | 208/9864 [53:42<40:43:20, 15.18s/it]
  2%|▏         | 209/9864 [53:56<39:36:02, 14.77s/it]
  2%|▏         | 210/9864 [54:10<39:16:35, 14.65s/it]
                                                     
{'loss': '0.2063', 'grad_norm': '6.693', 'learning_rate': '1.412e-05', 'epoch': '0.02129'}

  2%|▏         | 210/9864 [54:10<39:16:35, 14.65s/it]
  2%|▏         | 211/9864 [54:26<40:13:54, 15.00s/it]
  2%|▏         | 212/9864 [54:42<40:45:59, 15.21s/it]
  2%|▏         | 213/9864 [54:55<39:23:12, 14.69s/it]
  2%|▏         | 214/9864 [55:09<38:54:18, 14.51s/it]
  2%|▏         | 215/9864 [55:24<39:11:36, 14.62s/it]
  2%|▏         | 216/9864 [55:38<38:47:07, 14.47s/it]
  2%|▏         | 217/9864 [55:54<39:35:19, 14.77s/it]
  2%|▏         | 218/9864 [56:09<40:22:24, 15.07s/it]
  2%|▏         | 219/9864 [56:23<39:28:59, 14.74s/it]
  2%|▏         | 220/9864 [56:39<40:09:05, 14.99s/it]
                                                     
{'loss': '0.1935', 'grad_norm': '7.695', 'learning_rate': '1.48e-05', 'epoch': '0.02231'}

  2%|▏         | 220/9864 [56:39<40:09:05, 14.99s/it]
  2%|▏         | 221/9864 [56:53<39:24:53, 14.71s/it]
  2%|▏         | 222/9864 [57:07<38:56:18, 14.54s/it]
  2%|▏         | 223/9864 [57:22<38:57:03, 14.54s/it]
  2%|▏         | 224/9864 [57:39<41:18:52, 15.43s/it]
  2%|▏         | 225/9864 [57:55<41:18:26, 15.43s/it]
  2%|▏         | 226/9864 [58:08<40:04:39, 14.97s/it]
  2%|▏         | 227/9864 [58:22<39:02:51, 14.59s/it]
  2%|▏         | 228/9864 [58:38<39:41:21, 14.83s/it]
  2%|▏         | 229/9864 [58:52<39:15:51, 14.67s/it]
  2%|▏         | 230/9864 [59:06<39:01:15, 14.58s/it]
                                                     
{'loss': '0.1831', 'grad_norm': '7.409', 'learning_rate': '1.547e-05', 'epoch': '0.02332'}

  2%|▏         | 230/9864 [59:06<39:01:15, 14.58s/it]
  2%|▏         | 231/9864 [59:22<40:13:56, 15.04s/it]
  2%|▏         | 232/9864 [59:38<40:42:35, 15.22s/it]
  2%|▏         | 233/9864 [59:53<40:54:09, 15.29s/it]
  2%|▏         | 234/9864 [1:00:08<39:57:30, 14.94s/it]
  2%|▏         | 235/9864 [1:00:24<40:49:05, 15.26s/it]
  2%|▏         | 236/9864 [1:00:38<40:05:07, 14.99s/it]
  2%|▏         | 237/9864 [1:00:52<39:42:39, 14.85s/it]
  2%|▏         | 238/9864 [1:01:06<38:53:21, 14.54s/it]
  2%|▏         | 239/9864 [1:01:23<40:54:04, 15.30s/it]
  2%|▏         | 240/9864 [1:01:39<40:55:14, 15.31s/it]
                                                       
{'loss': '0.1826', 'grad_norm': '5.283', 'learning_rate': '1.615e-05', 'epoch': '0.02433'}

  2%|▏         | 240/9864 [1:01:39<40:55:14, 15.31s/it]
  2%|▏         | 241/9864 [1:01:53<39:47:31, 14.89s/it]
  2%|▏         | 242/9864 [1:02:07<39:22:59, 14.73s/it]
  2%|▏         | 243/9864 [1:02:23<40:38:21, 15.21s/it]
  2%|▏         | 244/9864 [1:02:38<40:07:31, 15.02s/it]
  2%|▏         | 245/9864 [1:02:52<39:32:06, 14.80s/it]
  2%|▏         | 246/9864 [1:03:08<40:13:52, 15.06s/it]
  3%|▎         | 247/9864 [1:03:25<42:19:00, 15.84s/it]
  3%|▎         | 248/9864 [1:03:40<41:02:49, 15.37s/it]
  3%|▎         | 249/9864 [1:03:54<40:06:12, 15.02s/it]
  3%|▎         | 250/9864 [1:04:10<40:39:00, 15.22s/it]
                                                       
{'loss': '0.1997', 'grad_norm': '8.46', 'learning_rate': '1.682e-05', 'epoch': '0.02535'}

  3%|▎         | 250/9864 [1:04:10<40:39:00, 15.22s/it]
  3%|▎         | 251/9864 [1:04:24<39:51:36, 14.93s/it]
  3%|▎         | 252/9864 [1:04:38<38:50:18, 14.55s/it]
  3%|▎         | 253/9864 [1:04:52<38:32:30, 14.44s/it]
  3%|▎         | 254/9864 [1:05:09<40:28:04, 15.16s/it]
  3%|▎         | 255/9864 [1:05:22<38:48:21, 14.54s/it]
  3%|▎         | 256/9864 [1:05:37<39:47:09, 14.91s/it]
  3%|▎         | 257/9864 [1:05:52<39:30:35, 14.81s/it]
  3%|▎         | 258/9864 [1:06:08<40:09:45, 15.05s/it]
  3%|▎         | 259/9864 [1:06:21<38:53:02, 14.57s/it]
  3%|▎         | 260/9864 [1:06:35<38:17:09, 14.35s/it]
                                                       
{'loss': '0.1989', 'grad_norm': '10.05', 'learning_rate': '1.75e-05', 'epoch': '0.02636'}

  3%|▎         | 260/9864 [1:06:35<38:17:09, 14.35s/it]
  3%|▎         | 261/9864 [1:06:48<37:35:22, 14.09s/it]
  3%|▎         | 262/9864 [1:07:04<38:38:34, 14.49s/it]
  3%|▎         | 263/9864 [1:07:20<39:40:32, 14.88s/it]
  3%|▎         | 264/9864 [1:07:36<40:37:51, 15.24s/it]
  3%|▎         | 265/9864 [1:07:50<40:04:01, 15.03s/it]
  3%|▎         | 266/9864 [1:08:06<40:45:34, 15.29s/it]
  3%|▎         | 267/9864 [1:08:20<39:37:00, 14.86s/it]
  3%|▎         | 268/9864 [1:08:35<39:34:46, 14.85s/it]
  3%|▎         | 269/9864 [1:08:48<38:26:55, 14.43s/it]
  3%|▎         | 270/9864 [1:09:06<40:48:01, 15.31s/it]
                                                       
{'loss': '0.1938', 'grad_norm': '7.927', 'learning_rate': '1.818e-05', 'epoch': '0.02737'}

  3%|▎         | 270/9864 [1:09:06<40:48:01, 15.31s/it]
  3%|▎         | 271/9864 [1:09:20<39:51:34, 14.96s/it]
  3%|▎         | 272/9864 [1:09:36<40:48:50, 15.32s/it]
  3%|▎         | 273/9864 [1:09:50<39:41:03, 14.90s/it]
  3%|▎         | 274/9864 [1:10:06<40:38:41, 15.26s/it]
  3%|▎         | 275/9864 [1:10:20<39:43:20, 14.91s/it]
  3%|▎         | 276/9864 [1:10:34<38:56:46, 14.62s/it]
  3%|▎         | 277/9864 [1:10:48<38:36:11, 14.50s/it]
  3%|▎         | 278/9864 [1:11:06<41:19:30, 15.52s/it]
  3%|▎         | 279/9864 [1:11:20<40:06:40, 15.07s/it]
  3%|▎         | 280/9864 [1:11:36<41:00:36, 15.40s/it]
                                                       
{'loss': '0.2036', 'grad_norm': '7.093', 'learning_rate': '1.885e-05', 'epoch': '0.02839'}

  3%|▎         | 280/9864 [1:11:36<41:00:36, 15.40s/it]
  3%|▎         | 281/9864 [1:11:50<39:50:51, 14.97s/it]
  3%|▎         | 282/9864 [1:12:06<40:23:03, 15.17s/it]
  3%|▎         | 283/9864 [1:12:20<39:48:47, 14.96s/it]
  3%|▎         | 284/9864 [1:12:34<38:56:20, 14.63s/it]
  3%|▎         | 285/9864 [1:12:51<40:48:25, 15.34s/it]
  3%|▎         | 286/9864 [1:13:05<39:22:43, 14.80s/it]
  3%|▎         | 287/9864 [1:13:21<40:12:29, 15.11s/it]
  3%|▎         | 288/9864 [1:13:34<38:48:59, 14.59s/it]
  3%|▎         | 289/9864 [1:13:49<38:50:27, 14.60s/it]
  3%|▎         | 290/9864 [1:14:04<39:48:08, 14.97s/it]
                                                       
{'loss': '0.191', 'grad_norm': '3.758', 'learning_rate': '1.953e-05', 'epoch': '0.0294'}

  3%|▎         | 290/9864 [1:14:04<39:48:08, 14.97s/it]
  3%|▎         | 291/9864 [1:14:18<39:00:04, 14.67s/it]
  3%|▎         | 292/9864 [1:14:32<38:14:52, 14.38s/it]
  3%|▎         | 293/9864 [1:14:48<39:29:36, 14.85s/it]
  3%|▎         | 294/9864 [1:15:04<40:15:55, 15.15s/it]
  3%|▎         | 295/9864 [1:15:18<39:33:17, 14.88s/it]
  3%|▎         | 296/9864 [1:15:33<39:56:53, 15.03s/it]
  3%|▎         | 297/9864 [1:15:48<39:21:12, 14.81s/it]
  3%|▎         | 298/9864 [1:16:04<40:10:37, 15.12s/it]
  3%|▎         | 299/9864 [1:16:18<39:17:22, 14.79s/it]
  3%|▎         | 300/9864 [1:16:33<40:03:58, 15.08s/it]
                                                       
{'loss': '0.1768', 'grad_norm': '5.62', 'learning_rate': '2e-05', 'epoch': '0.03042'}

  3%|▎         | 300/9864 [1:16:33<40:03:58, 15.08s/it]
  3%|▎         | 301/9864 [1:16:49<40:47:44, 15.36s/it]
  3%|▎         | 302/9864 [1:17:03<39:43:21, 14.96s/it]
  3%|▎         | 303/9864 [1:17:20<40:37:42, 15.30s/it]
  3%|▎         | 304/9864 [1:17:33<39:29:12, 14.87s/it]
  3%|▎         | 305/9864 [1:17:47<38:52:02, 14.64s/it]
  3%|▎         | 306/9864 [1:18:03<39:50:32, 15.01s/it]
  3%|▎         | 307/9864 [1:18:17<38:32:54, 14.52s/it]
  3%|▎         | 308/9864 [1:18:32<39:14:15, 14.78s/it]
  3%|▎         | 309/9864 [1:18:48<39:54:31, 15.04s/it]
  3%|▎         | 310/9864 [1:19:02<39:18:04, 14.81s/it]
                                                       
{'loss': '0.1869', 'grad_norm': '5.622', 'learning_rate': '2e-05', 'epoch': '0.03143'}

  3%|▎         | 310/9864 [1:19:02<39:18:04, 14.81s/it]
  3%|▎         | 311/9864 [1:19:16<38:37:47, 14.56s/it]
  3%|▎         | 312/9864 [1:19:32<39:50:21, 15.01s/it]
  3%|▎         | 313/9864 [1:19:46<38:53:02, 14.66s/it]
  3%|▎         | 314/9864 [1:20:02<40:04:54, 15.11s/it]
  3%|▎         | 315/9864 [1:20:18<40:38:40, 15.32s/it]
  3%|▎         | 316/9864 [1:20:32<39:19:20, 14.83s/it]
  3%|▎         | 317/9864 [1:20:47<40:05:58, 15.12s/it]
  3%|▎         | 318/9864 [1:21:01<39:11:36, 14.78s/it]
  3%|▎         | 319/9864 [1:21:15<38:33:31, 14.54s/it]
  3%|▎         | 320/9864 [1:21:31<39:45:14, 15.00s/it]
                                                       
{'loss': '0.1867', 'grad_norm': '4.246', 'learning_rate': '2e-05', 'epoch': '0.03244'}

  3%|▎         | 320/9864 [1:21:31<39:45:14, 15.00s/it]
  3%|▎         | 321/9864 [1:21:46<39:08:22, 14.77s/it]
  3%|▎         | 322/9864 [1:22:01<39:39:15, 14.96s/it]
  3%|▎         | 323/9864 [1:22:15<38:58:25, 14.71s/it]
  3%|▎         | 324/9864 [1:22:31<39:43:10, 14.99s/it]
  3%|▎         | 325/9864 [1:22:47<40:44:52, 15.38s/it]
  3%|▎         | 326/9864 [1:23:01<39:17:31, 14.83s/it]
  3%|▎         | 327/9864 [1:23:15<38:46:38, 14.64s/it]
  3%|▎         | 328/9864 [1:23:31<39:46:11, 15.01s/it]
  3%|▎         | 329/9864 [1:23:44<38:33:17, 14.56s/it]
  3%|▎         | 330/9864 [1:24:00<39:30:25, 14.92s/it]
                                                       
{'loss': '0.1738', 'grad_norm': '3.363', 'learning_rate': '2e-05', 'epoch': '0.03346'}

  3%|▎         | 330/9864 [1:24:00<39:30:25, 14.92s/it]
  3%|▎         | 331/9864 [1:24:14<38:43:07, 14.62s/it]
  3%|▎         | 332/9864 [1:24:30<39:31:49, 14.93s/it]
  3%|▎         | 333/9864 [1:24:45<40:11:12, 15.18s/it]
  3%|▎         | 334/9864 [1:25:00<39:31:56, 14.93s/it]
  3%|▎         | 335/9864 [1:25:15<40:02:05, 15.12s/it]
  3%|▎         | 336/9864 [1:25:29<39:12:33, 14.81s/it]
  3%|▎         | 337/9864 [1:25:45<39:47:45, 15.04s/it]
  3%|▎         | 338/9864 [1:25:58<38:36:38, 14.59s/it]
  3%|▎         | 339/9864 [1:26:14<39:24:49, 14.90s/it]
  3%|▎         | 340/9864 [1:26:29<39:47:17, 15.04s/it]
                                                       
{'loss': '0.1736', 'grad_norm': '6.952', 'learning_rate': '2e-05', 'epoch': '0.03447'}

  3%|▎         | 340/9864 [1:26:29<39:47:17, 15.04s/it]
  3%|▎         | 341/9864 [1:26:43<39:00:20, 14.75s/it]
  3%|▎         | 342/9864 [1:26:57<37:57:42, 14.35s/it]
  3%|▎         | 343/9864 [1:27:13<38:57:01, 14.73s/it]
  3%|▎         | 344/9864 [1:27:27<38:24:06, 14.52s/it]
  3%|▎         | 345/9864 [1:27:42<39:14:23, 14.84s/it]
  4%|▎         | 346/9864 [1:27:58<40:04:17, 15.16s/it]
  4%|▎         | 347/9864 [1:28:14<40:24:45, 15.29s/it]
  4%|▎         | 348/9864 [1:28:28<39:47:21, 15.05s/it]
  4%|▎         | 349/9864 [1:28:43<39:21:38, 14.89s/it]
  4%|▎         | 350/9864 [1:28:57<39:09:12, 14.82s/it]
                                                       
{'loss': '0.1696', 'grad_norm': '6.773', 'learning_rate': '2e-05', 'epoch': '0.03549'}

  4%|▎         | 350/9864 [1:28:57<39:09:12, 14.82s/it]
  4%|▎         | 351/9864 [1:29:13<39:51:14, 15.08s/it]
  4%|▎         | 352/9864 [1:29:27<39:12:47, 14.84s/it]
  4%|▎         | 353/9864 [1:29:43<39:58:25, 15.13s/it]
  4%|▎         | 354/9864 [1:30:01<41:48:26, 15.83s/it]
  4%|▎         | 355/9864 [1:30:15<40:32:38, 15.35s/it]
  4%|▎         | 356/9864 [1:30:29<39:40:46, 15.02s/it]
  4%|▎         | 357/9864 [1:30:43<39:08:56, 14.82s/it]
  4%|▎         | 358/9864 [1:30:57<38:23:58, 14.54s/it]
  4%|▎         | 359/9864 [1:31:13<39:25:21, 14.93s/it]
  4%|▎         | 360/9864 [1:31:27<38:38:00, 14.63s/it]
                                                       
{'loss': '0.1876', 'grad_norm': '4.116', 'learning_rate': '2e-05', 'epoch': '0.0365'}

  4%|▎         | 360/9864 [1:31:27<38:38:00, 14.63s/it]
  4%|▎         | 361/9864 [1:31:42<38:59:10, 14.77s/it]
  4%|▎         | 362/9864 [1:31:58<39:47:04, 15.07s/it]
  4%|▎         | 363/9864 [1:32:14<40:31:36, 15.36s/it]
  4%|▎         | 364/9864 [1:32:27<39:06:28, 14.82s/it]
  4%|▎         | 365/9864 [1:32:42<38:57:35, 14.77s/it]
  4%|▎         | 366/9864 [1:32:58<39:27:20, 14.95s/it]
  4%|▎         | 367/9864 [1:33:11<38:30:44, 14.60s/it]
  4%|▎         | 368/9864 [1:33:26<38:16:41, 14.51s/it]
  4%|▎         | 369/9864 [1:33:41<38:40:42, 14.66s/it]
  4%|▍         | 370/9864 [1:33:56<39:18:59, 14.91s/it]
                                                       
{'loss': '0.1922', 'grad_norm': '10.04', 'learning_rate': '2e-05', 'epoch': '0.03751'}

  4%|▍         | 370/9864 [1:33:56<39:18:59, 14.91s/it]
  4%|▍         | 371/9864 [1:34:12<39:46:15, 15.08s/it]
  4%|▍         | 372/9864 [1:34:25<38:42:31, 14.68s/it]
  4%|▍         | 373/9864 [1:34:39<38:09:23, 14.47s/it]
  4%|▍         | 374/9864 [1:34:55<39:01:52, 14.81s/it]
  4%|▍         | 375/9864 [1:35:09<38:26:25, 14.58s/it]
  4%|▍         | 376/9864 [1:35:25<39:25:23, 14.96s/it]
  4%|▍         | 377/9864 [1:35:41<40:05:23, 15.21s/it]
  4%|▍         | 378/9864 [1:35:55<39:23:47, 14.95s/it]
  4%|▍         | 379/9864 [1:36:11<40:34:33, 15.40s/it]
  4%|▍         | 380/9864 [1:36:26<39:51:31, 15.13s/it]
                                                       
{'loss': '0.1758', 'grad_norm': '5.396', 'learning_rate': '2e-05', 'epoch': '0.03853'}

  4%|▍         | 380/9864 [1:36:26<39:51:31, 15.13s/it]
  4%|▍         | 381/9864 [1:36:40<38:42:26, 14.69s/it]
  4%|▍         | 382/9864 [1:36:56<39:47:23, 15.11s/it]
  4%|▍         | 383/9864 [1:37:10<39:17:44, 14.92s/it]
  4%|▍         | 384/9864 [1:37:27<41:01:12, 15.58s/it]
  4%|▍         | 385/9864 [1:37:41<39:57:15, 15.17s/it]
  4%|▍         | 386/9864 [1:37:57<39:52:13, 15.14s/it]
  4%|▍         | 387/9864 [1:38:11<39:14:45, 14.91s/it]
  4%|▍         | 388/9864 [1:38:25<38:43:49, 14.71s/it]
  4%|▍         | 389/9864 [1:38:39<37:47:51, 14.36s/it]
  4%|▍         | 390/9864 [1:38:54<38:53:48, 14.78s/it]
                                                       
{'loss': '0.1947', 'grad_norm': '5.242', 'learning_rate': '2e-05', 'epoch': '0.03954'}

  4%|▍         | 390/9864 [1:38:54<38:53:48, 14.78s/it]
  4%|▍         | 391/9864 [1:39:10<39:45:04, 15.11s/it]
  4%|▍         | 392/9864 [1:39:26<39:57:30, 15.19s/it]
  4%|▍         | 393/9864 [1:39:40<39:00:04, 14.82s/it]
  4%|▍         | 394/9864 [1:39:55<39:03:06, 14.85s/it]
  4%|▍         | 395/9864 [1:40:09<38:36:15, 14.68s/it]
  4%|▍         | 396/9864 [1:40:23<38:09:56, 14.51s/it]
  4%|▍         | 397/9864 [1:40:37<37:51:54, 14.40s/it]
  4%|▍         | 398/9864 [1:40:53<38:44:04, 14.73s/it]
  4%|▍         | 399/9864 [1:41:08<39:27:52, 15.01s/it]
  4%|▍         | 400/9864 [1:41:24<40:16:52, 15.32s/it]
                                                       
{'loss': '0.1818', 'grad_norm': '7.072', 'learning_rate': '1.999e-05', 'epoch': '0.04055'}

  4%|▍         | 400/9864 [1:41:24<40:16:52, 15.32s/it]
  4%|▍         | 401/9864 [1:41:40<40:46:32, 15.51s/it]
  4%|▍         | 402/9864 [1:41:54<39:27:39, 15.01s/it]
  4%|▍         | 403/9864 [1:42:08<38:30:07, 14.65s/it]
  4%|▍         | 404/9864 [1:42:21<37:25:29, 14.24s/it]
  4%|▍         | 405/9864 [1:42:37<38:53:12, 14.80s/it]
  4%|▍         | 406/9864 [1:42:51<38:07:00, 14.51s/it]
  4%|▍         | 407/9864 [1:43:07<38:48:56, 14.78s/it]
  4%|▍         | 408/9864 [1:43:23<39:48:02, 15.15s/it]
  4%|▍         | 409/9864 [1:43:38<40:17:26, 15.34s/it]
  4%|▍         | 410/9864 [1:43:52<39:09:13, 14.91s/it]
                                                       
{'loss': '0.176', 'grad_norm': '3.386', 'learning_rate': '1.999e-05', 'epoch': '0.04157'}

  4%|▍         | 410/9864 [1:43:52<39:09:13, 14.91s/it]
  4%|▍         | 411/9864 [1:44:07<38:37:35, 14.71s/it]
  4%|▍         | 412/9864 [1:44:21<38:23:48, 14.62s/it]
  4%|▍         | 413/9864 [1:44:37<39:30:52, 15.05s/it]
  4%|▍         | 414/9864 [1:44:53<40:07:48, 15.29s/it]
  4%|▍         | 415/9864 [1:45:07<39:06:01, 14.90s/it]
  4%|▍         | 416/9864 [1:45:23<39:47:48, 15.16s/it]
  4%|▍         | 417/9864 [1:45:38<40:02:58, 15.26s/it]
  4%|▍         | 418/9864 [1:45:53<39:22:15, 15.00s/it]
  4%|▍         | 419/9864 [1:46:06<38:27:59, 14.66s/it]
  4%|▍         | 420/9864 [1:46:22<39:21:20, 15.00s/it]
                                                       
{'loss': '0.2052', 'grad_norm': '8.581', 'learning_rate': '1.999e-05', 'epoch': '0.04258'}

  4%|▍         | 420/9864 [1:46:22<39:21:20, 15.00s/it]
  4%|▍         | 421/9864 [1:46:38<39:52:21, 15.20s/it]
  4%|▍         | 422/9864 [1:46:52<39:06:25, 14.91s/it]
  4%|▍         | 423/9864 [1:47:06<38:22:22, 14.63s/it]
  4%|▍         | 424/9864 [1:47:21<39:00:35, 14.88s/it]
  4%|▍         | 425/9864 [1:47:37<39:49:26, 15.19s/it]
  4%|▍         | 426/9864 [1:47:52<39:13:24, 14.96s/it]
  4%|▍         | 427/9864 [1:48:06<38:44:42, 14.78s/it]
  4%|▍         | 428/9864 [1:48:22<39:49:25, 15.19s/it]
  4%|▍         | 429/9864 [1:48:38<40:18:34, 15.38s/it]
  4%|▍         | 430/9864 [1:48:52<38:59:32, 14.88s/it]
                                                       
{'loss': '0.178', 'grad_norm': '4.776', 'learning_rate': '1.999e-05', 'epoch': '0.0436'}

  4%|▍         | 430/9864 [1:48:52<38:59:32, 14.88s/it]
  4%|▍         | 431/9864 [1:49:06<38:05:53, 14.54s/it]
  4%|▍         | 432/9864 [1:49:19<37:17:30, 14.23s/it]
  4%|▍         | 433/9864 [1:49:37<39:59:31, 15.27s/it]
  4%|▍         | 434/9864 [1:49:51<38:58:28, 14.88s/it]
  4%|▍         | 435/9864 [1:50:05<38:40:18, 14.76s/it]
  4%|▍         | 436/9864 [1:50:22<40:03:06, 15.29s/it]
  4%|▍         | 437/9864 [1:50:37<40:07:04, 15.32s/it]
  4%|▍         | 438/9864 [1:50:51<38:56:41, 14.87s/it]
  4%|▍         | 439/9864 [1:51:05<38:26:01, 14.68s/it]
  4%|▍         | 440/9864 [1:51:19<37:49:49, 14.45s/it]
                                                       
{'loss': '0.1841', 'grad_norm': '3.677', 'learning_rate': '1.999e-05', 'epoch': '0.04461'}

  4%|▍         | 440/9864 [1:51:19<37:49:49, 14.45s/it]
  4%|▍         | 441/9864 [1:51:35<39:01:15, 14.91s/it]
  4%|▍         | 442/9864 [1:51:51<39:44:40, 15.19s/it]
  4%|▍         | 443/9864 [1:52:05<38:58:06, 14.89s/it]
  5%|▍         | 444/9864 [1:52:21<39:49:13, 15.22s/it]
  5%|▍         | 445/9864 [1:52:37<40:19:31, 15.41s/it]
  5%|▍         | 446/9864 [1:52:51<39:08:10, 14.96s/it]
  5%|▍         | 447/9864 [1:53:05<38:19:44, 14.65s/it]
  5%|▍         | 448/9864 [1:53:18<37:29:49, 14.34s/it]
  5%|▍         | 449/9864 [1:53:35<38:48:52, 14.84s/it]
  5%|▍         | 450/9864 [1:53:50<39:27:37, 15.09s/it]
                                                       
{'loss': '0.183', 'grad_norm': '8.513', 'learning_rate': '1.999e-05', 'epoch': '0.04562'}

  5%|▍         | 450/9864 [1:53:50<39:27:37, 15.09s/it]
  5%|▍         | 451/9864 [1:54:06<40:01:45, 15.31s/it]
  5%|▍         | 452/9864 [1:54:22<40:33:42, 15.51s/it]
  5%|▍         | 453/9864 [1:54:36<39:29:57, 15.11s/it]
  5%|▍         | 454/9864 [1:54:50<38:07:16, 14.58s/it]
  5%|▍         | 455/9864 [1:55:04<37:49:33, 14.47s/it]
  5%|▍         | 456/9864 [1:55:18<37:26:41, 14.33s/it]
  5%|▍         | 457/9864 [1:55:34<38:41:17, 14.81s/it]
  5%|▍         | 458/9864 [1:55:50<39:44:08, 15.21s/it]
  5%|▍         | 459/9864 [1:56:06<40:16:13, 15.41s/it]
  5%|▍         | 460/9864 [1:56:21<40:16:13, 15.42s/it]
                                                       
{'loss': '0.1881', 'grad_norm': '5.914', 'learning_rate': '1.999e-05', 'epoch': '0.04664'}

  5%|▍         | 460/9864 [1:56:21<40:16:13, 15.42s/it]
  5%|▍         | 461/9864 [1:56:35<39:14:17, 15.02s/it]
  5%|▍         | 462/9864 [1:56:49<38:20:14, 14.68s/it]
  5%|▍         | 463/9864 [1:57:03<37:59:51, 14.55s/it]
  5%|▍         | 464/9864 [1:57:17<37:19:38, 14.30s/it]
  5%|▍         | 465/9864 [1:57:34<39:29:24, 15.13s/it]
  5%|▍         | 466/9864 [1:57:48<38:47:43, 14.86s/it]
  5%|▍         | 467/9864 [1:58:04<39:38:37, 15.19s/it]
  5%|▍         | 468/9864 [1:58:20<39:57:29, 15.31s/it]
  5%|▍         | 469/9864 [1:58:34<38:52:53, 14.90s/it]
  5%|▍         | 470/9864 [1:58:48<38:14:22, 14.65s/it]
                                                       
{'loss': '0.1889', 'grad_norm': '3.995', 'learning_rate': '1.998e-05', 'epoch': '0.04765'}

  5%|▍         | 470/9864 [1:58:48<38:14:22, 14.65s/it]
  5%|▍         | 471/9864 [1:59:02<37:43:07, 14.46s/it]
  5%|▍         | 472/9864 [1:59:17<38:10:07, 14.63s/it]
  5%|▍         | 473/9864 [1:59:32<38:52:50, 14.90s/it]
  5%|▍         | 474/9864 [1:59:47<38:23:42, 14.72s/it]
  5%|▍         | 475/9864 [2:00:01<37:52:30, 14.52s/it]
  5%|▍         | 476/9864 [2:00:16<38:11:39, 14.65s/it]
  5%|▍         | 477/9864 [2:00:32<39:42:48, 15.23s/it]
  5%|▍         | 478/9864 [2:00:47<38:56:06, 14.93s/it]
  5%|▍         | 479/9864 [2:01:01<38:26:05, 14.74s/it]
  5%|▍         | 480/9864 [2:01:16<38:51:04, 14.90s/it]
                                                       
{'loss': '0.1847', 'grad_norm': '6.4', 'learning_rate': '1.998e-05', 'epoch': '0.04867'}

  5%|▍         | 480/9864 [2:01:16<38:51:04, 14.90s/it]
  5%|▍         | 481/9864 [2:01:32<39:52:59, 15.30s/it]
  5%|▍         | 482/9864 [2:01:46<38:46:30, 14.88s/it]
  5%|▍         | 483/9864 [2:02:02<39:08:27, 15.02s/it]
  5%|▍         | 484/9864 [2:02:17<39:10:50, 15.04s/it]
  5%|▍         | 485/9864 [2:02:31<38:39:39, 14.84s/it]
  5%|▍         | 486/9864 [2:02:46<38:21:40, 14.73s/it]
  5%|▍         | 487/9864 [2:03:00<37:54:02, 14.55s/it]
  5%|▍         | 488/9864 [2:03:16<38:54:23, 14.94s/it]
  5%|▍         | 489/9864 [2:03:30<38:52:10, 14.93s/it]
  5%|▍         | 490/9864 [2:03:45<38:17:39, 14.71s/it]
                                                       
{'loss': '0.2007', 'grad_norm': '2.674', 'learning_rate': '1.998e-05', 'epoch': '0.04968'}

  5%|▍         | 490/9864 [2:03:45<38:17:39, 14.71s/it]
  5%|▍         | 491/9864 [2:04:01<39:18:40, 15.10s/it]
  5%|▍         | 492/9864 [2:04:16<39:50:18, 15.30s/it]
  5%|▍         | 493/9864 [2:04:30<38:46:26, 14.90s/it]
  5%|▌         | 494/9864 [2:04:45<38:37:03, 14.84s/it]
  5%|▌         | 495/9864 [2:04:59<38:01:34, 14.61s/it]
  5%|▌         | 496/9864 [2:05:15<38:37:02, 14.84s/it]
  5%|▌         | 497/9864 [2:05:31<39:44:56, 15.28s/it]
  5%|▌         | 498/9864 [2:05:45<38:57:06, 14.97s/it]
  5%|▌         | 499/9864 [2:06:01<39:34:28, 15.21s/it]
  5%|▌         | 500/9864 [2:06:17<40:14:32, 15.47s/it]
                                                       
{'loss': '0.1932', 'grad_norm': '5.935', 'learning_rate': '1.998e-05', 'epoch': '0.05069'}

  5%|▌         | 500/9864 [2:06:17<40:14:32, 15.47s/it]
  5%|▌         | 501/9864 [2:06:31<39:25:26, 15.16s/it]
  5%|▌         | 502/9864 [2:06:45<38:22:36, 14.76s/it]
  5%|▌         | 503/9864 [2:06:59<37:51:19, 14.56s/it]
  5%|▌         | 504/9864 [2:07:15<38:48:08, 14.92s/it]
  5%|▌         | 505/9864 [2:07:31<39:13:12, 15.09s/it]
  5%|▌         | 506/9864 [2:07:45<38:35:12, 14.84s/it]
  5%|▌         | 507/9864 [2:08:02<40:25:02, 15.55s/it]
  5%|▌         | 508/9864 [2:08:16<39:12:37, 15.09s/it]
  5%|▌         | 509/9864 [2:08:30<38:05:36, 14.66s/it]
  5%|▌         | 510/9864 [2:08:44<37:27:08, 14.41s/it]
                                                       
{'loss': '0.1829', 'grad_norm': '2.892', 'learning_rate': '1.998e-05', 'epoch': '0.05171'}

  5%|▌         | 510/9864 [2:08:44<37:27:08, 14.41s/it]
  5%|▌         | 511/9864 [2:08:57<36:50:00, 14.18s/it]
  5%|▌         | 512/9864 [2:09:12<37:35:08, 14.47s/it]
  5%|▌         | 513/9864 [2:09:29<38:57:47, 15.00s/it]
  5%|▌         | 514/9864 [2:09:43<38:27:08, 14.81s/it]
  5%|▌         | 515/9864 [2:10:00<40:29:27, 15.59s/it]
  5%|▌         | 516/9864 [2:10:14<39:04:53, 15.05s/it]
  5%|▌         | 517/9864 [2:10:28<38:30:18, 14.83s/it]
  5%|▌         | 518/9864 [2:10:42<37:43:55, 14.53s/it]
  5%|▌         | 519/9864 [2:10:57<38:14:17, 14.73s/it]
  5%|▌         | 520/9864 [2:11:13<38:56:04, 15.00s/it]
                                                       
{'loss': '0.1827', 'grad_norm': '3.642', 'learning_rate': '1.997e-05', 'epoch': '0.05272'}

  5%|▌         | 520/9864 [2:11:13<38:56:04, 15.00s/it]
  5%|▌         | 521/9864 [2:11:27<38:06:07, 14.68s/it]
  5%|▌         | 522/9864 [2:11:45<40:42:47, 15.69s/it]
  5%|▌         | 523/9864 [2:11:59<39:20:42, 15.16s/it]
  5%|▌         | 524/9864 [2:12:13<38:35:52, 14.88s/it]
  5%|▌         | 525/9864 [2:12:27<37:46:42, 14.56s/it]
  5%|▌         | 526/9864 [2:12:41<37:32:55, 14.48s/it]
  5%|▌         | 527/9864 [2:12:57<38:27:10, 14.83s/it]
  5%|▌         | 528/9864 [2:13:13<39:15:36, 15.14s/it]
  5%|▌         | 529/9864 [2:13:29<39:46:59, 15.34s/it]
  5%|▌         | 530/9864 [2:13:45<40:11:16, 15.50s/it]
                                                       
{'loss': '0.1724', 'grad_norm': '6.453', 'learning_rate': '1.997e-05', 'epoch': '0.05374'}

  5%|▌         | 530/9864 [2:13:45<40:11:16, 15.50s/it]
  5%|▌         | 531/9864 [2:13:59<39:04:25, 15.07s/it]
  5%|▌         | 532/9864 [2:14:12<37:58:10, 14.65s/it]
  5%|▌         | 533/9864 [2:14:26<37:22:33, 14.42s/it]
  5%|▌         | 534/9864 [2:14:40<37:08:19, 14.33s/it]
  5%|▌         | 535/9864 [2:14:56<38:10:56, 14.73s/it]
  5%|▌         | 536/9864 [2:15:11<38:48:58, 14.98s/it]
  5%|▌         | 537/9864 [2:15:27<39:17:02, 15.16s/it]
  5%|▌         | 538/9864 [2:15:41<38:29:17, 14.86s/it]
  5%|▌         | 539/9864 [2:15:57<39:04:14, 15.08s/it]
  5%|▌         | 540/9864 [2:16:11<38:20:59, 14.81s/it]
                                                       
{'loss': '0.1623', 'grad_norm': '3.412', 'learning_rate': '1.997e-05', 'epoch': '0.05475'}

  5%|▌         | 540/9864 [2:16:11<38:20:59, 14.81s/it]
  5%|▌         | 541/9864 [2:16:25<37:37:10, 14.53s/it]
  5%|▌         | 542/9864 [2:16:39<37:13:38, 14.38s/it]
  6%|▌         | 543/9864 [2:16:53<37:00:13, 14.29s/it]
  6%|▌         | 544/9864 [2:17:10<39:11:08, 15.14s/it]
  6%|▌         | 545/9864 [2:17:26<39:37:09, 15.31s/it]
  6%|▌         | 546/9864 [2:17:39<38:05:45, 14.72s/it]
  6%|▌         | 547/9864 [2:17:55<39:06:31, 15.11s/it]
  6%|▌         | 548/9864 [2:18:09<37:58:19, 14.67s/it]
  6%|▌         | 549/9864 [2:18:23<37:24:53, 14.46s/it]
  6%|▌         | 550/9864 [2:18:37<37:09:52, 14.36s/it]
                                                       
{'loss': '0.1654', 'grad_norm': '5.395', 'learning_rate': '1.997e-05', 'epoch': '0.05576'}

  6%|▌         | 550/9864 [2:18:37<37:09:52, 14.36s/it]
  6%|▌         | 551/9864 [2:18:52<37:53:01, 14.64s/it]
  6%|▌         | 552/9864 [2:19:08<38:41:43, 14.96s/it]
  6%|▌         | 553/9864 [2:19:24<39:16:23, 15.18s/it]
  6%|▌         | 554/9864 [2:19:38<38:29:49, 14.89s/it]
  6%|▌         | 555/9864 [2:19:53<38:55:33, 15.05s/it]
  6%|▌         | 556/9864 [2:20:08<38:34:37, 14.92s/it]
  6%|▌         | 557/9864 [2:20:22<37:55:37, 14.67s/it]
  6%|▌         | 558/9864 [2:20:36<37:11:18, 14.39s/it]
  6%|▌         | 559/9864 [2:20:50<37:01:17, 14.32s/it]
  6%|▌         | 560/9864 [2:21:08<40:19:52, 15.61s/it]
                                                       
{'loss': '0.1789', 'grad_norm': '4.032', 'learning_rate': '1.996e-05', 'epoch': '0.05678'}

  6%|▌         | 560/9864 [2:21:08<40:19:52, 15.61s/it]
  6%|▌         | 561/9864 [2:21:23<39:19:08, 15.22s/it]
  6%|▌         | 562/9864 [2:21:38<39:36:52, 15.33s/it]
  6%|▌         | 563/9864 [2:21:52<38:21:24, 14.85s/it]
  6%|▌         | 564/9864 [2:22:06<37:20:22, 14.45s/it]
  6%|▌         | 565/9864 [2:22:19<36:47:22, 14.24s/it]
  6%|▌         | 566/9864 [2:22:33<36:06:14, 13.98s/it]
  6%|▌         | 567/9864 [2:22:50<38:24:05, 14.87s/it]
  6%|▌         | 568/9864 [2:23:03<37:33:00, 14.54s/it]
  6%|▌         | 569/9864 [2:23:21<39:38:53, 15.36s/it]
  6%|▌         | 570/9864 [2:23:35<38:36:45, 14.96s/it]
                                                       
{'loss': '0.1801', 'grad_norm': '3.744', 'learning_rate': '1.996e-05', 'epoch': '0.05779'}

  6%|▌         | 570/9864 [2:23:35<38:36:45, 14.96s/it]
  6%|▌         | 571/9864 [2:23:49<37:59:39, 14.72s/it]
  6%|▌         | 572/9864 [2:24:03<37:33:28, 14.55s/it]
  6%|▌         | 573/9864 [2:24:17<36:56:51, 14.32s/it]
  6%|▌         | 574/9864 [2:24:31<36:58:22, 14.33s/it]
  6%|▌         | 575/9864 [2:24:47<37:58:19, 14.72s/it]
  6%|▌         | 576/9864 [2:25:04<40:03:47, 15.53s/it]
  6%|▌         | 577/9864 [2:25:20<40:18:30, 15.63s/it]
  6%|▌         | 578/9864 [2:25:34<39:07:00, 15.16s/it]
  6%|▌         | 579/9864 [2:25:48<37:59:25, 14.73s/it]
  6%|▌         | 580/9864 [2:26:02<37:22:20, 14.49s/it]
                                                       
{'loss': '0.1741', 'grad_norm': '4.006', 'learning_rate': '1.996e-05', 'epoch': '0.0588'}

  6%|▌         | 580/9864 [2:26:02<37:22:20, 14.49s/it]
  6%|▌         | 581/9864 [2:26:15<36:42:41, 14.24s/it]
  6%|▌         | 582/9864 [2:26:29<36:18:27, 14.08s/it]
  6%|▌         | 583/9864 [2:26:45<37:54:33, 14.70s/it]
  6%|▌         | 584/9864 [2:27:02<39:39:22, 15.38s/it]
  6%|▌         | 585/9864 [2:27:18<39:42:29, 15.41s/it]
  6%|▌         | 586/9864 [2:27:32<38:46:20, 15.04s/it]
  6%|▌         | 587/9864 [2:27:46<37:46:08, 14.66s/it]
  6%|▌         | 588/9864 [2:28:00<37:44:35, 14.65s/it]
  6%|▌         | 589/9864 [2:28:14<37:09:42, 14.42s/it]
  6%|▌         | 590/9864 [2:28:28<36:37:28, 14.22s/it]
                                                       
{'loss': '0.1591', 'grad_norm': '4.614', 'learning_rate': '1.995e-05', 'epoch': '0.05982'}

  6%|▌         | 590/9864 [2:28:28<36:37:28, 14.22s/it]
  6%|▌         | 591/9864 [2:28:44<37:57:03, 14.73s/it]
  6%|▌         | 592/9864 [2:29:00<38:58:54, 15.14s/it]
  6%|▌         | 593/9864 [2:29:16<39:38:19, 15.39s/it]
  6%|▌         | 594/9864 [2:29:32<40:10:33, 15.60s/it]
  6%|▌         | 595/9864 [2:29:46<38:59:44, 15.15s/it]
  6%|▌         | 596/9864 [2:29:59<37:26:11, 14.54s/it]
  6%|▌         | 597/9864 [2:30:14<37:11:23, 14.45s/it]
  6%|▌         | 598/9864 [2:30:30<38:25:26, 14.93s/it]
  6%|▌         | 599/9864 [2:30:46<39:31:24, 15.36s/it]
  6%|▌         | 600/9864 [2:31:00<38:49:29, 15.09s/it]
                                                       
{'loss': '0.1846', 'grad_norm': '4.469', 'learning_rate': '1.995e-05', 'epoch': '0.06083'}

  6%|▌         | 600/9864 [2:31:00<38:49:29, 15.09s/it]
  6%|▌         | 601/9864 [2:31:18<40:47:38, 15.85s/it]
  6%|▌         | 602/9864 [2:31:32<39:23:57, 15.31s/it]
  6%|▌         | 603/9864 [2:31:46<38:21:26, 14.91s/it]
  6%|▌         | 604/9864 [2:32:00<37:58:34, 14.76s/it]
  6%|▌         | 605/9864 [2:32:16<38:50:02, 15.10s/it]
  6%|▌         | 606/9864 [2:32:30<37:50:45, 14.72s/it]
  6%|▌         | 607/9864 [2:32:46<39:02:16, 15.18s/it]
  6%|▌         | 608/9864 [2:33:02<39:15:33, 15.27s/it]
  6%|▌         | 609/9864 [2:33:17<38:53:54, 15.13s/it]
  6%|▌         | 610/9864 [2:33:33<39:26:59, 15.35s/it]
                                                       
{'loss': '0.1863', 'grad_norm': '4.247', 'learning_rate': '1.995e-05', 'epoch': '0.06185'}

  6%|▌         | 610/9864 [2:33:33<39:26:59, 15.35s/it]
  6%|▌         | 611/9864 [2:33:47<38:34:12, 15.01s/it]
  6%|▌         | 612/9864 [2:34:01<38:02:43, 14.80s/it]
  6%|▌         | 613/9864 [2:34:17<38:52:41, 15.13s/it]
  6%|▌         | 614/9864 [2:34:31<37:54:42, 14.75s/it]
  6%|▌         | 615/9864 [2:34:48<39:42:29, 15.46s/it]
  6%|▌         | 616/9864 [2:35:02<38:33:23, 15.01s/it]
  6%|▋         | 617/9864 [2:35:16<38:04:15, 14.82s/it]
  6%|▋         | 618/9864 [2:35:32<38:23:47, 14.95s/it]
  6%|▋         | 619/9864 [2:35:45<37:32:53, 14.62s/it]
  6%|▋         | 620/9864 [2:36:02<38:42:20, 15.07s/it]
                                                       
{'loss': '0.1806', 'grad_norm': '3.259', 'learning_rate': '1.994e-05', 'epoch': '0.06286'}

  6%|▋         | 620/9864 [2:36:02<38:42:20, 15.07s/it]
  6%|▋         | 621/9864 [2:36:15<37:46:06, 14.71s/it]
  6%|▋         | 622/9864 [2:36:31<38:34:32, 15.03s/it]
  6%|▋         | 623/9864 [2:36:47<38:51:31, 15.14s/it]
  6%|▋         | 624/9864 [2:37:01<38:09:20, 14.87s/it]
  6%|▋         | 625/9864 [2:37:15<37:29:12, 14.61s/it]
  6%|▋         | 626/9864 [2:37:30<37:53:04, 14.76s/it]
  6%|▋         | 627/9864 [2:37:44<37:16:30, 14.53s/it]
  6%|▋         | 628/9864 [2:37:58<36:41:03, 14.30s/it]
  6%|▋         | 629/9864 [2:38:16<39:35:50, 15.44s/it]
  6%|▋         | 630/9864 [2:38:30<38:33:59, 15.04s/it]
                                                       
{'loss': '0.172', 'grad_norm': '4.703', 'learning_rate': '1.994e-05', 'epoch': '0.06387'}

  6%|▋         | 630/9864 [2:38:30<38:33:59, 15.04s/it]
  6%|▋         | 631/9864 [2:38:46<39:11:59, 15.28s/it]
  6%|▋         | 632/9864 [2:39:00<38:05:08, 14.85s/it]
  6%|▋         | 633/9864 [2:39:14<37:23:02, 14.58s/it]
  6%|▋         | 634/9864 [2:39:29<37:58:24, 14.81s/it]
  6%|▋         | 635/9864 [2:39:43<37:19:52, 14.56s/it]
  6%|▋         | 636/9864 [2:39:59<38:19:21, 14.95s/it]
  6%|▋         | 637/9864 [2:40:14<38:34:48, 15.05s/it]
  6%|▋         | 638/9864 [2:40:30<39:17:27, 15.33s/it]
  6%|▋         | 639/9864 [2:40:44<38:09:19, 14.89s/it]
  6%|▋         | 640/9864 [2:40:58<37:38:01, 14.69s/it]
                                                       
{'loss': '0.1879', 'grad_norm': '4.637', 'learning_rate': '1.994e-05', 'epoch': '0.06489'}

  6%|▋         | 640/9864 [2:40:58<37:38:01, 14.69s/it]
  6%|▋         | 641/9864 [2:41:12<37:00:21, 14.44s/it]
  7%|▋         | 642/9864 [2:41:28<38:03:05, 14.85s/it]
  7%|▋         | 643/9864 [2:41:42<37:31:30, 14.65s/it]
  7%|▋         | 644/9864 [2:41:58<38:16:33, 14.95s/it]
  7%|▋         | 645/9864 [2:42:13<38:54:10, 15.19s/it]
  7%|▋         | 646/9864 [2:42:29<39:20:07, 15.36s/it]
  7%|▋         | 647/9864 [2:42:44<38:38:01, 15.09s/it]
  7%|▋         | 648/9864 [2:42:58<38:00:04, 14.84s/it]
  7%|▋         | 649/9864 [2:43:12<37:32:52, 14.67s/it]
  7%|▋         | 650/9864 [2:43:28<38:34:42, 15.07s/it]
                                                       
{'loss': '0.1756', 'grad_norm': '5.629', 'learning_rate': '1.993e-05', 'epoch': '0.0659'}

  7%|▋         | 650/9864 [2:43:28<38:34:42, 15.07s/it]
  7%|▋         | 651/9864 [2:43:42<37:46:02, 14.76s/it]
  7%|▋         | 652/9864 [2:43:58<38:35:26, 15.08s/it]
  7%|▋         | 653/9864 [2:44:13<38:51:57, 15.19s/it]
  7%|▋         | 654/9864 [2:44:27<37:54:48, 14.82s/it]
  7%|▋         | 655/9864 [2:44:43<38:31:44, 15.06s/it]
  7%|▋         | 656/9864 [2:44:57<37:45:32, 14.76s/it]
  7%|▋         | 657/9864 [2:45:11<36:56:34, 14.44s/it]
  7%|▋         | 658/9864 [2:45:26<37:21:52, 14.61s/it]
  7%|▋         | 659/9864 [2:45:40<36:48:32, 14.40s/it]
  7%|▋         | 660/9864 [2:45:57<39:05:24, 15.29s/it]
                                                       
{'loss': '0.1727', 'grad_norm': '6.254', 'learning_rate': '1.993e-05', 'epoch': '0.06692'}

  7%|▋         | 660/9864 [2:45:57<39:05:24, 15.29s/it]
  7%|▋         | 661/9864 [2:46:11<37:52:08, 14.81s/it]
  7%|▋         | 662/9864 [2:46:26<38:19:26, 14.99s/it]
  7%|▋         | 663/9864 [2:46:40<37:33:22, 14.69s/it]
  7%|▋         | 664/9864 [2:46:54<36:52:51, 14.43s/it]
  7%|▋         | 665/9864 [2:47:08<36:15:46, 14.19s/it]
  7%|▋         | 666/9864 [2:47:24<37:46:06, 14.78s/it]
  7%|▋         | 667/9864 [2:47:38<37:14:34, 14.58s/it]
  7%|▋         | 668/9864 [2:47:56<39:37:42, 15.51s/it]
  7%|▋         | 669/9864 [2:48:10<38:34:05, 15.10s/it]
  7%|▋         | 670/9864 [2:48:25<39:00:00, 15.27s/it]
                                                       
{'loss': '0.146', 'grad_norm': '4.36', 'learning_rate': '1.993e-05', 'epoch': '0.06793'}

  7%|▋         | 670/9864 [2:48:25<39:00:00, 15.27s/it]
  7%|▋         | 671/9864 [2:48:39<38:02:18, 14.90s/it]
  7%|▋         | 672/9864 [2:48:53<37:12:29, 14.57s/it]
  7%|▋         | 673/9864 [2:49:09<38:15:50, 14.99s/it]
  7%|▋         | 674/9864 [2:49:23<37:18:47, 14.62s/it]
  7%|▋         | 675/9864 [2:49:38<38:03:51, 14.91s/it]
  7%|▋         | 676/9864 [2:49:54<38:36:48, 15.13s/it]
  7%|▋         | 677/9864 [2:50:08<37:35:05, 14.73s/it]
  7%|▋         | 678/9864 [2:50:24<38:48:53, 15.21s/it]
  7%|▋         | 679/9864 [2:50:38<38:00:30, 14.90s/it]
  7%|▋         | 680/9864 [2:50:52<37:11:00, 14.58s/it]
                                                       
{'loss': '0.1931', 'grad_norm': '3.918', 'learning_rate': '1.992e-05', 'epoch': '0.06894'}

  7%|▋         | 680/9864 [2:50:52<37:11:00, 14.58s/it]
  7%|▋         | 681/9864 [2:51:08<37:44:53, 14.80s/it]
  7%|▋         | 682/9864 [2:51:23<38:37:25, 15.14s/it]
  7%|▋         | 683/9864 [2:51:39<39:01:57, 15.31s/it]
  7%|▋         | 684/9864 [2:51:54<38:25:59, 15.07s/it]
  7%|▋         | 685/9864 [2:52:07<36:59:10, 14.51s/it]
  7%|▋         | 686/9864 [2:52:22<37:47:27, 14.82s/it]
  7%|▋         | 687/9864 [2:52:37<37:16:34, 14.62s/it]
  7%|▋         | 688/9864 [2:52:51<36:56:45, 14.49s/it]
  7%|▋         | 689/9864 [2:53:06<37:39:10, 14.77s/it]
  7%|▋         | 690/9864 [2:53:21<37:17:12, 14.63s/it]
                                                       
{'loss': '0.1809', 'grad_norm': '5.484', 'learning_rate': '1.992e-05', 'epoch': '0.06996'}

  7%|▋         | 690/9864 [2:53:21<37:17:12, 14.63s/it]
  7%|▋         | 691/9864 [2:53:38<39:25:15, 15.47s/it]
  7%|▋         | 692/9864 [2:53:52<38:05:55, 14.95s/it]
  7%|▋         | 693/9864 [2:54:06<37:18:16, 14.64s/it]
  7%|▋         | 694/9864 [2:54:20<36:53:03, 14.48s/it]
  7%|▋         | 695/9864 [2:54:35<37:39:47, 14.79s/it]
  7%|▋         | 696/9864 [2:54:49<37:13:30, 14.62s/it]
  7%|▋         | 697/9864 [2:55:05<37:59:15, 14.92s/it]
  7%|▋         | 698/9864 [2:55:19<37:08:06, 14.59s/it]
  7%|▋         | 699/9864 [2:55:36<38:50:57, 15.26s/it]
  7%|▋         | 700/9864 [2:55:50<38:09:51, 14.99s/it]
                                                       
{'loss': '0.1838', 'grad_norm': '3.734', 'learning_rate': '1.991e-05', 'epoch': '0.07097'}

  7%|▋         | 700/9864 [2:55:50<38:09:51, 14.99s/it]
  7%|▋         | 701/9864 [2:56:04<37:25:38, 14.70s/it]
  7%|▋         | 702/9864 [2:56:20<38:29:20, 15.12s/it]
  7%|▋         | 703/9864 [2:56:34<37:35:28, 14.77s/it]
  7%|▋         | 704/9864 [2:56:48<37:07:57, 14.59s/it]
  7%|▋         | 705/9864 [2:57:04<37:40:03, 14.81s/it]
  7%|▋         | 706/9864 [2:57:17<36:52:22, 14.49s/it]
  7%|▋         | 707/9864 [2:57:34<38:14:17, 15.03s/it]
  7%|▋         | 708/9864 [2:57:47<37:09:18, 14.61s/it]
  7%|▋         | 709/9864 [2:58:01<36:39:19, 14.41s/it]
  7%|▋         | 710/9864 [2:58:17<37:35:19, 14.78s/it]
                                                       
{'loss': '0.1834', 'grad_norm': '5.262', 'learning_rate': '1.991e-05', 'epoch': '0.07198'}

  7%|▋         | 710/9864 [2:58:17<37:35:19, 14.78s/it]
  7%|▋         | 711/9864 [2:58:31<37:11:43, 14.63s/it]
  7%|▋         | 712/9864 [2:58:45<36:25:35, 14.33s/it]
  7%|▋         | 713/9864 [2:59:01<37:43:29, 14.84s/it]
  7%|▋         | 714/9864 [2:59:15<37:14:17, 14.65s/it]
  7%|▋         | 715/9864 [2:59:33<39:30:44, 15.55s/it]
  7%|▋         | 716/9864 [2:59:47<38:24:31, 15.11s/it]
  7%|▋         | 717/9864 [3:00:01<37:23:18, 14.72s/it]
  7%|▋         | 718/9864 [3:00:15<36:56:40, 14.54s/it]
  7%|▋         | 719/9864 [3:00:31<37:55:33, 14.93s/it]
  7%|▋         | 720/9864 [3:00:45<37:27:31, 14.75s/it]
                                                       
{'loss': '0.1723', 'grad_norm': '5.594', 'learning_rate': '1.99e-05', 'epoch': '0.073'}

  7%|▋         | 720/9864 [3:00:45<37:27:31, 14.75s/it]
  7%|▋         | 721/9864 [3:01:00<38:06:30, 15.00s/it]
  7%|▋         | 722/9864 [3:01:18<40:16:27, 15.86s/it]
  7%|▋         | 723/9864 [3:01:33<39:09:57, 15.42s/it]
  7%|▋         | 724/9864 [3:01:47<38:05:23, 15.00s/it]
  7%|▋         | 725/9864 [3:02:01<37:21:27, 14.72s/it]
  7%|▋         | 726/9864 [3:02:17<38:21:40, 15.11s/it]
  7%|▋         | 727/9864 [3:02:31<37:24:01, 14.74s/it]
  7%|▋         | 728/9864 [3:02:45<36:49:13, 14.51s/it]
  7%|▋         | 729/9864 [3:02:58<36:03:26, 14.21s/it]
  7%|▋         | 730/9864 [3:03:16<38:25:30, 15.14s/it]
                                                       
{'loss': '0.1669', 'grad_norm': '7.496', 'learning_rate': '1.99e-05', 'epoch': '0.07401'}

  7%|▋         | 730/9864 [3:03:16<38:25:30, 15.14s/it]
  7%|▋         | 731/9864 [3:03:31<38:43:52, 15.27s/it]
  7%|▋         | 732/9864 [3:03:45<37:45:33, 14.89s/it]
  7%|▋         | 733/9864 [3:04:00<38:01:13, 14.99s/it]
  7%|▋         | 734/9864 [3:04:14<36:51:20, 14.53s/it]
  7%|▋         | 735/9864 [3:04:28<36:43:43, 14.48s/it]
  7%|▋         | 736/9864 [3:04:42<36:19:13, 14.32s/it]
  7%|▋         | 737/9864 [3:04:58<37:12:45, 14.68s/it]
  7%|▋         | 738/9864 [3:05:12<36:59:26, 14.59s/it]
  7%|▋         | 739/9864 [3:05:30<39:24:10, 15.55s/it]
  8%|▊         | 740/9864 [3:05:44<38:13:38, 15.08s/it]
                                                       
{'loss': '0.1978', 'grad_norm': '9.116', 'learning_rate': '1.989e-05', 'epoch': '0.07503'}

  8%|▊         | 740/9864 [3:05:44<38:13:38, 15.08s/it]
  8%|▊         | 741/9864 [3:06:00<38:47:23, 15.31s/it]
  8%|▊         | 742/9864 [3:06:13<37:37:27, 14.85s/it]
  8%|▊         | 743/9864 [3:06:28<37:05:33, 14.64s/it]
  8%|▊         | 744/9864 [3:06:42<36:38:15, 14.46s/it]
  8%|▊         | 745/9864 [3:06:57<37:31:30, 14.81s/it]
  8%|▊         | 746/9864 [3:07:13<38:26:32, 15.18s/it]
  8%|▊         | 747/9864 [3:07:27<37:25:35, 14.78s/it]
  8%|▊         | 748/9864 [3:07:45<39:43:46, 15.69s/it]
  8%|▊         | 749/9864 [3:07:59<38:29:15, 15.20s/it]
  8%|▊         | 750/9864 [3:08:13<37:37:30, 14.86s/it]
                                                       
{'loss': '0.1746', 'grad_norm': '3.155', 'learning_rate': '1.989e-05', 'epoch': '0.07604'}

  8%|▊         | 750/9864 [3:08:13<37:37:30, 14.86s/it]
  8%|▊         | 751/9864 [3:08:27<36:57:33, 14.60s/it]
  8%|▊         | 752/9864 [3:08:41<36:47:04, 14.53s/it]
  8%|▊         | 753/9864 [3:08:57<37:26:40, 14.80s/it]
  8%|▊         | 754/9864 [3:09:12<37:45:15, 14.92s/it]
  8%|▊         | 755/9864 [3:09:27<38:08:04, 15.07s/it]
  8%|▊         | 756/9864 [3:09:43<38:27:33, 15.20s/it]
  8%|▊         | 757/9864 [3:09:57<37:37:18, 14.87s/it]
  8%|▊         | 758/9864 [3:10:11<37:09:56, 14.69s/it]
  8%|▊         | 759/9864 [3:10:26<36:54:00, 14.59s/it]
  8%|▊         | 760/9864 [3:10:39<36:17:34, 14.35s/it]
                                                       
{'loss': '0.1784', 'grad_norm': '3.177', 'learning_rate': '1.988e-05', 'epoch': '0.07705'}

  8%|▊         | 760/9864 [3:10:39<36:17:34, 14.35s/it]
  8%|▊         | 761/9864 [3:10:57<38:25:28, 15.20s/it]
  8%|▊         | 762/9864 [3:11:12<38:51:22, 15.37s/it]
  8%|▊         | 763/9864 [3:11:26<37:48:52, 14.96s/it]
  8%|▊         | 764/9864 [3:11:42<38:35:47, 15.27s/it]
  8%|▊         | 765/9864 [3:11:57<37:45:43, 14.94s/it]
  8%|▊         | 766/9864 [3:12:11<37:18:44, 14.76s/it]
  8%|▊         | 767/9864 [3:12:25<36:47:29, 14.56s/it]
  8%|▊         | 768/9864 [3:12:39<36:26:51, 14.43s/it]
  8%|▊         | 769/9864 [3:12:57<38:42:07, 15.32s/it]
  8%|▊         | 770/9864 [3:13:12<38:54:01, 15.40s/it]
                                                       
{'loss': '0.1761', 'grad_norm': '4.683', 'learning_rate': '1.988e-05', 'epoch': '0.07807'}

  8%|▊         | 770/9864 [3:13:12<38:54:01, 15.40s/it]
  8%|▊         | 771/9864 [3:13:26<37:44:49, 14.94s/it]
  8%|▊         | 772/9864 [3:13:41<37:54:59, 15.01s/it]
  8%|▊         | 773/9864 [3:13:55<37:16:01, 14.76s/it]
  8%|▊         | 774/9864 [3:14:10<37:02:33, 14.67s/it]
  8%|▊         | 775/9864 [3:14:24<36:54:26, 14.62s/it]
  8%|▊         | 776/9864 [3:14:38<36:28:01, 14.45s/it]
  8%|▊         | 777/9864 [3:14:56<38:39:51, 15.32s/it]
  8%|▊         | 778/9864 [3:15:12<39:18:13, 15.57s/it]
  8%|▊         | 779/9864 [3:15:26<38:12:58, 15.14s/it]
  8%|▊         | 780/9864 [3:15:42<38:33:32, 15.28s/it]
                                                       
{'loss': '0.1781', 'grad_norm': '4.257', 'learning_rate': '1.987e-05', 'epoch': '0.07908'}

  8%|▊         | 780/9864 [3:15:42<38:33:32, 15.28s/it]
  8%|▊         | 781/9864 [3:15:56<37:46:26, 14.97s/it]
  8%|▊         | 782/9864 [3:16:10<37:24:41, 14.83s/it]
  8%|▊         | 783/9864 [3:16:24<36:44:49, 14.57s/it]
  8%|▊         | 784/9864 [3:16:39<36:31:17, 14.48s/it]
  8%|▊         | 785/9864 [3:16:56<38:38:32, 15.32s/it]
  8%|▊         | 786/9864 [3:17:11<38:21:04, 15.21s/it]
  8%|▊         | 787/9864 [3:17:26<38:34:15, 15.30s/it]
  8%|▊         | 788/9864 [3:17:40<37:34:48, 14.91s/it]
  8%|▊         | 789/9864 [3:17:54<36:38:22, 14.53s/it]
  8%|▊         | 790/9864 [3:18:08<36:09:57, 14.35s/it]
                                                       
{'loss': '0.1755', 'grad_norm': '2.143', 'learning_rate': '1.987e-05', 'epoch': '0.0801'}

  8%|▊         | 790/9864 [3:18:08<36:09:57, 14.35s/it]
  8%|▊         | 791/9864 [3:18:22<36:05:59, 14.32s/it]
  8%|▊         | 792/9864 [3:18:40<38:32:58, 15.30s/it]
  8%|▊         | 793/9864 [3:18:54<37:24:58, 14.85s/it]
  8%|▊         | 794/9864 [3:19:11<39:05:08, 15.51s/it]
  8%|▊         | 795/9864 [3:19:24<37:47:07, 15.00s/it]
  8%|▊         | 796/9864 [3:19:38<36:25:23, 14.46s/it]
  8%|▊         | 797/9864 [3:19:51<35:47:48, 14.21s/it]
  8%|▊         | 798/9864 [3:20:04<35:02:46, 13.92s/it]
  8%|▊         | 799/9864 [3:20:20<36:07:59, 14.35s/it]
  8%|▊         | 800/9864 [3:20:36<37:24:25, 14.86s/it]
                                                       
{'loss': '0.186', 'grad_norm': '5.081', 'learning_rate': '1.986e-05', 'epoch': '0.08111'}

  8%|▊         | 800/9864 [3:20:36<37:24:25, 14.86s/it]
  8%|▊         | 801/9864 [3:20:52<38:05:00, 15.13s/it]
  8%|▊         | 802/9864 [3:21:05<37:00:28, 14.70s/it]
  8%|▊         | 803/9864 [3:21:21<37:34:01, 14.93s/it]
  8%|▊         | 804/9864 [3:21:35<37:00:27, 14.71s/it]
  8%|▊         | 805/9864 [3:21:49<36:37:40, 14.56s/it]
  8%|▊         | 806/9864 [3:22:03<36:26:11, 14.48s/it]
  8%|▊         | 807/9864 [3:22:19<37:26:32, 14.88s/it]
  8%|▊         | 808/9864 [3:22:35<38:14:39, 15.20s/it]
  8%|▊         | 809/9864 [3:22:50<37:58:55, 15.10s/it]
  8%|▊         | 810/9864 [3:23:04<36:59:26, 14.71s/it]
                                                       
{'loss': '0.1755', 'grad_norm': '3.466', 'learning_rate': '1.986e-05', 'epoch': '0.08212'}

  8%|▊         | 810/9864 [3:23:04<36:59:26, 14.71s/it]
  8%|▊         | 811/9864 [3:23:20<37:56:38, 15.09s/it]
  8%|▊         | 812/9864 [3:23:34<37:27:11, 14.90s/it]
  8%|▊         | 813/9864 [3:23:49<36:59:45, 14.72s/it]
  8%|▊         | 814/9864 [3:24:03<36:29:03, 14.51s/it]
  8%|▊         | 815/9864 [3:24:20<38:57:30, 15.50s/it]
  8%|▊         | 816/9864 [3:24:36<39:00:25, 15.52s/it]
  8%|▊         | 817/9864 [3:24:50<37:43:09, 15.01s/it]
  8%|▊         | 818/9864 [3:25:05<38:01:27, 15.13s/it]
  8%|▊         | 819/9864 [3:25:20<37:24:08, 14.89s/it]
  8%|▊         | 820/9864 [3:25:34<36:55:09, 14.70s/it]
                                                       
{'loss': '0.1794', 'grad_norm': '2.468', 'learning_rate': '1.985e-05', 'epoch': '0.08314'}

  8%|▊         | 820/9864 [3:25:34<36:55:09, 14.70s/it]
  8%|▊         | 821/9864 [3:25:48<36:25:42, 14.50s/it]
  8%|▊         | 822/9864 [3:26:03<37:14:29, 14.83s/it]
  8%|▊         | 823/9864 [3:26:20<38:29:19, 15.33s/it]
  8%|▊         | 824/9864 [3:26:34<37:32:51, 14.95s/it]
  8%|▊         | 825/9864 [3:26:50<37:58:19, 15.12s/it]
  8%|▊         | 826/9864 [3:27:06<38:35:30, 15.37s/it]
  8%|▊         | 827/9864 [3:27:20<37:51:05, 15.08s/it]
  8%|▊         | 828/9864 [3:27:34<36:51:17, 14.68s/it]
  8%|▊         | 829/9864 [3:27:49<37:26:29, 14.92s/it]
  8%|▊         | 830/9864 [3:28:04<37:02:00, 14.76s/it]
                                                       
{'loss': '0.1679', 'grad_norm': '2.249', 'learning_rate': '1.985e-05', 'epoch': '0.08415'}

  8%|▊         | 830/9864 [3:28:04<37:02:00, 14.76s/it]
  8%|▊         | 831/9864 [3:28:20<37:57:14, 15.13s/it]
  8%|▊         | 832/9864 [3:28:34<37:20:56, 14.89s/it]
  8%|▊         | 833/9864 [3:28:50<38:12:55, 15.23s/it]
  8%|▊         | 834/9864 [3:29:04<37:16:08, 14.86s/it]
  8%|▊         | 835/9864 [3:29:20<38:02:57, 15.17s/it]
  8%|▊         | 836/9864 [3:29:33<36:49:26, 14.68s/it]
  8%|▊         | 837/9864 [3:29:49<37:51:15, 15.10s/it]
  8%|▊         | 838/9864 [3:30:04<37:08:42, 14.82s/it]
  9%|▊         | 839/9864 [3:30:19<37:58:59, 15.15s/it]
  9%|▊         | 840/9864 [3:30:35<38:18:14, 15.28s/it]
                                                       
{'loss': '0.1787', 'grad_norm': '5.11', 'learning_rate': '1.984e-05', 'epoch': '0.08517'}

  9%|▊         | 840/9864 [3:30:35<38:18:14, 15.28s/it]
  9%|▊         | 841/9864 [3:30:49<37:16:35, 14.87s/it]
  9%|▊         | 842/9864 [3:31:04<37:27:20, 14.95s/it]
  9%|▊         | 843/9864 [3:31:19<37:10:56, 14.84s/it]
  9%|▊         | 844/9864 [3:31:34<37:35:48, 15.01s/it]
  9%|▊         | 845/9864 [3:31:48<36:55:45, 14.74s/it]
  9%|▊         | 846/9864 [3:32:02<36:25:24, 14.54s/it]
  9%|▊         | 847/9864 [3:32:17<36:49:28, 14.70s/it]
  9%|▊         | 848/9864 [3:32:33<37:29:06, 14.97s/it]
  9%|▊         | 849/9864 [3:32:47<36:30:26, 14.58s/it]
  9%|▊         | 850/9864 [3:33:02<37:17:02, 14.89s/it]
                                                       
{'loss': '0.1731', 'grad_norm': '3.999', 'learning_rate': '1.984e-05', 'epoch': '0.08618'}

  9%|▊         | 850/9864 [3:33:02<37:17:02, 14.89s/it]
  9%|▊         | 851/9864 [3:33:16<36:12:38, 14.46s/it]
  9%|▊         | 852/9864 [3:33:32<37:15:04, 14.88s/it]
  9%|▊         | 853/9864 [3:33:46<37:03:06, 14.80s/it]
  9%|▊         | 854/9864 [3:34:02<37:40:05, 15.05s/it]
  9%|▊         | 855/9864 [3:34:16<36:45:46, 14.69s/it]
  9%|▊         | 856/9864 [3:34:31<37:29:31, 14.98s/it]
  9%|▊         | 857/9864 [3:34:46<36:59:58, 14.79s/it]
  9%|▊         | 858/9864 [3:35:01<37:14:51, 14.89s/it]
  9%|▊         | 859/9864 [3:35:15<36:34:33, 14.62s/it]
  9%|▊         | 860/9864 [3:35:30<36:42:06, 14.67s/it]
                                                       
{'loss': '0.1718', 'grad_norm': '4.167', 'learning_rate': '1.983e-05', 'epoch': '0.08719'}

  9%|▊         | 860/9864 [3:35:30<36:42:06, 14.67s/it]
  9%|▊         | 861/9864 [3:35:45<37:14:47, 14.89s/it]
  9%|▊         | 862/9864 [3:36:01<38:02:36, 15.21s/it]
  9%|▊         | 863/9864 [3:36:17<38:27:44, 15.38s/it]
  9%|▉         | 864/9864 [3:36:31<37:33:34, 15.02s/it]
  9%|▉         | 865/9864 [3:36:45<36:54:21, 14.76s/it]
  9%|▉         | 866/9864 [3:37:01<37:41:03, 15.08s/it]
  9%|▉         | 867/9864 [3:37:15<36:37:44, 14.66s/it]
  9%|▉         | 868/9864 [3:37:29<36:19:10, 14.53s/it]
  9%|▉         | 869/9864 [3:37:46<38:03:27, 15.23s/it]
  9%|▉         | 870/9864 [3:37:59<36:52:48, 14.76s/it]
                                                       
{'loss': '0.1653', 'grad_norm': '3.849', 'learning_rate': '1.982e-05', 'epoch': '0.08821'}

  9%|▉         | 870/9864 [3:37:59<36:52:48, 14.76s/it]
  9%|▉         | 871/9864 [3:38:15<37:43:46, 15.10s/it]
  9%|▉         | 872/9864 [3:38:29<37:02:23, 14.83s/it]
  9%|▉         | 873/9864 [3:38:45<37:35:34, 15.05s/it]
  9%|▉         | 874/9864 [3:38:59<37:03:42, 14.84s/it]
  9%|▉         | 875/9864 [3:39:14<36:41:51, 14.70s/it]
  9%|▉         | 876/9864 [3:39:28<36:37:17, 14.67s/it]
  9%|▉         | 877/9864 [3:39:46<38:49:39, 15.55s/it]
  9%|▉         | 878/9864 [3:40:00<37:44:43, 15.12s/it]
  9%|▉         | 879/9864 [3:40:14<37:03:05, 14.85s/it]
  9%|▉         | 880/9864 [3:40:32<38:53:32, 15.58s/it]
                                                       
{'loss': '0.1564', 'grad_norm': '2.824', 'learning_rate': '1.982e-05', 'epoch': '0.08922'}

  9%|▉         | 880/9864 [3:40:32<38:53:32, 15.58s/it]
  9%|▉         | 881/9864 [3:40:45<37:39:30, 15.09s/it]
  9%|▉         | 882/9864 [3:41:00<37:20:44, 14.97s/it]
  9%|▉         | 883/9864 [3:41:14<36:33:34, 14.65s/it]
  9%|▉         | 884/9864 [3:41:28<36:19:45, 14.56s/it]
  9%|▉         | 885/9864 [3:41:44<37:20:05, 14.97s/it]
  9%|▉         | 886/9864 [3:42:00<37:54:40, 15.20s/it]
  9%|▉         | 887/9864 [3:42:15<37:51:02, 15.18s/it]
  9%|▉         | 888/9864 [3:42:31<37:59:09, 15.23s/it]
  9%|▉         | 889/9864 [3:42:45<37:05:09, 14.88s/it]
  9%|▉         | 890/9864 [3:42:59<36:37:56, 14.70s/it]
                                                       
{'loss': '0.1732', 'grad_norm': '5.243', 'learning_rate': '1.981e-05', 'epoch': '0.09023'}

  9%|▉         | 890/9864 [3:42:59<36:37:56, 14.70s/it]
  9%|▉         | 891/9864 [3:43:13<36:12:31, 14.53s/it]
  9%|▉         | 892/9864 [3:43:27<35:59:35, 14.44s/it]
  9%|▉         | 893/9864 [3:43:42<36:24:53, 14.61s/it]
  9%|▉         | 894/9864 [3:43:59<38:15:24, 15.35s/it]
  9%|▉         | 895/9864 [3:44:14<37:36:14, 15.09s/it]
  9%|▉         | 896/9864 [3:44:30<38:26:44, 15.43s/it]
  9%|▉         | 897/9864 [3:44:44<37:24:57, 15.02s/it]
  9%|▉         | 898/9864 [3:44:58<36:43:08, 14.74s/it]
  9%|▉         | 899/9864 [3:45:13<36:34:06, 14.68s/it]
  9%|▉         | 900/9864 [3:45:26<35:23:23, 14.21s/it]
                                                       
{'loss': '0.1709', 'grad_norm': '5.152', 'learning_rate': '1.98e-05', 'epoch': '0.09125'}

  9%|▉         | 900/9864 [3:45:26<35:23:23, 14.21s/it]
  9%|▉         | 901/9864 [3:45:42<36:39:11, 14.72s/it]
  9%|▉         | 902/9864 [3:45:59<38:30:36, 15.47s/it]
  9%|▉         | 903/9864 [3:46:13<37:32:55, 15.08s/it]
  9%|▉         | 904/9864 [3:46:29<38:05:35, 15.31s/it]
  9%|▉         | 905/9864 [3:46:43<37:11:29, 14.94s/it]
  9%|▉         | 906/9864 [3:46:57<36:30:17, 14.67s/it]
  9%|▉         | 907/9864 [3:47:11<35:58:12, 14.46s/it]
  9%|▉         | 908/9864 [3:47:25<35:54:45, 14.44s/it]
  9%|▉         | 909/9864 [3:47:40<35:37:26, 14.32s/it]
  9%|▉         | 910/9864 [3:47:59<39:06:24, 15.72s/it]
                                                       
{'loss': '0.1776', 'grad_norm': '5.263', 'learning_rate': '1.98e-05', 'epoch': '0.09226'}

  9%|▉         | 910/9864 [3:47:59<39:06:24, 15.72s/it]
  9%|▉         | 911/9864 [3:48:12<37:47:11, 15.19s/it]
  9%|▉         | 912/9864 [3:48:28<37:44:33, 15.18s/it]
  9%|▉         | 913/9864 [3:48:41<36:19:48, 14.61s/it]
  9%|▉         | 914/9864 [3:48:55<35:52:16, 14.43s/it]
  9%|▉         | 915/9864 [3:49:09<35:32:49, 14.30s/it]
  9%|▉         | 916/9864 [3:49:23<35:11:51, 14.16s/it]
  9%|▉         | 917/9864 [3:49:39<36:22:25, 14.64s/it]
  9%|▉         | 918/9864 [3:49:56<38:38:59, 15.55s/it]
  9%|▉         | 919/9864 [3:50:12<38:38:28, 15.55s/it]
  9%|▉         | 920/9864 [3:50:26<37:41:20, 15.17s/it]
                                                       
{'loss': '0.1658', 'grad_norm': '3.443', 'learning_rate': '1.979e-05', 'epoch': '0.09328'}

  9%|▉         | 920/9864 [3:50:26<37:41:20, 15.17s/it]
  9%|▉         | 921/9864 [3:50:40<36:30:04, 14.69s/it]
  9%|▉         | 922/9864 [3:50:54<35:58:53, 14.49s/it]
  9%|▉         | 923/9864 [3:51:08<35:39:29, 14.36s/it]
  9%|▉         | 924/9864 [3:51:22<35:31:12, 14.30s/it]
  9%|▉         | 925/9864 [3:51:38<36:40:23, 14.77s/it]
  9%|▉         | 926/9864 [3:51:55<38:49:50, 15.64s/it]
  9%|▉         | 927/9864 [3:52:11<38:55:19, 15.68s/it]
  9%|▉         | 928/9864 [3:52:25<37:47:25, 15.22s/it]
  9%|▉         | 929/9864 [3:52:39<36:50:47, 14.85s/it]
  9%|▉         | 930/9864 [3:52:53<36:11:02, 14.58s/it]
                                                       
{'loss': '0.1781', 'grad_norm': '3.129', 'learning_rate': '1.978e-05', 'epoch': '0.09429'}

  9%|▉         | 930/9864 [3:52:53<36:11:02, 14.58s/it]
  9%|▉         | 931/9864 [3:53:07<35:21:17, 14.25s/it]
  9%|▉         | 932/9864 [3:53:19<34:07:36, 13.75s/it]
  9%|▉         | 933/9864 [3:53:39<38:13:34, 15.41s/it]
  9%|▉         | 934/9864 [3:53:54<38:30:21, 15.52s/it]
  9%|▉         | 935/9864 [3:54:08<37:23:25, 15.08s/it]
  9%|▉         | 936/9864 [3:54:23<36:45:05, 14.82s/it]
  9%|▉         | 937/9864 [3:54:36<35:38:20, 14.37s/it]
 10%|▉         | 938/9864 [3:54:50<35:18:28, 14.24s/it]
 10%|▉         | 939/9864 [3:55:04<35:19:16, 14.25s/it]
 10%|▉         | 940/9864 [3:55:20<36:22:56, 14.68s/it]
                                                       
{'loss': '0.1732', 'grad_norm': '5.565', 'learning_rate': '1.978e-05', 'epoch': '0.0953'}

 10%|▉         | 940/9864 [3:55:20<36:22:56, 14.68s/it]
 10%|▉         | 941/9864 [3:55:37<38:23:12, 15.49s/it]
 10%|▉         | 942/9864 [3:55:51<37:27:54, 15.12s/it]
 10%|▉         | 943/9864 [3:56:07<37:48:31, 15.26s/it]
 10%|▉         | 944/9864 [3:56:22<37:14:57, 15.03s/it]
 10%|▉         | 945/9864 [3:56:36<36:37:53, 14.79s/it]
 10%|▉         | 946/9864 [3:56:49<35:43:26, 14.42s/it]
 10%|▉         | 947/9864 [3:57:05<36:53:02, 14.89s/it]
 10%|▉         | 948/9864 [3:57:21<37:07:07, 14.99s/it]
 10%|▉         | 949/9864 [3:57:36<37:45:37, 15.25s/it]
 10%|▉         | 950/9864 [3:57:50<36:54:59, 14.91s/it]
                                                       
{'loss': '0.1769', 'grad_norm': '7.805', 'learning_rate': '1.977e-05', 'epoch': '0.09632'}

 10%|▉         | 950/9864 [3:57:50<36:54:59, 14.91s/it]
 10%|▉         | 951/9864 [3:58:06<37:35:39, 15.18s/it]
 10%|▉         | 952/9864 [3:58:21<36:57:42, 14.93s/it]
 10%|▉         | 953/9864 [3:58:35<36:18:41, 14.67s/it]
 10%|▉         | 954/9864 [3:58:50<36:50:18, 14.88s/it]
 10%|▉         | 955/9864 [3:59:05<37:11:05, 15.03s/it]
 10%|▉         | 956/9864 [3:59:19<35:56:30, 14.53s/it]
 10%|▉         | 957/9864 [3:59:34<36:38:22, 14.81s/it]
 10%|▉         | 958/9864 [3:59:50<37:19:34, 15.09s/it]
 10%|▉         | 959/9864 [4:00:04<36:21:59, 14.70s/it]
 10%|▉         | 960/9864 [4:00:18<35:56:30, 14.53s/it]
                                                       
{'loss': '0.1497', 'grad_norm': '4.372', 'learning_rate': '1.976e-05', 'epoch': '0.09733'}

 10%|▉         | 960/9864 [4:00:18<35:56:30, 14.53s/it]
 10%|▉         | 961/9864 [4:00:32<35:20:42, 14.29s/it]
 10%|▉         | 962/9864 [4:00:49<37:20:03, 15.10s/it]
 10%|▉         | 963/9864 [4:01:03<36:23:39, 14.72s/it]
 10%|▉         | 964/9864 [4:01:16<35:17:40, 14.28s/it]
 10%|▉         | 965/9864 [4:01:31<36:18:09, 14.69s/it]
 10%|▉         | 966/9864 [4:01:47<37:15:45, 15.08s/it]
 10%|▉         | 967/9864 [4:02:02<36:36:58, 14.82s/it]
 10%|▉         | 968/9864 [4:02:15<35:55:12, 14.54s/it]
 10%|▉         | 969/9864 [4:02:29<35:16:42, 14.28s/it]
 10%|▉         | 970/9864 [4:02:45<36:23:08, 14.73s/it]
                                                       
{'loss': '0.1807', 'grad_norm': '6.302', 'learning_rate': '1.976e-05', 'epoch': '0.09835'}

 10%|▉         | 970/9864 [4:02:45<36:23:08, 14.73s/it]
 10%|▉         | 971/9864 [4:03:01<37:07:39, 15.03s/it]
 10%|▉         | 972/9864 [4:03:15<36:14:31, 14.67s/it]
 10%|▉         | 973/9864 [4:03:32<38:18:06, 15.51s/it]
 10%|▉         | 974/9864 [4:03:46<37:03:01, 15.00s/it]
 10%|▉         | 975/9864 [4:04:00<36:17:36, 14.70s/it]
 10%|▉         | 976/9864 [4:04:14<35:52:58, 14.53s/it]
 10%|▉         | 977/9864 [4:04:30<37:03:51, 15.01s/it]
 10%|▉         | 978/9864 [4:04:44<36:26:22, 14.76s/it]
 10%|▉         | 979/9864 [4:05:00<37:26:07, 15.17s/it]
 10%|▉         | 980/9864 [4:05:16<37:25:37, 15.17s/it]
                                                       
{'loss': '0.1696', 'grad_norm': '3.544', 'learning_rate': '1.975e-05', 'epoch': '0.09936'}

 10%|▉         | 980/9864 [4:05:16<37:25:37, 15.17s/it]
 10%|▉         | 981/9864 [4:05:31<37:25:02, 15.16s/it]
 10%|▉         | 982/9864 [4:05:45<36:43:16, 14.88s/it]
 10%|▉         | 983/9864 [4:05:59<36:08:36, 14.65s/it]
 10%|▉         | 984/9864 [4:06:13<35:37:01, 14.44s/it]
 10%|▉         | 985/9864 [4:06:29<36:36:55, 14.85s/it]
 10%|▉         | 986/9864 [4:06:43<36:12:05, 14.68s/it]
 10%|█         | 987/9864 [4:06:58<36:45:28, 14.91s/it]
 10%|█         | 988/9864 [4:07:16<38:39:28, 15.68s/it]
 10%|█         | 989/9864 [4:07:30<37:45:22, 15.32s/it]
 10%|█         | 990/9864 [4:07:45<36:54:35, 14.97s/it]
                                                       
{'loss': '0.1703', 'grad_norm': '6.872', 'learning_rate': '1.974e-05', 'epoch': '0.1004'}

 10%|█         | 990/9864 [4:07:45<36:54:35, 14.97s/it]
 10%|█         | 991/9864 [4:07:58<35:56:46, 14.58s/it]
 10%|█         | 992/9864 [4:08:13<35:47:23, 14.52s/it]
 10%|█         | 993/9864 [4:08:28<36:33:57, 14.84s/it]
 10%|█         | 994/9864 [4:08:44<37:16:54, 15.13s/it]
 10%|█         | 995/9864 [4:09:00<37:40:59, 15.30s/it]
 10%|█         | 996/9864 [4:09:16<38:16:34, 15.54s/it]
 10%|█         | 997/9864 [4:09:30<37:02:40, 15.04s/it]
 10%|█         | 998/9864 [4:09:44<36:09:14, 14.68s/it]
 10%|█         | 999/9864 [4:09:57<35:29:57, 14.42s/it]
 10%|█         | 1000/9864 [4:10:12<35:25:05, 14.38s/it]
                                                        
{'loss': '0.185', 'grad_norm': '3.822', 'learning_rate': '1.973e-05', 'epoch': '0.1014'}

 10%|█         | 1000/9864 [4:10:12<35:25:05, 14.38s/it]

Writing model shards:   0%|          | 0/1 [00:00<?, ?it/s]

Writing model shards: 100%|██████████| 1/1 [00:10<00:00, 10.96s/it]
Writing model shards: 100%|██████████| 1/1 [00:10<00:00, 10.97s/it]
/home/dungpq6/Project/SpatialStack/src/qwen_vl/model/geometry_encoders/vggt_encoder.py:285: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(dtype=dtype):

 10%|█         | 1001/9864 [4:11:37<87:46:16, 35.65s/it]
 10%|█         | 1002/9864 [4:11:51<72:07:04, 29.30s/it]
 10%|█         | 1003/9864 [4:12:07<61:52:53, 25.14s/it]
 10%|█         | 1004/9864 [4:12:23<54:58:03, 22.33s/it]
 10%|█         | 1005/9864 [4:12:36<48:27:57, 19.69s/it]
 10%|█         | 1006/9864 [4:12:50<44:19:26, 18.01s/it]
 10%|█         | 1007/9864 [4:13:04<41:10:18, 16.73s/it]
 10%|█         | 1008/9864 [4:13:22<41:47:07, 16.99s/it]
 10%|█         | 1009/9864 [4:13:36<39:55:56, 16.23s/it]
 10%|█         | 1010/9864 [4:13:50<38:18:35, 15.58s/it]
                                                        
{'loss': '0.1662', 'grad_norm': '3.338', 'learning_rate': '1.973e-05', 'epoch': '0.1024'}

 10%|█         | 1010/9864 [4:13:50<38:18:35, 15.58s/it]
 10%|█         | 1011/9864 [4:14:06<38:12:51, 15.54s/it]
 10%|█         | 1012/9864 [4:14:21<38:18:57, 15.58s/it]
 10%|█         | 1013/9864 [4:14:35<36:47:53, 14.97s/it]
 10%|█         | 1014/9864 [4:14:49<36:15:15, 14.75s/it]
 10%|█         | 1015/9864 [4:15:05<37:10:04, 15.12s/it]
 10%|█         | 1016/9864 [4:15:21<37:39:26, 15.32s/it]
 10%|█         | 1017/9864 [4:15:35<37:04:26, 15.09s/it]
 10%|█         | 1018/9864 [4:15:49<36:21:25, 14.80s/it]
 10%|█         | 1019/9864 [4:16:05<36:44:43, 14.96s/it]
 10%|█         | 1020/9864 [4:16:21<37:26:35, 15.24s/it]
                                                        
{'loss': '0.1811', 'grad_norm': '4.984', 'learning_rate': '1.972e-05', 'epoch': '0.1034'}

 10%|█         | 1020/9864 [4:16:21<37:26:35, 15.24s/it]
 10%|█         | 1021/9864 [4:16:35<36:31:27, 14.87s/it]
 10%|█         | 1022/9864 [4:16:49<35:47:52, 14.58s/it]
 10%|█         | 1023/9864 [4:17:03<35:24:32, 14.42s/it]
 10%|█         | 1024/9864 [4:17:20<37:29:58, 15.27s/it]
 10%|█         | 1025/9864 [4:17:34<36:51:35, 15.01s/it]
 10%|█         | 1026/9864 [4:17:48<36:14:46, 14.76s/it]
 10%|█         | 1027/9864 [4:18:04<36:54:45, 15.04s/it]
 10%|█         | 1028/9864 [4:18:18<36:22:39, 14.82s/it]
 10%|█         | 1029/9864 [4:18:34<37:02:08, 15.09s/it]
 10%|█         | 1030/9864 [4:18:49<36:35:38, 14.91s/it]
                                                        
{'loss': '0.1803', 'grad_norm': '8.099', 'learning_rate': '1.971e-05', 'epoch': '0.1044'}

 10%|█         | 1030/9864 [4:18:49<36:35:38, 14.91s/it]
 10%|█         | 1031/9864 [4:19:04<37:06:43, 15.13s/it]
 10%|█         | 1032/9864 [4:19:20<37:24:52, 15.25s/it]
 10%|█         | 1033/9864 [4:19:34<36:33:18, 14.90s/it]
 10%|█         | 1034/9864 [4:19:48<36:08:21, 14.73s/it]
 10%|█         | 1035/9864 [4:20:03<36:04:05, 14.71s/it]
 11%|█         | 1036/9864 [4:20:21<38:18:49, 15.62s/it]
 11%|█         | 1037/9864 [4:20:35<37:25:55, 15.27s/it]
 11%|█         | 1038/9864 [4:20:49<36:45:39, 14.99s/it]
 11%|█         | 1039/9864 [4:21:07<38:22:58, 15.66s/it]
 11%|█         | 1040/9864 [4:21:21<37:24:41, 15.26s/it]
                                                        
{'loss': '0.1535', 'grad_norm': '4.455', 'learning_rate': '1.97e-05', 'epoch': '0.1054'}

 11%|█         | 1040/9864 [4:21:21<37:24:41, 15.26s/it]
 11%|█         | 1041/9864 [4:21:35<36:27:00, 14.87s/it]
 11%|█         | 1042/9864 [4:21:49<35:59:27, 14.69s/it]
 11%|█         | 1043/9864 [4:22:06<37:46:07, 15.41s/it]
 11%|█         | 1044/9864 [4:22:21<36:58:34, 15.09s/it]
 11%|█         | 1045/9864 [4:22:35<36:04:13, 14.72s/it]
 11%|█         | 1046/9864 [4:22:48<35:25:34, 14.46s/it]
 11%|█         | 1047/9864 [4:23:06<37:25:54, 15.28s/it]
 11%|█         | 1048/9864 [4:23:20<36:44:01, 15.00s/it]
 11%|█         | 1049/9864 [4:23:34<35:50:47, 14.64s/it]
 11%|█         | 1050/9864 [4:23:50<36:51:58, 15.06s/it]
                                                        
{'loss': '0.1627', 'grad_norm': '2.393', 'learning_rate': '1.97e-05', 'epoch': '0.1065'}

 11%|█         | 1050/9864 [4:23:50<36:51:58, 15.06s/it]
 11%|█         | 1051/9864 [4:24:04<36:13:42, 14.80s/it]
 11%|█         | 1052/9864 [4:24:20<36:45:36, 15.02s/it]
 11%|█         | 1053/9864 [4:24:33<35:31:57, 14.52s/it]
 11%|█         | 1054/9864 [4:24:49<36:28:55, 14.91s/it]
 11%|█         | 1055/9864 [4:25:04<36:47:39, 15.04s/it]
 11%|█         | 1056/9864 [4:25:18<36:05:11, 14.75s/it]
 11%|█         | 1057/9864 [4:25:34<36:52:24, 15.07s/it]
 11%|█         | 1058/9864 [4:25:48<36:13:40, 14.81s/it]
 11%|█         | 1059/9864 [4:26:02<35:23:53, 14.47s/it]
 11%|█         | 1060/9864 [4:26:18<36:39:44, 14.99s/it]
                                                        
{'loss': '0.1516', 'grad_norm': '2.939', 'learning_rate': '1.969e-05', 'epoch': '0.1075'}

 11%|█         | 1060/9864 [4:26:18<36:39:44, 14.99s/it]
 11%|█         | 1061/9864 [4:26:32<35:56:07, 14.70s/it]
 11%|█         | 1062/9864 [4:26:48<36:34:20, 14.96s/it]
 11%|█         | 1063/9864 [4:27:03<37:14:40, 15.23s/it]
 11%|█         | 1064/9864 [4:27:18<36:25:01, 14.90s/it]
 11%|█         | 1065/9864 [4:27:33<36:41:31, 15.01s/it]
 11%|█         | 1066/9864 [4:27:47<35:49:51, 14.66s/it]
 11%|█         | 1067/9864 [4:28:02<36:35:25, 14.97s/it]
 11%|█         | 1068/9864 [4:28:16<35:37:48, 14.58s/it]
 11%|█         | 1069/9864 [4:28:30<35:19:46, 14.46s/it]
 11%|█         | 1070/9864 [4:28:47<37:14:40, 15.25s/it]
                                                        
{'loss': '0.1865', 'grad_norm': '3.59', 'learning_rate': '1.968e-05', 'epoch': '0.1085'}

 11%|█         | 1070/9864 [4:28:47<37:14:40, 15.25s/it]
 11%|█         | 1071/9864 [4:29:02<36:37:09, 14.99s/it]
 11%|█         | 1072/9864 [4:29:16<35:47:49, 14.66s/it]
 11%|█         | 1073/9864 [4:29:31<36:33:51, 14.97s/it]
 11%|█         | 1074/9864 [4:29:46<36:10:24, 14.82s/it]
 11%|█         | 1075/9864 [4:30:01<36:44:08, 15.05s/it]
 11%|█         | 1076/9864 [4:30:15<35:54:22, 14.71s/it]
 11%|█         | 1077/9864 [4:30:29<35:29:21, 14.54s/it]
 11%|█         | 1078/9864 [4:30:45<36:28:29, 14.95s/it]
 11%|█         | 1079/9864 [4:31:01<37:07:09, 15.21s/it]
 11%|█         | 1080/9864 [4:31:17<37:35:51, 15.41s/it]
                                                        
{'loss': '0.1708', 'grad_norm': '5.389', 'learning_rate': '1.967e-05', 'epoch': '0.1095'}

 11%|█         | 1080/9864 [4:31:17<37:35:51, 15.41s/it]
 11%|█         | 1081/9864 [4:31:31<36:22:58, 14.91s/it]
 11%|█         | 1082/9864 [4:31:47<37:03:46, 15.19s/it]
 11%|█         | 1083/9864 [4:32:01<36:12:47, 14.85s/it]
 11%|█         | 1084/9864 [4:32:14<35:15:11, 14.45s/it]
 11%|█         | 1085/9864 [4:32:28<35:02:27, 14.37s/it]
 11%|█         | 1086/9864 [4:32:44<36:03:53, 14.79s/it]
 11%|█         | 1087/9864 [4:33:02<37:57:43, 15.57s/it]
 11%|█         | 1088/9864 [4:33:16<37:02:55, 15.20s/it]
 11%|█         | 1089/9864 [4:33:30<36:08:27, 14.83s/it]
 11%|█         | 1090/9864 [4:33:45<36:23:17, 14.93s/it]
                                                        
{'loss': '0.1734', 'grad_norm': '4.639', 'learning_rate': '1.966e-05', 'epoch': '0.1105'}

 11%|█         | 1090/9864 [4:33:45<36:23:17, 14.93s/it]
 11%|█         | 1091/9864 [4:33:59<35:51:32, 14.71s/it]
 11%|█         | 1092/9864 [4:34:13<35:18:06, 14.49s/it]
 11%|█         | 1093/9864 [4:34:27<35:01:26, 14.38s/it]
 11%|█         | 1094/9864 [4:34:44<37:04:35, 15.22s/it]
 11%|█         | 1095/9864 [4:35:00<37:15:35, 15.30s/it]
 11%|█         | 1096/9864 [4:35:14<36:39:13, 15.05s/it]
 11%|█         | 1097/9864 [4:35:28<35:41:40, 14.66s/it]
 11%|█         | 1098/9864 [4:35:44<36:53:46, 15.15s/it]
 11%|█         | 1099/9864 [4:35:59<36:05:59, 14.83s/it]
 11%|█         | 1100/9864 [4:36:12<35:14:02, 14.47s/it]
                                                        
{'loss': '0.1916', 'grad_norm': '2.324', 'learning_rate': '1.965e-05', 'epoch': '0.1115'}

 11%|█         | 1100/9864 [4:36:12<35:14:02, 14.47s/it]
 11%|█         | 1101/9864 [4:36:26<34:43:00, 14.26s/it]
 11%|█         | 1102/9864 [4:36:42<36:12:45, 14.88s/it]
 11%|█         | 1103/9864 [4:36:59<37:50:45, 15.55s/it]
 11%|█         | 1104/9864 [4:37:13<36:48:06, 15.12s/it]
 11%|█         | 1105/9864 [4:37:29<37:08:15, 15.26s/it]
 11%|█         | 1106/9864 [4:37:43<36:05:37, 14.84s/it]
 11%|█         | 1107/9864 [4:37:57<35:41:12, 14.67s/it]
 11%|█         | 1108/9864 [4:38:11<35:13:27, 14.48s/it]
 11%|█         | 1109/9864 [4:38:25<34:46:35, 14.30s/it]
 11%|█▏        | 1110/9864 [4:38:42<37:00:46, 15.22s/it]
                                                        
{'loss': '0.1603', 'grad_norm': '4.086', 'learning_rate': '1.965e-05', 'epoch': '0.1125'}

 11%|█▏        | 1110/9864 [4:38:43<37:00:46, 15.22s/it]
 11%|█▏        | 1111/9864 [4:38:56<35:51:58, 14.75s/it]
 11%|█▏        | 1112/9864 [4:39:12<36:54:22, 15.18s/it]
 11%|█▏        | 1113/9864 [4:39:28<37:21:01, 15.37s/it]
 11%|█▏        | 1114/9864 [4:39:42<36:23:09, 14.97s/it]
 11%|█▏        | 1115/9864 [4:39:56<35:47:26, 14.73s/it]
 11%|█▏        | 1116/9864 [4:40:10<35:17:00, 14.52s/it]
 11%|█▏        | 1117/9864 [4:40:28<37:23:25, 15.39s/it]
 11%|█▏        | 1118/9864 [4:40:42<36:30:28, 15.03s/it]
 11%|█▏        | 1119/9864 [4:40:58<36:55:35, 15.20s/it]
 11%|█▏        | 1120/9864 [4:41:11<35:52:38, 14.77s/it]
                                                        
{'loss': '0.1606', 'grad_norm': '1.908', 'learning_rate': '1.964e-05', 'epoch': '0.1136'}

 11%|█▏        | 1120/9864 [4:41:11<35:52:38, 14.77s/it]
 11%|█▏        | 1121/9864 [4:41:25<35:13:25, 14.50s/it]
 11%|█▏        | 1122/9864 [4:41:41<36:14:10, 14.92s/it]
 11%|█▏        | 1123/9864 [4:41:55<35:43:40, 14.71s/it]
 11%|█▏        | 1124/9864 [4:42:10<35:34:46, 14.66s/it]
 11%|█▏        | 1125/9864 [4:42:26<36:35:06, 15.07s/it]
 11%|█▏        | 1126/9864 [4:42:39<35:28:20, 14.61s/it]
 11%|█▏        | 1127/9864 [4:42:55<36:12:33, 14.92s/it]
 11%|█▏        | 1128/9864 [4:43:09<35:27:32, 14.61s/it]
 11%|█▏        | 1129/9864 [4:43:23<35:05:39, 14.46s/it]
 11%|█▏        | 1130/9864 [4:43:39<36:09:03, 14.90s/it]
                                                        
{'loss': '0.1791', 'grad_norm': '3.333', 'learning_rate': '1.963e-05', 'epoch': '0.1146'}

 11%|█▏        | 1130/9864 [4:43:39<36:09:03, 14.90s/it]
 11%|█▏        | 1131/9864 [4:43:53<35:39:36, 14.70s/it]
 11%|█▏        | 1132/9864 [4:44:07<35:04:05, 14.46s/it]
 11%|█▏        | 1133/9864 [4:44:22<35:38:00, 14.69s/it]
 11%|█▏        | 1134/9864 [4:44:40<37:42:03, 15.55s/it]
 12%|█▏        | 1135/9864 [4:44:53<36:12:57, 14.94s/it]
 12%|█▏        | 1136/9864 [4:45:08<35:53:00, 14.80s/it]
 12%|█▏        | 1137/9864 [4:45:22<35:24:52, 14.61s/it]
 12%|█▏        | 1138/9864 [4:45:38<36:10:04, 14.92s/it]
 12%|█▏        | 1139/9864 [4:45:52<35:55:09, 14.82s/it]
 12%|█▏        | 1140/9864 [4:46:06<35:08:25, 14.50s/it]
                                                        
{'loss': '0.173', 'grad_norm': '5.628', 'learning_rate': '1.962e-05', 'epoch': '0.1156'}

 12%|█▏        | 1140/9864 [4:46:06<35:08:25, 14.50s/it]
 12%|█▏        | 1141/9864 [4:46:25<38:15:50, 15.79s/it]
 12%|█▏        | 1142/9864 [4:46:39<36:49:11, 15.20s/it]
 12%|█▏        | 1143/9864 [4:46:52<35:42:30, 14.74s/it]
 12%|█▏        | 1144/9864 [4:47:06<35:07:32, 14.50s/it]
 12%|█▏        | 1145/9864 [4:47:20<34:39:26, 14.31s/it]
 12%|█▏        | 1146/9864 [4:47:37<36:07:04, 14.91s/it]
 12%|█▏        | 1147/9864 [4:47:51<35:29:26, 14.66s/it]
 12%|█▏        | 1148/9864 [4:48:06<36:22:37, 15.02s/it]
 12%|█▏        | 1149/9864 [4:48:24<38:10:20, 15.77s/it]
 12%|█▏        | 1150/9864 [4:48:38<37:03:28, 15.31s/it]
                                                        
{'loss': '0.1602', 'grad_norm': '3.31', 'learning_rate': '1.961e-05', 'epoch': '0.1166'}

 12%|█▏        | 1150/9864 [4:48:38<37:03:28, 15.31s/it]
 12%|█▏        | 1151/9864 [4:48:52<36:12:41, 14.96s/it]
 12%|█▏        | 1152/9864 [4:49:07<35:45:37, 14.78s/it]
 12%|█▏        | 1153/9864 [4:49:21<35:07:54, 14.52s/it]
 12%|█▏        | 1154/9864 [4:49:37<36:09:09, 14.94s/it]
 12%|█▏        | 1155/9864 [4:49:52<36:37:05, 15.14s/it]
 12%|█▏        | 1156/9864 [4:50:06<35:29:35, 14.67s/it]
 12%|█▏        | 1157/9864 [4:50:23<37:33:59, 15.53s/it]
 12%|█▏        | 1158/9864 [4:50:37<36:12:49, 14.97s/it]
 12%|█▏        | 1159/9864 [4:50:51<35:28:39, 14.67s/it]
 12%|█▏        | 1160/9864 [4:51:05<34:46:53, 14.39s/it]
                                                        
{'loss': '0.1559', 'grad_norm': '2.382', 'learning_rate': '1.96e-05', 'epoch': '0.1176'}

 12%|█▏        | 1160/9864 [4:51:05<34:46:53, 14.39s/it]
 12%|█▏        | 1161/9864 [4:51:19<34:40:47, 14.35s/it]
 12%|█▏        | 1162/9864 [4:51:35<35:42:19, 14.77s/it]
 12%|█▏        | 1163/9864 [4:51:49<35:17:33, 14.60s/it]
 12%|█▏        | 1164/9864 [4:52:05<36:18:26, 15.02s/it]
 12%|█▏        | 1165/9864 [4:52:21<37:08:56, 15.37s/it]
 12%|█▏        | 1166/9864 [4:52:35<36:20:07, 15.04s/it]
 12%|█▏        | 1167/9864 [4:52:49<35:30:09, 14.70s/it]
 12%|█▏        | 1168/9864 [4:53:03<34:59:10, 14.48s/it]
 12%|█▏        | 1169/9864 [4:53:19<36:12:09, 14.99s/it]
 12%|█▏        | 1170/9864 [4:53:33<35:23:19, 14.65s/it]
                                                        
{'loss': '0.1916', 'grad_norm': '6.424', 'learning_rate': '1.959e-05', 'epoch': '0.1186'}

 12%|█▏        | 1170/9864 [4:53:33<35:23:19, 14.65s/it]
 12%|█▏        | 1171/9864 [4:53:47<34:54:46, 14.46s/it]
 12%|█▏        | 1172/9864 [4:54:04<36:46:30, 15.23s/it]
 12%|█▏        | 1173/9864 [4:54:18<35:43:10, 14.80s/it]
 12%|█▏        | 1174/9864 [4:54:34<36:52:13, 15.27s/it]
 12%|█▏        | 1175/9864 [4:54:49<36:13:16, 15.01s/it]
 12%|█▏        | 1176/9864 [4:55:03<35:40:54, 14.79s/it]
 12%|█▏        | 1177/9864 [4:55:19<36:09:18, 14.98s/it]
 12%|█▏        | 1178/9864 [4:55:33<35:38:58, 14.78s/it]
 12%|█▏        | 1179/9864 [4:55:47<35:02:57, 14.53s/it]
 12%|█▏        | 1180/9864 [4:56:04<37:05:43, 15.38s/it]
                                                        
{'loss': '0.164', 'grad_norm': '2.443', 'learning_rate': '1.958e-05', 'epoch': '0.1196'}

 12%|█▏        | 1180/9864 [4:56:04<37:05:43, 15.38s/it]
 12%|█▏        | 1181/9864 [4:56:18<36:10:25, 15.00s/it]
 12%|█▏        | 1182/9864 [4:56:34<36:47:47, 15.26s/it]
 12%|█▏        | 1183/9864 [4:56:48<35:56:37, 14.91s/it]
 12%|█▏        | 1184/9864 [4:57:03<36:14:23, 15.03s/it]
 12%|█▏        | 1185/9864 [4:57:18<35:38:58, 14.79s/it]
 12%|█▏        | 1186/9864 [4:57:32<35:14:53, 14.62s/it]
 12%|█▏        | 1187/9864 [4:57:48<36:09:12, 15.00s/it]
 12%|█▏        | 1188/9864 [4:58:04<36:49:18, 15.28s/it]
 12%|█▏        | 1189/9864 [4:58:18<35:59:28, 14.94s/it]
 12%|█▏        | 1190/9864 [4:58:32<35:22:49, 14.68s/it]
                                                        
{'loss': '0.1613', 'grad_norm': '1.744', 'learning_rate': '1.957e-05', 'epoch': '0.1207'}

 12%|█▏        | 1190/9864 [4:58:32<35:22:49, 14.68s/it]
 12%|█▏        | 1191/9864 [4:58:48<36:11:53, 15.03s/it]
 12%|█▏        | 1192/9864 [4:59:03<36:27:36, 15.14s/it]
 12%|█▏        | 1193/9864 [4:59:17<35:44:37, 14.84s/it]
 12%|█▏        | 1194/9864 [4:59:31<35:03:39, 14.56s/it]
 12%|█▏        | 1195/9864 [4:59:49<37:06:52, 15.41s/it]
 12%|█▏        | 1196/9864 [5:00:03<36:07:57, 15.01s/it]
 12%|█▏        | 1197/9864 [5:00:16<35:06:49, 14.59s/it]
 12%|█▏        | 1198/9864 [5:00:30<34:45:47, 14.44s/it]
 12%|█▏        | 1199/9864 [5:00:46<35:42:52, 14.84s/it]
 12%|█▏        | 1200/9864 [5:01:02<36:25:25, 15.13s/it]
                                                        
{'loss': '0.1808', 'grad_norm': '1.732', 'learning_rate': '1.956e-05', 'epoch': '0.1217'}

 12%|█▏        | 1200/9864 [5:01:02<36:25:25, 15.13s/it]
 12%|█▏        | 1201/9864 [5:01:16<35:36:39, 14.80s/it]
 12%|█▏        | 1202/9864 [5:01:32<36:22:49, 15.12s/it]
 12%|█▏        | 1203/9864 [5:01:48<36:48:10, 15.30s/it]
 12%|█▏        | 1204/9864 [5:02:02<35:48:42, 14.89s/it]
 12%|█▏        | 1205/9864 [5:02:16<35:28:41, 14.75s/it]
 12%|█▏        | 1206/9864 [5:02:32<36:09:07, 15.03s/it]
 12%|█▏        | 1207/9864 [5:02:47<36:32:37, 15.20s/it]
 12%|█▏        | 1208/9864 [5:03:01<35:47:27, 14.89s/it]
 12%|█▏        | 1209/9864 [5:03:15<35:00:00, 14.56s/it]
 12%|█▏        | 1210/9864 [5:03:30<35:30:19, 14.77s/it]
                                                        
{'loss': '0.1593', 'grad_norm': '1.566', 'learning_rate': '1.955e-05', 'epoch': '0.1227'}

 12%|█▏        | 1210/9864 [5:03:30<35:30:19, 14.77s/it]
 12%|█▏        | 1211/9864 [5:03:46<36:05:39, 15.02s/it]
 12%|█▏        | 1212/9864 [5:04:00<35:28:02, 14.76s/it]
 12%|█▏        | 1213/9864 [5:04:14<34:45:34, 14.46s/it]
 12%|█▏        | 1214/9864 [5:04:31<36:47:03, 15.31s/it]
 12%|█▏        | 1215/9864 [5:04:45<35:35:58, 14.82s/it]
 12%|█▏        | 1216/9864 [5:04:59<34:54:45, 14.53s/it]
 12%|█▏        | 1217/9864 [5:05:15<35:57:27, 14.97s/it]
 12%|█▏        | 1218/9864 [5:05:29<35:15:22, 14.68s/it]
 12%|█▏        | 1219/9864 [5:05:45<36:03:25, 15.02s/it]
 12%|█▏        | 1220/9864 [5:05:59<35:22:12, 14.73s/it]
                                                        
{'loss': '0.1683', 'grad_norm': '3.489', 'learning_rate': '1.954e-05', 'epoch': '0.1237'}

 12%|█▏        | 1220/9864 [5:05:59<35:22:12, 14.73s/it]
 12%|█▏        | 1221/9864 [5:06:15<36:30:20, 15.21s/it]
 12%|█▏        | 1222/9864 [5:06:30<36:43:18, 15.30s/it]
 12%|█▏        | 1223/9864 [5:06:45<35:51:13, 14.94s/it]
 12%|█▏        | 1224/9864 [5:07:00<36:30:20, 15.21s/it]
 12%|█▏        | 1225/9864 [5:07:15<35:46:49, 14.91s/it]
 12%|█▏        | 1226/9864 [5:07:29<35:15:17, 14.69s/it]
 12%|█▏        | 1227/9864 [5:07:44<35:27:48, 14.78s/it]
 12%|█▏        | 1228/9864 [5:07:58<34:59:37, 14.59s/it]
 12%|█▏        | 1229/9864 [5:08:15<37:04:02, 15.45s/it]
 12%|█▏        | 1230/9864 [5:08:30<36:20:34, 15.15s/it]
                                                        
{'loss': '0.1765', 'grad_norm': '5.326', 'learning_rate': '1.953e-05', 'epoch': '0.1247'}

 12%|█▏        | 1230/9864 [5:08:30<36:20:34, 15.15s/it]
 12%|█▏        | 1231/9864 [5:08:44<35:46:55, 14.92s/it]
 12%|█▏        | 1232/9864 [5:09:01<36:55:38, 15.40s/it]
 12%|█▎        | 1233/9864 [5:09:15<36:05:58, 15.06s/it]
 13%|█▎        | 1234/9864 [5:09:29<35:18:25, 14.73s/it]
 13%|█▎        | 1235/9864 [5:09:43<34:48:18, 14.52s/it]
 13%|█▎        | 1236/9864 [5:10:02<38:10:52, 15.93s/it]
 13%|█▎        | 1237/9864 [5:10:16<36:40:38, 15.31s/it]
 13%|█▎        | 1238/9864 [5:10:30<35:49:42, 14.95s/it]
 13%|█▎        | 1239/9864 [5:10:44<35:05:55, 14.65s/it]
 13%|█▎        | 1240/9864 [5:10:58<34:21:43, 14.34s/it]
                                                        
{'loss': '0.1745', 'grad_norm': '2.051', 'learning_rate': '1.952e-05', 'epoch': '0.1257'}

 13%|█▎        | 1240/9864 [5:10:58<34:21:43, 14.34s/it]
 13%|█▎        | 1241/9864 [5:11:14<35:31:26, 14.83s/it]
 13%|█▎        | 1242/9864 [5:11:28<34:56:41, 14.59s/it]
 13%|█▎        | 1243/9864 [5:11:42<34:19:24, 14.33s/it]
 13%|█▎        | 1244/9864 [5:11:58<35:33:17, 14.85s/it]
 13%|█▎        | 1245/9864 [5:12:15<37:25:11, 15.63s/it]
 13%|█▎        | 1246/9864 [5:12:29<36:28:43, 15.24s/it]
 13%|█▎        | 1247/9864 [5:12:44<35:41:02, 14.91s/it]
 13%|█▎        | 1248/9864 [5:12:57<34:58:33, 14.61s/it]
 13%|█▎        | 1249/9864 [5:13:13<35:44:07, 14.93s/it]
 13%|█▎        | 1250/9864 [5:13:27<35:05:01, 14.66s/it]
                                                        
{'loss': '0.1788', 'grad_norm': '4.164', 'learning_rate': '1.951e-05', 'epoch': '0.1267'}

 13%|█▎        | 1250/9864 [5:13:27<35:05:01, 14.66s/it]
 13%|█▎        | 1251/9864 [5:13:41<34:16:47, 14.33s/it]
 13%|█▎        | 1252/9864 [5:13:56<35:16:46, 14.75s/it]
 13%|█▎        | 1253/9864 [5:14:14<37:16:29, 15.58s/it]
 13%|█▎        | 1254/9864 [5:14:28<36:00:06, 15.05s/it]
 13%|█▎        | 1255/9864 [5:14:42<35:23:35, 14.80s/it]
 13%|█▎        | 1256/9864 [5:14:56<34:43:50, 14.52s/it]
 13%|█▎        | 1257/9864 [5:15:10<34:33:19, 14.45s/it]
 13%|█▎        | 1258/9864 [5:15:26<35:37:51, 14.90s/it]
 13%|█▎        | 1259/9864 [5:15:42<36:07:46, 15.12s/it]
 13%|█▎        | 1260/9864 [5:15:58<36:52:03, 15.43s/it]
                                                        
{'loss': '0.1734', 'grad_norm': '5.887', 'learning_rate': '1.95e-05', 'epoch': '0.1277'}

 13%|█▎        | 1260/9864 [5:15:58<36:52:03, 15.43s/it]
 13%|█▎        | 1261/9864 [5:16:13<36:40:12, 15.34s/it]
 13%|█▎        | 1262/9864 [5:16:27<35:23:09, 14.81s/it]
 13%|█▎        | 1263/9864 [5:16:41<34:57:25, 14.63s/it]
 13%|█▎        | 1264/9864 [5:16:55<34:39:42, 14.51s/it]
 13%|█▎        | 1265/9864 [5:17:09<34:23:11, 14.40s/it]
 13%|█▎        | 1266/9864 [5:17:25<35:09:01, 14.72s/it]
 13%|█▎        | 1267/9864 [5:17:39<34:46:49, 14.56s/it]
 13%|█▎        | 1268/9864 [5:17:54<35:27:34, 14.85s/it]
 13%|█▎        | 1269/9864 [5:18:12<37:20:19, 15.64s/it]
 13%|█▎        | 1270/9864 [5:18:26<36:19:51, 15.22s/it]
                                                        
{'loss': '0.1876', 'grad_norm': '2.772', 'learning_rate': '1.949e-05', 'epoch': '0.1288'}

 13%|█▎        | 1270/9864 [5:18:26<36:19:51, 15.22s/it]
 13%|█▎        | 1271/9864 [5:18:40<35:44:00, 14.97s/it]
 13%|█▎        | 1272/9864 [5:18:54<34:55:11, 14.63s/it]
 13%|█▎        | 1273/9864 [5:19:08<34:31:46, 14.47s/it]
 13%|█▎        | 1274/9864 [5:19:24<35:15:35, 14.78s/it]
 13%|█▎        | 1275/9864 [5:19:38<34:41:50, 14.54s/it]
 13%|█▎        | 1276/9864 [5:19:54<35:28:20, 14.87s/it]
 13%|█▎        | 1277/9864 [5:20:08<34:56:46, 14.65s/it]
 13%|█▎        | 1278/9864 [5:20:25<36:44:15, 15.40s/it]
 13%|█▎        | 1279/9864 [5:20:39<35:36:42, 14.93s/it]
 13%|█▎        | 1280/9864 [5:20:52<34:37:22, 14.52s/it]
                                                        
{'loss': '0.1606', 'grad_norm': '3.726', 'learning_rate': '1.948e-05', 'epoch': '0.1298'}

 13%|█▎        | 1280/9864 [5:20:52<34:37:22, 14.52s/it]
 13%|█▎        | 1281/9864 [5:21:08<35:33:40, 14.92s/it]
 13%|█▎        | 1282/9864 [5:21:22<35:10:26, 14.75s/it]
 13%|█▎        | 1283/9864 [5:21:37<34:53:00, 14.63s/it]
 13%|█▎        | 1284/9864 [5:21:52<35:17:02, 14.80s/it]
 13%|█▎        | 1285/9864 [5:22:06<34:38:58, 14.54s/it]
 13%|█▎        | 1286/9864 [5:22:24<36:55:03, 15.49s/it]
 13%|█▎        | 1287/9864 [5:22:37<35:30:23, 14.90s/it]
 13%|█▎        | 1288/9864 [5:22:51<34:39:03, 14.55s/it]
 13%|█▎        | 1289/9864 [5:23:07<35:43:12, 15.00s/it]
 13%|█▎        | 1290/9864 [5:23:21<35:24:45, 14.87s/it]
                                                        
{'loss': '0.148', 'grad_norm': '2.552', 'learning_rate': '1.947e-05', 'epoch': '0.1308'}

 13%|█▎        | 1290/9864 [5:23:21<35:24:45, 14.87s/it]
 13%|█▎        | 1291/9864 [5:23:36<34:58:39, 14.69s/it]
 13%|█▎        | 1292/9864 [5:23:52<35:57:09, 15.10s/it]
 13%|█▎        | 1293/9864 [5:24:06<35:01:52, 14.71s/it]
 13%|█▎        | 1294/9864 [5:24:23<37:00:36, 15.55s/it]
 13%|█▎        | 1295/9864 [5:24:37<35:57:04, 15.10s/it]
 13%|█▎        | 1296/9864 [5:24:51<35:00:02, 14.71s/it]
 13%|█▎        | 1297/9864 [5:25:04<34:09:44, 14.36s/it]
 13%|█▎        | 1298/9864 [5:25:21<35:30:43, 14.92s/it]
 13%|█▎        | 1299/9864 [5:25:35<34:57:32, 14.69s/it]
 13%|█▎        | 1300/9864 [5:25:51<35:47:12, 15.04s/it]
                                                        
{'loss': '0.1739', 'grad_norm': '4.825', 'learning_rate': '1.946e-05', 'epoch': '0.1318'}

 13%|█▎        | 1300/9864 [5:25:51<35:47:12, 15.04s/it]
 13%|█▎        | 1301/9864 [5:26:07<36:28:54, 15.34s/it]
 13%|█▎        | 1302/9864 [5:26:23<36:45:16, 15.45s/it]
 13%|█▎        | 1303/9864 [5:26:37<36:01:03, 15.15s/it]
 13%|█▎        | 1304/9864 [5:26:51<35:27:01, 14.91s/it]
 13%|█▎        | 1305/9864 [5:27:06<34:59:13, 14.72s/it]
 13%|█▎        | 1306/9864 [5:27:21<35:31:41, 14.95s/it]
 13%|█▎        | 1307/9864 [5:27:35<34:55:56, 14.70s/it]
 13%|█▎        | 1308/9864 [5:27:50<35:16:33, 14.84s/it]
 13%|█▎        | 1309/9864 [5:28:06<36:09:41, 15.22s/it]
 13%|█▎        | 1310/9864 [5:28:21<35:38:16, 15.00s/it]
                                                        
{'loss': '0.1571', 'grad_norm': '1.722', 'learning_rate': '1.945e-05', 'epoch': '0.1328'}

 13%|█▎        | 1310/9864 [5:28:21<35:38:16, 15.00s/it]
 13%|█▎        | 1311/9864 [5:28:36<35:59:06, 15.15s/it]
 13%|█▎        | 1312/9864 [5:28:51<35:22:11, 14.89s/it]
 13%|█▎        | 1313/9864 [5:29:07<36:07:42, 15.21s/it]
 13%|█▎        | 1314/9864 [5:29:20<35:08:18, 14.80s/it]
 13%|█▎        | 1315/9864 [5:29:34<34:17:16, 14.44s/it]
 13%|█▎        | 1316/9864 [5:29:52<36:33:15, 15.39s/it]
 13%|█▎        | 1317/9864 [5:30:05<35:14:14, 14.84s/it]
 13%|█▎        | 1318/9864 [5:30:19<34:12:16, 14.41s/it]
 13%|█▎        | 1319/9864 [5:30:34<35:03:14, 14.77s/it]
 13%|█▎        | 1320/9864 [5:30:50<35:41:36, 15.04s/it]
                                                        
{'loss': '0.1674', 'grad_norm': '2.394', 'learning_rate': '1.944e-05', 'epoch': '0.1338'}

 13%|█▎        | 1320/9864 [5:30:50<35:41:36, 15.04s/it]
 13%|█▎        | 1321/9864 [5:31:04<35:02:05, 14.76s/it]
 13%|█▎        | 1322/9864 [5:31:18<34:14:12, 14.43s/it]
 13%|█▎        | 1323/9864 [5:31:31<33:43:00, 14.21s/it]
 13%|█▎        | 1324/9864 [5:31:48<35:07:25, 14.81s/it]
 13%|█▎        | 1325/9864 [5:32:03<35:31:20, 14.98s/it]
 13%|█▎        | 1326/9864 [5:32:17<34:59:26, 14.75s/it]
 13%|█▎        | 1327/9864 [5:32:31<34:37:18, 14.60s/it]
 13%|█▎        | 1328/9864 [5:32:47<35:25:42, 14.94s/it]
 13%|█▎        | 1329/9864 [5:33:01<34:47:10, 14.67s/it]
 13%|█▎        | 1330/9864 [5:33:15<34:19:02, 14.48s/it]
                                                        
{'loss': '0.1669', 'grad_norm': '3.466', 'learning_rate': '1.943e-05', 'epoch': '0.1348'}

 13%|█▎        | 1330/9864 [5:33:15<34:19:02, 14.48s/it]
 13%|█▎        | 1331/9864 [5:33:29<34:05:15, 14.38s/it]
 14%|█▎        | 1332/9864 [5:33:45<34:55:45, 14.74s/it]
 14%|█▎        | 1333/9864 [5:34:01<35:50:46, 15.13s/it]
 14%|█▎        | 1334/9864 [5:34:15<35:00:55, 14.78s/it]
 14%|█▎        | 1335/9864 [5:34:31<35:56:10, 15.17s/it]
 14%|█▎        | 1336/9864 [5:34:45<34:48:49, 14.70s/it]
 14%|█▎        | 1337/9864 [5:35:00<35:29:10, 14.98s/it]
 14%|█▎        | 1338/9864 [5:35:15<35:01:30, 14.79s/it]
 14%|█▎        | 1339/9864 [5:35:29<34:23:27, 14.52s/it]
 14%|█▎        | 1340/9864 [5:35:44<35:17:20, 14.90s/it]
                                                        
{'loss': '0.1736', 'grad_norm': '2.046', 'learning_rate': '1.942e-05', 'epoch': '0.1359'}

 14%|█▎        | 1340/9864 [5:35:44<35:17:20, 14.90s/it]
 14%|█▎        | 1341/9864 [5:36:00<35:37:53, 15.05s/it]
 14%|█▎        | 1342/9864 [5:36:13<34:32:47, 14.59s/it]
 14%|█▎        | 1343/9864 [5:36:29<35:21:01, 14.94s/it]
 14%|█▎        | 1344/9864 [5:36:44<35:25:00, 14.96s/it]
 14%|█▎        | 1345/9864 [5:36:58<34:41:05, 14.66s/it]
 14%|█▎        | 1346/9864 [5:37:12<34:11:57, 14.45s/it]
 14%|█▎        | 1347/9864 [5:37:28<35:20:59, 14.94s/it]
 14%|█▎        | 1348/9864 [5:37:42<34:52:19, 14.74s/it]
 14%|█▎        | 1349/9864 [5:37:58<35:49:20, 15.15s/it]
 14%|█▎        | 1350/9864 [5:38:14<36:17:33, 15.35s/it]
                                                        
{'loss': '0.1546', 'grad_norm': '3.893', 'learning_rate': '1.941e-05', 'epoch': '0.1369'}

 14%|█▎        | 1350/9864 [5:38:14<36:17:33, 15.35s/it]
 14%|█▎        | 1351/9864 [5:38:29<35:53:07, 15.18s/it]
 14%|█▎        | 1352/9864 [5:38:43<35:15:25, 14.91s/it]
 14%|█▎        | 1353/9864 [5:38:59<35:55:55, 15.20s/it]
 14%|█▎        | 1354/9864 [5:39:13<35:15:27, 14.92s/it]
 14%|█▎        | 1355/9864 [5:39:29<35:58:36, 15.22s/it]
 14%|█▎        | 1356/9864 [5:39:45<36:37:02, 15.49s/it]
 14%|█▍        | 1357/9864 [5:39:59<35:33:48, 15.05s/it]
 14%|█▍        | 1358/9864 [5:40:15<36:04:51, 15.27s/it]
 14%|█▍        | 1359/9864 [5:40:29<35:08:23, 14.87s/it]
 14%|█▍        | 1360/9864 [5:40:45<35:26:43, 15.01s/it]
                                                        
{'loss': '0.1634', 'grad_norm': '4.678', 'learning_rate': '1.94e-05', 'epoch': '0.1379'}

 14%|█▍        | 1360/9864 [5:40:45<35:26:43, 15.01s/it]
 14%|█▍        | 1361/9864 [5:40:59<34:48:21, 14.74s/it]
 14%|█▍        | 1362/9864 [5:41:12<34:08:48, 14.46s/it]
 14%|█▍        | 1363/9864 [5:41:28<35:08:02, 14.88s/it]
 14%|█▍        | 1364/9864 [5:41:44<35:28:28, 15.02s/it]
 14%|█▍        | 1365/9864 [5:41:58<34:41:46, 14.70s/it]
 14%|█▍        | 1366/9864 [5:42:13<35:01:22, 14.84s/it]
 14%|█▍        | 1367/9864 [5:42:27<34:21:17, 14.56s/it]
 14%|█▍        | 1368/9864 [5:42:42<34:37:19, 14.67s/it]
 14%|█▍        | 1369/9864 [5:42:56<34:37:38, 14.67s/it]
 14%|█▍        | 1370/9864 [5:43:10<34:08:10, 14.47s/it]
                                                        
{'loss': '0.1756', 'grad_norm': '3.018', 'learning_rate': '1.939e-05', 'epoch': '0.1389'}

 14%|█▍        | 1370/9864 [5:43:10<34:08:10, 14.47s/it]
 14%|█▍        | 1371/9864 [5:43:26<34:57:05, 14.82s/it]
 14%|█▍        | 1372/9864 [5:43:41<35:18:28, 14.97s/it]
 14%|█▍        | 1373/9864 [5:43:55<34:44:18, 14.73s/it]
 14%|█▍        | 1374/9864 [5:44:10<34:22:34, 14.58s/it]
 14%|█▍        | 1375/9864 [5:44:25<35:15:20, 14.95s/it]
 14%|█▍        | 1376/9864 [5:44:41<35:55:21, 15.24s/it]
 14%|█▍        | 1377/9864 [5:44:55<35:09:36, 14.91s/it]
 14%|█▍        | 1378/9864 [5:45:09<34:27:04, 14.62s/it]
 14%|█▍        | 1379/9864 [5:45:26<35:31:34, 15.07s/it]
 14%|█▍        | 1380/9864 [5:45:40<34:52:12, 14.80s/it]
                                                        
{'loss': '0.1623', 'grad_norm': '4.071', 'learning_rate': '1.937e-05', 'epoch': '0.1399'}

 14%|█▍        | 1380/9864 [5:45:40<34:52:12, 14.80s/it]
 14%|█▍        | 1381/9864 [5:45:56<35:42:35, 15.15s/it]
 14%|█▍        | 1382/9864 [5:46:10<35:03:02, 14.88s/it]
 14%|█▍        | 1383/9864 [5:46:27<36:53:37, 15.66s/it]
 14%|█▍        | 1384/9864 [5:46:42<35:54:17, 15.24s/it]
 14%|█▍        | 1385/9864 [5:46:56<35:00:10, 14.86s/it]
 14%|█▍        | 1386/9864 [5:47:11<35:33:47, 15.10s/it]
 14%|█▍        | 1387/9864 [5:47:25<34:38:35, 14.71s/it]
 14%|█▍        | 1388/9864 [5:47:39<33:56:30, 14.42s/it]
 14%|█▍        | 1389/9864 [5:47:55<35:05:14, 14.90s/it]
 14%|█▍        | 1390/9864 [5:48:11<35:46:52, 15.20s/it]
                                                        
{'loss': '0.1822', 'grad_norm': '6.952', 'learning_rate': '1.936e-05', 'epoch': '0.1409'}

 14%|█▍        | 1390/9864 [5:48:11<35:46:52, 15.20s/it]
 14%|█▍        | 1391/9864 [5:48:27<36:08:58, 15.36s/it]
 14%|█▍        | 1392/9864 [5:48:41<35:32:47, 15.10s/it]
 14%|█▍        | 1393/9864 [5:48:55<34:52:32, 14.82s/it]
 14%|█▍        | 1394/9864 [5:49:11<35:48:00, 15.22s/it]
 14%|█▍        | 1395/9864 [5:49:26<35:05:08, 14.91s/it]
 14%|█▍        | 1396/9864 [5:49:41<35:18:46, 15.01s/it]
 14%|█▍        | 1397/9864 [5:49:55<34:25:19, 14.64s/it]
 14%|█▍        | 1398/9864 [5:50:11<35:24:44, 15.06s/it]
 14%|█▍        | 1399/9864 [5:50:26<35:51:36, 15.25s/it]
 14%|█▍        | 1400/9864 [5:50:40<35:05:29, 14.93s/it]
                                                        
{'loss': '0.1686', 'grad_norm': '2.84', 'learning_rate': '1.935e-05', 'epoch': '0.1419'}

 14%|█▍        | 1400/9864 [5:50:40<35:05:29, 14.93s/it]
 14%|█▍        | 1401/9864 [5:50:55<34:31:38, 14.69s/it]
 14%|█▍        | 1402/9864 [5:51:09<34:21:07, 14.61s/it]
 14%|█▍        | 1403/9864 [5:51:25<35:06:34, 14.94s/it]
 14%|█▍        | 1404/9864 [5:51:41<35:49:00, 15.24s/it]
 14%|█▍        | 1405/9864 [5:51:55<35:18:47, 15.03s/it]
 14%|█▍        | 1406/9864 [5:52:11<35:44:43, 15.21s/it]
 14%|█▍        | 1407/9864 [5:52:25<35:14:31, 15.00s/it]
 14%|█▍        | 1408/9864 [5:52:42<36:04:08, 15.36s/it]
 14%|█▍        | 1409/9864 [5:52:55<34:59:40, 14.90s/it]
 14%|█▍        | 1410/9864 [5:53:09<34:23:51, 14.65s/it]
                                                        
{'loss': '0.1586', 'grad_norm': '3.131', 'learning_rate': '1.934e-05', 'epoch': '0.143'}

 14%|█▍        | 1410/9864 [5:53:09<34:23:51, 14.65s/it]
 14%|█▍        | 1411/9864 [5:53:27<36:23:27, 15.50s/it]
 14%|█▍        | 1412/9864 [5:53:41<35:19:11, 15.04s/it]
 14%|█▍        | 1413/9864 [5:53:57<35:54:08, 15.29s/it]
 14%|█▍        | 1414/9864 [5:54:11<34:53:19, 14.86s/it]
 14%|█▍        | 1415/9864 [5:54:26<35:32:09, 15.14s/it]
 14%|█▍        | 1416/9864 [5:54:40<34:45:24, 14.81s/it]
 14%|█▍        | 1417/9864 [5:54:55<34:26:53, 14.68s/it]
 14%|█▍        | 1418/9864 [5:55:09<34:16:45, 14.61s/it]
 14%|█▍        | 1419/9864 [5:55:26<36:00:38, 15.35s/it]
 14%|█▍        | 1420/9864 [5:55:41<35:11:29, 15.00s/it]
                                                        
{'loss': '0.1809', 'grad_norm': '2.073', 'learning_rate': '1.933e-05', 'epoch': '0.144'}

 14%|█▍        | 1420/9864 [5:55:41<35:11:29, 15.00s/it]
 14%|█▍        | 1421/9864 [5:55:56<35:42:10, 15.22s/it]
 14%|█▍        | 1422/9864 [5:56:10<34:51:57, 14.87s/it]
 14%|█▍        | 1423/9864 [5:56:26<35:24:00, 15.10s/it]
 14%|█▍        | 1424/9864 [5:56:40<34:52:09, 14.87s/it]
 14%|█▍        | 1425/9864 [5:56:54<34:18:50, 14.64s/it]
 14%|█▍        | 1426/9864 [5:57:09<34:33:58, 14.75s/it]
 14%|█▍        | 1427/9864 [5:57:25<34:55:19, 14.90s/it]
 14%|█▍        | 1428/9864 [5:57:40<35:27:30, 15.13s/it]
 14%|█▍        | 1429/9864 [5:57:54<34:32:24, 14.74s/it]
 14%|█▍        | 1430/9864 [5:58:08<34:00:01, 14.51s/it]
                                                        
{'loss': '0.1734', 'grad_norm': '3.105', 'learning_rate': '1.932e-05', 'epoch': '0.145'}

 14%|█▍        | 1430/9864 [5:58:08<34:00:01, 14.51s/it]
 15%|█▍        | 1431/9864 [5:58:24<34:49:03, 14.86s/it]
 15%|█▍        | 1432/9864 [5:58:38<34:31:41, 14.74s/it]
 15%|█▍        | 1433/9864 [5:58:52<33:58:59, 14.51s/it]
 15%|█▍        | 1434/9864 [5:59:06<33:32:48, 14.33s/it]
 15%|█▍        | 1435/9864 [5:59:23<35:24:53, 15.13s/it]
 15%|█▍        | 1436/9864 [5:59:37<34:26:43, 14.71s/it]
 15%|█▍        | 1437/9864 [5:59:53<35:12:56, 15.04s/it]
 15%|█▍        | 1438/9864 [6:00:07<34:38:41, 14.80s/it]
 15%|█▍        | 1439/9864 [6:00:23<35:34:39, 15.20s/it]
 15%|█▍        | 1440/9864 [6:00:37<34:55:56, 14.93s/it]
                                                        
{'loss': '0.1647', 'grad_norm': '2.964', 'learning_rate': '1.93e-05', 'epoch': '0.146'}

 15%|█▍        | 1440/9864 [6:00:37<34:55:56, 14.93s/it]
 15%|█▍        | 1441/9864 [6:00:52<34:33:22, 14.77s/it]
 15%|█▍        | 1442/9864 [6:01:07<35:10:50, 15.04s/it]
 15%|█▍        | 1443/9864 [6:01:23<35:36:41, 15.22s/it]
 15%|█▍        | 1444/9864 [6:01:37<34:32:01, 14.76s/it]
 15%|█▍        | 1445/9864 [6:01:53<35:13:48, 15.06s/it]
 15%|█▍        | 1446/9864 [6:02:07<34:36:22, 14.80s/it]
 15%|█▍        | 1447/9864 [6:02:23<35:18:30, 15.10s/it]
 15%|█▍        | 1448/9864 [6:02:36<34:28:10, 14.74s/it]
 15%|█▍        | 1449/9864 [6:02:50<33:45:41, 14.44s/it]
 15%|█▍        | 1450/9864 [6:03:06<34:48:59, 14.90s/it]
                                                        
{'loss': '0.1541', 'grad_norm': '2.731', 'learning_rate': '1.929e-05', 'epoch': '0.147'}

 15%|█▍        | 1450/9864 [6:03:06<34:48:59, 14.90s/it]
 15%|█▍        | 1451/9864 [6:03:22<35:29:22, 15.19s/it]
 15%|█▍        | 1452/9864 [6:03:37<35:39:49, 15.26s/it]
 15%|█▍        | 1453/9864 [6:03:51<34:39:21, 14.83s/it]
 15%|█▍        | 1454/9864 [6:04:05<33:55:00, 14.52s/it]
 15%|█▍        | 1455/9864 [6:04:21<34:47:15, 14.89s/it]
 15%|█▍        | 1456/9864 [6:04:35<34:17:07, 14.68s/it]
 15%|█▍        | 1457/9864 [6:04:49<33:47:00, 14.47s/it]
 15%|█▍        | 1458/9864 [6:05:05<34:40:54, 14.85s/it]
 15%|█▍        | 1459/9864 [6:05:22<36:23:47, 15.59s/it]
 15%|█▍        | 1460/9864 [6:05:37<35:39:49, 15.28s/it]
                                                        
{'loss': '0.1796', 'grad_norm': '2.984', 'learning_rate': '1.928e-05', 'epoch': '0.148'}

 15%|█▍        | 1460/9864 [6:05:37<35:39:49, 15.28s/it]
 15%|█▍        | 1461/9864 [6:05:51<34:56:47, 14.97s/it]
 15%|█▍        | 1462/9864 [6:06:05<34:31:52, 14.80s/it]
 15%|█▍        | 1463/9864 [6:06:21<35:11:57, 15.08s/it]
 15%|█▍        | 1464/9864 [6:06:35<34:40:10, 14.86s/it]
 15%|█▍        | 1465/9864 [6:06:51<35:11:14, 15.08s/it]
 15%|█▍        | 1466/9864 [6:07:05<34:41:13, 14.87s/it]
 15%|█▍        | 1467/9864 [6:07:23<36:22:01, 15.59s/it]
 15%|█▍        | 1468/9864 [6:07:36<35:00:26, 15.01s/it]
 15%|█▍        | 1469/9864 [6:07:50<34:25:04, 14.76s/it]
 15%|█▍        | 1470/9864 [6:08:06<35:11:43, 15.09s/it]
                                                        
{'loss': '0.1694', 'grad_norm': '1.617', 'learning_rate': '1.927e-05', 'epoch': '0.149'}

 15%|█▍        | 1470/9864 [6:08:06<35:11:43, 15.09s/it]
 15%|█▍        | 1471/9864 [6:08:21<34:43:03, 14.89s/it]
 15%|█▍        | 1472/9864 [6:08:35<34:03:02, 14.61s/it]
 15%|█▍        | 1473/9864 [6:08:50<34:51:30, 14.96s/it]
 15%|█▍        | 1474/9864 [6:09:06<35:06:32, 15.06s/it]
 15%|█▍        | 1475/9864 [6:09:21<35:13:36, 15.12s/it]
 15%|█▍        | 1476/9864 [6:09:35<34:41:25, 14.89s/it]
 15%|█▍        | 1477/9864 [6:09:51<35:03:49, 15.05s/it]
 15%|█▍        | 1478/9864 [6:10:05<34:36:02, 14.85s/it]
 15%|█▍        | 1479/9864 [6:10:19<34:03:44, 14.62s/it]
 15%|█▌        | 1480/9864 [6:10:33<33:40:59, 14.46s/it]
                                                        
{'loss': '0.1662', 'grad_norm': '3.482', 'learning_rate': '1.926e-05', 'epoch': '0.1501'}

 15%|█▌        | 1480/9864 [6:10:33<33:40:59, 14.46s/it]
 15%|█▌        | 1481/9864 [6:10:49<34:32:28, 14.83s/it]
 15%|█▌        | 1482/9864 [6:11:05<35:33:56, 15.28s/it]
 15%|█▌        | 1483/9864 [6:11:21<35:38:11, 15.31s/it]
 15%|█▌        | 1484/9864 [6:11:35<34:51:01, 14.97s/it]
 15%|█▌        | 1485/9864 [6:11:49<34:10:21, 14.68s/it]
 15%|█▌        | 1486/9864 [6:12:05<35:15:26, 15.15s/it]
 15%|█▌        | 1487/9864 [6:12:19<34:29:16, 14.82s/it]
 15%|█▌        | 1488/9864 [6:12:33<34:05:23, 14.65s/it]
 15%|█▌        | 1489/9864 [6:12:48<33:47:04, 14.52s/it]
 15%|█▌        | 1490/9864 [6:13:06<36:40:12, 15.76s/it]
                                                        
{'loss': '0.1685', 'grad_norm': '2.367', 'learning_rate': '1.924e-05', 'epoch': '0.1511'}

 15%|█▌        | 1490/9864 [6:13:06<36:40:12, 15.76s/it]
 15%|█▌        | 1491/9864 [6:13:21<35:40:29, 15.34s/it]
 15%|█▌        | 1492/9864 [6:13:35<34:44:38, 14.94s/it]
 15%|█▌        | 1493/9864 [6:13:49<34:09:39, 14.69s/it]
 15%|█▌        | 1494/9864 [6:14:03<33:48:12, 14.54s/it]
 15%|█▌        | 1495/9864 [6:14:19<34:51:25, 14.99s/it]
 15%|█▌        | 1496/9864 [6:14:33<34:12:44, 14.72s/it]
 15%|█▌        | 1497/9864 [6:14:50<35:54:12, 15.45s/it]
 15%|█▌        | 1498/9864 [6:15:06<36:24:38, 15.67s/it]
 15%|█▌        | 1499/9864 [6:15:20<35:01:04, 15.07s/it]
 15%|█▌        | 1500/9864 [6:15:34<34:24:26, 14.81s/it]
                                                        
{'loss': '0.1834', 'grad_norm': '3.015', 'learning_rate': '1.923e-05', 'epoch': '0.1521'}

 15%|█▌        | 1500/9864 [6:15:34<34:24:26, 14.81s/it]
 15%|█▌        | 1501/9864 [6:15:48<33:42:04, 14.51s/it]
 15%|█▌        | 1502/9864 [6:16:02<33:31:59, 14.44s/it]
 15%|█▌        | 1503/9864 [6:16:18<34:11:39, 14.72s/it]
 15%|█▌        | 1504/9864 [6:16:32<33:50:49, 14.58s/it]
 15%|█▌        | 1505/9864 [6:16:48<34:33:33, 14.88s/it]
 15%|█▌        | 1506/9864 [6:17:04<35:29:45, 15.29s/it]
 15%|█▌        | 1507/9864 [6:17:18<34:47:54, 14.99s/it]
 15%|█▌        | 1508/9864 [6:17:32<34:09:36, 14.72s/it]
 15%|█▌        | 1509/9864 [6:17:46<33:27:52, 14.42s/it]
 15%|█▌        | 1510/9864 [6:18:02<34:36:29, 14.91s/it]
                                                        
{'loss': '0.1762', 'grad_norm': '2.32', 'learning_rate': '1.922e-05', 'epoch': '0.1531'}

 15%|█▌        | 1510/9864 [6:18:02<34:36:29, 14.91s/it]
 15%|█▌        | 1511/9864 [6:18:16<34:10:41, 14.73s/it]
 15%|█▌        | 1512/9864 [6:18:32<34:44:09, 14.97s/it]
 15%|█▌        | 1513/9864 [6:18:49<36:04:36, 15.55s/it]
 15%|█▌        | 1514/9864 [6:19:03<35:14:10, 15.19s/it]
 15%|█▌        | 1515/9864 [6:19:17<34:29:37, 14.87s/it]
 15%|█▌        | 1516/9864 [6:19:31<33:49:19, 14.59s/it]
 15%|█▌        | 1517/9864 [6:19:47<34:36:54, 14.93s/it]
 15%|█▌        | 1518/9864 [6:20:02<34:25:25, 14.85s/it]
 15%|█▌        | 1519/9864 [6:20:16<33:51:57, 14.61s/it]
 15%|█▌        | 1520/9864 [6:20:32<35:23:41, 15.27s/it]
                                                        
{'loss': '0.1637', 'grad_norm': '4.603', 'learning_rate': '1.92e-05', 'epoch': '0.1541'}

 15%|█▌        | 1520/9864 [6:20:32<35:23:41, 15.27s/it]
 15%|█▌        | 1521/9864 [6:20:48<35:55:01, 15.50s/it]
 15%|█▌        | 1522/9864 [6:21:03<35:11:25, 15.19s/it]
 15%|█▌        | 1523/9864 [6:21:17<34:29:22, 14.89s/it]
 15%|█▌        | 1524/9864 [6:21:31<33:53:29, 14.63s/it]
 15%|█▌        | 1525/9864 [6:21:47<34:31:42, 14.91s/it]
 15%|█▌        | 1526/9864 [6:22:01<33:48:33, 14.60s/it]
 15%|█▌        | 1527/9864 [6:22:17<34:49:40, 15.04s/it]
 15%|█▌        | 1528/9864 [6:22:32<35:17:09, 15.24s/it]
 16%|█▌        | 1529/9864 [6:22:48<35:33:07, 15.36s/it]
 16%|█▌        | 1530/9864 [6:23:02<34:38:20, 14.96s/it]
                                                        
{'loss': '0.1647', 'grad_norm': '1.714', 'learning_rate': '1.919e-05', 'epoch': '0.1551'}

 16%|█▌        | 1530/9864 [6:23:02<34:38:20, 14.96s/it]
 16%|█▌        | 1531/9864 [6:23:16<34:14:26, 14.79s/it]
 16%|█▌        | 1532/9864 [6:23:30<33:44:42, 14.58s/it]
 16%|█▌        | 1533/9864 [6:23:46<34:29:40, 14.91s/it]
 16%|█▌        | 1534/9864 [6:24:00<33:44:06, 14.58s/it]
 16%|█▌        | 1535/9864 [6:24:15<34:09:44, 14.77s/it]
 16%|█▌        | 1536/9864 [6:24:32<35:51:13, 15.50s/it]
 16%|█▌        | 1537/9864 [6:24:46<34:39:16, 14.98s/it]
 16%|█▌        | 1538/9864 [6:25:00<34:05:56, 14.74s/it]
 16%|█▌        | 1539/9864 [6:25:15<33:50:24, 14.63s/it]
 16%|█▌        | 1540/9864 [6:25:30<34:16:33, 14.82s/it]
                                                        
{'loss': '0.166', 'grad_norm': '4.208', 'learning_rate': '1.918e-05', 'epoch': '0.1561'}

 16%|█▌        | 1540/9864 [6:25:30<34:16:33, 14.82s/it]
 16%|█▌        | 1541/9864 [6:25:44<33:45:43, 14.60s/it]
 16%|█▌        | 1542/9864 [6:26:00<34:28:30, 14.91s/it]
 16%|█▌        | 1543/9864 [6:26:13<33:30:21, 14.50s/it]
 16%|█▌        | 1544/9864 [6:26:30<35:24:53, 15.32s/it]
 16%|█▌        | 1545/9864 [6:26:45<34:33:31, 14.96s/it]
 16%|█▌        | 1546/9864 [6:26:59<33:59:16, 14.71s/it]
 16%|█▌        | 1547/9864 [6:27:13<33:43:43, 14.60s/it]
 16%|█▌        | 1548/9864 [6:27:28<34:02:07, 14.73s/it]
 16%|█▌        | 1549/9864 [6:27:42<33:20:41, 14.44s/it]
 16%|█▌        | 1550/9864 [6:27:57<33:54:15, 14.68s/it]
                                                        
{'loss': '0.1651', 'grad_norm': '2.51', 'learning_rate': '1.917e-05', 'epoch': '0.1572'}

 16%|█▌        | 1550/9864 [6:27:57<33:54:15, 14.68s/it]
 16%|█▌        | 1551/9864 [6:28:12<33:47:20, 14.63s/it]
 16%|█▌        | 1552/9864 [6:28:29<35:21:05, 15.31s/it]
 16%|█▌        | 1553/9864 [6:28:43<34:43:12, 15.04s/it]
 16%|█▌        | 1554/9864 [6:28:57<34:07:20, 14.78s/it]
 16%|█▌        | 1555/9864 [6:29:13<34:38:08, 15.01s/it]
 16%|█▌        | 1556/9864 [6:29:27<33:57:23, 14.71s/it]
 16%|█▌        | 1557/9864 [6:29:41<33:40:44, 14.60s/it]
 16%|█▌        | 1558/9864 [6:29:55<33:10:12, 14.38s/it]
 16%|█▌        | 1559/9864 [6:30:11<34:03:49, 14.77s/it]
 16%|█▌        | 1560/9864 [6:30:28<36:03:41, 15.63s/it]
                                                        
{'loss': '0.1572', 'grad_norm': '6.519', 'learning_rate': '1.915e-05', 'epoch': '0.1582'}

 16%|█▌        | 1560/9864 [6:30:28<36:03:41, 15.63s/it]
 16%|█▌        | 1561/9864 [6:30:43<35:17:02, 15.30s/it]
 16%|█▌        | 1562/9864 [6:30:58<35:30:32, 15.40s/it]
 16%|█▌        | 1563/9864 [6:31:13<34:42:23, 15.05s/it]
 16%|█▌        | 1564/9864 [6:31:27<34:01:34, 14.76s/it]
 16%|█▌        | 1565/9864 [6:31:40<33:16:55, 14.44s/it]
 16%|█▌        | 1566/9864 [6:31:55<33:05:53, 14.36s/it]
 16%|█▌        | 1567/9864 [6:32:10<33:49:34, 14.68s/it]
 16%|█▌        | 1568/9864 [6:32:27<35:45:32, 15.52s/it]
 16%|█▌        | 1569/9864 [6:32:42<34:52:26, 15.14s/it]
 16%|█▌        | 1570/9864 [6:32:57<35:18:03, 15.32s/it]
                                                        
{'loss': '0.164', 'grad_norm': '5.83', 'learning_rate': '1.914e-05', 'epoch': '0.1592'}

 16%|█▌        | 1570/9864 [6:32:57<35:18:03, 15.32s/it]
 16%|█▌        | 1571/9864 [6:33:12<34:34:24, 15.01s/it]
 16%|█▌        | 1572/9864 [6:33:26<33:55:15, 14.73s/it]
 16%|█▌        | 1573/9864 [6:33:40<33:45:30, 14.66s/it]
 16%|█▌        | 1574/9864 [6:33:54<33:15:32, 14.44s/it]
 16%|█▌        | 1575/9864 [6:34:09<33:46:53, 14.67s/it]
 16%|█▌        | 1576/9864 [6:34:27<35:43:36, 15.52s/it]
 16%|█▌        | 1577/9864 [6:34:43<36:17:29, 15.77s/it]
 16%|█▌        | 1578/9864 [6:34:57<35:12:16, 15.30s/it]
 16%|█▌        | 1579/9864 [6:35:11<34:16:32, 14.89s/it]
 16%|█▌        | 1580/9864 [6:35:26<33:45:22, 14.67s/it]
                                                        
{'loss': '0.1591', 'grad_norm': '5.105', 'learning_rate': '1.913e-05', 'epoch': '0.1602'}

 16%|█▌        | 1580/9864 [6:35:26<33:45:22, 14.67s/it]
 16%|█▌        | 1581/9864 [6:35:40<33:31:54, 14.57s/it]
 16%|█▌        | 1582/9864 [6:35:56<34:30:40, 15.00s/it]
 16%|█▌        | 1583/9864 [6:36:12<34:56:42, 15.19s/it]
 16%|█▌        | 1584/9864 [6:36:27<35:17:56, 15.35s/it]
 16%|█▌        | 1585/9864 [6:36:43<35:41:34, 15.52s/it]
 16%|█▌        | 1586/9864 [6:36:57<34:45:39, 15.12s/it]
 16%|█▌        | 1587/9864 [6:37:12<34:22:54, 14.95s/it]
 16%|█▌        | 1588/9864 [6:37:26<33:56:43, 14.77s/it]
 16%|█▌        | 1589/9864 [6:37:41<33:36:02, 14.62s/it]
 16%|█▌        | 1590/9864 [6:37:56<33:56:37, 14.77s/it]
                                                        
{'loss': '0.1796', 'grad_norm': '2.474', 'learning_rate': '1.911e-05', 'epoch': '0.1612'}

 16%|█▌        | 1590/9864 [6:37:56<33:56:37, 14.77s/it]
 16%|█▌        | 1591/9864 [6:38:10<33:35:23, 14.62s/it]
 16%|█▌        | 1592/9864 [6:38:29<36:37:21, 15.94s/it]
 16%|█▌        | 1593/9864 [6:38:44<35:52:05, 15.61s/it]
 16%|█▌        | 1594/9864 [6:38:58<34:47:30, 15.15s/it]
 16%|█▌        | 1595/9864 [6:39:12<34:09:31, 14.87s/it]
 16%|█▌        | 1596/9864 [6:39:26<33:31:20, 14.60s/it]
 16%|█▌        | 1597/9864 [6:39:40<33:16:13, 14.49s/it]
 16%|█▌        | 1598/9864 [6:39:54<33:03:24, 14.40s/it]
 16%|█▌        | 1599/9864 [6:40:14<36:33:30, 15.92s/it]
 16%|█▌        | 1600/9864 [6:40:28<35:01:39, 15.26s/it]
                                                        
{'loss': '0.161', 'grad_norm': '3.146', 'learning_rate': '1.91e-05', 'epoch': '0.1622'}

 16%|█▌        | 1600/9864 [6:40:28<35:01:39, 15.26s/it]
 16%|█▌        | 1601/9864 [6:40:43<35:14:23, 15.35s/it]
 16%|█▌        | 1602/9864 [6:40:57<34:18:56, 14.95s/it]
 16%|█▋        | 1603/9864 [6:41:12<34:03:03, 14.84s/it]
 16%|█▋        | 1604/9864 [6:41:26<33:20:28, 14.53s/it]
 16%|█▋        | 1605/9864 [6:41:40<33:13:24, 14.48s/it]
 16%|█▋        | 1606/9864 [6:41:55<33:26:25, 14.58s/it]
 16%|█▋        | 1607/9864 [6:42:12<35:16:36, 15.38s/it]
 16%|█▋        | 1608/9864 [6:42:28<35:45:48, 15.59s/it]
 16%|█▋        | 1609/9864 [6:42:42<34:43:15, 15.14s/it]
 16%|█▋        | 1610/9864 [6:42:57<34:28:20, 15.04s/it]
                                                        
{'loss': '0.1645', 'grad_norm': '1.862', 'learning_rate': '1.909e-05', 'epoch': '0.1632'}

 16%|█▋        | 1610/9864 [6:42:57<34:28:20, 15.04s/it]
 16%|█▋        | 1611/9864 [6:43:11<33:36:41, 14.66s/it]
 16%|█▋        | 1612/9864 [6:43:25<33:04:31, 14.43s/it]
 16%|█▋        | 1613/9864 [6:43:40<34:01:13, 14.84s/it]
 16%|█▋        | 1614/9864 [6:43:56<34:49:14, 15.19s/it]
 16%|█▋        | 1615/9864 [6:44:12<35:10:32, 15.35s/it]
 16%|█▋        | 1616/9864 [6:44:27<35:02:26, 15.29s/it]
 16%|█▋        | 1617/9864 [6:44:42<34:15:32, 14.95s/it]
 16%|█▋        | 1618/9864 [6:44:55<33:28:31, 14.61s/it]
 16%|█▋        | 1619/9864 [6:45:09<32:58:25, 14.40s/it]
 16%|█▋        | 1620/9864 [6:45:23<32:41:53, 14.28s/it]
                                                        
{'loss': '0.1624', 'grad_norm': '4.69', 'learning_rate': '1.907e-05', 'epoch': '0.1642'}

 16%|█▋        | 1620/9864 [6:45:23<32:41:53, 14.28s/it]
 16%|█▋        | 1621/9864 [6:45:40<34:42:20, 15.16s/it]
 16%|█▋        | 1622/9864 [6:45:54<33:48:09, 14.76s/it]
 16%|█▋        | 1623/9864 [6:46:09<33:25:01, 14.60s/it]
 16%|█▋        | 1624/9864 [6:46:26<35:07:05, 15.34s/it]
 16%|█▋        | 1625/9864 [6:46:39<34:01:36, 14.87s/it]
 16%|█▋        | 1626/9864 [6:46:54<33:47:22, 14.77s/it]
 16%|█▋        | 1627/9864 [6:47:08<33:11:56, 14.51s/it]
 17%|█▋        | 1628/9864 [6:47:21<32:34:40, 14.24s/it]
 17%|█▋        | 1629/9864 [6:47:39<34:42:42, 15.17s/it]
 17%|█▋        | 1630/9864 [6:47:53<33:45:58, 14.76s/it]
                                                        
{'loss': '0.1436', 'grad_norm': '3.018', 'learning_rate': '1.906e-05', 'epoch': '0.1653'}

 17%|█▋        | 1630/9864 [6:47:53<33:45:58, 14.76s/it]
 17%|█▋        | 1631/9864 [6:48:06<33:01:49, 14.44s/it]
 17%|█▋        | 1632/9864 [6:48:24<35:04:01, 15.34s/it]
 17%|█▋        | 1633/9864 [6:48:38<34:23:54, 15.04s/it]
 17%|█▋        | 1634/9864 [6:48:52<33:38:32, 14.72s/it]
 17%|█▋        | 1635/9864 [6:49:06<33:11:38, 14.52s/it]
 17%|█▋        | 1636/9864 [6:49:20<33:05:19, 14.48s/it]
 17%|█▋        | 1637/9864 [6:49:36<33:59:52, 14.88s/it]
 17%|█▋        | 1638/9864 [6:49:52<34:29:14, 15.09s/it]
 17%|█▋        | 1639/9864 [6:50:07<34:46:26, 15.22s/it]
 17%|█▋        | 1640/9864 [6:50:23<35:23:04, 15.49s/it]
                                                        
{'loss': '0.1688', 'grad_norm': '3.981', 'learning_rate': '1.904e-05', 'epoch': '0.1663'}

 17%|█▋        | 1640/9864 [6:50:23<35:23:04, 15.49s/it]
 17%|█▋        | 1641/9864 [6:50:37<34:19:39, 15.03s/it]
 17%|█▋        | 1642/9864 [6:50:51<33:38:28, 14.73s/it]
 17%|█▋        | 1643/9864 [6:51:06<33:29:52, 14.67s/it]
 17%|█▋        | 1644/9864 [6:51:19<32:42:05, 14.32s/it]
 17%|█▋        | 1645/9864 [6:51:36<34:31:19, 15.12s/it]
 17%|█▋        | 1646/9864 [6:51:50<33:45:18, 14.79s/it]
 17%|█▋        | 1647/9864 [6:52:08<35:21:58, 15.49s/it]
 17%|█▋        | 1648/9864 [6:52:22<34:24:03, 15.07s/it]
 17%|█▋        | 1649/9864 [6:52:36<33:39:47, 14.75s/it]
 17%|█▋        | 1650/9864 [6:52:49<32:53:35, 14.42s/it]
                                                        
{'loss': '0.1649', 'grad_norm': '1.926', 'learning_rate': '1.903e-05', 'epoch': '0.1673'}

 17%|█▋        | 1650/9864 [6:52:49<32:53:35, 14.42s/it]
 17%|█▋        | 1651/9864 [6:53:03<32:30:06, 14.25s/it]
 17%|█▋        | 1652/9864 [6:53:21<34:41:50, 15.21s/it]
 17%|█▋        | 1653/9864 [6:53:35<33:56:17, 14.88s/it]
 17%|█▋        | 1654/9864 [6:53:49<33:35:32, 14.73s/it]
 17%|█▋        | 1655/9864 [6:54:05<34:02:54, 14.93s/it]
 17%|█▋        | 1656/9864 [6:54:20<34:42:08, 15.22s/it]
 17%|█▋        | 1657/9864 [6:54:34<33:47:50, 14.83s/it]
 17%|█▋        | 1658/9864 [6:54:49<33:32:46, 14.72s/it]
 17%|█▋        | 1659/9864 [6:55:03<33:13:54, 14.58s/it]
 17%|█▋        | 1660/9864 [6:55:20<34:52:06, 15.30s/it]
                                                        
{'loss': '0.1772', 'grad_norm': '3.811', 'learning_rate': '1.902e-05', 'epoch': '0.1683'}

 17%|█▋        | 1660/9864 [6:55:20<34:52:06, 15.30s/it]
 17%|█▋        | 1661/9864 [6:55:34<34:05:43, 14.96s/it]
 17%|█▋        | 1662/9864 [6:55:48<33:09:39, 14.55s/it]
 17%|█▋        | 1663/9864 [6:56:03<33:42:05, 14.79s/it]
 17%|█▋        | 1664/9864 [6:56:17<33:14:47, 14.60s/it]
 17%|█▋        | 1665/9864 [6:56:33<33:47:40, 14.84s/it]
 17%|█▋        | 1666/9864 [6:56:47<33:18:06, 14.62s/it]
 17%|█▋        | 1667/9864 [6:57:01<32:57:43, 14.48s/it]
 17%|█▋        | 1668/9864 [6:57:18<34:50:31, 15.30s/it]
 17%|█▋        | 1669/9864 [6:57:32<33:35:13, 14.75s/it]
 17%|█▋        | 1670/9864 [6:57:46<33:18:10, 14.63s/it]
                                                        
{'loss': '0.1754', 'grad_norm': '4.403', 'learning_rate': '1.9e-05', 'epoch': '0.1693'}

 17%|█▋        | 1670/9864 [6:57:46<33:18:10, 14.63s/it]
 17%|█▋        | 1671/9864 [6:58:02<33:56:58, 14.92s/it]
 17%|█▋        | 1672/9864 [6:58:17<34:23:34, 15.11s/it]
 17%|█▋        | 1673/9864 [6:58:31<33:48:42, 14.86s/it]
 17%|█▋        | 1674/9864 [6:58:46<33:24:55, 14.69s/it]
 17%|█▋        | 1675/9864 [6:59:03<35:20:37, 15.54s/it]
 17%|█▋        | 1676/9864 [6:59:17<34:23:48, 15.12s/it]
 17%|█▋        | 1677/9864 [6:59:32<33:42:28, 14.82s/it]
 17%|█▋        | 1678/9864 [6:59:46<33:07:58, 14.57s/it]
 17%|█▋        | 1679/9864 [7:00:01<33:38:10, 14.79s/it]
 17%|█▋        | 1680/9864 [7:00:15<33:06:30, 14.56s/it]
                                                        
{'loss': '0.1707', 'grad_norm': '1.899', 'learning_rate': '1.899e-05', 'epoch': '0.1703'}

 17%|█▋        | 1680/9864 [7:00:15<33:06:30, 14.56s/it]
 17%|█▋        | 1681/9864 [7:00:31<33:58:59, 14.95s/it]
 17%|█▋        | 1682/9864 [7:00:44<33:06:07, 14.56s/it]
 17%|█▋        | 1683/9864 [7:01:00<34:04:33, 14.99s/it]
 17%|█▋        | 1684/9864 [7:01:16<34:47:11, 15.31s/it]
 17%|█▋        | 1685/9864 [7:01:30<33:54:19, 14.92s/it]
 17%|█▋        | 1686/9864 [7:01:44<33:10:03, 14.60s/it]
 17%|█▋        | 1687/9864 [7:02:00<33:39:02, 14.81s/it]
 17%|█▋        | 1688/9864 [7:02:13<33:00:15, 14.53s/it]
 17%|█▋        | 1689/9864 [7:02:29<33:57:46, 14.96s/it]
 17%|█▋        | 1690/9864 [7:02:44<33:41:01, 14.84s/it]
                                                        
{'loss': '0.1549', 'grad_norm': '2.249', 'learning_rate': '1.897e-05', 'epoch': '0.1713'}

 17%|█▋        | 1690/9864 [7:02:44<33:41:01, 14.84s/it]
 17%|█▋        | 1691/9864 [7:02:59<34:01:13, 14.99s/it]
 17%|█▋        | 1692/9864 [7:03:15<34:18:53, 15.12s/it]
 17%|█▋        | 1693/9864 [7:03:29<33:45:51, 14.88s/it]
 17%|█▋        | 1694/9864 [7:03:44<33:30:09, 14.76s/it]
 17%|█▋        | 1695/9864 [7:03:59<34:03:12, 15.01s/it]
 17%|█▋        | 1696/9864 [7:04:15<34:25:28, 15.17s/it]
 17%|█▋        | 1697/9864 [7:04:29<33:54:12, 14.94s/it]
 17%|█▋        | 1698/9864 [7:04:43<33:31:15, 14.78s/it]
 17%|█▋        | 1699/9864 [7:04:59<34:17:12, 15.12s/it]
 17%|█▋        | 1700/9864 [7:05:15<34:33:17, 15.24s/it]
                                                        
{'loss': '0.1658', 'grad_norm': '2.413', 'learning_rate': '1.896e-05', 'epoch': '0.1724'}

 17%|█▋        | 1700/9864 [7:05:15<34:33:17, 15.24s/it]
 17%|█▋        | 1701/9864 [7:05:29<33:56:04, 14.97s/it]
 17%|█▋        | 1702/9864 [7:05:43<33:06:40, 14.60s/it]
 17%|█▋        | 1703/9864 [7:05:59<34:14:54, 15.11s/it]
 17%|█▋        | 1704/9864 [7:06:13<33:28:54, 14.77s/it]
 17%|█▋        | 1705/9864 [7:06:29<34:10:57, 15.08s/it]
 17%|█▋        | 1706/9864 [7:06:45<34:34:46, 15.26s/it]
 17%|█▋        | 1707/9864 [7:07:00<34:38:02, 15.29s/it]
 17%|█▋        | 1708/9864 [7:07:14<34:00:14, 15.01s/it]
 17%|█▋        | 1709/9864 [7:07:28<33:11:35, 14.65s/it]
 17%|█▋        | 1710/9864 [7:07:42<32:45:19, 14.46s/it]
                                                        
{'loss': '0.1669', 'grad_norm': '2.567', 'learning_rate': '1.894e-05', 'epoch': '0.1734'}

 17%|█▋        | 1710/9864 [7:07:42<32:45:19, 14.46s/it]
 17%|█▋        | 1711/9864 [7:07:58<33:35:42, 14.83s/it]
 17%|█▋        | 1712/9864 [7:08:14<34:03:35, 15.04s/it]
 17%|█▋        | 1713/9864 [7:08:27<33:09:37, 14.65s/it]
 17%|█▋        | 1714/9864 [7:08:43<33:54:21, 14.98s/it]
 17%|█▋        | 1715/9864 [7:08:59<34:43:42, 15.34s/it]
 17%|█▋        | 1716/9864 [7:09:14<34:04:33, 15.06s/it]
 17%|█▋        | 1717/9864 [7:09:27<33:07:52, 14.64s/it]
 17%|█▋        | 1718/9864 [7:09:43<33:46:32, 14.93s/it]
 17%|█▋        | 1719/9864 [7:09:59<34:29:19, 15.24s/it]
 17%|█▋        | 1720/9864 [7:10:13<33:51:49, 14.97s/it]
                                                        
{'loss': '0.174', 'grad_norm': '1.597', 'learning_rate': '1.893e-05', 'epoch': '0.1744'}

 17%|█▋        | 1720/9864 [7:10:13<33:51:49, 14.97s/it]
 17%|█▋        | 1721/9864 [7:10:27<33:00:03, 14.59s/it]
 17%|█▋        | 1722/9864 [7:10:43<33:53:59, 14.99s/it]
 17%|█▋        | 1723/9864 [7:10:57<33:08:43, 14.66s/it]
 17%|█▋        | 1724/9864 [7:11:12<33:47:18, 14.94s/it]
 17%|█▋        | 1725/9864 [7:11:26<32:54:52, 14.56s/it]
 17%|█▋        | 1726/9864 [7:11:42<33:56:48, 15.02s/it]
 18%|█▊        | 1727/9864 [7:11:57<34:10:43, 15.12s/it]
 18%|█▊        | 1728/9864 [7:12:12<33:49:03, 14.96s/it]
 18%|█▊        | 1729/9864 [7:12:25<32:46:45, 14.51s/it]
 18%|█▊        | 1730/9864 [7:12:41<33:30:50, 14.83s/it]
                                                        
{'loss': '0.1682', 'grad_norm': '3.327', 'learning_rate': '1.891e-05', 'epoch': '0.1754'}

 18%|█▊        | 1730/9864 [7:12:41<33:30:50, 14.83s/it]
 18%|█▊        | 1731/9864 [7:12:55<32:41:45, 14.47s/it]
 18%|█▊        | 1732/9864 [7:13:11<33:44:40, 14.94s/it]
 18%|█▊        | 1733/9864 [7:13:25<33:09:45, 14.68s/it]
 18%|█▊        | 1734/9864 [7:13:40<33:48:40, 14.97s/it]
 18%|█▊        | 1735/9864 [7:13:54<32:51:39, 14.55s/it]
 18%|█▊        | 1736/9864 [7:14:08<32:14:06, 14.28s/it]
 18%|█▊        | 1737/9864 [7:14:21<31:54:06, 14.13s/it]
 18%|█▊        | 1738/9864 [7:14:37<32:55:21, 14.59s/it]
 18%|█▊        | 1739/9864 [7:14:52<32:51:38, 14.56s/it]
 18%|█▊        | 1740/9864 [7:15:08<34:16:29, 15.19s/it]
                                                        
{'loss': '0.1575', 'grad_norm': '4.122', 'learning_rate': '1.89e-05', 'epoch': '0.1764'}

 18%|█▊        | 1740/9864 [7:15:08<34:16:29, 15.19s/it]
 18%|█▊        | 1741/9864 [7:15:22<33:36:32, 14.90s/it]
 18%|█▊        | 1742/9864 [7:15:40<35:37:57, 15.79s/it]
 18%|█▊        | 1743/9864 [7:15:54<34:27:14, 15.27s/it]
 18%|█▊        | 1744/9864 [7:16:09<33:53:41, 15.03s/it]
 18%|█▊        | 1745/9864 [7:16:23<33:17:00, 14.76s/it]
 18%|█▊        | 1746/9864 [7:16:39<33:53:12, 15.03s/it]
 18%|█▊        | 1747/9864 [7:16:53<33:09:44, 14.71s/it]
 18%|█▊        | 1748/9864 [7:17:08<33:25:21, 14.83s/it]
 18%|█▊        | 1749/9864 [7:17:22<32:54:25, 14.60s/it]
 18%|█▊        | 1750/9864 [7:17:39<34:57:19, 15.51s/it]
                                                        
{'loss': '0.1741', 'grad_norm': '2.795', 'learning_rate': '1.888e-05', 'epoch': '0.1774'}

 18%|█▊        | 1750/9864 [7:17:39<34:57:19, 15.51s/it]
 18%|█▊        | 1751/9864 [7:17:53<33:57:49, 15.07s/it]
 18%|█▊        | 1752/9864 [7:18:07<33:05:01, 14.68s/it]
 18%|█▊        | 1753/9864 [7:18:21<32:43:07, 14.52s/it]
 18%|█▊        | 1754/9864 [7:18:37<33:33:40, 14.90s/it]
 18%|█▊        | 1755/9864 [7:18:53<34:06:05, 15.14s/it]
 18%|█▊        | 1756/9864 [7:19:07<33:34:59, 14.91s/it]
 18%|█▊        | 1757/9864 [7:19:21<32:45:22, 14.55s/it]
 18%|█▊        | 1758/9864 [7:19:38<34:45:03, 15.43s/it]
 18%|█▊        | 1759/9864 [7:19:53<33:56:17, 15.07s/it]
 18%|█▊        | 1760/9864 [7:20:06<32:53:26, 14.61s/it]
                                                        
{'loss': '0.1748', 'grad_norm': '4.459', 'learning_rate': '1.887e-05', 'epoch': '0.1784'}

 18%|█▊        | 1760/9864 [7:20:06<32:53:26, 14.61s/it]
 18%|█▊        | 1761/9864 [7:20:20<32:14:31, 14.32s/it]
 18%|█▊        | 1762/9864 [7:20:36<33:12:43, 14.76s/it]
 18%|█▊        | 1763/9864 [7:20:51<33:41:21, 14.97s/it]
 18%|█▊        | 1764/9864 [7:21:05<32:57:12, 14.65s/it]
 18%|█▊        | 1765/9864 [7:21:20<33:30:44, 14.90s/it]
 18%|█▊        | 1766/9864 [7:21:35<33:08:55, 14.74s/it]
 18%|█▊        | 1767/9864 [7:21:51<33:54:41, 15.08s/it]
 18%|█▊        | 1768/9864 [7:22:05<33:04:28, 14.71s/it]
 18%|█▊        | 1769/9864 [7:22:18<32:28:59, 14.45s/it]
 18%|█▊        | 1770/9864 [7:22:36<34:28:41, 15.34s/it]
                                                        
{'loss': '0.1445', 'grad_norm': '3.754', 'learning_rate': '1.885e-05', 'epoch': '0.1795'}

 18%|█▊        | 1770/9864 [7:22:36<34:28:41, 15.34s/it]
 18%|█▊        | 1771/9864 [7:22:50<33:56:47, 15.10s/it]
 18%|█▊        | 1772/9864 [7:23:04<33:11:58, 14.77s/it]
 18%|█▊        | 1773/9864 [7:23:20<33:52:41, 15.07s/it]
 18%|█▊        | 1774/9864 [7:23:34<33:03:09, 14.71s/it]
 18%|█▊        | 1775/9864 [7:23:50<33:37:23, 14.96s/it]
 18%|█▊        | 1776/9864 [7:24:04<33:10:26, 14.77s/it]
 18%|█▊        | 1777/9864 [7:24:21<35:01:44, 15.59s/it]
 18%|█▊        | 1778/9864 [7:24:36<34:05:53, 15.18s/it]
 18%|█▊        | 1779/9864 [7:24:49<33:05:37, 14.74s/it]
 18%|█▊        | 1780/9864 [7:25:03<32:37:51, 14.53s/it]
                                                        
{'loss': '0.1467', 'grad_norm': '1.943', 'learning_rate': '1.884e-05', 'epoch': '0.1805'}

 18%|█▊        | 1780/9864 [7:25:03<32:37:51, 14.53s/it]
 18%|█▊        | 1781/9864 [7:25:19<33:20:51, 14.85s/it]
 18%|█▊        | 1782/9864 [7:25:33<32:38:01, 14.54s/it]
 18%|█▊        | 1783/9864 [7:25:49<33:35:22, 14.96s/it]
 18%|█▊        | 1784/9864 [7:26:03<33:02:59, 14.73s/it]
 18%|█▊        | 1785/9864 [7:26:19<33:46:30, 15.05s/it]
 18%|█▊        | 1786/9864 [7:26:34<34:17:05, 15.28s/it]
 18%|█▊        | 1787/9864 [7:26:49<33:38:18, 14.99s/it]
 18%|█▊        | 1788/9864 [7:27:03<33:07:13, 14.76s/it]
 18%|█▊        | 1789/9864 [7:27:19<33:38:28, 15.00s/it]
 18%|█▊        | 1790/9864 [7:27:34<34:05:48, 15.20s/it]
                                                        
{'loss': '0.164', 'grad_norm': '5.04', 'learning_rate': '1.882e-05', 'epoch': '0.1815'}

 18%|█▊        | 1790/9864 [7:27:34<34:05:48, 15.20s/it]
 18%|█▊        | 1791/9864 [7:27:49<33:30:32, 14.94s/it]
 18%|█▊        | 1792/9864 [7:28:03<33:17:16, 14.85s/it]
 18%|█▊        | 1793/9864 [7:28:19<34:00:57, 15.17s/it]
 18%|█▊        | 1794/9864 [7:28:35<34:09:24, 15.24s/it]
 18%|█▊        | 1795/9864 [7:28:48<33:17:56, 14.86s/it]
 18%|█▊        | 1796/9864 [7:29:04<33:54:29, 15.13s/it]
 18%|█▊        | 1797/9864 [7:29:19<33:21:54, 14.89s/it]
 18%|█▊        | 1798/9864 [7:29:35<34:12:55, 15.27s/it]
 18%|█▊        | 1799/9864 [7:29:49<33:17:03, 14.86s/it]
 18%|█▊        | 1800/9864 [7:30:03<33:02:12, 14.75s/it]
                                                        
{'loss': '0.1804', 'grad_norm': '5.832', 'learning_rate': '1.881e-05', 'epoch': '0.1825'}

 18%|█▊        | 1800/9864 [7:30:03<33:02:12, 14.75s/it]
 18%|█▊        | 1801/9864 [7:30:20<34:46:39, 15.53s/it]
 18%|█▊        | 1802/9864 [7:30:35<33:48:46, 15.10s/it]
 18%|█▊        | 1803/9864 [7:30:48<32:55:12, 14.70s/it]
 18%|█▊        | 1804/9864 [7:31:03<32:38:07, 14.58s/it]
 18%|█▊        | 1805/9864 [7:31:19<34:03:20, 15.21s/it]
 18%|█▊        | 1806/9864 [7:31:34<33:55:43, 15.16s/it]
 18%|█▊        | 1807/9864 [7:31:53<35:56:02, 16.06s/it]
 18%|█▊        | 1808/9864 [7:32:06<34:20:03, 15.34s/it]
 18%|█▊        | 1809/9864 [7:32:20<33:20:27, 14.90s/it]
 18%|█▊        | 1810/9864 [7:32:40<36:32:51, 16.34s/it]
                                                        
{'loss': '0.1557', 'grad_norm': '2.055', 'learning_rate': '1.879e-05', 'epoch': '0.1835'}

 18%|█▊        | 1810/9864 [7:32:40<36:32:51, 16.34s/it]
 18%|█▊        | 1811/9864 [7:32:54<35:14:16, 15.75s/it]
 18%|█▊        | 1812/9864 [7:33:11<35:56:50, 16.07s/it]
 18%|█▊        | 1813/9864 [7:33:25<34:19:44, 15.35s/it]
 18%|█▊        | 1814/9864 [7:33:41<35:04:51, 15.69s/it]
 18%|█▊        | 1815/9864 [7:33:59<36:14:07, 16.21s/it]
 18%|█▊        | 1816/9864 [7:34:13<35:13:53, 15.76s/it]
 18%|█▊        | 1817/9864 [7:34:32<37:17:06, 16.68s/it]
 18%|█▊        | 1818/9864 [7:34:46<35:38:58, 15.95s/it]
 18%|█▊        | 1819/9864 [7:35:02<35:21:19, 15.82s/it]
 18%|█▊        | 1820/9864 [7:35:18<35:27:43, 15.87s/it]
                                                        
{'loss': '0.1721', 'grad_norm': '2.902', 'learning_rate': '1.878e-05', 'epoch': '0.1845'}

 18%|█▊        | 1820/9864 [7:35:18<35:27:43, 15.87s/it]
 18%|█▊        | 1821/9864 [7:35:32<34:15:17, 15.33s/it]
 18%|█▊        | 1822/9864 [7:35:50<36:05:03, 16.15s/it]
 18%|█▊        | 1823/9864 [7:36:04<34:44:09, 15.55s/it]
 18%|█▊        | 1824/9864 [7:36:18<33:24:13, 14.96s/it]
 19%|█▊        | 1825/9864 [7:36:36<35:26:05, 15.87s/it]
 19%|█▊        | 1826/9864 [7:36:51<35:24:11, 15.86s/it]
 19%|█▊        | 1827/9864 [7:37:08<35:43:18, 16.00s/it]
 19%|█▊        | 1828/9864 [7:37:23<34:55:24, 15.65s/it]
 19%|█▊        | 1829/9864 [7:37:39<35:04:30, 15.72s/it]
 19%|█▊        | 1830/9864 [7:37:53<34:07:25, 15.29s/it]
                                                        
{'loss': '0.1622', 'grad_norm': '3.652', 'learning_rate': '1.876e-05', 'epoch': '0.1855'}

 19%|█▊        | 1830/9864 [7:37:53<34:07:25, 15.29s/it]
 19%|█▊        | 1831/9864 [7:38:07<33:07:04, 14.84s/it]
 19%|█▊        | 1832/9864 [7:38:21<32:42:51, 14.66s/it]
 19%|█▊        | 1833/9864 [7:38:39<34:45:10, 15.58s/it]
 19%|█▊        | 1834/9864 [7:38:53<33:46:25, 15.14s/it]
 19%|█▊        | 1835/9864 [7:39:09<34:19:22, 15.39s/it]
 19%|█▊        | 1836/9864 [7:39:24<34:21:27, 15.41s/it]
 19%|█▊        | 1837/9864 [7:39:38<33:38:49, 15.09s/it]
 19%|█▊        | 1838/9864 [7:39:53<33:08:46, 14.87s/it]
 19%|█▊        | 1839/9864 [7:40:07<32:30:26, 14.58s/it]
 19%|█▊        | 1840/9864 [7:40:21<32:11:50, 14.45s/it]
                                                        
{'loss': '0.144', 'grad_norm': '2.174', 'learning_rate': '1.874e-05', 'epoch': '0.1866'}

 19%|█▊        | 1840/9864 [7:40:21<32:11:50, 14.45s/it]
 19%|█▊        | 1841/9864 [7:40:38<34:16:57, 15.38s/it]
 19%|█▊        | 1842/9864 [7:40:53<33:23:54, 14.99s/it]
 19%|█▊        | 1843/9864 [7:41:08<33:44:41, 15.15s/it]
 19%|█▊        | 1844/9864 [7:41:22<33:14:09, 14.92s/it]
 19%|█▊        | 1845/9864 [7:41:39<34:12:17, 15.36s/it]
 19%|█▊        | 1846/9864 [7:41:55<34:44:07, 15.60s/it]
 19%|█▊        | 1847/9864 [7:42:11<35:17:40, 15.85s/it]
 19%|█▊        | 1848/9864 [7:42:27<35:11:19, 15.80s/it]
 19%|█▊        | 1849/9864 [7:42:42<34:52:14, 15.66s/it]
 19%|█▉        | 1850/9864 [7:43:01<36:34:29, 16.43s/it]
                                                        
{'loss': '0.1588', 'grad_norm': '1.425', 'learning_rate': '1.873e-05', 'epoch': '0.1876'}

 19%|█▉        | 1850/9864 [7:43:01<36:34:29, 16.43s/it]
 19%|█▉        | 1851/9864 [7:43:19<37:51:39, 17.01s/it]
 19%|█▉        | 1852/9864 [7:43:36<38:10:29, 17.15s/it]
 19%|█▉        | 1853/9864 [7:43:54<38:35:19, 17.34s/it]
 19%|█▉        | 1854/9864 [7:44:09<37:09:51, 16.70s/it]
 19%|█▉        | 1855/9864 [7:44:26<37:18:50, 16.77s/it]
 19%|█▉        | 1856/9864 [7:44:41<35:57:00, 16.16s/it]
 19%|█▉        | 1857/9864 [7:44:59<37:19:26, 16.78s/it]
 19%|█▉        | 1858/9864 [7:45:16<37:01:07, 16.65s/it]
 19%|█▉        | 1859/9864 [7:45:34<38:24:38, 17.27s/it]
 19%|█▉        | 1860/9864 [7:45:50<37:34:32, 16.90s/it]
                                                        
{'loss': '0.1697', 'grad_norm': '2.907', 'learning_rate': '1.871e-05', 'epoch': '0.1886'}

 19%|█▉        | 1860/9864 [7:45:50<37:34:32, 16.90s/it]
 19%|█▉        | 1861/9864 [7:46:06<36:49:20, 16.56s/it]
 19%|█▉        | 1862/9864 [7:46:23<36:38:52, 16.49s/it]
 19%|█▉        | 1863/9864 [7:46:40<37:24:38, 16.83s/it]
 19%|█▉        | 1864/9864 [7:46:57<37:35:47, 16.92s/it]
 19%|█▉        | 1865/9864 [7:47:12<36:19:49, 16.35s/it]
 19%|█▉        | 1866/9864 [7:47:27<35:24:25, 15.94s/it]
 19%|█▉        | 1867/9864 [7:47:47<37:35:14, 16.92s/it]
 19%|█▉        | 1868/9864 [7:48:01<35:47:23, 16.11s/it]
 19%|█▉        | 1869/9864 [7:48:16<35:24:57, 15.95s/it]
 19%|█▉        | 1870/9864 [7:48:35<36:55:22, 16.63s/it]
                                                        
{'loss': '0.1441', 'grad_norm': '3.168', 'learning_rate': '1.87e-05', 'epoch': '0.1896'}

 19%|█▉        | 1870/9864 [7:48:35<36:55:22, 16.63s/it]
 19%|█▉        | 1871/9864 [7:48:52<37:25:53, 16.86s/it]
 19%|█▉        | 1872/9864 [7:49:09<37:22:17, 16.83s/it]
 19%|█▉        | 1873/9864 [7:49:25<36:48:15, 16.58s/it]
 19%|█▉        | 1874/9864 [7:49:41<36:37:04, 16.50s/it]
 19%|█▉        | 1875/9864 [7:49:57<36:18:33, 16.36s/it]
 19%|█▉        | 1876/9864 [7:50:12<35:06:14, 15.82s/it]
 19%|█▉        | 1877/9864 [7:50:26<33:58:57, 15.32s/it]
 19%|█▉        | 1878/9864 [7:50:44<35:43:22, 16.10s/it]
 19%|█▉        | 1879/9864 [7:50:57<34:10:50, 15.41s/it]
 19%|█▉        | 1880/9864 [7:51:11<33:10:47, 14.96s/it]
                                                        
{'loss': '0.17', 'grad_norm': '2.321', 'learning_rate': '1.868e-05', 'epoch': '0.1906'}

 19%|█▉        | 1880/9864 [7:51:11<33:10:47, 14.96s/it]
 19%|█▉        | 1881/9864 [7:51:25<32:17:47, 14.56s/it]
 19%|█▉        | 1882/9864 [7:51:43<34:42:46, 15.66s/it]
 19%|█▉        | 1883/9864 [7:51:57<33:31:19, 15.12s/it]
 19%|█▉        | 1884/9864 [7:52:12<33:09:56, 14.96s/it]
 19%|█▉        | 1885/9864 [7:52:26<32:51:41, 14.83s/it]
 19%|█▉        | 1886/9864 [7:52:44<35:08:28, 15.86s/it]
 19%|█▉        | 1887/9864 [7:52:58<33:38:12, 15.18s/it]
 19%|█▉        | 1888/9864 [7:53:12<32:50:49, 14.83s/it]
 19%|█▉        | 1889/9864 [7:53:27<32:43:53, 14.78s/it]
 19%|█▉        | 1890/9864 [7:53:43<33:57:24, 15.33s/it]
                                                        
{'loss': '0.1691', 'grad_norm': '2.872', 'learning_rate': '1.866e-05', 'epoch': '0.1916'}

 19%|█▉        | 1890/9864 [7:53:43<33:57:24, 15.33s/it]
 19%|█▉        | 1891/9864 [7:54:00<34:59:36, 15.80s/it]
 19%|█▉        | 1892/9864 [7:54:15<33:58:51, 15.35s/it]
 19%|█▉        | 1893/9864 [7:54:33<35:49:27, 16.18s/it]
 19%|█▉        | 1894/9864 [7:54:47<34:46:39, 15.71s/it]
 19%|█▉        | 1895/9864 [7:55:02<34:22:23, 15.53s/it]
 19%|█▉        | 1896/9864 [7:55:17<33:40:54, 15.22s/it]
 19%|█▉        | 1897/9864 [7:55:33<34:22:19, 15.53s/it]
 19%|█▉        | 1898/9864 [7:55:49<34:55:07, 15.78s/it]
 19%|█▉        | 1899/9864 [7:56:04<34:02:50, 15.39s/it]
 19%|█▉        | 1900/9864 [7:56:20<34:46:05, 15.72s/it]
                                                        
{'loss': '0.1594', 'grad_norm': '2.358', 'learning_rate': '1.865e-05', 'epoch': '0.1926'}

 19%|█▉        | 1900/9864 [7:56:20<34:46:05, 15.72s/it]
 19%|█▉        | 1901/9864 [7:56:37<35:08:02, 15.88s/it]
 19%|█▉        | 1902/9864 [7:56:51<34:13:10, 15.47s/it]
 19%|█▉        | 1903/9864 [7:57:05<33:15:22, 15.04s/it]
 19%|█▉        | 1904/9864 [7:57:19<32:36:57, 14.75s/it]
 19%|█▉        | 1905/9864 [7:57:37<34:22:18, 15.55s/it]
 19%|█▉        | 1906/9864 [7:57:51<33:40:37, 15.23s/it]
 19%|█▉        | 1907/9864 [7:58:05<32:50:40, 14.86s/it]
 19%|█▉        | 1908/9864 [7:58:21<33:34:03, 15.19s/it]
 19%|█▉        | 1909/9864 [7:58:37<33:55:46, 15.35s/it]
 19%|█▉        | 1910/9864 [7:58:51<32:58:14, 14.92s/it]
                                                        
{'loss': '0.1736', 'grad_norm': '2.794', 'learning_rate': '1.863e-05', 'epoch': '0.1936'}

 19%|█▉        | 1910/9864 [7:58:51<32:58:14, 14.92s/it]
 19%|█▉        | 1911/9864 [7:59:06<32:49:33, 14.86s/it]
 19%|█▉        | 1912/9864 [7:59:20<32:47:44, 14.85s/it]
 19%|█▉        | 1913/9864 [7:59:38<34:41:20, 15.71s/it]
 19%|█▉        | 1914/9864 [7:59:52<33:48:49, 15.31s/it]
 19%|█▉        | 1915/9864 [8:00:09<34:24:28, 15.58s/it]
 19%|█▉        | 1916/9864 [8:00:23<33:24:33, 15.13s/it]
 19%|█▉        | 1917/9864 [8:00:38<33:46:07, 15.30s/it]
 19%|█▉        | 1918/9864 [8:00:53<33:08:22, 15.01s/it]
 19%|█▉        | 1919/9864 [8:01:07<32:52:15, 14.89s/it]
 19%|█▉        | 1920/9864 [8:01:23<33:28:08, 15.17s/it]
                                                        
{'loss': '0.1547', 'grad_norm': '3.896', 'learning_rate': '1.861e-05', 'epoch': '0.1947'}

 19%|█▉        | 1920/9864 [8:01:23<33:28:08, 15.17s/it]
 19%|█▉        | 1921/9864 [8:01:40<34:22:03, 15.58s/it]
 19%|█▉        | 1922/9864 [8:01:54<33:32:26, 15.20s/it]
 19%|█▉        | 1923/9864 [8:02:09<33:21:40, 15.12s/it]
 20%|█▉        | 1924/9864 [8:02:26<34:21:43, 15.58s/it]
 20%|█▉        | 1925/9864 [8:02:42<35:01:44, 15.88s/it]
 20%|█▉        | 1926/9864 [8:02:56<33:36:56, 15.25s/it]
 20%|█▉        | 1927/9864 [8:03:13<34:29:47, 15.65s/it]
 20%|█▉        | 1928/9864 [8:03:27<33:29:10, 15.19s/it]
 20%|█▉        | 1929/9864 [8:03:43<33:59:07, 15.42s/it]
 20%|█▉        | 1930/9864 [8:03:57<33:24:31, 15.16s/it]
                                                        
{'loss': '0.1576', 'grad_norm': '3.467', 'learning_rate': '1.86e-05', 'epoch': '0.1957'}

 20%|█▉        | 1930/9864 [8:03:57<33:24:31, 15.16s/it]
 20%|█▉        | 1931/9864 [8:04:12<32:57:51, 14.96s/it]
 20%|█▉        | 1932/9864 [8:04:31<35:30:32, 16.12s/it]
 20%|█▉        | 1933/9864 [8:04:45<34:31:52, 15.67s/it]
 20%|█▉        | 1934/9864 [8:05:00<33:46:58, 15.34s/it]
 20%|█▉        | 1935/9864 [8:05:16<34:09:56, 15.51s/it]
 20%|█▉        | 1936/9864 [8:05:31<34:18:45, 15.58s/it]
 20%|█▉        | 1937/9864 [8:05:46<33:32:38, 15.23s/it]
 20%|█▉        | 1938/9864 [8:06:00<33:05:48, 15.03s/it]
 20%|█▉        | 1939/9864 [8:06:16<33:47:16, 15.35s/it]
 20%|█▉        | 1940/9864 [8:06:32<34:06:31, 15.50s/it]
                                                        
{'loss': '0.1713', 'grad_norm': '5.196', 'learning_rate': '1.858e-05', 'epoch': '0.1967'}

 20%|█▉        | 1940/9864 [8:06:32<34:06:31, 15.50s/it]
 20%|█▉        | 1941/9864 [8:06:47<33:18:33, 15.13s/it]
 20%|█▉        | 1942/9864 [8:07:01<32:50:09, 14.92s/it]
 20%|█▉        | 1943/9864 [8:07:18<33:55:11, 15.42s/it]
 20%|█▉        | 1944/9864 [8:07:31<32:50:30, 14.93s/it]
 20%|█▉        | 1945/9864 [8:07:47<33:35:24, 15.27s/it]
 20%|█▉        | 1946/9864 [8:08:01<32:45:00, 14.89s/it]
 20%|█▉        | 1947/9864 [8:08:18<33:48:29, 15.37s/it]
 20%|█▉        | 1948/9864 [8:08:34<34:09:41, 15.54s/it]
 20%|█▉        | 1949/9864 [8:08:48<33:17:51, 15.14s/it]
 20%|█▉        | 1950/9864 [8:09:02<32:31:56, 14.80s/it]
                                                        
{'loss': '0.1587', 'grad_norm': '3.148', 'learning_rate': '1.856e-05', 'epoch': '0.1977'}

 20%|█▉        | 1950/9864 [8:09:02<32:31:56, 14.80s/it]
 20%|█▉        | 1951/9864 [8:09:18<33:23:45, 15.19s/it]
 20%|█▉        | 1952/9864 [8:09:34<34:04:49, 15.51s/it]
 20%|█▉        | 1953/9864 [8:09:49<33:29:44, 15.24s/it]
 20%|█▉        | 1954/9864 [8:10:04<32:58:14, 15.01s/it]
 20%|█▉        | 1955/9864 [8:10:19<33:27:31, 15.23s/it]
 20%|█▉        | 1956/9864 [8:10:35<33:57:47, 15.46s/it]
 20%|█▉        | 1957/9864 [8:10:50<33:15:38, 15.14s/it]
 20%|█▉        | 1958/9864 [8:11:04<32:52:16, 14.97s/it]
 20%|█▉        | 1959/9864 [8:11:22<34:49:38, 15.86s/it]
 20%|█▉        | 1960/9864 [8:11:37<34:07:06, 15.54s/it]
                                                        
{'loss': '0.1402', 'grad_norm': '2.082', 'learning_rate': '1.855e-05', 'epoch': '0.1987'}

 20%|█▉        | 1960/9864 [8:11:37<34:07:06, 15.54s/it]
 20%|█▉        | 1961/9864 [8:11:51<33:07:33, 15.09s/it]
 20%|█▉        | 1962/9864 [8:12:05<32:13:32, 14.68s/it]
 20%|█▉        | 1963/9864 [8:12:21<33:05:24, 15.08s/it]
 20%|█▉        | 1964/9864 [8:12:35<32:14:52, 14.70s/it]
 20%|█▉        | 1965/9864 [8:12:51<33:12:28, 15.13s/it]
 20%|█▉        | 1966/9864 [8:13:05<32:21:52, 14.75s/it]
 20%|█▉        | 1967/9864 [8:13:23<34:35:07, 15.77s/it]
 20%|█▉        | 1968/9864 [8:13:37<33:32:25, 15.29s/it]
 20%|█▉        | 1969/9864 [8:13:51<32:52:48, 14.99s/it]
 20%|█▉        | 1970/9864 [8:14:06<32:31:48, 14.84s/it]
                                                        
{'loss': '0.1734', 'grad_norm': '2.81', 'learning_rate': '1.853e-05', 'epoch': '0.1997'}

 20%|█▉        | 1970/9864 [8:14:06<32:31:48, 14.84s/it]
 20%|█▉        | 1971/9864 [8:14:22<33:34:11, 15.31s/it]
 20%|█▉        | 1972/9864 [8:14:38<34:17:08, 15.64s/it]
 20%|██        | 1973/9864 [8:14:53<33:29:39, 15.28s/it]
 20%|██        | 1974/9864 [8:15:07<32:41:22, 14.92s/it]
 20%|██        | 1975/9864 [8:15:25<34:37:45, 15.80s/it]
 20%|██        | 1976/9864 [8:15:39<33:28:18, 15.28s/it]
 20%|██        | 1977/9864 [8:15:53<32:41:34, 14.92s/it]
 20%|██        | 1978/9864 [8:16:07<32:15:18, 14.72s/it]
 20%|██        | 1979/9864 [8:16:26<35:02:40, 16.00s/it]
 20%|██        | 1980/9864 [8:16:40<33:53:17, 15.47s/it]
                                                        
{'loss': '0.1583', 'grad_norm': '5.52', 'learning_rate': '1.851e-05', 'epoch': '0.2007'}

 20%|██        | 1980/9864 [8:16:40<33:53:17, 15.47s/it]
 20%|██        | 1981/9864 [8:16:55<33:05:37, 15.11s/it]
 20%|██        | 1982/9864 [8:17:11<33:50:42, 15.46s/it]
 20%|██        | 1983/9864 [8:17:27<34:22:36, 15.70s/it]
 20%|██        | 1984/9864 [8:17:42<33:38:46, 15.37s/it]
 20%|██        | 1985/9864 [8:17:57<33:15:13, 15.19s/it]
 20%|██        | 1986/9864 [8:18:15<35:22:50, 16.17s/it]
 20%|██        | 1987/9864 [8:18:30<34:15:53, 15.66s/it]
 20%|██        | 1988/9864 [8:18:45<33:58:22, 15.53s/it]
 20%|██        | 1989/9864 [8:19:02<34:44:23, 15.88s/it]
 20%|██        | 1990/9864 [8:19:16<33:51:18, 15.48s/it]
                                                        
{'loss': '0.1731', 'grad_norm': '4.164', 'learning_rate': '1.849e-05', 'epoch': '0.2018'}

 20%|██        | 1990/9864 [8:19:16<33:51:18, 15.48s/it]
 20%|██        | 1991/9864 [8:19:32<34:21:58, 15.71s/it]
 20%|██        | 1992/9864 [8:19:47<33:40:21, 15.40s/it]
 20%|██        | 1993/9864 [8:20:02<33:13:19, 15.19s/it]
 20%|██        | 1994/9864 [8:20:20<35:12:05, 16.10s/it]
 20%|██        | 1995/9864 [8:20:34<34:09:18, 15.63s/it]
 20%|██        | 1996/9864 [8:20:49<33:15:34, 15.22s/it]
 20%|██        | 1997/9864 [8:21:04<33:36:57, 15.38s/it]
 20%|██        | 1998/9864 [8:21:19<32:53:53, 15.06s/it]
 20%|██        | 1999/9864 [8:21:35<33:42:44, 15.43s/it]
 20%|██        | 2000/9864 [8:21:49<32:54:57, 15.07s/it]
                                                        
{'loss': '0.1567', 'grad_norm': '2.276', 'learning_rate': '1.848e-05', 'epoch': '0.2028'}

 20%|██        | 2000/9864 [8:21:49<32:54:57, 15.07s/it]

Writing model shards:   0%|          | 0/1 [00:00<?, ?it/s]

Writing model shards: 100%|██████████| 1/1 [00:13<00:00, 13.09s/it]
Writing model shards: 100%|██████████| 1/1 [00:13<00:00, 13.09s/it]
/home/dungpq6/Project/SpatialStack/src/qwen_vl/model/geometry_encoders/vggt_encoder.py:285: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(dtype=dtype):

 20%|██        | 2001/9864 [8:23:04<72:11:28, 33.05s/it]
 20%|██        | 2002/9864 [8:23:21<61:17:47, 28.07s/it]
 20%|██        | 2003/9864 [8:23:37<53:42:05, 24.59s/it]
 20%|██        | 2004/9864 [8:23:53<47:40:22, 21.83s/it]
 20%|██        | 2005/9864 [8:24:09<43:49:59, 20.08s/it]
 20%|██        | 2006/9864 [8:24:26<42:07:52, 19.30s/it]
 20%|██        | 2007/9864 [8:24:41<39:12:35, 17.97s/it]
 20%|██        | 2008/9864 [8:24:57<37:50:44, 17.34s/it]
 20%|██        | 2009/9864 [8:25:13<36:49:09, 16.87s/it]
 20%|██        | 2010/9864 [8:25:27<35:11:45, 16.13s/it]
                                                        
{'loss': '0.1574', 'grad_norm': '2.297', 'learning_rate': '1.846e-05', 'epoch': '0.2038'}

 20%|██        | 2010/9864 [8:25:27<35:11:45, 16.13s/it]
 20%|██        | 2011/9864 [8:25:42<34:06:49, 15.64s/it]
 20%|██        | 2012/9864 [8:25:55<32:56:04, 15.10s/it]
 20%|██        | 2013/9864 [8:26:13<34:53:22, 16.00s/it]
 20%|██        | 2014/9864 [8:26:28<34:09:44, 15.67s/it]
 20%|██        | 2015/9864 [8:26:43<33:37:04, 15.42s/it]
 20%|██        | 2016/9864 [8:27:00<34:44:59, 15.94s/it]
 20%|██        | 2017/9864 [8:27:17<35:20:06, 16.21s/it]
 20%|██        | 2018/9864 [8:27:32<34:31:40, 15.84s/it]
 20%|██        | 2019/9864 [8:27:47<33:40:01, 15.45s/it]
 20%|██        | 2020/9864 [8:28:03<34:15:30, 15.72s/it]
                                                        
{'loss': '0.1794', 'grad_norm': '7.117', 'learning_rate': '1.844e-05', 'epoch': '0.2048'}

 20%|██        | 2020/9864 [8:28:03<34:15:30, 15.72s/it]
 20%|██        | 2021/9864 [8:28:19<34:34:02, 15.87s/it]
 20%|██        | 2022/9864 [8:28:34<33:34:55, 15.42s/it]
 21%|██        | 2023/9864 [8:28:50<34:07:37, 15.67s/it]
 21%|██        | 2024/9864 [8:29:04<32:55:14, 15.12s/it]
 21%|██        | 2025/9864 [8:29:20<33:30:08, 15.39s/it]
 21%|██        | 2026/9864 [8:29:34<32:47:39, 15.06s/it]
 21%|██        | 2027/9864 [8:29:50<33:35:38, 15.43s/it]
 21%|██        | 2028/9864 [8:30:04<32:41:07, 15.02s/it]
 21%|██        | 2029/9864 [8:30:20<33:09:33, 15.24s/it]
 21%|██        | 2030/9864 [8:30:35<32:43:26, 15.04s/it]
                                                        
{'loss': '0.1549', 'grad_norm': '3.581', 'learning_rate': '1.842e-05', 'epoch': '0.2058'}

 21%|██        | 2030/9864 [8:30:35<32:43:26, 15.04s/it]
 21%|██        | 2031/9864 [8:30:51<33:38:31, 15.46s/it]
 21%|██        | 2032/9864 [8:31:08<34:14:43, 15.74s/it]
 21%|██        | 2033/9864 [8:31:22<33:05:25, 15.21s/it]
 21%|██        | 2034/9864 [8:31:36<32:31:26, 14.95s/it]
 21%|██        | 2035/9864 [8:31:52<33:04:09, 15.21s/it]
 21%|██        | 2036/9864 [8:32:05<32:08:36, 14.78s/it]
 21%|██        | 2037/9864 [8:32:22<33:13:44, 15.28s/it]
 21%|██        | 2038/9864 [8:32:39<34:18:11, 15.78s/it]
 21%|██        | 2039/9864 [8:32:55<34:16:39, 15.77s/it]
 21%|██        | 2040/9864 [8:33:09<33:33:05, 15.44s/it]
                                                        
{'loss': '0.1753', 'grad_norm': '2.777', 'learning_rate': '1.841e-05', 'epoch': '0.2068'}

 21%|██        | 2040/9864 [8:33:09<33:33:05, 15.44s/it]
 21%|██        | 2041/9864 [8:33:24<32:50:30, 15.11s/it]
 21%|██        | 2042/9864 [8:33:40<33:25:50, 15.39s/it]
 21%|██        | 2043/9864 [8:33:54<32:43:41, 15.06s/it]
 21%|██        | 2044/9864 [8:34:08<32:10:53, 14.82s/it]
 21%|██        | 2045/9864 [8:34:24<32:39:30, 15.04s/it]
 21%|██        | 2046/9864 [8:34:40<33:32:47, 15.45s/it]
 21%|██        | 2047/9864 [8:34:56<33:50:24, 15.58s/it]
 21%|██        | 2048/9864 [8:35:10<32:36:48, 15.02s/it]
 21%|██        | 2049/9864 [8:35:24<32:15:34, 14.86s/it]
 21%|██        | 2050/9864 [8:35:39<32:08:51, 14.81s/it]
                                                        
{'loss': '0.1679', 'grad_norm': '3.576', 'learning_rate': '1.839e-05', 'epoch': '0.2078'}

 21%|██        | 2050/9864 [8:35:39<32:08:51, 14.81s/it]
 21%|██        | 2051/9864 [8:35:55<32:51:35, 15.14s/it]
 21%|██        | 2052/9864 [8:36:09<32:06:32, 14.80s/it]
 21%|██        | 2053/9864 [8:36:25<33:14:32, 15.32s/it]
 21%|██        | 2054/9864 [8:36:42<34:17:19, 15.81s/it]
 21%|██        | 2055/9864 [8:36:59<34:38:09, 15.97s/it]
 21%|██        | 2056/9864 [8:37:13<33:27:08, 15.42s/it]
 21%|██        | 2057/9864 [8:37:27<32:37:59, 15.05s/it]
 21%|██        | 2058/9864 [8:37:41<32:04:52, 14.80s/it]
 21%|██        | 2059/9864 [8:37:57<32:50:00, 15.14s/it]
 21%|██        | 2060/9864 [8:38:12<32:29:54, 14.99s/it]
                                                        
{'loss': '0.169', 'grad_norm': '2.853', 'learning_rate': '1.837e-05', 'epoch': '0.2089'}

 21%|██        | 2060/9864 [8:38:12<32:29:54, 14.99s/it]
 21%|██        | 2061/9864 [8:38:28<33:35:00, 15.49s/it]
 21%|██        | 2062/9864 [8:38:45<34:06:52, 15.74s/it]
 21%|██        | 2063/9864 [8:39:02<34:45:22, 16.04s/it]
 21%|██        | 2064/9864 [8:39:16<33:46:32, 15.59s/it]
 21%|██        | 2065/9864 [8:39:31<33:02:41, 15.25s/it]
 21%|██        | 2066/9864 [8:39:45<32:45:33, 15.12s/it]
 21%|██        | 2067/9864 [8:40:02<33:49:33, 15.62s/it]
 21%|██        | 2068/9864 [8:40:19<34:39:30, 16.00s/it]
 21%|██        | 2069/9864 [8:40:33<33:38:50, 15.54s/it]
 21%|██        | 2070/9864 [8:40:50<33:58:41, 15.69s/it]
                                                        
{'loss': '0.1598', 'grad_norm': '5.639', 'learning_rate': '1.835e-05', 'epoch': '0.2099'}

 21%|██        | 2070/9864 [8:40:50<33:58:41, 15.69s/it]
 21%|██        | 2071/9864 [8:41:04<33:06:42, 15.30s/it]
 21%|██        | 2072/9864 [8:41:20<33:44:45, 15.59s/it]
 21%|██        | 2073/9864 [8:41:35<33:06:09, 15.30s/it]
 21%|██        | 2074/9864 [8:41:49<32:39:32, 15.09s/it]
 21%|██        | 2075/9864 [8:42:06<33:40:30, 15.56s/it]
 21%|██        | 2076/9864 [8:42:21<33:05:04, 15.29s/it]
 21%|██        | 2077/9864 [8:42:37<33:38:15, 15.55s/it]
 21%|██        | 2078/9864 [8:42:51<33:01:27, 15.27s/it]
 21%|██        | 2079/9864 [8:43:05<31:58:30, 14.79s/it]
 21%|██        | 2080/9864 [8:43:21<32:28:05, 15.02s/it]
                                                        
{'loss': '0.1622', 'grad_norm': '2.385', 'learning_rate': '1.833e-05', 'epoch': '0.2109'}

 21%|██        | 2080/9864 [8:43:21<32:28:05, 15.02s/it]
 21%|██        | 2081/9864 [8:43:35<31:53:13, 14.75s/it]
 21%|██        | 2082/9864 [8:43:49<31:27:07, 14.55s/it]
 21%|██        | 2083/9864 [8:44:05<32:28:19, 15.02s/it]
 21%|██        | 2084/9864 [8:44:21<33:06:24, 15.32s/it]
 21%|██        | 2085/9864 [8:44:37<33:33:33, 15.53s/it]
 21%|██        | 2086/9864 [8:44:51<32:33:27, 15.07s/it]
 21%|██        | 2087/9864 [8:45:06<32:16:29, 14.94s/it]
 21%|██        | 2088/9864 [8:45:20<32:07:52, 14.88s/it]
 21%|██        | 2089/9864 [8:45:37<32:58:29, 15.27s/it]
 21%|██        | 2090/9864 [8:45:51<32:04:51, 14.86s/it]
                                                        
{'loss': '0.1626', 'grad_norm': '4.645', 'learning_rate': '1.832e-05', 'epoch': '0.2119'}

 21%|██        | 2090/9864 [8:45:51<32:04:51, 14.86s/it]
 21%|██        | 2091/9864 [8:46:05<31:44:54, 14.70s/it]
 21%|██        | 2092/9864 [8:46:23<33:47:03, 15.65s/it]
 21%|██        | 2093/9864 [8:46:39<34:16:27, 15.88s/it]
 21%|██        | 2094/9864 [8:46:54<33:25:20, 15.49s/it]
 21%|██        | 2095/9864 [8:47:08<32:50:25, 15.22s/it]
 21%|██        | 2096/9864 [8:47:23<32:30:42, 15.07s/it]
 21%|██▏       | 2097/9864 [8:47:39<33:11:28, 15.38s/it]
 21%|██▏       | 2098/9864 [8:47:53<32:26:13, 15.04s/it]
 21%|██▏       | 2099/9864 [8:48:07<31:35:59, 14.65s/it]
 21%|██▏       | 2100/9864 [8:48:23<32:16:13, 14.96s/it]
                                                        
{'loss': '0.1747', 'grad_norm': '5.632', 'learning_rate': '1.83e-05', 'epoch': '0.2129'}

 21%|██▏       | 2100/9864 [8:48:23<32:16:13, 14.96s/it]
 21%|██▏       | 2101/9864 [8:48:41<34:07:15, 15.82s/it]
 21%|██▏       | 2102/9864 [8:48:55<32:51:57, 15.24s/it]
 21%|██▏       | 2103/9864 [8:49:09<32:09:51, 14.92s/it]
 21%|██▏       | 2104/9864 [8:49:25<32:54:48, 15.27s/it]
 21%|██▏       | 2105/9864 [8:49:39<32:22:24, 15.02s/it]
 21%|██▏       | 2106/9864 [8:49:53<31:46:31, 14.75s/it]
 21%|██▏       | 2107/9864 [8:50:09<32:38:03, 15.15s/it]
 21%|██▏       | 2108/9864 [8:50:23<31:44:47, 14.74s/it]
 21%|██▏       | 2109/9864 [8:50:41<33:41:05, 15.64s/it]
 21%|██▏       | 2110/9864 [8:50:55<32:43:39, 15.19s/it]
                                                        
{'loss': '0.1587', 'grad_norm': '2.35', 'learning_rate': '1.828e-05', 'epoch': '0.2139'}

 21%|██▏       | 2110/9864 [8:50:55<32:43:39, 15.19s/it]
 21%|██▏       | 2111/9864 [8:51:09<32:10:23, 14.94s/it]
 21%|██▏       | 2112/9864 [8:51:26<33:00:54, 15.33s/it]
 21%|██▏       | 2113/9864 [8:51:40<32:20:48, 15.02s/it]
 21%|██▏       | 2114/9864 [8:51:56<33:14:48, 15.44s/it]
 21%|██▏       | 2115/9864 [8:52:10<32:18:06, 15.01s/it]
 21%|██▏       | 2116/9864 [8:52:26<32:33:26, 15.13s/it]
 21%|██▏       | 2117/9864 [8:52:41<32:49:41, 15.26s/it]
 21%|██▏       | 2118/9864 [8:52:56<32:14:35, 14.99s/it]
 21%|██▏       | 2119/9864 [8:53:10<31:32:00, 14.66s/it]
 21%|██▏       | 2120/9864 [8:53:26<32:44:23, 15.22s/it]
                                                        
{'loss': '0.1572', 'grad_norm': '4.24', 'learning_rate': '1.826e-05', 'epoch': '0.2149'}

 21%|██▏       | 2120/9864 [8:53:26<32:44:23, 15.22s/it]
 22%|██▏       | 2121/9864 [8:53:40<32:11:44, 14.97s/it]
 22%|██▏       | 2122/9864 [8:53:56<32:43:45, 15.22s/it]
 22%|██▏       | 2123/9864 [8:54:11<32:10:17, 14.96s/it]
 22%|██▏       | 2124/9864 [8:54:25<31:48:57, 14.80s/it]
 22%|██▏       | 2125/9864 [8:54:42<33:02:39, 15.37s/it]
 22%|██▏       | 2126/9864 [8:54:57<32:40:01, 15.20s/it]
 22%|██▏       | 2127/9864 [8:55:13<33:37:34, 15.65s/it]
 22%|██▏       | 2128/9864 [8:55:27<32:42:16, 15.22s/it]
 22%|██▏       | 2129/9864 [8:55:44<33:23:14, 15.54s/it]
 22%|██▏       | 2130/9864 [8:55:58<32:32:25, 15.15s/it]
                                                        
{'loss': '0.1803', 'grad_norm': '6.479', 'learning_rate': '1.824e-05', 'epoch': '0.216'}

 22%|██▏       | 2130/9864 [8:55:58<32:32:25, 15.15s/it]
 22%|██▏       | 2131/9864 [8:56:12<31:51:26, 14.83s/it]
 22%|██▏       | 2132/9864 [8:56:28<32:51:29, 15.30s/it]
 22%|██▏       | 2133/9864 [8:56:44<32:56:48, 15.34s/it]
 22%|██▏       | 2134/9864 [8:56:58<32:07:02, 14.96s/it]
 22%|██▏       | 2135/9864 [8:57:12<31:42:21, 14.77s/it]
 22%|██▏       | 2136/9864 [8:57:29<32:49:39, 15.29s/it]
 22%|██▏       | 2137/9864 [8:57:45<33:28:38, 15.60s/it]
 22%|██▏       | 2138/9864 [8:58:00<32:50:16, 15.30s/it]
 22%|██▏       | 2139/9864 [8:58:13<31:47:27, 14.82s/it]
 22%|██▏       | 2140/9864 [8:58:29<32:23:02, 15.09s/it]
                                                        
{'loss': '0.1612', 'grad_norm': '2.919', 'learning_rate': '1.822e-05', 'epoch': '0.217'}

 22%|██▏       | 2140/9864 [8:58:29<32:23:02, 15.09s/it]
 22%|██▏       | 2141/9864 [8:58:46<33:12:49, 15.48s/it]
 22%|██▏       | 2142/9864 [8:59:00<32:27:29, 15.13s/it]
 22%|██▏       | 2143/9864 [8:59:15<32:26:05, 15.12s/it]
 22%|██▏       | 2144/9864 [8:59:31<33:10:40, 15.47s/it]
 22%|██▏       | 2145/9864 [8:59:47<33:39:10, 15.70s/it]
 22%|██▏       | 2146/9864 [9:00:02<33:04:45, 15.43s/it]
 22%|██▏       | 2147/9864 [9:00:18<33:32:19, 15.65s/it]
 22%|██▏       | 2148/9864 [9:00:33<32:35:22, 15.21s/it]
 22%|██▏       | 2149/9864 [9:00:49<33:02:39, 15.42s/it]
 22%|██▏       | 2150/9864 [9:01:03<32:13:07, 15.04s/it]
                                                        
{'loss': '0.1675', 'grad_norm': '5.072', 'learning_rate': '1.821e-05', 'epoch': '0.218'}

 22%|██▏       | 2150/9864 [9:01:03<32:13:07, 15.04s/it]
 22%|██▏       | 2151/9864 [9:01:17<31:38:20, 14.77s/it]
 22%|██▏       | 2152/9864 [9:01:34<33:11:23, 15.49s/it]
 22%|██▏       | 2153/9864 [9:01:48<32:10:12, 15.02s/it]
 22%|██▏       | 2154/9864 [9:02:03<31:55:56, 14.91s/it]
 22%|██▏       | 2155/9864 [9:02:17<31:51:14, 14.88s/it]
 22%|██▏       | 2156/9864 [9:02:34<32:52:03, 15.35s/it]
 22%|██▏       | 2157/9864 [9:02:50<33:12:12, 15.51s/it]
 22%|██▏       | 2158/9864 [9:03:04<32:39:48, 15.26s/it]
 22%|██▏       | 2159/9864 [9:03:20<33:09:36, 15.49s/it]
 22%|██▏       | 2160/9864 [9:03:37<33:37:32, 15.71s/it]
                                                        
{'loss': '0.175', 'grad_norm': '2.511', 'learning_rate': '1.819e-05', 'epoch': '0.219'}

 22%|██▏       | 2160/9864 [9:03:37<33:37:32, 15.71s/it]
 22%|██▏       | 2161/9864 [9:03:52<33:21:43, 15.59s/it]
 22%|██▏       | 2162/9864 [9:04:06<32:32:07, 15.21s/it]
 22%|██▏       | 2163/9864 [9:04:22<32:58:53, 15.42s/it]
 22%|██▏       | 2164/9864 [9:04:37<32:28:10, 15.18s/it]
 22%|██▏       | 2165/9864 [9:04:54<33:26:41, 15.64s/it]
 22%|██▏       | 2166/9864 [9:05:08<32:37:27, 15.26s/it]
 22%|██▏       | 2167/9864 [9:05:22<32:10:14, 15.05s/it]
 22%|██▏       | 2168/9864 [9:05:38<32:47:07, 15.34s/it]
 22%|██▏       | 2169/9864 [9:05:55<33:33:14, 15.70s/it]
 22%|██▏       | 2170/9864 [9:06:09<32:39:47, 15.28s/it]
                                                        
{'loss': '0.1739', 'grad_norm': '2.879', 'learning_rate': '1.817e-05', 'epoch': '0.22'}

 22%|██▏       | 2170/9864 [9:06:09<32:39:47, 15.28s/it]
 22%|██▏       | 2171/9864 [9:06:24<32:12:35, 15.07s/it]
 22%|██▏       | 2172/9864 [9:06:41<33:12:37, 15.54s/it]
 22%|██▏       | 2173/9864 [9:06:57<33:29:29, 15.68s/it]
 22%|██▏       | 2174/9864 [9:07:11<32:52:48, 15.39s/it]
 22%|██▏       | 2175/9864 [9:07:26<32:13:05, 15.08s/it]
 22%|██▏       | 2176/9864 [9:07:42<33:01:57, 15.47s/it]
 22%|██▏       | 2177/9864 [9:07:58<33:38:48, 15.76s/it]
 22%|██▏       | 2178/9864 [9:08:13<32:54:23, 15.41s/it]
 22%|██▏       | 2179/9864 [9:08:27<32:15:53, 15.11s/it]
 22%|██▏       | 2180/9864 [9:08:44<32:53:38, 15.41s/it]
                                                        
{'loss': '0.1594', 'grad_norm': '3.171', 'learning_rate': '1.815e-05', 'epoch': '0.221'}

 22%|██▏       | 2180/9864 [9:08:44<32:53:38, 15.41s/it]
 22%|██▏       | 2181/9864 [9:09:00<33:24:41, 15.66s/it]
 22%|██▏       | 2182/9864 [9:09:15<33:06:44, 15.52s/it]
 22%|██▏       | 2183/9864 [9:09:29<32:26:09, 15.20s/it]
 22%|██▏       | 2184/9864 [9:09:46<33:06:16, 15.52s/it]
 22%|██▏       | 2185/9864 [9:10:00<32:17:03, 15.14s/it]
 22%|██▏       | 2186/9864 [9:10:15<31:55:34, 14.97s/it]
 22%|██▏       | 2187/9864 [9:10:29<31:52:54, 14.95s/it]
 22%|██▏       | 2188/9864 [9:10:46<32:58:33, 15.47s/it]
 22%|██▏       | 2189/9864 [9:11:02<33:23:21, 15.66s/it]
 22%|██▏       | 2190/9864 [9:11:17<32:53:17, 15.43s/it]
                                                        
{'loss': '0.1561', 'grad_norm': '2.032', 'learning_rate': '1.813e-05', 'epoch': '0.222'}

 22%|██▏       | 2190/9864 [9:11:17<32:53:17, 15.43s/it]
 22%|██▏       | 2191/9864 [9:11:32<32:24:28, 15.21s/it]
 22%|██▏       | 2192/9864 [9:11:49<33:49:23, 15.87s/it]
 22%|██▏       | 2193/9864 [9:12:03<32:44:12, 15.36s/it]
 22%|██▏       | 2194/9864 [9:12:18<32:29:53, 15.25s/it]
 22%|██▏       | 2195/9864 [9:12:35<33:20:17, 15.65s/it]
 22%|██▏       | 2196/9864 [9:12:50<32:45:00, 15.38s/it]
 22%|██▏       | 2197/9864 [9:13:04<32:16:01, 15.15s/it]
 22%|██▏       | 2198/9864 [9:13:21<33:00:10, 15.50s/it]
 22%|██▏       | 2199/9864 [9:13:35<32:14:40, 15.14s/it]
 22%|██▏       | 2200/9864 [9:13:52<33:45:18, 15.86s/it]
                                                        
{'loss': '0.1642', 'grad_norm': '2.144', 'learning_rate': '1.811e-05', 'epoch': '0.2231'}

 22%|██▏       | 2200/9864 [9:13:52<33:45:18, 15.86s/it]
 22%|██▏       | 2201/9864 [9:14:07<33:00:32, 15.51s/it]
 22%|██▏       | 2202/9864 [9:14:22<32:35:41, 15.31s/it]
 22%|██▏       | 2203/9864 [9:14:38<33:07:08, 15.56s/it]
 22%|██▏       | 2204/9864 [9:14:52<31:50:25, 14.96s/it]
 22%|██▏       | 2205/9864 [9:15:09<33:15:05, 15.63s/it]
 22%|██▏       | 2206/9864 [9:15:24<32:48:35, 15.42s/it]
 22%|██▏       | 2207/9864 [9:15:42<34:27:51, 16.20s/it]
 22%|██▏       | 2208/9864 [9:15:57<33:33:43, 15.78s/it]
 22%|██▏       | 2209/9864 [9:16:11<32:44:24, 15.40s/it]
 22%|██▏       | 2210/9864 [9:16:25<32:00:32, 15.06s/it]
                                                        
{'loss': '0.1661', 'grad_norm': '3.237', 'learning_rate': '1.809e-05', 'epoch': '0.2241'}

 22%|██▏       | 2210/9864 [9:16:25<32:00:32, 15.06s/it]
 22%|██▏       | 2211/9864 [9:16:42<33:07:06, 15.58s/it]
 22%|██▏       | 2212/9864 [9:16:59<33:36:25, 15.81s/it]
 22%|██▏       | 2213/9864 [9:17:14<33:03:41, 15.56s/it]
 22%|██▏       | 2214/9864 [9:17:28<32:32:56, 15.32s/it]
 22%|██▏       | 2215/9864 [9:17:46<33:44:49, 15.88s/it]
 22%|██▏       | 2216/9864 [9:18:00<32:58:30, 15.52s/it]
 22%|██▏       | 2217/9864 [9:18:15<32:35:26, 15.34s/it]
 22%|██▏       | 2218/9864 [9:18:30<32:14:19, 15.18s/it]
 22%|██▏       | 2219/9864 [9:18:47<33:16:16, 15.67s/it]
 23%|██▎       | 2220/9864 [9:19:03<33:45:34, 15.90s/it]
                                                        
{'loss': '0.1978', 'grad_norm': '9.454', 'learning_rate': '1.807e-05', 'epoch': '0.2251'}

 23%|██▎       | 2220/9864 [9:19:03<33:45:34, 15.90s/it]
 23%|██▎       | 2221/9864 [9:19:18<33:12:03, 15.64s/it]
 23%|██▎       | 2222/9864 [9:19:35<33:43:30, 15.89s/it]
 23%|██▎       | 2223/9864 [9:19:51<34:00:45, 16.02s/it]
 23%|██▎       | 2224/9864 [9:20:06<33:11:26, 15.64s/it]
 23%|██▎       | 2225/9864 [9:20:20<32:32:47, 15.34s/it]
 23%|██▎       | 2226/9864 [9:20:35<32:08:01, 15.15s/it]
 23%|██▎       | 2227/9864 [9:20:51<32:53:28, 15.50s/it]
 23%|██▎       | 2228/9864 [9:21:08<33:34:56, 15.83s/it]
 23%|██▎       | 2229/9864 [9:21:23<32:50:21, 15.48s/it]
 23%|██▎       | 2230/9864 [9:21:41<34:22:00, 16.21s/it]
                                                        
{'loss': '0.1656', 'grad_norm': '1.389', 'learning_rate': '1.805e-05', 'epoch': '0.2261'}

 23%|██▎       | 2230/9864 [9:21:41<34:22:00, 16.21s/it]
 23%|██▎       | 2231/9864 [9:21:55<33:28:48, 15.79s/it]
 23%|██▎       | 2232/9864 [9:22:10<32:47:57, 15.47s/it]
 23%|██▎       | 2233/9864 [9:22:24<32:02:38, 15.12s/it]
 23%|██▎       | 2234/9864 [9:22:40<32:37:05, 15.39s/it]
 23%|██▎       | 2235/9864 [9:22:55<32:11:32, 15.19s/it]
 23%|██▎       | 2236/9864 [9:23:12<33:01:19, 15.58s/it]
 23%|██▎       | 2237/9864 [9:23:26<32:30:16, 15.34s/it]
 23%|██▎       | 2238/9864 [9:23:45<34:31:27, 16.30s/it]
 23%|██▎       | 2239/9864 [9:23:59<33:12:39, 15.68s/it]
 23%|██▎       | 2240/9864 [9:24:14<32:44:58, 15.46s/it]
                                                        
{'loss': '0.1689', 'grad_norm': '6.099', 'learning_rate': '1.803e-05', 'epoch': '0.2271'}

 23%|██▎       | 2240/9864 [9:24:14<32:44:58, 15.46s/it]
 23%|██▎       | 2241/9864 [9:24:28<31:54:13, 15.07s/it]
 23%|██▎       | 2242/9864 [9:24:43<31:46:44, 15.01s/it]
 23%|██▎       | 2243/9864 [9:24:59<32:31:01, 15.36s/it]
 23%|██▎       | 2244/9864 [9:25:16<33:00:28, 15.59s/it]
 23%|██▎       | 2245/9864 [9:25:32<33:22:24, 15.77s/it]
 23%|██▎       | 2246/9864 [9:25:48<34:00:14, 16.07s/it]
 23%|██▎       | 2247/9864 [9:26:03<32:52:56, 15.54s/it]
 23%|██▎       | 2248/9864 [9:26:17<32:05:25, 15.17s/it]
 23%|██▎       | 2249/9864 [9:26:32<31:40:38, 14.98s/it]
 23%|██▎       | 2250/9864 [9:26:46<31:32:12, 14.91s/it]
                                                        
{'loss': '0.1559', 'grad_norm': '2.94', 'learning_rate': '1.801e-05', 'epoch': '0.2281'}

 23%|██▎       | 2250/9864 [9:26:46<31:32:12, 14.91s/it]
 23%|██▎       | 2251/9864 [9:27:05<33:36:14, 15.89s/it]
 23%|██▎       | 2252/9864 [9:27:21<33:48:56, 15.99s/it]
 23%|██▎       | 2253/9864 [9:27:37<34:15:23, 16.20s/it]
 23%|██▎       | 2254/9864 [9:27:52<33:05:43, 15.66s/it]
 23%|██▎       | 2255/9864 [9:28:06<32:02:48, 15.16s/it]
 23%|██▎       | 2256/9864 [9:28:20<31:31:31, 14.92s/it]
 23%|██▎       | 2257/9864 [9:28:34<31:04:29, 14.71s/it]
 23%|██▎       | 2258/9864 [9:28:51<32:16:27, 15.28s/it]
 23%|██▎       | 2259/9864 [9:29:08<33:40:48, 15.94s/it]
 23%|██▎       | 2260/9864 [9:29:23<32:36:46, 15.44s/it]
                                                        
{'loss': '0.155', 'grad_norm': '2.499', 'learning_rate': '1.799e-05', 'epoch': '0.2291'}

 23%|██▎       | 2260/9864 [9:29:23<32:36:46, 15.44s/it]
 23%|██▎       | 2261/9864 [9:29:39<33:12:49, 15.73s/it]
 23%|██▎       | 2262/9864 [9:29:53<32:08:15, 15.22s/it]
 23%|██▎       | 2263/9864 [9:30:08<31:40:11, 15.00s/it]
 23%|██▎       | 2264/9864 [9:30:23<31:33:51, 14.95s/it]
 23%|██▎       | 2265/9864 [9:30:37<31:22:19, 14.86s/it]
 23%|██▎       | 2266/9864 [9:30:54<32:41:27, 15.49s/it]
 23%|██▎       | 2267/9864 [9:31:12<34:15:45, 16.24s/it]
 23%|██▎       | 2268/9864 [9:31:27<33:22:00, 15.81s/it]
 23%|██▎       | 2269/9864 [9:31:44<33:50:42, 16.04s/it]
 23%|██▎       | 2270/9864 [9:31:58<32:52:33, 15.59s/it]
                                                        
{'loss': '0.1482', 'grad_norm': '2.469', 'learning_rate': '1.797e-05', 'epoch': '0.2301'}

 23%|██▎       | 2270/9864 [9:31:58<32:52:33, 15.59s/it]
 23%|██▎       | 2271/9864 [9:32:13<32:12:21, 15.27s/it]
 23%|██▎       | 2272/9864 [9:32:27<31:36:26, 14.99s/it]
 23%|██▎       | 2273/9864 [9:32:42<31:24:14, 14.89s/it]
 23%|██▎       | 2274/9864 [9:32:58<32:30:50, 15.42s/it]
 23%|██▎       | 2275/9864 [9:33:17<34:34:11, 16.40s/it]
 23%|██▎       | 2276/9864 [9:33:31<33:08:27, 15.72s/it]
 23%|██▎       | 2277/9864 [9:33:47<33:25:17, 15.86s/it]
 23%|██▎       | 2278/9864 [9:34:02<32:42:11, 15.52s/it]
 23%|██▎       | 2279/9864 [9:34:17<32:08:33, 15.26s/it]
 23%|██▎       | 2280/9864 [9:34:31<31:46:05, 15.08s/it]
                                                        
{'loss': '0.1461', 'grad_norm': '3.44', 'learning_rate': '1.795e-05', 'epoch': '0.2312'}

 23%|██▎       | 2280/9864 [9:34:31<31:46:05, 15.08s/it]
 23%|██▎       | 2281/9864 [9:34:46<31:33:42, 14.98s/it]
 23%|██▎       | 2282/9864 [9:35:00<30:50:50, 14.65s/it]
 23%|██▎       | 2283/9864 [9:35:19<33:54:08, 16.10s/it]
 23%|██▎       | 2284/9864 [9:35:35<33:54:57, 16.11s/it]
 23%|██▎       | 2285/9864 [9:35:50<32:51:11, 15.61s/it]
 23%|██▎       | 2286/9864 [9:36:04<32:05:36, 15.25s/it]
 23%|██▎       | 2287/9864 [9:36:19<31:32:07, 14.98s/it]
 23%|██▎       | 2288/9864 [9:36:33<31:03:22, 14.76s/it]
 23%|██▎       | 2289/9864 [9:36:47<30:42:13, 14.59s/it]
 23%|██▎       | 2290/9864 [9:37:03<31:21:19, 14.90s/it]
                                                        
{'loss': '0.1641', 'grad_norm': '3.79', 'learning_rate': '1.793e-05', 'epoch': '0.2322'}

 23%|██▎       | 2290/9864 [9:37:03<31:21:19, 14.90s/it]
 23%|██▎       | 2291/9864 [9:37:19<32:15:33, 15.34s/it]
 23%|██▎       | 2292/9864 [9:37:35<32:41:40, 15.54s/it]
 23%|██▎       | 2293/9864 [9:37:52<33:21:27, 15.86s/it]
 23%|██▎       | 2294/9864 [9:38:06<32:12:45, 15.32s/it]
 23%|██▎       | 2295/9864 [9:38:20<31:28:56, 14.97s/it]
 23%|██▎       | 2296/9864 [9:38:34<30:52:51, 14.69s/it]
 23%|██▎       | 2297/9864 [9:38:47<30:07:13, 14.33s/it]
 23%|██▎       | 2298/9864 [9:39:04<31:22:37, 14.93s/it]
 23%|██▎       | 2299/9864 [9:39:21<33:04:35, 15.74s/it]
 23%|██▎       | 2300/9864 [9:39:38<33:28:04, 15.93s/it]
                                                        
{'loss': '0.1703', 'grad_norm': '2.501', 'learning_rate': '1.791e-05', 'epoch': '0.2332'}

 23%|██▎       | 2300/9864 [9:39:38<33:28:04, 15.93s/it]
 23%|██▎       | 2301/9864 [9:39:52<32:35:50, 15.52s/it]
 23%|██▎       | 2302/9864 [9:40:07<31:54:27, 15.19s/it]
 23%|██▎       | 2303/9864 [9:40:22<31:39:04, 15.07s/it]
 23%|██▎       | 2304/9864 [9:40:36<30:59:23, 14.76s/it]
 23%|██▎       | 2305/9864 [9:40:49<30:22:41, 14.47s/it]
 23%|██▎       | 2306/9864 [9:41:06<31:29:24, 15.00s/it]
 23%|██▎       | 2307/9864 [9:41:20<30:57:58, 14.75s/it]
 23%|██▎       | 2308/9864 [9:41:38<32:52:18, 15.66s/it]
 23%|██▎       | 2309/9864 [9:41:52<31:51:01, 15.18s/it]
 23%|██▎       | 2310/9864 [9:42:05<30:52:13, 14.71s/it]
                                                        
{'loss': '0.1596', 'grad_norm': '1.854', 'learning_rate': '1.789e-05', 'epoch': '0.2342'}

 23%|██▎       | 2310/9864 [9:42:05<30:52:13, 14.71s/it]
 23%|██▎       | 2311/9864 [9:42:19<30:26:57, 14.51s/it]
 23%|██▎       | 2312/9864 [9:42:34<30:30:59, 14.55s/it]
 23%|██▎       | 2313/9864 [9:42:50<31:44:50, 15.14s/it]
 23%|██▎       | 2314/9864 [9:43:05<31:13:46, 14.89s/it]
 23%|██▎       | 2315/9864 [9:43:21<31:59:18, 15.25s/it]
 23%|██▎       | 2316/9864 [9:43:40<34:06:18, 16.27s/it]
 23%|██▎       | 2317/9864 [9:43:54<32:57:07, 15.72s/it]
 23%|██▎       | 2318/9864 [9:44:09<32:14:00, 15.38s/it]
 24%|██▎       | 2319/9864 [9:44:23<31:28:23, 15.02s/it]
 24%|██▎       | 2320/9864 [9:44:38<31:20:16, 14.95s/it]
                                                        
{'loss': '0.1741', 'grad_norm': '3.528', 'learning_rate': '1.787e-05', 'epoch': '0.2352'}

 24%|██▎       | 2320/9864 [9:44:38<31:20:16, 14.95s/it]
 24%|██▎       | 2321/9864 [9:44:52<30:59:53, 14.79s/it]
 24%|██▎       | 2322/9864 [9:45:10<33:14:49, 15.87s/it]
 24%|██▎       | 2323/9864 [9:45:28<34:26:16, 16.44s/it]
 24%|██▎       | 2324/9864 [9:45:43<33:25:02, 15.96s/it]
 24%|██▎       | 2325/9864 [9:45:57<32:23:53, 15.47s/it]
 24%|██▎       | 2326/9864 [9:46:12<31:49:40, 15.20s/it]
 24%|██▎       | 2327/9864 [9:46:26<31:21:17, 14.98s/it]
 24%|██▎       | 2328/9864 [9:46:41<31:10:39, 14.89s/it]
 24%|██▎       | 2329/9864 [9:46:57<31:59:55, 15.29s/it]
 24%|██▎       | 2330/9864 [9:47:12<31:36:51, 15.11s/it]
                                                        
{'loss': '0.1568', 'grad_norm': '6.209', 'learning_rate': '1.785e-05', 'epoch': '0.2362'}

 24%|██▎       | 2330/9864 [9:47:12<31:36:51, 15.11s/it]
 24%|██▎       | 2331/9864 [9:47:30<33:31:09, 16.02s/it]
 24%|██▎       | 2332/9864 [9:47:47<33:50:18, 16.17s/it]
 24%|██▎       | 2333/9864 [9:48:01<32:34:56, 15.58s/it]
 24%|██▎       | 2334/9864 [9:48:16<32:16:26, 15.43s/it]
 24%|██▎       | 2335/9864 [9:48:31<31:52:11, 15.24s/it]
 24%|██▎       | 2336/9864 [9:48:45<31:14:44, 14.94s/it]
 24%|██▎       | 2337/9864 [9:49:01<32:04:16, 15.34s/it]
 24%|██▎       | 2338/9864 [9:49:17<32:31:36, 15.56s/it]
 24%|██▎       | 2339/9864 [9:49:35<34:06:26, 16.32s/it]
 24%|██▎       | 2340/9864 [9:49:50<32:53:23, 15.74s/it]
                                                        
{'loss': '0.1563', 'grad_norm': '1.899', 'learning_rate': '1.783e-05', 'epoch': '0.2372'}

 24%|██▎       | 2340/9864 [9:49:50<32:53:23, 15.74s/it]
 24%|██▎       | 2341/9864 [9:50:04<31:48:45, 15.22s/it]
 24%|██▎       | 2342/9864 [9:50:19<31:32:50, 15.10s/it]
 24%|██▍       | 2343/9864 [9:50:33<31:25:32, 15.04s/it]
 24%|██▍       | 2344/9864 [9:50:48<31:00:45, 14.85s/it]
 24%|██▍       | 2345/9864 [9:51:04<31:48:18, 15.23s/it]
 24%|██▍       | 2346/9864 [9:51:22<33:29:08, 16.03s/it]
 24%|██▍       | 2347/9864 [9:51:36<32:24:44, 15.52s/it]
 24%|██▍       | 2348/9864 [9:51:53<32:57:49, 15.79s/it]
 24%|██▍       | 2349/9864 [9:52:07<32:04:05, 15.36s/it]
 24%|██▍       | 2350/9864 [9:52:22<31:46:15, 15.22s/it]
                                                        
{'loss': '0.161', 'grad_norm': '1.534', 'learning_rate': '1.781e-05', 'epoch': '0.2383'}

 24%|██▍       | 2350/9864 [9:52:22<31:46:15, 15.22s/it]
 24%|██▍       | 2351/9864 [9:52:36<31:18:35, 15.00s/it]
 24%|██▍       | 2352/9864 [9:52:51<30:51:14, 14.79s/it]
 24%|██▍       | 2353/9864 [9:53:06<31:29:50, 15.10s/it]
 24%|██▍       | 2354/9864 [9:53:23<32:33:49, 15.61s/it]
 24%|██▍       | 2355/9864 [9:53:40<33:19:33, 15.98s/it]
 24%|██▍       | 2356/9864 [9:53:56<33:05:23, 15.87s/it]
 24%|██▍       | 2357/9864 [9:54:10<32:21:32, 15.52s/it]
 24%|██▍       | 2358/9864 [9:54:25<31:36:10, 15.16s/it]
 24%|██▍       | 2359/9864 [9:54:40<31:35:45, 15.16s/it]
 24%|██▍       | 2360/9864 [9:54:55<31:30:15, 15.11s/it]
                                                        
{'loss': '0.1524', 'grad_norm': '2.973', 'learning_rate': '1.779e-05', 'epoch': '0.2393'}

 24%|██▍       | 2360/9864 [9:54:55<31:30:15, 15.11s/it]
 24%|██▍       | 2361/9864 [9:55:13<33:14:24, 15.95s/it]
 24%|██▍       | 2362/9864 [9:55:30<33:47:17, 16.21s/it]
 24%|██▍       | 2363/9864 [9:55:45<33:34:35, 16.11s/it]
 24%|██▍       | 2364/9864 [9:56:01<32:55:25, 15.80s/it]
 24%|██▍       | 2365/9864 [9:56:15<31:53:22, 15.31s/it]
 24%|██▍       | 2366/9864 [9:56:29<31:21:14, 15.05s/it]
 24%|██▍       | 2367/9864 [9:56:44<31:15:06, 15.01s/it]
 24%|██▍       | 2368/9864 [9:57:00<31:45:59, 15.26s/it]
 24%|██▍       | 2369/9864 [9:57:16<32:08:02, 15.43s/it]
 24%|██▍       | 2370/9864 [9:57:30<31:37:41, 15.19s/it]
                                                        
{'loss': '0.1542', 'grad_norm': '3.727', 'learning_rate': '1.777e-05', 'epoch': '0.2403'}

 24%|██▍       | 2370/9864 [9:57:30<31:37:41, 15.19s/it]
 24%|██▍       | 2371/9864 [9:57:48<33:26:03, 16.06s/it]
 24%|██▍       | 2372/9864 [9:58:03<32:30:15, 15.62s/it]
 24%|██▍       | 2373/9864 [9:58:18<31:52:49, 15.32s/it]
 24%|██▍       | 2374/9864 [9:58:32<31:16:24, 15.03s/it]
 24%|██▍       | 2375/9864 [9:58:46<30:48:05, 14.81s/it]
 24%|██▍       | 2376/9864 [9:59:05<32:54:26, 15.82s/it]
 24%|██▍       | 2377/9864 [9:59:19<32:11:43, 15.48s/it]
 24%|██▍       | 2378/9864 [9:59:38<33:58:57, 16.34s/it]
 24%|██▍       | 2379/9864 [9:59:52<33:01:44, 15.89s/it]
 24%|██▍       | 2380/9864 [10:00:07<32:21:31, 15.57s/it]
                                                         
{'loss': '0.1505', 'grad_norm': '2.47', 'learning_rate': '1.775e-05', 'epoch': '0.2413'}

 24%|██▍       | 2380/9864 [10:00:07<32:21:31, 15.57s/it]
 24%|██▍       | 2381/9864 [10:00:22<31:42:22, 15.25s/it]
 24%|██▍       | 2382/9864 [10:00:36<31:17:48, 15.06s/it]
 24%|██▍       | 2383/9864 [10:00:53<32:09:39, 15.48s/it]
 24%|██▍       | 2384/9864 [10:01:09<32:54:07, 15.84s/it]
 24%|██▍       | 2385/9864 [10:01:26<33:21:35, 16.06s/it]
 24%|██▍       | 2386/9864 [10:01:41<32:34:25, 15.68s/it]
 24%|██▍       | 2387/9864 [10:01:57<32:48:51, 15.80s/it]
 24%|██▍       | 2388/9864 [10:02:12<32:03:48, 15.44s/it]
 24%|██▍       | 2389/9864 [10:02:26<31:28:51, 15.16s/it]
 24%|██▍       | 2390/9864 [10:02:41<31:11:38, 15.03s/it]
                                                         
{'loss': '0.1593', 'grad_norm': '2.023', 'learning_rate': '1.773e-05', 'epoch': '0.2423'}

 24%|██▍       | 2390/9864 [10:02:41<31:11:38, 15.03s/it]
 24%|██▍       | 2391/9864 [10:02:59<33:03:13, 15.92s/it]
 24%|██▍       | 2392/9864 [10:03:14<32:24:56, 15.62s/it]
 24%|██▍       | 2393/9864 [10:03:29<32:02:52, 15.44s/it]
 24%|██▍       | 2394/9864 [10:03:45<32:29:43, 15.66s/it]
 24%|██▍       | 2395/9864 [10:04:01<32:47:26, 15.80s/it]
 24%|██▍       | 2396/9864 [10:04:16<32:12:00, 15.52s/it]
 24%|██▍       | 2397/9864 [10:04:31<31:45:40, 15.31s/it]
 24%|██▍       | 2398/9864 [10:04:47<32:20:42, 15.60s/it]
 24%|██▍       | 2399/9864 [10:05:04<33:06:58, 15.97s/it]
 24%|██▍       | 2400/9864 [10:05:18<32:17:08, 15.57s/it]
                                                         
{'loss': '0.1525', 'grad_norm': '1.796', 'learning_rate': '1.771e-05', 'epoch': '0.2433'}

 24%|██▍       | 2400/9864 [10:05:18<32:17:08, 15.57s/it]
 24%|██▍       | 2401/9864 [10:05:33<31:50:35, 15.36s/it]
 24%|██▍       | 2402/9864 [10:05:49<32:10:34, 15.52s/it]
 24%|██▍       | 2403/9864 [10:06:06<32:44:07, 15.80s/it]
 24%|██▍       | 2404/9864 [10:06:20<32:02:03, 15.46s/it]
 24%|██▍       | 2405/9864 [10:06:35<31:38:35, 15.27s/it]
 24%|██▍       | 2406/9864 [10:06:53<33:18:09, 16.08s/it]
 24%|██▍       | 2407/9864 [10:07:08<32:17:29, 15.59s/it]
 24%|██▍       | 2408/9864 [10:07:22<31:46:21, 15.34s/it]
 24%|██▍       | 2409/9864 [10:07:37<31:11:17, 15.06s/it]
 24%|██▍       | 2410/9864 [10:07:53<31:49:27, 15.37s/it]
                                                         
{'loss': '0.1575', 'grad_norm': '1.804', 'learning_rate': '1.769e-05', 'epoch': '0.2443'}

 24%|██▍       | 2410/9864 [10:07:53<31:49:27, 15.37s/it]
 24%|██▍       | 2411/9864 [10:08:09<32:34:33, 15.74s/it]
 24%|██▍       | 2412/9864 [10:08:24<31:36:22, 15.27s/it]
 24%|██▍       | 2413/9864 [10:08:40<32:04:32, 15.50s/it]
 24%|██▍       | 2414/9864 [10:08:55<32:13:37, 15.57s/it]
 24%|██▍       | 2415/9864 [10:09:10<31:32:06, 15.24s/it]
 24%|██▍       | 2416/9864 [10:09:24<31:11:00, 15.07s/it]
 25%|██▍       | 2417/9864 [10:09:39<30:56:01, 14.95s/it]
 25%|██▍       | 2418/9864 [10:09:56<31:57:49, 15.45s/it]
 25%|██▍       | 2419/9864 [10:10:12<32:23:51, 15.67s/it]
 25%|██▍       | 2420/9864 [10:10:27<32:04:37, 15.51s/it]
                                                         
{'loss': '0.1423', 'grad_norm': '2.88', 'learning_rate': '1.767e-05', 'epoch': '0.2454'}

 25%|██▍       | 2420/9864 [10:10:27<32:04:37, 15.51s/it]
 25%|██▍       | 2421/9864 [10:10:43<32:13:42, 15.59s/it]
 25%|██▍       | 2422/9864 [10:10:58<31:38:39, 15.31s/it]
 25%|██▍       | 2423/9864 [10:11:14<32:38:21, 15.79s/it]
 25%|██▍       | 2424/9864 [10:11:28<31:17:10, 15.14s/it]
 25%|██▍       | 2425/9864 [10:11:42<30:36:34, 14.81s/it]
 25%|██▍       | 2426/9864 [10:12:00<32:32:42, 15.75s/it]
 25%|██▍       | 2427/9864 [10:12:14<31:30:07, 15.25s/it]
 25%|██▍       | 2428/9864 [10:12:30<31:59:49, 15.49s/it]
 25%|██▍       | 2429/9864 [10:12:45<31:18:13, 15.16s/it]
 25%|██▍       | 2430/9864 [10:13:00<31:10:09, 15.09s/it]
                                                         
{'loss': '0.163', 'grad_norm': '4.262', 'learning_rate': '1.765e-05', 'epoch': '0.2464'}

 25%|██▍       | 2430/9864 [10:13:00<31:10:09, 15.09s/it]
 25%|██▍       | 2431/9864 [10:13:16<31:50:27, 15.42s/it]
 25%|██▍       | 2432/9864 [10:13:30<31:19:46, 15.18s/it]
 25%|██▍       | 2433/9864 [10:13:45<30:51:59, 14.95s/it]
 25%|██▍       | 2434/9864 [10:14:03<33:13:29, 16.10s/it]
 25%|██▍       | 2435/9864 [10:14:18<32:10:52, 15.59s/it]
 25%|██▍       | 2436/9864 [10:14:34<32:44:59, 15.87s/it]
 25%|██▍       | 2437/9864 [10:14:49<31:47:47, 15.41s/it]
 25%|██▍       | 2438/9864 [10:15:03<31:12:56, 15.13s/it]
 25%|██▍       | 2439/9864 [10:15:19<31:40:08, 15.35s/it]
 25%|██▍       | 2440/9864 [10:15:34<31:20:15, 15.20s/it]
                                                         
{'loss': '0.172', 'grad_norm': '2.441', 'learning_rate': '1.762e-05', 'epoch': '0.2474'}

 25%|██▍       | 2440/9864 [10:15:34<31:20:15, 15.20s/it]
 25%|██▍       | 2441/9864 [10:15:49<30:59:59, 15.03s/it]
 25%|██▍       | 2442/9864 [10:16:07<33:13:36, 16.12s/it]
 25%|██▍       | 2443/9864 [10:16:21<32:02:23, 15.54s/it]
 25%|██▍       | 2444/9864 [10:16:38<32:50:30, 15.93s/it]
 25%|██▍       | 2445/9864 [10:16:53<31:50:11, 15.45s/it]
 25%|██▍       | 2446/9864 [10:17:07<31:22:48, 15.23s/it]
 25%|██▍       | 2447/9864 [10:17:24<32:00:57, 15.54s/it]
 25%|██▍       | 2448/9864 [10:17:38<31:31:21, 15.30s/it]
 25%|██▍       | 2449/9864 [10:17:54<32:00:06, 15.54s/it]
 25%|██▍       | 2450/9864 [10:18:09<31:18:58, 15.21s/it]
                                                         
{'loss': '0.154', 'grad_norm': '2.778', 'learning_rate': '1.76e-05', 'epoch': '0.2484'}

 25%|██▍       | 2450/9864 [10:18:09<31:18:58, 15.21s/it]
 25%|██▍       | 2451/9864 [10:18:28<33:37:51, 16.33s/it]
 25%|██▍       | 2452/9864 [10:18:42<32:13:06, 15.65s/it]
 25%|██▍       | 2453/9864 [10:18:56<31:23:13, 15.25s/it]
 25%|██▍       | 2454/9864 [10:19:11<30:58:26, 15.05s/it]
 25%|██▍       | 2455/9864 [10:19:27<31:44:15, 15.42s/it]
 25%|██▍       | 2456/9864 [10:19:42<31:16:53, 15.20s/it]
 25%|██▍       | 2457/9864 [10:19:59<32:16:03, 15.68s/it]
 25%|██▍       | 2458/9864 [10:20:15<32:41:44, 15.89s/it]
 25%|██▍       | 2459/9864 [10:20:32<33:07:03, 16.10s/it]
 25%|██▍       | 2460/9864 [10:20:45<31:41:11, 15.41s/it]
                                                         
{'loss': '0.1719', 'grad_norm': '3.268', 'learning_rate': '1.758e-05', 'epoch': '0.2494'}

 25%|██▍       | 2460/9864 [10:20:45<31:41:11, 15.41s/it]
 25%|██▍       | 2461/9864 [10:21:00<30:57:06, 15.05s/it]
 25%|██▍       | 2462/9864 [10:21:14<30:33:15, 14.86s/it]
 25%|██▍       | 2463/9864 [10:21:31<31:35:25, 15.37s/it]
 25%|██▍       | 2464/9864 [10:21:45<31:10:47, 15.17s/it]
 25%|██▍       | 2465/9864 [10:22:00<30:46:56, 14.98s/it]
 25%|██▌       | 2466/9864 [10:22:18<32:34:54, 15.85s/it]
 25%|██▌       | 2467/9864 [10:22:34<33:00:54, 16.07s/it]
 25%|██▌       | 2468/9864 [10:22:49<32:21:42, 15.75s/it]
 25%|██▌       | 2469/9864 [10:23:04<31:40:18, 15.42s/it]
 25%|██▌       | 2470/9864 [10:23:20<31:57:36, 15.56s/it]
                                                         
{'loss': '0.1576', 'grad_norm': '2.02', 'learning_rate': '1.756e-05', 'epoch': '0.2504'}

 25%|██▌       | 2470/9864 [10:23:20<31:57:36, 15.56s/it]
 25%|██▌       | 2471/9864 [10:23:34<31:07:07, 15.15s/it]
 25%|██▌       | 2472/9864 [10:23:48<30:30:46, 14.86s/it]
 25%|██▌       | 2473/9864 [10:24:04<31:10:25, 15.18s/it]
 25%|██▌       | 2474/9864 [10:24:21<32:26:00, 15.80s/it]
 25%|██▌       | 2475/9864 [10:24:36<31:45:18, 15.47s/it]
 25%|██▌       | 2476/9864 [10:24:50<31:02:12, 15.12s/it]
 25%|██▌       | 2477/9864 [10:25:05<30:38:09, 14.93s/it]
 25%|██▌       | 2478/9864 [10:25:21<31:24:06, 15.31s/it]
 25%|██▌       | 2479/9864 [10:25:35<30:41:48, 14.96s/it]
 25%|██▌       | 2480/9864 [10:25:50<30:26:53, 14.84s/it]
                                                         
{'loss': '0.1624', 'grad_norm': '2.172', 'learning_rate': '1.754e-05', 'epoch': '0.2514'}

 25%|██▌       | 2480/9864 [10:25:50<30:26:53, 14.84s/it]
 25%|██▌       | 2481/9864 [10:26:08<32:26:00, 15.81s/it]
 25%|██▌       | 2482/9864 [10:26:22<31:38:34, 15.43s/it]
 25%|██▌       | 2483/9864 [10:26:39<32:18:58, 15.76s/it]
 25%|██▌       | 2484/9864 [10:26:53<31:20:42, 15.29s/it]
 25%|██▌       | 2485/9864 [10:27:10<32:04:13, 15.65s/it]
 25%|██▌       | 2486/9864 [10:27:24<31:12:57, 15.23s/it]
 25%|██▌       | 2487/9864 [10:27:38<30:22:59, 14.83s/it]
 25%|██▌       | 2488/9864 [10:27:53<30:52:53, 15.07s/it]
 25%|██▌       | 2489/9864 [10:28:10<31:34:46, 15.42s/it]
 25%|██▌       | 2490/9864 [10:28:26<31:56:54, 15.60s/it]
                                                         
{'loss': '0.1612', 'grad_norm': '4.62', 'learning_rate': '1.752e-05', 'epoch': '0.2525'}

 25%|██▌       | 2490/9864 [10:28:26<31:56:54, 15.60s/it]
 25%|██▌       | 2491/9864 [10:28:40<31:29:55, 15.38s/it]
 25%|██▌       | 2492/9864 [10:28:57<31:57:18, 15.60s/it]
 25%|██▌       | 2493/9864 [10:29:11<31:05:25, 15.18s/it]
 25%|██▌       | 2494/9864 [10:29:25<30:32:03, 14.91s/it]
 25%|██▌       | 2495/9864 [10:29:40<30:24:10, 14.85s/it]
 25%|██▌       | 2496/9864 [10:29:56<31:02:53, 15.17s/it]
 25%|██▌       | 2497/9864 [10:30:10<30:36:05, 14.95s/it]
 25%|██▌       | 2498/9864 [10:30:27<31:35:53, 15.44s/it]
 25%|██▌       | 2499/9864 [10:30:41<31:02:56, 15.18s/it]
 25%|██▌       | 2500/9864 [10:30:57<31:38:19, 15.47s/it]
                                                         
{'loss': '0.151', 'grad_norm': '4.868', 'learning_rate': '1.75e-05', 'epoch': '0.2535'}

 25%|██▌       | 2500/9864 [10:30:57<31:38:19, 15.47s/it]
 25%|██▌       | 2501/9864 [10:31:12<31:07:14, 15.22s/it]
 25%|██▌       | 2502/9864 [10:31:27<30:40:45, 15.00s/it]
 25%|██▌       | 2503/9864 [10:31:41<30:18:26, 14.82s/it]
 25%|██▌       | 2504/9864 [10:31:59<32:16:57, 15.79s/it]
 25%|██▌       | 2505/9864 [10:32:15<32:25:59, 15.87s/it]
 25%|██▌       | 2506/9864 [10:32:30<31:39:15, 15.49s/it]
 25%|██▌       | 2507/9864 [10:32:46<32:10:33, 15.74s/it]
 25%|██▌       | 2508/9864 [10:33:00<31:10:56, 15.26s/it]
 25%|██▌       | 2509/9864 [10:33:14<30:13:59, 14.80s/it]
 25%|██▌       | 2510/9864 [10:33:28<30:07:16, 14.75s/it]
                                                         
{'loss': '0.1619', 'grad_norm': '3.797', 'learning_rate': '1.747e-05', 'epoch': '0.2545'}

 25%|██▌       | 2510/9864 [10:33:28<30:07:16, 14.75s/it]
 25%|██▌       | 2511/9864 [10:33:45<30:56:03, 15.15s/it]
 25%|██▌       | 2512/9864 [10:34:01<31:44:34, 15.54s/it]
 25%|██▌       | 2513/9864 [10:34:17<31:50:42, 15.60s/it]
 25%|██▌       | 2514/9864 [10:34:31<31:07:13, 15.24s/it]
 25%|██▌       | 2515/9864 [10:34:47<31:29:29, 15.43s/it]
 26%|██▌       | 2516/9864 [10:35:02<31:05:24, 15.23s/it]
 26%|██▌       | 2517/9864 [10:35:16<30:10:49, 14.79s/it]
 26%|██▌       | 2518/9864 [10:35:30<29:55:20, 14.66s/it]
 26%|██▌       | 2519/9864 [10:35:44<29:47:25, 14.60s/it]
 26%|██▌       | 2520/9864 [10:36:04<32:53:51, 16.13s/it]
                                                         
{'loss': '0.1574', 'grad_norm': '3.036', 'learning_rate': '1.745e-05', 'epoch': '0.2555'}

 26%|██▌       | 2520/9864 [10:36:04<32:53:51, 16.13s/it]
 26%|██▌       | 2521/9864 [10:36:18<31:36:43, 15.50s/it]
 26%|██▌       | 2522/9864 [10:36:32<30:52:51, 15.14s/it]
 26%|██▌       | 2523/9864 [10:36:46<30:13:29, 14.82s/it]
 26%|██▌       | 2524/9864 [10:37:02<30:45:07, 15.08s/it]
 26%|██▌       | 2525/9864 [10:37:16<30:06:59, 14.77s/it]
 26%|██▌       | 2526/9864 [10:37:31<29:53:00, 14.66s/it]
 26%|██▌       | 2527/9864 [10:37:48<31:46:08, 15.59s/it]
 26%|██▌       | 2528/9864 [10:38:03<31:16:50, 15.35s/it]
 26%|██▌       | 2529/9864 [10:38:19<31:29:37, 15.46s/it]
 26%|██▌       | 2530/9864 [10:38:33<30:37:05, 15.03s/it]
                                                         
{'loss': '0.1595', 'grad_norm': '2.356', 'learning_rate': '1.743e-05', 'epoch': '0.2565'}

 26%|██▌       | 2530/9864 [10:38:33<30:37:05, 15.03s/it]
 26%|██▌       | 2531/9864 [10:38:49<31:20:24, 15.39s/it]
 26%|██▌       | 2532/9864 [10:39:04<30:48:16, 15.13s/it]
 26%|██▌       | 2533/9864 [10:39:18<30:17:20, 14.87s/it]
 26%|██▌       | 2534/9864 [10:39:32<29:47:28, 14.63s/it]
 26%|██▌       | 2535/9864 [10:39:50<31:49:40, 15.63s/it]
 26%|██▌       | 2536/9864 [10:40:04<30:55:38, 15.19s/it]
 26%|██▌       | 2537/9864 [10:40:20<31:33:30, 15.51s/it]
 26%|██▌       | 2538/9864 [10:40:35<31:03:24, 15.26s/it]
 26%|██▌       | 2539/9864 [10:40:51<31:32:44, 15.50s/it]
 26%|██▌       | 2540/9864 [10:41:05<30:48:24, 15.14s/it]
                                                         
{'loss': '0.1569', 'grad_norm': '2.252', 'learning_rate': '1.741e-05', 'epoch': '0.2575'}

 26%|██▌       | 2540/9864 [10:41:05<30:48:24, 15.14s/it]
 26%|██▌       | 2541/9864 [10:41:20<30:27:12, 14.97s/it]
 26%|██▌       | 2542/9864 [10:41:36<31:23:06, 15.43s/it]
 26%|██▌       | 2543/9864 [10:41:52<31:37:13, 15.55s/it]
 26%|██▌       | 2544/9864 [10:42:06<30:35:15, 15.04s/it]
 26%|██▌       | 2545/9864 [10:42:22<31:11:29, 15.34s/it]
 26%|██▌       | 2546/9864 [10:42:36<30:29:43, 15.00s/it]
 26%|██▌       | 2547/9864 [10:42:53<31:16:49, 15.39s/it]
 26%|██▌       | 2548/9864 [10:43:07<30:37:47, 15.07s/it]
 26%|██▌       | 2549/9864 [10:43:21<30:13:32, 14.88s/it]
 26%|██▌       | 2550/9864 [10:43:40<32:29:40, 15.99s/it]
                                                         
{'loss': '0.1534', 'grad_norm': '2.088', 'learning_rate': '1.739e-05', 'epoch': '0.2585'}

 26%|██▌       | 2550/9864 [10:43:40<32:29:40, 15.99s/it]
 26%|██▌       | 2551/9864 [10:43:54<31:17:56, 15.41s/it]
 26%|██▌       | 2552/9864 [10:44:09<30:44:01, 15.13s/it]
 26%|██▌       | 2553/9864 [10:44:23<30:11:36, 14.87s/it]
 26%|██▌       | 2554/9864 [10:44:39<30:53:21, 15.21s/it]
 26%|██▌       | 2555/9864 [10:44:53<30:25:06, 14.98s/it]
 26%|██▌       | 2556/9864 [10:45:09<31:02:22, 15.29s/it]
 26%|██▌       | 2557/9864 [10:45:24<30:32:41, 15.05s/it]
 26%|██▌       | 2558/9864 [10:45:40<31:25:52, 15.49s/it]
 26%|██▌       | 2559/9864 [10:45:57<31:51:36, 15.70s/it]
 26%|██▌       | 2560/9864 [10:46:11<30:58:15, 15.26s/it]
                                                         
{'loss': '0.1684', 'grad_norm': '4.148', 'learning_rate': '1.736e-05', 'epoch': '0.2596'}

 26%|██▌       | 2560/9864 [10:46:11<30:58:15, 15.26s/it]
 26%|██▌       | 2561/9864 [10:46:25<30:21:23, 14.96s/it]
 26%|██▌       | 2562/9864 [10:46:41<30:55:02, 15.24s/it]
 26%|██▌       | 2563/9864 [10:46:57<31:14:08, 15.40s/it]
 26%|██▌       | 2564/9864 [10:47:11<30:34:06, 15.07s/it]
 26%|██▌       | 2565/9864 [10:47:27<31:21:29, 15.47s/it]
 26%|██▌       | 2566/9864 [10:47:42<30:50:35, 15.21s/it]
 26%|██▌       | 2567/9864 [10:47:56<30:22:39, 14.99s/it]
 26%|██▌       | 2568/9864 [10:48:13<31:02:57, 15.32s/it]
 26%|██▌       | 2569/9864 [10:48:27<30:24:45, 15.01s/it]
 26%|██▌       | 2570/9864 [10:48:43<31:03:29, 15.33s/it]
                                                         
{'loss': '0.1569', 'grad_norm': '2.486', 'learning_rate': '1.734e-05', 'epoch': '0.2606'}

 26%|██▌       | 2570/9864 [10:48:43<31:03:29, 15.33s/it]
 26%|██▌       | 2571/9864 [10:48:59<31:35:25, 15.59s/it]
 26%|██▌       | 2572/9864 [10:49:13<30:37:14, 15.12s/it]
 26%|██▌       | 2573/9864 [10:49:29<31:11:45, 15.40s/it]
 26%|██▌       | 2574/9864 [10:49:44<30:43:34, 15.17s/it]
 26%|██▌       | 2575/9864 [10:49:58<30:07:13, 14.88s/it]
 26%|██▌       | 2576/9864 [10:50:14<30:54:35, 15.27s/it]
 26%|██▌       | 2577/9864 [10:50:31<31:39:33, 15.64s/it]
 26%|██▌       | 2578/9864 [10:50:45<30:49:15, 15.23s/it]
 26%|██▌       | 2579/9864 [10:50:59<29:56:27, 14.80s/it]
 26%|██▌       | 2580/9864 [10:51:15<30:41:34, 15.17s/it]
                                                         
{'loss': '0.1781', 'grad_norm': '2.367', 'learning_rate': '1.732e-05', 'epoch': '0.2616'}

 26%|██▌       | 2580/9864 [10:51:15<30:41:34, 15.17s/it]
 26%|██▌       | 2581/9864 [10:51:32<31:45:12, 15.70s/it]
 26%|██▌       | 2582/9864 [10:51:46<31:10:51, 15.41s/it]
 26%|██▌       | 2583/9864 [10:52:01<30:33:55, 15.11s/it]
 26%|██▌       | 2584/9864 [10:52:18<31:35:59, 15.63s/it]
 26%|██▌       | 2585/9864 [10:52:34<32:07:06, 15.88s/it]
 26%|██▌       | 2586/9864 [10:52:49<31:27:58, 15.56s/it]
 26%|██▌       | 2587/9864 [10:53:04<31:04:37, 15.37s/it]
 26%|██▌       | 2588/9864 [10:53:23<33:23:23, 16.52s/it]
 26%|██▌       | 2589/9864 [10:53:38<32:06:17, 15.89s/it]
 26%|██▋       | 2590/9864 [10:53:53<31:37:11, 15.65s/it]
                                                         
{'loss': '0.1629', 'grad_norm': '2.512', 'learning_rate': '1.73e-05', 'epoch': '0.2626'}

 26%|██▋       | 2590/9864 [10:53:53<31:37:11, 15.65s/it]
 26%|██▋       | 2591/9864 [10:54:09<32:06:46, 15.90s/it]
 26%|██▋       | 2592/9864 [10:54:23<31:04:58, 15.39s/it]
 26%|██▋       | 2593/9864 [10:54:39<31:23:57, 15.55s/it]
 26%|██▋       | 2594/9864 [10:54:54<30:37:36, 15.17s/it]
 26%|██▋       | 2595/9864 [10:55:09<30:55:57, 15.32s/it]
 26%|██▋       | 2596/9864 [10:55:26<31:36:43, 15.66s/it]
 26%|██▋       | 2597/9864 [10:55:40<30:55:50, 15.32s/it]
 26%|██▋       | 2598/9864 [10:55:55<30:25:49, 15.08s/it]
 26%|██▋       | 2599/9864 [10:56:11<31:13:48, 15.48s/it]
 26%|██▋       | 2600/9864 [10:56:27<31:42:45, 15.72s/it]
                                                         
{'loss': '0.1493', 'grad_norm': '2.657', 'learning_rate': '1.727e-05', 'epoch': '0.2636'}

 26%|██▋       | 2600/9864 [10:56:27<31:42:45, 15.72s/it]
 26%|██▋       | 2601/9864 [10:56:41<30:41:12, 15.21s/it]
 26%|██▋       | 2602/9864 [10:56:56<30:09:31, 14.95s/it]
 26%|██▋       | 2603/9864 [10:57:12<31:05:36, 15.42s/it]
 26%|██▋       | 2604/9864 [10:57:29<31:57:17, 15.85s/it]
 26%|██▋       | 2605/9864 [10:57:44<31:15:01, 15.50s/it]
 26%|██▋       | 2606/9864 [10:57:58<30:44:33, 15.25s/it]
 26%|██▋       | 2607/9864 [10:58:15<31:36:09, 15.68s/it]
 26%|██▋       | 2608/9864 [10:58:32<32:09:11, 15.95s/it]
 26%|██▋       | 2609/9864 [10:58:46<31:01:52, 15.40s/it]
 26%|██▋       | 2610/9864 [10:59:00<30:29:11, 15.13s/it]
                                                         
{'loss': '0.1669', 'grad_norm': '4.216', 'learning_rate': '1.725e-05', 'epoch': '0.2646'}

 26%|██▋       | 2610/9864 [10:59:00<30:29:11, 15.13s/it]
 26%|██▋       | 2611/9864 [10:59:17<31:15:16, 15.51s/it]
 26%|██▋       | 2612/9864 [10:59:33<31:42:06, 15.74s/it]
 26%|██▋       | 2613/9864 [10:59:47<30:40:49, 15.23s/it]
 27%|██▋       | 2614/9864 [11:00:01<30:08:33, 14.97s/it]
 27%|██▋       | 2615/9864 [11:00:20<32:06:13, 15.94s/it]
 27%|██▋       | 2616/9864 [11:00:34<30:54:59, 15.36s/it]
 27%|██▋       | 2617/9864 [11:00:48<30:23:00, 15.09s/it]
 27%|██▋       | 2618/9864 [11:01:03<30:05:00, 14.95s/it]
 27%|██▋       | 2619/9864 [11:01:19<31:00:47, 15.41s/it]
 27%|██▋       | 2620/9864 [11:01:36<31:36:55, 15.71s/it]
                                                         
{'loss': '0.1495', 'grad_norm': '2.347', 'learning_rate': '1.723e-05', 'epoch': '0.2656'}

 27%|██▋       | 2620/9864 [11:01:36<31:36:55, 15.71s/it]
 27%|██▋       | 2621/9864 [11:01:50<30:37:15, 15.22s/it]
 27%|██▋       | 2622/9864 [11:02:06<31:13:01, 15.52s/it]
 27%|██▋       | 2623/9864 [11:02:22<31:47:57, 15.81s/it]
 27%|██▋       | 2624/9864 [11:02:37<30:56:08, 15.38s/it]
 27%|██▋       | 2625/9864 [11:02:51<29:58:35, 14.91s/it]
 27%|██▋       | 2626/9864 [11:03:07<30:51:24, 15.35s/it]
 27%|██▋       | 2627/9864 [11:03:23<31:34:25, 15.71s/it]
 27%|██▋       | 2628/9864 [11:03:37<30:14:01, 15.04s/it]
 27%|██▋       | 2629/9864 [11:03:52<29:58:01, 14.91s/it]
 27%|██▋       | 2630/9864 [11:04:08<30:35:36, 15.22s/it]
                                                         
{'loss': '0.1566', 'grad_norm': '2.335', 'learning_rate': '1.721e-05', 'epoch': '0.2666'}

 27%|██▋       | 2630/9864 [11:04:08<30:35:36, 15.22s/it]
 27%|██▋       | 2631/9864 [11:04:23<31:00:57, 15.44s/it]
 27%|██▋       | 2632/9864 [11:04:38<30:39:01, 15.26s/it]
 27%|██▋       | 2633/9864 [11:04:53<30:21:53, 15.12s/it]
 27%|██▋       | 2634/9864 [11:05:10<31:11:02, 15.53s/it]
 27%|██▋       | 2635/9864 [11:05:26<31:40:17, 15.77s/it]
 27%|██▋       | 2636/9864 [11:05:40<30:56:23, 15.41s/it]
 27%|██▋       | 2637/9864 [11:05:57<31:24:08, 15.64s/it]
 27%|██▋       | 2638/9864 [11:06:11<30:43:51, 15.31s/it]
 27%|██▋       | 2639/9864 [11:06:27<31:03:52, 15.48s/it]
 27%|██▋       | 2640/9864 [11:06:42<30:45:23, 15.33s/it]
                                                         
{'loss': '0.1633', 'grad_norm': '4.483', 'learning_rate': '1.718e-05', 'epoch': '0.2677'}

 27%|██▋       | 2640/9864 [11:06:42<30:45:23, 15.33s/it]
 27%|██▋       | 2641/9864 [11:06:59<31:30:06, 15.70s/it]
 27%|██▋       | 2642/9864 [11:07:15<32:05:28, 16.00s/it]
 27%|██▋       | 2643/9864 [11:07:30<31:19:23, 15.62s/it]
 27%|██▋       | 2644/9864 [11:07:45<30:50:58, 15.38s/it]
 27%|██▋       | 2645/9864 [11:08:01<31:29:45, 15.71s/it]
 27%|██▋       | 2646/9864 [11:08:16<30:51:26, 15.39s/it]
 27%|██▋       | 2647/9864 [11:08:30<30:14:06, 15.08s/it]
 27%|██▋       | 2648/9864 [11:08:49<32:15:56, 16.10s/it]
 27%|██▋       | 2649/9864 [11:09:04<31:37:47, 15.78s/it]
 27%|██▋       | 2650/9864 [11:09:20<31:53:12, 15.91s/it]
                                                         
{'loss': '0.1586', 'grad_norm': '3.613', 'learning_rate': '1.716e-05', 'epoch': '0.2687'}

 27%|██▋       | 2650/9864 [11:09:20<31:53:12, 15.91s/it]
 27%|██▋       | 2651/9864 [11:09:35<31:22:09, 15.66s/it]
 27%|██▋       | 2652/9864 [11:09:51<31:44:10, 15.84s/it]
 27%|██▋       | 2653/9864 [11:10:05<30:36:11, 15.28s/it]
 27%|██▋       | 2654/9864 [11:10:20<29:59:09, 14.97s/it]
 27%|██▋       | 2655/9864 [11:10:34<29:34:08, 14.77s/it]
 27%|██▋       | 2656/9864 [11:10:52<31:18:46, 15.64s/it]
 27%|██▋       | 2657/9864 [11:11:06<30:23:46, 15.18s/it]
 27%|██▋       | 2658/9864 [11:11:22<30:50:26, 15.41s/it]
 27%|██▋       | 2659/9864 [11:11:36<30:19:21, 15.15s/it]
 27%|██▋       | 2660/9864 [11:11:52<30:41:57, 15.34s/it]
                                                         
{'loss': '0.1456', 'grad_norm': '2.017', 'learning_rate': '1.714e-05', 'epoch': '0.2697'}

 27%|██▋       | 2660/9864 [11:11:52<30:41:57, 15.34s/it]
 27%|██▋       | 2661/9864 [11:12:06<29:55:16, 14.95s/it]
 27%|██▋       | 2662/9864 [11:12:20<29:36:35, 14.80s/it]
 27%|██▋       | 2663/9864 [11:12:35<29:32:37, 14.77s/it]
 27%|██▋       | 2664/9864 [11:12:53<31:34:03, 15.78s/it]
 27%|██▋       | 2665/9864 [11:13:07<30:26:28, 15.22s/it]
 27%|██▋       | 2666/9864 [11:13:24<31:04:01, 15.54s/it]
 27%|██▋       | 2667/9864 [11:13:38<30:11:48, 15.10s/it]
 27%|██▋       | 2668/9864 [11:13:54<30:40:41, 15.35s/it]
 27%|██▋       | 2669/9864 [11:14:08<29:57:09, 14.99s/it]
 27%|██▋       | 2670/9864 [11:14:22<29:30:33, 14.77s/it]
                                                         
{'loss': '0.1541', 'grad_norm': '2.336', 'learning_rate': '1.712e-05', 'epoch': '0.2707'}

 27%|██▋       | 2670/9864 [11:14:22<29:30:33, 14.77s/it]
 27%|██▋       | 2671/9864 [11:14:39<30:36:27, 15.32s/it]
 27%|██▋       | 2672/9864 [11:14:55<31:10:10, 15.60s/it]
 27%|██▋       | 2673/9864 [11:15:09<30:17:08, 15.16s/it]
 27%|██▋       | 2674/9864 [11:15:25<30:46:43, 15.41s/it]
 27%|██▋       | 2675/9864 [11:15:40<30:48:10, 15.42s/it]
 27%|██▋       | 2676/9864 [11:15:55<30:17:43, 15.17s/it]
 27%|██▋       | 2677/9864 [11:16:09<29:46:54, 14.92s/it]
 27%|██▋       | 2678/9864 [11:16:24<29:31:50, 14.79s/it]
 27%|██▋       | 2679/9864 [11:16:42<31:42:21, 15.89s/it]
 27%|██▋       | 2680/9864 [11:16:57<30:49:24, 15.45s/it]
                                                         
{'loss': '0.1688', 'grad_norm': '1.664', 'learning_rate': '1.709e-05', 'epoch': '0.2717'}

 27%|██▋       | 2680/9864 [11:16:57<30:49:24, 15.45s/it]
 27%|██▋       | 2681/9864 [11:17:13<31:29:13, 15.78s/it]
 27%|██▋       | 2682/9864 [11:17:29<31:39:41, 15.87s/it]
 27%|██▋       | 2683/9864 [11:17:43<30:25:50, 15.26s/it]
 27%|██▋       | 2684/9864 [11:17:58<29:58:12, 15.03s/it]
 27%|██▋       | 2685/9864 [11:18:12<29:34:13, 14.83s/it]
 27%|██▋       | 2686/9864 [11:18:27<29:40:35, 14.88s/it]
 27%|██▋       | 2687/9864 [11:18:45<31:39:57, 15.88s/it]
 27%|██▋       | 2688/9864 [11:19:00<30:53:42, 15.50s/it]
 27%|██▋       | 2689/9864 [11:19:16<31:25:52, 15.77s/it]
 27%|██▋       | 2690/9864 [11:19:32<31:42:08, 15.91s/it]
                                                         
{'loss': '0.1501', 'grad_norm': '5.445', 'learning_rate': '1.707e-05', 'epoch': '0.2727'}

 27%|██▋       | 2690/9864 [11:19:32<31:42:08, 15.91s/it]
 27%|██▋       | 2691/9864 [11:19:47<30:51:40, 15.49s/it]
 27%|██▋       | 2692/9864 [11:20:02<30:23:40, 15.26s/it]
 27%|██▋       | 2693/9864 [11:20:16<29:53:03, 15.00s/it]
 27%|██▋       | 2694/9864 [11:20:30<29:28:49, 14.80s/it]
 27%|██▋       | 2695/9864 [11:20:48<31:17:25, 15.71s/it]
 27%|██▋       | 2696/9864 [11:21:05<31:48:29, 15.98s/it]
 27%|██▋       | 2697/9864 [11:21:21<32:08:20, 16.14s/it]
 27%|██▋       | 2698/9864 [11:21:36<31:25:08, 15.78s/it]
 27%|██▋       | 2699/9864 [11:21:51<30:37:26, 15.39s/it]
 27%|██▋       | 2700/9864 [11:22:05<30:05:16, 15.12s/it]
                                                         
{'loss': '0.1657', 'grad_norm': '1.37', 'learning_rate': '1.705e-05', 'epoch': '0.2737'}

 27%|██▋       | 2700/9864 [11:22:05<30:05:16, 15.12s/it]
 27%|██▋       | 2701/9864 [11:22:19<29:30:56, 14.83s/it]
 27%|██▋       | 2702/9864 [11:22:33<28:53:53, 14.53s/it]
 27%|██▋       | 2703/9864 [11:22:51<30:59:08, 15.58s/it]
 27%|██▋       | 2704/9864 [11:23:08<31:30:58, 15.85s/it]
 27%|██▋       | 2705/9864 [11:23:22<30:43:21, 15.45s/it]
 27%|██▋       | 2706/9864 [11:23:39<31:23:27, 15.79s/it]
 27%|██▋       | 2707/9864 [11:23:54<30:52:29, 15.53s/it]
 27%|██▋       | 2708/9864 [11:24:09<30:29:42, 15.34s/it]
 27%|██▋       | 2709/9864 [11:24:23<29:48:15, 15.00s/it]
 27%|██▋       | 2710/9864 [11:24:37<29:27:09, 14.82s/it]
                                                         
{'loss': '0.1512', 'grad_norm': '3.886', 'learning_rate': '1.702e-05', 'epoch': '0.2748'}

 27%|██▋       | 2710/9864 [11:24:37<29:27:09, 14.82s/it]
 27%|██▋       | 2711/9864 [11:24:55<31:16:21, 15.74s/it]
 27%|██▋       | 2712/9864 [11:25:12<32:05:12, 16.15s/it]
 28%|██▊       | 2713/9864 [11:25:27<30:57:21, 15.58s/it]
 28%|██▊       | 2714/9864 [11:25:43<31:18:13, 15.76s/it]
 28%|██▊       | 2715/9864 [11:25:57<30:24:22, 15.31s/it]
 28%|██▊       | 2716/9864 [11:26:11<29:55:11, 15.07s/it]
 28%|██▊       | 2717/9864 [11:26:26<29:51:40, 15.04s/it]
 28%|██▊       | 2718/9864 [11:26:45<31:58:11, 16.11s/it]
 28%|██▊       | 2719/9864 [11:27:02<32:27:14, 16.35s/it]
 28%|██▊       | 2720/9864 [11:27:16<31:15:01, 15.75s/it]
                                                         
{'loss': '0.1661', 'grad_norm': '3.179', 'learning_rate': '1.7e-05', 'epoch': '0.2758'}

 28%|██▊       | 2720/9864 [11:27:16<31:15:01, 15.75s/it]
 28%|██▊       | 2721/9864 [11:27:31<30:42:44, 15.48s/it]
 28%|██▊       | 2722/9864 [11:27:46<30:12:37, 15.23s/it]
 28%|██▊       | 2723/9864 [11:28:02<30:51:30, 15.56s/it]
 28%|██▊       | 2724/9864 [11:28:16<30:07:50, 15.19s/it]
 28%|██▊       | 2725/9864 [11:28:35<31:49:38, 16.05s/it]
 28%|██▊       | 2726/9864 [11:28:49<30:52:40, 15.57s/it]
 28%|██▊       | 2727/9864 [11:29:05<31:19:07, 15.80s/it]
 28%|██▊       | 2728/9864 [11:29:19<30:17:11, 15.28s/it]
 28%|██▊       | 2729/9864 [11:29:34<29:49:22, 15.05s/it]
 28%|██▊       | 2730/9864 [11:29:49<29:37:13, 14.95s/it]
                                                         
{'loss': '0.1785', 'grad_norm': '4.557', 'learning_rate': '1.698e-05', 'epoch': '0.2768'}

 28%|██▊       | 2730/9864 [11:29:49<29:37:13, 14.95s/it]
 28%|██▊       | 2731/9864 [11:30:05<30:18:06, 15.29s/it]
 28%|██▊       | 2732/9864 [11:30:19<29:56:45, 15.12s/it]
 28%|██▊       | 2733/9864 [11:30:38<31:50:24, 16.07s/it]
 28%|██▊       | 2734/9864 [11:30:52<30:55:41, 15.62s/it]
 28%|██▊       | 2735/9864 [11:31:09<31:24:13, 15.86s/it]
 28%|██▊       | 2736/9864 [11:31:23<30:23:31, 15.35s/it]
 28%|██▊       | 2737/9864 [11:31:37<29:55:40, 15.12s/it]
 28%|██▊       | 2738/9864 [11:31:54<30:37:22, 15.47s/it]
 28%|██▊       | 2739/9864 [11:32:09<30:19:03, 15.32s/it]
 28%|██▊       | 2740/9864 [11:32:25<30:43:41, 15.53s/it]
                                                         
{'loss': '0.1631', 'grad_norm': '2.765', 'learning_rate': '1.695e-05', 'epoch': '0.2778'}

 28%|██▊       | 2740/9864 [11:32:25<30:43:41, 15.53s/it]
 28%|██▊       | 2741/9864 [11:32:41<31:24:43, 15.88s/it]
 28%|██▊       | 2742/9864 [11:32:57<31:29:10, 15.92s/it]
 28%|██▊       | 2743/9864 [11:33:12<30:39:48, 15.50s/it]
 28%|██▊       | 2744/9864 [11:33:26<29:59:41, 15.17s/it]
 28%|██▊       | 2745/9864 [11:33:41<29:50:45, 15.09s/it]
 28%|██▊       | 2746/9864 [11:33:57<30:30:26, 15.43s/it]
 28%|██▊       | 2747/9864 [11:34:12<29:58:54, 15.17s/it]
 28%|██▊       | 2748/9864 [11:34:28<30:34:46, 15.47s/it]
 28%|██▊       | 2749/9864 [11:34:45<31:12:27, 15.79s/it]
 28%|██▊       | 2750/9864 [11:35:01<31:27:54, 15.92s/it]
                                                         
{'loss': '0.1655', 'grad_norm': '2.147', 'learning_rate': '1.693e-05', 'epoch': '0.2788'}

 28%|██▊       | 2750/9864 [11:35:01<31:27:54, 15.92s/it]
 28%|██▊       | 2751/9864 [11:35:16<30:39:54, 15.52s/it]
 28%|██▊       | 2752/9864 [11:35:30<29:57:03, 15.16s/it]
 28%|██▊       | 2753/9864 [11:35:44<29:27:45, 14.92s/it]
 28%|██▊       | 2754/9864 [11:36:00<30:10:11, 15.28s/it]
 28%|██▊       | 2755/9864 [11:36:15<29:59:31, 15.19s/it]
 28%|██▊       | 2756/9864 [11:36:31<30:22:01, 15.38s/it]
 28%|██▊       | 2757/9864 [11:36:46<29:48:08, 15.10s/it]
 28%|██▊       | 2758/9864 [11:37:04<31:37:31, 16.02s/it]
 28%|██▊       | 2759/9864 [11:37:19<31:01:05, 15.72s/it]
 28%|██▊       | 2760/9864 [11:37:33<30:24:22, 15.41s/it]
                                                         
{'loss': '0.1649', 'grad_norm': '3.08', 'learning_rate': '1.69e-05', 'epoch': '0.2798'}

 28%|██▊       | 2760/9864 [11:37:33<30:24:22, 15.41s/it]
 28%|██▊       | 2761/9864 [11:37:48<29:39:32, 15.03s/it]
 28%|██▊       | 2762/9864 [11:38:04<30:29:06, 15.45s/it]
 28%|██▊       | 2763/9864 [11:38:20<30:39:18, 15.54s/it]
 28%|██▊       | 2764/9864 [11:38:34<30:04:55, 15.25s/it]
 28%|██▊       | 2765/9864 [11:38:49<29:47:37, 15.11s/it]
 28%|██▊       | 2766/9864 [11:39:08<31:54:21, 16.18s/it]
 28%|██▊       | 2767/9864 [11:39:22<30:37:53, 15.54s/it]
 28%|██▊       | 2768/9864 [11:39:36<29:48:25, 15.12s/it]
 28%|██▊       | 2769/9864 [11:39:51<29:36:05, 15.02s/it]
 28%|██▊       | 2770/9864 [11:40:08<30:37:39, 15.54s/it]
                                                         
{'loss': '0.1623', 'grad_norm': '4.967', 'learning_rate': '1.688e-05', 'epoch': '0.2808'}

 28%|██▊       | 2770/9864 [11:40:08<30:37:39, 15.54s/it]
 28%|██▊       | 2771/9864 [11:40:24<31:05:39, 15.78s/it]
 28%|██▊       | 2772/9864 [11:40:38<30:12:13, 15.33s/it]
 28%|██▊       | 2773/9864 [11:40:54<30:28:55, 15.48s/it]
 28%|██▊       | 2774/9864 [11:41:10<31:00:29, 15.74s/it]
 28%|██▊       | 2775/9864 [11:41:25<30:21:28, 15.42s/it]
 28%|██▊       | 2776/9864 [11:41:39<29:32:16, 15.00s/it]
 28%|██▊       | 2777/9864 [11:41:55<30:18:59, 15.40s/it]
 28%|██▊       | 2778/9864 [11:42:11<30:42:53, 15.60s/it]
 28%|██▊       | 2779/9864 [11:42:27<30:31:07, 15.51s/it]
 28%|██▊       | 2780/9864 [11:42:42<30:39:37, 15.58s/it]
                                                         
{'loss': '0.1578', 'grad_norm': '4.413', 'learning_rate': '1.686e-05', 'epoch': '0.2819'}

 28%|██▊       | 2780/9864 [11:42:42<30:39:37, 15.58s/it]
 28%|██▊       | 2781/9864 [11:42:57<30:05:26, 15.29s/it]
 28%|██▊       | 2782/9864 [11:43:14<31:12:14, 15.86s/it]
 28%|██▊       | 2783/9864 [11:43:29<30:24:33, 15.46s/it]
 28%|██▊       | 2784/9864 [11:43:43<29:41:37, 15.10s/it]
 28%|██▊       | 2785/9864 [11:43:59<30:24:47, 15.47s/it]
 28%|██▊       | 2786/9864 [11:44:16<31:00:36, 15.77s/it]
 28%|██▊       | 2787/9864 [11:44:30<30:17:41, 15.41s/it]
 28%|██▊       | 2788/9864 [11:44:47<31:15:07, 15.90s/it]
 28%|██▊       | 2789/9864 [11:45:03<31:10:48, 15.87s/it]
 28%|██▊       | 2790/9864 [11:45:18<30:19:57, 15.44s/it]
                                                         
{'loss': '0.1602', 'grad_norm': '2.946', 'learning_rate': '1.683e-05', 'epoch': '0.2829'}

 28%|██▊       | 2790/9864 [11:45:18<30:19:57, 15.44s/it]
 28%|██▊       | 2791/9864 [11:45:32<29:46:46, 15.16s/it]
 28%|██▊       | 2792/9864 [11:45:47<29:47:17, 15.16s/it]
 28%|██▊       | 2793/9864 [11:46:04<30:43:26, 15.64s/it]
 28%|██▊       | 2794/9864 [11:46:20<30:56:35, 15.76s/it]
 28%|██▊       | 2795/9864 [11:46:34<29:49:16, 15.19s/it]
 28%|██▊       | 2796/9864 [11:46:52<31:39:16, 16.12s/it]
 28%|██▊       | 2797/9864 [11:47:07<30:49:34, 15.70s/it]
 28%|██▊       | 2798/9864 [11:47:21<29:58:14, 15.27s/it]
 28%|██▊       | 2799/9864 [11:47:36<29:23:17, 14.97s/it]
 28%|██▊       | 2800/9864 [11:47:51<29:23:59, 14.98s/it]
                                                         
{'loss': '0.1488', 'grad_norm': '2.983', 'learning_rate': '1.681e-05', 'epoch': '0.2839'}

 28%|██▊       | 2800/9864 [11:47:51<29:23:59, 14.98s/it]
 28%|██▊       | 2801/9864 [11:48:07<30:14:46, 15.42s/it]
 28%|██▊       | 2802/9864 [11:48:23<30:41:47, 15.65s/it]
 28%|██▊       | 2803/9864 [11:48:38<30:11:27, 15.39s/it]
 28%|██▊       | 2804/9864 [11:48:56<31:43:35, 16.18s/it]
 28%|██▊       | 2805/9864 [11:49:10<30:19:48, 15.47s/it]
 28%|██▊       | 2806/9864 [11:49:25<29:52:32, 15.24s/it]
 28%|██▊       | 2807/9864 [11:49:39<29:36:59, 15.11s/it]
 28%|██▊       | 2808/9864 [11:49:54<29:25:40, 15.01s/it]
 28%|██▊       | 2809/9864 [11:50:10<30:00:33, 15.31s/it]
 28%|██▊       | 2810/9864 [11:50:26<30:23:50, 15.51s/it]
                                                         
{'loss': '0.1591', 'grad_norm': '4.775', 'learning_rate': '1.678e-05', 'epoch': '0.2849'}

 28%|██▊       | 2810/9864 [11:50:26<30:23:50, 15.51s/it]
 28%|██▊       | 2811/9864 [11:50:42<30:43:09, 15.68s/it]
 29%|██▊       | 2812/9864 [11:50:58<31:03:57, 15.86s/it]
 29%|██▊       | 2813/9864 [11:51:13<30:14:59, 15.44s/it]
 29%|██▊       | 2814/9864 [11:51:28<29:49:15, 15.23s/it]
 29%|██▊       | 2815/9864 [11:51:42<29:29:05, 15.06s/it]
 29%|██▊       | 2816/9864 [11:51:57<29:11:26, 14.91s/it]
 29%|██▊       | 2817/9864 [11:52:12<29:04:55, 14.86s/it]
 29%|██▊       | 2818/9864 [11:52:29<30:37:32, 15.65s/it]
 29%|██▊       | 2819/9864 [11:52:44<29:59:06, 15.32s/it]
 29%|██▊       | 2820/9864 [11:53:02<31:41:49, 16.20s/it]
                                                         
{'loss': '0.1399', 'grad_norm': '2.85', 'learning_rate': '1.676e-05', 'epoch': '0.2859'}

 29%|██▊       | 2820/9864 [11:53:02<31:41:49, 16.20s/it]
 29%|██▊       | 2821/9864 [11:53:17<31:00:45, 15.85s/it]
 29%|██▊       | 2822/9864 [11:53:31<30:04:55, 15.38s/it]
 29%|██▊       | 2823/9864 [11:53:46<29:43:13, 15.20s/it]
 29%|██▊       | 2824/9864 [11:54:01<29:37:04, 15.15s/it]
 29%|██▊       | 2825/9864 [11:54:16<29:33:40, 15.12s/it]
 29%|██▊       | 2826/9864 [11:54:34<31:25:59, 16.08s/it]
 29%|██▊       | 2827/9864 [11:54:50<31:25:19, 16.07s/it]
 29%|██▊       | 2828/9864 [11:55:08<32:02:57, 16.40s/it]
 29%|██▊       | 2829/9864 [11:55:22<30:43:15, 15.72s/it]
 29%|██▊       | 2830/9864 [11:55:36<29:46:40, 15.24s/it]
                                                         
{'loss': '0.1595', 'grad_norm': '3.156', 'learning_rate': '1.674e-05', 'epoch': '0.2869'}

 29%|██▊       | 2830/9864 [11:55:36<29:46:40, 15.24s/it]
 29%|██▊       | 2831/9864 [11:55:51<29:26:58, 15.07s/it]
 29%|██▊       | 2832/9864 [11:56:05<29:13:41, 14.96s/it]
 29%|██▊       | 2833/9864 [11:56:24<31:18:55, 16.03s/it]
 29%|██▊       | 2834/9864 [11:56:38<30:18:23, 15.52s/it]
 29%|██▊       | 2835/9864 [11:56:54<30:40:24, 15.71s/it]
 29%|██▉       | 2836/9864 [11:57:09<30:03:39, 15.40s/it]
 29%|██▉       | 2837/9864 [11:57:26<30:49:10, 15.79s/it]
 29%|██▉       | 2838/9864 [11:57:40<29:55:24, 15.33s/it]
 29%|██▉       | 2839/9864 [11:57:55<29:48:43, 15.28s/it]
 29%|██▉       | 2840/9864 [11:58:12<30:41:45, 15.73s/it]
                                                         
{'loss': '0.1498', 'grad_norm': '1.983', 'learning_rate': '1.671e-05', 'epoch': '0.2879'}

 29%|██▉       | 2840/9864 [11:58:12<30:41:45, 15.73s/it]
 29%|██▉       | 2841/9864 [11:58:26<29:54:09, 15.33s/it]
 29%|██▉       | 2842/9864 [11:58:42<30:17:12, 15.53s/it]
 29%|██▉       | 2843/9864 [11:58:59<30:58:48, 15.89s/it]
 29%|██▉       | 2844/9864 [11:59:16<31:29:55, 16.15s/it]
 29%|██▉       | 2845/9864 [11:59:30<30:19:30, 15.55s/it]
 29%|██▉       | 2846/9864 [11:59:44<29:21:05, 15.06s/it]
 29%|██▉       | 2847/9864 [11:59:59<29:10:34, 14.97s/it]
 29%|██▉       | 2848/9864 [12:00:14<29:42:04, 15.24s/it]
 29%|██▉       | 2849/9864 [12:00:29<29:14:21, 15.01s/it]
 29%|██▉       | 2850/9864 [12:00:47<31:04:48, 15.95s/it]
                                                         
{'loss': '0.1673', 'grad_norm': '4.588', 'learning_rate': '1.669e-05', 'epoch': '0.289'}

 29%|██▉       | 2850/9864 [12:00:47<31:04:48, 15.95s/it]
 29%|██▉       | 2851/9864 [12:01:01<29:45:56, 15.28s/it]
 29%|██▉       | 2852/9864 [12:01:17<30:32:12, 15.68s/it]
 29%|██▉       | 2853/9864 [12:01:32<29:40:11, 15.23s/it]
 29%|██▉       | 2854/9864 [12:01:45<28:50:25, 14.81s/it]
 29%|██▉       | 2855/9864 [12:02:00<28:43:37, 14.76s/it]
 29%|██▉       | 2856/9864 [12:02:15<28:34:35, 14.68s/it]
 29%|██▉       | 2857/9864 [12:02:33<30:57:23, 15.90s/it]
 29%|██▉       | 2858/9864 [12:02:49<31:06:14, 15.98s/it]
 29%|██▉       | 2859/9864 [12:03:06<31:16:35, 16.07s/it]
 29%|██▉       | 2860/9864 [12:03:20<30:06:24, 15.47s/it]
                                                         
{'loss': '0.148', 'grad_norm': '2.469', 'learning_rate': '1.666e-05', 'epoch': '0.29'}

 29%|██▉       | 2860/9864 [12:03:20<30:06:24, 15.47s/it]
 29%|██▉       | 2861/9864 [12:03:34<29:22:13, 15.10s/it]
 29%|██▉       | 2862/9864 [12:03:48<28:52:26, 14.85s/it]
 29%|██▉       | 2863/9864 [12:04:03<28:44:36, 14.78s/it]
 29%|██▉       | 2864/9864 [12:04:19<29:35:51, 15.22s/it]
 29%|██▉       | 2865/9864 [12:04:35<30:13:22, 15.55s/it]
 29%|██▉       | 2866/9864 [12:04:53<31:07:39, 16.01s/it]
 29%|██▉       | 2867/9864 [12:05:11<32:28:45, 16.71s/it]
 29%|██▉       | 2868/9864 [12:05:26<31:43:07, 16.32s/it]
 29%|██▉       | 2869/9864 [12:05:40<30:19:13, 15.60s/it]
 29%|██▉       | 2870/9864 [12:05:56<30:13:30, 15.56s/it]
                                                         
{'loss': '0.1673', 'grad_norm': '1.699', 'learning_rate': '1.664e-05', 'epoch': '0.291'}

 29%|██▉       | 2870/9864 [12:05:56<30:13:30, 15.56s/it]
 29%|██▉       | 2871/9864 [12:06:12<30:38:01, 15.77s/it]
 29%|██▉       | 2872/9864 [12:06:28<30:49:18, 15.87s/it]
 29%|██▉       | 2873/9864 [12:06:45<31:15:14, 16.09s/it]
 29%|██▉       | 2874/9864 [12:07:01<31:06:14, 16.02s/it]
 29%|██▉       | 2875/9864 [12:07:15<29:59:52, 15.45s/it]
 29%|██▉       | 2876/9864 [12:07:31<30:15:08, 15.59s/it]
 29%|██▉       | 2877/9864 [12:07:45<29:29:08, 15.19s/it]
 29%|██▉       | 2878/9864 [12:07:59<28:51:11, 14.87s/it]
 29%|██▉       | 2879/9864 [12:08:13<28:26:46, 14.66s/it]
 29%|██▉       | 2880/9864 [12:08:31<30:14:46, 15.59s/it]
                                                         
{'loss': '0.1586', 'grad_norm': '4.821', 'learning_rate': '1.661e-05', 'epoch': '0.292'}

 29%|██▉       | 2880/9864 [12:08:31<30:14:46, 15.59s/it]
 29%|██▉       | 2881/9864 [12:08:45<29:13:12, 15.06s/it]
 29%|██▉       | 2882/9864 [12:09:00<29:35:44, 15.26s/it]
 29%|██▉       | 2883/9864 [12:09:17<30:15:35, 15.60s/it]
 29%|██▉       | 2884/9864 [12:09:31<29:17:57, 15.11s/it]
 29%|██▉       | 2885/9864 [12:09:45<28:42:08, 14.81s/it]
 29%|██▉       | 2886/9864 [12:09:59<28:18:00, 14.60s/it]
 29%|██▉       | 2887/9864 [12:10:16<29:32:09, 15.24s/it]
 29%|██▉       | 2888/9864 [12:10:31<29:24:15, 15.17s/it]
 29%|██▉       | 2889/9864 [12:10:47<29:50:13, 15.40s/it]
 29%|██▉       | 2890/9864 [12:11:05<31:18:42, 16.16s/it]
                                                         
{'loss': '0.1572', 'grad_norm': '2.657', 'learning_rate': '1.659e-05', 'epoch': '0.293'}

 29%|██▉       | 2890/9864 [12:11:05<31:18:42, 16.16s/it]
 29%|██▉       | 2891/9864 [12:11:19<30:12:07, 15.59s/it]
 29%|██▉       | 2892/9864 [12:11:33<29:29:42, 15.23s/it]
 29%|██▉       | 2893/9864 [12:11:47<28:49:30, 14.89s/it]
 29%|██▉       | 2894/9864 [12:12:03<29:27:30, 15.22s/it]
 29%|██▉       | 2895/9864 [12:12:18<29:15:13, 15.11s/it]
 29%|██▉       | 2896/9864 [12:12:32<28:44:39, 14.85s/it]
 29%|██▉       | 2897/9864 [12:12:48<29:23:21, 15.19s/it]
 29%|██▉       | 2898/9864 [12:13:06<30:49:08, 15.93s/it]
 29%|██▉       | 2899/9864 [12:13:20<29:52:17, 15.44s/it]
 29%|██▉       | 2900/9864 [12:13:35<29:28:48, 15.24s/it]
                                                         
{'loss': '0.1557', 'grad_norm': '1.808', 'learning_rate': '1.656e-05', 'epoch': '0.294'}

 29%|██▉       | 2900/9864 [12:13:35<29:28:48, 15.24s/it]
 29%|██▉       | 2901/9864 [12:13:49<28:45:36, 14.87s/it]
 29%|██▉       | 2902/9864 [12:14:04<28:32:31, 14.76s/it]
 29%|██▉       | 2903/9864 [12:14:19<29:06:19, 15.05s/it]
 29%|██▉       | 2904/9864 [12:14:34<28:58:40, 14.99s/it]
 29%|██▉       | 2905/9864 [12:14:50<29:25:08, 15.22s/it]
 29%|██▉       | 2906/9864 [12:15:08<30:54:48, 15.99s/it]
 29%|██▉       | 2907/9864 [12:15:22<30:04:01, 15.56s/it]
 29%|██▉       | 2908/9864 [12:15:37<29:28:19, 15.25s/it]
 29%|██▉       | 2909/9864 [12:15:51<28:40:25, 14.84s/it]
 30%|██▉       | 2910/9864 [12:16:07<29:27:20, 15.25s/it]
                                                         
{'loss': '0.1683', 'grad_norm': '4.116', 'learning_rate': '1.654e-05', 'epoch': '0.295'}

 30%|██▉       | 2910/9864 [12:16:07<29:27:20, 15.25s/it]
 30%|██▉       | 2911/9864 [12:16:21<28:48:47, 14.92s/it]
 30%|██▉       | 2912/9864 [12:16:37<29:37:18, 15.34s/it]
 30%|██▉       | 2913/9864 [12:16:53<29:40:31, 15.37s/it]
 30%|██▉       | 2914/9864 [12:17:09<30:14:53, 15.67s/it]
 30%|██▉       | 2915/9864 [12:17:23<29:09:50, 15.11s/it]
 30%|██▉       | 2916/9864 [12:17:38<28:47:48, 14.92s/it]
 30%|██▉       | 2917/9864 [12:17:52<28:20:01, 14.68s/it]
 30%|██▉       | 2918/9864 [12:18:08<29:08:02, 15.10s/it]
 30%|██▉       | 2919/9864 [12:18:22<28:38:48, 14.85s/it]
 30%|██▉       | 2920/9864 [12:18:40<30:32:20, 15.83s/it]
                                                         
{'loss': '0.1712', 'grad_norm': '6.093', 'learning_rate': '1.651e-05', 'epoch': '0.2961'}

 30%|██▉       | 2920/9864 [12:18:40<30:32:20, 15.83s/it]
 30%|██▉       | 2921/9864 [12:18:56<30:47:27, 15.97s/it]
 30%|██▉       | 2922/9864 [12:19:11<29:48:19, 15.46s/it]
 30%|██▉       | 2923/9864 [12:19:25<29:02:35, 15.06s/it]
 30%|██▉       | 2924/9864 [12:19:39<28:35:43, 14.83s/it]
 30%|██▉       | 2925/9864 [12:19:55<28:58:54, 15.04s/it]
 30%|██▉       | 2926/9864 [12:20:09<28:41:40, 14.89s/it]
 30%|██▉       | 2927/9864 [12:20:25<29:19:01, 15.21s/it]
 30%|██▉       | 2928/9864 [12:20:41<29:45:13, 15.44s/it]
 30%|██▉       | 2929/9864 [12:20:57<30:12:31, 15.68s/it]
 30%|██▉       | 2930/9864 [12:21:12<29:37:16, 15.38s/it]
                                                         
{'loss': '0.161', 'grad_norm': '4.736', 'learning_rate': '1.649e-05', 'epoch': '0.2971'}

 30%|██▉       | 2930/9864 [12:21:12<29:37:16, 15.38s/it]
 30%|██▉       | 2931/9864 [12:21:26<29:02:01, 15.08s/it]
 30%|██▉       | 2932/9864 [12:21:41<28:39:14, 14.88s/it]
 30%|██▉       | 2933/9864 [12:21:55<28:21:59, 14.73s/it]
 30%|██▉       | 2934/9864 [12:22:11<28:42:36, 14.91s/it]
 30%|██▉       | 2935/9864 [12:22:27<29:33:35, 15.36s/it]
 30%|██▉       | 2936/9864 [12:22:45<31:15:44, 16.24s/it]
 30%|██▉       | 2937/9864 [12:23:00<30:08:11, 15.66s/it]
 30%|██▉       | 2938/9864 [12:23:14<29:32:20, 15.35s/it]
 30%|██▉       | 2939/9864 [12:23:29<28:56:56, 15.05s/it]
 30%|██▉       | 2940/9864 [12:23:43<28:41:01, 14.91s/it]
                                                         
{'loss': '0.166', 'grad_norm': '3.479', 'learning_rate': '1.646e-05', 'epoch': '0.2981'}

 30%|██▉       | 2940/9864 [12:23:43<28:41:01, 14.91s/it]
 30%|██▉       | 2941/9864 [12:23:59<29:23:18, 15.28s/it]
 30%|██▉       | 2942/9864 [12:24:13<28:30:20, 14.83s/it]
 30%|██▉       | 2943/9864 [12:24:29<29:02:34, 15.11s/it]
 30%|██▉       | 2944/9864 [12:24:45<29:24:00, 15.29s/it]
 30%|██▉       | 2945/9864 [12:24:59<28:54:10, 15.04s/it]
 30%|██▉       | 2946/9864 [12:25:14<28:36:48, 14.89s/it]
 30%|██▉       | 2947/9864 [12:25:28<28:09:39, 14.66s/it]
 30%|██▉       | 2948/9864 [12:25:41<27:34:32, 14.35s/it]
 30%|██▉       | 2949/9864 [12:25:57<28:28:32, 14.82s/it]
 30%|██▉       | 2950/9864 [12:26:13<29:14:14, 15.22s/it]
                                                         
{'loss': '0.1627', 'grad_norm': '3.375', 'learning_rate': '1.644e-05', 'epoch': '0.2991'}

 30%|██▉       | 2950/9864 [12:26:13<29:14:14, 15.22s/it]
 30%|██▉       | 2951/9864 [12:26:29<29:41:14, 15.46s/it]
 30%|██▉       | 2952/9864 [12:26:45<29:47:00, 15.51s/it]
 30%|██▉       | 2953/9864 [12:26:59<29:08:56, 15.18s/it]
 30%|██▉       | 2954/9864 [12:27:14<28:39:26, 14.93s/it]
 30%|██▉       | 2955/9864 [12:27:28<28:11:28, 14.69s/it]
 30%|██▉       | 2956/9864 [12:27:42<27:58:46, 14.58s/it]
 30%|██▉       | 2957/9864 [12:28:00<29:43:47, 15.50s/it]
 30%|██▉       | 2958/9864 [12:28:14<28:57:42, 15.10s/it]
 30%|██▉       | 2959/9864 [12:28:29<29:05:18, 15.17s/it]
 30%|███       | 2960/9864 [12:28:45<29:28:12, 15.37s/it]
                                                         
{'loss': '0.1465', 'grad_norm': '3.868', 'learning_rate': '1.641e-05', 'epoch': '0.3001'}

 30%|███       | 2960/9864 [12:28:45<29:28:12, 15.37s/it]
 30%|███       | 2961/9864 [12:28:59<28:39:15, 14.94s/it]
 30%|███       | 2962/9864 [12:29:14<28:21:04, 14.79s/it]
 30%|███       | 2963/9864 [12:29:28<28:11:35, 14.71s/it]
 30%|███       | 2964/9864 [12:29:44<29:08:38, 15.21s/it]
 30%|███       | 2965/9864 [12:29:58<28:26:44, 14.84s/it]
 30%|███       | 2966/9864 [12:30:14<28:51:11, 15.06s/it]
 30%|███       | 2967/9864 [12:30:29<29:03:54, 15.17s/it]
 30%|███       | 2968/9864 [12:30:45<29:31:40, 15.41s/it]
 30%|███       | 2969/9864 [12:31:00<28:47:43, 15.03s/it]
 30%|███       | 2970/9864 [12:31:14<28:24:55, 14.84s/it]
                                                         
{'loss': '0.1614', 'grad_norm': '2.424', 'learning_rate': '1.639e-05', 'epoch': '0.3011'}

 30%|███       | 2970/9864 [12:31:14<28:24:55, 14.84s/it]
 30%|███       | 2971/9864 [12:31:30<28:52:12, 15.08s/it]
 30%|███       | 2972/9864 [12:31:43<28:00:41, 14.63s/it]
 30%|███       | 2973/9864 [12:31:59<28:37:04, 14.95s/it]
 30%|███       | 2974/9864 [12:32:15<29:14:46, 15.28s/it]
 30%|███       | 2975/9864 [12:32:29<28:31:25, 14.91s/it]
 30%|███       | 2976/9864 [12:32:45<29:00:45, 15.16s/it]
 30%|███       | 2977/9864 [12:32:59<28:29:02, 14.89s/it]
 30%|███       | 2978/9864 [12:33:13<28:06:40, 14.70s/it]
 30%|███       | 2979/9864 [12:33:29<28:54:27, 15.12s/it]
 30%|███       | 2980/9864 [12:33:45<29:17:54, 15.32s/it]
                                                         
{'loss': '0.1525', 'grad_norm': '3.017', 'learning_rate': '1.636e-05', 'epoch': '0.3021'}

 30%|███       | 2980/9864 [12:33:45<29:17:54, 15.32s/it]
 30%|███       | 2981/9864 [12:34:01<29:23:34, 15.37s/it]
 30%|███       | 2982/9864 [12:34:15<28:46:38, 15.05s/it]
 30%|███       | 2983/9864 [12:34:29<28:19:51, 14.82s/it]
 30%|███       | 2984/9864 [12:34:45<29:06:03, 15.23s/it]
 30%|███       | 2985/9864 [12:35:00<28:32:17, 14.93s/it]
 30%|███       | 2986/9864 [12:35:14<28:02:45, 14.68s/it]
 30%|███       | 2987/9864 [12:35:30<28:56:54, 15.15s/it]
 30%|███       | 2988/9864 [12:35:48<30:40:10, 16.06s/it]
 30%|███       | 2989/9864 [12:36:02<29:38:42, 15.52s/it]
 30%|███       | 2990/9864 [12:36:16<28:45:31, 15.06s/it]
                                                         
{'loss': '0.1459', 'grad_norm': '1.952', 'learning_rate': '1.634e-05', 'epoch': '0.3031'}

 30%|███       | 2990/9864 [12:36:16<28:45:31, 15.06s/it]
 30%|███       | 2991/9864 [12:36:33<29:24:37, 15.40s/it]
 30%|███       | 2992/9864 [12:36:47<28:42:26, 15.04s/it]
 30%|███       | 2993/9864 [12:37:01<28:23:10, 14.87s/it]
 30%|███       | 2994/9864 [12:37:16<28:05:39, 14.72s/it]
 30%|███       | 2995/9864 [12:37:32<28:55:59, 15.16s/it]
 30%|███       | 2996/9864 [12:37:47<29:07:51, 15.27s/it]
 30%|███       | 2997/9864 [12:38:03<29:28:56, 15.46s/it]
 30%|███       | 2998/9864 [12:38:17<28:41:04, 15.04s/it]
 30%|███       | 2999/9864 [12:38:33<28:54:06, 15.16s/it]
 30%|███       | 3000/9864 [12:38:47<28:21:32, 14.87s/it]
                                                         
{'loss': '0.157', 'grad_norm': '2.12', 'learning_rate': '1.631e-05', 'epoch': '0.3042'}

 30%|███       | 3000/9864 [12:38:47<28:21:32, 14.87s/it]

Writing model shards:   0%|          | 0/1 [00:00<?, ?it/s]

Writing model shards: 100%|██████████| 1/1 [00:10<00:00, 10.00s/it]
Writing model shards: 100%|██████████| 1/1 [00:10<00:00, 10.01s/it]
/home/dungpq6/Project/SpatialStack/src/qwen_vl/model/geometry_encoders/vggt_encoder.py:285: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(dtype=dtype):

 30%|███       | 3001/9864 [12:40:00<61:24:34, 32.21s/it]
 30%|███       | 3002/9864 [12:40:16<52:12:32, 27.39s/it]
 30%|███       | 3003/9864 [12:40:34<47:01:40, 24.68s/it]
 30%|███       | 3004/9864 [12:40:50<41:44:32, 21.91s/it]
 30%|███       | 3005/9864 [12:41:06<38:48:49, 20.37s/it]
 30%|███       | 3006/9864 [12:41:22<36:22:39, 19.10s/it]
 30%|███       | 3007/9864 [12:41:36<33:28:25, 17.57s/it]
 30%|███       | 3008/9864 [12:41:51<31:28:05, 16.52s/it]
 31%|███       | 3009/9864 [12:42:05<30:10:39, 15.85s/it]
 31%|███       | 3010/9864 [12:42:19<28:58:17, 15.22s/it]
                                                         
{'loss': '0.1673', 'grad_norm': '5.109', 'learning_rate': '1.629e-05', 'epoch': '0.3052'}

 31%|███       | 3010/9864 [12:42:19<28:58:17, 15.22s/it]
 31%|███       | 3011/9864 [12:42:36<30:14:03, 15.88s/it]
 31%|███       | 3012/9864 [12:42:50<29:23:37, 15.44s/it]
 31%|███       | 3013/9864 [12:43:06<29:33:31, 15.53s/it]
 31%|███       | 3014/9864 [12:43:22<29:31:05, 15.51s/it]
 31%|███       | 3015/9864 [12:43:35<28:29:37, 14.98s/it]
 31%|███       | 3016/9864 [12:43:49<27:50:17, 14.63s/it]
 31%|███       | 3017/9864 [12:44:03<27:36:29, 14.52s/it]
 31%|███       | 3018/9864 [12:44:19<28:27:48, 14.97s/it]
 31%|███       | 3019/9864 [12:44:35<28:56:28, 15.22s/it]
 31%|███       | 3020/9864 [12:44:51<29:22:02, 15.45s/it]
                                                         
{'loss': '0.1427', 'grad_norm': '1.823', 'learning_rate': '1.626e-05', 'epoch': '0.3062'}

 31%|███       | 3020/9864 [12:44:51<29:22:02, 15.45s/it]
 31%|███       | 3021/9864 [12:45:06<28:51:35, 15.18s/it]
 31%|███       | 3022/9864 [12:45:21<29:08:08, 15.33s/it]
 31%|███       | 3023/9864 [12:45:35<28:22:09, 14.93s/it]
 31%|███       | 3024/9864 [12:45:50<28:05:14, 14.78s/it]
 31%|███       | 3025/9864 [12:46:06<28:58:28, 15.25s/it]
 31%|███       | 3026/9864 [12:46:21<28:25:27, 14.96s/it]
 31%|███       | 3027/9864 [12:46:36<28:51:26, 15.19s/it]
 31%|███       | 3028/9864 [12:46:51<28:22:58, 14.95s/it]
 31%|███       | 3029/9864 [12:47:06<28:53:49, 15.22s/it]
 31%|███       | 3030/9864 [12:47:22<29:17:01, 15.43s/it]
                                                         
{'loss': '0.1494', 'grad_norm': '3.893', 'learning_rate': '1.624e-05', 'epoch': '0.3072'}

 31%|███       | 3030/9864 [12:47:22<29:17:01, 15.43s/it]
 31%|███       | 3031/9864 [12:47:37<28:45:48, 15.15s/it]
 31%|███       | 3032/9864 [12:47:51<28:14:56, 14.89s/it]
 31%|███       | 3033/9864 [12:48:07<28:47:46, 15.18s/it]
 31%|███       | 3034/9864 [12:48:21<28:13:59, 14.88s/it]
 31%|███       | 3035/9864 [12:48:38<29:05:59, 15.34s/it]
 31%|███       | 3036/9864 [12:48:52<28:20:27, 14.94s/it]
 31%|███       | 3037/9864 [12:49:09<29:53:20, 15.76s/it]
 31%|███       | 3038/9864 [12:49:24<29:01:31, 15.31s/it]
 31%|███       | 3039/9864 [12:49:37<28:12:17, 14.88s/it]
 31%|███       | 3040/9864 [12:49:51<27:42:19, 14.62s/it]
                                                         
{'loss': '0.1522', 'grad_norm': '2.442', 'learning_rate': '1.621e-05', 'epoch': '0.3082'}

 31%|███       | 3040/9864 [12:49:51<27:42:19, 14.62s/it]
 31%|███       | 3041/9864 [12:50:07<28:18:10, 14.93s/it]
 31%|███       | 3042/9864 [12:50:21<27:53:39, 14.72s/it]
 31%|███       | 3043/9864 [12:50:37<28:28:28, 15.03s/it]
 31%|███       | 3044/9864 [12:50:51<27:56:31, 14.75s/it]
 31%|███       | 3045/9864 [12:51:07<28:40:40, 15.14s/it]
 31%|███       | 3046/9864 [12:51:23<28:58:54, 15.30s/it]
 31%|███       | 3047/9864 [12:51:37<28:27:32, 15.03s/it]
 31%|███       | 3048/9864 [12:51:51<27:54:35, 14.74s/it]
 31%|███       | 3049/9864 [12:52:07<28:24:14, 15.00s/it]
 31%|███       | 3050/9864 [12:52:23<28:57:16, 15.30s/it]
                                                         
{'loss': '0.15', 'grad_norm': '3.393', 'learning_rate': '1.619e-05', 'epoch': '0.3092'}

 31%|███       | 3050/9864 [12:52:23<28:57:16, 15.30s/it]
 31%|███       | 3051/9864 [12:52:37<28:19:50, 14.97s/it]
 31%|███       | 3052/9864 [12:52:51<27:38:03, 14.60s/it]
 31%|███       | 3053/9864 [12:53:08<29:05:06, 15.37s/it]
 31%|███       | 3054/9864 [12:53:23<28:32:37, 15.09s/it]
 31%|███       | 3055/9864 [12:53:37<27:58:54, 14.79s/it]
 31%|███       | 3056/9864 [12:53:51<27:27:36, 14.52s/it]
 31%|███       | 3057/9864 [12:54:07<28:18:26, 14.97s/it]
 31%|███       | 3058/9864 [12:54:22<28:49:32, 15.25s/it]
 31%|███       | 3059/9864 [12:54:36<28:05:52, 14.86s/it]
 31%|███       | 3060/9864 [12:54:52<28:34:49, 15.12s/it]
                                                         
{'loss': '0.1566', 'grad_norm': '3.244', 'learning_rate': '1.616e-05', 'epoch': '0.3102'}

 31%|███       | 3060/9864 [12:54:52<28:34:49, 15.12s/it]
 31%|███       | 3061/9864 [12:55:07<28:11:21, 14.92s/it]
 31%|███       | 3062/9864 [12:55:21<28:03:25, 14.85s/it]
 31%|███       | 3063/9864 [12:55:35<27:40:01, 14.65s/it]
 31%|███       | 3064/9864 [12:55:50<27:27:47, 14.54s/it]
 31%|███       | 3065/9864 [12:56:05<28:09:06, 14.91s/it]
 31%|███       | 3066/9864 [12:56:20<27:43:05, 14.68s/it]
 31%|███       | 3067/9864 [12:56:38<29:34:58, 15.67s/it]
 31%|███       | 3068/9864 [12:56:54<29:48:06, 15.79s/it]
 31%|███       | 3069/9864 [12:57:08<28:52:32, 15.30s/it]
 31%|███       | 3070/9864 [12:57:22<28:15:09, 14.97s/it]
                                                         
{'loss': '0.1581', 'grad_norm': '2.389', 'learning_rate': '1.613e-05', 'epoch': '0.3113'}

 31%|███       | 3070/9864 [12:57:22<28:15:09, 14.97s/it]
 31%|███       | 3071/9864 [12:57:36<27:44:16, 14.70s/it]
 31%|███       | 3072/9864 [12:57:50<27:33:07, 14.60s/it]
 31%|███       | 3073/9864 [12:58:07<28:23:30, 15.05s/it]
 31%|███       | 3074/9864 [12:58:23<28:55:04, 15.33s/it]
 31%|███       | 3075/9864 [12:58:40<30:14:07, 16.03s/it]
 31%|███       | 3076/9864 [12:58:54<29:08:22, 15.45s/it]
 31%|███       | 3077/9864 [12:59:09<28:25:55, 15.08s/it]
 31%|███       | 3078/9864 [12:59:23<28:17:16, 15.01s/it]
 31%|███       | 3079/9864 [12:59:37<27:46:01, 14.73s/it]
 31%|███       | 3080/9864 [12:59:53<27:56:38, 14.83s/it]
                                                         
{'loss': '0.1652', 'grad_norm': '2.407', 'learning_rate': '1.611e-05', 'epoch': '0.3123'}

 31%|███       | 3080/9864 [12:59:53<27:56:38, 14.83s/it]
 31%|███       | 3081/9864 [13:00:08<28:22:25, 15.06s/it]
 31%|███       | 3082/9864 [13:00:26<29:52:29, 15.86s/it]
 31%|███▏      | 3083/9864 [13:00:40<28:54:49, 15.35s/it]
 31%|███▏      | 3084/9864 [13:00:56<29:15:33, 15.54s/it]
 31%|███▏      | 3085/9864 [13:01:10<28:24:29, 15.09s/it]
 31%|███▏      | 3086/9864 [13:01:24<28:01:56, 14.89s/it]
 31%|███▏      | 3087/9864 [13:01:39<27:37:21, 14.67s/it]
 31%|███▏      | 3088/9864 [13:01:53<27:27:57, 14.59s/it]
 31%|███▏      | 3089/9864 [13:02:11<29:14:45, 15.54s/it]
 31%|███▏      | 3090/9864 [13:02:26<29:20:28, 15.59s/it]
                                                         
{'loss': '0.1613', 'grad_norm': '2.143', 'learning_rate': '1.608e-05', 'epoch': '0.3133'}

 31%|███▏      | 3090/9864 [13:02:26<29:20:28, 15.59s/it]
 31%|███▏      | 3091/9864 [13:02:40<28:21:09, 15.07s/it]
 31%|███▏      | 3092/9864 [13:02:57<29:11:04, 15.51s/it]
 31%|███▏      | 3093/9864 [13:03:10<28:00:07, 14.89s/it]
 31%|███▏      | 3094/9864 [13:03:25<27:37:33, 14.69s/it]
 31%|███▏      | 3095/9864 [13:03:39<27:16:55, 14.51s/it]
 31%|███▏      | 3096/9864 [13:03:56<28:43:38, 15.28s/it]
 31%|███▏      | 3097/9864 [13:04:12<29:07:36, 15.50s/it]
 31%|███▏      | 3098/9864 [13:04:26<28:26:22, 15.13s/it]
 31%|███▏      | 3099/9864 [13:04:42<28:56:58, 15.41s/it]
 31%|███▏      | 3100/9864 [13:04:56<28:22:11, 15.10s/it]
                                                         
{'loss': '0.1435', 'grad_norm': '1.309', 'learning_rate': '1.606e-05', 'epoch': '0.3143'}

 31%|███▏      | 3100/9864 [13:04:56<28:22:11, 15.10s/it]
 31%|███▏      | 3101/9864 [13:05:10<27:41:59, 14.74s/it]
 31%|███▏      | 3102/9864 [13:05:25<27:25:32, 14.60s/it]
 31%|███▏      | 3103/9864 [13:05:39<27:08:19, 14.45s/it]
 31%|███▏      | 3104/9864 [13:05:54<27:41:40, 14.75s/it]
 31%|███▏      | 3105/9864 [13:06:10<28:23:24, 15.12s/it]
 31%|███▏      | 3106/9864 [13:06:26<28:49:55, 15.36s/it]
 31%|███▏      | 3107/9864 [13:06:42<29:00:15, 15.45s/it]
 32%|███▏      | 3108/9864 [13:06:56<28:04:14, 14.96s/it]
 32%|███▏      | 3109/9864 [13:07:10<27:43:23, 14.77s/it]
 32%|███▏      | 3110/9864 [13:07:24<27:13:42, 14.51s/it]
                                                         
{'loss': '0.1571', 'grad_norm': '1.913', 'learning_rate': '1.603e-05', 'epoch': '0.3153'}

 32%|███▏      | 3110/9864 [13:07:24<27:13:42, 14.51s/it]
 32%|███▏      | 3111/9864 [13:07:38<27:19:01, 14.56s/it]
 32%|███▏      | 3112/9864 [13:07:57<29:31:33, 15.74s/it]
 32%|███▏      | 3113/9864 [13:08:13<29:50:31, 15.91s/it]
 32%|███▏      | 3114/9864 [13:08:28<29:23:57, 15.68s/it]
 32%|███▏      | 3115/9864 [13:08:42<28:04:14, 14.97s/it]
 32%|███▏      | 3116/9864 [13:08:56<27:31:43, 14.69s/it]
 32%|███▏      | 3117/9864 [13:09:11<27:34:07, 14.71s/it]
 32%|███▏      | 3118/9864 [13:09:25<27:17:10, 14.56s/it]
 32%|███▏      | 3119/9864 [13:09:40<27:51:14, 14.87s/it]
 32%|███▏      | 3120/9864 [13:09:56<28:31:17, 15.23s/it]
                                                         
{'loss': '0.1601', 'grad_norm': '2.611', 'learning_rate': '1.6e-05', 'epoch': '0.3163'}

 32%|███▏      | 3120/9864 [13:09:56<28:31:17, 15.23s/it]
 32%|███▏      | 3121/9864 [13:10:12<28:46:06, 15.36s/it]
 32%|███▏      | 3122/9864 [13:10:28<29:13:26, 15.60s/it]
 32%|███▏      | 3123/9864 [13:10:42<28:11:38, 15.06s/it]
 32%|███▏      | 3124/9864 [13:10:56<27:41:18, 14.79s/it]
 32%|███▏      | 3125/9864 [13:11:10<27:12:46, 14.54s/it]
 32%|███▏      | 3126/9864 [13:11:24<26:36:21, 14.22s/it]
 32%|███▏      | 3127/9864 [13:11:39<27:26:03, 14.66s/it]
 32%|███▏      | 3128/9864 [13:11:57<29:13:54, 15.62s/it]
 32%|███▏      | 3129/9864 [13:12:13<29:23:27, 15.71s/it]
 32%|███▏      | 3130/9864 [13:12:27<28:24:30, 15.19s/it]
                                                         
{'loss': '0.138', 'grad_norm': '3.087', 'learning_rate': '1.598e-05', 'epoch': '0.3173'}

 32%|███▏      | 3130/9864 [13:12:27<28:24:30, 15.19s/it]
 32%|███▏      | 3131/9864 [13:12:41<27:40:45, 14.80s/it]
 32%|███▏      | 3132/9864 [13:12:55<27:23:44, 14.65s/it]
 32%|███▏      | 3133/9864 [13:13:09<26:44:31, 14.30s/it]
 32%|███▏      | 3134/9864 [13:13:23<26:41:43, 14.28s/it]
 32%|███▏      | 3135/9864 [13:13:40<28:30:14, 15.25s/it]
 32%|███▏      | 3136/9864 [13:13:56<28:46:21, 15.40s/it]
 32%|███▏      | 3137/9864 [13:14:12<29:12:57, 15.64s/it]
 32%|███▏      | 3138/9864 [13:14:27<28:27:18, 15.23s/it]
 32%|███▏      | 3139/9864 [13:14:41<27:58:13, 14.97s/it]
 32%|███▏      | 3140/9864 [13:14:55<27:32:22, 14.74s/it]
                                                         
{'loss': '0.1523', 'grad_norm': '3.355', 'learning_rate': '1.595e-05', 'epoch': '0.3184'}

 32%|███▏      | 3140/9864 [13:14:55<27:32:22, 14.74s/it]
 32%|███▏      | 3141/9864 [13:15:10<27:20:47, 14.64s/it]
 32%|███▏      | 3142/9864 [13:15:27<29:02:35, 15.55s/it]
 32%|███▏      | 3143/9864 [13:15:42<28:25:09, 15.22s/it]
 32%|███▏      | 3144/9864 [13:15:56<27:51:51, 14.93s/it]
 32%|███▏      | 3145/9864 [13:16:14<29:29:55, 15.81s/it]
 32%|███▏      | 3146/9864 [13:16:28<28:42:39, 15.39s/it]
 32%|███▏      | 3147/9864 [13:16:42<27:48:23, 14.90s/it]
 32%|███▏      | 3148/9864 [13:16:57<27:34:46, 14.78s/it]
 32%|███▏      | 3149/9864 [13:17:13<28:15:54, 15.15s/it]
 32%|███▏      | 3150/9864 [13:17:29<28:43:46, 15.40s/it]
                                                         
{'loss': '0.1358', 'grad_norm': '2.468', 'learning_rate': '1.592e-05', 'epoch': '0.3194'}

 32%|███▏      | 3150/9864 [13:17:29<28:43:46, 15.40s/it]
 32%|███▏      | 3151/9864 [13:17:43<28:14:12, 15.14s/it]
 32%|███▏      | 3152/9864 [13:17:59<28:44:37, 15.42s/it]
 32%|███▏      | 3153/9864 [13:18:15<28:57:17, 15.53s/it]
 32%|███▏      | 3154/9864 [13:18:29<28:06:54, 15.08s/it]
 32%|███▏      | 3155/9864 [13:18:43<27:46:05, 14.90s/it]
 32%|███▏      | 3156/9864 [13:18:58<27:27:28, 14.74s/it]
 32%|███▏      | 3157/9864 [13:19:14<28:00:02, 15.03s/it]
 32%|███▏      | 3158/9864 [13:19:29<28:16:31, 15.18s/it]
 32%|███▏      | 3159/9864 [13:19:45<28:44:05, 15.43s/it]
 32%|███▏      | 3160/9864 [13:19:59<27:56:00, 15.00s/it]
                                                         
{'loss': '0.1568', 'grad_norm': '4.817', 'learning_rate': '1.59e-05', 'epoch': '0.3204'}

 32%|███▏      | 3160/9864 [13:19:59<27:56:00, 15.00s/it]
 32%|███▏      | 3161/9864 [13:20:15<28:18:49, 15.21s/it]
 32%|███▏      | 3162/9864 [13:20:28<27:22:53, 14.71s/it]
 32%|███▏      | 3163/9864 [13:20:42<27:02:45, 14.53s/it]
 32%|███▏      | 3164/9864 [13:20:56<26:43:51, 14.36s/it]
 32%|███▏      | 3165/9864 [13:21:14<28:48:26, 15.48s/it]
 32%|███▏      | 3166/9864 [13:21:30<28:59:45, 15.58s/it]
 32%|███▏      | 3167/9864 [13:21:45<28:14:45, 15.18s/it]
 32%|███▏      | 3168/9864 [13:22:01<28:46:06, 15.47s/it]
 32%|███▏      | 3169/9864 [13:22:14<27:43:50, 14.91s/it]
 32%|███▏      | 3170/9864 [13:22:28<27:14:45, 14.65s/it]
                                                         
{'loss': '0.1621', 'grad_norm': '2.076', 'learning_rate': '1.587e-05', 'epoch': '0.3214'}

 32%|███▏      | 3170/9864 [13:22:28<27:14:45, 14.65s/it]
 32%|███▏      | 3171/9864 [13:22:43<27:03:11, 14.55s/it]
 32%|███▏      | 3172/9864 [13:22:59<27:46:58, 14.95s/it]
 32%|███▏      | 3173/9864 [13:23:15<28:23:57, 15.28s/it]
 32%|███▏      | 3174/9864 [13:23:31<29:06:05, 15.66s/it]
 32%|███▏      | 3175/9864 [13:23:45<28:10:41, 15.17s/it]
 32%|███▏      | 3176/9864 [13:23:59<27:34:50, 14.85s/it]
 32%|███▏      | 3177/9864 [13:24:15<28:01:25, 15.09s/it]
 32%|███▏      | 3178/9864 [13:24:29<27:17:24, 14.69s/it]
 32%|███▏      | 3179/9864 [13:24:43<26:58:34, 14.53s/it]
 32%|███▏      | 3180/9864 [13:24:58<27:33:10, 14.84s/it]
                                                         
{'loss': '0.142', 'grad_norm': '2.546', 'learning_rate': '1.584e-05', 'epoch': '0.3224'}

 32%|███▏      | 3180/9864 [13:24:58<27:33:10, 14.84s/it]
 32%|███▏      | 3181/9864 [13:25:14<27:56:43, 15.05s/it]
 32%|███▏      | 3182/9864 [13:25:28<27:30:06, 14.82s/it]
 32%|███▏      | 3183/9864 [13:25:42<27:00:06, 14.55s/it]
 32%|███▏      | 3184/9864 [13:25:57<26:56:43, 14.52s/it]
 32%|███▏      | 3185/9864 [13:26:11<27:02:03, 14.57s/it]
 32%|███▏      | 3186/9864 [13:26:27<27:44:19, 14.95s/it]
 32%|███▏      | 3187/9864 [13:26:41<27:18:06, 14.72s/it]
 32%|███▏      | 3188/9864 [13:26:57<27:54:09, 15.05s/it]
 32%|███▏      | 3189/9864 [13:27:15<29:23:59, 15.86s/it]
 32%|███▏      | 3190/9864 [13:27:29<28:14:30, 15.23s/it]
                                                         
{'loss': '0.15', 'grad_norm': '3.855', 'learning_rate': '1.582e-05', 'epoch': '0.3234'}

 32%|███▏      | 3190/9864 [13:27:29<28:14:30, 15.23s/it]
 32%|███▏      | 3191/9864 [13:27:43<27:34:21, 14.88s/it]
 32%|███▏      | 3192/9864 [13:27:57<27:10:37, 14.66s/it]
 32%|███▏      | 3193/9864 [13:28:13<27:49:58, 15.02s/it]
 32%|███▏      | 3194/9864 [13:28:27<27:24:48, 14.80s/it]
 32%|███▏      | 3195/9864 [13:28:41<27:12:16, 14.69s/it]
 32%|███▏      | 3196/9864 [13:28:57<27:49:49, 15.03s/it]
 32%|███▏      | 3197/9864 [13:29:15<29:17:06, 15.81s/it]
 32%|███▏      | 3198/9864 [13:29:29<28:17:27, 15.28s/it]
 32%|███▏      | 3199/9864 [13:29:43<27:26:51, 14.83s/it]
 32%|███▏      | 3200/9864 [13:29:56<26:43:03, 14.43s/it]
                                                         
{'loss': '0.1558', 'grad_norm': '2.975', 'learning_rate': '1.579e-05', 'epoch': '0.3244'}

 32%|███▏      | 3200/9864 [13:29:56<26:43:03, 14.43s/it]
 32%|███▏      | 3201/9864 [13:30:12<27:26:24, 14.83s/it]
 32%|███▏      | 3202/9864 [13:30:26<26:49:08, 14.49s/it]
 32%|███▏      | 3203/9864 [13:30:42<28:00:19, 15.14s/it]
 32%|███▏      | 3204/9864 [13:30:58<28:32:51, 15.43s/it]
 32%|███▏      | 3205/9864 [13:31:14<28:41:58, 15.52s/it]
 33%|███▎      | 3206/9864 [13:31:28<27:59:05, 15.13s/it]
 33%|███▎      | 3207/9864 [13:31:43<27:39:54, 14.96s/it]
 33%|███▎      | 3208/9864 [13:31:57<27:13:50, 14.73s/it]
 33%|███▎      | 3209/9864 [13:32:13<27:51:41, 15.07s/it]
 33%|███▎      | 3210/9864 [13:32:29<28:17:01, 15.30s/it]
                                                         
{'loss': '0.1487', 'grad_norm': '5.263', 'learning_rate': '1.576e-05', 'epoch': '0.3255'}

 33%|███▎      | 3210/9864 [13:32:29<28:17:01, 15.30s/it]
 33%|███▎      | 3211/9864 [13:32:43<27:51:03, 15.07s/it]
 33%|███▎      | 3212/9864 [13:32:58<27:30:40, 14.89s/it]
 33%|███▎      | 3213/9864 [13:33:13<27:51:05, 15.08s/it]
 33%|███▎      | 3214/9864 [13:33:27<27:01:48, 14.63s/it]
 33%|███▎      | 3215/9864 [13:33:41<26:45:05, 14.48s/it]
 33%|███▎      | 3216/9864 [13:33:55<26:35:48, 14.40s/it]
 33%|███▎      | 3217/9864 [13:34:11<27:09:22, 14.71s/it]
 33%|███▎      | 3218/9864 [13:34:26<27:34:28, 14.94s/it]
 33%|███▎      | 3219/9864 [13:34:40<26:50:15, 14.54s/it]
 33%|███▎      | 3220/9864 [13:34:54<26:35:21, 14.41s/it]
                                                         
{'loss': '0.1464', 'grad_norm': '3.999', 'learning_rate': '1.574e-05', 'epoch': '0.3265'}

 33%|███▎      | 3220/9864 [13:34:54<26:35:21, 14.41s/it]
 33%|███▎      | 3221/9864 [13:35:11<28:14:01, 15.30s/it]
 33%|███▎      | 3222/9864 [13:35:26<27:41:44, 15.01s/it]
 33%|███▎      | 3223/9864 [13:35:39<27:05:21, 14.68s/it]
 33%|███▎      | 3224/9864 [13:35:54<26:50:35, 14.55s/it]
 33%|███▎      | 3225/9864 [13:36:11<28:20:47, 15.37s/it]
 33%|███▎      | 3226/9864 [13:36:25<27:39:08, 15.00s/it]
 33%|███▎      | 3227/9864 [13:36:39<26:57:20, 14.62s/it]
 33%|███▎      | 3228/9864 [13:36:56<28:30:18, 15.46s/it]
 33%|███▎      | 3229/9864 [13:37:11<27:56:23, 15.16s/it]
 33%|███▎      | 3230/9864 [13:37:25<27:25:18, 14.88s/it]
                                                         
{'loss': '0.149', 'grad_norm': '4.411', 'learning_rate': '1.571e-05', 'epoch': '0.3275'}

 33%|███▎      | 3230/9864 [13:37:25<27:25:18, 14.88s/it]
 33%|███▎      | 3231/9864 [13:37:39<27:00:32, 14.66s/it]
 33%|███▎      | 3232/9864 [13:37:55<27:34:16, 14.97s/it]
 33%|███▎      | 3233/9864 [13:38:11<28:08:04, 15.27s/it]
 33%|███▎      | 3234/9864 [13:38:25<27:23:31, 14.87s/it]
 33%|███▎      | 3235/9864 [13:38:41<28:03:23, 15.24s/it]
 33%|███▎      | 3236/9864 [13:38:57<28:18:03, 15.37s/it]
 33%|███▎      | 3237/9864 [13:39:11<27:32:22, 14.96s/it]
 33%|███▎      | 3238/9864 [13:39:24<26:57:24, 14.65s/it]
 33%|███▎      | 3239/9864 [13:39:40<27:31:08, 14.95s/it]
 33%|███▎      | 3240/9864 [13:39:54<27:11:44, 14.78s/it]
                                                         
{'loss': '0.1536', 'grad_norm': '1.968', 'learning_rate': '1.568e-05', 'epoch': '0.3285'}

 33%|███▎      | 3240/9864 [13:39:54<27:11:44, 14.78s/it]
 33%|███▎      | 3241/9864 [13:40:09<26:54:12, 14.62s/it]
 33%|███▎      | 3242/9864 [13:40:24<27:17:13, 14.83s/it]
 33%|███▎      | 3243/9864 [13:40:40<27:57:15, 15.20s/it]
 33%|███▎      | 3244/9864 [13:40:55<28:02:18, 15.25s/it]
 33%|███▎      | 3245/9864 [13:41:10<27:30:11, 14.96s/it]
 33%|███▎      | 3246/9864 [13:41:26<28:03:00, 15.26s/it]
 33%|███▎      | 3247/9864 [13:41:40<27:18:00, 14.85s/it]
 33%|███▎      | 3248/9864 [13:41:54<27:00:53, 14.70s/it]
 33%|███▎      | 3249/9864 [13:42:10<27:50:09, 15.15s/it]
 33%|███▎      | 3250/9864 [13:42:25<27:39:23, 15.05s/it]
                                                         
{'loss': '0.1471', 'grad_norm': '2.869', 'learning_rate': '1.566e-05', 'epoch': '0.3295'}

 33%|███▎      | 3250/9864 [13:42:25<27:39:23, 15.05s/it]
 33%|███▎      | 3251/9864 [13:42:41<28:01:27, 15.26s/it]
 33%|███▎      | 3252/9864 [13:42:57<28:38:40, 15.60s/it]
 33%|███▎      | 3253/9864 [13:43:12<28:06:26, 15.31s/it]
 33%|███▎      | 3254/9864 [13:43:27<28:20:01, 15.43s/it]
 33%|███▎      | 3255/9864 [13:43:42<27:42:01, 15.09s/it]
 33%|███▎      | 3256/9864 [13:43:58<28:19:36, 15.43s/it]
 33%|███▎      | 3257/9864 [13:44:12<27:29:55, 14.98s/it]
 33%|███▎      | 3258/9864 [13:44:28<28:02:56, 15.29s/it]
 33%|███▎      | 3259/9864 [13:44:42<27:26:10, 14.95s/it]
 33%|███▎      | 3260/9864 [13:44:58<28:01:00, 15.27s/it]
                                                         
{'loss': '0.166', 'grad_norm': '3.166', 'learning_rate': '1.563e-05', 'epoch': '0.3305'}

 33%|███▎      | 3260/9864 [13:44:58<28:01:00, 15.27s/it]
 33%|███▎      | 3261/9864 [13:45:12<27:29:19, 14.99s/it]
 33%|███▎      | 3262/9864 [13:45:28<28:00:32, 15.27s/it]
 33%|███▎      | 3263/9864 [13:45:43<27:35:09, 15.04s/it]
 33%|███▎      | 3264/9864 [13:45:57<26:57:53, 14.71s/it]
 33%|███▎      | 3265/9864 [13:46:15<29:03:09, 15.85s/it]
 33%|███▎      | 3266/9864 [13:46:29<28:02:10, 15.30s/it]
 33%|███▎      | 3267/9864 [13:46:45<28:02:15, 15.30s/it]
 33%|███▎      | 3268/9864 [13:46:59<27:41:11, 15.11s/it]
 33%|███▎      | 3269/9864 [13:47:13<27:07:11, 14.80s/it]
 33%|███▎      | 3270/9864 [13:47:29<27:36:30, 15.07s/it]
                                                         
{'loss': '0.1509', 'grad_norm': '3.816', 'learning_rate': '1.56e-05', 'epoch': '0.3315'}

 33%|███▎      | 3270/9864 [13:47:29<27:36:30, 15.07s/it]
 33%|███▎      | 3271/9864 [13:47:43<27:05:52, 14.80s/it]
 33%|███▎      | 3272/9864 [13:48:00<27:53:41, 15.23s/it]
 33%|███▎      | 3273/9864 [13:48:14<27:18:38, 14.92s/it]
 33%|███▎      | 3274/9864 [13:48:29<27:46:13, 15.17s/it]
 33%|███▎      | 3275/9864 [13:48:45<28:00:33, 15.30s/it]
 33%|███▎      | 3276/9864 [13:48:59<27:22:10, 14.96s/it]
 33%|███▎      | 3277/9864 [13:49:15<27:58:15, 15.29s/it]
 33%|███▎      | 3278/9864 [13:49:29<27:18:14, 14.92s/it]
 33%|███▎      | 3279/9864 [13:49:46<28:06:16, 15.36s/it]
 33%|███▎      | 3280/9864 [13:50:00<27:27:23, 15.01s/it]
                                                         
{'loss': '0.1452', 'grad_norm': '2.207', 'learning_rate': '1.557e-05', 'epoch': '0.3325'}

 33%|███▎      | 3280/9864 [13:50:00<27:27:23, 15.01s/it]
 33%|███▎      | 3281/9864 [13:50:16<28:05:07, 15.36s/it]
 33%|███▎      | 3282/9864 [13:50:30<27:26:40, 15.01s/it]
 33%|███▎      | 3283/9864 [13:50:45<27:07:40, 14.84s/it]
 33%|███▎      | 3284/9864 [13:51:01<27:44:06, 15.17s/it]
 33%|███▎      | 3285/9864 [13:51:15<27:04:21, 14.81s/it]
 33%|███▎      | 3286/9864 [13:51:30<27:13:13, 14.90s/it]
 33%|███▎      | 3287/9864 [13:51:46<27:54:19, 15.27s/it]
 33%|███▎      | 3288/9864 [13:52:00<27:11:27, 14.89s/it]
 33%|███▎      | 3289/9864 [13:52:14<26:59:50, 14.78s/it]
 33%|███▎      | 3290/9864 [13:52:31<27:45:41, 15.20s/it]
                                                         
{'loss': '0.1525', 'grad_norm': '2.301', 'learning_rate': '1.555e-05', 'epoch': '0.3336'}

 33%|███▎      | 3290/9864 [13:52:31<27:45:41, 15.20s/it]
 33%|███▎      | 3291/9864 [13:52:44<27:01:58, 14.81s/it]
 33%|███▎      | 3292/9864 [13:53:00<27:38:30, 15.14s/it]
 33%|███▎      | 3293/9864 [13:53:16<28:02:11, 15.36s/it]
 33%|███▎      | 3294/9864 [13:53:31<27:36:21, 15.13s/it]
 33%|███▎      | 3295/9864 [13:53:47<28:13:42, 15.47s/it]
 33%|███▎      | 3296/9864 [13:54:02<27:37:26, 15.14s/it]
 33%|███▎      | 3297/9864 [13:54:16<27:15:44, 14.95s/it]
 33%|███▎      | 3298/9864 [13:54:30<26:53:41, 14.75s/it]
 33%|███▎      | 3299/9864 [13:54:46<27:26:26, 15.05s/it]
 33%|███▎      | 3300/9864 [13:55:02<27:50:29, 15.27s/it]
                                                         
{'loss': '0.1477', 'grad_norm': '1.733', 'learning_rate': '1.552e-05', 'epoch': '0.3346'}

 33%|███▎      | 3300/9864 [13:55:02<27:50:29, 15.27s/it]
 33%|███▎      | 3301/9864 [13:55:16<27:24:20, 15.03s/it]
 33%|███▎      | 3302/9864 [13:55:34<28:58:13, 15.89s/it]
 33%|███▎      | 3303/9864 [13:55:49<28:09:25, 15.45s/it]
 33%|███▎      | 3304/9864 [13:56:02<26:59:17, 14.81s/it]
 34%|███▎      | 3305/9864 [13:56:16<26:44:09, 14.67s/it]
 34%|███▎      | 3306/9864 [13:56:31<26:34:47, 14.59s/it]
 34%|███▎      | 3307/9864 [13:56:47<27:19:28, 15.00s/it]
 34%|███▎      | 3308/9864 [13:57:03<27:48:15, 15.27s/it]
 34%|███▎      | 3309/9864 [13:57:17<27:24:50, 15.06s/it]
 34%|███▎      | 3310/9864 [13:57:35<29:11:06, 16.03s/it]
                                                         
{'loss': '0.1586', 'grad_norm': '3.202', 'learning_rate': '1.549e-05', 'epoch': '0.3356'}

 34%|███▎      | 3310/9864 [13:57:35<29:11:06, 16.03s/it]
 34%|███▎      | 3311/9864 [13:57:50<28:16:30, 15.53s/it]
 34%|███▎      | 3312/9864 [13:58:04<27:47:31, 15.27s/it]
 34%|███▎      | 3313/9864 [13:58:18<27:03:11, 14.87s/it]
 34%|███▎      | 3314/9864 [13:58:34<27:28:23, 15.10s/it]
 34%|███▎      | 3315/9864 [13:58:49<27:12:57, 14.96s/it]
 34%|███▎      | 3316/9864 [13:59:05<27:47:57, 15.28s/it]
 34%|███▎      | 3317/9864 [13:59:19<27:05:00, 14.89s/it]
 34%|███▎      | 3318/9864 [13:59:36<28:28:09, 15.66s/it]
 34%|███▎      | 3319/9864 [13:59:50<27:29:18, 15.12s/it]
 34%|███▎      | 3320/9864 [14:00:04<26:53:06, 14.79s/it]
                                                         
{'loss': '0.1601', 'grad_norm': '4.577', 'learning_rate': '1.547e-05', 'epoch': '0.3366'}

 34%|███▎      | 3320/9864 [14:00:04<26:53:06, 14.79s/it]
 34%|███▎      | 3321/9864 [14:00:18<26:30:47, 14.59s/it]
 34%|███▎      | 3322/9864 [14:00:32<26:09:31, 14.39s/it]
 34%|███▎      | 3323/9864 [14:00:49<27:33:14, 15.17s/it]
 34%|███▎      | 3324/9864 [14:01:03<27:05:26, 14.91s/it]
 34%|███▎      | 3325/9864 [14:01:20<27:49:22, 15.32s/it]
 34%|███▎      | 3326/9864 [14:01:35<28:02:21, 15.44s/it]
 34%|███▎      | 3327/9864 [14:01:50<27:21:26, 15.07s/it]
 34%|███▎      | 3328/9864 [14:02:04<26:52:39, 14.80s/it]
 34%|███▎      | 3329/9864 [14:02:18<26:26:00, 14.56s/it]
 34%|███▍      | 3330/9864 [14:02:33<27:05:55, 14.93s/it]
                                                         
{'loss': '0.1619', 'grad_norm': '5.188', 'learning_rate': '1.544e-05', 'epoch': '0.3376'}

 34%|███▍      | 3330/9864 [14:02:33<27:05:55, 14.93s/it]
 34%|███▍      | 3331/9864 [14:02:48<26:53:22, 14.82s/it]
 34%|███▍      | 3332/9864 [14:03:02<26:39:51, 14.70s/it]
 34%|███▍      | 3333/9864 [14:03:20<28:08:57, 15.52s/it]
 34%|███▍      | 3334/9864 [14:03:35<27:42:00, 15.27s/it]
 34%|███▍      | 3335/9864 [14:03:49<27:05:02, 14.93s/it]
 34%|███▍      | 3336/9864 [14:04:03<26:33:47, 14.65s/it]
 34%|███▍      | 3337/9864 [14:04:17<26:06:11, 14.40s/it]
 34%|███▍      | 3338/9864 [14:04:33<27:00:52, 14.90s/it]
 34%|███▍      | 3339/9864 [14:04:48<27:32:26, 15.19s/it]
 34%|███▍      | 3340/9864 [14:05:04<27:53:10, 15.39s/it]
                                                         
{'loss': '0.1574', 'grad_norm': '2.946', 'learning_rate': '1.541e-05', 'epoch': '0.3386'}

 34%|███▍      | 3340/9864 [14:05:04<27:53:10, 15.39s/it]
 34%|███▍      | 3341/9864 [14:05:20<27:49:38, 15.36s/it]
 34%|███▍      | 3342/9864 [14:05:34<27:12:49, 15.02s/it]
 34%|███▍      | 3343/9864 [14:05:48<26:56:11, 14.87s/it]
 34%|███▍      | 3344/9864 [14:06:03<26:38:20, 14.71s/it]
 34%|███▍      | 3345/9864 [14:06:17<26:30:14, 14.64s/it]
 34%|███▍      | 3346/9864 [14:06:33<26:54:22, 14.86s/it]
 34%|███▍      | 3347/9864 [14:06:50<28:16:12, 15.62s/it]
 34%|███▍      | 3348/9864 [14:07:06<28:38:26, 15.82s/it]
 34%|███▍      | 3349/9864 [14:07:21<27:54:58, 15.43s/it]
 34%|███▍      | 3350/9864 [14:07:35<27:11:49, 15.03s/it]
                                                         
{'loss': '0.1414', 'grad_norm': '2.485', 'learning_rate': '1.538e-05', 'epoch': '0.3396'}

 34%|███▍      | 3350/9864 [14:07:35<27:11:49, 15.03s/it]
 34%|███▍      | 3351/9864 [14:07:49<26:46:38, 14.80s/it]
 34%|███▍      | 3352/9864 [14:08:04<26:47:39, 14.81s/it]
 34%|███▍      | 3353/9864 [14:08:20<27:12:59, 15.05s/it]
 34%|███▍      | 3354/9864 [14:08:34<26:41:25, 14.76s/it]
 34%|███▍      | 3355/9864 [14:08:51<28:08:35, 15.57s/it]
 34%|███▍      | 3356/9864 [14:09:05<27:16:53, 15.09s/it]
 34%|███▍      | 3357/9864 [14:09:21<27:45:38, 15.36s/it]
 34%|███▍      | 3358/9864 [14:09:35<27:12:43, 15.06s/it]
 34%|███▍      | 3359/9864 [14:09:50<26:41:39, 14.77s/it]
 34%|███▍      | 3360/9864 [14:10:04<26:24:45, 14.62s/it]
                                                         
{'loss': '0.1426', 'grad_norm': '3.501', 'learning_rate': '1.535e-05', 'epoch': '0.3407'}

 34%|███▍      | 3360/9864 [14:10:04<26:24:45, 14.62s/it]
 34%|███▍      | 3361/9864 [14:10:18<26:06:34, 14.45s/it]
 34%|███▍      | 3362/9864 [14:10:34<26:56:40, 14.92s/it]
 34%|███▍      | 3363/9864 [14:10:51<28:17:26, 15.67s/it]
 34%|███▍      | 3364/9864 [14:11:06<27:32:35, 15.25s/it]
 34%|███▍      | 3365/9864 [14:11:21<27:42:13, 15.35s/it]
 34%|███▍      | 3366/9864 [14:11:35<27:10:04, 15.05s/it]
 34%|███▍      | 3367/9864 [14:11:49<26:29:52, 14.68s/it]
 34%|███▍      | 3368/9864 [14:12:04<26:15:50, 14.56s/it]
 34%|███▍      | 3369/9864 [14:12:18<25:57:29, 14.39s/it]
 34%|███▍      | 3370/9864 [14:12:33<26:27:06, 14.66s/it]
                                                         
{'loss': '0.1624', 'grad_norm': '4.271', 'learning_rate': '1.533e-05', 'epoch': '0.3417'}

 34%|███▍      | 3370/9864 [14:12:33<26:27:06, 14.66s/it]
 34%|███▍      | 3371/9864 [14:12:49<27:08:01, 15.04s/it]
 34%|███▍      | 3372/9864 [14:13:06<28:22:33, 15.74s/it]
 34%|███▍      | 3373/9864 [14:13:20<27:19:44, 15.16s/it]
 34%|███▍      | 3374/9864 [14:13:34<26:42:03, 14.81s/it]
 34%|███▍      | 3375/9864 [14:13:48<26:29:22, 14.70s/it]
 34%|███▍      | 3376/9864 [14:14:02<26:05:08, 14.47s/it]
 34%|███▍      | 3377/9864 [14:14:18<26:52:46, 14.92s/it]
 34%|███▍      | 3378/9864 [14:14:33<26:40:32, 14.81s/it]
 34%|███▍      | 3379/9864 [14:14:47<26:18:49, 14.61s/it]
 34%|███▍      | 3380/9864 [14:15:06<28:35:13, 15.87s/it]
                                                         
{'loss': '0.152', 'grad_norm': '4.93', 'learning_rate': '1.53e-05', 'epoch': '0.3427'}

 34%|███▍      | 3380/9864 [14:15:06<28:35:13, 15.87s/it]
 34%|███▍      | 3381/9864 [14:15:20<27:41:21, 15.38s/it]
 34%|███▍      | 3382/9864 [14:15:34<27:01:04, 15.01s/it]
 34%|███▍      | 3383/9864 [14:15:48<26:33:47, 14.76s/it]
 34%|███▍      | 3384/9864 [14:16:02<26:10:24, 14.54s/it]
 34%|███▍      | 3385/9864 [14:16:18<26:58:54, 14.99s/it]
 34%|███▍      | 3386/9864 [14:16:32<26:27:54, 14.71s/it]
 34%|███▍      | 3387/9864 [14:16:48<27:07:13, 15.07s/it]
 34%|███▍      | 3388/9864 [14:17:05<27:40:52, 15.39s/it]
 34%|███▍      | 3389/9864 [14:17:18<26:43:01, 14.85s/it]
 34%|███▍      | 3390/9864 [14:17:32<26:18:11, 14.63s/it]
                                                         
{'loss': '0.1577', 'grad_norm': '4.246', 'learning_rate': '1.527e-05', 'epoch': '0.3437'}

 34%|███▍      | 3390/9864 [14:17:32<26:18:11, 14.63s/it]
 34%|███▍      | 3391/9864 [14:17:46<25:52:34, 14.39s/it]
 34%|███▍      | 3392/9864 [14:18:00<25:52:46, 14.40s/it]
 34%|███▍      | 3393/9864 [14:18:16<26:35:30, 14.79s/it]
 34%|███▍      | 3394/9864 [14:18:31<26:25:40, 14.70s/it]
 34%|███▍      | 3395/9864 [14:18:48<27:56:59, 15.55s/it]
 34%|███▍      | 3396/9864 [14:19:04<28:18:50, 15.76s/it]
 34%|███▍      | 3397/9864 [14:19:18<27:15:24, 15.17s/it]
 34%|███▍      | 3398/9864 [14:19:32<26:38:35, 14.83s/it]
 34%|███▍      | 3399/9864 [14:19:47<26:26:13, 14.72s/it]
 34%|███▍      | 3400/9864 [14:20:01<26:19:22, 14.66s/it]
                                                         
{'loss': '0.1428', 'grad_norm': '4.005', 'learning_rate': '1.524e-05', 'epoch': '0.3447'}

 34%|███▍      | 3400/9864 [14:20:01<26:19:22, 14.66s/it]
 34%|███▍      | 3401/9864 [14:20:17<26:57:35, 15.02s/it]
 34%|███▍      | 3402/9864 [14:20:31<26:19:16, 14.66s/it]
 34%|███▍      | 3403/9864 [14:20:49<27:59:41, 15.60s/it]
 35%|███▍      | 3404/9864 [14:21:05<28:10:29, 15.70s/it]
 35%|███▍      | 3405/9864 [14:21:18<27:03:17, 15.08s/it]
 35%|███▍      | 3406/9864 [14:21:32<26:27:23, 14.75s/it]
 35%|███▍      | 3407/9864 [14:21:46<25:58:14, 14.48s/it]
 35%|███▍      | 3408/9864 [14:22:02<26:52:35, 14.99s/it]
 35%|███▍      | 3409/9864 [14:22:16<26:16:33, 14.65s/it]
 35%|███▍      | 3410/9864 [14:22:32<26:54:05, 15.01s/it]
                                                         
{'loss': '0.1585', 'grad_norm': '3.175', 'learning_rate': '1.522e-05', 'epoch': '0.3457'}

 35%|███▍      | 3410/9864 [14:22:32<26:54:05, 15.01s/it]
 35%|███▍      | 3411/9864 [14:22:48<27:15:40, 15.21s/it]
 35%|███▍      | 3412/9864 [14:23:04<27:58:38, 15.61s/it]
 35%|███▍      | 3413/9864 [14:23:19<27:21:59, 15.27s/it]
 35%|███▍      | 3414/9864 [14:23:33<26:51:14, 14.99s/it]
 35%|███▍      | 3415/9864 [14:23:49<27:17:46, 15.24s/it]
 35%|███▍      | 3416/9864 [14:24:03<26:42:55, 14.92s/it]
 35%|███▍      | 3417/9864 [14:24:17<25:58:54, 14.51s/it]
 35%|███▍      | 3418/9864 [14:24:34<27:34:12, 15.40s/it]
 35%|███▍      | 3419/9864 [14:24:48<26:35:37, 14.85s/it]
 35%|███▍      | 3420/9864 [14:25:03<26:59:22, 15.08s/it]
                                                         
{'loss': '0.1474', 'grad_norm': '2.028', 'learning_rate': '1.519e-05', 'epoch': '0.3467'}

 35%|███▍      | 3420/9864 [14:25:03<26:59:22, 15.08s/it]
 35%|███▍      | 3421/9864 [14:25:17<26:17:54, 14.69s/it]
 35%|███▍      | 3422/9864 [14:25:31<25:55:09, 14.48s/it]
 35%|███▍      | 3423/9864 [14:25:46<26:25:31, 14.77s/it]
 35%|███▍      | 3424/9864 [14:26:01<26:10:05, 14.63s/it]
 35%|███▍      | 3425/9864 [14:26:16<26:42:20, 14.93s/it]
 35%|███▍      | 3426/9864 [14:26:33<27:28:11, 15.36s/it]
 35%|███▍      | 3427/9864 [14:26:47<26:49:02, 15.00s/it]
 35%|███▍      | 3428/9864 [14:27:03<27:34:45, 15.43s/it]
 35%|███▍      | 3429/9864 [14:27:18<26:55:34, 15.06s/it]
 35%|███▍      | 3430/9864 [14:27:31<26:16:22, 14.70s/it]
                                                         
{'loss': '0.1497', 'grad_norm': '2.483', 'learning_rate': '1.516e-05', 'epoch': '0.3478'}

 35%|███▍      | 3430/9864 [14:27:31<26:16:22, 14.70s/it]
 35%|███▍      | 3431/9864 [14:27:48<27:10:14, 15.21s/it]
 35%|███▍      | 3432/9864 [14:28:04<27:33:40, 15.43s/it]
 35%|███▍      | 3433/9864 [14:28:20<27:53:31, 15.61s/it]
 35%|███▍      | 3434/9864 [14:28:34<27:00:31, 15.12s/it]
 35%|███▍      | 3435/9864 [14:28:48<26:24:44, 14.79s/it]
 35%|███▍      | 3436/9864 [14:29:04<27:05:36, 15.17s/it]
 35%|███▍      | 3437/9864 [14:29:17<26:06:12, 14.62s/it]
 35%|███▍      | 3438/9864 [14:29:31<25:32:55, 14.31s/it]
 35%|███▍      | 3439/9864 [14:29:48<27:17:52, 15.30s/it]
 35%|███▍      | 3440/9864 [14:30:03<26:45:13, 14.99s/it]
                                                         
{'loss': '0.1611', 'grad_norm': '2.828', 'learning_rate': '1.513e-05', 'epoch': '0.3488'}

 35%|███▍      | 3440/9864 [14:30:03<26:45:13, 14.99s/it]
 35%|███▍      | 3441/9864 [14:30:19<27:23:43, 15.35s/it]
 35%|███▍      | 3442/9864 [14:30:32<26:24:23, 14.80s/it]
 35%|███▍      | 3443/9864 [14:30:46<25:59:26, 14.57s/it]
 35%|███▍      | 3444/9864 [14:31:02<26:41:53, 14.97s/it]
 35%|███▍      | 3445/9864 [14:31:17<26:23:40, 14.80s/it]
 35%|███▍      | 3446/9864 [14:31:31<25:58:42, 14.57s/it]
 35%|███▍      | 3447/9864 [14:31:47<26:36:28, 14.93s/it]
 35%|███▍      | 3448/9864 [14:32:02<27:07:39, 15.22s/it]
 35%|███▍      | 3449/9864 [14:32:18<27:30:47, 15.44s/it]
 35%|███▍      | 3450/9864 [14:32:33<26:49:01, 15.05s/it]
                                                         
{'loss': '0.1485', 'grad_norm': '6.131', 'learning_rate': '1.51e-05', 'epoch': '0.3498'}

 35%|███▍      | 3450/9864 [14:32:33<26:49:01, 15.05s/it]
 35%|███▍      | 3451/9864 [14:32:49<27:24:51, 15.39s/it]
 35%|███▍      | 3452/9864 [14:33:03<26:54:58, 15.11s/it]
 35%|███▌      | 3453/9864 [14:33:18<26:31:14, 14.89s/it]
 35%|███▌      | 3454/9864 [14:33:32<26:22:56, 14.82s/it]
 35%|███▌      | 3455/9864 [14:33:50<28:00:53, 15.74s/it]
 35%|███▌      | 3456/9864 [14:34:06<28:04:43, 15.77s/it]
 35%|███▌      | 3457/9864 [14:34:20<27:14:02, 15.30s/it]
 35%|███▌      | 3458/9864 [14:34:35<26:46:12, 15.04s/it]
 35%|███▌      | 3459/9864 [14:34:51<27:26:19, 15.42s/it]
 35%|███▌      | 3460/9864 [14:35:05<26:32:23, 14.92s/it]
                                                         
{'loss': '0.1517', 'grad_norm': '2.462', 'learning_rate': '1.507e-05', 'epoch': '0.3508'}

 35%|███▌      | 3460/9864 [14:35:05<26:32:23, 14.92s/it]
 35%|███▌      | 3461/9864 [14:35:19<26:08:03, 14.69s/it]
 35%|███▌      | 3462/9864 [14:35:33<25:56:54, 14.59s/it]
 35%|███▌      | 3463/9864 [14:35:47<25:40:24, 14.44s/it]
 35%|███▌      | 3464/9864 [14:36:05<27:21:17, 15.39s/it]
 35%|███▌      | 3465/9864 [14:36:21<27:35:11, 15.52s/it]
 35%|███▌      | 3466/9864 [14:36:36<27:41:52, 15.58s/it]
 35%|███▌      | 3467/9864 [14:36:50<26:47:10, 15.07s/it]
 35%|███▌      | 3468/9864 [14:37:04<26:09:58, 14.73s/it]
 35%|███▌      | 3469/9864 [14:37:18<25:39:21, 14.44s/it]
 35%|███▌      | 3470/9864 [14:37:32<25:12:38, 14.19s/it]
                                                         
{'loss': '0.1384', 'grad_norm': '3.342', 'learning_rate': '1.505e-05', 'epoch': '0.3518'}

 35%|███▌      | 3470/9864 [14:37:32<25:12:38, 14.19s/it]
 35%|███▌      | 3471/9864 [14:37:47<26:00:46, 14.65s/it]
 35%|███▌      | 3472/9864 [14:38:03<26:36:07, 14.98s/it]
 35%|███▌      | 3473/9864 [14:38:18<26:22:00, 14.85s/it]
 35%|███▌      | 3474/9864 [14:38:31<25:39:47, 14.46s/it]
 35%|███▌      | 3475/9864 [14:38:47<26:28:53, 14.92s/it]
 35%|███▌      | 3476/9864 [14:39:01<26:02:33, 14.68s/it]
 35%|███▌      | 3477/9864 [14:39:16<25:49:54, 14.56s/it]
 35%|███▌      | 3478/9864 [14:39:30<25:44:17, 14.51s/it]
 35%|███▌      | 3479/9864 [14:39:46<26:21:41, 14.86s/it]
 35%|███▌      | 3480/9864 [14:40:02<27:07:26, 15.30s/it]
                                                         
{'loss': '0.154', 'grad_norm': '3.996', 'learning_rate': '1.502e-05', 'epoch': '0.3528'}

 35%|███▌      | 3480/9864 [14:40:02<27:07:26, 15.30s/it]
 35%|███▌      | 3481/9864 [14:40:18<27:22:29, 15.44s/it]
 35%|███▌      | 3482/9864 [14:40:34<27:42:52, 15.63s/it]
 35%|███▌      | 3483/9864 [14:40:48<27:01:10, 15.24s/it]
 35%|███▌      | 3484/9864 [14:41:02<26:13:16, 14.80s/it]
 35%|███▌      | 3485/9864 [14:41:16<25:56:15, 14.64s/it]
 35%|███▌      | 3486/9864 [14:41:30<25:47:00, 14.55s/it]
 35%|███▌      | 3487/9864 [14:41:47<26:43:08, 15.08s/it]
 35%|███▌      | 3488/9864 [14:42:02<26:58:10, 15.23s/it]
 35%|███▌      | 3489/9864 [14:42:18<27:19:35, 15.43s/it]
 35%|███▌      | 3490/9864 [14:42:34<27:22:56, 15.47s/it]
                                                         
{'loss': '0.1477', 'grad_norm': '1.432', 'learning_rate': '1.499e-05', 'epoch': '0.3538'}

 35%|███▌      | 3490/9864 [14:42:34<27:22:56, 15.47s/it]
 35%|███▌      | 3491/9864 [14:42:48<26:35:47, 15.02s/it]
 35%|███▌      | 3492/9864 [14:43:02<25:52:55, 14.62s/it]
 35%|███▌      | 3493/9864 [14:43:16<25:34:52, 14.45s/it]
 35%|███▌      | 3494/9864 [14:43:30<25:22:38, 14.34s/it]
 35%|███▌      | 3495/9864 [14:43:45<25:49:58, 14.60s/it]
 35%|███▌      | 3496/9864 [14:44:01<26:23:32, 14.92s/it]
 35%|███▌      | 3497/9864 [14:44:16<26:50:50, 15.18s/it]
 35%|███▌      | 3498/9864 [14:44:33<27:24:08, 15.50s/it]
 35%|███▌      | 3499/9864 [14:44:47<26:38:37, 15.07s/it]
 35%|███▌      | 3500/9864 [14:45:00<25:57:44, 14.69s/it]
                                                         
{'loss': '0.1436', 'grad_norm': '4.457', 'learning_rate': '1.496e-05', 'epoch': '0.3549'}

 35%|███▌      | 3500/9864 [14:45:00<25:57:44, 14.69s/it]
 35%|███▌      | 3501/9864 [14:45:15<25:45:05, 14.57s/it]
 36%|███▌      | 3502/9864 [14:45:29<25:36:21, 14.49s/it]
 36%|███▌      | 3503/9864 [14:45:44<25:44:33, 14.57s/it]
 36%|███▌      | 3504/9864 [14:46:00<26:48:39, 15.18s/it]
 36%|███▌      | 3505/9864 [14:46:18<28:17:41, 16.02s/it]
 36%|███▌      | 3506/9864 [14:46:35<28:24:18, 16.08s/it]
 36%|███▌      | 3507/9864 [14:46:49<27:22:03, 15.50s/it]
 36%|███▌      | 3508/9864 [14:47:03<26:47:28, 15.17s/it]
 36%|███▌      | 3509/9864 [14:47:17<26:12:37, 14.85s/it]
 36%|███▌      | 3510/9864 [14:47:31<25:48:19, 14.62s/it]
                                                         
{'loss': '0.1583', 'grad_norm': '2.159', 'learning_rate': '1.493e-05', 'epoch': '0.3559'}

 36%|███▌      | 3510/9864 [14:47:31<25:48:19, 14.62s/it]
 36%|███▌      | 3511/9864 [14:47:47<26:28:28, 15.00s/it]
 36%|███▌      | 3512/9864 [14:48:01<25:52:55, 14.67s/it]
 36%|███▌      | 3513/9864 [14:48:19<27:41:15, 15.69s/it]
 36%|███▌      | 3514/9864 [14:48:36<28:03:42, 15.91s/it]
 36%|███▌      | 3515/9864 [14:48:50<27:11:16, 15.42s/it]
 36%|███▌      | 3516/9864 [14:49:04<26:25:51, 14.99s/it]
 36%|███▌      | 3517/9864 [14:49:18<26:13:53, 14.88s/it]
 36%|███▌      | 3518/9864 [14:49:34<26:47:31, 15.20s/it]
 36%|███▌      | 3519/9864 [14:49:48<26:10:58, 14.86s/it]
 36%|███▌      | 3520/9864 [14:50:04<26:27:27, 15.01s/it]
                                                         
{'loss': '0.1552', 'grad_norm': '4.59', 'learning_rate': '1.49e-05', 'epoch': '0.3569'}

 36%|███▌      | 3520/9864 [14:50:04<26:27:27, 15.01s/it]
 36%|███▌      | 3521/9864 [14:50:21<27:45:37, 15.76s/it]
 36%|███▌      | 3522/9864 [14:50:35<26:51:20, 15.24s/it]
 36%|███▌      | 3523/9864 [14:50:50<26:27:32, 15.02s/it]
 36%|███▌      | 3524/9864 [14:51:04<25:53:38, 14.70s/it]
 36%|███▌      | 3525/9864 [14:51:20<26:33:59, 15.09s/it]
 36%|███▌      | 3526/9864 [14:51:34<26:09:30, 14.86s/it]
 36%|███▌      | 3527/9864 [14:51:50<26:28:04, 15.04s/it]
 36%|███▌      | 3528/9864 [14:52:04<25:53:07, 14.71s/it]
 36%|███▌      | 3529/9864 [14:52:21<27:25:17, 15.58s/it]
 36%|███▌      | 3530/9864 [14:52:35<26:38:56, 15.15s/it]
                                                         
{'loss': '0.1481', 'grad_norm': '2.359', 'learning_rate': '1.488e-05', 'epoch': '0.3579'}

 36%|███▌      | 3530/9864 [14:52:35<26:38:56, 15.15s/it]
 36%|███▌      | 3531/9864 [14:52:50<26:20:33, 14.97s/it]
 36%|███▌      | 3532/9864 [14:53:04<25:56:19, 14.75s/it]
 36%|███▌      | 3533/9864 [14:53:19<26:14:43, 14.92s/it]
 36%|███▌      | 3534/9864 [14:53:34<25:51:00, 14.70s/it]
 36%|███▌      | 3535/9864 [14:53:50<26:40:15, 15.17s/it]
 36%|███▌      | 3536/9864 [14:54:05<26:33:17, 15.11s/it]
 36%|███▌      | 3537/9864 [14:54:21<26:52:28, 15.29s/it]
 36%|███▌      | 3538/9864 [14:54:34<26:08:34, 14.88s/it]
 36%|███▌      | 3539/9864 [14:54:48<25:25:52, 14.47s/it]
 36%|███▌      | 3540/9864 [14:55:02<25:02:06, 14.25s/it]
                                                         
{'loss': '0.1425', 'grad_norm': '1.928', 'learning_rate': '1.485e-05', 'epoch': '0.3589'}

 36%|███▌      | 3540/9864 [14:55:02<25:02:06, 14.25s/it]
 36%|███▌      | 3541/9864 [14:55:18<25:58:26, 14.79s/it]
 36%|███▌      | 3542/9864 [14:55:34<26:44:13, 15.23s/it]
 36%|███▌      | 3543/9864 [14:55:50<27:02:42, 15.40s/it]
 36%|███▌      | 3544/9864 [14:56:06<27:11:59, 15.49s/it]
 36%|███▌      | 3545/9864 [14:56:20<26:39:56, 15.19s/it]
 36%|███▌      | 3546/9864 [14:56:34<26:06:09, 14.87s/it]
 36%|███▌      | 3547/9864 [14:56:48<25:41:15, 14.64s/it]
 36%|███▌      | 3548/9864 [14:57:05<26:33:56, 15.14s/it]
 36%|███▌      | 3549/9864 [14:57:19<25:58:28, 14.81s/it]
 36%|███▌      | 3550/9864 [14:57:36<27:27:28, 15.66s/it]
                                                         
{'loss': '0.1664', 'grad_norm': '3.939', 'learning_rate': '1.482e-05', 'epoch': '0.3599'}

 36%|███▌      | 3550/9864 [14:57:36<27:27:28, 15.66s/it]
 36%|███▌      | 3551/9864 [14:57:50<26:40:04, 15.21s/it]
 36%|███▌      | 3552/9864 [14:58:05<26:07:08, 14.90s/it]
 36%|███▌      | 3553/9864 [14:58:20<26:23:50, 15.06s/it]
 36%|███▌      | 3554/9864 [14:58:34<25:49:34, 14.73s/it]
 36%|███▌      | 3555/9864 [14:58:48<25:35:15, 14.60s/it]
 36%|███▌      | 3556/9864 [14:59:04<26:05:42, 14.89s/it]
 36%|███▌      | 3557/9864 [14:59:18<25:33:56, 14.59s/it]
 36%|███▌      | 3558/9864 [14:59:33<26:08:46, 14.93s/it]
 36%|███▌      | 3559/9864 [14:59:50<26:45:10, 15.28s/it]
 36%|███▌      | 3560/9864 [15:00:06<27:17:32, 15.59s/it]
                                                         
{'loss': '0.1625', 'grad_norm': '2.086', 'learning_rate': '1.479e-05', 'epoch': '0.3609'}

 36%|███▌      | 3560/9864 [15:00:06<27:17:32, 15.59s/it]
 36%|███▌      | 3561/9864 [15:00:20<26:25:35, 15.09s/it]
 36%|███▌      | 3562/9864 [15:00:34<25:52:40, 14.78s/it]
 36%|███▌      | 3563/9864 [15:00:48<25:39:28, 14.66s/it]
 36%|███▌      | 3564/9864 [15:01:05<26:34:48, 15.19s/it]
 36%|███▌      | 3565/9864 [15:01:19<26:17:54, 15.03s/it]
 36%|███▌      | 3566/9864 [15:01:35<26:40:22, 15.25s/it]
 36%|███▌      | 3567/9864 [15:01:51<27:00:31, 15.44s/it]
 36%|███▌      | 3568/9864 [15:02:06<27:00:09, 15.44s/it]
 36%|███▌      | 3569/9864 [15:02:21<26:25:39, 15.11s/it]
 36%|███▌      | 3570/9864 [15:02:35<25:55:07, 14.82s/it]
                                                         
{'loss': '0.149', 'grad_norm': '2.388', 'learning_rate': '1.476e-05', 'epoch': '0.362'}

 36%|███▌      | 3570/9864 [15:02:35<25:55:07, 14.82s/it]
 36%|███▌      | 3571/9864 [15:02:50<26:16:17, 15.03s/it]
 36%|███▌      | 3572/9864 [15:03:05<25:50:28, 14.79s/it]
 36%|███▌      | 3573/9864 [15:03:18<25:11:26, 14.42s/it]
 36%|███▌      | 3574/9864 [15:03:33<25:33:22, 14.63s/it]
 36%|███▌      | 3575/9864 [15:03:49<26:15:03, 15.03s/it]
 36%|███▋      | 3576/9864 [15:04:05<26:47:09, 15.34s/it]
 36%|███▋      | 3577/9864 [15:04:20<26:16:09, 15.04s/it]
 36%|███▋      | 3578/9864 [15:04:35<26:37:07, 15.24s/it]
 36%|███▋      | 3579/9864 [15:04:49<25:50:06, 14.80s/it]
 36%|███▋      | 3580/9864 [15:05:03<25:29:03, 14.60s/it]
                                                         
{'loss': '0.1473', 'grad_norm': '4.008', 'learning_rate': '1.473e-05', 'epoch': '0.363'}

 36%|███▋      | 3580/9864 [15:05:03<25:29:03, 14.60s/it]
 36%|███▋      | 3581/9864 [15:05:19<25:59:49, 14.90s/it]
 36%|███▋      | 3582/9864 [15:05:35<26:32:24, 15.21s/it]
 36%|███▋      | 3583/9864 [15:05:49<25:53:34, 14.84s/it]
 36%|███▋      | 3584/9864 [15:06:04<26:15:05, 15.05s/it]
 36%|███▋      | 3585/9864 [15:06:20<26:40:08, 15.29s/it]
 36%|███▋      | 3586/9864 [15:06:34<26:05:39, 14.96s/it]
 36%|███▋      | 3587/9864 [15:06:49<25:42:09, 14.74s/it]
 36%|███▋      | 3588/9864 [15:07:03<25:31:35, 14.64s/it]
 36%|███▋      | 3589/9864 [15:07:21<27:08:25, 15.57s/it]
 36%|███▋      | 3590/9864 [15:07:35<26:13:39, 15.05s/it]
                                                         
{'loss': '0.1574', 'grad_norm': '3.675', 'learning_rate': '1.47e-05', 'epoch': '0.364'}

 36%|███▋      | 3590/9864 [15:07:35<26:13:39, 15.05s/it]
 36%|███▋      | 3591/9864 [15:07:49<25:46:50, 14.80s/it]
 36%|███▋      | 3592/9864 [15:08:03<25:41:47, 14.75s/it]
 36%|███▋      | 3593/9864 [15:08:19<26:04:59, 14.97s/it]
 36%|███▋      | 3594/9864 [15:08:35<26:26:26, 15.18s/it]
 36%|███▋      | 3595/9864 [15:08:49<25:59:04, 14.92s/it]
 36%|███▋      | 3596/9864 [15:09:05<26:28:18, 15.20s/it]
 36%|███▋      | 3597/9864 [15:09:21<26:49:29, 15.41s/it]
 36%|███▋      | 3598/9864 [15:09:35<26:04:59, 14.99s/it]
 36%|███▋      | 3599/9864 [15:09:49<25:39:43, 14.75s/it]
 36%|███▋      | 3600/9864 [15:10:03<25:24:10, 14.60s/it]
                                                         
{'loss': '0.1386', 'grad_norm': '3.958', 'learning_rate': '1.467e-05', 'epoch': '0.365'}

 36%|███▋      | 3600/9864 [15:10:03<25:24:10, 14.60s/it]
 37%|███▋      | 3601/9864 [15:10:21<27:03:52, 15.56s/it]
 37%|███▋      | 3602/9864 [15:10:35<26:25:09, 15.19s/it]
 37%|███▋      | 3603/9864 [15:10:49<25:56:21, 14.91s/it]
 37%|███▋      | 3604/9864 [15:11:03<25:27:54, 14.64s/it]
 37%|███▋      | 3605/9864 [15:11:21<26:47:12, 15.41s/it]
 37%|███▋      | 3606/9864 [15:11:35<26:11:38, 15.07s/it]
 37%|███▋      | 3607/9864 [15:11:48<25:17:49, 14.55s/it]
 37%|███▋      | 3608/9864 [15:12:04<26:03:12, 14.99s/it]
 37%|███▋      | 3609/9864 [15:12:19<25:39:45, 14.77s/it]
 37%|███▋      | 3610/9864 [15:12:32<25:08:08, 14.47s/it]
                                                         
{'loss': '0.1466', 'grad_norm': '5.128', 'learning_rate': '1.464e-05', 'epoch': '0.366'}

 37%|███▋      | 3610/9864 [15:12:32<25:08:08, 14.47s/it]
 37%|███▋      | 3611/9864 [15:12:47<25:06:46, 14.46s/it]
 37%|███▋      | 3612/9864 [15:13:01<25:10:05, 14.49s/it]
 37%|███▋      | 3613/9864 [15:13:17<25:42:31, 14.81s/it]
 37%|███▋      | 3614/9864 [15:13:33<26:18:57, 15.16s/it]
 37%|███▋      | 3615/9864 [15:13:49<26:38:27, 15.35s/it]
 37%|███▋      | 3616/9864 [15:14:03<26:00:49, 14.99s/it]
 37%|███▋      | 3617/9864 [15:14:18<26:17:17, 15.15s/it]
 37%|███▋      | 3618/9864 [15:14:32<25:37:16, 14.77s/it]
 37%|███▋      | 3619/9864 [15:14:46<25:16:23, 14.57s/it]
 37%|███▋      | 3620/9864 [15:15:02<25:53:37, 14.93s/it]
                                                         
{'loss': '0.1651', 'grad_norm': '1.54', 'learning_rate': '1.462e-05', 'epoch': '0.367'}

 37%|███▋      | 3620/9864 [15:15:02<25:53:37, 14.93s/it]
 37%|███▋      | 3621/9864 [15:15:18<26:11:38, 15.10s/it]
 37%|███▋      | 3622/9864 [15:15:33<26:27:16, 15.26s/it]
 37%|███▋      | 3623/9864 [15:15:48<26:13:10, 15.12s/it]
 37%|███▋      | 3624/9864 [15:16:02<25:40:22, 14.81s/it]
 37%|███▋      | 3625/9864 [15:16:18<26:04:44, 15.05s/it]
 37%|███▋      | 3626/9864 [15:16:32<25:46:57, 14.88s/it]
 37%|███▋      | 3627/9864 [15:16:48<26:13:52, 15.14s/it]
 37%|███▋      | 3628/9864 [15:17:02<25:45:50, 14.87s/it]
 37%|███▋      | 3629/9864 [15:17:18<26:11:52, 15.13s/it]
 37%|███▋      | 3630/9864 [15:17:32<25:42:04, 14.84s/it]
                                                         
{'loss': '0.1587', 'grad_norm': '3.901', 'learning_rate': '1.459e-05', 'epoch': '0.368'}

 37%|███▋      | 3630/9864 [15:17:32<25:42:04, 14.84s/it]
 37%|███▋      | 3631/9864 [15:17:48<26:13:25, 15.15s/it]
 37%|███▋      | 3632/9864 [15:18:02<25:43:35, 14.86s/it]
 37%|███▋      | 3633/9864 [15:18:17<25:36:11, 14.79s/it]
 37%|███▋      | 3634/9864 [15:18:33<26:16:11, 15.18s/it]
 37%|███▋      | 3635/9864 [15:18:49<26:40:48, 15.42s/it]
 37%|███▋      | 3636/9864 [15:19:03<25:54:26, 14.98s/it]
 37%|███▋      | 3637/9864 [15:19:19<26:40:58, 15.43s/it]
 37%|███▋      | 3638/9864 [15:19:35<26:47:16, 15.49s/it]
 37%|███▋      | 3639/9864 [15:19:49<26:14:26, 15.18s/it]
 37%|███▋      | 3640/9864 [15:20:03<25:42:30, 14.87s/it]
                                                         
{'loss': '0.1574', 'grad_norm': '3.288', 'learning_rate': '1.456e-05', 'epoch': '0.369'}

 37%|███▋      | 3640/9864 [15:20:03<25:42:30, 14.87s/it]
 37%|███▋      | 3641/9864 [15:20:18<25:29:30, 14.75s/it]
 37%|███▋      | 3642/9864 [15:20:34<25:59:18, 15.04s/it]
 37%|███▋      | 3643/9864 [15:20:50<26:38:02, 15.41s/it]
 37%|███▋      | 3644/9864 [15:21:05<26:21:49, 15.26s/it]
 37%|███▋      | 3645/9864 [15:21:20<26:12:24, 15.17s/it]
 37%|███▋      | 3646/9864 [15:21:35<26:27:40, 15.32s/it]
 37%|███▋      | 3647/9864 [15:21:50<25:56:59, 15.03s/it]
 37%|███▋      | 3648/9864 [15:22:04<25:39:18, 14.86s/it]
 37%|███▋      | 3649/9864 [15:22:18<25:19:21, 14.67s/it]
 37%|███▋      | 3650/9864 [15:22:35<26:05:24, 15.12s/it]
                                                         
{'loss': '0.1495', 'grad_norm': '2.323', 'learning_rate': '1.453e-05', 'epoch': '0.3701'}

 37%|███▋      | 3650/9864 [15:22:35<26:05:24, 15.12s/it]
 37%|███▋      | 3651/9864 [15:22:49<25:38:19, 14.86s/it]
 37%|███▋      | 3652/9864 [15:23:05<26:12:39, 15.19s/it]
 37%|███▋      | 3653/9864 [15:23:19<25:35:12, 14.83s/it]
 37%|███▋      | 3654/9864 [15:23:36<26:56:26, 15.62s/it]
 37%|███▋      | 3655/9864 [15:23:50<26:10:34, 15.18s/it]
 37%|███▋      | 3656/9864 [15:24:05<25:35:29, 14.84s/it]
 37%|███▋      | 3657/9864 [15:24:20<26:04:57, 15.13s/it]
 37%|███▋      | 3658/9864 [15:24:34<25:32:51, 14.82s/it]
 37%|███▋      | 3659/9864 [15:24:48<25:03:44, 14.54s/it]
 37%|███▋      | 3660/9864 [15:25:04<25:37:15, 14.87s/it]
                                                         
{'loss': '0.1378', 'grad_norm': '2.559', 'learning_rate': '1.45e-05', 'epoch': '0.3711'}

 37%|███▋      | 3660/9864 [15:25:04<25:37:15, 14.87s/it]
 37%|███▋      | 3661/9864 [15:25:22<27:06:18, 15.73s/it]
 37%|███▋      | 3662/9864 [15:25:35<26:05:30, 15.15s/it]
 37%|███▋      | 3663/9864 [15:25:50<25:43:17, 14.93s/it]
 37%|███▋      | 3664/9864 [15:26:06<26:19:23, 15.28s/it]
 37%|███▋      | 3665/9864 [15:26:20<25:46:41, 14.97s/it]
 37%|███▋      | 3666/9864 [15:26:34<25:14:59, 14.67s/it]
 37%|███▋      | 3667/9864 [15:26:48<24:54:47, 14.47s/it]
 37%|███▋      | 3668/9864 [15:27:04<25:37:58, 14.89s/it]
 37%|███▋      | 3669/9864 [15:27:20<26:14:54, 15.25s/it]
 37%|███▋      | 3670/9864 [15:27:36<26:23:07, 15.34s/it]
                                                         
{'loss': '0.1672', 'grad_norm': '2.131', 'learning_rate': '1.447e-05', 'epoch': '0.3721'}

 37%|███▋      | 3670/9864 [15:27:36<26:23:07, 15.34s/it]
 37%|███▋      | 3671/9864 [15:27:52<26:48:58, 15.59s/it]
 37%|███▋      | 3672/9864 [15:28:06<26:10:34, 15.22s/it]
 37%|███▋      | 3673/9864 [15:28:20<25:37:25, 14.90s/it]
 37%|███▋      | 3674/9864 [15:28:35<25:15:10, 14.69s/it]
 37%|███▋      | 3675/9864 [15:28:49<24:56:31, 14.51s/it]
 37%|███▋      | 3676/9864 [15:29:05<25:49:02, 15.02s/it]
 37%|███▋      | 3677/9864 [15:29:22<27:00:01, 15.71s/it]
 37%|███▋      | 3678/9864 [15:29:36<26:02:51, 15.16s/it]
 37%|███▋      | 3679/9864 [15:29:50<25:39:12, 14.93s/it]
 37%|███▋      | 3680/9864 [15:30:07<26:14:30, 15.28s/it]
                                                         
{'loss': '0.1523', 'grad_norm': '3.654', 'learning_rate': '1.444e-05', 'epoch': '0.3731'}

 37%|███▋      | 3680/9864 [15:30:07<26:14:30, 15.28s/it]
 37%|███▋      | 3681/9864 [15:30:21<25:44:12, 14.99s/it]
 37%|███▋      | 3682/9864 [15:30:35<25:07:54, 14.64s/it]
 37%|███▋      | 3683/9864 [15:30:49<24:53:21, 14.50s/it]
 37%|███▋      | 3684/9864 [15:31:07<26:29:48, 15.44s/it]
 37%|███▋      | 3685/9864 [15:31:23<27:01:38, 15.75s/it]
 37%|███▋      | 3686/9864 [15:31:38<26:25:05, 15.39s/it]
 37%|███▋      | 3687/9864 [15:31:53<26:39:59, 15.54s/it]
 37%|███▋      | 3688/9864 [15:32:08<25:59:17, 15.15s/it]
 37%|███▋      | 3689/9864 [15:32:22<25:32:00, 14.89s/it]
 37%|███▋      | 3690/9864 [15:32:36<24:59:18, 14.57s/it]
                                                         
{'loss': '0.1432', 'grad_norm': '2.381', 'learning_rate': '1.441e-05', 'epoch': '0.3741'}

 37%|███▋      | 3690/9864 [15:32:36<24:59:18, 14.57s/it]
 37%|███▋      | 3691/9864 [15:32:50<24:50:28, 14.49s/it]
 37%|███▋      | 3692/9864 [15:33:06<25:43:18, 15.00s/it]
 37%|███▋      | 3693/9864 [15:33:22<26:14:10, 15.31s/it]
 37%|███▋      | 3694/9864 [15:33:40<27:27:18, 16.02s/it]
 37%|███▋      | 3695/9864 [15:33:54<26:25:40, 15.42s/it]
 37%|███▋      | 3696/9864 [15:34:08<25:54:52, 15.13s/it]
 37%|███▋      | 3697/9864 [15:34:23<25:28:45, 14.87s/it]
 37%|███▋      | 3698/9864 [15:34:37<25:03:58, 14.63s/it]
 38%|███▊      | 3699/9864 [15:34:53<26:02:34, 15.21s/it]
 38%|███▊      | 3700/9864 [15:35:09<26:16:54, 15.35s/it]
                                                         
{'loss': '0.1353', 'grad_norm': '2.266', 'learning_rate': '1.438e-05', 'epoch': '0.3751'}

 38%|███▊      | 3700/9864 [15:35:09<26:16:54, 15.35s/it]
 38%|███▊      | 3701/9864 [15:35:23<25:32:17, 14.92s/it]
 38%|███▊      | 3702/9864 [15:35:40<26:42:13, 15.60s/it]
 38%|███▊      | 3703/9864 [15:35:54<26:00:22, 15.20s/it]
 38%|███▊      | 3704/9864 [15:36:09<25:42:58, 15.03s/it]
 38%|███▊      | 3705/9864 [15:36:23<25:12:19, 14.73s/it]
 38%|███▊      | 3706/9864 [15:36:39<25:52:42, 15.13s/it]
 38%|███▊      | 3707/9864 [15:36:53<25:14:44, 14.76s/it]
 38%|███▊      | 3708/9864 [15:37:09<26:01:10, 15.22s/it]
 38%|███▊      | 3709/9864 [15:37:25<26:27:28, 15.47s/it]
 38%|███▊      | 3710/9864 [15:37:41<26:38:04, 15.58s/it]
                                                         
{'loss': '0.1489', 'grad_norm': '3.277', 'learning_rate': '1.435e-05', 'epoch': '0.3761'}

 38%|███▊      | 3710/9864 [15:37:41<26:38:04, 15.58s/it]
 38%|███▊      | 3711/9864 [15:37:55<25:54:10, 15.16s/it]
 38%|███▊      | 3712/9864 [15:38:09<24:58:57, 14.62s/it]
 38%|███▊      | 3713/9864 [15:38:24<25:27:40, 14.90s/it]
 38%|███▊      | 3714/9864 [15:38:38<24:54:01, 14.58s/it]
 38%|███▊      | 3715/9864 [15:38:54<25:29:53, 14.93s/it]
 38%|███▊      | 3716/9864 [15:39:08<24:52:04, 14.56s/it]
 38%|███▊      | 3717/9864 [15:39:23<25:22:57, 14.87s/it]
 38%|███▊      | 3718/9864 [15:39:39<25:52:11, 15.15s/it]
 38%|███▊      | 3719/9864 [15:39:52<25:02:01, 14.67s/it]
 38%|███▊      | 3720/9864 [15:40:07<24:58:49, 14.64s/it]
                                                         
{'loss': '0.157', 'grad_norm': '3.62', 'learning_rate': '1.432e-05', 'epoch': '0.3772'}

 38%|███▊      | 3720/9864 [15:40:07<24:58:49, 14.64s/it]
 38%|███▊      | 3721/9864 [15:40:23<25:27:55, 14.92s/it]
 38%|███▊      | 3722/9864 [15:40:39<26:13:30, 15.37s/it]
 38%|███▊      | 3723/9864 [15:40:53<25:38:25, 15.03s/it]
 38%|███▊      | 3724/9864 [15:41:08<25:30:41, 14.96s/it]
 38%|███▊      | 3725/9864 [15:41:23<25:13:53, 14.80s/it]
 38%|███▊      | 3726/9864 [15:41:39<25:56:16, 15.21s/it]
 38%|███▊      | 3727/9864 [15:41:55<26:14:19, 15.39s/it]
 38%|███▊      | 3728/9864 [15:42:08<25:27:54, 14.94s/it]
 38%|███▊      | 3729/9864 [15:42:24<25:56:24, 15.22s/it]
 38%|███▊      | 3730/9864 [15:42:41<26:36:31, 15.62s/it]
                                                         
{'loss': '0.1508', 'grad_norm': '3.723', 'learning_rate': '1.429e-05', 'epoch': '0.3782'}

 38%|███▊      | 3730/9864 [15:42:41<26:36:31, 15.62s/it]
 38%|███▊      | 3731/9864 [15:42:55<25:49:31, 15.16s/it]
 38%|███▊      | 3732/9864 [15:43:09<25:25:30, 14.93s/it]
 38%|███▊      | 3733/9864 [15:43:23<25:00:28, 14.68s/it]
 38%|███▊      | 3734/9864 [15:43:39<25:41:20, 15.09s/it]
 38%|███▊      | 3735/9864 [15:43:55<26:08:26, 15.35s/it]
 38%|███▊      | 3736/9864 [15:44:09<25:10:12, 14.79s/it]
 38%|███▊      | 3737/9864 [15:44:27<26:49:50, 15.76s/it]
 38%|███▊      | 3738/9864 [15:44:41<26:09:51, 15.38s/it]
 38%|███▊      | 3739/9864 [15:44:55<25:23:34, 14.92s/it]
 38%|███▊      | 3740/9864 [15:45:10<25:07:43, 14.77s/it]
                                                         
{'loss': '0.1409', 'grad_norm': '3.057', 'learning_rate': '1.426e-05', 'epoch': '0.3792'}

 38%|███▊      | 3740/9864 [15:45:10<25:07:43, 14.77s/it]
 38%|███▊      | 3741/9864 [15:45:24<24:47:49, 14.58s/it]
 38%|███▊      | 3742/9864 [15:45:40<25:47:01, 15.16s/it]
 38%|███▊      | 3743/9864 [15:45:57<26:19:27, 15.48s/it]
 38%|███▊      | 3744/9864 [15:46:12<26:23:45, 15.53s/it]
 38%|███▊      | 3745/9864 [15:46:29<26:53:59, 15.83s/it]
 38%|███▊      | 3746/9864 [15:46:43<26:11:27, 15.41s/it]
 38%|███▊      | 3747/9864 [15:46:57<25:23:47, 14.95s/it]
 38%|███▊      | 3748/9864 [15:47:11<25:05:07, 14.77s/it]
 38%|███▊      | 3749/9864 [15:47:25<24:39:50, 14.52s/it]
 38%|███▊      | 3750/9864 [15:47:41<25:27:26, 14.99s/it]
                                                         
{'loss': '0.1633', 'grad_norm': '4.184', 'learning_rate': '1.423e-05', 'epoch': '0.3802'}

 38%|███▊      | 3750/9864 [15:47:41<25:27:26, 14.99s/it]
 38%|███▊      | 3751/9864 [15:47:58<26:19:18, 15.50s/it]
 38%|███▊      | 3752/9864 [15:48:12<25:44:35, 15.16s/it]
 38%|███▊      | 3753/9864 [15:48:30<27:01:32, 15.92s/it]
 38%|███▊      | 3754/9864 [15:48:44<26:10:49, 15.43s/it]
 38%|███▊      | 3755/9864 [15:48:59<25:33:09, 15.06s/it]
 38%|███▊      | 3756/9864 [15:49:13<25:02:49, 14.76s/it]
 38%|███▊      | 3757/9864 [15:49:27<24:38:57, 14.53s/it]
 38%|███▊      | 3758/9864 [15:49:41<24:27:45, 14.42s/it]
 38%|███▊      | 3759/9864 [15:49:57<25:30:55, 15.05s/it]
 38%|███▊      | 3760/9864 [15:50:13<25:59:36, 15.33s/it]
                                                         
{'loss': '0.1634', 'grad_norm': '1.627', 'learning_rate': '1.42e-05', 'epoch': '0.3812'}

 38%|███▊      | 3760/9864 [15:50:13<25:59:36, 15.33s/it]
 38%|███▊      | 3761/9864 [15:50:30<26:26:34, 15.60s/it]
 38%|███▊      | 3762/9864 [15:50:45<26:32:56, 15.66s/it]
 38%|███▊      | 3763/9864 [15:50:59<25:43:51, 15.18s/it]
 38%|███▊      | 3764/9864 [15:51:14<25:13:58, 14.89s/it]
 38%|███▊      | 3765/9864 [15:51:29<25:24:27, 15.00s/it]
 38%|███▊      | 3766/9864 [15:51:43<24:58:33, 14.74s/it]
 38%|███▊      | 3767/9864 [15:52:01<26:37:48, 15.72s/it]
 38%|███▊      | 3768/9864 [15:52:15<25:53:19, 15.29s/it]
 38%|███▊      | 3769/9864 [15:52:31<26:15:30, 15.51s/it]
 38%|███▊      | 3770/9864 [15:52:48<26:35:53, 15.71s/it]
                                                         
{'loss': '0.1488', 'grad_norm': '2.369', 'learning_rate': '1.417e-05', 'epoch': '0.3822'}

 38%|███▊      | 3770/9864 [15:52:48<26:35:53, 15.71s/it]
 38%|███▊      | 3771/9864 [15:53:02<25:47:23, 15.24s/it]
 38%|███▊      | 3772/9864 [15:53:16<25:29:30, 15.06s/it]
 38%|███▊      | 3773/9864 [15:53:31<25:09:36, 14.87s/it]
 38%|███▊      | 3774/9864 [15:53:45<24:49:39, 14.68s/it]
 38%|███▊      | 3775/9864 [15:54:01<25:31:45, 15.09s/it]
 38%|███▊      | 3776/9864 [15:54:17<26:08:49, 15.46s/it]
 38%|███▊      | 3777/9864 [15:54:33<26:13:31, 15.51s/it]
 38%|███▊      | 3778/9864 [15:54:49<26:31:32, 15.69s/it]
 38%|███▊      | 3779/9864 [15:55:03<25:47:25, 15.26s/it]
 38%|███▊      | 3780/9864 [15:55:17<25:08:30, 14.88s/it]
                                                         
{'loss': '0.1461', 'grad_norm': '3.347', 'learning_rate': '1.414e-05', 'epoch': '0.3832'}

 38%|███▊      | 3780/9864 [15:55:17<25:08:30, 14.88s/it]
 38%|███▊      | 3781/9864 [15:55:32<24:50:38, 14.70s/it]
 38%|███▊      | 3782/9864 [15:55:46<24:42:11, 14.62s/it]
 38%|███▊      | 3783/9864 [15:56:02<25:22:31, 15.02s/it]
 38%|███▊      | 3784/9864 [15:56:18<25:48:14, 15.28s/it]
 38%|███▊      | 3785/9864 [15:56:36<27:01:03, 16.00s/it]
 38%|███▊      | 3786/9864 [15:56:50<26:16:42, 15.56s/it]
 38%|███▊      | 3787/9864 [15:57:04<25:31:39, 15.12s/it]
 38%|███▊      | 3788/9864 [15:57:19<25:11:21, 14.92s/it]
 38%|███▊      | 3789/9864 [15:57:32<24:35:04, 14.57s/it]
 38%|███▊      | 3790/9864 [15:57:48<25:09:39, 14.91s/it]
                                                         
{'loss': '0.1589', 'grad_norm': '2.684', 'learning_rate': '1.411e-05', 'epoch': '0.3843'}

 38%|███▊      | 3790/9864 [15:57:48<25:09:39, 14.91s/it]
 38%|███▊      | 3791/9864 [15:58:04<25:37:21, 15.19s/it]
 38%|███▊      | 3792/9864 [15:58:20<26:05:16, 15.47s/it]
 38%|███▊      | 3793/9864 [15:58:36<26:20:32, 15.62s/it]
 38%|███▊      | 3794/9864 [15:58:50<25:33:47, 15.16s/it]
 38%|███▊      | 3795/9864 [15:59:05<25:14:31, 14.97s/it]
 38%|███▊      | 3796/9864 [15:59:19<25:05:16, 14.88s/it]
 38%|███▊      | 3797/9864 [15:59:34<24:45:35, 14.69s/it]
 39%|███▊      | 3798/9864 [15:59:50<25:24:02, 15.07s/it]
 39%|███▊      | 3799/9864 [16:00:04<24:52:32, 14.77s/it]
 39%|███▊      | 3800/9864 [16:00:21<26:20:52, 15.64s/it]
                                                         
{'loss': '0.1599', 'grad_norm': '2.977', 'learning_rate': '1.408e-05', 'epoch': '0.3853'}

 39%|███▊      | 3800/9864 [16:00:21<26:20:52, 15.64s/it]
 39%|███▊      | 3801/9864 [16:00:37<26:35:14, 15.79s/it]
 39%|███▊      | 3802/9864 [16:00:52<25:58:44, 15.43s/it]
 39%|███▊      | 3803/9864 [16:01:06<25:23:08, 15.08s/it]
 39%|███▊      | 3804/9864 [16:01:21<25:07:33, 14.93s/it]
 39%|███▊      | 3805/9864 [16:01:35<24:45:39, 14.71s/it]
 39%|███▊      | 3806/9864 [16:01:51<25:08:44, 14.94s/it]
 39%|███▊      | 3807/9864 [16:02:06<25:23:02, 15.09s/it]
 39%|███▊      | 3808/9864 [16:02:24<26:47:26, 15.93s/it]
 39%|███▊      | 3809/9864 [16:02:38<25:50:05, 15.36s/it]
 39%|███▊      | 3810/9864 [16:02:52<25:05:37, 14.92s/it]
                                                         
{'loss': '0.1543', 'grad_norm': '2.117', 'learning_rate': '1.405e-05', 'epoch': '0.3863'}

 39%|███▊      | 3810/9864 [16:02:52<25:05:37, 14.92s/it]
 39%|███▊      | 3811/9864 [16:03:06<24:50:13, 14.77s/it]
 39%|███▊      | 3812/9864 [16:03:20<24:27:12, 14.55s/it]
 39%|███▊      | 3813/9864 [16:03:34<24:03:47, 14.32s/it]
 39%|███▊      | 3814/9864 [16:03:50<24:55:57, 14.84s/it]
 39%|███▊      | 3815/9864 [16:04:06<25:25:45, 15.13s/it]
 39%|███▊      | 3816/9864 [16:04:22<25:41:36, 15.29s/it]
 39%|███▊      | 3817/9864 [16:04:37<25:58:18, 15.46s/it]
 39%|███▊      | 3818/9864 [16:04:52<25:28:01, 15.16s/it]
 39%|███▊      | 3819/9864 [16:05:06<25:04:16, 14.93s/it]
 39%|███▊      | 3820/9864 [16:05:20<24:39:51, 14.69s/it]
                                                         
{'loss': '0.1488', 'grad_norm': '3.053', 'learning_rate': '1.402e-05', 'epoch': '0.3873'}

 39%|███▊      | 3820/9864 [16:05:20<24:39:51, 14.69s/it]
 39%|███▊      | 3821/9864 [16:05:34<24:10:54, 14.41s/it]
 39%|███▊      | 3822/9864 [16:05:50<25:04:21, 14.94s/it]
 39%|███▉      | 3823/9864 [16:06:08<26:19:47, 15.69s/it]
 39%|███▉      | 3824/9864 [16:06:22<25:36:02, 15.26s/it]
 39%|███▉      | 3825/9864 [16:06:37<25:38:38, 15.29s/it]
 39%|███▉      | 3826/9864 [16:06:51<24:58:08, 14.89s/it]
 39%|███▉      | 3827/9864 [16:07:06<24:46:14, 14.77s/it]
 39%|███▉      | 3828/9864 [16:07:20<24:27:35, 14.59s/it]
 39%|███▉      | 3829/9864 [16:07:36<25:11:09, 15.02s/it]
 39%|███▉      | 3830/9864 [16:07:50<24:41:30, 14.73s/it]
                                                         
{'loss': '0.1498', 'grad_norm': '2.674', 'learning_rate': '1.399e-05', 'epoch': '0.3883'}

 39%|███▉      | 3830/9864 [16:07:50<24:41:30, 14.73s/it]
 39%|███▉      | 3831/9864 [16:08:05<24:56:10, 14.88s/it]
 39%|███▉      | 3832/9864 [16:08:23<26:13:09, 15.65s/it]
 39%|███▉      | 3833/9864 [16:08:37<25:35:18, 15.27s/it]
 39%|███▉      | 3834/9864 [16:08:51<24:57:06, 14.90s/it]
 39%|███▉      | 3835/9864 [16:09:06<24:41:32, 14.74s/it]
 39%|███▉      | 3836/9864 [16:09:21<25:14:50, 15.08s/it]
 39%|███▉      | 3837/9864 [16:09:36<25:01:43, 14.95s/it]
 39%|███▉      | 3838/9864 [16:09:52<25:33:10, 15.27s/it]
 39%|███▉      | 3839/9864 [16:10:08<25:54:41, 15.48s/it]
 39%|███▉      | 3840/9864 [16:10:23<25:23:14, 15.17s/it]
                                                         
{'loss': '0.1593', 'grad_norm': '2.299', 'learning_rate': '1.396e-05', 'epoch': '0.3893'}

 39%|███▉      | 3840/9864 [16:10:23<25:23:14, 15.17s/it]
 39%|███▉      | 3841/9864 [16:10:38<25:38:55, 15.33s/it]
 39%|███▉      | 3842/9864 [16:10:52<25:04:30, 14.99s/it]
 39%|███▉      | 3843/9864 [16:11:08<25:37:20, 15.32s/it]
 39%|███▉      | 3844/9864 [16:11:22<24:48:47, 14.84s/it]
 39%|███▉      | 3845/9864 [16:11:36<24:30:35, 14.66s/it]
 39%|███▉      | 3846/9864 [16:11:53<25:31:46, 15.27s/it]
 39%|███▉      | 3847/9864 [16:12:09<25:52:11, 15.48s/it]
 39%|███▉      | 3848/9864 [16:12:24<25:21:10, 15.17s/it]
 39%|███▉      | 3849/9864 [16:12:40<26:03:42, 15.60s/it]
 39%|███▉      | 3850/9864 [16:12:54<25:07:56, 15.04s/it]
                                                         
{'loss': '0.1512', 'grad_norm': '1.721', 'learning_rate': '1.393e-05', 'epoch': '0.3903'}

 39%|███▉      | 3850/9864 [16:12:54<25:07:56, 15.04s/it]
 39%|███▉      | 3851/9864 [16:13:08<24:45:15, 14.82s/it]
 39%|███▉      | 3852/9864 [16:13:24<25:24:31, 15.21s/it]
 39%|███▉      | 3853/9864 [16:13:38<24:48:30, 14.86s/it]
 39%|███▉      | 3854/9864 [16:13:55<25:29:44, 15.27s/it]
 39%|███▉      | 3855/9864 [16:14:09<24:59:51, 14.98s/it]
 39%|███▉      | 3856/9864 [16:14:24<25:13:14, 15.11s/it]
 39%|███▉      | 3857/9864 [16:14:41<25:48:40, 15.47s/it]
 39%|███▉      | 3858/9864 [16:14:55<25:16:44, 15.15s/it]
 39%|███▉      | 3859/9864 [16:15:09<24:45:37, 14.84s/it]
 39%|███▉      | 3860/9864 [16:15:25<25:26:26, 15.25s/it]
                                                         
{'loss': '0.1571', 'grad_norm': '4.039', 'learning_rate': '1.39e-05', 'epoch': '0.3914'}

 39%|███▉      | 3860/9864 [16:15:25<25:26:26, 15.25s/it]
 39%|███▉      | 3861/9864 [16:15:40<25:06:15, 15.06s/it]
 39%|███▉      | 3862/9864 [16:15:56<25:34:00, 15.33s/it]
 39%|███▉      | 3863/9864 [16:16:10<24:45:31, 14.85s/it]
 39%|███▉      | 3864/9864 [16:16:25<25:08:49, 15.09s/it]
 39%|███▉      | 3865/9864 [16:16:40<24:44:00, 14.84s/it]
 39%|███▉      | 3866/9864 [16:16:56<25:26:57, 15.27s/it]
 39%|███▉      | 3867/9864 [16:17:10<24:57:26, 14.98s/it]
 39%|███▉      | 3868/9864 [16:17:26<25:21:31, 15.23s/it]
 39%|███▉      | 3869/9864 [16:17:42<25:35:56, 15.37s/it]
 39%|███▉      | 3870/9864 [16:17:56<24:49:57, 14.91s/it]
                                                         
{'loss': '0.1561', 'grad_norm': '2.126', 'learning_rate': '1.387e-05', 'epoch': '0.3924'}

 39%|███▉      | 3870/9864 [16:17:56<24:49:57, 14.91s/it]
 39%|███▉      | 3871/9864 [16:18:10<24:29:02, 14.71s/it]
 39%|███▉      | 3872/9864 [16:18:24<24:08:18, 14.50s/it]
 39%|███▉      | 3873/9864 [16:18:41<25:32:46, 15.35s/it]
 39%|███▉      | 3874/9864 [16:18:56<25:04:46, 15.07s/it]
 39%|███▉      | 3875/9864 [16:19:11<25:22:23, 15.25s/it]
 39%|███▉      | 3876/9864 [16:19:27<25:41:02, 15.44s/it]
 39%|███▉      | 3877/9864 [16:19:42<25:22:46, 15.26s/it]
 39%|███▉      | 3878/9864 [16:19:56<24:51:44, 14.95s/it]
 39%|███▉      | 3879/9864 [16:20:11<24:40:56, 14.85s/it]
 39%|███▉      | 3880/9864 [16:20:28<25:58:52, 15.63s/it]
                                                         
{'loss': '0.1474', 'grad_norm': '2.066', 'learning_rate': '1.384e-05', 'epoch': '0.3934'}

 39%|███▉      | 3880/9864 [16:20:28<25:58:52, 15.63s/it]
 39%|███▉      | 3881/9864 [16:20:42<25:03:18, 15.08s/it]
 39%|███▉      | 3882/9864 [16:20:57<24:56:58, 15.01s/it]
 39%|███▉      | 3883/9864 [16:21:13<25:16:47, 15.22s/it]
 39%|███▉      | 3884/9864 [16:21:29<25:42:29, 15.48s/it]
 39%|███▉      | 3885/9864 [16:21:43<25:05:00, 15.10s/it]
 39%|███▉      | 3886/9864 [16:21:57<24:48:45, 14.94s/it]
 39%|███▉      | 3887/9864 [16:22:12<24:23:42, 14.69s/it]
 39%|███▉      | 3888/9864 [16:22:27<25:01:04, 15.07s/it]
 39%|███▉      | 3889/9864 [16:22:43<25:19:30, 15.26s/it]
 39%|███▉      | 3890/9864 [16:22:59<25:41:00, 15.48s/it]
                                                         
{'loss': '0.1386', 'grad_norm': '4.123', 'learning_rate': '1.381e-05', 'epoch': '0.3944'}

 39%|███▉      | 3890/9864 [16:22:59<25:41:00, 15.48s/it]
 39%|███▉      | 3891/9864 [16:23:15<25:45:16, 15.52s/it]
 39%|███▉      | 3892/9864 [16:23:29<24:54:50, 15.02s/it]
 39%|███▉      | 3893/9864 [16:23:43<24:24:14, 14.71s/it]
 39%|███▉      | 3894/9864 [16:23:57<24:02:03, 14.49s/it]
 39%|███▉      | 3895/9864 [16:24:11<23:59:04, 14.47s/it]
 39%|███▉      | 3896/9864 [16:24:27<24:41:06, 14.89s/it]
 40%|███▉      | 3897/9864 [16:24:42<25:01:16, 15.10s/it]
 40%|███▉      | 3898/9864 [16:24:58<25:23:47, 15.32s/it]
 40%|███▉      | 3899/9864 [16:25:15<25:49:31, 15.59s/it]
 40%|███▉      | 3900/9864 [16:25:28<24:57:28, 15.07s/it]
                                                         
{'loss': '0.1439', 'grad_norm': '5.117', 'learning_rate': '1.378e-05', 'epoch': '0.3954'}

 40%|███▉      | 3900/9864 [16:25:28<24:57:28, 15.07s/it]
 40%|███▉      | 3901/9864 [16:25:42<24:28:30, 14.78s/it]
 40%|███▉      | 3902/9864 [16:25:57<24:06:28, 14.56s/it]
 40%|███▉      | 3903/9864 [16:26:10<23:43:09, 14.32s/it]
 40%|███▉      | 3904/9864 [16:26:27<24:41:18, 14.91s/it]
 40%|███▉      | 3905/9864 [16:26:43<25:23:50, 15.34s/it]
 40%|███▉      | 3906/9864 [16:26:58<25:25:34, 15.36s/it]
 40%|███▉      | 3907/9864 [16:27:14<25:36:59, 15.48s/it]
 40%|███▉      | 3908/9864 [16:27:28<24:56:42, 15.08s/it]
 40%|███▉      | 3909/9864 [16:27:42<24:25:53, 14.77s/it]
 40%|███▉      | 3910/9864 [16:27:56<24:04:23, 14.56s/it]
                                                         
{'loss': '0.1469', 'grad_norm': '2.265', 'learning_rate': '1.375e-05', 'epoch': '0.3964'}

 40%|███▉      | 3910/9864 [16:27:56<24:04:23, 14.56s/it]
 40%|███▉      | 3911/9864 [16:28:11<23:53:05, 14.44s/it]
 40%|███▉      | 3912/9864 [16:28:28<25:34:05, 15.46s/it]
 40%|███▉      | 3913/9864 [16:28:43<25:10:26, 15.23s/it]
 40%|███▉      | 3914/9864 [16:29:00<26:13:24, 15.87s/it]
 40%|███▉      | 3915/9864 [16:29:14<25:13:39, 15.27s/it]
 40%|███▉      | 3916/9864 [16:29:28<24:34:36, 14.88s/it]
 40%|███▉      | 3917/9864 [16:29:43<24:24:11, 14.77s/it]
 40%|███▉      | 3918/9864 [16:29:57<24:13:54, 14.67s/it]
 40%|███▉      | 3919/9864 [16:30:13<24:44:16, 14.98s/it]
 40%|███▉      | 3920/9864 [16:30:29<25:18:12, 15.33s/it]
                                                         
{'loss': '0.1559', 'grad_norm': '3.995', 'learning_rate': '1.372e-05', 'epoch': '0.3974'}

 40%|███▉      | 3920/9864 [16:30:29<25:18:12, 15.33s/it]
 40%|███▉      | 3921/9864 [16:30:43<24:45:49, 15.00s/it]
 40%|███▉      | 3922/9864 [16:30:58<24:31:31, 14.86s/it]
 40%|███▉      | 3923/9864 [16:31:16<25:55:17, 15.71s/it]
 40%|███▉      | 3924/9864 [16:31:30<25:13:37, 15.29s/it]
 40%|███▉      | 3925/9864 [16:31:44<24:28:01, 14.83s/it]
 40%|███▉      | 3926/9864 [16:31:58<24:11:56, 14.67s/it]
 40%|███▉      | 3927/9864 [16:32:16<25:46:38, 15.63s/it]
 40%|███▉      | 3928/9864 [16:32:30<25:07:19, 15.24s/it]
 40%|███▉      | 3929/9864 [16:32:45<24:47:51, 15.04s/it]
 40%|███▉      | 3930/9864 [16:33:01<25:21:50, 15.39s/it]
                                                         
{'loss': '0.1542', 'grad_norm': '1.974', 'learning_rate': '1.369e-05', 'epoch': '0.3985'}

 40%|███▉      | 3930/9864 [16:33:01<25:21:50, 15.39s/it]
 40%|███▉      | 3931/9864 [16:33:15<24:41:26, 14.98s/it]
 40%|███▉      | 3932/9864 [16:33:30<24:34:29, 14.91s/it]
 40%|███▉      | 3933/9864 [16:33:44<24:09:05, 14.66s/it]
 40%|███▉      | 3934/9864 [16:34:00<24:51:19, 15.09s/it]
 40%|███▉      | 3935/9864 [16:34:16<25:12:36, 15.31s/it]
 40%|███▉      | 3936/9864 [16:34:30<24:35:41, 14.94s/it]
 40%|███▉      | 3937/9864 [16:34:43<24:00:18, 14.58s/it]
 40%|███▉      | 3938/9864 [16:34:59<24:33:48, 14.92s/it]
 40%|███▉      | 3939/9864 [16:35:16<25:16:22, 15.36s/it]
 40%|███▉      | 3940/9864 [16:35:30<24:35:38, 14.95s/it]
                                                         
{'loss': '0.1505', 'grad_norm': '3.718', 'learning_rate': '1.366e-05', 'epoch': '0.3995'}

 40%|███▉      | 3940/9864 [16:35:30<24:35:38, 14.95s/it]
 40%|███▉      | 3941/9864 [16:35:46<25:12:25, 15.32s/it]
 40%|███▉      | 3942/9864 [16:36:00<24:45:09, 15.05s/it]
 40%|███▉      | 3943/9864 [16:36:16<25:08:21, 15.28s/it]
 40%|███▉      | 3944/9864 [16:36:31<24:48:15, 15.08s/it]
 40%|███▉      | 3945/9864 [16:36:45<24:26:30, 14.87s/it]
 40%|████      | 3946/9864 [16:37:01<24:59:21, 15.20s/it]
 40%|████      | 3947/9864 [16:37:17<25:20:30, 15.42s/it]
 40%|████      | 3948/9864 [16:37:32<25:23:47, 15.45s/it]
 40%|████      | 3949/9864 [16:37:47<24:48:34, 15.10s/it]
 40%|████      | 3950/9864 [16:38:03<25:16:32, 15.39s/it]
                                                         
{'loss': '0.1485', 'grad_norm': '4.076', 'learning_rate': '1.363e-05', 'epoch': '0.4005'}

 40%|████      | 3950/9864 [16:38:03<25:16:32, 15.39s/it]
 40%|████      | 3951/9864 [16:38:17<24:42:28, 15.04s/it]
 40%|████      | 3952/9864 [16:38:31<24:25:16, 14.87s/it]
 40%|████      | 3953/9864 [16:38:46<24:05:26, 14.67s/it]
 40%|████      | 3954/9864 [16:39:02<24:41:41, 15.04s/it]
 40%|████      | 3955/9864 [16:39:15<24:09:22, 14.72s/it]
 40%|████      | 3956/9864 [16:39:31<24:44:31, 15.08s/it]
 40%|████      | 3957/9864 [16:39:46<24:22:32, 14.86s/it]
 40%|████      | 3958/9864 [16:40:01<24:48:22, 15.12s/it]
 40%|████      | 3959/9864 [16:40:16<24:34:43, 14.98s/it]
 40%|████      | 3960/9864 [16:40:31<24:18:00, 14.82s/it]
                                                         
{'loss': '0.1406', 'grad_norm': '3.133', 'learning_rate': '1.36e-05', 'epoch': '0.4015'}

 40%|████      | 3960/9864 [16:40:31<24:18:00, 14.82s/it]
 40%|████      | 3961/9864 [16:40:44<23:49:40, 14.53s/it]
 40%|████      | 3962/9864 [16:41:03<25:34:53, 15.60s/it]
 40%|████      | 3963/9864 [16:41:17<24:57:46, 15.23s/it]
 40%|████      | 3964/9864 [16:41:31<24:34:23, 14.99s/it]
 40%|████      | 3965/9864 [16:41:46<24:32:29, 14.98s/it]
 40%|████      | 3966/9864 [16:42:03<25:09:58, 15.36s/it]
 40%|████      | 3967/9864 [16:42:17<24:28:58, 14.95s/it]
 40%|████      | 3968/9864 [16:42:31<24:13:23, 14.79s/it]
 40%|████      | 3969/9864 [16:42:47<24:38:56, 15.05s/it]
 40%|████      | 3970/9864 [16:43:02<25:00:49, 15.28s/it]
                                                         
{'loss': '0.1483', 'grad_norm': '2.179', 'learning_rate': '1.357e-05', 'epoch': '0.4025'}

 40%|████      | 3970/9864 [16:43:02<25:00:49, 15.28s/it]
 40%|████      | 3971/9864 [16:43:17<24:36:33, 15.03s/it]
 40%|████      | 3972/9864 [16:43:31<24:14:49, 14.81s/it]
 40%|████      | 3973/9864 [16:43:48<25:04:38, 15.32s/it]
 40%|████      | 3974/9864 [16:44:03<25:14:48, 15.43s/it]
 40%|████      | 3975/9864 [16:44:18<24:41:10, 15.09s/it]
 40%|████      | 3976/9864 [16:44:34<25:11:09, 15.40s/it]
 40%|████      | 3977/9864 [16:44:48<24:27:29, 14.96s/it]
 40%|████      | 3978/9864 [16:45:03<24:41:26, 15.10s/it]
 40%|████      | 3979/9864 [16:45:17<24:07:42, 14.76s/it]
 40%|████      | 3980/9864 [16:45:31<23:47:15, 14.55s/it]
                                                         
{'loss': '0.1475', 'grad_norm': '2.844', 'learning_rate': '1.354e-05', 'epoch': '0.4035'}

 40%|████      | 3980/9864 [16:45:31<23:47:15, 14.55s/it]
 40%|████      | 3981/9864 [16:45:49<25:15:56, 15.46s/it]
 40%|████      | 3982/9864 [16:46:03<24:51:39, 15.22s/it]
 40%|████      | 3983/9864 [16:46:18<24:20:20, 14.90s/it]
 40%|████      | 3984/9864 [16:46:33<24:42:16, 15.13s/it]
 40%|████      | 3985/9864 [16:46:49<25:10:42, 15.42s/it]
 40%|████      | 3986/9864 [16:47:04<24:45:01, 15.16s/it]
 40%|████      | 3987/9864 [16:47:18<24:19:00, 14.90s/it]
 40%|████      | 3988/9864 [16:47:36<25:38:24, 15.71s/it]
 40%|████      | 3989/9864 [16:47:50<25:02:31, 15.34s/it]
 40%|████      | 3990/9864 [16:48:04<24:23:41, 14.95s/it]
                                                         
{'loss': '0.1415', 'grad_norm': '2.165', 'learning_rate': '1.351e-05', 'epoch': '0.4045'}

 40%|████      | 3990/9864 [16:48:04<24:23:41, 14.95s/it]
 40%|████      | 3991/9864 [16:48:19<24:04:46, 14.76s/it]
 40%|████      | 3992/9864 [16:48:35<24:45:49, 15.18s/it]
 40%|████      | 3993/9864 [16:48:51<25:20:41, 15.54s/it]
 40%|████      | 3994/9864 [16:49:05<24:36:21, 15.09s/it]
 41%|████      | 3995/9864 [16:49:19<23:54:08, 14.66s/it]
 41%|████      | 3996/9864 [16:49:35<24:23:06, 14.96s/it]
 41%|████      | 3997/9864 [16:49:50<24:51:08, 15.25s/it]
 41%|████      | 3998/9864 [16:50:05<24:28:55, 15.02s/it]
 41%|████      | 3999/9864 [16:50:19<24:13:50, 14.87s/it]
 41%|████      | 4000/9864 [16:50:37<25:27:13, 15.63s/it]
                                                         
{'loss': '0.1725', 'grad_norm': '3.709', 'learning_rate': '1.348e-05', 'epoch': '0.4055'}

 41%|████      | 4000/9864 [16:50:37<25:27:13, 15.63s/it]

Writing model shards:   0%|          | 0/1 [00:00<?, ?it/s]

Writing model shards: 100%|██████████| 1/1 [00:08<00:00,  8.84s/it]
Writing model shards: 100%|██████████| 1/1 [00:08<00:00,  8.84s/it]
/home/dungpq6/Project/SpatialStack/src/qwen_vl/model/geometry_encoders/vggt_encoder.py:285: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(dtype=dtype):

 41%|████      | 4001/9864 [16:51:58<57:38:39, 35.39s/it]
 41%|████      | 4002/9864 [16:52:13<47:28:10, 29.15s/it]
 41%|████      | 4003/9864 [16:52:29<40:52:50, 25.11s/it]
 41%|████      | 4004/9864 [16:52:43<35:23:16, 21.74s/it]
 41%|████      | 4005/9864 [16:52:59<32:39:07, 20.06s/it]
 41%|████      | 4006/9864 [16:53:13<30:00:47, 18.44s/it]
 41%|████      | 4007/9864 [16:53:27<27:47:44, 17.08s/it]
 41%|████      | 4008/9864 [16:53:43<27:21:16, 16.82s/it]
 41%|████      | 4009/9864 [16:53:59<26:57:43, 16.58s/it]
 41%|████      | 4010/9864 [16:54:13<25:29:17, 15.67s/it]
                                                         
{'loss': '0.155', 'grad_norm': '1.602', 'learning_rate': '1.344e-05', 'epoch': '0.4066'}

 41%|████      | 4010/9864 [16:54:13<25:29:17, 15.67s/it]
 41%|████      | 4011/9864 [16:54:29<25:47:43, 15.87s/it]
 41%|████      | 4012/9864 [16:54:45<25:47:43, 15.87s/it]
 41%|████      | 4013/9864 [16:54:59<24:38:14, 15.16s/it]
 41%|████      | 4014/9864 [16:55:13<24:11:22, 14.89s/it]
 41%|████      | 4015/9864 [16:55:27<23:52:11, 14.69s/it]
 41%|████      | 4016/9864 [16:55:43<24:25:39, 15.04s/it]
 41%|████      | 4017/9864 [16:55:59<24:59:08, 15.38s/it]
 41%|████      | 4018/9864 [16:56:14<24:36:23, 15.15s/it]
 41%|████      | 4019/9864 [16:56:32<25:52:07, 15.93s/it]
 41%|████      | 4020/9864 [16:56:46<25:01:08, 15.41s/it]
                                                         
{'loss': '0.1341', 'grad_norm': '1.862', 'learning_rate': '1.341e-05', 'epoch': '0.4076'}

 41%|████      | 4020/9864 [16:56:46<25:01:08, 15.41s/it]
 41%|████      | 4021/9864 [16:57:00<24:18:59, 14.98s/it]
 41%|████      | 4022/9864 [16:57:14<24:01:14, 14.80s/it]
 41%|████      | 4023/9864 [16:57:28<23:38:19, 14.57s/it]
 41%|████      | 4024/9864 [16:57:46<25:03:30, 15.45s/it]
 41%|████      | 4025/9864 [16:58:00<24:43:32, 15.24s/it]
 41%|████      | 4026/9864 [16:58:17<25:13:53, 15.56s/it]
 41%|████      | 4027/9864 [16:58:32<25:18:48, 15.61s/it]
 41%|████      | 4028/9864 [16:58:46<24:29:15, 15.11s/it]
 41%|████      | 4029/9864 [16:59:00<23:59:49, 14.81s/it]
 41%|████      | 4030/9864 [16:59:15<23:42:18, 14.63s/it]
                                                         
{'loss': '0.1546', 'grad_norm': '3.319', 'learning_rate': '1.338e-05', 'epoch': '0.4086'}

 41%|████      | 4030/9864 [16:59:15<23:42:18, 14.63s/it]
 41%|████      | 4031/9864 [16:59:29<23:29:05, 14.49s/it]
 41%|████      | 4032/9864 [16:59:47<25:01:44, 15.45s/it]
 41%|████      | 4033/9864 [17:00:01<24:30:39, 15.13s/it]
 41%|████      | 4034/9864 [17:00:16<24:41:11, 15.24s/it]
 41%|████      | 4035/9864 [17:00:32<25:01:10, 15.45s/it]
 41%|████      | 4036/9864 [17:00:47<24:28:05, 15.11s/it]
 41%|████      | 4037/9864 [17:01:01<24:11:25, 14.95s/it]
 41%|████      | 4038/9864 [17:01:15<23:44:42, 14.67s/it]
 41%|████      | 4039/9864 [17:01:30<23:34:56, 14.57s/it]
 41%|████      | 4040/9864 [17:01:46<24:25:12, 15.09s/it]
                                                         
{'loss': '0.1341', 'grad_norm': '2.376', 'learning_rate': '1.335e-05', 'epoch': '0.4096'}

 41%|████      | 4040/9864 [17:01:46<24:25:12, 15.09s/it]
 41%|████      | 4041/9864 [17:02:04<25:44:40, 15.92s/it]
 41%|████      | 4042/9864 [17:02:20<25:52:06, 16.00s/it]
 41%|████      | 4043/9864 [17:02:34<24:45:15, 15.31s/it]
 41%|████      | 4044/9864 [17:02:48<24:01:45, 14.86s/it]
 41%|████      | 4045/9864 [17:03:01<23:34:32, 14.59s/it]
 41%|████      | 4046/9864 [17:03:16<23:19:14, 14.43s/it]
 41%|████      | 4047/9864 [17:03:32<24:13:18, 14.99s/it]
 41%|████      | 4048/9864 [17:03:46<24:03:15, 14.89s/it]
 41%|████      | 4049/9864 [17:04:05<25:46:24, 15.96s/it]
 41%|████      | 4050/9864 [17:04:21<25:52:37, 16.02s/it]
                                                         
{'loss': '0.1555', 'grad_norm': '4.954', 'learning_rate': '1.332e-05', 'epoch': '0.4106'}

 41%|████      | 4050/9864 [17:04:21<25:52:37, 16.02s/it]
 41%|████      | 4051/9864 [17:04:36<25:08:20, 15.57s/it]
 41%|████      | 4052/9864 [17:04:50<24:30:35, 15.18s/it]
 41%|████      | 4053/9864 [17:05:04<24:08:23, 14.95s/it]
 41%|████      | 4054/9864 [17:05:18<23:22:35, 14.48s/it]
 41%|████      | 4055/9864 [17:05:34<24:14:16, 15.02s/it]
 41%|████      | 4056/9864 [17:05:48<23:47:27, 14.75s/it]
 41%|████      | 4057/9864 [17:06:06<25:14:21, 15.65s/it]
 41%|████      | 4058/9864 [17:06:22<25:22:31, 15.73s/it]
 41%|████      | 4059/9864 [17:06:36<24:32:11, 15.22s/it]
 41%|████      | 4060/9864 [17:06:50<24:05:47, 14.95s/it]
                                                         
{'loss': '0.165', 'grad_norm': '2.115', 'learning_rate': '1.329e-05', 'epoch': '0.4116'}

 41%|████      | 4060/9864 [17:06:50<24:05:47, 14.95s/it]
 41%|████      | 4061/9864 [17:07:04<23:29:09, 14.57s/it]
 41%|████      | 4062/9864 [17:07:18<23:06:51, 14.34s/it]
 41%|████      | 4063/9864 [17:07:33<23:47:14, 14.76s/it]
 41%|████      | 4064/9864 [17:07:47<23:28:59, 14.58s/it]
 41%|████      | 4065/9864 [17:08:05<24:53:10, 15.45s/it]
 41%|████      | 4066/9864 [17:08:21<25:01:19, 15.54s/it]
 41%|████      | 4067/9864 [17:08:35<24:30:44, 15.22s/it]
 41%|████      | 4068/9864 [17:08:49<24:00:15, 14.91s/it]
 41%|████▏     | 4069/9864 [17:09:04<23:43:45, 14.74s/it]
 41%|████▏     | 4070/9864 [17:09:18<23:16:22, 14.46s/it]
                                                         
{'loss': '0.1465', 'grad_norm': '3.756', 'learning_rate': '1.326e-05', 'epoch': '0.4126'}

 41%|████▏     | 4070/9864 [17:09:18<23:16:22, 14.46s/it]
 41%|████▏     | 4071/9864 [17:09:34<24:08:19, 15.00s/it]
 41%|████▏     | 4072/9864 [17:09:48<23:42:21, 14.73s/it]
 41%|████▏     | 4073/9864 [17:10:06<25:16:08, 15.71s/it]
 41%|████▏     | 4074/9864 [17:10:22<25:26:28, 15.82s/it]
 41%|████▏     | 4075/9864 [17:10:37<24:49:49, 15.44s/it]
 41%|████▏     | 4076/9864 [17:10:51<24:14:36, 15.08s/it]
 41%|████▏     | 4077/9864 [17:11:05<23:42:45, 14.75s/it]
 41%|████▏     | 4078/9864 [17:11:21<24:23:49, 15.18s/it]
 41%|████▏     | 4079/9864 [17:11:35<23:56:13, 14.90s/it]
 41%|████▏     | 4080/9864 [17:11:51<24:15:28, 15.10s/it]
                                                         
{'loss': '0.1561', 'grad_norm': '1.418', 'learning_rate': '1.323e-05', 'epoch': '0.4137'}

 41%|████▏     | 4080/9864 [17:11:51<24:15:28, 15.10s/it]
 41%|████▏     | 4081/9864 [17:12:09<25:38:54, 15.97s/it]
 41%|████▏     | 4082/9864 [17:12:23<24:36:33, 15.32s/it]
 41%|████▏     | 4083/9864 [17:12:37<23:58:03, 14.93s/it]
 41%|████▏     | 4084/9864 [17:12:51<23:33:25, 14.67s/it]
 41%|████▏     | 4085/9864 [17:13:07<24:13:51, 15.09s/it]
 41%|████▏     | 4086/9864 [17:13:21<23:42:27, 14.77s/it]
 41%|████▏     | 4087/9864 [17:13:35<23:29:56, 14.64s/it]
 41%|████▏     | 4088/9864 [17:13:51<24:13:03, 15.09s/it]
 41%|████▏     | 4089/9864 [17:14:09<25:31:58, 15.92s/it]
 41%|████▏     | 4090/9864 [17:14:24<24:51:13, 15.50s/it]
                                                         
{'loss': '0.1488', 'grad_norm': '2.224', 'learning_rate': '1.32e-05', 'epoch': '0.4147'}

 41%|████▏     | 4090/9864 [17:14:24<24:51:13, 15.50s/it]
 41%|████▏     | 4091/9864 [17:14:38<24:23:21, 15.21s/it]
 41%|████▏     | 4092/9864 [17:14:54<24:39:36, 15.38s/it]
 41%|████▏     | 4093/9864 [17:15:08<23:51:56, 14.89s/it]
 42%|████▏     | 4094/9864 [17:15:22<23:37:53, 14.74s/it]
 42%|████▏     | 4095/9864 [17:15:36<23:15:56, 14.52s/it]
 42%|████▏     | 4096/9864 [17:15:50<22:59:09, 14.35s/it]
 42%|████▏     | 4097/9864 [17:16:07<24:27:27, 15.27s/it]
 42%|████▏     | 4098/9864 [17:16:23<24:37:12, 15.37s/it]
 42%|████▏     | 4099/9864 [17:16:39<24:53:03, 15.54s/it]
 42%|████▏     | 4100/9864 [17:16:53<24:15:23, 15.15s/it]
                                                         
{'loss': '0.1324', 'grad_norm': '2.731', 'learning_rate': '1.317e-05', 'epoch': '0.4157'}

 42%|████▏     | 4100/9864 [17:16:53<24:15:23, 15.15s/it]
 42%|████▏     | 4101/9864 [17:17:07<23:49:06, 14.88s/it]
 42%|████▏     | 4102/9864 [17:17:22<23:29:28, 14.68s/it]
 42%|████▏     | 4103/9864 [17:17:36<23:27:47, 14.66s/it]
 42%|████▏     | 4104/9864 [17:17:51<23:17:45, 14.56s/it]
 42%|████▏     | 4105/9864 [17:18:08<24:41:54, 15.44s/it]
 42%|████▏     | 4106/9864 [17:18:24<24:43:01, 15.45s/it]
 42%|████▏     | 4107/9864 [17:18:39<24:49:19, 15.52s/it]
 42%|████▏     | 4108/9864 [17:18:53<24:08:16, 15.10s/it]
 42%|████▏     | 4109/9864 [17:19:07<23:26:10, 14.66s/it]
 42%|████▏     | 4110/9864 [17:19:21<23:11:29, 14.51s/it]
                                                         
{'loss': '0.1505', 'grad_norm': '1.659', 'learning_rate': '1.313e-05', 'epoch': '0.4167'}

 42%|████▏     | 4110/9864 [17:19:21<23:11:29, 14.51s/it]
 42%|████▏     | 4111/9864 [17:19:35<22:57:55, 14.37s/it]
 42%|████▏     | 4112/9864 [17:19:51<23:35:10, 14.76s/it]
 42%|████▏     | 4113/9864 [17:20:07<24:06:27, 15.09s/it]
 42%|████▏     | 4114/9864 [17:20:21<23:46:04, 14.88s/it]
 42%|████▏     | 4115/9864 [17:20:37<24:23:03, 15.27s/it]
 42%|████▏     | 4116/9864 [17:20:53<24:35:22, 15.40s/it]
 42%|████▏     | 4117/9864 [17:21:07<23:50:27, 14.93s/it]
 42%|████▏     | 4118/9864 [17:21:21<23:16:08, 14.58s/it]
 42%|████▏     | 4119/9864 [17:21:35<23:08:12, 14.50s/it]
 42%|████▏     | 4120/9864 [17:21:51<23:50:06, 14.94s/it]
                                                         
{'loss': '0.1399', 'grad_norm': '3.929', 'learning_rate': '1.31e-05', 'epoch': '0.4177'}

 42%|████▏     | 4120/9864 [17:21:51<23:50:06, 14.94s/it]
 42%|████▏     | 4121/9864 [17:22:05<23:28:21, 14.71s/it]
 42%|████▏     | 4122/9864 [17:22:22<24:41:40, 15.48s/it]
 42%|████▏     | 4123/9864 [17:22:36<23:58:27, 15.03s/it]
 42%|████▏     | 4124/9864 [17:22:52<24:23:36, 15.30s/it]
 42%|████▏     | 4125/9864 [17:23:06<23:48:38, 14.94s/it]
 42%|████▏     | 4126/9864 [17:23:20<23:22:12, 14.66s/it]
 42%|████▏     | 4127/9864 [17:23:35<23:11:48, 14.56s/it]
 42%|████▏     | 4128/9864 [17:23:50<23:40:32, 14.86s/it]
 42%|████▏     | 4129/9864 [17:24:04<23:19:47, 14.64s/it]
 42%|████▏     | 4130/9864 [17:24:22<24:58:16, 15.68s/it]
                                                         
{'loss': '0.1358', 'grad_norm': '2.296', 'learning_rate': '1.307e-05', 'epoch': '0.4187'}

 42%|████▏     | 4130/9864 [17:24:22<24:58:16, 15.68s/it]
 42%|████▏     | 4131/9864 [17:24:36<24:09:53, 15.17s/it]
 42%|████▏     | 4132/9864 [17:24:51<23:48:44, 14.96s/it]
 42%|████▏     | 4133/9864 [17:25:07<24:23:23, 15.32s/it]
 42%|████▏     | 4134/9864 [17:25:21<23:42:38, 14.90s/it]
 42%|████▏     | 4135/9864 [17:25:35<23:22:04, 14.68s/it]
 42%|████▏     | 4136/9864 [17:25:51<23:56:39, 15.05s/it]
 42%|████▏     | 4137/9864 [17:26:08<24:58:23, 15.70s/it]
 42%|████▏     | 4138/9864 [17:26:22<24:14:29, 15.24s/it]
 42%|████▏     | 4139/9864 [17:26:37<23:52:30, 15.01s/it]
 42%|████▏     | 4140/9864 [17:26:53<24:20:39, 15.31s/it]
                                                         
{'loss': '0.1291', 'grad_norm': '2.793', 'learning_rate': '1.304e-05', 'epoch': '0.4197'}

 42%|████▏     | 4140/9864 [17:26:53<24:20:39, 15.31s/it]
 42%|████▏     | 4141/9864 [17:27:07<23:53:24, 15.03s/it]
 42%|████▏     | 4142/9864 [17:27:21<23:22:00, 14.70s/it]
 42%|████▏     | 4143/9864 [17:27:37<24:05:02, 15.16s/it]
 42%|████▏     | 4144/9864 [17:27:53<24:14:40, 15.26s/it]
 42%|████▏     | 4145/9864 [17:28:07<23:30:45, 14.80s/it]
 42%|████▏     | 4146/9864 [17:28:23<24:01:40, 15.13s/it]
 42%|████▏     | 4147/9864 [17:28:38<24:22:43, 15.35s/it]
 42%|████▏     | 4148/9864 [17:28:53<23:52:44, 15.04s/it]
 42%|████▏     | 4149/9864 [17:29:07<23:28:59, 14.79s/it]
 42%|████▏     | 4150/9864 [17:29:21<23:16:58, 14.67s/it]
                                                         
{'loss': '0.144', 'grad_norm': '1.945', 'learning_rate': '1.301e-05', 'epoch': '0.4208'}

 42%|████▏     | 4150/9864 [17:29:21<23:16:58, 14.67s/it]
 42%|████▏     | 4151/9864 [17:29:39<24:41:50, 15.56s/it]
 42%|████▏     | 4152/9864 [17:29:53<24:08:41, 15.22s/it]
 42%|████▏     | 4153/9864 [17:30:07<23:31:27, 14.83s/it]
 42%|████▏     | 4154/9864 [17:30:23<24:07:00, 15.21s/it]
 42%|████▏     | 4155/9864 [17:30:38<23:50:52, 15.04s/it]
 42%|████▏     | 4156/9864 [17:30:54<24:23:56, 15.39s/it]
 42%|████▏     | 4157/9864 [17:31:09<24:01:12, 15.15s/it]
 42%|████▏     | 4158/9864 [17:31:25<24:30:51, 15.47s/it]
 42%|████▏     | 4159/9864 [17:31:41<24:30:01, 15.46s/it]
 42%|████▏     | 4160/9864 [17:31:55<23:50:11, 15.04s/it]
                                                         
{'loss': '0.1472', 'grad_norm': '1.772', 'learning_rate': '1.298e-05', 'epoch': '0.4218'}

 42%|████▏     | 4160/9864 [17:31:55<23:50:11, 15.04s/it]
 42%|████▏     | 4161/9864 [17:32:10<24:06:43, 15.22s/it]
 42%|████▏     | 4162/9864 [17:32:25<23:43:40, 14.98s/it]
 42%|████▏     | 4163/9864 [17:32:40<24:04:58, 15.21s/it]
 42%|████▏     | 4164/9864 [17:32:55<23:40:29, 14.95s/it]
 42%|████▏     | 4165/9864 [17:33:11<24:18:29, 15.36s/it]
 42%|████▏     | 4166/9864 [17:33:28<24:51:33, 15.71s/it]
 42%|████▏     | 4167/9864 [17:33:42<24:09:17, 15.26s/it]
 42%|████▏     | 4168/9864 [17:33:58<24:33:43, 15.52s/it]
 42%|████▏     | 4169/9864 [17:34:12<23:52:25, 15.09s/it]
 42%|████▏     | 4170/9864 [17:34:26<23:12:23, 14.67s/it]
                                                         
{'loss': '0.1375', 'grad_norm': '3.175', 'learning_rate': '1.295e-05', 'epoch': '0.4228'}

 42%|████▏     | 4170/9864 [17:34:26<23:12:23, 14.67s/it]
 42%|████▏     | 4171/9864 [17:34:41<23:30:27, 14.87s/it]
 42%|████▏     | 4172/9864 [17:34:55<23:12:06, 14.67s/it]
 42%|████▏     | 4173/9864 [17:35:11<23:55:18, 15.13s/it]
 42%|████▏     | 4174/9864 [17:35:27<24:20:34, 15.40s/it]
 42%|████▏     | 4175/9864 [17:35:42<23:43:25, 15.01s/it]
 42%|████▏     | 4176/9864 [17:35:58<24:14:10, 15.34s/it]
 42%|████▏     | 4177/9864 [17:36:12<23:49:33, 15.08s/it]
 42%|████▏     | 4178/9864 [17:36:27<23:31:26, 14.89s/it]
 42%|████▏     | 4179/9864 [17:36:42<23:46:15, 15.05s/it]
 42%|████▏     | 4180/9864 [17:36:59<24:26:46, 15.48s/it]
                                                         
{'loss': '0.1442', 'grad_norm': '3.69', 'learning_rate': '1.292e-05', 'epoch': '0.4238'}

 42%|████▏     | 4180/9864 [17:36:59<24:26:46, 15.48s/it]
 42%|████▏     | 4181/9864 [17:37:14<24:37:47, 15.60s/it]
 42%|████▏     | 4182/9864 [17:37:28<23:46:26, 15.06s/it]
 42%|████▏     | 4183/9864 [17:37:43<23:28:09, 14.87s/it]
 42%|████▏     | 4184/9864 [17:37:58<23:49:07, 15.10s/it]
 42%|████▏     | 4185/9864 [17:38:13<23:30:06, 14.90s/it]
 42%|████▏     | 4186/9864 [17:38:27<23:05:52, 14.64s/it]
 42%|████▏     | 4187/9864 [17:38:43<23:46:09, 15.07s/it]
 42%|████▏     | 4188/9864 [17:39:01<25:08:10, 15.94s/it]
 42%|████▏     | 4189/9864 [17:39:15<24:22:23, 15.46s/it]
 42%|████▏     | 4190/9864 [17:39:29<23:45:29, 15.07s/it]
                                                         
{'loss': '0.1435', 'grad_norm': '1.521', 'learning_rate': '1.288e-05', 'epoch': '0.4248'}

 42%|████▏     | 4190/9864 [17:39:29<23:45:29, 15.07s/it]
 42%|████▏     | 4191/9864 [17:39:43<23:20:03, 14.81s/it]
 42%|████▏     | 4192/9864 [17:39:59<23:41:07, 15.03s/it]
 43%|████▎     | 4193/9864 [17:40:13<23:18:48, 14.80s/it]
 43%|████▎     | 4194/9864 [17:40:29<23:48:35, 15.12s/it]
 43%|████▎     | 4195/9864 [17:40:45<24:11:00, 15.36s/it]
 43%|████▎     | 4196/9864 [17:40:59<23:38:15, 15.01s/it]
 43%|████▎     | 4197/9864 [17:41:15<23:55:31, 15.20s/it]
 43%|████▎     | 4198/9864 [17:41:29<23:22:35, 14.85s/it]
 43%|████▎     | 4199/9864 [17:41:45<23:47:41, 15.12s/it]
 43%|████▎     | 4200/9864 [17:41:59<23:17:50, 14.81s/it]
                                                         
{'loss': '0.1451', 'grad_norm': '2.571', 'learning_rate': '1.285e-05', 'epoch': '0.4258'}

 43%|████▎     | 4200/9864 [17:41:59<23:17:50, 14.81s/it]
 43%|████▎     | 4201/9864 [17:42:13<23:04:05, 14.66s/it]
 43%|████▎     | 4202/9864 [17:42:31<24:48:57, 15.78s/it]
 43%|████▎     | 4203/9864 [17:42:46<24:02:00, 15.28s/it]
 43%|████▎     | 4204/9864 [17:43:00<23:31:26, 14.96s/it]
 43%|████▎     | 4205/9864 [17:43:16<24:00:09, 15.27s/it]
 43%|████▎     | 4206/9864 [17:43:32<24:19:03, 15.47s/it]
 43%|████▎     | 4207/9864 [17:43:46<23:32:43, 14.98s/it]
 43%|████▎     | 4208/9864 [17:44:01<23:35:38, 15.02s/it]
 43%|████▎     | 4209/9864 [17:44:16<23:52:30, 15.20s/it]
 43%|████▎     | 4210/9864 [17:44:32<24:13:01, 15.42s/it]
                                                         
{'loss': '0.152', 'grad_norm': '2.347', 'learning_rate': '1.282e-05', 'epoch': '0.4268'}

 43%|████▎     | 4210/9864 [17:44:32<24:13:01, 15.42s/it]
 43%|████▎     | 4211/9864 [17:44:47<23:52:51, 15.21s/it]
 43%|████▎     | 4212/9864 [17:45:01<23:16:46, 14.83s/it]
 43%|████▎     | 4213/9864 [17:45:16<23:23:45, 14.90s/it]
 43%|████▎     | 4214/9864 [17:45:32<24:08:02, 15.38s/it]
 43%|████▎     | 4215/9864 [17:45:49<24:29:53, 15.61s/it]
 43%|████▎     | 4216/9864 [17:46:02<23:37:07, 15.05s/it]
 43%|████▎     | 4217/9864 [17:46:16<23:09:26, 14.76s/it]
 43%|████▎     | 4218/9864 [17:46:33<24:00:59, 15.31s/it]
 43%|████▎     | 4219/9864 [17:46:49<24:31:55, 15.64s/it]
 43%|████▎     | 4220/9864 [17:47:03<23:41:45, 15.11s/it]
                                                         
{'loss': '0.152', 'grad_norm': '1.605', 'learning_rate': '1.279e-05', 'epoch': '0.4279'}

 43%|████▎     | 4220/9864 [17:47:03<23:41:45, 15.11s/it]
 43%|████▎     | 4221/9864 [17:47:19<24:07:42, 15.39s/it]
 43%|████▎     | 4222/9864 [17:47:34<23:36:45, 15.07s/it]
 43%|████▎     | 4223/9864 [17:47:49<23:54:38, 15.26s/it]
 43%|████▎     | 4224/9864 [17:48:04<23:29:51, 15.00s/it]
 43%|████▎     | 4225/9864 [17:48:18<22:54:15, 14.62s/it]
 43%|████▎     | 4226/9864 [17:48:34<23:40:15, 15.11s/it]
 43%|████▎     | 4227/9864 [17:48:50<24:01:26, 15.34s/it]
 43%|████▎     | 4228/9864 [17:49:04<23:25:57, 14.97s/it]
 43%|████▎     | 4229/9864 [17:49:18<23:09:28, 14.79s/it]
 43%|████▎     | 4230/9864 [17:49:34<23:30:19, 15.02s/it]
                                                         
{'loss': '0.1436', 'grad_norm': '5.959', 'learning_rate': '1.276e-05', 'epoch': '0.4289'}

 43%|████▎     | 4230/9864 [17:49:34<23:30:19, 15.02s/it]
 43%|████▎     | 4231/9864 [17:49:50<24:04:45, 15.39s/it]
 43%|████▎     | 4232/9864 [17:50:04<23:39:39, 15.12s/it]
 43%|████▎     | 4233/9864 [17:50:19<23:13:36, 14.85s/it]
 43%|████▎     | 4234/9864 [17:50:34<23:29:07, 15.02s/it]
 43%|████▎     | 4235/9864 [17:50:50<23:55:59, 15.31s/it]
 43%|████▎     | 4236/9864 [17:51:04<23:07:58, 14.80s/it]
 43%|████▎     | 4237/9864 [17:51:20<23:48:56, 15.24s/it]
 43%|████▎     | 4238/9864 [17:51:34<23:03:30, 14.75s/it]
 43%|████▎     | 4239/9864 [17:51:49<23:26:59, 15.01s/it]
 43%|████▎     | 4240/9864 [17:52:03<22:57:57, 14.70s/it]
                                                         
{'loss': '0.1476', 'grad_norm': '3.09', 'learning_rate': '1.273e-05', 'epoch': '0.4299'}

 43%|████▎     | 4240/9864 [17:52:03<22:57:57, 14.70s/it]
 43%|████▎     | 4241/9864 [17:52:18<22:49:10, 14.61s/it]
 43%|████▎     | 4242/9864 [17:52:33<23:25:05, 15.00s/it]
 43%|████▎     | 4243/9864 [17:52:48<23:11:03, 14.85s/it]
 43%|████▎     | 4244/9864 [17:53:04<23:45:20, 15.22s/it]
 43%|████▎     | 4245/9864 [17:53:18<23:16:04, 14.91s/it]
 43%|████▎     | 4246/9864 [17:53:36<24:34:53, 15.75s/it]
 43%|████▎     | 4247/9864 [17:53:50<23:35:26, 15.12s/it]
 43%|████▎     | 4248/9864 [17:54:04<23:08:21, 14.83s/it]
 43%|████▎     | 4249/9864 [17:54:18<22:42:24, 14.56s/it]
 43%|████▎     | 4250/9864 [17:54:33<23:08:43, 14.84s/it]
                                                         
{'loss': '0.124', 'grad_norm': '2.794', 'learning_rate': '1.269e-05', 'epoch': '0.4309'}

 43%|████▎     | 4250/9864 [17:54:33<23:08:43, 14.84s/it]
 43%|████▎     | 4251/9864 [17:54:49<23:43:27, 15.22s/it]
 43%|████▎     | 4252/9864 [17:55:04<23:16:34, 14.93s/it]
 43%|████▎     | 4253/9864 [17:55:20<23:47:34, 15.27s/it]
 43%|████▎     | 4254/9864 [17:55:35<23:54:13, 15.34s/it]
 43%|████▎     | 4255/9864 [17:55:49<23:16:31, 14.94s/it]
 43%|████▎     | 4256/9864 [17:56:03<22:43:55, 14.59s/it]
 43%|████▎     | 4257/9864 [17:56:17<22:43:14, 14.59s/it]
 43%|████▎     | 4258/9864 [17:56:33<23:15:25, 14.94s/it]
 43%|████▎     | 4259/9864 [17:56:49<23:44:00, 15.24s/it]
 43%|████▎     | 4260/9864 [17:57:05<23:53:01, 15.34s/it]
                                                         
{'loss': '0.1431', 'grad_norm': '2.49', 'learning_rate': '1.266e-05', 'epoch': '0.4319'}

 43%|████▎     | 4260/9864 [17:57:05<23:53:01, 15.34s/it]
 43%|████▎     | 4261/9864 [17:57:21<24:22:11, 15.66s/it]
 43%|████▎     | 4262/9864 [17:57:35<23:27:07, 15.07s/it]
 43%|████▎     | 4263/9864 [17:57:49<22:56:36, 14.75s/it]
 43%|████▎     | 4264/9864 [17:58:03<22:37:44, 14.55s/it]
 43%|████▎     | 4265/9864 [17:58:17<22:29:24, 14.46s/it]
 43%|████▎     | 4266/9864 [17:58:32<22:42:10, 14.60s/it]
 43%|████▎     | 4267/9864 [17:58:48<23:09:44, 14.90s/it]
 43%|████▎     | 4268/9864 [17:59:04<23:42:17, 15.25s/it]
 43%|████▎     | 4269/9864 [17:59:20<24:03:42, 15.48s/it]
 43%|████▎     | 4270/9864 [17:59:34<23:28:34, 15.11s/it]
                                                         
{'loss': '0.139', 'grad_norm': '1.906', 'learning_rate': '1.263e-05', 'epoch': '0.4329'}

 43%|████▎     | 4270/9864 [17:59:34<23:28:34, 15.11s/it]
 43%|████▎     | 4271/9864 [17:59:49<23:16:28, 14.98s/it]
 43%|████▎     | 4272/9864 [18:00:03<22:52:51, 14.73s/it]
 43%|████▎     | 4273/9864 [18:00:17<22:37:50, 14.57s/it]
 43%|████▎     | 4274/9864 [18:00:35<24:03:09, 15.49s/it]
 43%|████▎     | 4275/9864 [18:00:49<23:24:54, 15.08s/it]
 43%|████▎     | 4276/9864 [18:01:04<23:37:49, 15.22s/it]
 43%|████▎     | 4277/9864 [18:01:20<23:47:11, 15.33s/it]
 43%|████▎     | 4278/9864 [18:01:33<22:44:50, 14.66s/it]
 43%|████▎     | 4279/9864 [18:01:47<22:27:20, 14.47s/it]
 43%|████▎     | 4280/9864 [18:02:02<22:26:39, 14.47s/it]
                                                         
{'loss': '0.1547', 'grad_norm': '3.211', 'learning_rate': '1.26e-05', 'epoch': '0.4339'}

 43%|████▎     | 4280/9864 [18:02:02<22:26:39, 14.47s/it]
 43%|████▎     | 4281/9864 [18:02:17<23:05:41, 14.89s/it]
 43%|████▎     | 4282/9864 [18:02:33<23:26:36, 15.12s/it]
 43%|████▎     | 4283/9864 [18:02:47<22:48:29, 14.71s/it]
 43%|████▎     | 4284/9864 [18:03:02<23:10:35, 14.95s/it]
 43%|████▎     | 4285/9864 [18:03:18<23:27:10, 15.13s/it]
 43%|████▎     | 4286/9864 [18:03:32<22:52:49, 14.77s/it]
 43%|████▎     | 4287/9864 [18:03:46<22:28:24, 14.51s/it]
 43%|████▎     | 4288/9864 [18:04:02<23:24:39, 15.11s/it]
 43%|████▎     | 4289/9864 [18:04:17<23:11:59, 14.98s/it]
 43%|████▎     | 4290/9864 [18:04:33<23:42:37, 15.31s/it]
                                                         
{'loss': '0.1432', 'grad_norm': '2.759', 'learning_rate': '1.257e-05', 'epoch': '0.435'}

 43%|████▎     | 4290/9864 [18:04:33<23:42:37, 15.31s/it]
 44%|████▎     | 4291/9864 [18:04:47<23:06:43, 14.93s/it]
 44%|████▎     | 4292/9864 [18:05:03<23:43:19, 15.33s/it]
 44%|████▎     | 4293/9864 [18:05:17<22:48:58, 14.74s/it]
 44%|████▎     | 4294/9864 [18:05:32<23:10:25, 14.98s/it]
 44%|████▎     | 4295/9864 [18:05:46<22:50:33, 14.77s/it]
 44%|████▎     | 4296/9864 [18:06:01<22:37:31, 14.63s/it]
 44%|████▎     | 4297/9864 [18:06:17<23:19:19, 15.08s/it]
 44%|████▎     | 4298/9864 [18:06:31<22:47:26, 14.74s/it]
 44%|████▎     | 4299/9864 [18:06:48<23:48:27, 15.40s/it]
 44%|████▎     | 4300/9864 [18:07:02<23:26:14, 15.16s/it]
                                                         
{'loss': '0.1428', 'grad_norm': '4.019', 'learning_rate': '1.254e-05', 'epoch': '0.436'}

 44%|████▎     | 4300/9864 [18:07:02<23:26:14, 15.16s/it]
 44%|████▎     | 4301/9864 [18:07:17<23:01:09, 14.90s/it]
 44%|████▎     | 4302/9864 [18:07:33<23:29:36, 15.21s/it]
 44%|████▎     | 4303/9864 [18:07:47<23:07:42, 14.97s/it]
 44%|████▎     | 4304/9864 [18:08:00<22:10:58, 14.36s/it]
 44%|████▎     | 4305/9864 [18:08:16<22:52:22, 14.81s/it]
 44%|████▎     | 4306/9864 [18:08:30<22:43:18, 14.72s/it]
 44%|████▎     | 4307/9864 [18:08:48<24:01:11, 15.56s/it]
 44%|████▎     | 4308/9864 [18:09:02<23:27:51, 15.20s/it]
 44%|████▎     | 4309/9864 [18:09:19<23:58:12, 15.53s/it]
 44%|████▎     | 4310/9864 [18:09:33<23:34:40, 15.28s/it]
                                                         
{'loss': '0.1495', 'grad_norm': '2.293', 'learning_rate': '1.25e-05', 'epoch': '0.437'}

 44%|████▎     | 4310/9864 [18:09:33<23:34:40, 15.28s/it]
 44%|████▎     | 4311/9864 [18:09:47<22:56:13, 14.87s/it]
 44%|████▎     | 4312/9864 [18:10:01<22:28:25, 14.57s/it]
 44%|████▎     | 4313/9864 [18:10:15<22:22:43, 14.51s/it]
 44%|████▎     | 4314/9864 [18:10:33<23:54:44, 15.51s/it]
 44%|████▎     | 4315/9864 [18:10:50<24:19:04, 15.78s/it]
 44%|████▍     | 4316/9864 [18:11:04<23:27:35, 15.22s/it]
 44%|████▍     | 4317/9864 [18:11:19<23:44:30, 15.41s/it]
 44%|████▍     | 4318/9864 [18:11:34<23:14:29, 15.09s/it]
 44%|████▍     | 4319/9864 [18:11:48<22:38:23, 14.70s/it]
 44%|████▍     | 4320/9864 [18:12:01<22:15:16, 14.45s/it]
                                                         
{'loss': '0.1593', 'grad_norm': '2.746', 'learning_rate': '1.247e-05', 'epoch': '0.438'}

 44%|████▍     | 4320/9864 [18:12:01<22:15:16, 14.45s/it]
 44%|████▍     | 4321/9864 [18:12:17<22:54:50, 14.88s/it]
 44%|████▍     | 4322/9864 [18:12:33<23:16:55, 15.12s/it]
 44%|████▍     | 4323/9864 [18:12:49<23:41:09, 15.39s/it]
 44%|████▍     | 4324/9864 [18:13:03<23:02:58, 14.98s/it]
 44%|████▍     | 4325/9864 [18:13:17<22:25:43, 14.58s/it]
 44%|████▍     | 4326/9864 [18:13:32<22:46:43, 14.81s/it]
 44%|████▍     | 4327/9864 [18:13:46<22:35:22, 14.69s/it]
 44%|████▍     | 4328/9864 [18:14:02<22:58:18, 14.94s/it]
 44%|████▍     | 4329/9864 [18:14:16<22:30:49, 14.64s/it]
 44%|████▍     | 4330/9864 [18:14:32<23:00:55, 14.97s/it]
                                                         
{'loss': '0.1456', 'grad_norm': '3.796', 'learning_rate': '1.244e-05', 'epoch': '0.439'}

 44%|████▍     | 4330/9864 [18:14:32<23:00:55, 14.97s/it]
 44%|████▍     | 4331/9864 [18:14:47<23:15:32, 15.13s/it]
 44%|████▍     | 4332/9864 [18:15:01<22:50:18, 14.86s/it]
 44%|████▍     | 4333/9864 [18:15:15<22:17:45, 14.51s/it]
 44%|████▍     | 4334/9864 [18:15:29<22:05:41, 14.38s/it]
 44%|████▍     | 4335/9864 [18:15:45<22:46:55, 14.83s/it]
 44%|████▍     | 4336/9864 [18:15:59<22:26:39, 14.62s/it]
 44%|████▍     | 4337/9864 [18:16:15<22:54:14, 14.92s/it]
 44%|████▍     | 4338/9864 [18:16:30<23:13:37, 15.13s/it]
 44%|████▍     | 4339/9864 [18:16:46<23:39:41, 15.42s/it]
 44%|████▍     | 4340/9864 [18:17:00<23:00:16, 14.99s/it]
                                                         
{'loss': '0.1389', 'grad_norm': '2.552', 'learning_rate': '1.241e-05', 'epoch': '0.44'}

 44%|████▍     | 4340/9864 [18:17:00<23:00:16, 14.99s/it]
 44%|████▍     | 4341/9864 [18:17:15<22:42:01, 14.80s/it]
 44%|████▍     | 4342/9864 [18:17:31<23:19:40, 15.21s/it]
 44%|████▍     | 4343/9864 [18:17:45<22:51:04, 14.90s/it]
 44%|████▍     | 4344/9864 [18:17:59<22:32:21, 14.70s/it]
 44%|████▍     | 4345/9864 [18:18:17<23:53:44, 15.59s/it]
 44%|████▍     | 4346/9864 [18:18:32<23:40:52, 15.45s/it]
 44%|████▍     | 4347/9864 [18:18:47<23:13:27, 15.15s/it]
 44%|████▍     | 4348/9864 [18:19:01<22:43:26, 14.83s/it]
 44%|████▍     | 4349/9864 [18:19:17<23:10:00, 15.12s/it]
 44%|████▍     | 4350/9864 [18:19:31<22:38:50, 14.79s/it]
                                                         
{'loss': '0.1621', 'grad_norm': '4.952', 'learning_rate': '1.238e-05', 'epoch': '0.441'}

 44%|████▍     | 4350/9864 [18:19:31<22:38:50, 14.79s/it]
 44%|████▍     | 4351/9864 [18:19:44<22:09:42, 14.47s/it]
 44%|████▍     | 4352/9864 [18:19:59<22:04:25, 14.42s/it]
 44%|████▍     | 4353/9864 [18:20:16<23:26:35, 15.31s/it]
 44%|████▍     | 4354/9864 [18:20:32<23:37:09, 15.43s/it]
 44%|████▍     | 4355/9864 [18:20:46<22:58:32, 15.01s/it]
 44%|████▍     | 4356/9864 [18:21:00<22:47:14, 14.89s/it]
 44%|████▍     | 4357/9864 [18:21:16<23:13:03, 15.18s/it]
 44%|████▍     | 4358/9864 [18:21:30<22:36:23, 14.78s/it]
 44%|████▍     | 4359/9864 [18:21:43<21:58:27, 14.37s/it]
 44%|████▍     | 4360/9864 [18:21:57<21:41:42, 14.19s/it]
                                                         
{'loss': '0.1447', 'grad_norm': '4.953', 'learning_rate': '1.235e-05', 'epoch': '0.442'}

 44%|████▍     | 4360/9864 [18:21:57<21:41:42, 14.19s/it]
 44%|████▍     | 4361/9864 [18:22:16<23:59:22, 15.69s/it]
 44%|████▍     | 4362/9864 [18:22:30<23:11:21, 15.17s/it]
 44%|████▍     | 4363/9864 [18:22:45<22:48:05, 14.92s/it]
 44%|████▍     | 4364/9864 [18:22:59<22:35:56, 14.79s/it]
 44%|████▍     | 4365/9864 [18:23:15<23:02:05, 15.08s/it]
 44%|████▍     | 4366/9864 [18:23:29<22:24:39, 14.67s/it]
 44%|████▍     | 4367/9864 [18:23:43<22:25:48, 14.69s/it]
 44%|████▍     | 4368/9864 [18:23:59<23:00:38, 15.07s/it]
 44%|████▍     | 4369/9864 [18:24:15<23:24:03, 15.33s/it]
 44%|████▍     | 4370/9864 [18:24:31<23:39:25, 15.50s/it]
                                                         
{'loss': '0.1418', 'grad_norm': '3.079', 'learning_rate': '1.231e-05', 'epoch': '0.4431'}

 44%|████▍     | 4370/9864 [18:24:31<23:39:25, 15.50s/it]
 44%|████▍     | 4371/9864 [18:24:46<23:12:35, 15.21s/it]
 44%|████▍     | 4372/9864 [18:25:00<22:43:45, 14.90s/it]
 44%|████▍     | 4373/9864 [18:25:16<23:23:40, 15.34s/it]
 44%|████▍     | 4374/9864 [18:25:30<22:47:46, 14.95s/it]
 44%|████▍     | 4375/9864 [18:25:45<22:27:33, 14.73s/it]
 44%|████▍     | 4376/9864 [18:26:02<23:55:41, 15.70s/it]
 44%|████▍     | 4377/9864 [18:26:17<23:22:38, 15.34s/it]
 44%|████▍     | 4378/9864 [18:26:33<23:48:14, 15.62s/it]
 44%|████▍     | 4379/9864 [18:26:48<23:14:28, 15.25s/it]
 44%|████▍     | 4380/9864 [18:27:01<22:35:32, 14.83s/it]
                                                         
{'loss': '0.1377', 'grad_norm': '2.386', 'learning_rate': '1.228e-05', 'epoch': '0.4441'}

 44%|████▍     | 4380/9864 [18:27:01<22:35:32, 14.83s/it]
 44%|████▍     | 4381/9864 [18:27:17<23:05:38, 15.16s/it]
 44%|████▍     | 4382/9864 [18:27:31<22:34:49, 14.83s/it]
 44%|████▍     | 4383/9864 [18:27:48<23:15:14, 15.27s/it]
 44%|████▍     | 4384/9864 [18:28:02<22:47:38, 14.97s/it]
 44%|████▍     | 4385/9864 [18:28:17<22:34:53, 14.84s/it]
 44%|████▍     | 4386/9864 [18:28:31<22:19:07, 14.67s/it]
 44%|████▍     | 4387/9864 [18:28:47<22:47:05, 14.98s/it]
 44%|████▍     | 4388/9864 [18:29:01<22:20:23, 14.69s/it]
 44%|████▍     | 4389/9864 [18:29:17<23:04:56, 15.18s/it]
 45%|████▍     | 4390/9864 [18:29:31<22:39:55, 14.91s/it]
                                                         
{'loss': '0.1304', 'grad_norm': '4.552', 'learning_rate': '1.225e-05', 'epoch': '0.4451'}

 45%|████▍     | 4390/9864 [18:29:31<22:39:55, 14.91s/it]
 45%|████▍     | 4391/9864 [18:29:49<24:13:55, 15.94s/it]
 45%|████▍     | 4392/9864 [18:30:04<23:39:57, 15.57s/it]
 45%|████▍     | 4393/9864 [18:30:18<23:00:05, 15.14s/it]
 45%|████▍     | 4394/9864 [18:30:34<23:11:50, 15.27s/it]
 45%|████▍     | 4395/9864 [18:30:48<22:39:07, 14.91s/it]
 45%|████▍     | 4396/9864 [18:31:04<23:10:00, 15.25s/it]
 45%|████▍     | 4397/9864 [18:31:18<22:38:20, 14.91s/it]
 45%|████▍     | 4398/9864 [18:31:33<22:23:30, 14.75s/it]
 45%|████▍     | 4399/9864 [18:31:50<23:48:50, 15.69s/it]
 45%|████▍     | 4400/9864 [18:32:04<23:02:22, 15.18s/it]
                                                         
{'loss': '0.1429', 'grad_norm': '2.437', 'learning_rate': '1.222e-05', 'epoch': '0.4461'}

 45%|████▍     | 4400/9864 [18:32:04<23:02:22, 15.18s/it]
 45%|████▍     | 4401/9864 [18:32:18<22:32:53, 14.86s/it]
 45%|████▍     | 4402/9864 [18:32:34<22:54:59, 15.10s/it]
 45%|████▍     | 4403/9864 [18:32:48<22:18:53, 14.71s/it]
 45%|████▍     | 4404/9864 [18:33:04<22:56:53, 15.13s/it]
 45%|████▍     | 4405/9864 [18:33:19<22:37:53, 14.92s/it]
 45%|████▍     | 4406/9864 [18:33:33<22:23:30, 14.77s/it]
 45%|████▍     | 4407/9864 [18:33:48<22:43:16, 14.99s/it]
 45%|████▍     | 4408/9864 [18:34:05<23:14:22, 15.33s/it]
 45%|████▍     | 4409/9864 [18:34:19<22:46:24, 15.03s/it]
 45%|████▍     | 4410/9864 [18:34:35<23:04:38, 15.23s/it]
                                                         
{'loss': '0.1459', 'grad_norm': '5.102', 'learning_rate': '1.219e-05', 'epoch': '0.4471'}

 45%|████▍     | 4410/9864 [18:34:35<23:04:38, 15.23s/it]
 45%|████▍     | 4411/9864 [18:34:48<22:20:32, 14.75s/it]
 45%|████▍     | 4412/9864 [18:35:04<22:48:52, 15.06s/it]
 45%|████▍     | 4413/9864 [18:35:18<22:19:55, 14.75s/it]
 45%|████▍     | 4414/9864 [18:35:32<21:56:04, 14.49s/it]
 45%|████▍     | 4415/9864 [18:35:48<22:52:11, 15.11s/it]
 45%|████▍     | 4416/9864 [18:36:04<23:06:48, 15.27s/it]
 45%|████▍     | 4417/9864 [18:36:19<22:43:27, 15.02s/it]
 45%|████▍     | 4418/9864 [18:36:35<23:15:11, 15.37s/it]
 45%|████▍     | 4419/9864 [18:36:51<23:37:25, 15.62s/it]
 45%|████▍     | 4420/9864 [18:37:05<23:04:34, 15.26s/it]
                                                         
{'loss': '0.1413', 'grad_norm': '4.126', 'learning_rate': '1.215e-05', 'epoch': '0.4481'}

 45%|████▍     | 4420/9864 [18:37:05<23:04:34, 15.26s/it]
 45%|████▍     | 4421/9864 [18:37:19<22:24:23, 14.82s/it]
 45%|████▍     | 4422/9864 [18:37:35<22:58:57, 15.20s/it]
 45%|████▍     | 4423/9864 [18:37:51<23:17:16, 15.41s/it]
 45%|████▍     | 4424/9864 [18:38:05<22:29:42, 14.89s/it]
 45%|████▍     | 4425/9864 [18:38:19<22:05:52, 14.63s/it]
 45%|████▍     | 4426/9864 [18:38:35<22:40:03, 15.01s/it]
 45%|████▍     | 4427/9864 [18:38:51<23:07:07, 15.31s/it]
 45%|████▍     | 4428/9864 [18:39:05<22:33:51, 14.94s/it]
 45%|████▍     | 4429/9864 [18:39:21<23:15:26, 15.41s/it]
 45%|████▍     | 4430/9864 [18:39:36<22:44:52, 15.07s/it]
                                                         
{'loss': '0.1588', 'grad_norm': '3.859', 'learning_rate': '1.212e-05', 'epoch': '0.4491'}

 45%|████▍     | 4430/9864 [18:39:36<22:44:52, 15.07s/it]
 45%|████▍     | 4431/9864 [18:39:51<23:04:54, 15.29s/it]
 45%|████▍     | 4432/9864 [18:40:06<22:48:04, 15.11s/it]
 45%|████▍     | 4433/9864 [18:40:21<22:29:29, 14.91s/it]
 45%|████▍     | 4434/9864 [18:40:37<23:02:50, 15.28s/it]
 45%|████▍     | 4435/9864 [18:40:52<23:15:51, 15.43s/it]
 45%|████▍     | 4436/9864 [18:41:07<22:44:25, 15.08s/it]
 45%|████▍     | 4437/9864 [18:41:23<23:04:03, 15.30s/it]
 45%|████▍     | 4438/9864 [18:41:39<23:23:40, 15.52s/it]
 45%|████▌     | 4439/9864 [18:41:53<22:51:43, 15.17s/it]
 45%|████▌     | 4440/9864 [18:42:07<22:18:52, 14.81s/it]
                                                         
{'loss': '0.1411', 'grad_norm': '4.083', 'learning_rate': '1.209e-05', 'epoch': '0.4502'}

 45%|████▌     | 4440/9864 [18:42:07<22:18:52, 14.81s/it]
 45%|████▌     | 4441/9864 [18:42:21<21:55:36, 14.56s/it]
 45%|████▌     | 4442/9864 [18:42:37<22:33:28, 14.98s/it]
 45%|████▌     | 4443/9864 [18:42:53<23:10:29, 15.39s/it]
 45%|████▌     | 4444/9864 [18:43:09<23:32:18, 15.63s/it]
 45%|████▌     | 4445/9864 [18:43:26<23:46:28, 15.79s/it]
 45%|████▌     | 4446/9864 [18:43:40<23:09:29, 15.39s/it]
 45%|████▌     | 4447/9864 [18:43:54<22:33:35, 14.99s/it]
 45%|████▌     | 4448/9864 [18:44:08<22:17:45, 14.82s/it]
 45%|████▌     | 4449/9864 [18:44:24<22:48:56, 15.17s/it]
 45%|████▌     | 4450/9864 [18:44:38<21:58:57, 14.62s/it]
                                                         
{'loss': '0.1375', 'grad_norm': '4.923', 'learning_rate': '1.206e-05', 'epoch': '0.4512'}

 45%|████▌     | 4450/9864 [18:44:38<21:58:57, 14.62s/it]
 45%|████▌     | 4451/9864 [18:44:54<22:31:05, 14.98s/it]
 45%|████▌     | 4452/9864 [18:45:09<22:54:40, 15.24s/it]
 45%|████▌     | 4453/9864 [18:45:24<22:24:48, 14.91s/it]
 45%|████▌     | 4454/9864 [18:45:39<22:25:01, 14.92s/it]
 45%|████▌     | 4455/9864 [18:45:53<22:05:12, 14.70s/it]
 45%|████▌     | 4456/9864 [18:46:07<21:49:56, 14.53s/it]
 45%|████▌     | 4457/9864 [18:46:23<22:30:38, 14.99s/it]
 45%|████▌     | 4458/9864 [18:46:37<22:07:53, 14.74s/it]
 45%|████▌     | 4459/9864 [18:46:53<22:44:48, 15.15s/it]
 45%|████▌     | 4460/9864 [18:47:09<23:05:43, 15.39s/it]
                                                         
{'loss': '0.1543', 'grad_norm': '5.178', 'learning_rate': '1.202e-05', 'epoch': '0.4522'}

 45%|████▌     | 4460/9864 [18:47:09<23:05:43, 15.39s/it]
 45%|████▌     | 4461/9864 [18:47:23<22:31:39, 15.01s/it]
 45%|████▌     | 4462/9864 [18:47:40<23:24:28, 15.60s/it]
 45%|████▌     | 4463/9864 [18:47:54<22:43:59, 15.15s/it]
 45%|████▌     | 4464/9864 [18:48:08<22:12:18, 14.80s/it]
 45%|████▌     | 4465/9864 [18:48:24<22:45:29, 15.17s/it]
 45%|████▌     | 4466/9864 [18:48:39<22:34:47, 15.06s/it]
 45%|████▌     | 4467/9864 [18:48:57<23:49:39, 15.89s/it]
 45%|████▌     | 4468/9864 [18:49:11<23:03:53, 15.39s/it]
 45%|████▌     | 4469/9864 [18:49:25<22:24:30, 14.95s/it]
 45%|████▌     | 4470/9864 [18:49:41<22:44:09, 15.17s/it]
                                                         
{'loss': '0.1497', 'grad_norm': '2.189', 'learning_rate': '1.199e-05', 'epoch': '0.4532'}

 45%|████▌     | 4470/9864 [18:49:41<22:44:09, 15.17s/it]
 45%|████▌     | 4471/9864 [18:49:55<22:15:35, 14.86s/it]
 45%|████▌     | 4472/9864 [18:50:09<21:58:49, 14.68s/it]
 45%|████▌     | 4473/9864 [18:50:25<22:18:52, 14.90s/it]
 45%|████▌     | 4474/9864 [18:50:40<22:32:42, 15.06s/it]
 45%|████▌     | 4475/9864 [18:50:54<22:14:09, 14.85s/it]
 45%|████▌     | 4476/9864 [18:51:10<22:37:21, 15.12s/it]
 45%|████▌     | 4477/9864 [18:51:24<22:07:55, 14.79s/it]
 45%|████▌     | 4478/9864 [18:51:38<21:38:17, 14.46s/it]
 45%|████▌     | 4479/9864 [18:51:54<22:12:51, 14.85s/it]
 45%|████▌     | 4480/9864 [18:52:10<22:44:37, 15.21s/it]
                                                         
{'loss': '0.1442', 'grad_norm': '2.981', 'learning_rate': '1.196e-05', 'epoch': '0.4542'}

 45%|████▌     | 4480/9864 [18:52:10<22:44:37, 15.21s/it]
 45%|████▌     | 4481/9864 [18:52:24<22:15:55, 14.89s/it]
 45%|████▌     | 4482/9864 [18:52:40<22:43:11, 15.20s/it]
 45%|████▌     | 4483/9864 [18:52:54<22:21:26, 14.96s/it]
 45%|████▌     | 4484/9864 [18:53:10<22:46:12, 15.24s/it]
 45%|████▌     | 4485/9864 [18:53:24<22:16:53, 14.91s/it]
 45%|████▌     | 4486/9864 [18:53:39<22:25:13, 15.01s/it]
 45%|████▌     | 4487/9864 [18:53:56<23:00:18, 15.40s/it]
 45%|████▌     | 4488/9864 [18:54:10<22:30:28, 15.07s/it]
 46%|████▌     | 4489/9864 [18:54:24<22:03:41, 14.78s/it]
 46%|████▌     | 4490/9864 [18:54:40<22:26:12, 15.03s/it]
                                                         
{'loss': '0.1388', 'grad_norm': '4.711', 'learning_rate': '1.193e-05', 'epoch': '0.4552'}

 46%|████▌     | 4490/9864 [18:54:40<22:26:12, 15.03s/it]
 46%|████▌     | 4491/9864 [18:54:54<22:00:28, 14.75s/it]
 46%|████▌     | 4492/9864 [18:55:08<21:52:43, 14.66s/it]
 46%|████▌     | 4493/9864 [18:55:25<22:38:02, 15.17s/it]
 46%|████▌     | 4494/9864 [18:55:43<23:57:23, 16.06s/it]
 46%|████▌     | 4495/9864 [18:55:57<23:09:47, 15.53s/it]
 46%|████▌     | 4496/9864 [18:56:11<22:23:15, 15.01s/it]
 46%|████▌     | 4497/9864 [18:56:25<21:49:43, 14.64s/it]
 46%|████▌     | 4498/9864 [18:56:40<22:20:14, 14.99s/it]
 46%|████▌     | 4499/9864 [18:56:54<21:39:28, 14.53s/it]
 46%|████▌     | 4500/9864 [18:57:10<22:14:35, 14.93s/it]
                                                         
{'loss': '0.1603', 'grad_norm': '2.469', 'learning_rate': '1.19e-05', 'epoch': '0.4562'}

 46%|████▌     | 4500/9864 [18:57:10<22:14:35, 14.93s/it]
 46%|████▌     | 4501/9864 [18:57:26<22:35:39, 15.17s/it]
 46%|████▌     | 4502/9864 [18:57:41<22:55:56, 15.40s/it]
 46%|████▌     | 4503/9864 [18:57:56<22:24:53, 15.05s/it]
 46%|████▌     | 4504/9864 [18:58:10<22:06:35, 14.85s/it]
 46%|████▌     | 4505/9864 [18:58:26<22:28:22, 15.10s/it]
 46%|████▌     | 4506/9864 [18:58:40<21:59:57, 14.78s/it]
 46%|████▌     | 4507/9864 [18:58:55<21:58:58, 14.77s/it]
 46%|████▌     | 4508/9864 [18:59:11<22:33:44, 15.17s/it]
 46%|████▌     | 4509/9864 [18:59:29<23:52:33, 16.05s/it]
 46%|████▌     | 4510/9864 [18:59:42<22:48:57, 15.34s/it]
                                                         
{'loss': '0.1495', 'grad_norm': '1.708', 'learning_rate': '1.186e-05', 'epoch': '0.4573'}

 46%|████▌     | 4510/9864 [18:59:42<22:48:57, 15.34s/it]
 46%|████▌     | 4511/9864 [18:59:57<22:18:25, 15.00s/it]
 46%|████▌     | 4512/9864 [19:00:11<21:52:15, 14.71s/it]
 46%|████▌     | 4513/9864 [19:00:27<22:27:29, 15.11s/it]
 46%|████▌     | 4514/9864 [19:00:41<22:08:34, 14.90s/it]
 46%|████▌     | 4515/9864 [19:00:57<22:24:57, 15.09s/it]
 46%|████▌     | 4516/9864 [19:01:15<23:42:58, 15.96s/it]
 46%|████▌     | 4517/9864 [19:01:29<23:06:24, 15.56s/it]
 46%|████▌     | 4518/9864 [19:01:44<22:39:24, 15.26s/it]
 46%|████▌     | 4519/9864 [19:01:58<22:16:44, 15.01s/it]
 46%|████▌     | 4520/9864 [19:02:14<22:37:53, 15.25s/it]
                                                         
{'loss': '0.1465', 'grad_norm': '1.938', 'learning_rate': '1.183e-05', 'epoch': '0.4583'}

 46%|████▌     | 4520/9864 [19:02:14<22:37:53, 15.25s/it]
 46%|████▌     | 4521/9864 [19:02:28<22:12:54, 14.97s/it]
 46%|████▌     | 4522/9864 [19:02:43<21:54:02, 14.76s/it]
 46%|████▌     | 4523/9864 [19:02:59<22:24:08, 15.10s/it]
 46%|████▌     | 4524/9864 [19:03:15<23:08:31, 15.60s/it]
 46%|████▌     | 4525/9864 [19:03:29<22:15:13, 15.01s/it]
 46%|████▌     | 4526/9864 [19:03:43<21:56:10, 14.79s/it]
 46%|████▌     | 4527/9864 [19:03:58<21:47:32, 14.70s/it]
 46%|████▌     | 4528/9864 [19:04:13<22:15:01, 15.01s/it]
 46%|████▌     | 4529/9864 [19:04:28<21:51:54, 14.75s/it]
 46%|████▌     | 4530/9864 [19:04:42<21:48:03, 14.71s/it]
                                                         
{'loss': '0.1284', 'grad_norm': '2.115', 'learning_rate': '1.18e-05', 'epoch': '0.4593'}

 46%|████▌     | 4530/9864 [19:04:42<21:48:03, 14.71s/it]
 46%|████▌     | 4531/9864 [19:05:00<23:09:22, 15.63s/it]
 46%|████▌     | 4532/9864 [19:05:16<23:09:04, 15.63s/it]
 46%|████▌     | 4533/9864 [19:05:29<22:16:59, 15.05s/it]
 46%|████▌     | 4534/9864 [19:05:44<21:59:46, 14.86s/it]
 46%|████▌     | 4535/9864 [19:05:58<21:50:23, 14.75s/it]
 46%|████▌     | 4536/9864 [19:06:13<21:39:07, 14.63s/it]
 46%|████▌     | 4537/9864 [19:06:28<22:13:02, 15.01s/it]
 46%|████▌     | 4538/9864 [19:06:43<22:00:14, 14.87s/it]
 46%|████▌     | 4539/9864 [19:07:01<23:09:31, 15.66s/it]
 46%|████▌     | 4540/9864 [19:07:17<23:20:48, 15.79s/it]
                                                         
{'loss': '0.1422', 'grad_norm': '2.4', 'learning_rate': '1.177e-05', 'epoch': '0.4603'}

 46%|████▌     | 4540/9864 [19:07:17<23:20:48, 15.79s/it]
 46%|████▌     | 4541/9864 [19:07:31<22:46:04, 15.40s/it]
 46%|████▌     | 4542/9864 [19:07:45<22:11:30, 15.01s/it]
 46%|████▌     | 4543/9864 [19:07:59<21:45:46, 14.72s/it]
 46%|████▌     | 4544/9864 [19:08:15<22:16:38, 15.07s/it]
 46%|████▌     | 4545/9864 [19:08:29<21:46:01, 14.73s/it]
 46%|████▌     | 4546/9864 [19:08:45<22:23:31, 15.16s/it]
 46%|████▌     | 4547/9864 [19:09:01<22:47:52, 15.44s/it]
 46%|████▌     | 4548/9864 [19:09:17<22:58:13, 15.56s/it]
 46%|████▌     | 4549/9864 [19:09:32<22:31:43, 15.26s/it]
 46%|████▌     | 4550/9864 [19:09:46<22:14:27, 15.07s/it]
                                                         
{'loss': '0.1496', 'grad_norm': '5.318', 'learning_rate': '1.173e-05', 'epoch': '0.4613'}

 46%|████▌     | 4550/9864 [19:09:46<22:14:27, 15.07s/it]
 46%|████▌     | 4551/9864 [19:10:01<22:02:09, 14.93s/it]
 46%|████▌     | 4552/9864 [19:10:17<22:26:48, 15.21s/it]
 46%|████▌     | 4553/9864 [19:10:31<22:00:38, 14.92s/it]
 46%|████▌     | 4554/9864 [19:10:47<22:39:53, 15.37s/it]
 46%|████▌     | 4555/9864 [19:11:05<23:28:28, 15.92s/it]
 46%|████▌     | 4556/9864 [19:11:19<22:38:21, 15.35s/it]
 46%|████▌     | 4557/9864 [19:11:33<22:04:55, 14.98s/it]
 46%|████▌     | 4558/9864 [19:11:47<21:52:45, 14.84s/it]
 46%|████▌     | 4559/9864 [19:12:02<21:43:45, 14.75s/it]
 46%|████▌     | 4560/9864 [19:12:17<22:05:55, 15.00s/it]
                                                         
{'loss': '0.1685', 'grad_norm': '1.895', 'learning_rate': '1.17e-05', 'epoch': '0.4623'}

 46%|████▌     | 4560/9864 [19:12:17<22:05:55, 15.00s/it]
 46%|████▌     | 4561/9864 [19:12:34<22:37:56, 15.36s/it]
 46%|████▌     | 4562/9864 [19:12:49<22:43:32, 15.43s/it]
 46%|████▋     | 4563/9864 [19:13:03<22:11:23, 15.07s/it]
 46%|████▋     | 4564/9864 [19:13:20<22:36:48, 15.36s/it]
 46%|████▋     | 4565/9864 [19:13:34<22:02:39, 14.98s/it]
 46%|████▋     | 4566/9864 [19:13:48<21:47:42, 14.81s/it]
 46%|████▋     | 4567/9864 [19:14:02<21:28:18, 14.59s/it]
 46%|████▋     | 4568/9864 [19:14:18<22:08:14, 15.05s/it]
 46%|████▋     | 4569/9864 [19:14:35<22:41:03, 15.42s/it]
 46%|████▋     | 4570/9864 [19:14:50<22:44:40, 15.47s/it]
                                                         
{'loss': '0.1445', 'grad_norm': '2.974', 'learning_rate': '1.167e-05', 'epoch': '0.4633'}

 46%|████▋     | 4570/9864 [19:14:50<22:44:40, 15.47s/it]
 46%|████▋     | 4571/9864 [19:15:04<22:02:53, 15.00s/it]
 46%|████▋     | 4572/9864 [19:15:20<22:30:01, 15.31s/it]
 46%|████▋     | 4573/9864 [19:15:34<22:05:35, 15.03s/it]
 46%|████▋     | 4574/9864 [19:15:48<21:34:57, 14.69s/it]
 46%|████▋     | 4575/9864 [19:16:02<21:16:49, 14.48s/it]
 46%|████▋     | 4576/9864 [19:16:18<22:00:48, 14.99s/it]
 46%|████▋     | 4577/9864 [19:16:34<22:28:02, 15.30s/it]
 46%|████▋     | 4578/9864 [19:16:51<22:51:12, 15.56s/it]
 46%|████▋     | 4579/9864 [19:17:05<22:18:36, 15.20s/it]
 46%|████▋     | 4580/9864 [19:17:21<22:28:45, 15.32s/it]
                                                         
{'loss': '0.1319', 'grad_norm': '2.825', 'learning_rate': '1.164e-05', 'epoch': '0.4644'}

 46%|████▋     | 4580/9864 [19:17:21<22:28:45, 15.32s/it]
 46%|████▋     | 4581/9864 [19:17:34<21:51:05, 14.89s/it]
 46%|████▋     | 4582/9864 [19:17:49<21:35:45, 14.72s/it]
 46%|████▋     | 4583/9864 [19:18:05<22:01:56, 15.02s/it]
 46%|████▋     | 4584/9864 [19:18:19<21:42:14, 14.80s/it]
 46%|████▋     | 4585/9864 [19:18:34<22:04:37, 15.06s/it]
 46%|████▋     | 4586/9864 [19:18:48<21:28:38, 14.65s/it]
 47%|████▋     | 4587/9864 [19:19:04<21:52:22, 14.92s/it]
 47%|████▋     | 4588/9864 [19:19:20<22:18:17, 15.22s/it]
 47%|████▋     | 4589/9864 [19:19:34<21:49:22, 14.89s/it]
 47%|████▋     | 4590/9864 [19:19:48<21:31:40, 14.69s/it]
                                                         
{'loss': '0.1598', 'grad_norm': '4.624', 'learning_rate': '1.161e-05', 'epoch': '0.4654'}

 47%|████▋     | 4590/9864 [19:19:48<21:31:40, 14.69s/it]
 47%|████▋     | 4591/9864 [19:20:04<22:02:49, 15.05s/it]
 47%|████▋     | 4592/9864 [19:20:18<21:35:32, 14.74s/it]
 47%|████▋     | 4593/9864 [19:20:33<21:57:04, 14.99s/it]
 47%|████▋     | 4594/9864 [19:20:50<22:24:19, 15.31s/it]
 47%|████▋     | 4595/9864 [19:21:05<22:32:46, 15.40s/it]
 47%|████▋     | 4596/9864 [19:21:19<21:55:17, 14.98s/it]
 47%|████▋     | 4597/9864 [19:21:34<21:39:18, 14.80s/it]
 47%|████▋     | 4598/9864 [19:21:50<22:11:19, 15.17s/it]
 47%|████▋     | 4599/9864 [19:22:04<21:39:48, 14.81s/it]
 47%|████▋     | 4600/9864 [19:22:18<21:28:29, 14.69s/it]
                                                         
{'loss': '0.1398', 'grad_norm': '1.53', 'learning_rate': '1.157e-05', 'epoch': '0.4664'}

 47%|████▋     | 4600/9864 [19:22:18<21:28:29, 14.69s/it]
 47%|████▋     | 4601/9864 [19:22:34<22:00:45, 15.06s/it]
 47%|████▋     | 4602/9864 [19:22:50<22:31:38, 15.41s/it]
 47%|████▋     | 4603/9864 [19:23:06<22:48:17, 15.61s/it]
 47%|████▋     | 4604/9864 [19:23:21<22:23:17, 15.32s/it]
 47%|████▋     | 4605/9864 [19:23:35<21:54:22, 15.00s/it]
 47%|████▋     | 4606/9864 [19:23:51<22:25:32, 15.35s/it]
 47%|████▋     | 4607/9864 [19:24:06<22:02:13, 15.09s/it]
 47%|████▋     | 4608/9864 [19:24:20<21:43:11, 14.88s/it]
 47%|████▋     | 4609/9864 [19:24:36<22:09:02, 15.17s/it]
 47%|████▋     | 4610/9864 [19:24:53<23:09:16, 15.87s/it]
                                                         
{'loss': '0.1391', 'grad_norm': '4.046', 'learning_rate': '1.154e-05', 'epoch': '0.4674'}

 47%|████▋     | 4610/9864 [19:24:53<23:09:16, 15.87s/it]
 47%|████▋     | 4611/9864 [19:25:07<22:13:45, 15.23s/it]
 47%|████▋     | 4612/9864 [19:25:21<21:36:46, 14.81s/it]
 47%|████▋     | 4613/9864 [19:25:37<22:03:15, 15.12s/it]
 47%|████▋     | 4614/9864 [19:25:51<21:34:59, 14.80s/it]
 47%|████▋     | 4615/9864 [19:26:04<20:59:51, 14.40s/it]
 47%|████▋     | 4616/9864 [19:26:18<20:42:35, 14.21s/it]
 47%|████▋     | 4617/9864 [19:26:34<21:31:34, 14.77s/it]
 47%|████▋     | 4618/9864 [19:26:52<22:52:19, 15.70s/it]
 47%|████▋     | 4619/9864 [19:27:07<22:22:33, 15.36s/it]
 47%|████▋     | 4620/9864 [19:27:21<21:45:53, 14.94s/it]
                                                         
{'loss': '0.1362', 'grad_norm': '2.776', 'learning_rate': '1.151e-05', 'epoch': '0.4684'}

 47%|████▋     | 4620/9864 [19:27:21<21:45:53, 14.94s/it]
 47%|████▋     | 4621/9864 [19:27:36<22:06:03, 15.18s/it]
 47%|████▋     | 4622/9864 [19:27:51<21:48:28, 14.98s/it]
 47%|████▋     | 4623/9864 [19:28:05<21:33:08, 14.80s/it]
 47%|████▋     | 4624/9864 [19:28:19<21:16:55, 14.62s/it]
 47%|████▋     | 4625/9864 [19:28:34<21:02:46, 14.46s/it]
 47%|████▋     | 4626/9864 [19:28:53<23:21:21, 16.05s/it]
 47%|████▋     | 4627/9864 [19:29:08<22:47:00, 15.66s/it]
 47%|████▋     | 4628/9864 [19:29:22<21:51:29, 15.03s/it]
 47%|████▋     | 4629/9864 [19:29:37<22:09:38, 15.24s/it]
 47%|████▋     | 4630/9864 [19:29:52<21:44:29, 14.95s/it]
                                                         
{'loss': '0.1413', 'grad_norm': '1.505', 'learning_rate': '1.148e-05', 'epoch': '0.4694'}

 47%|████▋     | 4630/9864 [19:29:52<21:44:29, 14.95s/it]
 47%|████▋     | 4631/9864 [19:30:06<21:28:13, 14.77s/it]
 47%|████▋     | 4632/9864 [19:30:20<21:19:10, 14.67s/it]
 47%|████▋     | 4633/9864 [19:30:35<21:10:05, 14.57s/it]
 47%|████▋     | 4634/9864 [19:30:54<23:18:05, 16.04s/it]
 47%|████▋     | 4635/9864 [19:31:09<22:33:17, 15.53s/it]
 47%|████▋     | 4636/9864 [19:31:24<22:38:07, 15.59s/it]
 47%|████▋     | 4637/9864 [19:31:38<21:50:40, 15.05s/it]
 47%|████▋     | 4638/9864 [19:31:52<21:28:56, 14.80s/it]
 47%|████▋     | 4639/9864 [19:32:07<21:19:19, 14.69s/it]
 47%|████▋     | 4640/9864 [19:32:22<21:23:51, 14.75s/it]
                                                         
{'loss': '0.1293', 'grad_norm': '1.975', 'learning_rate': '1.144e-05', 'epoch': '0.4704'}

 47%|████▋     | 4640/9864 [19:32:22<21:23:51, 14.75s/it]
 47%|████▋     | 4641/9864 [19:32:39<22:42:44, 15.65s/it]
 47%|████▋     | 4642/9864 [19:32:56<22:59:51, 15.85s/it]
 47%|████▋     | 4643/9864 [19:33:10<22:06:57, 15.25s/it]
 47%|████▋     | 4644/9864 [19:33:26<22:36:52, 15.60s/it]
 47%|████▋     | 4645/9864 [19:33:40<21:59:07, 15.17s/it]
 47%|████▋     | 4646/9864 [19:33:54<21:31:06, 14.85s/it]
 47%|████▋     | 4647/9864 [19:34:09<21:17:10, 14.69s/it]
 47%|████▋     | 4648/9864 [19:34:22<20:58:33, 14.48s/it]
 47%|████▋     | 4649/9864 [19:34:42<23:13:44, 16.04s/it]
 47%|████▋     | 4650/9864 [19:34:56<22:20:33, 15.43s/it]
                                                         
{'loss': '0.1477', 'grad_norm': '3.083', 'learning_rate': '1.141e-05', 'epoch': '0.4715'}

 47%|████▋     | 4650/9864 [19:34:56<22:20:33, 15.43s/it]
 47%|████▋     | 4651/9864 [19:35:12<22:39:51, 15.65s/it]
 47%|████▋     | 4652/9864 [19:35:27<22:09:46, 15.31s/it]
 47%|████▋     | 4653/9864 [19:35:41<21:38:46, 14.95s/it]
 47%|████▋     | 4654/9864 [19:35:55<21:18:08, 14.72s/it]
 47%|████▋     | 4655/9864 [19:36:09<21:01:10, 14.53s/it]
 47%|████▋     | 4656/9864 [19:36:25<21:40:11, 14.98s/it]
 47%|████▋     | 4657/9864 [19:36:43<22:55:41, 15.85s/it]
 47%|████▋     | 4658/9864 [19:36:57<22:11:10, 15.34s/it]
 47%|████▋     | 4659/9864 [19:37:13<22:27:01, 15.53s/it]
 47%|████▋     | 4660/9864 [19:37:27<21:46:41, 15.07s/it]
                                                         
{'loss': '0.1402', 'grad_norm': '4.581', 'learning_rate': '1.138e-05', 'epoch': '0.4725'}

 47%|████▋     | 4660/9864 [19:37:27<21:46:41, 15.07s/it]
 47%|████▋     | 4661/9864 [19:37:41<21:11:45, 14.67s/it]
 47%|████▋     | 4662/9864 [19:37:56<21:18:34, 14.75s/it]
 47%|████▋     | 4663/9864 [19:38:10<21:04:06, 14.58s/it]
 47%|████▋     | 4664/9864 [19:38:26<21:41:49, 15.02s/it]
 47%|████▋     | 4665/9864 [19:38:44<22:54:35, 15.86s/it]
 47%|████▋     | 4666/9864 [19:38:58<22:11:29, 15.37s/it]
 47%|████▋     | 4667/9864 [19:39:14<22:21:51, 15.49s/it]
 47%|████▋     | 4668/9864 [19:39:28<21:40:02, 15.01s/it]
 47%|████▋     | 4669/9864 [19:39:42<21:13:10, 14.70s/it]
 47%|████▋     | 4670/9864 [19:39:57<21:15:41, 14.74s/it]
                                                         
{'loss': '0.1326', 'grad_norm': '2.96', 'learning_rate': '1.135e-05', 'epoch': '0.4735'}

 47%|████▋     | 4670/9864 [19:39:57<21:15:41, 14.74s/it]
 47%|████▋     | 4671/9864 [19:40:13<21:49:18, 15.13s/it]
 47%|████▋     | 4672/9864 [19:40:30<22:52:55, 15.87s/it]
 47%|████▋     | 4673/9864 [19:40:45<22:17:20, 15.46s/it]
 47%|████▋     | 4674/9864 [19:41:01<22:23:58, 15.54s/it]
 47%|████▋     | 4675/9864 [19:41:14<21:35:40, 14.98s/it]
 47%|████▋     | 4676/9864 [19:41:29<21:20:00, 14.80s/it]
 47%|████▋     | 4677/9864 [19:41:43<21:09:16, 14.68s/it]
 47%|████▋     | 4678/9864 [19:41:57<21:00:13, 14.58s/it]
 47%|████▋     | 4679/9864 [19:42:14<21:57:48, 15.25s/it]
 47%|████▋     | 4680/9864 [19:42:30<22:20:41, 15.52s/it]
                                                         
{'loss': '0.1376', 'grad_norm': '2.249', 'learning_rate': '1.131e-05', 'epoch': '0.4745'}

 47%|████▋     | 4680/9864 [19:42:30<22:20:41, 15.52s/it]
 47%|████▋     | 4681/9864 [19:42:44<21:42:56, 15.08s/it]
 47%|████▋     | 4682/9864 [19:43:00<22:04:25, 15.33s/it]
 47%|████▋     | 4683/9864 [19:43:14<21:24:26, 14.87s/it]
 47%|████▋     | 4684/9864 [19:43:28<21:01:30, 14.61s/it]
 47%|████▋     | 4685/9864 [19:43:42<20:48:54, 14.47s/it]
 48%|████▊     | 4686/9864 [19:43:58<21:29:54, 14.95s/it]
 48%|████▊     | 4687/9864 [19:44:14<21:43:41, 15.11s/it]
 48%|████▊     | 4688/9864 [19:44:29<21:56:11, 15.26s/it]
 48%|████▊     | 4689/9864 [19:44:46<22:26:13, 15.61s/it]
 48%|████▊     | 4690/9864 [19:45:00<21:52:58, 15.23s/it]
                                                         
{'loss': '0.1434', 'grad_norm': '2.257', 'learning_rate': '1.128e-05', 'epoch': '0.4755'}

 48%|████▊     | 4690/9864 [19:45:00<21:52:58, 15.23s/it]
 48%|████▊     | 4691/9864 [19:45:15<21:30:37, 14.97s/it]
 48%|████▊     | 4692/9864 [19:45:28<20:58:40, 14.60s/it]
 48%|████▊     | 4693/9864 [19:45:42<20:45:29, 14.45s/it]
 48%|████▊     | 4694/9864 [19:45:57<20:37:50, 14.37s/it]
 48%|████▊     | 4695/9864 [19:46:14<22:05:16, 15.38s/it]
 48%|████▊     | 4696/9864 [19:46:32<22:58:51, 16.01s/it]
 48%|████▊     | 4697/9864 [19:46:46<22:12:54, 15.48s/it]
 48%|████▊     | 4698/9864 [19:47:00<21:38:35, 15.08s/it]
 48%|████▊     | 4699/9864 [19:47:14<21:17:31, 14.84s/it]
 48%|████▊     | 4700/9864 [19:47:29<21:02:44, 14.67s/it]
                                                         
{'loss': '0.1449', 'grad_norm': '1.889', 'learning_rate': '1.125e-05', 'epoch': '0.4765'}

 48%|████▊     | 4700/9864 [19:47:29<21:02:44, 14.67s/it]
 48%|████▊     | 4701/9864 [19:47:43<20:50:32, 14.53s/it]
 48%|████▊     | 4702/9864 [19:47:59<21:29:14, 14.99s/it]
 48%|████▊     | 4703/9864 [19:48:14<21:35:16, 15.06s/it]
 48%|████▊     | 4704/9864 [19:48:32<22:44:43, 15.87s/it]
 48%|████▊     | 4705/9864 [19:48:45<21:42:27, 15.15s/it]
 48%|████▊     | 4706/9864 [19:48:59<21:06:29, 14.73s/it]
 48%|████▊     | 4707/9864 [19:49:13<20:54:09, 14.59s/it]
 48%|████▊     | 4708/9864 [19:49:28<20:47:40, 14.52s/it]
 48%|████▊     | 4709/9864 [19:49:42<20:42:16, 14.46s/it]
 48%|████▊     | 4710/9864 [19:49:58<21:29:25, 15.01s/it]
                                                         
{'loss': '0.139', 'grad_norm': '2.585', 'learning_rate': '1.122e-05', 'epoch': '0.4775'}

 48%|████▊     | 4710/9864 [19:49:58<21:29:25, 15.01s/it]
 48%|████▊     | 4711/9864 [19:50:14<21:45:26, 15.20s/it]
 48%|████▊     | 4712/9864 [19:50:30<22:06:48, 15.45s/it]
 48%|████▊     | 4713/9864 [19:50:44<21:37:35, 15.11s/it]
 48%|████▊     | 4714/9864 [19:50:59<21:18:31, 14.90s/it]
 48%|████▊     | 4715/9864 [19:51:13<21:01:58, 14.71s/it]
 48%|████▊     | 4716/9864 [19:51:28<20:55:43, 14.64s/it]
 48%|████▊     | 4717/9864 [19:51:44<21:33:05, 15.07s/it]
 48%|████▊     | 4718/9864 [19:52:01<22:37:42, 15.83s/it]
 48%|████▊     | 4719/9864 [19:52:15<21:54:38, 15.33s/it]
 48%|████▊     | 4720/9864 [19:52:32<22:21:12, 15.64s/it]
                                                         
{'loss': '0.1284', 'grad_norm': '2.453', 'learning_rate': '1.118e-05', 'epoch': '0.4785'}

 48%|████▊     | 4720/9864 [19:52:32<22:21:12, 15.64s/it]
 48%|████▊     | 4721/9864 [19:52:46<21:55:18, 15.34s/it]
 48%|████▊     | 4722/9864 [19:53:00<21:18:28, 14.92s/it]
 48%|████▊     | 4723/9864 [19:53:14<20:51:52, 14.61s/it]
 48%|████▊     | 4724/9864 [19:53:30<21:27:38, 15.03s/it]
 48%|████▊     | 4725/9864 [19:53:44<21:04:11, 14.76s/it]
 48%|████▊     | 4726/9864 [19:54:02<22:18:30, 15.63s/it]
 48%|████▊     | 4727/9864 [19:54:16<21:48:00, 15.28s/it]
 48%|████▊     | 4728/9864 [19:54:32<21:56:59, 15.39s/it]
 48%|████▊     | 4729/9864 [19:54:46<21:23:43, 15.00s/it]
 48%|████▊     | 4730/9864 [19:55:00<20:40:58, 14.50s/it]
                                                         
{'loss': '0.1447', 'grad_norm': '2.331', 'learning_rate': '1.115e-05', 'epoch': '0.4796'}

 48%|████▊     | 4730/9864 [19:55:00<20:40:58, 14.50s/it]
 48%|████▊     | 4731/9864 [19:55:14<20:34:45, 14.43s/it]
 48%|████▊     | 4732/9864 [19:55:28<20:38:48, 14.48s/it]
 48%|████▊     | 4733/9864 [19:55:46<21:56:55, 15.40s/it]
 48%|████▊     | 4734/9864 [19:56:02<22:01:06, 15.45s/it]
 48%|████▊     | 4735/9864 [19:56:16<21:34:14, 15.14s/it]
 48%|████▊     | 4736/9864 [19:56:32<21:58:28, 15.43s/it]
 48%|████▊     | 4737/9864 [19:56:47<21:42:38, 15.24s/it]
 48%|████▊     | 4738/9864 [19:57:02<21:31:05, 15.11s/it]
 48%|████▊     | 4739/9864 [19:57:17<21:29:03, 15.09s/it]
 48%|████▊     | 4740/9864 [19:57:31<21:05:01, 14.81s/it]
                                                         
{'loss': '0.1367', 'grad_norm': '1.867', 'learning_rate': '1.112e-05', 'epoch': '0.4806'}

 48%|████▊     | 4740/9864 [19:57:31<21:05:01, 14.81s/it]
 48%|████▊     | 4741/9864 [19:57:48<22:12:02, 15.60s/it]
 48%|████▊     | 4742/9864 [19:58:02<21:24:54, 15.05s/it]
 48%|████▊     | 4743/9864 [19:58:18<21:51:06, 15.36s/it]
 48%|████▊     | 4744/9864 [19:58:34<22:07:01, 15.55s/it]
 48%|████▊     | 4745/9864 [19:58:48<21:28:27, 15.10s/it]
 48%|████▊     | 4746/9864 [19:59:03<21:14:41, 14.94s/it]
 48%|████▊     | 4747/9864 [19:59:17<21:03:12, 14.81s/it]
 48%|████▊     | 4748/9864 [19:59:31<20:40:32, 14.55s/it]
 48%|████▊     | 4749/9864 [19:59:47<21:21:22, 15.03s/it]
 48%|████▊     | 4750/9864 [20:00:03<21:30:52, 15.15s/it]
                                                         
{'loss': '0.1361', 'grad_norm': '3.968', 'learning_rate': '1.108e-05', 'epoch': '0.4816'}

 48%|████▊     | 4750/9864 [20:00:03<21:30:52, 15.15s/it]
 48%|████▊     | 4751/9864 [20:00:19<21:57:26, 15.46s/it]
 48%|████▊     | 4752/9864 [20:00:33<21:08:52, 14.89s/it]
 48%|████▊     | 4753/9864 [20:00:47<20:52:14, 14.70s/it]
 48%|████▊     | 4754/9864 [20:01:01<20:38:34, 14.54s/it]
 48%|████▊     | 4755/9864 [20:01:15<20:30:30, 14.45s/it]
 48%|████▊     | 4756/9864 [20:01:31<21:01:41, 14.82s/it]
 48%|████▊     | 4757/9864 [20:01:45<20:45:43, 14.64s/it]
 48%|████▊     | 4758/9864 [20:02:03<22:08:29, 15.61s/it]
 48%|████▊     | 4759/9864 [20:02:19<22:09:54, 15.63s/it]
 48%|████▊     | 4760/9864 [20:02:33<21:30:02, 15.17s/it]
                                                         
{'loss': '0.1276', 'grad_norm': '2.855', 'learning_rate': '1.105e-05', 'epoch': '0.4826'}

 48%|████▊     | 4760/9864 [20:02:33<21:30:02, 15.17s/it]
 48%|████▊     | 4761/9864 [20:02:46<20:51:19, 14.71s/it]
 48%|████▊     | 4762/9864 [20:03:01<20:46:37, 14.66s/it]
 48%|████▊     | 4763/9864 [20:03:15<20:22:38, 14.38s/it]
 48%|████▊     | 4764/9864 [20:03:31<21:01:12, 14.84s/it]
 48%|████▊     | 4765/9864 [20:03:45<20:44:26, 14.64s/it]
 48%|████▊     | 4766/9864 [20:04:02<21:53:04, 15.45s/it]
 48%|████▊     | 4767/9864 [20:04:18<22:12:55, 15.69s/it]
 48%|████▊     | 4768/9864 [20:04:32<21:31:39, 15.21s/it]
 48%|████▊     | 4769/9864 [20:04:47<21:11:53, 14.98s/it]
 48%|████▊     | 4770/9864 [20:05:01<20:48:50, 14.71s/it]
                                                         
{'loss': '0.1251', 'grad_norm': '5.791', 'learning_rate': '1.102e-05', 'epoch': '0.4836'}

 48%|████▊     | 4770/9864 [20:05:01<20:48:50, 14.71s/it]
 48%|████▊     | 4771/9864 [20:05:15<20:40:29, 14.61s/it]
 48%|████▊     | 4772/9864 [20:05:31<21:16:13, 15.04s/it]
 48%|████▊     | 4773/9864 [20:05:46<20:52:47, 14.76s/it]
 48%|████▊     | 4774/9864 [20:06:01<21:10:39, 14.98s/it]
 48%|████▊     | 4775/9864 [20:06:19<22:28:16, 15.90s/it]
 48%|████▊     | 4776/9864 [20:06:34<21:52:14, 15.47s/it]
 48%|████▊     | 4777/9864 [20:06:48<21:18:33, 15.08s/it]
 48%|████▊     | 4778/9864 [20:07:02<21:08:00, 14.96s/it]
 48%|████▊     | 4779/9864 [20:07:18<21:33:54, 15.27s/it]
 48%|████▊     | 4780/9864 [20:07:33<21:09:34, 14.98s/it]
                                                         
{'loss': '0.1518', 'grad_norm': '1.762', 'learning_rate': '1.099e-05', 'epoch': '0.4846'}

 48%|████▊     | 4780/9864 [20:07:33<21:09:34, 14.98s/it]
 48%|████▊     | 4781/9864 [20:07:47<20:51:26, 14.77s/it]
 48%|████▊     | 4782/9864 [20:08:02<21:04:51, 14.93s/it]
 48%|████▊     | 4783/9864 [20:08:20<22:12:29, 15.73s/it]
 48%|████▊     | 4784/9864 [20:08:34<21:19:54, 15.12s/it]
 49%|████▊     | 4785/9864 [20:08:47<20:31:21, 14.55s/it]
 49%|████▊     | 4786/9864 [20:09:03<21:07:55, 14.98s/it]
 49%|████▊     | 4787/9864 [20:09:17<20:58:31, 14.87s/it]
 49%|████▊     | 4788/9864 [20:09:31<20:36:49, 14.62s/it]
 49%|████▊     | 4789/9864 [20:09:47<21:14:09, 15.06s/it]
 49%|████▊     | 4790/9864 [20:10:03<21:27:52, 15.23s/it]
                                                         
{'loss': '0.1333', 'grad_norm': '3.788', 'learning_rate': '1.095e-05', 'epoch': '0.4856'}

 49%|████▊     | 4790/9864 [20:10:03<21:27:52, 15.23s/it]
 49%|████▊     | 4791/9864 [20:10:19<21:40:34, 15.38s/it]
 49%|████▊     | 4792/9864 [20:10:33<21:05:56, 14.98s/it]
 49%|████▊     | 4793/9864 [20:10:49<21:32:21, 15.29s/it]
 49%|████▊     | 4794/9864 [20:11:03<21:11:11, 15.04s/it]
 49%|████▊     | 4795/9864 [20:11:17<20:37:14, 14.64s/it]
 49%|████▊     | 4796/9864 [20:11:32<20:38:45, 14.67s/it]
 49%|████▊     | 4797/9864 [20:11:46<20:31:16, 14.58s/it]
 49%|████▊     | 4798/9864 [20:12:04<21:41:30, 15.41s/it]
 49%|████▊     | 4799/9864 [20:12:19<21:51:19, 15.53s/it]
 49%|████▊     | 4800/9864 [20:12:33<21:15:41, 15.11s/it]
                                                         
{'loss': '0.1438', 'grad_norm': '5.285', 'learning_rate': '1.092e-05', 'epoch': '0.4867'}

 49%|████▊     | 4800/9864 [20:12:33<21:15:41, 15.11s/it]
 49%|████▊     | 4801/9864 [20:12:50<21:40:12, 15.41s/it]
 49%|████▊     | 4802/9864 [20:13:04<21:05:33, 15.00s/it]
 49%|████▊     | 4803/9864 [20:13:18<20:42:43, 14.73s/it]
 49%|████▊     | 4804/9864 [20:13:32<20:39:39, 14.70s/it]
 49%|████▊     | 4805/9864 [20:13:46<20:20:04, 14.47s/it]
 49%|████▊     | 4806/9864 [20:14:04<21:35:41, 15.37s/it]
 49%|████▊     | 4807/9864 [20:14:19<21:44:02, 15.47s/it]
 49%|████▊     | 4808/9864 [20:14:35<21:47:02, 15.51s/it]
 49%|████▉     | 4809/9864 [20:14:49<21:09:06, 15.06s/it]
 49%|████▉     | 4810/9864 [20:15:03<20:34:03, 14.65s/it]
                                                         
{'loss': '0.154', 'grad_norm': '1.721', 'learning_rate': '1.089e-05', 'epoch': '0.4877'}

 49%|████▉     | 4810/9864 [20:15:03<20:34:03, 14.65s/it]
 49%|████▉     | 4811/9864 [20:15:17<20:26:55, 14.57s/it]
 49%|████▉     | 4812/9864 [20:15:31<20:09:07, 14.36s/it]
 49%|████▉     | 4813/9864 [20:15:45<20:04:10, 14.30s/it]
 49%|████▉     | 4814/9864 [20:16:01<20:43:05, 14.77s/it]
 49%|████▉     | 4815/9864 [20:16:17<21:10:45, 15.10s/it]
 49%|████▉     | 4816/9864 [20:16:32<21:16:06, 15.17s/it]
 49%|████▉     | 4817/9864 [20:16:46<20:43:18, 14.78s/it]
 49%|████▉     | 4818/9864 [20:17:00<20:12:40, 14.42s/it]
 49%|████▉     | 4819/9864 [20:17:14<20:15:26, 14.46s/it]
 49%|████▉     | 4820/9864 [20:17:28<20:04:36, 14.33s/it]
                                                         
{'loss': '0.1533', 'grad_norm': '4.172', 'learning_rate': '1.086e-05', 'epoch': '0.4887'}

 49%|████▉     | 4820/9864 [20:17:28<20:04:36, 14.33s/it]
 49%|████▉     | 4821/9864 [20:17:42<19:51:14, 14.17s/it]
 49%|████▉     | 4822/9864 [20:18:00<21:18:20, 15.21s/it]
 49%|████▉     | 4823/9864 [20:18:15<21:31:33, 15.37s/it]
 49%|████▉     | 4824/9864 [20:18:30<21:08:36, 15.10s/it]
 49%|████▉     | 4825/9864 [20:18:44<20:52:34, 14.91s/it]
 49%|████▉     | 4826/9864 [20:18:59<20:30:59, 14.66s/it]
 49%|████▉     | 4827/9864 [20:19:13<20:27:22, 14.62s/it]
 49%|████▉     | 4828/9864 [20:19:27<20:05:19, 14.36s/it]
 49%|████▉     | 4829/9864 [20:19:42<20:35:28, 14.72s/it]
 49%|████▉     | 4830/9864 [20:19:57<20:26:22, 14.62s/it]
                                                         
{'loss': '0.1267', 'grad_norm': '2.412', 'learning_rate': '1.082e-05', 'epoch': '0.4897'}

 49%|████▉     | 4830/9864 [20:19:57<20:26:22, 14.62s/it]
 49%|████▉     | 4831/9864 [20:20:12<20:49:33, 14.90s/it]
 49%|████▉     | 4832/9864 [20:20:28<21:11:45, 15.16s/it]
 49%|████▉     | 4833/9864 [20:20:42<20:39:19, 14.78s/it]
 49%|████▉     | 4834/9864 [20:20:56<20:25:49, 14.62s/it]
 49%|████▉     | 4835/9864 [20:21:10<20:08:28, 14.42s/it]
 49%|████▉     | 4836/9864 [20:21:24<20:05:18, 14.38s/it]
 49%|████▉     | 4837/9864 [20:21:39<20:00:40, 14.33s/it]
 49%|████▉     | 4838/9864 [20:21:56<21:17:33, 15.25s/it]
 49%|████▉     | 4839/9864 [20:22:12<21:30:38, 15.41s/it]
 49%|████▉     | 4840/9864 [20:22:26<21:04:13, 15.10s/it]
                                                         
{'loss': '0.1388', 'grad_norm': '2.012', 'learning_rate': '1.079e-05', 'epoch': '0.4907'}

 49%|████▉     | 4840/9864 [20:22:26<21:04:13, 15.10s/it]
 49%|████▉     | 4841/9864 [20:22:41<20:51:22, 14.95s/it]
 49%|████▉     | 4842/9864 [20:22:55<20:34:50, 14.75s/it]
 49%|████▉     | 4843/9864 [20:23:09<20:22:04, 14.60s/it]
 49%|████▉     | 4844/9864 [20:23:23<20:06:23, 14.42s/it]
 49%|████▉     | 4845/9864 [20:23:39<20:45:12, 14.89s/it]
 49%|████▉     | 4846/9864 [20:23:57<21:48:34, 15.65s/it]
 49%|████▉     | 4847/9864 [20:24:13<21:55:39, 15.73s/it]
 49%|████▉     | 4848/9864 [20:24:27<21:31:03, 15.44s/it]
 49%|████▉     | 4849/9864 [20:24:42<21:04:20, 15.13s/it]
 49%|████▉     | 4850/9864 [20:24:56<20:46:15, 14.91s/it]
                                                         
{'loss': '0.142', 'grad_norm': '3.797', 'learning_rate': '1.076e-05', 'epoch': '0.4917'}

 49%|████▉     | 4850/9864 [20:24:56<20:46:15, 14.91s/it]
 49%|████▉     | 4851/9864 [20:25:10<20:24:06, 14.65s/it]
 49%|████▉     | 4852/9864 [20:25:24<20:08:47, 14.47s/it]
 49%|████▉     | 4853/9864 [20:25:40<20:36:18, 14.80s/it]
 49%|████▉     | 4854/9864 [20:25:56<20:56:32, 15.05s/it]
 49%|████▉     | 4855/9864 [20:26:13<22:06:14, 15.89s/it]
 49%|████▉     | 4856/9864 [20:26:27<21:20:41, 15.34s/it]
 49%|████▉     | 4857/9864 [20:26:41<20:42:03, 14.88s/it]
 49%|████▉     | 4858/9864 [20:26:56<20:29:10, 14.73s/it]
 49%|████▉     | 4859/9864 [20:27:10<20:16:12, 14.58s/it]
 49%|████▉     | 4860/9864 [20:27:24<20:08:43, 14.49s/it]
                                                         
{'loss': '0.1327', 'grad_norm': '6.346', 'learning_rate': '1.073e-05', 'epoch': '0.4927'}

 49%|████▉     | 4860/9864 [20:27:24<20:08:43, 14.49s/it]
 49%|████▉     | 4861/9864 [20:27:41<20:55:30, 15.06s/it]
 49%|████▉     | 4862/9864 [20:27:58<21:52:12, 15.74s/it]
 49%|████▉     | 4863/9864 [20:28:14<22:02:48, 15.87s/it]
 49%|████▉     | 4864/9864 [20:28:28<21:21:44, 15.38s/it]
 49%|████▉     | 4865/9864 [20:28:43<20:56:55, 15.09s/it]
 49%|████▉     | 4866/9864 [20:28:57<20:32:09, 14.79s/it]
 49%|████▉     | 4867/9864 [20:29:11<20:21:58, 14.67s/it]
 49%|████▉     | 4868/9864 [20:29:27<20:38:42, 14.88s/it]
 49%|████▉     | 4869/9864 [20:29:41<20:25:49, 14.72s/it]
 49%|████▉     | 4870/9864 [20:30:00<22:17:41, 16.07s/it]
                                                         
{'loss': '0.1316', 'grad_norm': '4.616', 'learning_rate': '1.069e-05', 'epoch': '0.4938'}

 49%|████▉     | 4870/9864 [20:30:00<22:17:41, 16.07s/it]
 49%|████▉     | 4871/9864 [20:30:14<21:35:09, 15.56s/it]
 49%|████▉     | 4872/9864 [20:30:28<20:53:25, 15.07s/it]
 49%|████▉     | 4873/9864 [20:30:42<20:28:15, 14.77s/it]
 49%|████▉     | 4874/9864 [20:30:57<20:10:43, 14.56s/it]
 49%|████▉     | 4875/9864 [20:31:11<20:09:47, 14.55s/it]
 49%|████▉     | 4876/9864 [20:31:27<20:42:59, 14.95s/it]
 49%|████▉     | 4877/9864 [20:31:43<21:16:11, 15.35s/it]
 49%|████▉     | 4878/9864 [20:32:01<22:08:48, 15.99s/it]
 49%|████▉     | 4879/9864 [20:32:15<21:38:11, 15.63s/it]
 49%|████▉     | 4880/9864 [20:32:29<20:54:29, 15.10s/it]
                                                         
{'loss': '0.1409', 'grad_norm': '2.285', 'learning_rate': '1.066e-05', 'epoch': '0.4948'}

 49%|████▉     | 4880/9864 [20:32:29<20:54:29, 15.10s/it]
 49%|████▉     | 4881/9864 [20:32:44<20:43:26, 14.97s/it]
 49%|████▉     | 4882/9864 [20:32:59<20:42:02, 14.96s/it]
 50%|████▉     | 4883/9864 [20:33:13<20:28:01, 14.79s/it]
 50%|████▉     | 4884/9864 [20:33:28<20:11:52, 14.60s/it]
 50%|████▉     | 4885/9864 [20:33:46<21:58:41, 15.89s/it]
 50%|████▉     | 4886/9864 [20:34:00<21:12:49, 15.34s/it]
 50%|████▉     | 4887/9864 [20:34:16<21:16:22, 15.39s/it]
 50%|████▉     | 4888/9864 [20:34:31<21:10:36, 15.32s/it]
 50%|████▉     | 4889/9864 [20:34:47<21:25:07, 15.50s/it]
 50%|████▉     | 4890/9864 [20:35:02<21:01:38, 15.22s/it]
                                                         
{'loss': '0.1394', 'grad_norm': '6.528', 'learning_rate': '1.063e-05', 'epoch': '0.4958'}

 50%|████▉     | 4890/9864 [20:35:02<21:01:38, 15.22s/it]
 50%|████▉     | 4891/9864 [20:35:16<20:41:50, 14.98s/it]
 50%|████▉     | 4892/9864 [20:35:34<22:05:56, 16.00s/it]
 50%|████▉     | 4893/9864 [20:35:49<21:22:02, 15.47s/it]
 50%|████▉     | 4894/9864 [20:36:05<21:43:55, 15.74s/it]
 50%|████▉     | 4895/9864 [20:36:19<21:04:58, 15.27s/it]
 50%|████▉     | 4896/9864 [20:36:35<21:15:26, 15.40s/it]
 50%|████▉     | 4897/9864 [20:36:50<21:02:32, 15.25s/it]
 50%|████▉     | 4898/9864 [20:37:04<20:44:56, 15.04s/it]
 50%|████▉     | 4899/9864 [20:37:20<20:55:09, 15.17s/it]
 50%|████▉     | 4900/9864 [20:37:35<21:02:40, 15.26s/it]
                                                         
{'loss': '0.1357', 'grad_norm': '3.981', 'learning_rate': '1.059e-05', 'epoch': '0.4968'}

 50%|████▉     | 4900/9864 [20:37:35<21:02:40, 15.26s/it]
 50%|████▉     | 4901/9864 [20:37:51<21:18:46, 15.46s/it]
 50%|████▉     | 4902/9864 [20:38:05<20:35:17, 14.94s/it]
 50%|████▉     | 4903/9864 [20:38:19<20:05:05, 14.57s/it]
 50%|████▉     | 4904/9864 [20:38:35<20:43:38, 15.04s/it]
 50%|████▉     | 4905/9864 [20:38:49<20:28:13, 14.86s/it]
 50%|████▉     | 4906/9864 [20:39:03<19:59:43, 14.52s/it]
 50%|████▉     | 4907/9864 [20:39:19<20:33:00, 14.92s/it]
 50%|████▉     | 4908/9864 [20:39:36<21:35:54, 15.69s/it]
 50%|████▉     | 4909/9864 [20:39:51<21:05:55, 15.33s/it]
 50%|████▉     | 4910/9864 [20:40:05<20:37:55, 14.99s/it]
                                                         
{'loss': '0.1482', 'grad_norm': '2.011', 'learning_rate': '1.056e-05', 'epoch': '0.4978'}

 50%|████▉     | 4910/9864 [20:40:05<20:37:55, 14.99s/it]
 50%|████▉     | 4911/9864 [20:40:21<21:04:12, 15.31s/it]
 50%|████▉     | 4912/9864 [20:40:34<20:15:23, 14.73s/it]
 50%|████▉     | 4913/9864 [20:40:49<20:08:49, 14.65s/it]
 50%|████▉     | 4914/9864 [20:41:05<20:35:22, 14.97s/it]
 50%|████▉     | 4915/9864 [20:41:21<21:03:12, 15.31s/it]
 50%|████▉     | 4916/9864 [20:41:37<21:19:40, 15.52s/it]
 50%|████▉     | 4917/9864 [20:41:51<20:50:02, 15.16s/it]
 50%|████▉     | 4918/9864 [20:42:06<20:35:48, 14.99s/it]
 50%|████▉     | 4919/9864 [20:42:21<20:51:29, 15.18s/it]
 50%|████▉     | 4920/9864 [20:42:36<20:30:53, 14.94s/it]
                                                         
{'loss': '0.1407', 'grad_norm': '2.094', 'learning_rate': '1.053e-05', 'epoch': '0.4988'}

 50%|████▉     | 4920/9864 [20:42:36<20:30:53, 14.94s/it]
 50%|████▉     | 4921/9864 [20:42:50<20:06:50, 14.65s/it]
 50%|████▉     | 4922/9864 [20:43:04<19:53:38, 14.49s/it]
 50%|████▉     | 4923/9864 [20:43:22<21:18:56, 15.53s/it]
 50%|████▉     | 4924/9864 [20:43:36<20:47:14, 15.15s/it]
 50%|████▉     | 4925/9864 [20:43:52<21:05:34, 15.37s/it]
 50%|████▉     | 4926/9864 [20:44:06<20:40:22, 15.07s/it]
 50%|████▉     | 4927/9864 [20:44:22<21:01:08, 15.33s/it]
 50%|████▉     | 4928/9864 [20:44:36<20:29:52, 14.95s/it]
 50%|████▉     | 4929/9864 [20:44:51<20:14:30, 14.77s/it]
 50%|████▉     | 4930/9864 [20:45:06<20:39:57, 15.08s/it]
                                                         
{'loss': '0.1369', 'grad_norm': '2.686', 'learning_rate': '1.05e-05', 'epoch': '0.4998'}

 50%|████▉     | 4930/9864 [20:45:06<20:39:57, 15.08s/it]
 50%|████▉     | 4931/9864 [20:45:22<20:54:12, 15.25s/it]
 50%|█████     | 4932/9864 [20:45:38<21:18:57, 15.56s/it]
 50%|█████     | 4933/9864 [20:45:52<20:30:48, 14.98s/it]
 50%|█████     | 4934/9864 [20:46:06<19:56:19, 14.56s/it]
 50%|█████     | 4935/9864 [20:46:21<20:22:50, 14.89s/it]
 50%|█████     | 4936/9864 [20:46:35<20:06:20, 14.69s/it]
 50%|█████     | 4937/9864 [20:46:49<19:50:28, 14.50s/it]
 50%|█████     | 4938/9864 [20:47:03<19:36:28, 14.33s/it]
 50%|█████     | 4939/9864 [20:47:21<20:58:17, 15.33s/it]
 50%|█████     | 4940/9864 [20:47:37<21:21:38, 15.62s/it]
                                                         
{'loss': '0.1263', 'grad_norm': '3.243', 'learning_rate': '1.046e-05', 'epoch': '0.5009'}

 50%|█████     | 4940/9864 [20:47:37<21:21:38, 15.62s/it]
 50%|█████     | 4941/9864 [20:47:51<20:43:09, 15.15s/it]
 50%|█████     | 4942/9864 [20:48:06<20:29:04, 14.98s/it]
 50%|█████     | 4943/9864 [20:48:22<20:51:37, 15.26s/it]
 50%|█████     | 4944/9864 [20:48:36<20:23:37, 14.92s/it]
 50%|█████     | 4945/9864 [20:48:50<20:04:32, 14.69s/it]
 50%|█████     | 4946/9864 [20:49:06<20:30:21, 15.01s/it]
 50%|█████     | 4947/9864 [20:49:23<21:08:07, 15.47s/it]
 50%|█████     | 4948/9864 [20:49:39<21:25:38, 15.69s/it]
 50%|█████     | 4949/9864 [20:49:53<20:50:51, 15.27s/it]
 50%|█████     | 4950/9864 [20:50:09<21:04:30, 15.44s/it]
                                                         
{'loss': '0.1261', 'grad_norm': '2.4', 'learning_rate': '1.043e-05', 'epoch': '0.5019'}

 50%|█████     | 4950/9864 [20:50:09<21:04:30, 15.44s/it]
 50%|█████     | 4951/9864 [20:50:23<20:27:32, 14.99s/it]
 50%|█████     | 4952/9864 [20:50:37<20:12:09, 14.81s/it]
 50%|█████     | 4953/9864 [20:50:51<19:55:27, 14.61s/it]
 50%|█████     | 4954/9864 [20:51:07<20:27:44, 15.00s/it]
 50%|█████     | 4955/9864 [20:51:23<20:36:08, 15.11s/it]
 50%|█████     | 4956/9864 [20:51:39<21:00:52, 15.41s/it]
 50%|█████     | 4957/9864 [20:51:55<21:15:40, 15.60s/it]
 50%|█████     | 4958/9864 [20:52:08<20:27:35, 15.01s/it]
 50%|█████     | 4959/9864 [20:52:22<20:00:18, 14.68s/it]
 50%|█████     | 4960/9864 [20:52:36<19:45:57, 14.51s/it]
                                                         
{'loss': '0.1286', 'grad_norm': '2.637', 'learning_rate': '1.04e-05', 'epoch': '0.5029'}

 50%|█████     | 4960/9864 [20:52:36<19:45:57, 14.51s/it]
 50%|█████     | 4961/9864 [20:52:50<19:31:07, 14.33s/it]
 50%|█████     | 4962/9864 [20:53:06<20:12:56, 14.85s/it]
 50%|█████     | 4963/9864 [20:53:22<20:37:54, 15.15s/it]
 50%|█████     | 4964/9864 [20:53:38<21:00:19, 15.43s/it]
 50%|█████     | 4965/9864 [20:53:54<21:09:37, 15.55s/it]
 50%|█████     | 4966/9864 [20:54:08<20:32:54, 15.10s/it]
 50%|█████     | 4967/9864 [20:54:23<20:17:16, 14.91s/it]
 50%|█████     | 4968/9864 [20:54:37<19:59:11, 14.70s/it]
 50%|█████     | 4969/9864 [20:54:51<19:44:40, 14.52s/it]
 50%|█████     | 4970/9864 [20:55:07<20:11:14, 14.85s/it]
                                                         
{'loss': '0.1384', 'grad_norm': '2.582', 'learning_rate': '1.036e-05', 'epoch': '0.5039'}

 50%|█████     | 4970/9864 [20:55:07<20:11:14, 14.85s/it]
 50%|█████     | 4971/9864 [20:55:22<20:23:50, 15.01s/it]
 50%|█████     | 4972/9864 [20:55:38<20:45:38, 15.28s/it]
 50%|█████     | 4973/9864 [20:55:54<21:09:24, 15.57s/it]
 50%|█████     | 4974/9864 [20:56:08<20:36:59, 15.18s/it]
 50%|█████     | 4975/9864 [20:56:23<20:13:02, 14.89s/it]
 50%|█████     | 4976/9864 [20:56:37<19:52:37, 14.64s/it]
 50%|█████     | 4977/9864 [20:56:51<19:37:36, 14.46s/it]
 50%|█████     | 4978/9864 [20:57:06<20:00:04, 14.74s/it]
 50%|█████     | 4979/9864 [20:57:23<21:00:00, 15.48s/it]
 50%|█████     | 4980/9864 [20:57:37<20:20:44, 15.00s/it]
                                                         
{'loss': '0.1477', 'grad_norm': '3.758', 'learning_rate': '1.033e-05', 'epoch': '0.5049'}

 50%|█████     | 4980/9864 [20:57:37<20:20:44, 15.00s/it]
 50%|█████     | 4981/9864 [20:57:54<20:57:05, 15.45s/it]
 51%|█████     | 4982/9864 [20:58:08<20:23:59, 15.04s/it]
 51%|█████     | 4983/9864 [20:58:22<19:55:43, 14.70s/it]
 51%|█████     | 4984/9864 [20:58:36<19:41:00, 14.52s/it]
 51%|█████     | 4985/9864 [20:58:50<19:25:01, 14.33s/it]
 51%|█████     | 4986/9864 [20:59:07<20:38:19, 15.23s/it]
 51%|█████     | 4987/9864 [20:59:22<20:43:11, 15.29s/it]
 51%|█████     | 4988/9864 [20:59:38<20:53:17, 15.42s/it]
 51%|█████     | 4989/9864 [20:59:53<20:32:28, 15.17s/it]
 51%|█████     | 4990/9864 [21:00:06<19:51:56, 14.67s/it]
                                                         
{'loss': '0.1303', 'grad_norm': '1.728', 'learning_rate': '1.03e-05', 'epoch': '0.5059'}

 51%|█████     | 4990/9864 [21:00:06<19:51:56, 14.67s/it]
 51%|█████     | 4991/9864 [21:00:21<19:46:22, 14.61s/it]
 51%|█████     | 4992/9864 [21:00:35<19:33:34, 14.45s/it]
 51%|█████     | 4993/9864 [21:00:48<19:12:28, 14.20s/it]
 51%|█████     | 4994/9864 [21:01:06<20:36:21, 15.23s/it]
 51%|█████     | 4995/9864 [21:01:23<21:30:31, 15.90s/it]
 51%|█████     | 4996/9864 [21:01:37<20:31:44, 15.18s/it]
 51%|█████     | 4997/9864 [21:01:51<19:56:35, 14.75s/it]
 51%|█████     | 4998/9864 [21:02:05<19:52:11, 14.70s/it]
 51%|█████     | 4999/9864 [21:02:20<19:41:48, 14.58s/it]
 51%|█████     | 5000/9864 [21:02:34<19:30:25, 14.44s/it]
                                                         
{'loss': '0.1492', 'grad_norm': '3.559', 'learning_rate': '1.027e-05', 'epoch': '0.5069'}

 51%|█████     | 5000/9864 [21:02:34<19:30:25, 14.44s/it]

Writing model shards:   0%|          | 0/1 [00:00<?, ?it/s]

Writing model shards: 100%|██████████| 1/1 [00:08<00:00,  8.94s/it]
Writing model shards: 100%|██████████| 1/1 [00:08<00:00,  8.94s/it]
/home/dungpq6/Project/SpatialStack/src/qwen_vl/model/geometry_encoders/vggt_encoder.py:285: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(dtype=dtype):

 51%|█████     | 5001/9864 [21:03:53<45:48:30, 33.91s/it]
 51%|█████     | 5002/9864 [21:04:09<38:23:21, 28.42s/it]
 51%|█████     | 5003/9864 [21:04:25<33:29:31, 24.80s/it]
 51%|█████     | 5004/9864 [21:04:41<29:49:35, 22.09s/it]
 51%|█████     | 5005/9864 [21:04:55<26:30:32, 19.64s/it]
 51%|█████     | 5006/9864 [21:05:09<24:18:27, 18.01s/it]
 51%|█████     | 5007/9864 [21:05:23<22:46:46, 16.88s/it]
 51%|█████     | 5008/9864 [21:05:37<21:42:12, 16.09s/it]
 51%|█████     | 5009/9864 [21:05:52<20:58:12, 15.55s/it]
 51%|█████     | 5010/9864 [21:06:11<22:22:56, 16.60s/it]
                                                         
{'loss': '0.1289', 'grad_norm': '1.765', 'learning_rate': '1.023e-05', 'epoch': '0.5079'}

 51%|█████     | 5010/9864 [21:06:11<22:22:56, 16.60s/it]
 51%|█████     | 5011/9864 [21:06:25<21:23:49, 15.87s/it]
 51%|█████     | 5012/9864 [21:06:41<21:30:09, 15.95s/it]
 51%|█████     | 5013/9864 [21:06:55<20:47:28, 15.43s/it]
 51%|█████     | 5014/9864 [21:07:10<20:29:30, 15.21s/it]
 51%|█████     | 5015/9864 [21:07:24<19:54:49, 14.78s/it]
 51%|█████     | 5016/9864 [21:07:38<19:42:12, 14.63s/it]
 51%|█████     | 5017/9864 [21:07:55<20:26:13, 15.18s/it]
 51%|█████     | 5018/9864 [21:08:10<20:41:44, 15.37s/it]
 51%|█████     | 5019/9864 [21:08:26<20:57:28, 15.57s/it]
 51%|█████     | 5020/9864 [21:08:42<21:04:13, 15.66s/it]
                                                         
{'loss': '0.1341', 'grad_norm': '6.098', 'learning_rate': '1.02e-05', 'epoch': '0.509'}

 51%|█████     | 5020/9864 [21:08:42<21:04:13, 15.66s/it]
 51%|█████     | 5021/9864 [21:08:56<20:28:12, 15.22s/it]
 51%|█████     | 5022/9864 [21:09:11<20:01:11, 14.88s/it]
 51%|█████     | 5023/9864 [21:09:24<19:30:50, 14.51s/it]
 51%|█████     | 5024/9864 [21:09:39<19:34:05, 14.55s/it]
 51%|█████     | 5025/9864 [21:09:57<20:51:14, 15.51s/it]
 51%|█████     | 5026/9864 [21:10:10<20:10:09, 15.01s/it]
 51%|█████     | 5027/9864 [21:10:26<20:24:41, 15.19s/it]
 51%|█████     | 5028/9864 [21:10:42<20:46:32, 15.47s/it]
 51%|█████     | 5029/9864 [21:10:56<20:12:42, 15.05s/it]
 51%|█████     | 5030/9864 [21:11:11<20:02:24, 14.92s/it]
                                                         
{'loss': '0.1273', 'grad_norm': '3.11', 'learning_rate': '1.017e-05', 'epoch': '0.51'}

 51%|█████     | 5030/9864 [21:11:11<20:02:24, 14.92s/it]
 51%|█████     | 5031/9864 [21:11:25<19:43:16, 14.69s/it]
 51%|█████     | 5032/9864 [21:11:39<19:31:40, 14.55s/it]
 51%|█████     | 5033/9864 [21:11:58<21:09:29, 15.77s/it]
 51%|█████     | 5034/9864 [21:12:14<21:16:31, 15.86s/it]
 51%|█████     | 5035/9864 [21:12:30<21:21:20, 15.92s/it]
 51%|█████     | 5036/9864 [21:12:44<20:38:05, 15.39s/it]
 51%|█████     | 5037/9864 [21:12:58<20:10:23, 15.05s/it]
 51%|█████     | 5038/9864 [21:13:13<19:54:07, 14.85s/it]
 51%|█████     | 5039/9864 [21:13:27<19:43:05, 14.71s/it]
 51%|█████     | 5040/9864 [21:13:43<20:18:06, 15.15s/it]
                                                         
{'loss': '0.1373', 'grad_norm': '3.912', 'learning_rate': '1.013e-05', 'epoch': '0.511'}

 51%|█████     | 5040/9864 [21:13:43<20:18:06, 15.15s/it]
 51%|█████     | 5041/9864 [21:13:59<20:33:57, 15.35s/it]
 51%|█████     | 5042/9864 [21:14:15<20:34:48, 15.36s/it]
 51%|█████     | 5043/9864 [21:14:28<19:53:06, 14.85s/it]
 51%|█████     | 5044/9864 [21:14:44<20:14:55, 15.12s/it]
 51%|█████     | 5045/9864 [21:14:58<19:55:42, 14.89s/it]
 51%|█████     | 5046/9864 [21:15:12<19:33:47, 14.62s/it]
 51%|█████     | 5047/9864 [21:15:26<19:14:18, 14.38s/it]
 51%|█████     | 5048/9864 [21:15:42<19:51:59, 14.85s/it]
 51%|█████     | 5049/9864 [21:15:59<20:51:06, 15.59s/it]
 51%|█████     | 5050/9864 [21:16:14<20:24:06, 15.26s/it]
                                                         
{'loss': '0.1398', 'grad_norm': '3.035', 'learning_rate': '1.01e-05', 'epoch': '0.512'}

 51%|█████     | 5050/9864 [21:16:14<20:24:06, 15.26s/it]
 51%|█████     | 5051/9864 [21:16:28<20:05:00, 15.02s/it]
 51%|█████     | 5052/9864 [21:16:44<20:19:12, 15.20s/it]
 51%|█████     | 5053/9864 [21:16:58<19:59:45, 14.96s/it]
 51%|█████     | 5054/9864 [21:17:12<19:35:41, 14.67s/it]
 51%|█████     | 5055/9864 [21:17:26<19:09:03, 14.34s/it]
 51%|█████▏    | 5056/9864 [21:17:42<19:48:16, 14.83s/it]
 51%|█████▏    | 5057/9864 [21:18:00<20:56:51, 15.69s/it]
 51%|█████▏    | 5058/9864 [21:18:13<20:14:13, 15.16s/it]
 51%|█████▏    | 5059/9864 [21:18:29<20:30:44, 15.37s/it]
 51%|█████▏    | 5060/9864 [21:18:44<20:09:33, 15.11s/it]
                                                         
{'loss': '0.1273', 'grad_norm': '4.241', 'learning_rate': '1.007e-05', 'epoch': '0.513'}

 51%|█████▏    | 5060/9864 [21:18:44<20:09:33, 15.11s/it]
 51%|█████▏    | 5061/9864 [21:18:58<19:40:30, 14.75s/it]
 51%|█████▏    | 5062/9864 [21:19:12<19:22:45, 14.53s/it]
 51%|█████▏    | 5063/9864 [21:19:26<19:14:58, 14.43s/it]
 51%|█████▏    | 5064/9864 [21:19:40<19:12:24, 14.41s/it]
 51%|█████▏    | 5065/9864 [21:19:58<20:35:50, 15.45s/it]
 51%|█████▏    | 5066/9864 [21:20:14<20:47:18, 15.60s/it]
 51%|█████▏    | 5067/9864 [21:20:28<20:12:14, 15.16s/it]
 51%|█████▏    | 5068/9864 [21:20:44<20:27:29, 15.36s/it]
 51%|█████▏    | 5069/9864 [21:20:58<19:58:51, 15.00s/it]
 51%|█████▏    | 5070/9864 [21:21:12<19:38:10, 14.75s/it]
                                                         
{'loss': '0.1334', 'grad_norm': '2.949', 'learning_rate': '1.004e-05', 'epoch': '0.514'}

 51%|█████▏    | 5070/9864 [21:21:12<19:38:10, 14.75s/it]
 51%|█████▏    | 5071/9864 [21:21:27<19:26:45, 14.61s/it]
 51%|█████▏    | 5072/9864 [21:21:43<20:05:37, 15.10s/it]
 51%|█████▏    | 5073/9864 [21:21:59<20:34:26, 15.46s/it]
 51%|█████▏    | 5074/9864 [21:22:15<20:43:59, 15.58s/it]
 51%|█████▏    | 5075/9864 [21:22:31<20:54:30, 15.72s/it]
 51%|█████▏    | 5076/9864 [21:22:46<20:25:27, 15.36s/it]
 51%|█████▏    | 5077/9864 [21:23:00<20:02:54, 15.08s/it]
 51%|█████▏    | 5078/9864 [21:23:15<20:03:51, 15.09s/it]
 51%|█████▏    | 5079/9864 [21:23:31<20:18:19, 15.28s/it]
 52%|█████▏    | 5080/9864 [21:23:47<20:26:45, 15.39s/it]
                                                         
{'loss': '0.1369', 'grad_norm': '4.239', 'learning_rate': '1e-05', 'epoch': '0.515'}

 52%|█████▏    | 5080/9864 [21:23:47<20:26:45, 15.39s/it]
 52%|█████▏    | 5081/9864 [21:24:03<20:44:02, 15.61s/it]
 52%|█████▏    | 5082/9864 [21:24:17<20:13:56, 15.23s/it]
 52%|█████▏    | 5083/9864 [21:24:34<20:43:59, 15.61s/it]
 52%|█████▏    | 5084/9864 [21:24:48<20:09:54, 15.19s/it]
 52%|█████▏    | 5085/9864 [21:25:02<19:53:44, 14.99s/it]
 52%|█████▏    | 5086/9864 [21:25:16<19:27:30, 14.66s/it]
 52%|█████▏    | 5087/9864 [21:25:32<19:53:39, 14.99s/it]
 52%|█████▏    | 5088/9864 [21:25:47<20:04:11, 15.13s/it]
 52%|█████▏    | 5089/9864 [21:26:03<20:19:49, 15.33s/it]
 52%|█████▏    | 5090/9864 [21:26:19<20:41:47, 15.61s/it]
                                                         
{'loss': '0.1361', 'grad_norm': '2.966', 'learning_rate': '9.97e-06', 'epoch': '0.5161'}

 52%|█████▏    | 5090/9864 [21:26:19<20:41:47, 15.61s/it]
 52%|█████▏    | 5091/9864 [21:26:33<20:00:33, 15.09s/it]
 52%|█████▏    | 5092/9864 [21:26:48<19:41:31, 14.86s/it]
 52%|█████▏    | 5093/9864 [21:27:02<19:22:17, 14.62s/it]
 52%|█████▏    | 5094/9864 [21:27:16<19:10:44, 14.47s/it]
 52%|█████▏    | 5095/9864 [21:27:31<19:34:22, 14.78s/it]
 52%|█████▏    | 5096/9864 [21:27:47<20:01:40, 15.12s/it]
 52%|█████▏    | 5097/9864 [21:28:03<20:10:01, 15.23s/it]
 52%|█████▏    | 5098/9864 [21:28:19<20:24:27, 15.42s/it]
 52%|█████▏    | 5099/9864 [21:28:32<19:40:41, 14.87s/it]
 52%|█████▏    | 5100/9864 [21:28:47<19:31:45, 14.76s/it]
                                                         
{'loss': '0.1272', 'grad_norm': '2.652', 'learning_rate': '9.938e-06', 'epoch': '0.5171'}

 52%|█████▏    | 5100/9864 [21:28:47<19:31:45, 14.76s/it]
 52%|█████▏    | 5101/9864 [21:29:01<19:17:36, 14.58s/it]
 52%|█████▏    | 5102/9864 [21:29:17<19:51:28, 15.01s/it]
 52%|█████▏    | 5103/9864 [21:29:31<19:31:26, 14.76s/it]
 52%|█████▏    | 5104/9864 [21:29:47<20:06:44, 15.21s/it]
 52%|█████▏    | 5105/9864 [21:30:05<21:10:07, 16.01s/it]
 52%|█████▏    | 5106/9864 [21:30:19<20:19:13, 15.37s/it]
 52%|█████▏    | 5107/9864 [21:30:34<19:58:34, 15.12s/it]
 52%|█████▏    | 5108/9864 [21:30:47<19:29:16, 14.75s/it]
 52%|█████▏    | 5109/9864 [21:31:03<19:47:38, 14.99s/it]
 52%|█████▏    | 5110/9864 [21:31:17<19:25:16, 14.71s/it]
                                                         
{'loss': '0.1407', 'grad_norm': '2.992', 'learning_rate': '9.905e-06', 'epoch': '0.5181'}

 52%|█████▏    | 5110/9864 [21:31:17<19:25:16, 14.71s/it]
 52%|█████▏    | 5111/9864 [21:31:33<19:57:49, 15.12s/it]
 52%|█████▏    | 5112/9864 [21:31:48<19:40:52, 14.91s/it]
 52%|█████▏    | 5113/9864 [21:32:03<20:04:06, 15.21s/it]
 52%|█████▏    | 5114/9864 [21:32:20<20:40:17, 15.67s/it]
 52%|█████▏    | 5115/9864 [21:32:35<20:08:50, 15.27s/it]
 52%|█████▏    | 5116/9864 [21:32:49<19:39:18, 14.90s/it]
 52%|█████▏    | 5117/9864 [21:33:04<19:57:53, 15.14s/it]
 52%|█████▏    | 5118/9864 [21:33:19<19:39:53, 14.92s/it]
 52%|█████▏    | 5119/9864 [21:33:33<19:28:49, 14.78s/it]
 52%|█████▏    | 5120/9864 [21:33:51<20:51:50, 15.83s/it]
                                                         
{'loss': '0.1343', 'grad_norm': '1.702', 'learning_rate': '9.872e-06', 'epoch': '0.5191'}

 52%|█████▏    | 5120/9864 [21:33:51<20:51:50, 15.83s/it]
 52%|█████▏    | 5121/9864 [21:34:07<20:49:29, 15.81s/it]
 52%|█████▏    | 5122/9864 [21:34:21<20:06:08, 15.26s/it]
 52%|█████▏    | 5123/9864 [21:34:35<19:44:21, 14.99s/it]
 52%|█████▏    | 5124/9864 [21:34:50<19:27:37, 14.78s/it]
 52%|█████▏    | 5125/9864 [21:35:05<19:45:12, 15.01s/it]
 52%|█████▏    | 5126/9864 [21:35:20<19:31:22, 14.83s/it]
 52%|█████▏    | 5127/9864 [21:35:34<19:19:17, 14.68s/it]
 52%|█████▏    | 5128/9864 [21:35:50<19:47:42, 15.05s/it]
 52%|█████▏    | 5129/9864 [21:36:07<20:45:31, 15.78s/it]
 52%|█████▏    | 5130/9864 [21:36:22<20:11:31, 15.36s/it]
                                                         
{'loss': '0.1316', 'grad_norm': '2.134', 'learning_rate': '9.839e-06', 'epoch': '0.5201'}

 52%|█████▏    | 5130/9864 [21:36:22<20:11:31, 15.36s/it]
 52%|█████▏    | 5131/9864 [21:36:36<19:35:44, 14.90s/it]
 52%|█████▏    | 5132/9864 [21:36:51<19:36:23, 14.92s/it]
 52%|█████▏    | 5133/9864 [21:37:06<19:48:09, 15.07s/it]
 52%|█████▏    | 5134/9864 [21:37:20<19:32:51, 14.88s/it]
 52%|█████▏    | 5135/9864 [21:37:35<19:24:42, 14.78s/it]
 52%|█████▏    | 5136/9864 [21:37:51<20:00:12, 15.23s/it]
 52%|█████▏    | 5137/9864 [21:38:05<19:26:19, 14.80s/it]
 52%|█████▏    | 5138/9864 [21:38:20<19:17:14, 14.69s/it]
 52%|█████▏    | 5139/9864 [21:38:33<18:57:52, 14.45s/it]
 52%|█████▏    | 5140/9864 [21:38:48<18:52:43, 14.39s/it]
                                                         
{'loss': '0.1409', 'grad_norm': '3.561', 'learning_rate': '9.806e-06', 'epoch': '0.5211'}

 52%|█████▏    | 5140/9864 [21:38:48<18:52:43, 14.39s/it]
 52%|█████▏    | 5141/9864 [21:39:01<18:38:49, 14.21s/it]
 52%|█████▏    | 5142/9864 [21:39:17<19:05:31, 14.56s/it]
 52%|█████▏    | 5143/9864 [21:39:32<19:30:45, 14.88s/it]
 52%|█████▏    | 5144/9864 [21:39:49<20:02:07, 15.28s/it]
 52%|█████▏    | 5145/9864 [21:40:05<20:22:41, 15.55s/it]
 52%|█████▏    | 5146/9864 [21:40:19<19:43:53, 15.06s/it]
 52%|█████▏    | 5147/9864 [21:40:33<19:22:49, 14.79s/it]
 52%|█████▏    | 5148/9864 [21:40:47<19:08:12, 14.61s/it]
 52%|█████▏    | 5149/9864 [21:41:04<19:49:58, 15.14s/it]
 52%|█████▏    | 5150/9864 [21:41:19<20:02:30, 15.31s/it]
                                                         
{'loss': '0.1287', 'grad_norm': '3.841', 'learning_rate': '9.773e-06', 'epoch': '0.5221'}

 52%|█████▏    | 5150/9864 [21:41:19<20:02:30, 15.31s/it]
 52%|█████▏    | 5151/9864 [21:41:33<19:31:57, 14.92s/it]
 52%|█████▏    | 5152/9864 [21:41:49<19:45:58, 15.10s/it]
 52%|█████▏    | 5153/9864 [21:42:05<20:05:49, 15.36s/it]
 52%|█████▏    | 5154/9864 [21:42:19<19:48:08, 15.14s/it]
 52%|█████▏    | 5155/9864 [21:42:34<19:25:35, 14.85s/it]
 52%|█████▏    | 5156/9864 [21:42:48<19:07:31, 14.62s/it]
 52%|█████▏    | 5157/9864 [21:43:04<19:37:31, 15.01s/it]
 52%|█████▏    | 5158/9864 [21:43:20<20:01:40, 15.32s/it]
 52%|█████▏    | 5159/9864 [21:43:34<19:44:24, 15.10s/it]
 52%|█████▏    | 5160/9864 [21:43:51<20:13:44, 15.48s/it]
                                                         
{'loss': '0.1433', 'grad_norm': '3.775', 'learning_rate': '9.741e-06', 'epoch': '0.5232'}

 52%|█████▏    | 5160/9864 [21:43:51<20:13:44, 15.48s/it]
 52%|█████▏    | 5161/9864 [21:44:06<20:20:43, 15.57s/it]
 52%|█████▏    | 5162/9864 [21:44:21<19:49:49, 15.18s/it]
 52%|█████▏    | 5163/9864 [21:44:35<19:22:02, 14.83s/it]
 52%|█████▏    | 5164/9864 [21:44:48<19:00:15, 14.56s/it]
 52%|█████▏    | 5165/9864 [21:45:04<19:27:56, 14.91s/it]
 52%|█████▏    | 5166/9864 [21:45:18<19:10:35, 14.69s/it]
 52%|█████▏    | 5167/9864 [21:45:34<19:21:42, 14.84s/it]
 52%|█████▏    | 5168/9864 [21:45:47<18:56:50, 14.53s/it]
 52%|█████▏    | 5169/9864 [21:46:05<20:13:38, 15.51s/it]
 52%|█████▏    | 5170/9864 [21:46:19<19:43:45, 15.13s/it]
                                                         
{'loss': '0.1289', 'grad_norm': '2.878', 'learning_rate': '9.708e-06', 'epoch': '0.5242'}

 52%|█████▏    | 5170/9864 [21:46:19<19:43:45, 15.13s/it]
 52%|█████▏    | 5171/9864 [21:46:34<19:30:30, 14.96s/it]
 52%|█████▏    | 5172/9864 [21:46:48<19:08:13, 14.68s/it]
 52%|█████▏    | 5173/9864 [21:47:02<18:53:17, 14.50s/it]
 52%|█████▏    | 5174/9864 [21:47:18<19:24:57, 14.90s/it]
 52%|█████▏    | 5175/9864 [21:47:34<19:51:34, 15.25s/it]
 52%|█████▏    | 5176/9864 [21:47:48<19:25:21, 14.91s/it]
 52%|█████▏    | 5177/9864 [21:48:06<20:29:08, 15.73s/it]
 52%|█████▏    | 5178/9864 [21:48:20<19:54:06, 15.29s/it]
 53%|█████▎    | 5179/9864 [21:48:34<19:22:30, 14.89s/it]
 53%|█████▎    | 5180/9864 [21:48:48<19:10:48, 14.74s/it]
                                                         
{'loss': '0.1303', 'grad_norm': '3.133', 'learning_rate': '9.675e-06', 'epoch': '0.5252'}

 53%|█████▎    | 5180/9864 [21:48:48<19:10:48, 14.74s/it]
 53%|█████▎    | 5181/9864 [21:49:03<19:05:49, 14.68s/it]
 53%|█████▎    | 5182/9864 [21:49:19<19:39:44, 15.12s/it]
 53%|█████▎    | 5183/9864 [21:49:35<19:59:14, 15.37s/it]
 53%|█████▎    | 5184/9864 [21:49:52<20:29:27, 15.76s/it]
 53%|█████▎    | 5185/9864 [21:50:08<20:34:48, 15.83s/it]
 53%|█████▎    | 5186/9864 [21:50:22<19:47:46, 15.23s/it]
 53%|█████▎    | 5187/9864 [21:50:36<19:30:21, 15.01s/it]
 53%|█████▎    | 5188/9864 [21:50:50<19:14:50, 14.82s/it]
 53%|█████▎    | 5189/9864 [21:51:05<19:04:03, 14.68s/it]
 53%|█████▎    | 5190/9864 [21:51:21<19:48:41, 15.26s/it]
                                                         
{'loss': '0.1388', 'grad_norm': '3.253', 'learning_rate': '9.642e-06', 'epoch': '0.5262'}

 53%|█████▎    | 5190/9864 [21:51:21<19:48:41, 15.26s/it]
 53%|█████▎    | 5191/9864 [21:51:39<20:42:47, 15.96s/it]
 53%|█████▎    | 5192/9864 [21:51:53<19:56:26, 15.37s/it]
 53%|█████▎    | 5193/9864 [21:52:09<20:17:00, 15.63s/it]
 53%|█████▎    | 5194/9864 [21:52:23<19:33:06, 15.07s/it]
 53%|█████▎    | 5195/9864 [21:52:37<19:12:54, 14.82s/it]
 53%|█████▎    | 5196/9864 [21:52:51<19:00:08, 14.65s/it]
 53%|█████▎    | 5197/9864 [21:53:07<19:24:05, 14.97s/it]
 53%|█████▎    | 5198/9864 [21:53:22<19:11:52, 14.81s/it]
 53%|█████▎    | 5199/9864 [21:53:39<20:13:18, 15.61s/it]
 53%|█████▎    | 5200/9864 [21:53:55<20:15:36, 15.64s/it]
                                                         
{'loss': '0.1452', 'grad_norm': '2.436', 'learning_rate': '9.609e-06', 'epoch': '0.5272'}

 53%|█████▎    | 5200/9864 [21:53:55<20:15:36, 15.64s/it]
 53%|█████▎    | 5201/9864 [21:54:09<19:38:42, 15.17s/it]
 53%|█████▎    | 5202/9864 [21:54:23<19:07:55, 14.77s/it]
 53%|█████▎    | 5203/9864 [21:54:37<18:50:05, 14.55s/it]
 53%|█████▎    | 5204/9864 [21:54:50<18:27:42, 14.26s/it]
 53%|█████▎    | 5205/9864 [21:55:06<19:06:41, 14.77s/it]
 53%|█████▎    | 5206/9864 [21:55:20<18:49:25, 14.55s/it]
 53%|█████▎    | 5207/9864 [21:55:38<20:00:11, 15.46s/it]
 53%|█████▎    | 5208/9864 [21:55:54<20:13:49, 15.64s/it]
 53%|█████▎    | 5209/9864 [21:56:08<19:38:45, 15.19s/it]
 53%|█████▎    | 5210/9864 [21:56:22<19:13:00, 14.86s/it]
                                                         
{'loss': '0.1288', 'grad_norm': '1.944', 'learning_rate': '9.577e-06', 'epoch': '0.5282'}

 53%|█████▎    | 5210/9864 [21:56:22<19:13:00, 14.86s/it]
 53%|█████▎    | 5211/9864 [21:56:37<19:01:07, 14.71s/it]
 53%|█████▎    | 5212/9864 [21:56:50<18:39:27, 14.44s/it]
 53%|█████▎    | 5213/9864 [21:57:06<19:15:44, 14.91s/it]
 53%|█████▎    | 5214/9864 [21:57:23<19:48:14, 15.33s/it]
 53%|█████▎    | 5215/9864 [21:57:37<19:14:58, 14.91s/it]
 53%|█████▎    | 5216/9864 [21:57:53<19:52:33, 15.39s/it]
 53%|█████▎    | 5217/9864 [21:58:07<19:19:16, 14.97s/it]
 53%|█████▎    | 5218/9864 [21:58:21<18:59:02, 14.71s/it]
 53%|█████▎    | 5219/9864 [21:58:35<18:28:44, 14.32s/it]
 53%|█████▎    | 5220/9864 [21:58:50<18:54:12, 14.65s/it]
                                                         
{'loss': '0.1373', 'grad_norm': '2.796', 'learning_rate': '9.544e-06', 'epoch': '0.5292'}

 53%|█████▎    | 5220/9864 [21:58:50<18:54:12, 14.65s/it]
 53%|█████▎    | 5221/9864 [21:59:04<18:41:25, 14.49s/it]
 53%|█████▎    | 5222/9864 [21:59:20<19:11:38, 14.89s/it]
 53%|█████▎    | 5223/9864 [21:59:36<19:36:53, 15.22s/it]
 53%|█████▎    | 5224/9864 [21:59:51<19:21:33, 15.02s/it]
 53%|█████▎    | 5225/9864 [22:00:06<19:39:58, 15.26s/it]
 53%|█████▎    | 5226/9864 [22:00:21<19:15:20, 14.95s/it]
 53%|█████▎    | 5227/9864 [22:00:37<19:46:22, 15.35s/it]
 53%|█████▎    | 5228/9864 [22:00:51<19:27:45, 15.11s/it]
 53%|█████▎    | 5229/9864 [22:01:06<19:03:44, 14.81s/it]
 53%|█████▎    | 5230/9864 [22:01:22<19:31:10, 15.16s/it]
                                                         
{'loss': '0.1284', 'grad_norm': '2.715', 'learning_rate': '9.511e-06', 'epoch': '0.5303'}

 53%|█████▎    | 5230/9864 [22:01:22<19:31:10, 15.16s/it]
 53%|█████▎    | 5231/9864 [22:01:37<19:32:59, 15.19s/it]
 53%|█████▎    | 5232/9864 [22:01:51<19:16:04, 14.98s/it]
 53%|█████▎    | 5233/9864 [22:02:07<19:37:52, 15.26s/it]
 53%|█████▎    | 5234/9864 [22:02:21<19:13:20, 14.95s/it]
 53%|█████▎    | 5235/9864 [22:02:35<18:51:20, 14.66s/it]
 53%|█████▎    | 5236/9864 [22:02:51<19:19:00, 15.03s/it]
 53%|█████▎    | 5237/9864 [22:03:07<19:43:30, 15.35s/it]
 53%|█████▎    | 5238/9864 [22:03:21<19:13:16, 14.96s/it]
 53%|█████▎    | 5239/9864 [22:03:38<19:45:26, 15.38s/it]
 53%|█████▎    | 5240/9864 [22:03:54<19:58:26, 15.55s/it]
                                                         
{'loss': '0.1468', 'grad_norm': '2.408', 'learning_rate': '9.478e-06', 'epoch': '0.5313'}

 53%|█████▎    | 5240/9864 [22:03:54<19:58:26, 15.55s/it]
 53%|█████▎    | 5241/9864 [22:04:08<19:30:42, 15.19s/it]
 53%|█████▎    | 5242/9864 [22:04:23<19:15:22, 15.00s/it]
 53%|█████▎    | 5243/9864 [22:04:37<18:50:38, 14.68s/it]
 53%|█████▎    | 5244/9864 [22:04:53<19:31:41, 15.22s/it]
 53%|█████▎    | 5245/9864 [22:05:08<19:30:54, 15.21s/it]
 53%|█████▎    | 5246/9864 [22:05:23<19:10:57, 14.95s/it]
 53%|█████▎    | 5247/9864 [22:05:39<19:51:29, 15.48s/it]
 53%|█████▎    | 5248/9864 [22:05:56<20:09:11, 15.72s/it]
 53%|█████▎    | 5249/9864 [22:06:10<19:38:42, 15.32s/it]
 53%|█████▎    | 5250/9864 [22:06:24<19:17:33, 15.05s/it]
                                                         
{'loss': '0.1343', 'grad_norm': '2.484', 'learning_rate': '9.445e-06', 'epoch': '0.5323'}

 53%|█████▎    | 5250/9864 [22:06:24<19:17:33, 15.05s/it]
 53%|█████▎    | 5251/9864 [22:06:38<18:51:57, 14.72s/it]
 53%|█████▎    | 5252/9864 [22:06:54<19:05:10, 14.90s/it]
 53%|█████▎    | 5253/9864 [22:07:07<18:27:33, 14.41s/it]
 53%|█████▎    | 5254/9864 [22:07:23<19:09:17, 14.96s/it]
 53%|█████▎    | 5255/9864 [22:07:41<20:11:15, 15.77s/it]
 53%|█████▎    | 5256/9864 [22:07:55<19:33:18, 15.28s/it]
 53%|█████▎    | 5257/9864 [22:08:09<19:05:29, 14.92s/it]
 53%|█████▎    | 5258/9864 [22:08:23<18:52:49, 14.76s/it]
 53%|█████▎    | 5259/9864 [22:08:37<18:29:40, 14.46s/it]
 53%|█████▎    | 5260/9864 [22:08:54<19:14:26, 15.04s/it]
                                                         
{'loss': '0.1368', 'grad_norm': '3.722', 'learning_rate': '9.413e-06', 'epoch': '0.5333'}

 53%|█████▎    | 5260/9864 [22:08:54<19:14:26, 15.04s/it]
 53%|█████▎    | 5261/9864 [22:09:07<18:45:11, 14.67s/it]
 53%|█████▎    | 5262/9864 [22:09:26<20:04:48, 15.71s/it]
 53%|█████▎    | 5263/9864 [22:09:41<20:02:47, 15.69s/it]
 53%|█████▎    | 5264/9864 [22:09:55<19:20:03, 15.13s/it]
 53%|█████▎    | 5265/9864 [22:10:09<18:44:13, 14.67s/it]
 53%|█████▎    | 5266/9864 [22:10:23<18:29:55, 14.48s/it]
 53%|█████▎    | 5267/9864 [22:10:37<18:22:53, 14.39s/it]
 53%|█████▎    | 5268/9864 [22:10:53<18:55:12, 14.82s/it]
 53%|█████▎    | 5269/9864 [22:11:07<18:35:01, 14.56s/it]
 53%|█████▎    | 5270/9864 [22:11:26<20:31:15, 16.08s/it]
                                                         
{'loss': '0.1467', 'grad_norm': '2.543', 'learning_rate': '9.38e-06', 'epoch': '0.5343'}

 53%|█████▎    | 5270/9864 [22:11:26<20:31:15, 16.08s/it]
 53%|█████▎    | 5271/9864 [22:11:41<20:02:15, 15.71s/it]
 53%|█████▎    | 5272/9864 [22:11:55<19:33:24, 15.33s/it]
 53%|█████▎    | 5273/9864 [22:12:09<19:00:25, 14.90s/it]
 53%|█████▎    | 5274/9864 [22:12:24<18:44:07, 14.69s/it]
 53%|█████▎    | 5275/9864 [22:12:38<18:35:01, 14.58s/it]
 53%|█████▎    | 5276/9864 [22:12:54<19:17:14, 15.13s/it]
 53%|█████▎    | 5277/9864 [22:13:08<18:50:08, 14.78s/it]
 54%|█████▎    | 5278/9864 [22:13:26<20:02:37, 15.73s/it]
 54%|█████▎    | 5279/9864 [22:13:42<20:12:32, 15.87s/it]
 54%|█████▎    | 5280/9864 [22:13:57<19:38:03, 15.42s/it]
                                                         
{'loss': '0.1358', 'grad_norm': '2.207', 'learning_rate': '9.347e-06', 'epoch': '0.5353'}

 54%|█████▎    | 5280/9864 [22:13:57<19:38:03, 15.42s/it]
 54%|█████▎    | 5281/9864 [22:14:11<19:10:37, 15.06s/it]
 54%|█████▎    | 5282/9864 [22:14:26<19:00:15, 14.93s/it]
 54%|█████▎    | 5283/9864 [22:14:42<19:30:55, 15.34s/it]
 54%|█████▎    | 5284/9864 [22:14:56<19:01:19, 14.95s/it]
 54%|█████▎    | 5285/9864 [22:15:12<19:28:45, 15.31s/it]
 54%|█████▎    | 5286/9864 [22:15:28<19:42:19, 15.50s/it]
 54%|█████▎    | 5287/9864 [22:15:44<19:57:52, 15.70s/it]
 54%|█████▎    | 5288/9864 [22:15:58<19:23:26, 15.25s/it]
 54%|█████▎    | 5289/9864 [22:16:12<18:44:46, 14.75s/it]
 54%|█████▎    | 5290/9864 [22:16:28<19:15:52, 15.16s/it]
                                                         
{'loss': '0.1354', 'grad_norm': '2.456', 'learning_rate': '9.314e-06', 'epoch': '0.5363'}

 54%|█████▎    | 5290/9864 [22:16:28<19:15:52, 15.16s/it]
 54%|█████▎    | 5291/9864 [22:16:42<18:53:48, 14.88s/it]
 54%|█████▎    | 5292/9864 [22:16:57<18:37:39, 14.67s/it]
 54%|█████▎    | 5293/9864 [22:17:14<19:52:02, 15.65s/it]
 54%|█████▎    | 5294/9864 [22:17:31<20:00:46, 15.77s/it]
 54%|█████▎    | 5295/9864 [22:17:45<19:25:30, 15.31s/it]
 54%|█████▎    | 5296/9864 [22:17:59<19:10:24, 15.11s/it]
 54%|█████▎    | 5297/9864 [22:18:14<19:02:46, 15.01s/it]
 54%|█████▎    | 5298/9864 [22:18:30<19:23:36, 15.29s/it]
 54%|█████▎    | 5299/9864 [22:18:45<19:05:59, 15.06s/it]
 54%|█████▎    | 5300/9864 [22:19:00<19:22:42, 15.29s/it]
                                                         
{'loss': '0.1305', 'grad_norm': '2.156', 'learning_rate': '9.282e-06', 'epoch': '0.5374'}

 54%|█████▎    | 5300/9864 [22:19:00<19:22:42, 15.29s/it]
 54%|█████▎    | 5301/9864 [22:19:17<19:44:50, 15.58s/it]
 54%|█████▍    | 5302/9864 [22:19:32<19:45:38, 15.59s/it]
 54%|█████▍    | 5303/9864 [22:19:47<19:17:35, 15.23s/it]
 54%|█████▍    | 5304/9864 [22:20:01<18:52:22, 14.90s/it]
 54%|█████▍    | 5305/9864 [22:20:15<18:35:52, 14.69s/it]
 54%|█████▍    | 5306/9864 [22:20:31<19:04:09, 15.06s/it]
 54%|█████▍    | 5307/9864 [22:20:47<19:34:04, 15.46s/it]
 54%|█████▍    | 5308/9864 [22:21:01<18:51:43, 14.90s/it]
 54%|█████▍    | 5309/9864 [22:21:17<19:17:10, 15.24s/it]
 54%|█████▍    | 5310/9864 [22:21:34<19:44:43, 15.61s/it]
                                                         
{'loss': '0.1314', 'grad_norm': '2.633', 'learning_rate': '9.249e-06', 'epoch': '0.5384'}

 54%|█████▍    | 5310/9864 [22:21:34<19:44:43, 15.61s/it]
 54%|█████▍    | 5311/9864 [22:21:48<19:11:27, 15.17s/it]
 54%|█████▍    | 5312/9864 [22:22:02<18:58:32, 15.01s/it]
 54%|█████▍    | 5313/9864 [22:22:17<18:50:29, 14.90s/it]
 54%|█████▍    | 5314/9864 [22:22:33<19:13:38, 15.21s/it]
 54%|█████▍    | 5315/9864 [22:22:49<19:27:24, 15.40s/it]
 54%|█████▍    | 5316/9864 [22:23:05<19:39:05, 15.56s/it]
 54%|█████▍    | 5317/9864 [22:23:19<19:04:04, 15.10s/it]
 54%|█████▍    | 5318/9864 [22:23:35<19:28:05, 15.42s/it]
 54%|█████▍    | 5319/9864 [22:23:49<19:07:52, 15.15s/it]
 54%|█████▍    | 5320/9864 [22:24:04<18:50:43, 14.93s/it]
                                                         
{'loss': '0.1346', 'grad_norm': '5.04', 'learning_rate': '9.216e-06', 'epoch': '0.5394'}

 54%|█████▍    | 5320/9864 [22:24:04<18:50:43, 14.93s/it]
 54%|█████▍    | 5321/9864 [22:24:20<19:22:38, 15.36s/it]
 54%|█████▍    | 5322/9864 [22:24:35<19:14:18, 15.25s/it]
 54%|█████▍    | 5323/9864 [22:24:49<18:52:12, 14.96s/it]
 54%|█████▍    | 5324/9864 [22:25:05<19:13:23, 15.24s/it]
 54%|█████▍    | 5325/9864 [22:25:20<19:00:15, 15.07s/it]
 54%|█████▍    | 5326/9864 [22:25:37<19:33:46, 15.52s/it]
 54%|█████▍    | 5327/9864 [22:25:51<18:59:05, 15.06s/it]
 54%|█████▍    | 5328/9864 [22:26:05<18:37:35, 14.78s/it]
 54%|█████▍    | 5329/9864 [22:26:20<18:52:32, 14.98s/it]
 54%|█████▍    | 5330/9864 [22:26:35<18:59:18, 15.08s/it]
                                                         
{'loss': '0.1423', 'grad_norm': '3.199', 'learning_rate': '9.183e-06', 'epoch': '0.5404'}

 54%|█████▍    | 5330/9864 [22:26:35<18:59:18, 15.08s/it]
 54%|█████▍    | 5331/9864 [22:26:50<18:45:08, 14.89s/it]
 54%|█████▍    | 5332/9864 [22:27:06<19:07:10, 15.19s/it]
 54%|█████▍    | 5333/9864 [22:27:19<18:33:50, 14.75s/it]
 54%|█████▍    | 5334/9864 [22:27:34<18:20:36, 14.58s/it]
 54%|█████▍    | 5335/9864 [22:27:50<18:49:12, 14.96s/it]
 54%|█████▍    | 5336/9864 [22:28:04<18:45:35, 14.92s/it]
 54%|█████▍    | 5337/9864 [22:28:20<19:04:34, 15.17s/it]
 54%|█████▍    | 5338/9864 [22:28:36<19:19:03, 15.37s/it]
 54%|█████▍    | 5339/9864 [22:28:50<18:53:00, 15.02s/it]
 54%|█████▍    | 5340/9864 [22:29:06<19:15:00, 15.32s/it]
                                                         
{'loss': '0.1372', 'grad_norm': '2.682', 'learning_rate': '9.151e-06', 'epoch': '0.5414'}

 54%|█████▍    | 5340/9864 [22:29:06<19:15:00, 15.32s/it]
 54%|█████▍    | 5341/9864 [22:29:20<18:39:18, 14.85s/it]
 54%|█████▍    | 5342/9864 [22:29:35<18:54:46, 15.06s/it]
 54%|█████▍    | 5343/9864 [22:29:49<18:31:17, 14.75s/it]
 54%|█████▍    | 5344/9864 [22:30:03<18:11:54, 14.49s/it]
 54%|█████▍    | 5345/9864 [22:30:19<18:46:02, 14.95s/it]
 54%|█████▍    | 5346/9864 [22:30:35<18:51:17, 15.02s/it]
 54%|█████▍    | 5347/9864 [22:30:49<18:33:51, 14.80s/it]
 54%|█████▍    | 5348/9864 [22:31:03<18:22:04, 14.64s/it]
 54%|█████▍    | 5349/9864 [22:31:19<18:48:29, 15.00s/it]
 54%|█████▍    | 5350/9864 [22:31:33<18:36:40, 14.84s/it]
                                                         
{'loss': '0.1282', 'grad_norm': '3.894', 'learning_rate': '9.118e-06', 'epoch': '0.5424'}

 54%|█████▍    | 5350/9864 [22:31:33<18:36:40, 14.84s/it]
 54%|█████▍    | 5351/9864 [22:31:50<19:07:10, 15.25s/it]
 54%|█████▍    | 5352/9864 [22:32:04<18:51:09, 15.04s/it]
 54%|█████▍    | 5353/9864 [22:32:22<19:50:35, 15.84s/it]
 54%|█████▍    | 5354/9864 [22:32:36<19:20:34, 15.44s/it]
 54%|█████▍    | 5355/9864 [22:32:51<18:52:39, 15.07s/it]
 54%|█████▍    | 5356/9864 [22:33:07<19:11:00, 15.32s/it]
 54%|█████▍    | 5357/9864 [22:33:21<18:51:49, 15.07s/it]
 54%|█████▍    | 5358/9864 [22:33:37<19:14:55, 15.38s/it]
 54%|█████▍    | 5359/9864 [22:33:52<18:54:56, 15.12s/it]
 54%|█████▍    | 5360/9864 [22:34:09<19:56:47, 15.94s/it]
                                                         
{'loss': '0.1203', 'grad_norm': '2.476', 'learning_rate': '9.085e-06', 'epoch': '0.5434'}

 54%|█████▍    | 5360/9864 [22:34:09<19:56:47, 15.94s/it]
 54%|█████▍    | 5361/9864 [22:34:24<19:24:39, 15.52s/it]
 54%|█████▍    | 5362/9864 [22:34:38<18:55:15, 15.13s/it]
 54%|█████▍    | 5363/9864 [22:34:52<18:26:24, 14.75s/it]
 54%|█████▍    | 5364/9864 [22:35:08<18:57:20, 15.16s/it]
 54%|█████▍    | 5365/9864 [22:35:23<18:41:22, 14.96s/it]
 54%|█████▍    | 5366/9864 [22:35:38<18:55:20, 15.14s/it]
 54%|█████▍    | 5367/9864 [22:35:54<19:18:23, 15.46s/it]
 54%|█████▍    | 5368/9864 [22:36:11<19:33:39, 15.66s/it]
 54%|█████▍    | 5369/9864 [22:36:24<18:40:32, 14.96s/it]
 54%|█████▍    | 5370/9864 [22:36:38<18:14:38, 14.61s/it]
                                                         
{'loss': '0.1278', 'grad_norm': '3.561', 'learning_rate': '9.053e-06', 'epoch': '0.5444'}

 54%|█████▍    | 5370/9864 [22:36:38<18:14:38, 14.61s/it]
 54%|█████▍    | 5371/9864 [22:36:52<18:04:55, 14.49s/it]
 54%|█████▍    | 5372/9864 [22:37:08<18:42:10, 14.99s/it]
 54%|█████▍    | 5373/9864 [22:37:22<18:20:51, 14.71s/it]
 54%|█████▍    | 5374/9864 [22:37:38<18:43:57, 15.02s/it]
 54%|█████▍    | 5375/9864 [22:37:55<19:25:52, 15.58s/it]
 55%|█████▍    | 5376/9864 [22:38:09<18:46:29, 15.06s/it]
 55%|█████▍    | 5377/9864 [22:38:23<18:29:39, 14.84s/it]
 55%|█████▍    | 5378/9864 [22:38:37<18:10:49, 14.59s/it]
 55%|█████▍    | 5379/9864 [22:38:52<18:13:51, 14.63s/it]
 55%|█████▍    | 5380/9864 [22:39:07<18:35:51, 14.93s/it]
                                                         
{'loss': '0.1283', 'grad_norm': '1.651', 'learning_rate': '9.02e-06', 'epoch': '0.5455'}

 55%|█████▍    | 5380/9864 [22:39:07<18:35:51, 14.93s/it]
 55%|█████▍    | 5381/9864 [22:39:22<18:23:23, 14.77s/it]
 55%|█████▍    | 5382/9864 [22:39:39<19:20:41, 15.54s/it]
 55%|█████▍    | 5383/9864 [22:39:55<19:27:53, 15.64s/it]
 55%|█████▍    | 5384/9864 [22:40:09<18:57:30, 15.23s/it]
 55%|█████▍    | 5385/9864 [22:40:23<18:29:52, 14.87s/it]
 55%|█████▍    | 5386/9864 [22:40:37<18:12:18, 14.64s/it]
 55%|█████▍    | 5387/9864 [22:40:52<18:03:44, 14.52s/it]
 55%|█████▍    | 5388/9864 [22:41:07<18:28:04, 14.85s/it]
 55%|█████▍    | 5389/9864 [22:41:25<19:35:00, 15.75s/it]
 55%|█████▍    | 5390/9864 [22:41:39<18:59:40, 15.28s/it]
                                                         
{'loss': '0.1327', 'grad_norm': '3.509', 'learning_rate': '8.987e-06', 'epoch': '0.5465'}

 55%|█████▍    | 5390/9864 [22:41:39<18:59:40, 15.28s/it]
 55%|█████▍    | 5391/9864 [22:41:53<18:32:47, 14.93s/it]
 55%|█████▍    | 5392/9864 [22:42:09<18:49:39, 15.16s/it]
 55%|█████▍    | 5393/9864 [22:42:23<18:24:26, 14.82s/it]
 55%|█████▍    | 5394/9864 [22:42:37<18:12:17, 14.66s/it]
 55%|█████▍    | 5395/9864 [22:42:53<18:34:01, 14.96s/it]
 55%|█████▍    | 5396/9864 [22:43:09<18:57:52, 15.28s/it]
 55%|█████▍    | 5397/9864 [22:43:23<18:26:46, 14.87s/it]
 55%|█████▍    | 5398/9864 [22:43:39<18:43:36, 15.10s/it]
 55%|█████▍    | 5399/9864 [22:43:53<18:24:50, 14.85s/it]
 55%|█████▍    | 5400/9864 [22:44:09<18:51:22, 15.21s/it]
                                                         
{'loss': '0.1214', 'grad_norm': '2.93', 'learning_rate': '8.954e-06', 'epoch': '0.5475'}

 55%|█████▍    | 5400/9864 [22:44:09<18:51:22, 15.21s/it]
 55%|█████▍    | 5401/9864 [22:44:23<18:31:42, 14.95s/it]
 55%|█████▍    | 5402/9864 [22:44:37<18:15:48, 14.74s/it]
 55%|█████▍    | 5403/9864 [22:44:54<18:48:22, 15.18s/it]
 55%|█████▍    | 5404/9864 [22:45:10<19:09:30, 15.46s/it]
 55%|█████▍    | 5405/9864 [22:45:25<19:13:06, 15.52s/it]
 55%|█████▍    | 5406/9864 [22:45:40<18:41:08, 15.09s/it]
 55%|█████▍    | 5407/9864 [22:45:55<18:49:13, 15.20s/it]
 55%|█████▍    | 5408/9864 [22:46:09<18:32:53, 14.99s/it]
 55%|█████▍    | 5409/9864 [22:46:24<18:21:59, 14.84s/it]
 55%|█████▍    | 5410/9864 [22:46:38<18:07:01, 14.64s/it]
                                                         
{'loss': '0.1233', 'grad_norm': '3.651', 'learning_rate': '8.922e-06', 'epoch': '0.5485'}

 55%|█████▍    | 5410/9864 [22:46:38<18:07:01, 14.64s/it]
 55%|█████▍    | 5411/9864 [22:46:52<17:41:32, 14.30s/it]
 55%|█████▍    | 5412/9864 [22:47:11<19:34:10, 15.82s/it]
 55%|█████▍    | 5413/9864 [22:47:26<19:05:17, 15.44s/it]
 55%|█████▍    | 5414/9864 [22:47:40<18:41:24, 15.12s/it]
 55%|█████▍    | 5415/9864 [22:47:56<18:58:54, 15.36s/it]
 55%|█████▍    | 5416/9864 [22:48:10<18:31:11, 14.99s/it]
 55%|█████▍    | 5417/9864 [22:48:25<18:22:26, 14.87s/it]
 55%|█████▍    | 5418/9864 [22:48:39<18:06:19, 14.66s/it]
 55%|█████▍    | 5419/9864 [22:48:55<18:43:36, 15.17s/it]
 55%|█████▍    | 5420/9864 [22:49:13<19:43:40, 15.98s/it]
                                                         
{'loss': '0.1316', 'grad_norm': '2.782', 'learning_rate': '8.889e-06', 'epoch': '0.5495'}

 55%|█████▍    | 5420/9864 [22:49:13<19:43:40, 15.98s/it]
 55%|█████▍    | 5421/9864 [22:49:27<19:03:53, 15.45s/it]
 55%|█████▍    | 5422/9864 [22:49:41<18:34:20, 15.05s/it]
 55%|█████▍    | 5423/9864 [22:49:57<18:57:19, 15.37s/it]
 55%|█████▍    | 5424/9864 [22:50:11<18:26:38, 14.95s/it]
 55%|█████▍    | 5425/9864 [22:50:26<18:12:12, 14.76s/it]
 55%|█████▌    | 5426/9864 [22:50:40<18:01:40, 14.62s/it]
 55%|█████▌    | 5427/9864 [22:50:56<18:23:44, 14.93s/it]
 55%|█████▌    | 5428/9864 [22:51:12<18:47:12, 15.25s/it]
 55%|█████▌    | 5429/9864 [22:51:28<19:06:45, 15.51s/it]
 55%|█████▌    | 5430/9864 [22:51:44<19:27:21, 15.80s/it]
                                                         
{'loss': '0.1334', 'grad_norm': '4.684', 'learning_rate': '8.857e-06', 'epoch': '0.5505'}

 55%|█████▌    | 5430/9864 [22:51:44<19:27:21, 15.80s/it]
 55%|█████▌    | 5431/9864 [22:51:59<18:53:27, 15.34s/it]
 55%|█████▌    | 5432/9864 [22:52:13<18:25:21, 14.96s/it]
 55%|█████▌    | 5433/9864 [22:52:26<17:57:59, 14.60s/it]
 55%|█████▌    | 5434/9864 [22:52:40<17:36:48, 14.31s/it]
 55%|█████▌    | 5435/9864 [22:52:56<18:20:58, 14.92s/it]
 55%|█████▌    | 5436/9864 [22:53:14<19:32:37, 15.89s/it]
 55%|█████▌    | 5437/9864 [22:53:29<18:51:33, 15.34s/it]
 55%|█████▌    | 5438/9864 [22:53:42<18:16:00, 14.86s/it]
 55%|█████▌    | 5439/9864 [22:53:58<18:30:55, 15.06s/it]
 55%|█████▌    | 5440/9864 [22:54:12<18:17:57, 14.89s/it]
                                                         
{'loss': '0.1441', 'grad_norm': '3.076', 'learning_rate': '8.824e-06', 'epoch': '0.5515'}

 55%|█████▌    | 5440/9864 [22:54:12<18:17:57, 14.89s/it]
 55%|█████▌    | 5441/9864 [22:54:27<18:08:41, 14.77s/it]
 55%|█████▌    | 5442/9864 [22:54:41<18:04:12, 14.71s/it]
 55%|█████▌    | 5443/9864 [22:54:57<18:27:18, 15.03s/it]
 55%|█████▌    | 5444/9864 [22:55:13<18:41:08, 15.22s/it]
 55%|█████▌    | 5445/9864 [22:55:27<18:12:11, 14.83s/it]
 55%|█████▌    | 5446/9864 [22:55:42<18:22:51, 14.98s/it]
 55%|█████▌    | 5447/9864 [22:55:56<18:10:57, 14.82s/it]
 55%|█████▌    | 5448/9864 [22:56:11<17:57:00, 14.63s/it]
 55%|█████▌    | 5449/9864 [22:56:25<17:46:03, 14.49s/it]
 55%|█████▌    | 5450/9864 [22:56:39<17:42:51, 14.45s/it]
                                                         
{'loss': '0.1262', 'grad_norm': '3.874', 'learning_rate': '8.791e-06', 'epoch': '0.5526'}

 55%|█████▌    | 5450/9864 [22:56:39<17:42:51, 14.45s/it]
 55%|█████▌    | 5451/9864 [22:56:54<17:56:06, 14.63s/it]
 55%|█████▌    | 5452/9864 [22:57:11<18:43:41, 15.28s/it]
 55%|█████▌    | 5453/9864 [22:57:25<18:07:58, 14.80s/it]
 55%|█████▌    | 5454/9864 [22:57:41<18:29:41, 15.10s/it]
 55%|█████▌    | 5455/9864 [22:57:54<17:52:49, 14.60s/it]
 55%|█████▌    | 5456/9864 [22:58:08<17:44:31, 14.49s/it]
 55%|█████▌    | 5457/9864 [22:58:23<17:48:02, 14.54s/it]
 55%|█████▌    | 5458/9864 [22:58:37<17:39:01, 14.42s/it]
 55%|█████▌    | 5459/9864 [22:58:53<18:17:43, 14.95s/it]
 55%|█████▌    | 5460/9864 [22:59:11<19:19:44, 15.80s/it]
                                                         
{'loss': '0.1358', 'grad_norm': '3.078', 'learning_rate': '8.759e-06', 'epoch': '0.5536'}

 55%|█████▌    | 5460/9864 [22:59:11<19:19:44, 15.80s/it]
 55%|█████▌    | 5461/9864 [22:59:25<18:41:49, 15.29s/it]
 55%|█████▌    | 5462/9864 [22:59:41<18:59:45, 15.54s/it]
 55%|█████▌    | 5463/9864 [22:59:55<18:32:56, 15.17s/it]
 55%|█████▌    | 5464/9864 [23:00:10<18:07:48, 14.83s/it]
 55%|█████▌    | 5465/9864 [23:00:23<17:46:56, 14.55s/it]
 55%|█████▌    | 5466/9864 [23:00:39<18:10:36, 14.88s/it]
 55%|█████▌    | 5467/9864 [23:00:55<18:28:23, 15.12s/it]
 55%|█████▌    | 5468/9864 [23:01:11<18:50:37, 15.43s/it]
 55%|█████▌    | 5469/9864 [23:01:25<18:21:16, 15.03s/it]
 55%|█████▌    | 5470/9864 [23:01:41<18:35:52, 15.24s/it]
                                                         
{'loss': '0.1408', 'grad_norm': '2.461', 'learning_rate': '8.726e-06', 'epoch': '0.5546'}

 55%|█████▌    | 5470/9864 [23:01:41<18:35:52, 15.24s/it]
 55%|█████▌    | 5471/9864 [23:01:55<18:21:06, 15.04s/it]
 55%|█████▌    | 5472/9864 [23:02:09<17:52:04, 14.65s/it]
 55%|█████▌    | 5473/9864 [23:02:23<17:33:18, 14.39s/it]
 55%|█████▌    | 5474/9864 [23:02:39<18:18:36, 15.02s/it]
 56%|█████▌    | 5475/9864 [23:02:57<19:12:09, 15.75s/it]
 56%|█████▌    | 5476/9864 [23:03:11<18:36:38, 15.27s/it]
 56%|█████▌    | 5477/9864 [23:03:27<18:51:08, 15.47s/it]
 56%|█████▌    | 5478/9864 [23:03:41<18:14:03, 14.97s/it]
 56%|█████▌    | 5479/9864 [23:03:54<17:45:07, 14.57s/it]
 56%|█████▌    | 5480/9864 [23:04:08<17:31:10, 14.39s/it]
                                                         
{'loss': '0.1266', 'grad_norm': '2.964', 'learning_rate': '8.694e-06', 'epoch': '0.5556'}

 56%|█████▌    | 5480/9864 [23:04:08<17:31:10, 14.39s/it]
 56%|█████▌    | 5481/9864 [23:04:22<17:24:55, 14.30s/it]
 56%|█████▌    | 5482/9864 [23:04:39<18:08:21, 14.90s/it]
 56%|█████▌    | 5483/9864 [23:04:53<17:56:08, 14.74s/it]
 56%|█████▌    | 5484/9864 [23:05:12<19:36:32, 16.12s/it]
 56%|█████▌    | 5485/9864 [23:05:26<18:49:52, 15.48s/it]
 56%|█████▌    | 5486/9864 [23:05:40<18:19:07, 15.06s/it]
 56%|█████▌    | 5487/9864 [23:05:55<18:01:53, 14.83s/it]
 56%|█████▌    | 5488/9864 [23:06:09<17:46:22, 14.62s/it]
 56%|█████▌    | 5489/9864 [23:06:25<18:12:45, 14.99s/it]
 56%|█████▌    | 5490/9864 [23:06:39<17:55:42, 14.76s/it]
                                                         
{'loss': '0.1361', 'grad_norm': '2.266', 'learning_rate': '8.661e-06', 'epoch': '0.5566'}

 56%|█████▌    | 5490/9864 [23:06:39<17:55:42, 14.76s/it]
 56%|█████▌    | 5491/9864 [23:06:55<18:24:11, 15.15s/it]
 56%|█████▌    | 5492/9864 [23:07:12<19:14:22, 15.84s/it]
 56%|█████▌    | 5493/9864 [23:07:27<18:45:50, 15.45s/it]
 56%|█████▌    | 5494/9864 [23:07:42<18:27:10, 15.20s/it]
 56%|█████▌    | 5495/9864 [23:07:56<18:04:38, 14.90s/it]
 56%|█████▌    | 5496/9864 [23:08:10<17:46:52, 14.65s/it]
 56%|█████▌    | 5497/9864 [23:08:24<17:26:10, 14.37s/it]
 56%|█████▌    | 5498/9864 [23:08:42<18:50:22, 15.53s/it]
 56%|█████▌    | 5499/9864 [23:08:58<18:58:41, 15.65s/it]
 56%|█████▌    | 5500/9864 [23:09:13<19:00:05, 15.67s/it]
                                                         
{'loss': '0.1347', 'grad_norm': '2.947', 'learning_rate': '8.629e-06', 'epoch': '0.5576'}

 56%|█████▌    | 5500/9864 [23:09:13<19:00:05, 15.67s/it]
 56%|█████▌    | 5501/9864 [23:09:28<18:29:37, 15.26s/it]
 56%|█████▌    | 5502/9864 [23:09:42<18:00:30, 14.86s/it]
 56%|█████▌    | 5503/9864 [23:09:56<17:48:18, 14.70s/it]
 56%|█████▌    | 5504/9864 [23:10:10<17:27:35, 14.42s/it]
 56%|█████▌    | 5505/9864 [23:10:24<17:33:32, 14.50s/it]
 56%|█████▌    | 5506/9864 [23:10:40<18:00:23, 14.87s/it]
 56%|█████▌    | 5507/9864 [23:10:59<19:28:19, 16.09s/it]
 56%|█████▌    | 5508/9864 [23:11:14<18:52:04, 15.59s/it]
 56%|█████▌    | 5509/9864 [23:11:27<18:13:49, 15.07s/it]
 56%|█████▌    | 5510/9864 [23:11:42<18:07:27, 14.99s/it]
                                                         
{'loss': '0.1388', 'grad_norm': '4.114', 'learning_rate': '8.596e-06', 'epoch': '0.5586'}

 56%|█████▌    | 5510/9864 [23:11:42<18:07:27, 14.99s/it]
 56%|█████▌    | 5511/9864 [23:11:57<17:55:08, 14.82s/it]
 56%|█████▌    | 5512/9864 [23:12:11<17:40:23, 14.62s/it]
 56%|█████▌    | 5513/9864 [23:12:26<17:47:42, 14.72s/it]
 56%|█████▌    | 5514/9864 [23:12:41<18:03:10, 14.94s/it]
 56%|█████▌    | 5515/9864 [23:12:59<19:03:22, 15.77s/it]
 56%|█████▌    | 5516/9864 [23:13:15<18:59:17, 15.72s/it]
 56%|█████▌    | 5517/9864 [23:13:28<18:18:37, 15.16s/it]
 56%|█████▌    | 5518/9864 [23:13:43<18:01:38, 14.93s/it]
 56%|█████▌    | 5519/9864 [23:13:57<17:52:19, 14.81s/it]
 56%|█████▌    | 5520/9864 [23:14:12<17:44:25, 14.70s/it]
                                                         
{'loss': '0.1323', 'grad_norm': '4.622', 'learning_rate': '8.564e-06', 'epoch': '0.5597'}

 56%|█████▌    | 5520/9864 [23:14:12<17:44:25, 14.70s/it]
 56%|█████▌    | 5521/9864 [23:14:26<17:34:24, 14.57s/it]
 56%|█████▌    | 5522/9864 [23:14:44<18:43:12, 15.52s/it]
 56%|█████▌    | 5523/9864 [23:14:58<18:17:38, 15.17s/it]
 56%|█████▌    | 5524/9864 [23:15:16<19:10:51, 15.91s/it]
 56%|█████▌    | 5525/9864 [23:15:30<18:33:26, 15.40s/it]
 56%|█████▌    | 5526/9864 [23:15:44<18:01:03, 14.95s/it]
 56%|█████▌    | 5527/9864 [23:15:58<17:39:18, 14.65s/it]
 56%|█████▌    | 5528/9864 [23:16:12<17:33:47, 14.58s/it]
 56%|█████▌    | 5529/9864 [23:16:28<18:00:32, 14.96s/it]
 56%|█████▌    | 5530/9864 [23:16:43<17:52:36, 14.85s/it]
                                                         
{'loss': '0.1342', 'grad_norm': '2.691', 'learning_rate': '8.531e-06', 'epoch': '0.5607'}

 56%|█████▌    | 5530/9864 [23:16:43<17:52:36, 14.85s/it]
 56%|█████▌    | 5531/9864 [23:16:59<18:27:29, 15.34s/it]
 56%|█████▌    | 5532/9864 [23:17:15<18:40:24, 15.52s/it]
 56%|█████▌    | 5533/9864 [23:17:31<18:43:54, 15.57s/it]
 56%|█████▌    | 5534/9864 [23:17:46<18:27:21, 15.34s/it]
 56%|█████▌    | 5535/9864 [23:18:00<18:06:50, 15.06s/it]
 56%|█████▌    | 5536/9864 [23:18:16<18:25:07, 15.32s/it]
 56%|█████▌    | 5537/9864 [23:18:30<17:52:16, 14.87s/it]
 56%|█████▌    | 5538/9864 [23:18:44<17:32:44, 14.60s/it]
 56%|█████▌    | 5539/9864 [23:19:01<18:36:43, 15.49s/it]
 56%|█████▌    | 5540/9864 [23:19:15<18:06:36, 15.08s/it]
                                                         
{'loss': '0.1165', 'grad_norm': '2.983', 'learning_rate': '8.499e-06', 'epoch': '0.5617'}

 56%|█████▌    | 5540/9864 [23:19:15<18:06:36, 15.08s/it]
 56%|█████▌    | 5541/9864 [23:19:31<18:25:04, 15.34s/it]
 56%|█████▌    | 5542/9864 [23:19:46<18:09:24, 15.12s/it]
 56%|█████▌    | 5543/9864 [23:20:00<17:46:57, 14.82s/it]
 56%|█████▌    | 5544/9864 [23:20:16<18:00:56, 15.01s/it]
 56%|█████▌    | 5545/9864 [23:20:30<17:38:51, 14.71s/it]
 56%|█████▌    | 5546/9864 [23:20:46<18:07:12, 15.11s/it]
 56%|█████▌    | 5547/9864 [23:21:00<17:53:34, 14.92s/it]
 56%|█████▌    | 5548/9864 [23:21:15<18:04:12, 15.07s/it]
 56%|█████▋    | 5549/9864 [23:21:31<18:22:52, 15.34s/it]
 56%|█████▋    | 5550/9864 [23:21:46<17:56:59, 14.98s/it]
                                                         
{'loss': '0.1241', 'grad_norm': '2.936', 'learning_rate': '8.466e-06', 'epoch': '0.5627'}

 56%|█████▋    | 5550/9864 [23:21:46<17:56:59, 14.98s/it]
 56%|█████▋    | 5551/9864 [23:22:00<17:34:23, 14.67s/it]
 56%|█████▋    | 5552/9864 [23:22:15<17:58:23, 15.01s/it]
 56%|█████▋    | 5553/9864 [23:22:30<17:45:52, 14.83s/it]
 56%|█████▋    | 5554/9864 [23:22:46<18:05:49, 15.12s/it]
 56%|█████▋    | 5555/9864 [23:22:59<17:37:10, 14.72s/it]
 56%|█████▋    | 5556/9864 [23:23:17<18:37:52, 15.57s/it]
 56%|█████▋    | 5557/9864 [23:23:31<18:03:36, 15.10s/it]
 56%|█████▋    | 5558/9864 [23:23:45<17:52:30, 14.94s/it]
 56%|█████▋    | 5559/9864 [23:24:01<18:06:13, 15.14s/it]
 56%|█████▋    | 5560/9864 [23:24:15<17:45:28, 14.85s/it]
                                                         
{'loss': '0.1224', 'grad_norm': '4.407', 'learning_rate': '8.434e-06', 'epoch': '0.5637'}

 56%|█████▋    | 5560/9864 [23:24:15<17:45:28, 14.85s/it]
 56%|█████▋    | 5561/9864 [23:24:29<17:27:26, 14.61s/it]
 56%|█████▋    | 5562/9864 [23:24:46<18:04:01, 15.12s/it]
 56%|█████▋    | 5563/9864 [23:25:00<17:51:11, 14.94s/it]
 56%|█████▋    | 5564/9864 [23:25:16<18:07:42, 15.18s/it]
 56%|█████▋    | 5565/9864 [23:25:32<18:21:04, 15.37s/it]
 56%|█████▋    | 5566/9864 [23:25:46<17:56:32, 15.03s/it]
 56%|█████▋    | 5567/9864 [23:26:02<18:23:00, 15.40s/it]
 56%|█████▋    | 5568/9864 [23:26:17<18:05:15, 15.16s/it]
 56%|█████▋    | 5569/9864 [23:26:31<17:41:48, 14.83s/it]
 56%|█████▋    | 5570/9864 [23:26:47<18:11:16, 15.25s/it]
                                                         
{'loss': '0.1279', 'grad_norm': '2.8', 'learning_rate': '8.401e-06', 'epoch': '0.5647'}

 56%|█████▋    | 5570/9864 [23:26:47<18:11:16, 15.25s/it]
 56%|█████▋    | 5571/9864 [23:27:01<17:48:57, 14.94s/it]
 56%|█████▋    | 5572/9864 [23:27:17<18:11:02, 15.25s/it]
 56%|█████▋    | 5573/9864 [23:27:33<18:13:16, 15.29s/it]
 57%|█████▋    | 5574/9864 [23:27:49<18:28:49, 15.51s/it]
 57%|█████▋    | 5575/9864 [23:28:03<17:56:23, 15.06s/it]
 57%|█████▋    | 5576/9864 [23:28:17<17:39:14, 14.82s/it]
 57%|█████▋    | 5577/9864 [23:28:32<17:50:49, 14.99s/it]
 57%|█████▋    | 5578/9864 [23:28:46<17:30:41, 14.71s/it]
 57%|█████▋    | 5579/9864 [23:29:00<17:03:50, 14.34s/it]
 57%|█████▋    | 5580/9864 [23:29:18<18:22:47, 15.45s/it]
                                                         
{'loss': '0.1171', 'grad_norm': '2.423', 'learning_rate': '8.369e-06', 'epoch': '0.5657'}

 57%|█████▋    | 5580/9864 [23:29:18<18:22:47, 15.45s/it]
 57%|█████▋    | 5581/9864 [23:29:33<18:06:16, 15.22s/it]
 57%|█████▋    | 5582/9864 [23:29:47<17:42:04, 14.88s/it]
 57%|█████▋    | 5583/9864 [23:30:03<18:09:46, 15.27s/it]
 57%|█████▋    | 5584/9864 [23:30:17<17:47:43, 14.97s/it]
 57%|█████▋    | 5585/9864 [23:30:32<17:56:59, 15.10s/it]
 57%|█████▋    | 5586/9864 [23:30:47<17:39:37, 14.86s/it]
 57%|█████▋    | 5587/9864 [23:31:01<17:20:36, 14.60s/it]
 57%|█████▋    | 5588/9864 [23:31:15<17:15:57, 14.54s/it]
 57%|█████▋    | 5589/9864 [23:31:33<18:19:46, 15.44s/it]
 57%|█████▋    | 5590/9864 [23:31:49<18:37:34, 15.69s/it]
                                                         
{'loss': '0.1189', 'grad_norm': '3.121', 'learning_rate': '8.336e-06', 'epoch': '0.5668'}

 57%|█████▋    | 5590/9864 [23:31:49<18:37:34, 15.69s/it]
 57%|█████▋    | 5591/9864 [23:32:03<18:00:08, 15.17s/it]
 57%|█████▋    | 5592/9864 [23:32:19<18:17:16, 15.41s/it]
 57%|█████▋    | 5593/9864 [23:32:33<17:45:28, 14.97s/it]
 57%|█████▋    | 5594/9864 [23:32:47<17:23:55, 14.67s/it]
 57%|█████▋    | 5595/9864 [23:33:01<17:15:29, 14.55s/it]
 57%|█████▋    | 5596/9864 [23:33:15<17:09:01, 14.47s/it]
 57%|█████▋    | 5597/9864 [23:33:31<17:41:18, 14.92s/it]
 57%|█████▋    | 5598/9864 [23:33:49<18:44:11, 15.81s/it]
 57%|█████▋    | 5599/9864 [23:34:04<18:30:52, 15.63s/it]
 57%|█████▋    | 5600/9864 [23:34:19<18:03:48, 15.25s/it]
                                                         
{'loss': '0.134', 'grad_norm': '2.364', 'learning_rate': '8.304e-06', 'epoch': '0.5678'}

 57%|█████▋    | 5600/9864 [23:34:19<18:03:48, 15.25s/it]
 57%|█████▋    | 5601/9864 [23:34:33<17:46:23, 15.01s/it]
 57%|█████▋    | 5602/9864 [23:34:47<17:17:49, 14.61s/it]
 57%|█████▋    | 5603/9864 [23:35:01<17:14:10, 14.56s/it]
 57%|█████▋    | 5604/9864 [23:35:17<17:46:05, 15.02s/it]
 57%|█████▋    | 5605/9864 [23:35:32<17:33:16, 14.84s/it]
 57%|█████▋    | 5606/9864 [23:35:48<17:53:48, 15.13s/it]
 57%|█████▋    | 5607/9864 [23:36:03<18:02:47, 15.26s/it]
 57%|█████▋    | 5608/9864 [23:36:17<17:33:14, 14.85s/it]
 57%|█████▋    | 5609/9864 [23:36:31<17:13:24, 14.57s/it]
 57%|█████▋    | 5610/9864 [23:36:45<16:51:51, 14.27s/it]
                                                         
{'loss': '0.1258', 'grad_norm': '2.719', 'learning_rate': '8.272e-06', 'epoch': '0.5688'}

 57%|█████▋    | 5610/9864 [23:36:45<16:51:51, 14.27s/it]
 57%|█████▋    | 5611/9864 [23:36:59<16:48:42, 14.23s/it]
 57%|█████▋    | 5612/9864 [23:37:14<17:16:26, 14.63s/it]
 57%|█████▋    | 5613/9864 [23:37:29<17:07:12, 14.50s/it]
 57%|█████▋    | 5614/9864 [23:37:42<16:52:08, 14.29s/it]
 57%|█████▋    | 5615/9864 [23:38:02<18:48:39, 15.94s/it]
 57%|█████▋    | 5616/9864 [23:38:16<18:12:48, 15.44s/it]
 57%|█████▋    | 5617/9864 [23:38:31<17:48:32, 15.10s/it]
 57%|█████▋    | 5618/9864 [23:38:45<17:22:43, 14.73s/it]
 57%|█████▋    | 5619/9864 [23:38:59<17:07:04, 14.52s/it]
 57%|█████▋    | 5620/9864 [23:39:15<17:40:16, 14.99s/it]
                                                         
{'loss': '0.127', 'grad_norm': '2.785', 'learning_rate': '8.239e-06', 'epoch': '0.5698'}

 57%|█████▋    | 5620/9864 [23:39:15<17:40:16, 14.99s/it]
 57%|█████▋    | 5621/9864 [23:39:29<17:26:54, 14.80s/it]
 57%|█████▋    | 5622/9864 [23:39:46<18:18:49, 15.54s/it]
 57%|█████▋    | 5623/9864 [23:40:02<18:28:42, 15.69s/it]
 57%|█████▋    | 5624/9864 [23:40:17<17:58:23, 15.26s/it]
 57%|█████▋    | 5625/9864 [23:40:31<17:34:44, 14.93s/it]
 57%|█████▋    | 5626/9864 [23:40:45<17:30:25, 14.87s/it]
 57%|█████▋    | 5627/9864 [23:41:00<17:13:58, 14.64s/it]
 57%|█████▋    | 5628/9864 [23:41:16<17:50:29, 15.16s/it]
 57%|█████▋    | 5629/9864 [23:41:31<17:38:42, 15.00s/it]
 57%|█████▋    | 5630/9864 [23:41:49<18:42:51, 15.91s/it]
                                                         
{'loss': '0.1298', 'grad_norm': '3.798', 'learning_rate': '8.207e-06', 'epoch': '0.5708'}

 57%|█████▋    | 5630/9864 [23:41:49<18:42:51, 15.91s/it]
 57%|█████▋    | 5631/9864 [23:42:05<18:49:52, 16.02s/it]
 57%|█████▋    | 5632/9864 [23:42:19<17:59:34, 15.31s/it]
 57%|█████▋    | 5633/9864 [23:42:33<17:47:34, 15.14s/it]
 57%|█████▋    | 5634/9864 [23:42:47<17:20:12, 14.75s/it]
 57%|█████▋    | 5635/9864 [23:43:04<18:00:13, 15.33s/it]
 57%|█████▋    | 5636/9864 [23:43:17<17:24:13, 14.82s/it]
 57%|█████▋    | 5637/9864 [23:43:33<17:49:23, 15.18s/it]
 57%|█████▋    | 5638/9864 [23:43:51<18:38:01, 15.87s/it]
 57%|█████▋    | 5639/9864 [23:44:06<18:13:48, 15.53s/it]
 57%|█████▋    | 5640/9864 [23:44:20<17:52:56, 15.24s/it]
                                                         
{'loss': '0.129', 'grad_norm': '5.37', 'learning_rate': '8.175e-06', 'epoch': '0.5718'}

 57%|█████▋    | 5640/9864 [23:44:20<17:52:56, 15.24s/it]
 57%|█████▋    | 5641/9864 [23:44:35<17:44:19, 15.12s/it]
 57%|█████▋    | 5642/9864 [23:44:51<18:05:23, 15.42s/it]
 57%|█████▋    | 5643/9864 [23:45:06<17:53:33, 15.26s/it]
 57%|█████▋    | 5644/9864 [23:45:20<17:32:21, 14.96s/it]
 57%|█████▋    | 5645/9864 [23:45:36<17:51:11, 15.23s/it]
 57%|█████▋    | 5646/9864 [23:45:54<18:44:47, 16.00s/it]
 57%|█████▋    | 5647/9864 [23:46:08<17:59:19, 15.36s/it]
 57%|█████▋    | 5648/9864 [23:46:21<17:21:57, 14.83s/it]
 57%|█████▋    | 5649/9864 [23:46:38<17:47:33, 15.20s/it]
 57%|█████▋    | 5650/9864 [23:46:52<17:23:26, 14.86s/it]
                                                         
{'loss': '0.1197', 'grad_norm': '3.184', 'learning_rate': '8.143e-06', 'epoch': '0.5728'}

 57%|█████▋    | 5650/9864 [23:46:52<17:23:26, 14.86s/it]
 57%|█████▋    | 5651/9864 [23:47:05<16:54:58, 14.45s/it]
 57%|█████▋    | 5652/9864 [23:47:19<16:39:44, 14.24s/it]
 57%|█████▋    | 5653/9864 [23:47:35<17:10:02, 14.68s/it]
 57%|█████▋    | 5654/9864 [23:47:52<18:12:59, 15.58s/it]
 57%|█████▋    | 5655/9864 [23:48:07<17:53:30, 15.30s/it]
 57%|█████▋    | 5656/9864 [23:48:21<17:30:19, 14.98s/it]
 57%|█████▋    | 5657/9864 [23:48:37<17:49:55, 15.26s/it]
 57%|█████▋    | 5658/9864 [23:48:51<17:24:08, 14.90s/it]
 57%|█████▋    | 5659/9864 [23:49:05<17:12:46, 14.74s/it]
 57%|█████▋    | 5660/9864 [23:49:21<17:19:49, 14.84s/it]
                                                         
{'loss': '0.1144', 'grad_norm': '3.733', 'learning_rate': '8.11e-06', 'epoch': '0.5739'}

 57%|█████▋    | 5660/9864 [23:49:21<17:19:49, 14.84s/it]
 57%|█████▋    | 5661/9864 [23:49:36<17:34:04, 15.05s/it]
 57%|█████▋    | 5662/9864 [23:49:52<17:59:07, 15.41s/it]
 57%|█████▋    | 5663/9864 [23:50:07<17:36:23, 15.09s/it]
 57%|█████▋    | 5664/9864 [23:50:20<17:09:56, 14.71s/it]
 57%|█████▋    | 5665/9864 [23:50:35<17:04:18, 14.64s/it]
 57%|█████▋    | 5666/9864 [23:50:51<17:40:44, 15.16s/it]
 57%|█████▋    | 5667/9864 [23:51:05<17:12:29, 14.76s/it]
 57%|█████▋    | 5668/9864 [23:51:21<17:39:18, 15.15s/it]
 57%|█████▋    | 5669/9864 [23:51:39<18:34:01, 15.93s/it]
 57%|█████▋    | 5670/9864 [23:51:52<17:37:35, 15.13s/it]
                                                         
{'loss': '0.1084', 'grad_norm': '3.487', 'learning_rate': '8.078e-06', 'epoch': '0.5749'}

 57%|█████▋    | 5670/9864 [23:51:52<17:37:35, 15.13s/it]
 57%|█████▋    | 5671/9864 [23:52:06<17:13:59, 14.80s/it]
 58%|█████▊    | 5672/9864 [23:52:20<16:57:44, 14.57s/it]
 58%|█████▊    | 5673/9864 [23:52:36<17:21:05, 14.90s/it]
 58%|█████▊    | 5674/9864 [23:52:50<17:12:50, 14.79s/it]
 58%|█████▊    | 5675/9864 [23:53:06<17:23:22, 14.94s/it]
 58%|█████▊    | 5676/9864 [23:53:23<18:18:46, 15.74s/it]
 58%|█████▊    | 5677/9864 [23:53:38<17:45:14, 15.27s/it]
 58%|█████▊    | 5678/9864 [23:53:52<17:19:34, 14.90s/it]
 58%|█████▊    | 5679/9864 [23:54:06<17:02:34, 14.66s/it]
 58%|█████▊    | 5680/9864 [23:54:20<16:56:59, 14.58s/it]
                                                         
{'loss': '0.1161', 'grad_norm': '2.532', 'learning_rate': '8.046e-06', 'epoch': '0.5759'}

 58%|█████▊    | 5680/9864 [23:54:20<16:56:59, 14.58s/it]
 58%|█████▊    | 5681/9864 [23:54:36<17:26:14, 15.01s/it]
 58%|█████▊    | 5682/9864 [23:54:50<17:07:34, 14.74s/it]
 58%|█████▊    | 5683/9864 [23:55:06<17:29:21, 15.06s/it]
 58%|█████▊    | 5684/9864 [23:55:22<17:54:19, 15.42s/it]
 58%|█████▊    | 5685/9864 [23:55:38<18:01:21, 15.53s/it]
 58%|█████▊    | 5686/9864 [23:55:52<17:20:46, 14.95s/it]
 58%|█████▊    | 5687/9864 [23:56:06<17:14:32, 14.86s/it]
 58%|█████▊    | 5688/9864 [23:56:20<16:54:26, 14.58s/it]
 58%|█████▊    | 5689/9864 [23:56:36<17:20:21, 14.95s/it]
 58%|█████▊    | 5690/9864 [23:56:50<17:02:51, 14.70s/it]
                                                         
{'loss': '0.1261', 'grad_norm': '2.656', 'learning_rate': '8.014e-06', 'epoch': '0.5769'}

 58%|█████▊    | 5690/9864 [23:56:50<17:02:51, 14.70s/it]
 58%|█████▊    | 5691/9864 [23:57:06<17:29:21, 15.09s/it]
 58%|█████▊    | 5692/9864 [23:57:24<18:26:37, 15.92s/it]
 58%|█████▊    | 5693/9864 [23:57:38<17:38:37, 15.23s/it]
 58%|█████▊    | 5694/9864 [23:57:52<17:30:10, 15.11s/it]
 58%|█████▊    | 5695/9864 [23:58:07<17:14:24, 14.89s/it]
 58%|█████▊    | 5696/9864 [23:58:21<16:59:52, 14.68s/it]
 58%|█████▊    | 5697/9864 [23:58:37<17:27:08, 15.08s/it]
 58%|█████▊    | 5698/9864 [23:58:53<17:39:51, 15.26s/it]
 58%|█████▊    | 5699/9864 [23:59:11<18:33:23, 16.04s/it]
 58%|█████▊    | 5700/9864 [23:59:25<17:56:56, 15.52s/it]
                                                         
{'loss': '0.1376', 'grad_norm': '3.403', 'learning_rate': '7.982e-06', 'epoch': '0.5779'}

 58%|█████▊    | 5700/9864 [23:59:25<17:56:56, 15.52s/it]
 58%|█████▊    | 5701/9864 [23:59:39<17:19:22, 14.98s/it]
 58%|█████▊    | 5702/9864 [23:59:53<17:01:40, 14.73s/it]
 58%|█████▊    | 5703/9864 [24:00:07<17:00:02, 14.71s/it]
 58%|█████▊    | 5704/9864 [24:00:21<16:41:43, 14.45s/it]
 58%|█████▊    | 5705/9864 [24:00:37<17:04:58, 14.79s/it]
 58%|█████▊    | 5706/9864 [24:00:52<17:20:22, 15.01s/it]
 58%|█████▊    | 5707/9864 [24:01:10<18:21:40, 15.90s/it]
 58%|█████▊    | 5708/9864 [24:01:25<17:46:57, 15.40s/it]
 58%|█████▊    | 5709/9864 [24:01:39<17:21:38, 15.04s/it]
 58%|█████▊    | 5710/9864 [24:01:53<16:58:04, 14.70s/it]
                                                         
{'loss': '0.1155', 'grad_norm': '2.898', 'learning_rate': '7.949e-06', 'epoch': '0.5789'}

 58%|█████▊    | 5710/9864 [24:01:53<16:58:04, 14.70s/it]
 58%|█████▊    | 5711/9864 [24:02:07<16:52:22, 14.63s/it]
 58%|█████▊    | 5712/9864 [24:02:23<17:25:31, 15.11s/it]
 58%|█████▊    | 5713/9864 [24:02:38<17:17:06, 14.99s/it]
 58%|█████▊    | 5714/9864 [24:02:54<17:41:09, 15.34s/it]
 58%|█████▊    | 5715/9864 [24:03:10<17:57:14, 15.58s/it]
 58%|█████▊    | 5716/9864 [24:03:27<18:10:09, 15.77s/it]
 58%|█████▊    | 5717/9864 [24:03:41<17:34:44, 15.26s/it]
 58%|█████▊    | 5718/9864 [24:03:55<17:10:57, 14.92s/it]
 58%|█████▊    | 5719/9864 [24:04:09<16:59:43, 14.76s/it]
 58%|█████▊    | 5720/9864 [24:04:23<16:45:06, 14.55s/it]
                                                         
{'loss': '0.1351', 'grad_norm': '3.287', 'learning_rate': '7.917e-06', 'epoch': '0.5799'}

 58%|█████▊    | 5720/9864 [24:04:23<16:45:06, 14.55s/it]
 58%|█████▊    | 5721/9864 [24:04:39<17:07:39, 14.88s/it]
 58%|█████▊    | 5722/9864 [24:04:55<17:22:11, 15.10s/it]
 58%|█████▊    | 5723/9864 [24:05:10<17:38:36, 15.34s/it]
 58%|█████▊    | 5724/9864 [24:05:27<17:56:08, 15.60s/it]
 58%|█████▊    | 5725/9864 [24:05:41<17:32:22, 15.26s/it]
 58%|█████▊    | 5726/9864 [24:05:55<17:08:04, 14.91s/it]
 58%|█████▊    | 5727/9864 [24:06:09<16:42:42, 14.54s/it]
 58%|█████▊    | 5728/9864 [24:06:24<16:46:01, 14.59s/it]
 58%|█████▊    | 5729/9864 [24:06:40<17:13:35, 15.00s/it]
 58%|█████▊    | 5730/9864 [24:06:57<18:05:42, 15.76s/it]
                                                         
{'loss': '0.1249', 'grad_norm': '2.915', 'learning_rate': '7.885e-06', 'epoch': '0.5809'}

 58%|█████▊    | 5730/9864 [24:06:57<18:05:42, 15.76s/it]
 58%|█████▊    | 5731/9864 [24:07:13<18:12:56, 15.87s/it]
 58%|█████▊    | 5732/9864 [24:07:27<17:37:10, 15.35s/it]
 58%|█████▊    | 5733/9864 [24:07:42<17:15:59, 15.05s/it]
 58%|█████▊    | 5734/9864 [24:07:56<17:01:54, 14.85s/it]
 58%|█████▊    | 5735/9864 [24:08:10<16:48:08, 14.65s/it]
 58%|█████▊    | 5736/9864 [24:08:24<16:34:47, 14.46s/it]
 58%|█████▊    | 5737/9864 [24:08:40<16:58:47, 14.81s/it]
 58%|█████▊    | 5738/9864 [24:08:57<17:55:36, 15.64s/it]
 58%|█████▊    | 5739/9864 [24:09:11<17:19:46, 15.12s/it]
 58%|█████▊    | 5740/9864 [24:09:27<17:24:09, 15.19s/it]
                                                         
{'loss': '0.1209', 'grad_norm': '2.577', 'learning_rate': '7.853e-06', 'epoch': '0.582'}

 58%|█████▊    | 5740/9864 [24:09:27<17:24:09, 15.19s/it]
 58%|█████▊    | 5741/9864 [24:09:41<16:58:42, 14.82s/it]
 58%|█████▊    | 5742/9864 [24:09:55<16:47:19, 14.66s/it]
 58%|█████▊    | 5743/9864 [24:10:10<16:47:46, 14.67s/it]
 58%|█████▊    | 5744/9864 [24:10:25<17:11:07, 15.02s/it]
 58%|█████▊    | 5745/9864 [24:10:41<17:25:39, 15.23s/it]
 58%|█████▊    | 5746/9864 [24:10:57<17:46:11, 15.53s/it]
 58%|█████▊    | 5747/9864 [24:11:11<17:14:11, 15.07s/it]
 58%|█████▊    | 5748/9864 [24:11:27<17:16:29, 15.11s/it]
 58%|█████▊    | 5749/9864 [24:11:40<16:47:06, 14.68s/it]
 58%|█████▊    | 5750/9864 [24:11:54<16:21:33, 14.32s/it]
                                                         
{'loss': '0.1363', 'grad_norm': '3.481', 'learning_rate': '7.821e-06', 'epoch': '0.583'}

 58%|█████▊    | 5750/9864 [24:11:54<16:21:33, 14.32s/it]
 58%|█████▊    | 5751/9864 [24:12:07<16:06:40, 14.10s/it]
 58%|█████▊    | 5752/9864 [24:12:22<16:11:14, 14.17s/it]
 58%|█████▊    | 5753/9864 [24:12:42<18:13:51, 15.96s/it]
 58%|█████▊    | 5754/9864 [24:12:56<17:38:55, 15.46s/it]
 58%|█████▊    | 5755/9864 [24:13:10<17:10:06, 15.04s/it]
 58%|█████▊    | 5756/9864 [24:13:26<17:23:44, 15.24s/it]
 58%|█████▊    | 5757/9864 [24:13:40<17:06:25, 15.00s/it]
 58%|█████▊    | 5758/9864 [24:13:55<16:55:34, 14.84s/it]
 58%|█████▊    | 5759/9864 [24:14:09<16:39:43, 14.61s/it]
 58%|█████▊    | 5760/9864 [24:14:25<17:11:03, 15.07s/it]
                                                         
{'loss': '0.1154', 'grad_norm': '2.235', 'learning_rate': '7.789e-06', 'epoch': '0.584'}

 58%|█████▊    | 5760/9864 [24:14:25<17:11:03, 15.07s/it]
 58%|█████▊    | 5761/9864 [24:14:43<18:01:12, 15.81s/it]
 58%|█████▊    | 5762/9864 [24:14:57<17:26:33, 15.31s/it]
 58%|█████▊    | 5763/9864 [24:15:11<17:05:44, 15.01s/it]
 58%|█████▊    | 5764/9864 [24:15:27<17:17:24, 15.18s/it]
 58%|█████▊    | 5765/9864 [24:15:40<16:48:15, 14.76s/it]
 58%|█████▊    | 5766/9864 [24:15:54<16:28:04, 14.47s/it]
 58%|█████▊    | 5767/9864 [24:16:10<16:57:59, 14.91s/it]
 58%|█████▊    | 5768/9864 [24:16:28<17:55:19, 15.75s/it]
 58%|█████▊    | 5769/9864 [24:16:42<17:23:52, 15.29s/it]
 58%|█████▊    | 5770/9864 [24:16:56<16:56:12, 14.89s/it]
                                                         
{'loss': '0.1169', 'grad_norm': '3', 'learning_rate': '7.757e-06', 'epoch': '0.585'}

 58%|█████▊    | 5770/9864 [24:16:56<16:56:12, 14.89s/it]
 59%|█████▊    | 5771/9864 [24:17:12<17:11:20, 15.12s/it]
 59%|█████▊    | 5772/9864 [24:17:26<16:57:38, 14.92s/it]
 59%|█████▊    | 5773/9864 [24:17:40<16:36:58, 14.62s/it]
 59%|█████▊    | 5774/9864 [24:17:54<16:31:13, 14.54s/it]
 59%|█████▊    | 5775/9864 [24:18:11<17:03:35, 15.02s/it]
 59%|█████▊    | 5776/9864 [24:18:28<17:54:04, 15.76s/it]
 59%|█████▊    | 5777/9864 [24:18:42<17:14:25, 15.19s/it]
 59%|█████▊    | 5778/9864 [24:18:56<16:58:58, 14.96s/it]
 59%|█████▊    | 5779/9864 [24:19:12<17:20:24, 15.28s/it]
 59%|█████▊    | 5780/9864 [24:19:27<16:57:33, 14.95s/it]
                                                         
{'loss': '0.1214', 'grad_norm': '2.493', 'learning_rate': '7.725e-06', 'epoch': '0.586'}

 59%|█████▊    | 5780/9864 [24:19:27<16:57:33, 14.95s/it]
 59%|█████▊    | 5781/9864 [24:19:40<16:37:02, 14.65s/it]
 59%|█████▊    | 5782/9864 [24:19:55<16:29:13, 14.54s/it]
 59%|█████▊    | 5783/9864 [24:20:11<17:00:16, 15.00s/it]
 59%|█████▊    | 5784/9864 [24:20:27<17:13:58, 15.21s/it]
 59%|█████▊    | 5785/9864 [24:20:41<16:56:15, 14.95s/it]
 59%|█████▊    | 5786/9864 [24:20:55<16:38:01, 14.68s/it]
 59%|█████▊    | 5787/9864 [24:21:11<17:08:09, 15.13s/it]
 59%|█████▊    | 5788/9864 [24:21:25<16:49:28, 14.86s/it]
 59%|█████▊    | 5789/9864 [24:21:39<16:31:02, 14.59s/it]
 59%|█████▊    | 5790/9864 [24:21:54<16:29:49, 14.58s/it]
                                                         
{'loss': '0.1372', 'grad_norm': '3.159', 'learning_rate': '7.693e-06', 'epoch': '0.587'}

 59%|█████▊    | 5790/9864 [24:21:54<16:29:49, 14.58s/it]
 59%|█████▊    | 5791/9864 [24:22:12<17:32:44, 15.51s/it]
 59%|█████▊    | 5792/9864 [24:22:27<17:32:42, 15.51s/it]
 59%|█████▊    | 5793/9864 [24:22:42<17:15:47, 15.27s/it]
 59%|█████▊    | 5794/9864 [24:22:57<17:23:09, 15.38s/it]
 59%|█████▊    | 5795/9864 [24:23:11<16:50:42, 14.90s/it]
 59%|█████▉    | 5796/9864 [24:23:25<16:36:31, 14.70s/it]
 59%|█████▉    | 5797/9864 [24:23:40<16:28:04, 14.58s/it]
 59%|█████▉    | 5798/9864 [24:23:54<16:24:01, 14.52s/it]
 59%|█████▉    | 5799/9864 [24:24:12<17:36:24, 15.59s/it]
 59%|█████▉    | 5800/9864 [24:24:28<17:36:31, 15.60s/it]
                                                         
{'loss': '0.1183', 'grad_norm': '4.415', 'learning_rate': '7.661e-06', 'epoch': '0.588'}

 59%|█████▉    | 5800/9864 [24:24:28<17:36:31, 15.60s/it]
 59%|█████▉    | 5801/9864 [24:24:42<17:17:43, 15.32s/it]
 59%|█████▉    | 5802/9864 [24:24:58<17:28:49, 15.49s/it]
 59%|█████▉    | 5803/9864 [24:25:13<17:04:19, 15.13s/it]
 59%|█████▉    | 5804/9864 [24:25:27<16:47:05, 14.88s/it]
 59%|█████▉    | 5805/9864 [24:25:41<16:31:05, 14.65s/it]
 59%|█████▉    | 5806/9864 [24:25:57<17:05:27, 15.16s/it]
 59%|█████▉    | 5807/9864 [24:26:13<17:10:22, 15.24s/it]
 59%|█████▉    | 5808/9864 [24:26:27<16:46:13, 14.89s/it]
 59%|█████▉    | 5809/9864 [24:26:41<16:34:26, 14.71s/it]
 59%|█████▉    | 5810/9864 [24:26:57<16:58:10, 15.07s/it]
                                                         
{'loss': '0.1218', 'grad_norm': '4.384', 'learning_rate': '7.629e-06', 'epoch': '0.5891'}

 59%|█████▉    | 5810/9864 [24:26:57<16:58:10, 15.07s/it]
 59%|█████▉    | 5811/9864 [24:27:11<16:35:10, 14.73s/it]
 59%|█████▉    | 5812/9864 [24:27:26<16:29:31, 14.65s/it]
 59%|█████▉    | 5813/9864 [24:27:39<16:04:49, 14.29s/it]
 59%|█████▉    | 5814/9864 [24:27:53<16:02:11, 14.25s/it]
 59%|█████▉    | 5815/9864 [24:28:12<17:43:01, 15.75s/it]
 59%|█████▉    | 5816/9864 [24:28:27<17:17:11, 15.37s/it]
 59%|█████▉    | 5817/9864 [24:28:42<17:06:18, 15.22s/it]
 59%|█████▉    | 5818/9864 [24:28:58<17:18:41, 15.40s/it]
 59%|█████▉    | 5819/9864 [24:29:11<16:43:30, 14.89s/it]
 59%|█████▉    | 5820/9864 [24:29:25<16:28:36, 14.67s/it]
                                                         
{'loss': '0.1252', 'grad_norm': '3.094', 'learning_rate': '7.597e-06', 'epoch': '0.5901'}

 59%|█████▉    | 5820/9864 [24:29:25<16:28:36, 14.67s/it]
 59%|█████▉    | 5821/9864 [24:29:40<16:21:38, 14.57s/it]
 59%|█████▉    | 5822/9864 [24:29:57<17:25:01, 15.51s/it]
 59%|█████▉    | 5823/9864 [24:30:14<17:44:12, 15.80s/it]
 59%|█████▉    | 5824/9864 [24:30:28<17:08:30, 15.27s/it]
 59%|█████▉    | 5825/9864 [24:30:42<16:40:00, 14.86s/it]
 59%|█████▉    | 5826/9864 [24:30:58<16:59:05, 15.14s/it]
 59%|█████▉    | 5827/9864 [24:31:12<16:42:00, 14.89s/it]
 59%|█████▉    | 5828/9864 [24:31:26<16:19:52, 14.57s/it]
 59%|█████▉    | 5829/9864 [24:31:42<16:51:45, 15.04s/it]
 59%|█████▉    | 5830/9864 [24:32:00<17:48:15, 15.89s/it]
                                                         
{'loss': '0.1157', 'grad_norm': '2.731', 'learning_rate': '7.565e-06', 'epoch': '0.5911'}

 59%|█████▉    | 5830/9864 [24:32:00<17:48:15, 15.89s/it]
 59%|█████▉    | 5831/9864 [24:32:14<17:04:28, 15.24s/it]
 59%|█████▉    | 5832/9864 [24:32:28<16:41:59, 14.91s/it]
 59%|█████▉    | 5833/9864 [24:32:44<17:05:22, 15.26s/it]
 59%|█████▉    | 5834/9864 [24:32:57<16:32:57, 14.78s/it]
 59%|█████▉    | 5835/9864 [24:33:11<16:08:07, 14.42s/it]
 59%|█████▉    | 5836/9864 [24:33:25<15:57:37, 14.26s/it]
 59%|█████▉    | 5837/9864 [24:33:41<16:29:32, 14.74s/it]
 59%|█████▉    | 5838/9864 [24:33:58<17:18:49, 15.48s/it]
 59%|█████▉    | 5839/9864 [24:34:12<16:45:37, 14.99s/it]
 59%|█████▉    | 5840/9864 [24:34:26<16:22:36, 14.65s/it]
                                                         
{'loss': '0.1336', 'grad_norm': '4.333', 'learning_rate': '7.534e-06', 'epoch': '0.5921'}

 59%|█████▉    | 5840/9864 [24:34:26<16:22:36, 14.65s/it]
 59%|█████▉    | 5841/9864 [24:34:41<16:43:32, 14.97s/it]
 59%|█████▉    | 5842/9864 [24:34:55<16:23:49, 14.68s/it]
 59%|█████▉    | 5843/9864 [24:35:10<16:17:32, 14.59s/it]
 59%|█████▉    | 5844/9864 [24:35:25<16:37:14, 14.88s/it]
 59%|█████▉    | 5845/9864 [24:35:41<16:55:47, 15.16s/it]
 59%|█████▉    | 5846/9864 [24:35:57<17:06:23, 15.33s/it]
 59%|█████▉    | 5847/9864 [24:36:11<16:36:56, 14.89s/it]
 59%|█████▉    | 5848/9864 [24:36:25<16:20:25, 14.65s/it]
 59%|█████▉    | 5849/9864 [24:36:40<16:33:46, 14.85s/it]
 59%|█████▉    | 5850/9864 [24:36:54<16:14:12, 14.56s/it]
                                                         
{'loss': '0.1188', 'grad_norm': '3.205', 'learning_rate': '7.502e-06', 'epoch': '0.5931'}

 59%|█████▉    | 5850/9864 [24:36:54<16:14:12, 14.56s/it]
 59%|█████▉    | 5851/9864 [24:37:08<16:01:22, 14.37s/it]
 59%|█████▉    | 5852/9864 [24:37:24<16:37:41, 14.92s/it]
 59%|█████▉    | 5853/9864 [24:37:40<16:56:11, 15.20s/it]
 59%|█████▉    | 5854/9864 [24:37:56<17:12:01, 15.44s/it]
 59%|█████▉    | 5855/9864 [24:38:10<16:46:04, 15.06s/it]
 59%|█████▉    | 5856/9864 [24:38:26<17:03:41, 15.32s/it]
 59%|█████▉    | 5857/9864 [24:38:40<16:38:30, 14.95s/it]
 59%|█████▉    | 5858/9864 [24:38:54<16:19:42, 14.67s/it]
 59%|█████▉    | 5859/9864 [24:39:08<16:06:10, 14.47s/it]
 59%|█████▉    | 5860/9864 [24:39:26<17:11:15, 15.45s/it]
                                                         
{'loss': '0.1355', 'grad_norm': '3.766', 'learning_rate': '7.47e-06', 'epoch': '0.5941'}

 59%|█████▉    | 5860/9864 [24:39:26<17:11:15, 15.45s/it]
 59%|█████▉    | 5861/9864 [24:39:42<17:17:56, 15.56s/it]
 59%|█████▉    | 5862/9864 [24:39:56<17:00:15, 15.30s/it]
 59%|█████▉    | 5863/9864 [24:40:11<16:42:00, 15.03s/it]
 59%|█████▉    | 5864/9864 [24:40:27<17:04:21, 15.37s/it]
 59%|█████▉    | 5865/9864 [24:40:41<16:32:40, 14.89s/it]
 59%|█████▉    | 5866/9864 [24:40:55<16:10:28, 14.56s/it]
 59%|█████▉    | 5867/9864 [24:41:09<15:57:55, 14.38s/it]
 59%|█████▉    | 5868/9864 [24:41:27<17:15:20, 15.55s/it]
 59%|█████▉    | 5869/9864 [24:41:43<17:26:05, 15.71s/it]
 60%|█████▉    | 5870/9864 [24:41:57<17:00:03, 15.32s/it]
                                                         
{'loss': '0.1256', 'grad_norm': '2.515', 'learning_rate': '7.438e-06', 'epoch': '0.5951'}

 60%|█████▉    | 5870/9864 [24:41:57<17:00:03, 15.32s/it]
 60%|█████▉    | 5871/9864 [24:42:11<16:30:44, 14.89s/it]
 60%|█████▉    | 5872/9864 [24:42:27<16:41:06, 15.05s/it]
 60%|█████▉    | 5873/9864 [24:42:41<16:24:42, 14.80s/it]
 60%|█████▉    | 5874/9864 [24:42:55<16:11:20, 14.61s/it]
 60%|█████▉    | 5875/9864 [24:43:09<16:00:54, 14.45s/it]
 60%|█████▉    | 5876/9864 [24:43:27<17:13:34, 15.55s/it]
 60%|█████▉    | 5877/9864 [24:43:41<16:37:36, 15.01s/it]
 60%|█████▉    | 5878/9864 [24:43:55<16:25:38, 14.84s/it]
 60%|█████▉    | 5879/9864 [24:44:10<16:14:09, 14.67s/it]
 60%|█████▉    | 5880/9864 [24:44:26<16:50:20, 15.22s/it]
                                                         
{'loss': '0.1126', 'grad_norm': '3.134', 'learning_rate': '7.407e-06', 'epoch': '0.5962'}

 60%|█████▉    | 5880/9864 [24:44:26<16:50:20, 15.22s/it]
 60%|█████▉    | 5881/9864 [24:44:40<16:20:46, 14.77s/it]
 60%|█████▉    | 5882/9864 [24:44:54<16:10:34, 14.62s/it]
 60%|█████▉    | 5883/9864 [24:45:10<16:36:52, 15.02s/it]
 60%|█████▉    | 5884/9864 [24:45:28<17:24:10, 15.74s/it]
 60%|█████▉    | 5885/9864 [24:45:42<16:52:58, 15.27s/it]
 60%|█████▉    | 5886/9864 [24:45:56<16:33:02, 14.98s/it]
 60%|█████▉    | 5887/9864 [24:46:10<16:15:31, 14.72s/it]
 60%|█████▉    | 5888/9864 [24:46:26<16:39:06, 15.08s/it]
 60%|█████▉    | 5889/9864 [24:46:41<16:27:39, 14.91s/it]
 60%|█████▉    | 5890/9864 [24:46:57<16:49:00, 15.23s/it]
                                                         
{'loss': '0.1285', 'grad_norm': '3.356', 'learning_rate': '7.375e-06', 'epoch': '0.5972'}

 60%|█████▉    | 5890/9864 [24:46:57<16:49:00, 15.23s/it]
 60%|█████▉    | 5891/9864 [24:47:12<17:01:33, 15.43s/it]
 60%|█████▉    | 5892/9864 [24:47:29<17:17:17, 15.67s/it]
 60%|█████▉    | 5893/9864 [24:47:43<16:56:51, 15.36s/it]
 60%|█████▉    | 5894/9864 [24:47:58<16:40:16, 15.12s/it]
 60%|█████▉    | 5895/9864 [24:48:13<16:50:47, 15.28s/it]
 60%|█████▉    | 5896/9864 [24:48:27<16:16:34, 14.77s/it]
 60%|█████▉    | 5897/9864 [24:48:43<16:44:44, 15.20s/it]
 60%|█████▉    | 5898/9864 [24:48:57<16:19:32, 14.82s/it]
 60%|█████▉    | 5899/9864 [24:49:13<16:35:10, 15.06s/it]
 60%|█████▉    | 5900/9864 [24:49:29<16:53:28, 15.34s/it]
                                                         
{'loss': '0.1406', 'grad_norm': '3.389', 'learning_rate': '7.343e-06', 'epoch': '0.5982'}

 60%|█████▉    | 5900/9864 [24:49:29<16:53:28, 15.34s/it]
 60%|█████▉    | 5901/9864 [24:49:43<16:31:10, 15.01s/it]
 60%|█████▉    | 5902/9864 [24:49:58<16:21:06, 14.86s/it]
 60%|█████▉    | 5903/9864 [24:50:13<16:40:10, 15.15s/it]
 60%|█████▉    | 5904/9864 [24:50:28<16:23:58, 14.91s/it]
 60%|█████▉    | 5905/9864 [24:50:43<16:37:09, 15.11s/it]
 60%|█████▉    | 5906/9864 [24:51:00<17:00:06, 15.46s/it]
 60%|█████▉    | 5907/9864 [24:51:14<16:39:31, 15.16s/it]
 60%|█████▉    | 5908/9864 [24:51:30<16:56:10, 15.41s/it]
 60%|█████▉    | 5909/9864 [24:51:44<16:32:19, 15.05s/it]
 60%|█████▉    | 5910/9864 [24:52:00<16:43:45, 15.23s/it]
                                                         
{'loss': '0.1303', 'grad_norm': '3.807', 'learning_rate': '7.312e-06', 'epoch': '0.5992'}

 60%|█████▉    | 5910/9864 [24:52:00<16:43:45, 15.23s/it]
 60%|█████▉    | 5911/9864 [24:52:14<16:18:58, 14.86s/it]
 60%|█████▉    | 5912/9864 [24:52:28<15:55:26, 14.51s/it]
 60%|█████▉    | 5913/9864 [24:52:42<15:48:45, 14.41s/it]
 60%|█████▉    | 5914/9864 [24:52:59<16:41:26, 15.21s/it]
 60%|█████▉    | 5915/9864 [24:53:15<16:53:29, 15.40s/it]
 60%|█████▉    | 5916/9864 [24:53:29<16:32:59, 15.09s/it]
 60%|█████▉    | 5917/9864 [24:53:44<16:23:22, 14.95s/it]
 60%|█████▉    | 5918/9864 [24:54:00<16:41:58, 15.24s/it]
 60%|██████    | 5919/9864 [24:54:14<16:16:53, 14.86s/it]
 60%|██████    | 5920/9864 [24:54:27<15:55:59, 14.54s/it]
                                                         
{'loss': '0.1213', 'grad_norm': '3.072', 'learning_rate': '7.28e-06', 'epoch': '0.6002'}

 60%|██████    | 5920/9864 [24:54:27<15:55:59, 14.54s/it]
 60%|██████    | 5921/9864 [24:54:43<16:25:04, 14.99s/it]
 60%|██████    | 5922/9864 [24:54:59<16:33:47, 15.13s/it]
 60%|██████    | 5923/9864 [24:55:15<16:56:28, 15.48s/it]
 60%|██████    | 5924/9864 [24:55:29<16:33:27, 15.13s/it]
 60%|██████    | 5925/9864 [24:55:46<16:53:36, 15.44s/it]
 60%|██████    | 5926/9864 [24:55:59<16:16:03, 14.87s/it]
 60%|██████    | 5927/9864 [24:56:13<16:04:10, 14.69s/it]
 60%|██████    | 5928/9864 [24:56:27<15:38:52, 14.31s/it]
 60%|██████    | 5929/9864 [24:56:43<16:14:27, 14.86s/it]
 60%|██████    | 5930/9864 [24:57:01<17:12:19, 15.74s/it]
                                                         
{'loss': '0.1317', 'grad_norm': '3.555', 'learning_rate': '7.248e-06', 'epoch': '0.6012'}

 60%|██████    | 5930/9864 [24:57:01<17:12:19, 15.74s/it]
 60%|██████    | 5931/9864 [24:57:15<16:38:00, 15.23s/it]
 60%|██████    | 5932/9864 [24:57:28<16:07:11, 14.76s/it]
 60%|██████    | 5933/9864 [24:57:44<16:26:12, 15.05s/it]
 60%|██████    | 5934/9864 [24:57:58<15:55:22, 14.59s/it]
 60%|██████    | 5935/9864 [24:58:11<15:38:47, 14.34s/it]
 60%|██████    | 5936/9864 [24:58:27<16:07:02, 14.77s/it]
 60%|██████    | 5937/9864 [24:58:43<16:29:10, 15.11s/it]
 60%|██████    | 5938/9864 [24:58:57<16:05:16, 14.75s/it]
 60%|██████    | 5939/9864 [24:59:13<16:29:00, 15.12s/it]
 60%|██████    | 5940/9864 [24:59:28<16:17:08, 14.94s/it]
                                                         
{'loss': '0.1191', 'grad_norm': '2.42', 'learning_rate': '7.217e-06', 'epoch': '0.6022'}

 60%|██████    | 5940/9864 [24:59:28<16:17:08, 14.94s/it]
 60%|██████    | 5941/9864 [24:59:43<16:33:46, 15.20s/it]
 60%|██████    | 5942/9864 [24:59:58<16:18:08, 14.96s/it]
 60%|██████    | 5943/9864 [25:00:12<15:53:11, 14.59s/it]
 60%|██████    | 5944/9864 [25:00:27<16:04:49, 14.77s/it]
 60%|██████    | 5945/9864 [25:00:42<16:22:58, 15.05s/it]
 60%|██████    | 5946/9864 [25:00:58<16:35:22, 15.24s/it]
 60%|██████    | 5947/9864 [25:01:12<16:07:29, 14.82s/it]
 60%|██████    | 5948/9864 [25:01:27<16:02:18, 14.74s/it]
 60%|██████    | 5949/9864 [25:01:43<16:27:08, 15.13s/it]
 60%|██████    | 5950/9864 [25:01:57<16:11:15, 14.89s/it]
                                                         
{'loss': '0.1242', 'grad_norm': '3.486', 'learning_rate': '7.185e-06', 'epoch': '0.6033'}

 60%|██████    | 5950/9864 [25:01:57<16:11:15, 14.89s/it]
 60%|██████    | 5951/9864 [25:02:11<15:59:34, 14.71s/it]
 60%|██████    | 5952/9864 [25:02:27<16:22:58, 15.08s/it]
 60%|██████    | 5953/9864 [25:02:43<16:33:24, 15.24s/it]
 60%|██████    | 5954/9864 [25:02:57<16:06:16, 14.83s/it]
 60%|██████    | 5955/9864 [25:03:12<16:20:56, 15.06s/it]
 60%|██████    | 5956/9864 [25:03:28<16:37:46, 15.32s/it]
 60%|██████    | 5957/9864 [25:03:42<16:08:16, 14.87s/it]
 60%|██████    | 5958/9864 [25:03:56<15:49:25, 14.58s/it]
 60%|██████    | 5959/9864 [25:04:10<15:39:42, 14.44s/it]
 60%|██████    | 5960/9864 [25:04:26<16:09:53, 14.91s/it]
                                                         
{'loss': '0.0988', 'grad_norm': '2.947', 'learning_rate': '7.154e-06', 'epoch': '0.6043'}

 60%|██████    | 5960/9864 [25:04:26<16:09:53, 14.91s/it]
 60%|██████    | 5961/9864 [25:04:42<16:27:36, 15.18s/it]
 60%|██████    | 5962/9864 [25:04:56<16:06:31, 14.86s/it]
 60%|██████    | 5963/9864 [25:05:12<16:27:45, 15.19s/it]
 60%|██████    | 5964/9864 [25:05:27<16:30:20, 15.24s/it]
 60%|██████    | 5965/9864 [25:05:42<16:20:27, 15.09s/it]
 60%|██████    | 5966/9864 [25:05:56<16:08:37, 14.91s/it]
 60%|██████    | 5967/9864 [25:06:11<16:00:00, 14.78s/it]
 61%|██████    | 5968/9864 [25:06:26<16:08:18, 14.91s/it]
 61%|██████    | 5969/9864 [25:06:41<16:14:36, 15.01s/it]
 61%|██████    | 5970/9864 [25:06:56<15:59:47, 14.79s/it]
                                                         
{'loss': '0.1193', 'grad_norm': '3.777', 'learning_rate': '7.122e-06', 'epoch': '0.6053'}

 61%|██████    | 5970/9864 [25:06:56<15:59:47, 14.79s/it]
 61%|██████    | 5971/9864 [25:07:11<16:18:35, 15.08s/it]
 61%|██████    | 5972/9864 [25:07:28<16:41:14, 15.44s/it]
 61%|██████    | 5973/9864 [25:07:42<16:23:00, 15.16s/it]
 61%|██████    | 5974/9864 [25:07:56<16:02:23, 14.84s/it]
 61%|██████    | 5975/9864 [25:08:12<16:17:21, 15.08s/it]
 61%|██████    | 5976/9864 [25:08:26<16:02:41, 14.86s/it]
 61%|██████    | 5977/9864 [25:08:42<16:19:05, 15.11s/it]
 61%|██████    | 5978/9864 [25:08:56<15:50:33, 14.68s/it]
 61%|██████    | 5979/9864 [25:09:13<16:51:44, 15.63s/it]
 61%|██████    | 5980/9864 [25:09:28<16:23:46, 15.20s/it]
                                                         
{'loss': '0.1174', 'grad_norm': '2.784', 'learning_rate': '7.091e-06', 'epoch': '0.6063'}

 61%|██████    | 5980/9864 [25:09:28<16:23:46, 15.20s/it]
 61%|██████    | 5981/9864 [25:09:42<16:07:46, 14.95s/it]
 61%|██████    | 5982/9864 [25:09:58<16:19:29, 15.14s/it]
 61%|██████    | 5983/9864 [25:10:12<16:00:27, 14.85s/it]
 61%|██████    | 5984/9864 [25:10:27<16:12:55, 15.05s/it]
 61%|██████    | 5985/9864 [25:10:42<15:59:22, 14.84s/it]
 61%|██████    | 5986/9864 [25:10:58<16:25:35, 15.25s/it]
 61%|██████    | 5987/9864 [25:11:12<16:04:09, 14.92s/it]
 61%|██████    | 5988/9864 [25:11:28<16:23:11, 15.22s/it]
 61%|██████    | 5989/9864 [25:11:42<16:02:12, 14.90s/it]
 61%|██████    | 5990/9864 [25:11:58<16:17:54, 15.15s/it]
                                                         
{'loss': '0.1127', 'grad_norm': '3.34', 'learning_rate': '7.059e-06', 'epoch': '0.6073'}

 61%|██████    | 5990/9864 [25:11:58<16:17:54, 15.15s/it]
 61%|██████    | 5991/9864 [25:12:12<15:51:07, 14.73s/it]
 61%|██████    | 5992/9864 [25:12:26<15:44:14, 14.63s/it]
 61%|██████    | 5993/9864 [25:12:42<16:03:40, 14.94s/it]
 61%|██████    | 5994/9864 [25:12:58<16:27:10, 15.31s/it]
 61%|██████    | 5995/9864 [25:13:12<16:06:47, 14.99s/it]
 61%|██████    | 5996/9864 [25:13:28<16:16:00, 15.14s/it]
 61%|██████    | 5997/9864 [25:13:41<15:47:55, 14.71s/it]
 61%|██████    | 5998/9864 [25:13:57<16:06:08, 14.99s/it]
 61%|██████    | 5999/9864 [25:14:11<15:55:34, 14.83s/it]
 61%|██████    | 6000/9864 [25:14:26<15:44:16, 14.66s/it]
                                                         
{'loss': '0.1146', 'grad_norm': '3.766', 'learning_rate': '7.028e-06', 'epoch': '0.6083'}

 61%|██████    | 6000/9864 [25:14:26<15:44:16, 14.66s/it]

Writing model shards:   0%|          | 0/1 [00:00<?, ?it/s]

Writing model shards: 100%|██████████| 1/1 [00:08<00:00,  8.89s/it]
Writing model shards: 100%|██████████| 1/1 [00:08<00:00,  8.89s/it]
/home/dungpq6/Project/SpatialStack/src/qwen_vl/model/geometry_encoders/vggt_encoder.py:285: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(dtype=dtype):

 61%|██████    | 6001/9864 [25:15:44<36:17:04, 33.81s/it]
 61%|██████    | 6002/9864 [25:15:59<30:03:27, 28.02s/it]
 61%|██████    | 6003/9864 [25:16:15<26:09:20, 24.39s/it]
 61%|██████    | 6004/9864 [25:16:29<22:53:28, 21.35s/it]
 61%|██████    | 6005/9864 [25:16:43<20:32:54, 19.17s/it]
 61%|██████    | 6006/9864 [25:16:59<19:33:39, 18.25s/it]
 61%|██████    | 6007/9864 [25:17:13<18:17:21, 17.07s/it]
 61%|██████    | 6008/9864 [25:17:30<18:05:39, 16.89s/it]
 61%|██████    | 6009/9864 [25:17:46<17:56:04, 16.75s/it]
 61%|██████    | 6010/9864 [25:18:02<17:28:52, 16.33s/it]
                                                         
{'loss': '0.1165', 'grad_norm': '3.432', 'learning_rate': '6.997e-06', 'epoch': '0.6093'}

 61%|██████    | 6010/9864 [25:18:02<17:28:52, 16.33s/it]
 61%|██████    | 6011/9864 [25:18:15<16:39:00, 15.56s/it]
 61%|██████    | 6012/9864 [25:18:29<16:11:39, 15.13s/it]
 61%|██████    | 6013/9864 [25:18:44<15:52:10, 14.84s/it]
 61%|██████    | 6014/9864 [25:18:58<15:37:41, 14.61s/it]
 61%|██████    | 6015/9864 [25:19:14<16:09:23, 15.11s/it]
 61%|██████    | 6016/9864 [25:19:29<16:15:30, 15.21s/it]
 61%|██████    | 6017/9864 [25:19:45<16:17:53, 15.25s/it]
 61%|██████    | 6018/9864 [25:19:59<15:50:24, 14.83s/it]
 61%|██████    | 6019/9864 [25:20:14<16:02:21, 15.02s/it]
 61%|██████    | 6020/9864 [25:20:28<15:46:05, 14.77s/it]
                                                         
{'loss': '0.1093', 'grad_norm': '3.668', 'learning_rate': '6.965e-06', 'epoch': '0.6104'}

 61%|██████    | 6020/9864 [25:20:28<15:46:05, 14.77s/it]
 61%|██████    | 6021/9864 [25:20:43<15:43:15, 14.73s/it]
 61%|██████    | 6022/9864 [25:20:58<15:44:26, 14.75s/it]
 61%|██████    | 6023/9864 [25:21:15<16:36:11, 15.56s/it]
 61%|██████    | 6024/9864 [25:21:31<16:35:56, 15.56s/it]
 61%|██████    | 6025/9864 [25:21:46<16:23:09, 15.37s/it]
 61%|██████    | 6026/9864 [25:22:00<16:02:39, 15.05s/it]
 61%|██████    | 6027/9864 [25:22:15<16:07:56, 15.14s/it]
 61%|██████    | 6028/9864 [25:22:29<15:47:15, 14.82s/it]
 61%|██████    | 6029/9864 [25:22:44<15:38:24, 14.68s/it]
 61%|██████    | 6030/9864 [25:23:00<16:01:08, 15.04s/it]
                                                         
{'loss': '0.106', 'grad_norm': '3.843', 'learning_rate': '6.934e-06', 'epoch': '0.6114'}

 61%|██████    | 6030/9864 [25:23:00<16:01:08, 15.04s/it]
 61%|██████    | 6031/9864 [25:23:15<16:17:57, 15.31s/it]
 61%|██████    | 6032/9864 [25:23:31<16:26:14, 15.44s/it]
 61%|██████    | 6033/9864 [25:23:46<16:04:18, 15.10s/it]
 61%|██████    | 6034/9864 [25:24:01<16:03:12, 15.09s/it]
 61%|██████    | 6035/9864 [25:24:17<16:25:36, 15.44s/it]
 61%|██████    | 6036/9864 [25:24:31<16:01:28, 15.07s/it]
 61%|██████    | 6037/9864 [25:24:47<16:09:40, 15.20s/it]
 61%|██████    | 6038/9864 [25:25:01<15:46:20, 14.84s/it]
 61%|██████    | 6039/9864 [25:25:14<15:24:06, 14.50s/it]
 61%|██████    | 6040/9864 [25:25:32<16:18:58, 15.36s/it]
                                                         
{'loss': '0.1171', 'grad_norm': '4.442', 'learning_rate': '6.903e-06', 'epoch': '0.6124'}

 61%|██████    | 6040/9864 [25:25:32<16:18:58, 15.36s/it]
 61%|██████    | 6041/9864 [25:25:46<15:53:11, 14.96s/it]
 61%|██████▏   | 6042/9864 [25:26:00<15:32:17, 14.64s/it]
 61%|██████▏   | 6043/9864 [25:26:16<15:58:19, 15.05s/it]
 61%|██████▏   | 6044/9864 [25:26:31<16:12:50, 15.28s/it]
 61%|██████▏   | 6045/9864 [25:26:46<15:52:54, 14.97s/it]
 61%|██████▏   | 6046/9864 [25:27:00<15:36:38, 14.72s/it]
 61%|██████▏   | 6047/9864 [25:27:14<15:20:09, 14.46s/it]
 61%|██████▏   | 6048/9864 [25:27:31<16:12:20, 15.29s/it]
 61%|██████▏   | 6049/9864 [25:27:45<15:55:07, 15.02s/it]
 61%|██████▏   | 6050/9864 [25:27:59<15:37:04, 14.74s/it]
                                                         
{'loss': '0.1236', 'grad_norm': '3.45', 'learning_rate': '6.872e-06', 'epoch': '0.6134'}

 61%|██████▏   | 6050/9864 [25:27:59<15:37:04, 14.74s/it]
 61%|██████▏   | 6051/9864 [25:28:15<15:52:04, 14.98s/it]
 61%|██████▏   | 6052/9864 [25:28:29<15:44:53, 14.87s/it]
 61%|██████▏   | 6053/9864 [25:28:45<15:53:38, 15.01s/it]
 61%|██████▏   | 6054/9864 [25:28:59<15:36:36, 14.75s/it]
 61%|██████▏   | 6055/9864 [25:29:14<15:39:43, 14.80s/it]
 61%|██████▏   | 6056/9864 [25:29:30<15:57:36, 15.09s/it]
 61%|██████▏   | 6057/9864 [25:29:43<15:27:24, 14.62s/it]
 61%|██████▏   | 6058/9864 [25:29:59<15:53:51, 15.04s/it]
 61%|██████▏   | 6059/9864 [25:30:14<15:42:45, 14.87s/it]
 61%|██████▏   | 6060/9864 [25:30:27<15:15:01, 14.43s/it]
                                                         
{'loss': '0.1155', 'grad_norm': '2.494', 'learning_rate': '6.841e-06', 'epoch': '0.6144'}

 61%|██████▏   | 6060/9864 [25:30:27<15:15:01, 14.43s/it]
 61%|██████▏   | 6061/9864 [25:30:43<15:45:42, 14.92s/it]
 61%|██████▏   | 6062/9864 [25:30:57<15:34:20, 14.74s/it]
 61%|██████▏   | 6063/9864 [25:31:12<15:27:33, 14.64s/it]
 61%|██████▏   | 6064/9864 [25:31:29<16:19:55, 15.47s/it]
 61%|██████▏   | 6065/9864 [25:31:43<15:55:49, 15.10s/it]
 61%|██████▏   | 6066/9864 [25:31:57<15:32:54, 14.74s/it]
 62%|██████▏   | 6067/9864 [25:32:13<15:58:39, 15.15s/it]
 62%|██████▏   | 6068/9864 [25:32:29<16:05:31, 15.26s/it]
 62%|██████▏   | 6069/9864 [25:32:43<15:44:36, 14.93s/it]
 62%|██████▏   | 6070/9864 [25:32:57<15:29:20, 14.70s/it]
                                                         
{'loss': '0.1226', 'grad_norm': '2.815', 'learning_rate': '6.809e-06', 'epoch': '0.6154'}

 62%|██████▏   | 6070/9864 [25:32:57<15:29:20, 14.70s/it]
 62%|██████▏   | 6071/9864 [25:33:11<15:15:28, 14.48s/it]
 62%|██████▏   | 6072/9864 [25:33:29<16:10:52, 15.36s/it]
 62%|██████▏   | 6073/9864 [25:33:43<15:50:07, 15.04s/it]
 62%|██████▏   | 6074/9864 [25:33:59<16:13:03, 15.40s/it]
 62%|██████▏   | 6075/9864 [25:34:14<15:51:28, 15.07s/it]
 62%|██████▏   | 6076/9864 [25:34:29<16:01:07, 15.22s/it]
 62%|██████▏   | 6077/9864 [25:34:43<15:39:19, 14.88s/it]
 62%|██████▏   | 6078/9864 [25:34:57<15:16:10, 14.52s/it]
 62%|██████▏   | 6079/9864 [25:35:11<15:12:14, 14.46s/it]
 62%|██████▏   | 6080/9864 [25:35:27<15:43:19, 14.96s/it]
                                                         
{'loss': '0.107', 'grad_norm': '2.491', 'learning_rate': '6.778e-06', 'epoch': '0.6164'}

 62%|██████▏   | 6080/9864 [25:35:27<15:43:19, 14.96s/it]
 62%|██████▏   | 6081/9864 [25:35:46<16:44:11, 15.93s/it]
 62%|██████▏   | 6082/9864 [25:36:00<16:07:13, 15.34s/it]
 62%|██████▏   | 6083/9864 [25:36:15<16:18:34, 15.53s/it]
 62%|██████▏   | 6084/9864 [25:36:30<15:51:29, 15.10s/it]
 62%|██████▏   | 6085/9864 [25:36:43<15:28:17, 14.74s/it]
 62%|██████▏   | 6086/9864 [25:36:58<15:21:39, 14.64s/it]
 62%|██████▏   | 6087/9864 [25:37:12<15:08:36, 14.43s/it]
 62%|██████▏   | 6088/9864 [25:37:28<15:39:42, 14.93s/it]
 62%|██████▏   | 6089/9864 [25:37:43<15:50:33, 15.11s/it]
 62%|██████▏   | 6090/9864 [25:37:58<15:33:09, 14.84s/it]
                                                         
{'loss': '0.1134', 'grad_norm': '3.492', 'learning_rate': '6.747e-06', 'epoch': '0.6174'}

 62%|██████▏   | 6090/9864 [25:37:58<15:33:09, 14.84s/it]
 62%|██████▏   | 6091/9864 [25:38:13<15:52:00, 15.14s/it]
 62%|██████▏   | 6092/9864 [25:38:27<15:30:10, 14.80s/it]
 62%|██████▏   | 6093/9864 [25:38:41<15:11:43, 14.51s/it]
 62%|██████▏   | 6094/9864 [25:38:56<15:12:51, 14.53s/it]
 62%|██████▏   | 6095/9864 [25:39:10<15:06:21, 14.43s/it]
 62%|██████▏   | 6096/9864 [25:39:28<16:06:06, 15.38s/it]
 62%|██████▏   | 6097/9864 [25:39:44<16:16:09, 15.55s/it]
 62%|██████▏   | 6098/9864 [25:40:00<16:22:39, 15.66s/it]
 62%|██████▏   | 6099/9864 [25:40:14<15:56:10, 15.24s/it]
 62%|██████▏   | 6100/9864 [25:40:28<15:39:58, 14.98s/it]
                                                         
{'loss': '0.1371', 'grad_norm': '6.076', 'learning_rate': '6.716e-06', 'epoch': '0.6185'}

 62%|██████▏   | 6100/9864 [25:40:28<15:39:58, 14.98s/it]
 62%|██████▏   | 6101/9864 [25:40:42<15:17:47, 14.63s/it]
 62%|██████▏   | 6102/9864 [25:40:57<15:16:06, 14.61s/it]
 62%|██████▏   | 6103/9864 [25:41:12<15:38:15, 14.97s/it]
 62%|██████▏   | 6104/9864 [25:41:28<15:55:09, 15.24s/it]
 62%|██████▏   | 6105/9864 [25:41:44<16:08:44, 15.46s/it]
 62%|██████▏   | 6106/9864 [25:42:01<16:24:48, 15.72s/it]
 62%|██████▏   | 6107/9864 [25:42:15<15:55:19, 15.26s/it]
 62%|██████▏   | 6108/9864 [25:42:29<15:34:01, 14.92s/it]
 62%|██████▏   | 6109/9864 [25:42:43<15:24:47, 14.78s/it]
 62%|██████▏   | 6110/9864 [25:43:00<15:56:27, 15.29s/it]
                                                         
{'loss': '0.1218', 'grad_norm': '2.884', 'learning_rate': '6.685e-06', 'epoch': '0.6195'}

 62%|██████▏   | 6110/9864 [25:43:00<15:56:27, 15.29s/it]
 62%|██████▏   | 6111/9864 [25:43:14<15:37:07, 14.98s/it]
 62%|██████▏   | 6112/9864 [25:43:30<15:52:45, 15.24s/it]
 62%|██████▏   | 6113/9864 [25:43:46<16:10:17, 15.52s/it]
 62%|██████▏   | 6114/9864 [25:44:02<16:13:08, 15.57s/it]
 62%|██████▏   | 6115/9864 [25:44:16<15:54:11, 15.27s/it]
 62%|██████▏   | 6116/9864 [25:44:30<15:30:04, 14.89s/it]
 62%|██████▏   | 6117/9864 [25:44:44<15:08:24, 14.55s/it]
 62%|██████▏   | 6118/9864 [25:44:59<15:09:13, 14.56s/it]
 62%|██████▏   | 6119/9864 [25:45:16<16:02:50, 15.43s/it]
 62%|██████▏   | 6120/9864 [25:45:31<15:44:14, 15.13s/it]
                                                         
{'loss': '0.1303', 'grad_norm': '3.906', 'learning_rate': '6.654e-06', 'epoch': '0.6205'}

 62%|██████▏   | 6120/9864 [25:45:31<15:44:14, 15.13s/it]
 62%|██████▏   | 6121/9864 [25:45:47<16:06:16, 15.49s/it]
 62%|██████▏   | 6122/9864 [25:46:03<16:17:04, 15.67s/it]
 62%|██████▏   | 6123/9864 [25:46:16<15:35:03, 15.00s/it]
 62%|██████▏   | 6124/9864 [25:46:31<15:20:28, 14.77s/it]
 62%|██████▏   | 6125/9864 [25:46:45<15:10:29, 14.61s/it]
 62%|██████▏   | 6126/9864 [25:47:01<15:40:13, 15.09s/it]
 62%|██████▏   | 6127/9864 [25:47:17<16:00:34, 15.42s/it]
 62%|██████▏   | 6128/9864 [25:47:33<16:01:09, 15.44s/it]
 62%|██████▏   | 6129/9864 [25:47:47<15:39:51, 15.10s/it]
 62%|██████▏   | 6130/9864 [25:48:03<15:49:16, 15.25s/it]
                                                         
{'loss': '0.1307', 'grad_norm': '3.415', 'learning_rate': '6.623e-06', 'epoch': '0.6215'}

 62%|██████▏   | 6130/9864 [25:48:03<15:49:16, 15.25s/it]
 62%|██████▏   | 6131/9864 [25:48:17<15:35:25, 15.04s/it]
 62%|██████▏   | 6132/9864 [25:48:31<15:20:29, 14.80s/it]
 62%|██████▏   | 6133/9864 [25:48:46<15:12:41, 14.68s/it]
 62%|██████▏   | 6134/9864 [25:49:02<15:45:36, 15.21s/it]
 62%|██████▏   | 6135/9864 [25:49:18<15:57:32, 15.41s/it]
 62%|██████▏   | 6136/9864 [25:49:34<16:06:22, 15.55s/it]
 62%|██████▏   | 6137/9864 [25:49:48<15:33:45, 15.03s/it]
 62%|██████▏   | 6138/9864 [25:50:04<16:00:56, 15.47s/it]
 62%|██████▏   | 6139/9864 [25:50:19<15:36:48, 15.09s/it]
 62%|██████▏   | 6140/9864 [25:50:33<15:25:11, 14.91s/it]
                                                         
{'loss': '0.1287', 'grad_norm': '4.024', 'learning_rate': '6.592e-06', 'epoch': '0.6225'}

 62%|██████▏   | 6140/9864 [25:50:33<15:25:11, 14.91s/it]
 62%|██████▏   | 6141/9864 [25:50:47<15:09:36, 14.66s/it]
 62%|██████▏   | 6142/9864 [25:51:03<15:35:32, 15.08s/it]
 62%|██████▏   | 6143/9864 [25:51:19<15:55:44, 15.41s/it]
 62%|██████▏   | 6144/9864 [25:51:36<16:13:57, 15.71s/it]
 62%|██████▏   | 6145/9864 [25:51:50<15:47:58, 15.29s/it]
 62%|██████▏   | 6146/9864 [25:52:06<15:56:54, 15.44s/it]
 62%|██████▏   | 6147/9864 [25:52:20<15:30:16, 15.02s/it]
 62%|██████▏   | 6148/9864 [25:52:33<15:03:18, 14.59s/it]
 62%|██████▏   | 6149/9864 [25:52:47<14:47:59, 14.34s/it]
 62%|██████▏   | 6150/9864 [25:53:02<15:02:00, 14.57s/it]
                                                         
{'loss': '0.1095', 'grad_norm': '5.505', 'learning_rate': '6.562e-06', 'epoch': '0.6235'}

 62%|██████▏   | 6150/9864 [25:53:02<15:02:00, 14.57s/it]
 62%|██████▏   | 6151/9864 [25:53:18<15:21:03, 14.88s/it]
 62%|██████▏   | 6152/9864 [25:53:34<15:34:31, 15.11s/it]
 62%|██████▏   | 6153/9864 [25:53:49<15:45:05, 15.28s/it]
 62%|██████▏   | 6154/9864 [25:54:04<15:24:55, 14.96s/it]
 62%|██████▏   | 6155/9864 [25:54:17<14:57:52, 14.52s/it]
 62%|██████▏   | 6156/9864 [25:54:31<14:52:04, 14.43s/it]
 62%|██████▏   | 6157/9864 [25:54:46<14:48:53, 14.39s/it]
 62%|██████▏   | 6158/9864 [25:55:01<15:17:59, 14.86s/it]
 62%|██████▏   | 6159/9864 [25:55:17<15:33:34, 15.12s/it]
 62%|██████▏   | 6160/9864 [25:55:33<15:47:48, 15.35s/it]
                                                         
{'loss': '0.1243', 'grad_norm': '4.551', 'learning_rate': '6.531e-06', 'epoch': '0.6245'}

 62%|██████▏   | 6160/9864 [25:55:33<15:47:48, 15.35s/it]
 62%|██████▏   | 6161/9864 [25:55:49<15:57:34, 15.52s/it]
 62%|██████▏   | 6162/9864 [25:56:03<15:33:28, 15.13s/it]
 62%|██████▏   | 6163/9864 [25:56:18<15:18:09, 14.88s/it]
 62%|██████▏   | 6164/9864 [25:56:32<15:05:34, 14.68s/it]
 62%|██████▎   | 6165/9864 [25:56:46<14:55:20, 14.52s/it]
 63%|██████▎   | 6166/9864 [25:57:03<15:45:32, 15.34s/it]
 63%|██████▎   | 6167/9864 [25:57:18<15:42:55, 15.30s/it]
 63%|██████▎   | 6168/9864 [25:57:34<15:55:32, 15.51s/it]
 63%|██████▎   | 6169/9864 [25:57:48<15:29:06, 15.09s/it]
 63%|██████▎   | 6170/9864 [25:58:02<15:07:47, 14.74s/it]
                                                         
{'loss': '0.1203', 'grad_norm': '2.61', 'learning_rate': '6.5e-06', 'epoch': '0.6256'}

 63%|██████▎   | 6170/9864 [25:58:02<15:07:47, 14.74s/it]
 63%|██████▎   | 6171/9864 [25:58:16<14:51:57, 14.49s/it]
 63%|██████▎   | 6172/9864 [25:58:31<14:50:49, 14.48s/it]
 63%|██████▎   | 6173/9864 [25:58:46<15:10:10, 14.80s/it]
 63%|██████▎   | 6174/9864 [25:59:02<15:33:47, 15.18s/it]
 63%|██████▎   | 6175/9864 [25:59:20<16:13:50, 15.84s/it]
 63%|██████▎   | 6176/9864 [25:59:33<15:26:17, 15.07s/it]
 63%|██████▎   | 6177/9864 [25:59:47<15:00:12, 14.65s/it]
 63%|██████▎   | 6178/9864 [26:00:00<14:39:46, 14.32s/it]
 63%|██████▎   | 6179/9864 [26:00:14<14:34:26, 14.24s/it]
 63%|██████▎   | 6180/9864 [26:00:30<15:04:08, 14.73s/it]
                                                         
{'loss': '0.1074', 'grad_norm': '3.292', 'learning_rate': '6.469e-06', 'epoch': '0.6266'}

 63%|██████▎   | 6180/9864 [26:00:30<15:04:08, 14.73s/it]
 63%|██████▎   | 6181/9864 [26:00:46<15:24:46, 15.07s/it]
 63%|██████▎   | 6182/9864 [26:01:02<15:39:42, 15.31s/it]
 63%|██████▎   | 6183/9864 [26:01:17<15:39:21, 15.31s/it]
 63%|██████▎   | 6184/9864 [26:01:31<15:19:16, 14.99s/it]
 63%|██████▎   | 6185/9864 [26:01:45<14:59:32, 14.67s/it]
 63%|██████▎   | 6186/9864 [26:01:59<14:48:35, 14.50s/it]
 63%|██████▎   | 6187/9864 [26:02:13<14:35:11, 14.28s/it]
 63%|██████▎   | 6188/9864 [26:02:29<15:02:52, 14.74s/it]
 63%|██████▎   | 6189/9864 [26:02:45<15:20:31, 15.03s/it]
 63%|██████▎   | 6190/9864 [26:03:00<15:20:33, 15.03s/it]
                                                         
{'loss': '0.1332', 'grad_norm': '3.054', 'learning_rate': '6.439e-06', 'epoch': '0.6276'}

 63%|██████▎   | 6190/9864 [26:03:00<15:20:33, 15.03s/it]
 63%|██████▎   | 6191/9864 [26:03:16<15:38:29, 15.33s/it]
 63%|██████▎   | 6192/9864 [26:03:30<15:16:18, 14.97s/it]
 63%|██████▎   | 6193/9864 [26:03:44<14:53:59, 14.61s/it]
 63%|██████▎   | 6194/9864 [26:03:57<14:37:09, 14.34s/it]
 63%|██████▎   | 6195/9864 [26:04:12<14:32:20, 14.27s/it]
 63%|██████▎   | 6196/9864 [26:04:27<14:52:51, 14.61s/it]
 63%|██████▎   | 6197/9864 [26:04:42<15:06:25, 14.83s/it]
 63%|██████▎   | 6198/9864 [26:04:58<15:24:34, 15.13s/it]
 63%|██████▎   | 6199/9864 [26:05:14<15:35:32, 15.32s/it]
 63%|██████▎   | 6200/9864 [26:05:28<15:17:54, 15.03s/it]
                                                         
{'loss': '0.1125', 'grad_norm': '4.435', 'learning_rate': '6.408e-06', 'epoch': '0.6286'}

 63%|██████▎   | 6200/9864 [26:05:28<15:17:54, 15.03s/it]
 63%|██████▎   | 6201/9864 [26:05:42<14:49:30, 14.57s/it]
 63%|██████▎   | 6202/9864 [26:05:56<14:44:10, 14.49s/it]
 63%|██████▎   | 6203/9864 [26:06:12<15:14:33, 14.99s/it]
 63%|██████▎   | 6204/9864 [26:06:28<15:28:36, 15.22s/it]
 63%|██████▎   | 6205/9864 [26:06:42<15:09:09, 14.91s/it]
 63%|██████▎   | 6206/9864 [26:06:58<15:26:53, 15.20s/it]
 63%|██████▎   | 6207/9864 [26:07:13<15:26:43, 15.20s/it]
 63%|██████▎   | 6208/9864 [26:07:27<15:01:56, 14.80s/it]
 63%|██████▎   | 6209/9864 [26:07:41<14:52:16, 14.65s/it]
 63%|██████▎   | 6210/9864 [26:07:56<14:46:33, 14.56s/it]
                                                         
{'loss': '0.1164', 'grad_norm': '2.969', 'learning_rate': '6.377e-06', 'epoch': '0.6296'}

 63%|██████▎   | 6210/9864 [26:07:56<14:46:33, 14.56s/it]
 63%|██████▎   | 6211/9864 [26:08:11<15:07:34, 14.91s/it]
 63%|██████▎   | 6212/9864 [26:08:27<15:22:20, 15.15s/it]
 63%|██████▎   | 6213/9864 [26:08:41<14:58:22, 14.76s/it]
 63%|██████▎   | 6214/9864 [26:08:59<15:52:41, 15.66s/it]
 63%|██████▎   | 6215/9864 [26:09:13<15:27:36, 15.25s/it]
 63%|██████▎   | 6216/9864 [26:09:27<15:03:54, 14.87s/it]
 63%|██████▎   | 6217/9864 [26:09:41<14:49:34, 14.64s/it]
 63%|██████▎   | 6218/9864 [26:09:55<14:42:00, 14.51s/it]
 63%|██████▎   | 6219/9864 [26:10:11<15:01:56, 14.85s/it]
 63%|██████▎   | 6220/9864 [26:10:27<15:20:17, 15.15s/it]
                                                         
{'loss': '0.1235', 'grad_norm': '3.953', 'learning_rate': '6.347e-06', 'epoch': '0.6306'}

 63%|██████▎   | 6220/9864 [26:10:27<15:20:17, 15.15s/it]
 63%|██████▎   | 6221/9864 [26:10:42<15:24:17, 15.22s/it]
 63%|██████▎   | 6222/9864 [26:10:58<15:34:38, 15.40s/it]
 63%|██████▎   | 6223/9864 [26:11:12<15:04:12, 14.90s/it]
 63%|██████▎   | 6224/9864 [26:11:26<14:53:00, 14.72s/it]
 63%|██████▎   | 6225/9864 [26:11:40<14:43:47, 14.57s/it]
 63%|██████▎   | 6226/9864 [26:11:55<14:37:46, 14.48s/it]
 63%|██████▎   | 6227/9864 [26:12:10<15:02:56, 14.90s/it]
 63%|██████▎   | 6228/9864 [26:12:26<15:15:54, 15.11s/it]
 63%|██████▎   | 6229/9864 [26:12:42<15:30:50, 15.36s/it]
 63%|██████▎   | 6230/9864 [26:12:58<15:37:47, 15.48s/it]
                                                         
{'loss': '0.1105', 'grad_norm': '3.519', 'learning_rate': '6.316e-06', 'epoch': '0.6316'}

 63%|██████▎   | 6230/9864 [26:12:58<15:37:47, 15.48s/it]
 63%|██████▎   | 6231/9864 [26:13:12<15:12:16, 15.07s/it]
 63%|██████▎   | 6232/9864 [26:13:25<14:45:01, 14.62s/it]
 63%|██████▎   | 6233/9864 [26:13:39<14:31:05, 14.39s/it]
 63%|██████▎   | 6234/9864 [26:13:54<14:31:49, 14.41s/it]
 63%|██████▎   | 6235/9864 [26:14:09<14:49:37, 14.71s/it]
 63%|██████▎   | 6236/9864 [26:14:25<15:08:13, 15.02s/it]
 63%|██████▎   | 6237/9864 [26:14:40<15:16:05, 15.15s/it]
 63%|██████▎   | 6238/9864 [26:14:54<14:53:08, 14.78s/it]
 63%|██████▎   | 6239/9864 [26:15:09<14:44:14, 14.64s/it]
 63%|██████▎   | 6240/9864 [26:15:22<14:29:09, 14.39s/it]
                                                         
{'loss': '0.1175', 'grad_norm': '2.51', 'learning_rate': '6.286e-06', 'epoch': '0.6327'}

 63%|██████▎   | 6240/9864 [26:15:22<14:29:09, 14.39s/it]
 63%|██████▎   | 6241/9864 [26:15:36<14:23:09, 14.29s/it]
 63%|██████▎   | 6242/9864 [26:15:51<14:23:42, 14.31s/it]
 63%|██████▎   | 6243/9864 [26:16:08<15:13:27, 15.14s/it]
 63%|██████▎   | 6244/9864 [26:16:24<15:29:06, 15.40s/it]
 63%|██████▎   | 6245/9864 [26:16:39<15:26:11, 15.36s/it]
 63%|██████▎   | 6246/9864 [26:16:53<15:04:54, 15.01s/it]
 63%|██████▎   | 6247/9864 [26:17:08<14:52:13, 14.80s/it]
 63%|██████▎   | 6248/9864 [26:17:21<14:30:42, 14.45s/it]
 63%|██████▎   | 6249/9864 [26:17:35<14:25:12, 14.36s/it]
 63%|██████▎   | 6250/9864 [26:17:50<14:23:25, 14.33s/it]
                                                         
{'loss': '0.1077', 'grad_norm': '3.059', 'learning_rate': '6.255e-06', 'epoch': '0.6337'}

 63%|██████▎   | 6250/9864 [26:17:50<14:23:25, 14.33s/it]
 63%|██████▎   | 6251/9864 [26:18:06<15:01:46, 14.98s/it]
 63%|██████▎   | 6252/9864 [26:18:22<15:11:18, 15.14s/it]
 63%|██████▎   | 6253/9864 [26:18:37<15:15:10, 15.21s/it]
 63%|██████▎   | 6254/9864 [26:18:51<14:51:05, 14.81s/it]
 63%|██████▎   | 6255/9864 [26:19:05<14:33:17, 14.52s/it]
 63%|██████▎   | 6256/9864 [26:19:19<14:25:08, 14.39s/it]
 63%|██████▎   | 6257/9864 [26:19:33<14:21:18, 14.33s/it]
 63%|██████▎   | 6258/9864 [26:19:47<14:18:24, 14.28s/it]
 63%|██████▎   | 6259/9864 [26:20:07<15:51:17, 15.83s/it]
 63%|██████▎   | 6260/9864 [26:20:21<15:19:34, 15.31s/it]
                                                         
{'loss': '0.1102', 'grad_norm': '4.517', 'learning_rate': '6.225e-06', 'epoch': '0.6347'}

 63%|██████▎   | 6260/9864 [26:20:21<15:19:34, 15.31s/it]
 63%|██████▎   | 6261/9864 [26:20:37<15:33:12, 15.54s/it]
 63%|██████▎   | 6262/9864 [26:20:51<15:07:39, 15.12s/it]
 63%|██████▎   | 6263/9864 [26:21:05<14:48:07, 14.80s/it]
 64%|██████▎   | 6264/9864 [26:21:19<14:33:09, 14.55s/it]
 64%|██████▎   | 6265/9864 [26:21:33<14:26:16, 14.44s/it]
 64%|██████▎   | 6266/9864 [26:21:48<14:25:05, 14.43s/it]
 64%|██████▎   | 6267/9864 [26:22:07<15:46:06, 15.78s/it]
 64%|██████▎   | 6268/9864 [26:22:21<15:21:13, 15.37s/it]
 64%|██████▎   | 6269/9864 [26:22:37<15:30:38, 15.53s/it]
 64%|██████▎   | 6270/9864 [26:22:51<15:07:06, 15.14s/it]
                                                         
{'loss': '0.1156', 'grad_norm': '3.417', 'learning_rate': '6.194e-06', 'epoch': '0.6357'}

 64%|██████▎   | 6270/9864 [26:22:51<15:07:06, 15.14s/it]
 64%|██████▎   | 6271/9864 [26:23:05<14:50:30, 14.87s/it]
 64%|██████▎   | 6272/9864 [26:23:20<14:37:44, 14.66s/it]
 64%|██████▎   | 6273/9864 [26:23:33<14:21:50, 14.40s/it]
 64%|██████▎   | 6274/9864 [26:23:49<14:48:30, 14.85s/it]
 64%|██████▎   | 6275/9864 [26:24:06<15:28:01, 15.51s/it]
 64%|██████▎   | 6276/9864 [26:24:20<15:00:09, 15.05s/it]
 64%|██████▎   | 6277/9864 [26:24:34<14:43:21, 14.78s/it]
 64%|██████▎   | 6278/9864 [26:24:51<15:11:08, 15.24s/it]
 64%|██████▎   | 6279/9864 [26:25:05<14:51:11, 14.92s/it]
 64%|██████▎   | 6280/9864 [26:25:19<14:28:32, 14.54s/it]
                                                         
{'loss': '0.1129', 'grad_norm': '3.713', 'learning_rate': '6.164e-06', 'epoch': '0.6367'}

 64%|██████▎   | 6280/9864 [26:25:19<14:28:32, 14.54s/it]
 64%|██████▎   | 6281/9864 [26:25:32<14:17:31, 14.36s/it]
 64%|██████▎   | 6282/9864 [26:25:48<14:39:52, 14.74s/it]
 64%|██████▎   | 6283/9864 [26:26:04<14:52:53, 14.96s/it]
 64%|██████▎   | 6284/9864 [26:26:19<15:03:25, 15.14s/it]
 64%|██████▎   | 6285/9864 [26:26:33<14:38:26, 14.73s/it]
 64%|██████▎   | 6286/9864 [26:26:48<14:53:26, 14.98s/it]
 64%|██████▎   | 6287/9864 [26:27:02<14:32:47, 14.64s/it]
 64%|██████▎   | 6288/9864 [26:27:16<14:20:43, 14.44s/it]
 64%|██████▍   | 6289/9864 [26:27:32<14:35:52, 14.70s/it]
 64%|██████▍   | 6290/9864 [26:27:46<14:27:16, 14.56s/it]
                                                         
{'loss': '0.1266', 'grad_norm': '2.993', 'learning_rate': '6.134e-06', 'epoch': '0.6377'}

 64%|██████▍   | 6290/9864 [26:27:46<14:27:16, 14.56s/it]
 64%|██████▍   | 6291/9864 [26:28:04<15:22:55, 15.50s/it]
 64%|██████▍   | 6292/9864 [26:28:18<15:05:17, 15.21s/it]
 64%|██████▍   | 6293/9864 [26:28:34<15:17:27, 15.42s/it]
 64%|██████▍   | 6294/9864 [26:28:49<15:02:43, 15.17s/it]
 64%|██████▍   | 6295/9864 [26:29:03<14:48:31, 14.94s/it]
 64%|██████▍   | 6296/9864 [26:29:18<14:57:29, 15.09s/it]
 64%|██████▍   | 6297/9864 [26:29:33<14:40:27, 14.81s/it]
 64%|██████▍   | 6298/9864 [26:29:46<14:19:37, 14.46s/it]
 64%|██████▍   | 6299/9864 [26:30:04<15:13:50, 15.38s/it]
 64%|██████▍   | 6300/9864 [26:30:18<14:55:14, 15.07s/it]
                                                         
{'loss': '0.1115', 'grad_norm': '3.827', 'learning_rate': '6.104e-06', 'epoch': '0.6387'}

 64%|██████▍   | 6300/9864 [26:30:18<14:55:14, 15.07s/it]
 64%|██████▍   | 6301/9864 [26:30:34<15:08:43, 15.30s/it]
 64%|██████▍   | 6302/9864 [26:30:48<14:50:30, 15.00s/it]
 64%|██████▍   | 6303/9864 [26:31:03<14:39:59, 14.83s/it]
 64%|██████▍   | 6304/9864 [26:31:17<14:31:35, 14.69s/it]
 64%|██████▍   | 6305/9864 [26:31:33<14:51:54, 15.04s/it]
 64%|██████▍   | 6306/9864 [26:31:47<14:31:14, 14.69s/it]
 64%|██████▍   | 6307/9864 [26:32:04<15:10:57, 15.37s/it]
 64%|██████▍   | 6308/9864 [26:32:17<14:39:32, 14.84s/it]
 64%|██████▍   | 6309/9864 [26:32:33<14:53:02, 15.07s/it]
 64%|██████▍   | 6310/9864 [26:32:47<14:38:13, 14.83s/it]
                                                         
{'loss': '0.1219', 'grad_norm': '3.507', 'learning_rate': '6.073e-06', 'epoch': '0.6398'}

 64%|██████▍   | 6310/9864 [26:32:47<14:38:13, 14.83s/it]
 64%|██████▍   | 6311/9864 [26:33:01<14:24:26, 14.60s/it]
 64%|██████▍   | 6312/9864 [26:33:15<14:16:52, 14.47s/it]
 64%|██████▍   | 6313/9864 [26:33:31<14:39:04, 14.85s/it]
 64%|██████▍   | 6314/9864 [26:33:46<14:32:23, 14.74s/it]
 64%|██████▍   | 6315/9864 [26:34:03<15:17:26, 15.51s/it]
 64%|██████▍   | 6316/9864 [26:34:17<14:54:15, 15.12s/it]
 64%|██████▍   | 6317/9864 [26:34:33<15:05:30, 15.32s/it]
 64%|██████▍   | 6318/9864 [26:34:46<14:33:32, 14.78s/it]
 64%|██████▍   | 6319/9864 [26:35:00<14:15:31, 14.48s/it]
 64%|██████▍   | 6320/9864 [26:35:16<14:37:45, 14.86s/it]
                                                         
{'loss': '0.1129', 'grad_norm': '3.001', 'learning_rate': '6.043e-06', 'epoch': '0.6408'}

 64%|██████▍   | 6320/9864 [26:35:16<14:37:45, 14.86s/it]
 64%|██████▍   | 6321/9864 [26:35:30<14:28:01, 14.70s/it]
 64%|██████▍   | 6322/9864 [26:35:46<14:43:52, 14.97s/it]
 64%|██████▍   | 6323/9864 [26:36:01<14:50:46, 15.09s/it]
 64%|██████▍   | 6324/9864 [26:36:16<14:43:57, 14.98s/it]
 64%|██████▍   | 6325/9864 [26:36:32<15:00:35, 15.27s/it]
 64%|██████▍   | 6326/9864 [26:36:46<14:42:09, 14.96s/it]
 64%|██████▍   | 6327/9864 [26:37:01<14:32:20, 14.80s/it]
 64%|██████▍   | 6328/9864 [26:37:14<14:15:53, 14.52s/it]
 64%|██████▍   | 6329/9864 [26:37:30<14:30:24, 14.77s/it]
 64%|██████▍   | 6330/9864 [26:37:47<15:14:52, 15.53s/it]
                                                         
{'loss': '0.1093', 'grad_norm': '4.304', 'learning_rate': '6.013e-06', 'epoch': '0.6418'}

 64%|██████▍   | 6330/9864 [26:37:47<15:14:52, 15.53s/it]
 64%|██████▍   | 6331/9864 [26:38:01<14:50:42, 15.13s/it]
 64%|██████▍   | 6332/9864 [26:38:17<15:04:01, 15.36s/it]
 64%|██████▍   | 6333/9864 [26:38:32<14:54:09, 15.19s/it]
 64%|██████▍   | 6334/9864 [26:38:46<14:28:20, 14.76s/it]
 64%|██████▍   | 6335/9864 [26:39:00<14:12:42, 14.50s/it]
 64%|██████▍   | 6336/9864 [26:39:13<13:59:27, 14.28s/it]
 64%|██████▍   | 6337/9864 [26:39:29<14:27:16, 14.75s/it]
 64%|██████▍   | 6338/9864 [26:39:47<15:13:27, 15.54s/it]
 64%|██████▍   | 6339/9864 [26:40:01<14:54:58, 15.23s/it]
 64%|██████▍   | 6340/9864 [26:40:15<14:36:08, 14.92s/it]
                                                         
{'loss': '0.1205', 'grad_norm': '4.966', 'learning_rate': '5.983e-06', 'epoch': '0.6428'}

 64%|██████▍   | 6340/9864 [26:40:15<14:36:08, 14.92s/it]
 64%|██████▍   | 6341/9864 [26:40:31<14:49:09, 15.14s/it]
 64%|██████▍   | 6342/9864 [26:40:45<14:32:59, 14.87s/it]
 64%|██████▍   | 6343/9864 [26:41:00<14:22:50, 14.70s/it]
 64%|██████▍   | 6344/9864 [26:41:15<14:40:11, 15.00s/it]
 64%|██████▍   | 6345/9864 [26:41:29<14:22:01, 14.70s/it]
 64%|██████▍   | 6346/9864 [26:41:43<14:07:01, 14.45s/it]
 64%|██████▍   | 6347/9864 [26:42:01<15:01:31, 15.38s/it]
 64%|██████▍   | 6348/9864 [26:42:14<14:23:18, 14.73s/it]
 64%|██████▍   | 6349/9864 [26:42:30<14:44:44, 15.10s/it]
 64%|██████▍   | 6350/9864 [26:42:44<14:30:51, 14.87s/it]
                                                         
{'loss': '0.1082', 'grad_norm': '3.561', 'learning_rate': '5.953e-06', 'epoch': '0.6438'}

 64%|██████▍   | 6350/9864 [26:42:44<14:30:51, 14.87s/it]
 64%|██████▍   | 6351/9864 [26:43:00<14:49:02, 15.18s/it]
 64%|██████▍   | 6352/9864 [26:43:14<14:25:33, 14.79s/it]
 64%|██████▍   | 6353/9864 [26:43:28<14:12:06, 14.56s/it]
 64%|██████▍   | 6354/9864 [26:43:42<14:02:17, 14.40s/it]
 64%|██████▍   | 6355/9864 [26:44:00<14:56:03, 15.32s/it]
 64%|██████▍   | 6356/9864 [26:44:15<14:58:07, 15.36s/it]
 64%|██████▍   | 6357/9864 [26:44:29<14:26:11, 14.82s/it]
 64%|██████▍   | 6358/9864 [26:44:44<14:40:03, 15.06s/it]
 64%|██████▍   | 6359/9864 [26:44:58<14:24:50, 14.80s/it]
 64%|██████▍   | 6360/9864 [26:45:13<14:21:02, 14.74s/it]
                                                         
{'loss': '0.112', 'grad_norm': '4.201', 'learning_rate': '5.923e-06', 'epoch': '0.6448'}

 64%|██████▍   | 6360/9864 [26:45:13<14:21:02, 14.74s/it]
 64%|██████▍   | 6361/9864 [26:45:27<14:11:30, 14.58s/it]
 64%|██████▍   | 6362/9864 [26:45:43<14:26:43, 14.85s/it]
 65%|██████▍   | 6363/9864 [26:45:57<14:16:45, 14.68s/it]
 65%|██████▍   | 6364/9864 [26:46:13<14:36:50, 15.03s/it]
 65%|██████▍   | 6365/9864 [26:46:27<14:17:52, 14.71s/it]
 65%|██████▍   | 6366/9864 [26:46:42<14:35:22, 15.01s/it]
 65%|██████▍   | 6367/9864 [26:46:56<14:07:53, 14.55s/it]
 65%|██████▍   | 6368/9864 [26:47:10<13:57:04, 14.37s/it]
 65%|██████▍   | 6369/9864 [26:47:24<13:56:12, 14.36s/it]
 65%|██████▍   | 6370/9864 [26:47:38<13:44:52, 14.16s/it]
                                                         
{'loss': '0.1123', 'grad_norm': '5.024', 'learning_rate': '5.893e-06', 'epoch': '0.6458'}

 65%|██████▍   | 6370/9864 [26:47:38<13:44:52, 14.16s/it]
 65%|██████▍   | 6371/9864 [26:47:54<14:12:36, 14.65s/it]
 65%|██████▍   | 6372/9864 [26:48:09<14:22:57, 14.83s/it]
 65%|██████▍   | 6373/9864 [26:48:25<14:35:44, 15.05s/it]
 65%|██████▍   | 6374/9864 [26:48:40<14:51:36, 15.33s/it]
 65%|██████▍   | 6375/9864 [26:48:55<14:30:12, 14.96s/it]
 65%|██████▍   | 6376/9864 [26:49:08<14:08:53, 14.60s/it]
 65%|██████▍   | 6377/9864 [26:49:22<13:46:15, 14.22s/it]
 65%|██████▍   | 6378/9864 [26:49:36<13:43:52, 14.18s/it]
 65%|██████▍   | 6379/9864 [26:49:51<14:06:30, 14.57s/it]
 65%|██████▍   | 6380/9864 [26:50:08<14:51:47, 15.36s/it]
                                                         
{'loss': '0.1108', 'grad_norm': '3.008', 'learning_rate': '5.863e-06', 'epoch': '0.6468'}

 65%|██████▍   | 6380/9864 [26:50:08<14:51:47, 15.36s/it]
 65%|██████▍   | 6381/9864 [26:50:23<14:28:55, 14.97s/it]
 65%|██████▍   | 6382/9864 [26:50:38<14:41:07, 15.18s/it]
 65%|██████▍   | 6383/9864 [26:50:52<14:13:27, 14.71s/it]
 65%|██████▍   | 6384/9864 [26:51:06<14:01:17, 14.50s/it]
 65%|██████▍   | 6385/9864 [26:51:20<13:56:38, 14.43s/it]
 65%|██████▍   | 6386/9864 [26:51:34<13:43:58, 14.21s/it]
 65%|██████▍   | 6387/9864 [26:51:51<14:35:51, 15.11s/it]
 65%|██████▍   | 6388/9864 [26:52:06<14:40:18, 15.20s/it]
 65%|██████▍   | 6389/9864 [26:52:20<14:15:36, 14.77s/it]
 65%|██████▍   | 6390/9864 [26:52:36<14:29:40, 15.02s/it]
                                                         
{'loss': '0.1036', 'grad_norm': '3.097', 'learning_rate': '5.833e-06', 'epoch': '0.6479'}

 65%|██████▍   | 6390/9864 [26:52:36<14:29:40, 15.02s/it]
 65%|██████▍   | 6391/9864 [26:52:50<14:08:53, 14.67s/it]
 65%|██████▍   | 6392/9864 [26:53:04<14:07:55, 14.65s/it]
 65%|██████▍   | 6393/9864 [26:53:18<13:56:17, 14.46s/it]
 65%|██████▍   | 6394/9864 [26:53:34<14:19:40, 14.86s/it]
 65%|██████▍   | 6395/9864 [26:53:50<14:31:55, 15.08s/it]
 65%|██████▍   | 6396/9864 [26:54:06<14:49:41, 15.39s/it]
 65%|██████▍   | 6397/9864 [26:54:21<14:51:08, 15.42s/it]
 65%|██████▍   | 6398/9864 [26:54:35<14:26:00, 14.99s/it]
 65%|██████▍   | 6399/9864 [26:54:49<13:59:45, 14.54s/it]
 65%|██████▍   | 6400/9864 [26:55:03<13:48:31, 14.35s/it]
                                                         
{'loss': '0.1185', 'grad_norm': '4.345', 'learning_rate': '5.803e-06', 'epoch': '0.6489'}

 65%|██████▍   | 6400/9864 [26:55:03<13:48:31, 14.35s/it]
 65%|██████▍   | 6401/9864 [26:55:18<14:05:15, 14.65s/it]
 65%|██████▍   | 6402/9864 [26:55:32<13:54:04, 14.46s/it]
 65%|██████▍   | 6403/9864 [26:55:46<13:38:40, 14.19s/it]
 65%|██████▍   | 6404/9864 [26:56:03<14:33:06, 15.14s/it]
 65%|██████▍   | 6405/9864 [26:56:18<14:36:03, 15.20s/it]
 65%|██████▍   | 6406/9864 [26:56:32<14:15:57, 14.85s/it]
 65%|██████▍   | 6407/9864 [26:56:46<14:03:46, 14.64s/it]
 65%|██████▍   | 6408/9864 [26:57:00<13:50:30, 14.42s/it]
 65%|██████▍   | 6409/9864 [26:57:16<14:18:47, 14.91s/it]
 65%|██████▍   | 6410/9864 [26:57:30<13:58:49, 14.57s/it]
                                                         
{'loss': '0.09968', 'grad_norm': '2.603', 'learning_rate': '5.773e-06', 'epoch': '0.6499'}

 65%|██████▍   | 6410/9864 [26:57:30<13:58:49, 14.57s/it]
 65%|██████▍   | 6411/9864 [26:57:45<14:10:42, 14.78s/it]
 65%|██████▌   | 6412/9864 [26:58:03<14:52:53, 15.52s/it]
 65%|██████▌   | 6413/9864 [26:58:17<14:24:52, 15.04s/it]
 65%|██████▌   | 6414/9864 [26:58:31<14:17:26, 14.91s/it]
 65%|██████▌   | 6415/9864 [26:58:45<14:00:45, 14.63s/it]
 65%|██████▌   | 6416/9864 [26:59:01<14:12:58, 14.84s/it]
 65%|██████▌   | 6417/9864 [26:59:15<13:59:49, 14.62s/it]
 65%|██████▌   | 6418/9864 [26:59:29<13:52:48, 14.50s/it]
 65%|██████▌   | 6419/9864 [26:59:45<14:15:10, 14.89s/it]
 65%|██████▌   | 6420/9864 [27:00:02<15:03:26, 15.74s/it]
                                                         
{'loss': '0.1151', 'grad_norm': '3.106', 'learning_rate': '5.744e-06', 'epoch': '0.6509'}

 65%|██████▌   | 6420/9864 [27:00:02<15:03:26, 15.74s/it]
 65%|██████▌   | 6421/9864 [27:00:17<14:37:42, 15.30s/it]
 65%|██████▌   | 6422/9864 [27:00:31<14:20:28, 15.00s/it]
 65%|██████▌   | 6423/9864 [27:00:45<14:08:45, 14.80s/it]
 65%|██████▌   | 6424/9864 [27:01:00<14:01:46, 14.68s/it]
 65%|██████▌   | 6425/9864 [27:01:16<14:26:32, 15.12s/it]
 65%|██████▌   | 6426/9864 [27:01:31<14:33:50, 15.25s/it]
 65%|██████▌   | 6427/9864 [27:01:46<14:14:39, 14.92s/it]
 65%|██████▌   | 6428/9864 [27:02:03<15:01:04, 15.73s/it]
 65%|██████▌   | 6429/9864 [27:02:17<14:26:04, 15.13s/it]
 65%|██████▌   | 6430/9864 [27:02:31<14:01:39, 14.71s/it]
                                                         
{'loss': '0.09169', 'grad_norm': '3.692', 'learning_rate': '5.714e-06', 'epoch': '0.6519'}

 65%|██████▌   | 6430/9864 [27:02:31<14:01:39, 14.71s/it]
 65%|██████▌   | 6431/9864 [27:02:44<13:46:47, 14.45s/it]
 65%|██████▌   | 6432/9864 [27:02:59<13:45:03, 14.42s/it]
 65%|██████▌   | 6433/9864 [27:03:14<14:04:27, 14.77s/it]
 65%|██████▌   | 6434/9864 [27:03:30<14:23:44, 15.11s/it]
 65%|██████▌   | 6435/9864 [27:03:46<14:38:01, 15.36s/it]
 65%|██████▌   | 6436/9864 [27:04:02<14:47:06, 15.53s/it]
 65%|██████▌   | 6437/9864 [27:04:16<14:23:04, 15.11s/it]
 65%|██████▌   | 6438/9864 [27:04:31<14:08:37, 14.86s/it]
 65%|██████▌   | 6439/9864 [27:04:45<14:06:07, 14.82s/it]
 65%|██████▌   | 6440/9864 [27:05:01<14:25:14, 15.16s/it]
                                                         
{'loss': '0.1162', 'grad_norm': '4.18', 'learning_rate': '5.684e-06', 'epoch': '0.6529'}

 65%|██████▌   | 6440/9864 [27:05:01<14:25:14, 15.16s/it]
 65%|██████▌   | 6441/9864 [27:05:17<14:42:14, 15.46s/it]
 65%|██████▌   | 6442/9864 [27:05:33<14:46:56, 15.55s/it]
 65%|██████▌   | 6443/9864 [27:05:49<14:55:07, 15.70s/it]
 65%|██████▌   | 6444/9864 [27:06:03<14:29:22, 15.25s/it]
 65%|██████▌   | 6445/9864 [27:06:17<14:03:43, 14.81s/it]
 65%|██████▌   | 6446/9864 [27:06:31<13:51:10, 14.59s/it]
 65%|██████▌   | 6447/9864 [27:06:46<13:45:38, 14.50s/it]
 65%|██████▌   | 6448/9864 [27:07:02<14:10:19, 14.94s/it]
 65%|██████▌   | 6449/9864 [27:07:17<14:24:12, 15.18s/it]
 65%|██████▌   | 6450/9864 [27:07:33<14:41:05, 15.49s/it]
                                                         
{'loss': '0.1002', 'grad_norm': '3.318', 'learning_rate': '5.655e-06', 'epoch': '0.6539'}

 65%|██████▌   | 6450/9864 [27:07:33<14:41:05, 15.49s/it]
 65%|██████▌   | 6451/9864 [27:07:49<14:48:11, 15.61s/it]
 65%|██████▌   | 6452/9864 [27:08:03<14:17:21, 15.08s/it]
 65%|██████▌   | 6453/9864 [27:08:17<14:00:55, 14.79s/it]
 65%|██████▌   | 6454/9864 [27:08:32<13:57:14, 14.73s/it]
 65%|██████▌   | 6455/9864 [27:08:46<13:52:03, 14.64s/it]
 65%|██████▌   | 6456/9864 [27:09:04<14:37:28, 15.45s/it]
 65%|██████▌   | 6457/9864 [27:09:18<14:12:38, 15.02s/it]
 65%|██████▌   | 6458/9864 [27:09:32<13:55:53, 14.72s/it]
 65%|██████▌   | 6459/9864 [27:09:50<14:50:23, 15.69s/it]
 65%|██████▌   | 6460/9864 [27:10:04<14:27:10, 15.28s/it]
                                                         
{'loss': '0.107', 'grad_norm': '4.341', 'learning_rate': '5.625e-06', 'epoch': '0.655'}

 65%|██████▌   | 6460/9864 [27:10:04<14:27:10, 15.28s/it]
 66%|██████▌   | 6461/9864 [27:10:18<13:57:55, 14.77s/it]
 66%|██████▌   | 6462/9864 [27:10:32<13:48:36, 14.61s/it]
 66%|██████▌   | 6463/9864 [27:10:46<13:45:07, 14.56s/it]
 66%|██████▌   | 6464/9864 [27:11:02<14:09:21, 14.99s/it]
 66%|██████▌   | 6465/9864 [27:11:18<14:25:44, 15.28s/it]
 66%|██████▌   | 6466/9864 [27:11:33<14:09:48, 15.01s/it]
 66%|██████▌   | 6467/9864 [27:11:50<14:48:21, 15.69s/it]
 66%|██████▌   | 6468/9864 [27:12:04<14:26:31, 15.31s/it]
 66%|██████▌   | 6469/9864 [27:12:19<14:07:31, 14.98s/it]
 66%|██████▌   | 6470/9864 [27:12:32<13:44:58, 14.58s/it]
                                                         
{'loss': '0.1123', 'grad_norm': '2.846', 'learning_rate': '5.596e-06', 'epoch': '0.656'}

 66%|██████▌   | 6470/9864 [27:12:32<13:44:58, 14.58s/it]
 66%|██████▌   | 6471/9864 [27:12:48<13:59:33, 14.85s/it]
 66%|██████▌   | 6472/9864 [27:13:04<14:19:24, 15.20s/it]
 66%|██████▌   | 6473/9864 [27:13:18<14:08:38, 15.02s/it]
 66%|██████▌   | 6474/9864 [27:13:33<13:56:47, 14.81s/it]
 66%|██████▌   | 6475/9864 [27:13:50<14:43:09, 15.64s/it]
 66%|██████▌   | 6476/9864 [27:14:04<14:07:19, 15.01s/it]
 66%|██████▌   | 6477/9864 [27:14:18<13:53:27, 14.76s/it]
 66%|██████▌   | 6478/9864 [27:14:32<13:45:05, 14.62s/it]
 66%|██████▌   | 6479/9864 [27:14:50<14:36:28, 15.54s/it]
 66%|██████▌   | 6480/9864 [27:15:04<14:15:24, 15.17s/it]
                                                         
{'loss': '0.1096', 'grad_norm': '4.011', 'learning_rate': '5.566e-06', 'epoch': '0.657'}

 66%|██████▌   | 6480/9864 [27:15:04<14:15:24, 15.17s/it]
 66%|██████▌   | 6481/9864 [27:15:17<13:36:56, 14.49s/it]
 66%|██████▌   | 6482/9864 [27:15:31<13:22:47, 14.24s/it]
 66%|██████▌   | 6483/9864 [27:15:49<14:24:36, 15.34s/it]
 66%|██████▌   | 6484/9864 [27:16:03<14:07:57, 15.05s/it]
 66%|██████▌   | 6485/9864 [27:16:17<13:50:47, 14.75s/it]
 66%|██████▌   | 6486/9864 [27:16:31<13:40:44, 14.58s/it]
 66%|██████▌   | 6487/9864 [27:16:47<14:06:00, 15.03s/it]
 66%|██████▌   | 6488/9864 [27:17:03<14:21:42, 15.31s/it]
 66%|██████▌   | 6489/9864 [27:17:17<14:02:21, 14.98s/it]
 66%|██████▌   | 6490/9864 [27:17:32<13:49:15, 14.75s/it]
                                                         
{'loss': '0.1116', 'grad_norm': '4.013', 'learning_rate': '5.537e-06', 'epoch': '0.658'}

 66%|██████▌   | 6490/9864 [27:17:32<13:49:15, 14.75s/it]
 66%|██████▌   | 6491/9864 [27:17:49<14:36:37, 15.59s/it]
 66%|██████▌   | 6492/9864 [27:18:04<14:15:14, 15.22s/it]
 66%|██████▌   | 6493/9864 [27:18:18<13:56:36, 14.89s/it]
 66%|██████▌   | 6494/9864 [27:18:32<13:47:32, 14.73s/it]
 66%|██████▌   | 6495/9864 [27:18:50<14:34:23, 15.57s/it]
 66%|██████▌   | 6496/9864 [27:19:04<14:15:23, 15.24s/it]
 66%|██████▌   | 6497/9864 [27:19:18<13:54:19, 14.87s/it]
 66%|██████▌   | 6498/9864 [27:19:34<14:05:12, 15.07s/it]
 66%|██████▌   | 6499/9864 [27:19:49<14:18:39, 15.31s/it]
 66%|██████▌   | 6500/9864 [27:20:03<13:54:50, 14.89s/it]
                                                         
{'loss': '0.1062', 'grad_norm': '2.934', 'learning_rate': '5.508e-06', 'epoch': '0.659'}

 66%|██████▌   | 6500/9864 [27:20:03<13:54:50, 14.89s/it]
 66%|██████▌   | 6501/9864 [27:20:17<13:32:05, 14.49s/it]
 66%|██████▌   | 6502/9864 [27:20:33<13:49:35, 14.81s/it]
 66%|██████▌   | 6503/9864 [27:20:47<13:38:49, 14.62s/it]
 66%|██████▌   | 6504/9864 [27:21:02<13:57:06, 14.95s/it]
 66%|██████▌   | 6505/9864 [27:21:16<13:41:16, 14.67s/it]
 66%|██████▌   | 6506/9864 [27:21:34<14:35:20, 15.64s/it]
 66%|██████▌   | 6507/9864 [27:21:49<14:11:38, 15.22s/it]
 66%|██████▌   | 6508/9864 [27:22:03<13:54:39, 14.92s/it]
 66%|██████▌   | 6509/9864 [27:22:19<14:11:57, 15.24s/it]
 66%|██████▌   | 6510/9864 [27:22:33<13:57:45, 14.99s/it]
                                                         
{'loss': '0.1054', 'grad_norm': '4.472', 'learning_rate': '5.478e-06', 'epoch': '0.66'}

 66%|██████▌   | 6510/9864 [27:22:33<13:57:45, 14.99s/it]
 66%|██████▌   | 6511/9864 [27:22:49<14:13:19, 15.27s/it]
 66%|██████▌   | 6512/9864 [27:23:03<13:53:53, 14.93s/it]
 66%|██████▌   | 6513/9864 [27:23:18<13:59:24, 15.03s/it]
 66%|██████▌   | 6514/9864 [27:23:34<14:09:34, 15.22s/it]
 66%|██████▌   | 6515/9864 [27:23:48<13:42:47, 14.74s/it]
 66%|██████▌   | 6516/9864 [27:24:02<13:36:21, 14.63s/it]
 66%|██████▌   | 6517/9864 [27:24:18<14:01:56, 15.09s/it]
 66%|██████▌   | 6518/9864 [27:24:32<13:46:10, 14.81s/it]
 66%|██████▌   | 6519/9864 [27:24:48<13:58:14, 15.04s/it]
 66%|██████▌   | 6520/9864 [27:25:03<13:53:09, 14.95s/it]
                                                         
{'loss': '0.1102', 'grad_norm': '3.824', 'learning_rate': '5.449e-06', 'epoch': '0.661'}

 66%|██████▌   | 6520/9864 [27:25:03<13:53:09, 14.95s/it]
 66%|██████▌   | 6521/9864 [27:25:20<14:36:03, 15.72s/it]
 66%|██████▌   | 6522/9864 [27:25:35<14:13:40, 15.33s/it]
 66%|██████▌   | 6523/9864 [27:25:49<13:53:02, 14.96s/it]
 66%|██████▌   | 6524/9864 [27:26:03<13:40:06, 14.73s/it]
 66%|██████▌   | 6525/9864 [27:26:19<14:00:46, 15.11s/it]
 66%|██████▌   | 6526/9864 [27:26:35<14:12:39, 15.33s/it]
 66%|██████▌   | 6527/9864 [27:26:49<13:49:07, 14.91s/it]
 66%|██████▌   | 6528/9864 [27:27:03<13:36:57, 14.69s/it]
 66%|██████▌   | 6529/9864 [27:27:21<14:26:00, 15.58s/it]
 66%|██████▌   | 6530/9864 [27:27:35<14:03:03, 15.17s/it]
                                                         
{'loss': '0.1062', 'grad_norm': '3.117', 'learning_rate': '5.42e-06', 'epoch': '0.6621'}

 66%|██████▌   | 6530/9864 [27:27:35<14:03:03, 15.17s/it]
 66%|██████▌   | 6531/9864 [27:27:49<13:50:47, 14.96s/it]
 66%|██████▌   | 6532/9864 [27:28:03<13:27:38, 14.54s/it]
 66%|██████▌   | 6533/9864 [27:28:19<13:54:01, 15.02s/it]
 66%|██████▌   | 6534/9864 [27:28:34<14:00:05, 15.14s/it]
 66%|██████▋   | 6535/9864 [27:28:49<13:42:30, 14.82s/it]
 66%|██████▋   | 6536/9864 [27:29:02<13:21:43, 14.45s/it]
 66%|██████▋   | 6537/9864 [27:29:19<14:08:32, 15.30s/it]
 66%|██████▋   | 6538/9864 [27:29:34<13:50:52, 14.99s/it]
 66%|██████▋   | 6539/9864 [27:29:48<13:34:35, 14.70s/it]
 66%|██████▋   | 6540/9864 [27:30:04<13:59:20, 15.15s/it]
                                                         
{'loss': '0.1177', 'grad_norm': '3.252', 'learning_rate': '5.391e-06', 'epoch': '0.6631'}

 66%|██████▋   | 6540/9864 [27:30:04<13:59:20, 15.15s/it]
 66%|██████▋   | 6541/9864 [27:30:20<14:10:50, 15.36s/it]
 66%|██████▋   | 6542/9864 [27:30:34<13:44:44, 14.90s/it]
 66%|██████▋   | 6543/9864 [27:30:48<13:36:36, 14.75s/it]
 66%|██████▋   | 6544/9864 [27:31:02<13:27:43, 14.60s/it]
 66%|██████▋   | 6545/9864 [27:31:20<14:21:53, 15.58s/it]
 66%|██████▋   | 6546/9864 [27:31:34<13:54:32, 15.09s/it]
 66%|██████▋   | 6547/9864 [27:31:48<13:40:45, 14.85s/it]
 66%|██████▋   | 6548/9864 [27:32:03<13:30:21, 14.66s/it]
 66%|██████▋   | 6549/9864 [27:32:19<13:52:57, 15.08s/it]
 66%|██████▋   | 6550/9864 [27:32:34<14:05:57, 15.32s/it]
                                                         
{'loss': '0.1114', 'grad_norm': '2.606', 'learning_rate': '5.361e-06', 'epoch': '0.6641'}

 66%|██████▋   | 6550/9864 [27:32:34<14:05:57, 15.32s/it]
 66%|██████▋   | 6551/9864 [27:32:48<13:43:50, 14.92s/it]
 66%|██████▋   | 6552/9864 [27:33:05<14:02:58, 15.27s/it]
 66%|██████▋   | 6553/9864 [27:33:21<14:17:17, 15.54s/it]
 66%|██████▋   | 6554/9864 [27:33:35<14:00:04, 15.23s/it]
 66%|██████▋   | 6555/9864 [27:33:49<13:34:10, 14.76s/it]
 66%|██████▋   | 6556/9864 [27:34:03<13:26:37, 14.63s/it]
 66%|██████▋   | 6557/9864 [27:34:20<13:58:39, 15.22s/it]
 66%|██████▋   | 6558/9864 [27:34:35<14:05:26, 15.34s/it]
 66%|██████▋   | 6559/9864 [27:34:49<13:38:41, 14.86s/it]
 67%|██████▋   | 6560/9864 [27:35:05<13:59:55, 15.25s/it]
                                                         
{'loss': '0.1042', 'grad_norm': '3.35', 'learning_rate': '5.332e-06', 'epoch': '0.6651'}

 67%|██████▋   | 6560/9864 [27:35:05<13:59:55, 15.25s/it]
 67%|██████▋   | 6561/9864 [27:35:22<14:19:34, 15.61s/it]
 67%|██████▋   | 6562/9864 [27:35:36<13:55:31, 15.18s/it]
 67%|██████▋   | 6563/9864 [27:35:50<13:39:59, 14.90s/it]
 67%|██████▋   | 6564/9864 [27:36:06<13:54:21, 15.17s/it]
 67%|██████▋   | 6565/9864 [27:36:20<13:42:44, 14.96s/it]
 67%|██████▋   | 6566/9864 [27:36:36<13:58:05, 15.25s/it]
 67%|██████▋   | 6567/9864 [27:36:51<13:41:37, 14.95s/it]
 67%|██████▋   | 6568/9864 [27:37:06<13:54:09, 15.18s/it]
 67%|██████▋   | 6569/9864 [27:37:22<14:07:36, 15.43s/it]
 67%|██████▋   | 6570/9864 [27:37:36<13:40:44, 14.95s/it]
                                                         
{'loss': '0.1053', 'grad_norm': '4.169', 'learning_rate': '5.303e-06', 'epoch': '0.6661'}

 67%|██████▋   | 6570/9864 [27:37:36<13:40:44, 14.95s/it]
 67%|██████▋   | 6571/9864 [27:37:50<13:23:37, 14.64s/it]
 67%|██████▋   | 6572/9864 [27:38:06<13:46:02, 15.06s/it]
 67%|██████▋   | 6573/9864 [27:38:22<13:58:27, 15.29s/it]
 67%|██████▋   | 6574/9864 [27:38:36<13:41:33, 14.98s/it]
 67%|██████▋   | 6575/9864 [27:38:51<13:30:49, 14.79s/it]
 67%|██████▋   | 6576/9864 [27:39:07<13:54:51, 15.23s/it]
 67%|██████▋   | 6577/9864 [27:39:23<14:14:28, 15.60s/it]
 67%|██████▋   | 6578/9864 [27:39:37<13:47:36, 15.11s/it]
 67%|██████▋   | 6579/9864 [27:39:51<13:30:35, 14.81s/it]
 67%|██████▋   | 6580/9864 [27:40:08<13:52:07, 15.20s/it]
                                                         
{'loss': '0.1198', 'grad_norm': '5.181', 'learning_rate': '5.274e-06', 'epoch': '0.6671'}

 67%|██████▋   | 6580/9864 [27:40:08<13:52:07, 15.20s/it]
 67%|██████▋   | 6581/9864 [27:40:22<13:36:06, 14.92s/it]
 67%|██████▋   | 6582/9864 [27:40:36<13:21:26, 14.65s/it]
 67%|██████▋   | 6583/9864 [27:40:51<13:35:57, 14.92s/it]
 67%|██████▋   | 6584/9864 [27:41:07<13:51:43, 15.21s/it]
 67%|██████▋   | 6585/9864 [27:41:21<13:30:45, 14.84s/it]
 67%|██████▋   | 6586/9864 [27:41:36<13:23:09, 14.70s/it]
 67%|██████▋   | 6587/9864 [27:41:52<13:47:14, 15.15s/it]
 67%|██████▋   | 6588/9864 [27:42:06<13:30:44, 14.85s/it]
 67%|██████▋   | 6589/9864 [27:42:22<13:52:12, 15.25s/it]
 67%|██████▋   | 6590/9864 [27:42:36<13:26:53, 14.79s/it]
                                                         
{'loss': '0.105', 'grad_norm': '2.941', 'learning_rate': '5.246e-06', 'epoch': '0.6681'}

 67%|██████▋   | 6590/9864 [27:42:36<13:26:53, 14.79s/it]
 67%|██████▋   | 6591/9864 [27:42:52<13:46:00, 15.14s/it]
 67%|██████▋   | 6592/9864 [27:43:08<13:59:31, 15.39s/it]
 67%|██████▋   | 6593/9864 [27:43:22<13:40:28, 15.05s/it]
 67%|██████▋   | 6594/9864 [27:43:37<13:38:31, 15.02s/it]
 67%|██████▋   | 6595/9864 [27:43:51<13:21:18, 14.71s/it]
 67%|██████▋   | 6596/9864 [27:44:09<14:09:15, 15.59s/it]
 67%|██████▋   | 6597/9864 [27:44:22<13:36:12, 14.99s/it]
 67%|██████▋   | 6598/9864 [27:44:36<13:22:10, 14.74s/it]
 67%|██████▋   | 6599/9864 [27:44:54<14:07:17, 15.57s/it]
 67%|██████▋   | 6600/9864 [27:45:08<13:37:44, 15.03s/it]
                                                         
{'loss': '0.09505', 'grad_norm': '3.876', 'learning_rate': '5.217e-06', 'epoch': '0.6692'}

 67%|██████▋   | 6600/9864 [27:45:08<13:37:44, 15.03s/it]
 67%|██████▋   | 6601/9864 [27:45:22<13:23:19, 14.77s/it]
 67%|██████▋   | 6602/9864 [27:45:36<13:15:02, 14.62s/it]
 67%|██████▋   | 6603/9864 [27:45:52<13:30:29, 14.91s/it]
 67%|██████▋   | 6604/9864 [27:46:07<13:43:46, 15.16s/it]
 67%|██████▋   | 6605/9864 [27:46:22<13:29:06, 14.90s/it]
 67%|██████▋   | 6606/9864 [27:46:38<13:46:33, 15.22s/it]
 67%|██████▋   | 6607/9864 [27:46:53<13:56:17, 15.41s/it]
 67%|██████▋   | 6608/9864 [27:47:08<13:34:32, 15.01s/it]
 67%|██████▋   | 6609/9864 [27:47:22<13:22:04, 14.78s/it]
 67%|██████▋   | 6610/9864 [27:47:37<13:22:04, 14.79s/it]
                                                         
{'loss': '0.09402', 'grad_norm': '3.227', 'learning_rate': '5.188e-06', 'epoch': '0.6702'}

 67%|██████▋   | 6610/9864 [27:47:37<13:22:04, 14.79s/it]
 67%|██████▋   | 6611/9864 [27:47:54<14:08:09, 15.64s/it]
 67%|██████▋   | 6612/9864 [27:48:08<13:43:35, 15.20s/it]
 67%|██████▋   | 6613/9864 [27:48:22<13:23:55, 14.84s/it]
 67%|██████▋   | 6614/9864 [27:48:40<14:13:06, 15.75s/it]
 67%|██████▋   | 6615/9864 [27:48:54<13:43:13, 15.20s/it]
 67%|██████▋   | 6616/9864 [27:49:09<13:33:01, 15.02s/it]
 67%|██████▋   | 6617/9864 [27:49:23<13:23:20, 14.84s/it]
 67%|██████▋   | 6618/9864 [27:49:37<13:05:04, 14.51s/it]
 67%|██████▋   | 6619/9864 [27:49:55<14:03:04, 15.59s/it]
 67%|██████▋   | 6620/9864 [27:50:09<13:39:26, 15.16s/it]
                                                         
{'loss': '0.1161', 'grad_norm': '4.137', 'learning_rate': '5.159e-06', 'epoch': '0.6712'}

 67%|██████▋   | 6620/9864 [27:50:09<13:39:26, 15.16s/it]
 67%|██████▋   | 6621/9864 [27:50:26<14:08:25, 15.70s/it]
 67%|██████▋   | 6622/9864 [27:50:40<13:33:46, 15.06s/it]
 67%|██████▋   | 6623/9864 [27:50:54<13:19:50, 14.81s/it]
 67%|██████▋   | 6624/9864 [27:51:08<13:12:38, 14.68s/it]
 67%|██████▋   | 6625/9864 [27:51:22<13:01:19, 14.47s/it]
 67%|██████▋   | 6626/9864 [27:51:38<13:25:16, 14.92s/it]
 67%|██████▋   | 6627/9864 [27:51:53<13:19:54, 14.83s/it]
 67%|██████▋   | 6628/9864 [27:52:10<13:59:28, 15.57s/it]
 67%|██████▋   | 6629/9864 [27:52:24<13:32:02, 15.06s/it]
 67%|██████▋   | 6630/9864 [27:52:41<13:53:34, 15.47s/it]
                                                         
{'loss': '0.1024', 'grad_norm': '3.851', 'learning_rate': '5.13e-06', 'epoch': '0.6722'}

 67%|██████▋   | 6630/9864 [27:52:41<13:53:34, 15.47s/it]
 67%|██████▋   | 6631/9864 [27:52:55<13:34:20, 15.11s/it]
 67%|██████▋   | 6632/9864 [27:53:09<13:17:59, 14.81s/it]
 67%|██████▋   | 6633/9864 [27:53:23<13:07:19, 14.62s/it]
 67%|██████▋   | 6634/9864 [27:53:40<13:36:31, 15.17s/it]
 67%|██████▋   | 6635/9864 [27:53:54<13:25:12, 14.96s/it]
 67%|██████▋   | 6636/9864 [27:54:12<14:09:06, 15.78s/it]
 67%|██████▋   | 6637/9864 [27:54:27<14:06:14, 15.73s/it]
 67%|██████▋   | 6638/9864 [27:54:42<13:40:54, 15.27s/it]
 67%|██████▋   | 6639/9864 [27:54:56<13:21:48, 14.92s/it]
 67%|██████▋   | 6640/9864 [27:55:10<13:11:20, 14.73s/it]
                                                         
{'loss': '0.11', 'grad_norm': '4.273', 'learning_rate': '5.102e-06', 'epoch': '0.6732'}

 67%|██████▋   | 6640/9864 [27:55:10<13:11:20, 14.73s/it]
 67%|██████▋   | 6641/9864 [27:55:26<13:34:52, 15.17s/it]
 67%|██████▋   | 6642/9864 [27:55:40<13:21:49, 14.93s/it]
 67%|██████▋   | 6643/9864 [27:55:58<14:09:17, 15.82s/it]
 67%|██████▋   | 6644/9864 [27:56:13<13:45:47, 15.39s/it]
 67%|██████▋   | 6645/9864 [27:56:27<13:29:22, 15.09s/it]
 67%|██████▋   | 6646/9864 [27:56:43<13:36:24, 15.22s/it]
 67%|██████▋   | 6647/9864 [27:56:57<13:16:49, 14.86s/it]
 67%|██████▋   | 6648/9864 [27:57:12<13:28:55, 15.09s/it]
 67%|██████▋   | 6649/9864 [27:57:27<13:14:19, 14.82s/it]
 67%|██████▋   | 6650/9864 [27:57:43<13:34:25, 15.20s/it]
                                                         
{'loss': '0.1091', 'grad_norm': '3.258', 'learning_rate': '5.073e-06', 'epoch': '0.6742'}

 67%|██████▋   | 6650/9864 [27:57:43<13:34:25, 15.20s/it]
 67%|██████▋   | 6651/9864 [27:57:58<13:35:59, 15.24s/it]
 67%|██████▋   | 6652/9864 [27:58:12<13:21:29, 14.97s/it]
 67%|██████▋   | 6653/9864 [27:58:26<13:08:14, 14.73s/it]
 67%|██████▋   | 6654/9864 [27:58:42<13:29:08, 15.12s/it]
 67%|██████▋   | 6655/9864 [27:58:57<13:12:09, 14.81s/it]
 67%|██████▋   | 6656/9864 [27:59:13<13:41:23, 15.36s/it]
 67%|██████▋   | 6657/9864 [27:59:27<13:23:45, 15.04s/it]
 67%|██████▋   | 6658/9864 [27:59:43<13:23:44, 15.04s/it]
 68%|██████▊   | 6659/9864 [27:59:59<13:38:33, 15.32s/it]
 68%|██████▊   | 6660/9864 [28:00:13<13:25:00, 15.08s/it]
                                                         
{'loss': '0.107', 'grad_norm': '3.965', 'learning_rate': '5.045e-06', 'epoch': '0.6752'}

 68%|██████▊   | 6660/9864 [28:00:13<13:25:00, 15.08s/it]
 68%|██████▊   | 6661/9864 [28:00:27<13:05:23, 14.71s/it]
 68%|██████▊   | 6662/9864 [28:00:43<13:30:09, 15.18s/it]
 68%|██████▊   | 6663/9864 [28:00:59<13:42:54, 15.42s/it]
 68%|██████▊   | 6664/9864 [28:01:13<13:25:24, 15.10s/it]
 68%|██████▊   | 6665/9864 [28:01:29<13:34:27, 15.28s/it]
 68%|██████▊   | 6666/9864 [28:01:43<13:10:56, 14.84s/it]
 68%|██████▊   | 6667/9864 [28:01:59<13:22:12, 15.06s/it]
 68%|██████▊   | 6668/9864 [28:02:13<13:04:48, 14.73s/it]
 68%|██████▊   | 6669/9864 [28:02:28<13:14:05, 14.91s/it]
 68%|██████▊   | 6670/9864 [28:02:42<13:04:27, 14.74s/it]
                                                         
{'loss': '0.09193', 'grad_norm': '2.445', 'learning_rate': '5.016e-06', 'epoch': '0.6763'}

 68%|██████▊   | 6670/9864 [28:02:42<13:04:27, 14.74s/it]
 68%|██████▊   | 6671/9864 [28:02:58<13:17:52, 14.99s/it]
 68%|██████▊   | 6672/9864 [28:03:12<13:07:31, 14.80s/it]
 68%|██████▊   | 6673/9864 [28:03:27<13:03:32, 14.73s/it]
 68%|██████▊   | 6674/9864 [28:03:43<13:27:11, 15.18s/it]
 68%|██████▊   | 6675/9864 [28:03:59<13:38:19, 15.40s/it]
 68%|██████▊   | 6676/9864 [28:04:13<13:20:15, 15.06s/it]
 68%|██████▊   | 6677/9864 [28:04:29<13:26:23, 15.18s/it]
 68%|██████▊   | 6678/9864 [28:04:43<13:10:31, 14.89s/it]
 68%|██████▊   | 6679/9864 [28:04:57<12:55:15, 14.60s/it]
 68%|██████▊   | 6680/9864 [28:05:13<13:15:35, 14.99s/it]
                                                         
{'loss': '0.1004', 'grad_norm': '3.43', 'learning_rate': '4.988e-06', 'epoch': '0.6773'}

 68%|██████▊   | 6680/9864 [28:05:13<13:15:35, 14.99s/it]
 68%|██████▊   | 6681/9864 [28:05:27<12:58:53, 14.68s/it]
 68%|██████▊   | 6682/9864 [28:05:43<13:18:21, 15.05s/it]
 68%|██████▊   | 6683/9864 [28:05:59<13:34:54, 15.37s/it]
 68%|██████▊   | 6684/9864 [28:06:12<13:08:15, 14.87s/it]
 68%|██████▊   | 6685/9864 [28:06:28<13:21:48, 15.13s/it]
 68%|██████▊   | 6686/9864 [28:06:42<12:59:49, 14.72s/it]
 68%|██████▊   | 6687/9864 [28:06:56<12:51:31, 14.57s/it]
 68%|██████▊   | 6688/9864 [28:07:12<13:06:48, 14.86s/it]
 68%|██████▊   | 6689/9864 [28:07:27<13:15:21, 15.03s/it]
 68%|██████▊   | 6690/9864 [28:07:41<13:02:41, 14.80s/it]
                                                         
{'loss': '0.1094', 'grad_norm': '4.581', 'learning_rate': '4.959e-06', 'epoch': '0.6783'}

 68%|██████▊   | 6690/9864 [28:07:41<13:02:41, 14.80s/it]
 68%|██████▊   | 6691/9864 [28:07:57<13:09:59, 14.94s/it]
 68%|██████▊   | 6692/9864 [28:08:12<13:23:58, 15.21s/it]
 68%|██████▊   | 6693/9864 [28:08:27<13:12:33, 15.00s/it]
 68%|██████▊   | 6694/9864 [28:08:41<13:03:10, 14.82s/it]
 68%|██████▊   | 6695/9864 [28:08:55<12:52:30, 14.63s/it]
 68%|██████▊   | 6696/9864 [28:09:10<12:47:47, 14.54s/it]
 68%|██████▊   | 6697/9864 [28:09:27<13:36:14, 15.46s/it]
 68%|██████▊   | 6698/9864 [28:09:43<13:38:53, 15.52s/it]
 68%|██████▊   | 6699/9864 [28:09:57<13:15:53, 15.09s/it]
 68%|██████▊   | 6700/9864 [28:10:13<13:32:17, 15.40s/it]
                                                         
{'loss': '0.0966', 'grad_norm': '3.038', 'learning_rate': '4.931e-06', 'epoch': '0.6793'}

 68%|██████▊   | 6700/9864 [28:10:13<13:32:17, 15.40s/it]
 68%|██████▊   | 6701/9864 [28:10:27<13:07:20, 14.94s/it]
 68%|██████▊   | 6702/9864 [28:10:42<12:59:49, 14.80s/it]
 68%|██████▊   | 6703/9864 [28:10:55<12:44:12, 14.51s/it]
 68%|██████▊   | 6704/9864 [28:11:10<12:37:28, 14.38s/it]
 68%|██████▊   | 6705/9864 [28:11:29<14:00:01, 15.96s/it]
 68%|██████▊   | 6706/9864 [28:11:44<13:34:37, 15.48s/it]
 68%|██████▊   | 6707/9864 [28:11:58<13:19:41, 15.20s/it]
 68%|██████▊   | 6708/9864 [28:12:14<13:27:41, 15.36s/it]
 68%|██████▊   | 6709/9864 [28:12:28<13:07:33, 14.98s/it]
 68%|██████▊   | 6710/9864 [28:12:41<12:43:06, 14.52s/it]
                                                         
{'loss': '0.1075', 'grad_norm': '5.231', 'learning_rate': '4.903e-06', 'epoch': '0.6803'}

 68%|██████▊   | 6710/9864 [28:12:41<12:43:06, 14.52s/it]
 68%|██████▊   | 6711/9864 [28:12:56<12:38:20, 14.43s/it]
 68%|██████▊   | 6712/9864 [28:13:10<12:31:38, 14.31s/it]
 68%|██████▊   | 6713/9864 [28:13:27<13:22:53, 15.29s/it]
 68%|██████▊   | 6714/9864 [28:13:43<13:28:20, 15.40s/it]
 68%|██████▊   | 6715/9864 [28:13:57<13:10:24, 15.06s/it]
 68%|██████▊   | 6716/9864 [28:14:11<12:58:46, 14.84s/it]
 68%|██████▊   | 6717/9864 [28:14:27<13:08:19, 15.03s/it]
 68%|██████▊   | 6718/9864 [28:14:41<12:59:53, 14.87s/it]
 68%|██████▊   | 6719/9864 [28:14:56<12:48:56, 14.67s/it]
 68%|██████▊   | 6720/9864 [28:15:12<13:17:52, 15.23s/it]
                                                         
{'loss': '0.1029', 'grad_norm': '3.768', 'learning_rate': '4.874e-06', 'epoch': '0.6813'}

 68%|██████▊   | 6720/9864 [28:15:12<13:17:52, 15.23s/it]
 68%|██████▊   | 6721/9864 [28:15:29<13:48:31, 15.82s/it]
 68%|██████▊   | 6722/9864 [28:15:43<13:19:21, 15.26s/it]
 68%|██████▊   | 6723/9864 [28:15:58<13:06:29, 15.02s/it]
 68%|██████▊   | 6724/9864 [28:16:13<13:11:15, 15.12s/it]
 68%|██████▊   | 6725/9864 [28:16:27<12:57:16, 14.86s/it]
 68%|██████▊   | 6726/9864 [28:16:42<12:51:29, 14.75s/it]
 68%|██████▊   | 6727/9864 [28:16:58<13:06:05, 15.04s/it]
 68%|██████▊   | 6728/9864 [28:17:12<12:52:36, 14.78s/it]
 68%|██████▊   | 6729/9864 [28:17:29<13:37:09, 15.64s/it]
 68%|██████▊   | 6730/9864 [28:17:44<13:16:29, 15.25s/it]
                                                         
{'loss': '0.09774', 'grad_norm': '2.745', 'learning_rate': '4.846e-06', 'epoch': '0.6823'}

 68%|██████▊   | 6730/9864 [28:17:44<13:16:29, 15.25s/it]
 68%|██████▊   | 6731/9864 [28:17:58<13:06:28, 15.06s/it]
 68%|██████▊   | 6732/9864 [28:18:15<13:24:39, 15.41s/it]
 68%|██████▊   | 6733/9864 [28:18:28<12:58:09, 14.91s/it]
 68%|██████▊   | 6734/9864 [28:18:42<12:44:15, 14.65s/it]
 68%|██████▊   | 6735/9864 [28:18:58<13:04:30, 15.04s/it]
 68%|██████▊   | 6736/9864 [28:19:12<12:47:08, 14.71s/it]
 68%|██████▊   | 6737/9864 [28:19:30<13:36:05, 15.66s/it]
 68%|██████▊   | 6738/9864 [28:19:45<13:17:07, 15.30s/it]
 68%|██████▊   | 6739/9864 [28:19:59<13:09:50, 15.17s/it]
 68%|██████▊   | 6740/9864 [28:20:15<13:15:44, 15.28s/it]
                                                         
{'loss': '0.1153', 'grad_norm': '4.279', 'learning_rate': '4.818e-06', 'epoch': '0.6833'}

 68%|██████▊   | 6740/9864 [28:20:15<13:15:44, 15.28s/it]
 68%|██████▊   | 6741/9864 [28:20:30<13:07:21, 15.13s/it]
 68%|██████▊   | 6742/9864 [28:20:43<12:43:00, 14.66s/it]
 68%|██████▊   | 6743/9864 [28:20:59<13:05:44, 15.11s/it]
 68%|██████▊   | 6744/9864 [28:21:15<13:10:34, 15.20s/it]
 68%|██████▊   | 6745/9864 [28:21:31<13:26:11, 15.51s/it]
 68%|██████▊   | 6746/9864 [28:21:45<13:06:49, 15.14s/it]
 68%|██████▊   | 6747/9864 [28:22:00<12:55:37, 14.93s/it]
 68%|██████▊   | 6748/9864 [28:22:14<12:50:30, 14.84s/it]
 68%|██████▊   | 6749/9864 [28:22:31<13:10:26, 15.23s/it]
 68%|██████▊   | 6750/9864 [28:22:46<13:16:53, 15.35s/it]
                                                         
{'loss': '0.09707', 'grad_norm': '3.535', 'learning_rate': '4.79e-06', 'epoch': '0.6844'}

 68%|██████▊   | 6750/9864 [28:22:46<13:16:53, 15.35s/it]
 68%|██████▊   | 6751/9864 [28:23:00<12:59:10, 15.02s/it]
 68%|██████▊   | 6752/9864 [28:23:16<13:10:58, 15.25s/it]
 68%|██████▊   | 6753/9864 [28:23:32<13:12:29, 15.28s/it]
 68%|██████▊   | 6754/9864 [28:23:46<12:51:32, 14.89s/it]
 68%|██████▊   | 6755/9864 [28:23:59<12:35:37, 14.58s/it]
 68%|██████▊   | 6756/9864 [28:24:13<12:25:03, 14.38s/it]
 69%|██████▊   | 6757/9864 [28:24:29<12:45:30, 14.78s/it]
 69%|██████▊   | 6758/9864 [28:24:43<12:30:28, 14.50s/it]
 69%|██████▊   | 6759/9864 [28:25:00<13:15:30, 15.37s/it]
 69%|██████▊   | 6760/9864 [28:25:14<12:51:56, 14.92s/it]
                                                         
{'loss': '0.1059', 'grad_norm': '4.994', 'learning_rate': '4.762e-06', 'epoch': '0.6854'}

 69%|██████▊   | 6760/9864 [28:25:14<12:51:56, 14.92s/it]
 69%|██████▊   | 6761/9864 [28:25:30<12:59:42, 15.08s/it]
 69%|██████▊   | 6762/9864 [28:25:44<12:41:38, 14.73s/it]
 69%|██████▊   | 6763/9864 [28:25:58<12:37:27, 14.66s/it]
 69%|██████▊   | 6764/9864 [28:26:12<12:24:12, 14.40s/it]
 69%|██████▊   | 6765/9864 [28:26:28<12:43:27, 14.78s/it]
 69%|██████▊   | 6766/9864 [28:26:42<12:37:25, 14.67s/it]
 69%|██████▊   | 6767/9864 [28:27:00<13:23:25, 15.57s/it]
 69%|██████▊   | 6768/9864 [28:27:15<13:27:12, 15.64s/it]
 69%|██████▊   | 6769/9864 [28:27:29<13:01:57, 15.16s/it]
 69%|██████▊   | 6770/9864 [28:27:43<12:43:11, 14.80s/it]
                                                         
{'loss': '0.09864', 'grad_norm': '3.716', 'learning_rate': '4.734e-06', 'epoch': '0.6864'}

 69%|██████▊   | 6770/9864 [28:27:43<12:43:11, 14.80s/it]
 69%|██████▊   | 6771/9864 [28:27:58<12:32:14, 14.59s/it]
 69%|██████▊   | 6772/9864 [28:28:12<12:27:20, 14.50s/it]
 69%|██████▊   | 6773/9864 [28:28:28<12:50:01, 14.95s/it]
 69%|██████▊   | 6774/9864 [28:28:44<13:03:53, 15.22s/it]
 69%|██████▊   | 6775/9864 [28:28:59<13:11:47, 15.38s/it]
 69%|██████▊   | 6776/9864 [28:29:15<13:19:17, 15.53s/it]
 69%|██████▊   | 6777/9864 [28:29:30<13:01:11, 15.18s/it]
 69%|██████▊   | 6778/9864 [28:29:44<12:42:37, 14.83s/it]
 69%|██████▊   | 6779/9864 [28:29:58<12:27:06, 14.53s/it]
 69%|██████▊   | 6780/9864 [28:30:12<12:24:37, 14.49s/it]
                                                         
{'loss': '0.102', 'grad_norm': '3.569', 'learning_rate': '4.706e-06', 'epoch': '0.6874'}

 69%|██████▊   | 6780/9864 [28:30:12<12:24:37, 14.49s/it]
 69%|██████▊   | 6781/9864 [28:30:26<12:20:53, 14.42s/it]
 69%|██████▉   | 6782/9864 [28:30:44<13:07:10, 15.32s/it]
 69%|██████▉   | 6783/9864 [28:30:58<12:53:48, 15.07s/it]
 69%|██████▉   | 6784/9864 [28:31:14<13:05:38, 15.30s/it]
 69%|██████▉   | 6785/9864 [28:31:28<12:48:19, 14.97s/it]
 69%|██████▉   | 6786/9864 [28:31:42<12:35:39, 14.73s/it]
 69%|██████▉   | 6787/9864 [28:31:56<12:26:51, 14.56s/it]
 69%|██████▉   | 6788/9864 [28:32:11<12:24:10, 14.52s/it]
 69%|██████▉   | 6789/9864 [28:32:27<12:45:27, 14.94s/it]
 69%|██████▉   | 6790/9864 [28:32:44<13:24:20, 15.70s/it]
                                                         
{'loss': '0.09518', 'grad_norm': '3.72', 'learning_rate': '4.678e-06', 'epoch': '0.6884'}

 69%|██████▉   | 6790/9864 [28:32:44<13:24:20, 15.70s/it]
 69%|██████▉   | 6791/9864 [28:33:00<13:31:22, 15.84s/it]
 69%|██████▉   | 6792/9864 [28:33:14<13:02:26, 15.28s/it]
 69%|██████▉   | 6793/9864 [28:33:28<12:41:14, 14.87s/it]
 69%|██████▉   | 6794/9864 [28:33:42<12:24:53, 14.56s/it]
 69%|██████▉   | 6795/9864 [28:33:56<12:20:15, 14.47s/it]
 69%|██████▉   | 6796/9864 [28:34:11<12:17:05, 14.41s/it]
 69%|██████▉   | 6797/9864 [28:34:30<13:31:52, 15.88s/it]
 69%|██████▉   | 6798/9864 [28:34:44<12:59:23, 15.25s/it]
 69%|██████▉   | 6799/9864 [28:35:00<13:08:19, 15.43s/it]
 69%|██████▉   | 6800/9864 [28:35:14<12:44:05, 14.96s/it]
                                                         
{'loss': '0.09562', 'grad_norm': '3.87', 'learning_rate': '4.651e-06', 'epoch': '0.6894'}

 69%|██████▉   | 6800/9864 [28:35:14<12:44:05, 14.96s/it]
 69%|██████▉   | 6801/9864 [28:35:27<12:26:55, 14.63s/it]
 69%|██████▉   | 6802/9864 [28:35:42<12:21:31, 14.53s/it]
 69%|██████▉   | 6803/9864 [28:35:55<12:09:53, 14.31s/it]
 69%|██████▉   | 6804/9864 [28:36:10<12:05:53, 14.23s/it]
 69%|██████▉   | 6805/9864 [28:36:27<12:52:55, 15.16s/it]
 69%|██████▉   | 6806/9864 [28:36:44<13:30:26, 15.90s/it]
 69%|██████▉   | 6807/9864 [28:36:58<12:57:11, 15.25s/it]
 69%|██████▉   | 6808/9864 [28:37:12<12:41:07, 14.94s/it]
 69%|██████▉   | 6809/9864 [28:37:27<12:30:06, 14.73s/it]
 69%|██████▉   | 6810/9864 [28:37:40<12:15:50, 14.46s/it]
                                                         
{'loss': '0.1055', 'grad_norm': '3.987', 'learning_rate': '4.623e-06', 'epoch': '0.6904'}

 69%|██████▉   | 6810/9864 [28:37:40<12:15:50, 14.46s/it]
 69%|██████▉   | 6811/9864 [28:37:55<12:17:31, 14.49s/it]
 69%|██████▉   | 6812/9864 [28:38:09<12:16:14, 14.47s/it]
 69%|██████▉   | 6813/9864 [28:38:29<13:26:22, 15.86s/it]
 69%|██████▉   | 6814/9864 [28:38:45<13:32:42, 15.99s/it]
 69%|██████▉   | 6815/9864 [28:38:59<13:04:13, 15.43s/it]
 69%|██████▉   | 6816/9864 [28:39:14<12:50:20, 15.16s/it]
 69%|██████▉   | 6817/9864 [28:39:28<12:40:43, 14.98s/it]
 69%|██████▉   | 6818/9864 [28:39:42<12:29:05, 14.76s/it]
 69%|██████▉   | 6819/9864 [28:39:56<12:17:26, 14.53s/it]
 69%|██████▉   | 6820/9864 [28:40:10<12:07:23, 14.34s/it]
                                                         
{'loss': '0.102', 'grad_norm': '4', 'learning_rate': '4.595e-06', 'epoch': '0.6915'}

 69%|██████▉   | 6820/9864 [28:40:10<12:07:23, 14.34s/it]
 69%|██████▉   | 6821/9864 [28:40:29<13:20:21, 15.78s/it]
 69%|██████▉   | 6822/9864 [28:40:43<12:52:25, 15.24s/it]
 69%|██████▉   | 6823/9864 [28:40:58<12:36:47, 14.93s/it]
 69%|██████▉   | 6824/9864 [28:41:12<12:24:39, 14.70s/it]
 69%|██████▉   | 6825/9864 [28:41:26<12:15:13, 14.52s/it]
 69%|██████▉   | 6826/9864 [28:41:40<12:03:43, 14.29s/it]
 69%|██████▉   | 6827/9864 [28:41:54<11:59:56, 14.22s/it]
 69%|██████▉   | 6828/9864 [28:42:11<12:51:25, 15.25s/it]
 69%|██████▉   | 6829/9864 [28:42:29<13:35:30, 16.12s/it]
 69%|██████▉   | 6830/9864 [28:42:44<13:10:39, 15.64s/it]
                                                         
{'loss': '0.09798', 'grad_norm': '3.884', 'learning_rate': '4.568e-06', 'epoch': '0.6925'}

 69%|██████▉   | 6830/9864 [28:42:44<13:10:39, 15.64s/it]
 69%|██████▉   | 6831/9864 [28:42:59<12:54:51, 15.33s/it]
 69%|██████▉   | 6832/9864 [28:43:13<12:44:01, 15.12s/it]
 69%|██████▉   | 6833/9864 [28:43:28<12:35:01, 14.95s/it]
 69%|██████▉   | 6834/9864 [28:43:42<12:20:51, 14.67s/it]
 69%|██████▉   | 6835/9864 [28:43:56<12:17:59, 14.62s/it]
 69%|██████▉   | 6836/9864 [28:44:16<13:29:51, 16.05s/it]
 69%|██████▉   | 6837/9864 [28:44:32<13:31:05, 16.08s/it]
 69%|██████▉   | 6838/9864 [28:44:46<13:01:28, 15.50s/it]
 69%|██████▉   | 6839/9864 [28:45:00<12:39:46, 15.07s/it]
 69%|██████▉   | 6840/9864 [28:45:14<12:21:23, 14.71s/it]
                                                         
{'loss': '0.1027', 'grad_norm': '4.107', 'learning_rate': '4.54e-06', 'epoch': '0.6935'}

 69%|██████▉   | 6840/9864 [28:45:14<12:21:23, 14.71s/it]
 69%|██████▉   | 6841/9864 [28:45:29<12:22:15, 14.73s/it]
 69%|██████▉   | 6842/9864 [28:45:43<12:15:14, 14.60s/it]
 69%|██████▉   | 6843/9864 [28:45:57<12:09:19, 14.49s/it]
 69%|██████▉   | 6844/9864 [28:46:16<13:22:26, 15.94s/it]
 69%|██████▉   | 6845/9864 [28:46:33<13:25:16, 16.00s/it]
 69%|██████▉   | 6846/9864 [28:46:46<12:49:26, 15.30s/it]
 69%|██████▉   | 6847/9864 [28:47:00<12:28:52, 14.89s/it]
 69%|██████▉   | 6848/9864 [28:47:15<12:19:50, 14.72s/it]
 69%|██████▉   | 6849/9864 [28:47:28<12:07:23, 14.48s/it]
 69%|██████▉   | 6850/9864 [28:47:43<12:09:19, 14.52s/it]
                                                         
{'loss': '0.104', 'grad_norm': '4.02', 'learning_rate': '4.513e-06', 'epoch': '0.6945'}

 69%|██████▉   | 6850/9864 [28:47:43<12:09:19, 14.52s/it]
 69%|██████▉   | 6851/9864 [28:47:57<11:58:15, 14.30s/it]
 69%|██████▉   | 6852/9864 [28:48:16<13:15:42, 15.85s/it]
 69%|██████▉   | 6853/9864 [28:48:30<12:49:04, 15.33s/it]
 69%|██████▉   | 6854/9864 [28:48:44<12:24:57, 14.85s/it]
 69%|██████▉   | 6855/9864 [28:48:58<12:07:30, 14.51s/it]
 70%|██████▉   | 6856/9864 [28:49:12<12:04:11, 14.45s/it]
 70%|██████▉   | 6857/9864 [28:49:27<12:03:09, 14.43s/it]
 70%|██████▉   | 6858/9864 [28:49:41<12:05:34, 14.48s/it]
 70%|██████▉   | 6859/9864 [28:50:00<13:15:36, 15.89s/it]
 70%|██████▉   | 6860/9864 [28:50:16<13:15:46, 15.89s/it]
                                                         
{'loss': '0.09229', 'grad_norm': '3.843', 'learning_rate': '4.485e-06', 'epoch': '0.6955'}

 70%|██████▉   | 6860/9864 [28:50:16<13:15:46, 15.89s/it]
 70%|██████▉   | 6861/9864 [28:50:30<12:43:44, 15.26s/it]
 70%|██████▉   | 6862/9864 [28:50:44<12:21:59, 14.83s/it]
 70%|██████▉   | 6863/9864 [28:50:58<12:16:35, 14.73s/it]
 70%|██████▉   | 6864/9864 [28:51:13<12:13:20, 14.67s/it]
 70%|██████▉   | 6865/9864 [28:51:27<12:02:09, 14.45s/it]
 70%|██████▉   | 6866/9864 [28:51:44<12:45:29, 15.32s/it]
 70%|██████▉   | 6867/9864 [28:51:58<12:29:03, 15.00s/it]
 70%|██████▉   | 6868/9864 [28:52:15<12:48:37, 15.39s/it]
 70%|██████▉   | 6869/9864 [28:52:29<12:35:02, 15.13s/it]
 70%|██████▉   | 6870/9864 [28:52:43<12:19:33, 14.82s/it]
                                                         
{'loss': '0.1067', 'grad_norm': '5.432', 'learning_rate': '4.458e-06', 'epoch': '0.6965'}

 70%|██████▉   | 6870/9864 [28:52:43<12:19:33, 14.82s/it]
 70%|██████▉   | 6871/9864 [28:52:58<12:13:34, 14.71s/it]
 70%|██████▉   | 6872/9864 [28:53:12<12:07:27, 14.59s/it]
 70%|██████▉   | 6873/9864 [28:53:28<12:26:33, 14.98s/it]
 70%|██████▉   | 6874/9864 [28:53:44<12:45:12, 15.36s/it]
 70%|██████▉   | 6875/9864 [28:54:02<13:24:51, 16.16s/it]
 70%|██████▉   | 6876/9864 [28:54:16<12:55:46, 15.58s/it]
 70%|██████▉   | 6877/9864 [28:54:30<12:28:24, 15.03s/it]
 70%|██████▉   | 6878/9864 [28:54:44<12:09:33, 14.66s/it]
 70%|██████▉   | 6879/9864 [28:54:58<11:56:33, 14.40s/it]
 70%|██████▉   | 6880/9864 [28:55:14<12:17:11, 14.82s/it]
                                                         
{'loss': '0.0946', 'grad_norm': '3.582', 'learning_rate': '4.431e-06', 'epoch': '0.6975'}

 70%|██████▉   | 6880/9864 [28:55:14<12:17:11, 14.82s/it]
 70%|██████▉   | 6881/9864 [28:55:28<12:09:45, 14.68s/it]
 70%|██████▉   | 6882/9864 [28:55:45<12:48:35, 15.46s/it]
 70%|██████▉   | 6883/9864 [28:55:59<12:28:02, 15.06s/it]
 70%|██████▉   | 6884/9864 [28:56:16<12:48:02, 15.46s/it]
 70%|██████▉   | 6885/9864 [28:56:30<12:31:24, 15.13s/it]
 70%|██████▉   | 6886/9864 [28:56:45<12:22:18, 14.96s/it]
 70%|██████▉   | 6887/9864 [28:57:00<12:32:38, 15.17s/it]
 70%|██████▉   | 6888/9864 [28:57:15<12:18:00, 14.88s/it]
 70%|██████▉   | 6889/9864 [28:57:29<12:10:36, 14.73s/it]
 70%|██████▉   | 6890/9864 [28:57:45<12:29:40, 15.12s/it]
                                                         
{'loss': '0.08819', 'grad_norm': '5.048', 'learning_rate': '4.403e-06', 'epoch': '0.6986'}

 70%|██████▉   | 6890/9864 [28:57:45<12:29:40, 15.12s/it]
 70%|██████▉   | 6891/9864 [28:58:02<13:03:22, 15.81s/it]
 70%|██████▉   | 6892/9864 [28:58:17<12:40:28, 15.35s/it]
 70%|██████▉   | 6893/9864 [28:58:31<12:20:08, 14.95s/it]
 70%|██████▉   | 6894/9864 [28:58:47<12:37:21, 15.30s/it]
 70%|██████▉   | 6895/9864 [28:59:01<12:22:50, 15.01s/it]
 70%|██████▉   | 6896/9864 [28:59:16<12:13:30, 14.83s/it]
 70%|██████▉   | 6897/9864 [28:59:30<12:05:03, 14.66s/it]
 70%|██████▉   | 6898/9864 [28:59:48<12:49:49, 15.57s/it]
 70%|██████▉   | 6899/9864 [29:00:03<12:51:16, 15.61s/it]
 70%|██████▉   | 6900/9864 [29:00:17<12:24:16, 15.07s/it]
                                                         
{'loss': '0.09642', 'grad_norm': '3.8', 'learning_rate': '4.376e-06', 'epoch': '0.6996'}

 70%|██████▉   | 6900/9864 [29:00:17<12:24:16, 15.07s/it]
 70%|██████▉   | 6901/9864 [29:00:31<12:04:07, 14.66s/it]
 70%|██████▉   | 6902/9864 [29:00:45<11:54:46, 14.48s/it]
 70%|██████▉   | 6903/9864 [29:01:01<12:13:26, 14.86s/it]
 70%|██████▉   | 6904/9864 [29:01:15<12:04:24, 14.68s/it]
 70%|███████   | 6905/9864 [29:01:29<12:01:07, 14.62s/it]
 70%|███████   | 6906/9864 [29:01:46<12:37:59, 15.38s/it]
 70%|███████   | 6907/9864 [29:02:03<12:52:16, 15.67s/it]
 70%|███████   | 6908/9864 [29:02:17<12:31:13, 15.25s/it]
 70%|███████   | 6909/9864 [29:02:31<12:18:32, 15.00s/it]
 70%|███████   | 6910/9864 [29:02:47<12:27:56, 15.19s/it]
                                                         
{'loss': '0.09329', 'grad_norm': '5.539', 'learning_rate': '4.349e-06', 'epoch': '0.7006'}

 70%|███████   | 6910/9864 [29:02:47<12:27:56, 15.19s/it]
 70%|███████   | 6911/9864 [29:03:01<12:15:16, 14.94s/it]
 70%|███████   | 6912/9864 [29:03:15<11:57:55, 14.59s/it]
 70%|███████   | 6913/9864 [29:03:32<12:23:04, 15.11s/it]
 70%|███████   | 6914/9864 [29:03:46<12:08:23, 14.81s/it]
 70%|███████   | 6915/9864 [29:04:02<12:24:37, 15.15s/it]
 70%|███████   | 6916/9864 [29:04:17<12:32:40, 15.32s/it]
 70%|███████   | 6917/9864 [29:04:31<12:11:11, 14.89s/it]
 70%|███████   | 6918/9864 [29:04:47<12:24:27, 15.16s/it]
 70%|███████   | 6919/9864 [29:05:01<12:13:47, 14.95s/it]
 70%|███████   | 6920/9864 [29:05:16<12:09:59, 14.88s/it]
                                                         
{'loss': '0.09173', 'grad_norm': '4.343', 'learning_rate': '4.322e-06', 'epoch': '0.7016'}

 70%|███████   | 6920/9864 [29:05:16<12:09:59, 14.88s/it]
 70%|███████   | 6921/9864 [29:05:32<12:27:49, 15.25s/it]
 70%|███████   | 6922/9864 [29:05:46<12:05:28, 14.80s/it]
 70%|███████   | 6923/9864 [29:06:02<12:18:35, 15.07s/it]
 70%|███████   | 6924/9864 [29:06:16<12:02:16, 14.74s/it]
 70%|███████   | 6925/9864 [29:06:32<12:27:43, 15.26s/it]
 70%|███████   | 6926/9864 [29:06:48<12:33:12, 15.38s/it]
 70%|███████   | 6927/9864 [29:07:02<12:13:45, 14.99s/it]
 70%|███████   | 6928/9864 [29:07:18<12:23:38, 15.20s/it]
 70%|███████   | 6929/9864 [29:07:32<12:07:10, 14.87s/it]
 70%|███████   | 6930/9864 [29:07:47<12:20:43, 15.15s/it]
                                                         
{'loss': '0.1037', 'grad_norm': '5.085', 'learning_rate': '4.295e-06', 'epoch': '0.7026'}

 70%|███████   | 6930/9864 [29:07:47<12:20:43, 15.15s/it]
 70%|███████   | 6931/9864 [29:08:02<12:08:41, 14.91s/it]
 70%|███████   | 6932/9864 [29:08:16<11:59:51, 14.73s/it]
 70%|███████   | 6933/9864 [29:08:32<12:18:46, 15.12s/it]
 70%|███████   | 6934/9864 [29:08:48<12:30:10, 15.36s/it]
 70%|███████   | 6935/9864 [29:09:02<12:12:30, 15.01s/it]
 70%|███████   | 6936/9864 [29:09:16<11:59:24, 14.74s/it]
 70%|███████   | 6937/9864 [29:09:34<12:36:53, 15.52s/it]
 70%|███████   | 6938/9864 [29:09:48<12:18:10, 15.14s/it]
 70%|███████   | 6939/9864 [29:10:02<12:02:11, 14.81s/it]
 70%|███████   | 6940/9864 [29:10:16<11:51:12, 14.59s/it]
                                                         
{'loss': '0.08701', 'grad_norm': '4.198', 'learning_rate': '4.268e-06', 'epoch': '0.7036'}

 70%|███████   | 6940/9864 [29:10:16<11:51:12, 14.59s/it]
 70%|███████   | 6941/9864 [29:10:32<12:06:50, 14.92s/it]
 70%|███████   | 6942/9864 [29:10:48<12:25:06, 15.30s/it]
 70%|███████   | 6943/9864 [29:11:02<12:08:21, 14.96s/it]
 70%|███████   | 6944/9864 [29:11:16<11:53:07, 14.65s/it]
 70%|███████   | 6945/9864 [29:11:33<12:28:25, 15.38s/it]
 70%|███████   | 6946/9864 [29:11:47<12:06:10, 14.93s/it]
 70%|███████   | 6947/9864 [29:12:01<11:50:23, 14.61s/it]
 70%|███████   | 6948/9864 [29:12:15<11:37:42, 14.36s/it]
 70%|███████   | 6949/9864 [29:12:32<12:24:39, 15.33s/it]
 70%|███████   | 6950/9864 [29:12:46<11:59:36, 14.82s/it]
                                                         
{'loss': '0.1024', 'grad_norm': '4.155', 'learning_rate': '4.241e-06', 'epoch': '0.7046'}

 70%|███████   | 6950/9864 [29:12:46<11:59:36, 14.82s/it]
 70%|███████   | 6951/9864 [29:13:00<11:54:56, 14.73s/it]
 70%|███████   | 6952/9864 [29:13:15<11:59:17, 14.82s/it]
 70%|███████   | 6953/9864 [29:13:32<12:20:29, 15.26s/it]
 70%|███████   | 6954/9864 [29:13:47<12:16:09, 15.18s/it]
 71%|███████   | 6955/9864 [29:14:01<12:03:23, 14.92s/it]
 71%|███████   | 6956/9864 [29:14:15<11:53:38, 14.72s/it]
 71%|███████   | 6957/9864 [29:14:33<12:33:07, 15.54s/it]
 71%|███████   | 6958/9864 [29:14:47<12:11:17, 15.10s/it]
 71%|███████   | 6959/9864 [29:15:01<11:55:36, 14.78s/it]
 71%|███████   | 6960/9864 [29:15:17<12:15:34, 15.20s/it]
                                                         
{'loss': '0.09057', 'grad_norm': '3.718', 'learning_rate': '4.214e-06', 'epoch': '0.7057'}

 71%|███████   | 6960/9864 [29:15:17<12:15:34, 15.20s/it]
 71%|███████   | 6961/9864 [29:15:31<11:54:05, 14.76s/it]
 71%|███████   | 6962/9864 [29:15:47<12:08:04, 15.05s/it]
 71%|███████   | 6963/9864 [29:16:01<11:54:09, 14.77s/it]
 71%|███████   | 6964/9864 [29:16:15<11:50:44, 14.70s/it]
 71%|███████   | 6965/9864 [29:16:31<12:05:11, 15.01s/it]
 71%|███████   | 6966/9864 [29:16:47<12:19:12, 15.30s/it]
 71%|███████   | 6967/9864 [29:17:01<12:04:31, 15.01s/it]
 71%|███████   | 6968/9864 [29:17:17<12:19:55, 15.33s/it]
 71%|███████   | 6969/9864 [29:17:33<12:30:12, 15.55s/it]
 71%|███████   | 6970/9864 [29:17:48<12:11:42, 15.17s/it]
                                                         
{'loss': '0.09453', 'grad_norm': '4.422', 'learning_rate': '4.188e-06', 'epoch': '0.7067'}

 71%|███████   | 6970/9864 [29:17:48<12:11:42, 15.17s/it]
 71%|███████   | 6971/9864 [29:18:02<11:55:11, 14.83s/it]
 71%|███████   | 6972/9864 [29:18:15<11:37:39, 14.47s/it]
 71%|███████   | 6973/9864 [29:18:31<11:54:35, 14.83s/it]
 71%|███████   | 6974/9864 [29:18:47<12:16:21, 15.29s/it]
 71%|███████   | 6975/9864 [29:19:04<12:29:14, 15.56s/it]
 71%|███████   | 6976/9864 [29:19:19<12:32:42, 15.64s/it]
 71%|███████   | 6977/9864 [29:19:33<12:02:26, 15.01s/it]
 71%|███████   | 6978/9864 [29:19:47<11:48:21, 14.73s/it]
 71%|███████   | 6979/9864 [29:20:01<11:33:47, 14.43s/it]
 71%|███████   | 6980/9864 [29:20:17<11:57:46, 14.93s/it]
                                                         
{'loss': '0.1082', 'grad_norm': '3.085', 'learning_rate': '4.161e-06', 'epoch': '0.7077'}

 71%|███████   | 6980/9864 [29:20:17<11:57:46, 14.93s/it]
 71%|███████   | 6981/9864 [29:20:32<12:08:08, 15.15s/it]
 71%|███████   | 6982/9864 [29:20:46<11:50:25, 14.79s/it]
 71%|███████   | 6983/9864 [29:21:04<12:28:27, 15.59s/it]
 71%|███████   | 6984/9864 [29:21:18<12:02:03, 15.04s/it]
 71%|███████   | 6985/9864 [29:21:32<11:50:42, 14.81s/it]
 71%|███████   | 6986/9864 [29:21:47<11:48:22, 14.77s/it]
 71%|███████   | 6987/9864 [29:22:00<11:32:27, 14.44s/it]
 71%|███████   | 6988/9864 [29:22:18<12:20:46, 15.45s/it]
 71%|███████   | 6989/9864 [29:22:32<11:58:52, 15.00s/it]
 71%|███████   | 6990/9864 [29:22:46<11:44:34, 14.71s/it]
                                                         
{'loss': '0.09791', 'grad_norm': '2.554', 'learning_rate': '4.134e-06', 'epoch': '0.7087'}

 71%|███████   | 6990/9864 [29:22:46<11:44:34, 14.71s/it]
 71%|███████   | 6991/9864 [29:23:04<12:24:50, 15.56s/it]
 71%|███████   | 6992/9864 [29:23:18<12:03:37, 15.12s/it]
 71%|███████   | 6993/9864 [29:23:32<11:49:37, 14.83s/it]
 71%|███████   | 6994/9864 [29:23:46<11:38:57, 14.61s/it]
 71%|███████   | 6995/9864 [29:24:02<11:59:22, 15.04s/it]
 71%|███████   | 6996/9864 [29:24:18<12:13:01, 15.34s/it]
 71%|███████   | 6997/9864 [29:24:32<11:59:08, 15.05s/it]
 71%|███████   | 6998/9864 [29:24:46<11:42:04, 14.70s/it]
 71%|███████   | 6999/9864 [29:25:04<12:22:35, 15.55s/it]
 71%|███████   | 7000/9864 [29:25:17<11:54:46, 14.97s/it]
                                                         
{'loss': '0.105', 'grad_norm': '3.598', 'learning_rate': '4.108e-06', 'epoch': '0.7097'}

 71%|███████   | 7000/9864 [29:25:17<11:54:46, 14.97s/it]

Writing model shards:   0%|          | 0/1 [00:00<?, ?it/s]

Writing model shards: 100%|██████████| 1/1 [00:08<00:00,  8.95s/it]
Writing model shards: 100%|██████████| 1/1 [00:08<00:00,  8.95s/it]
/home/dungpq6/Project/SpatialStack/src/qwen_vl/model/geometry_encoders/vggt_encoder.py:285: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(dtype=dtype):

 71%|███████   | 7001/9864 [29:26:36<27:03:05, 34.02s/it]
 71%|███████   | 7002/9864 [29:26:50<22:19:27, 28.08s/it]
 71%|███████   | 7003/9864 [29:27:06<19:24:36, 24.42s/it]
 71%|███████   | 7004/9864 [29:27:20<16:58:33, 21.37s/it]
 71%|███████   | 7005/9864 [29:27:35<15:18:11, 19.27s/it]
 71%|███████   | 7006/9864 [29:27:51<14:36:03, 18.39s/it]
 71%|███████   | 7007/9864 [29:28:06<13:41:28, 17.25s/it]
 71%|███████   | 7008/9864 [29:28:22<13:26:30, 16.94s/it]
 71%|███████   | 7009/9864 [29:28:38<13:14:01, 16.69s/it]
 71%|███████   | 7010/9864 [29:28:52<12:43:47, 16.06s/it]
                                                         
{'loss': '0.08633', 'grad_norm': '3.338', 'learning_rate': '4.081e-06', 'epoch': '0.7107'}

 71%|███████   | 7010/9864 [29:28:52<12:43:47, 16.06s/it]
 71%|███████   | 7011/9864 [29:29:08<12:42:47, 16.04s/it]
 71%|███████   | 7012/9864 [29:29:23<12:15:02, 15.46s/it]
 71%|███████   | 7013/9864 [29:29:36<11:50:22, 14.95s/it]
 71%|███████   | 7014/9864 [29:29:53<12:12:43, 15.43s/it]
 71%|███████   | 7015/9864 [29:30:09<12:18:24, 15.55s/it]
 71%|███████   | 7016/9864 [29:30:23<11:58:21, 15.13s/it]
 71%|███████   | 7017/9864 [29:30:39<12:13:45, 15.46s/it]
 71%|███████   | 7018/9864 [29:30:53<11:52:46, 15.03s/it]
 71%|███████   | 7019/9864 [29:31:08<11:53:18, 15.04s/it]
 71%|███████   | 7020/9864 [29:31:23<11:44:50, 14.87s/it]
                                                         
{'loss': '0.1014', 'grad_norm': '3.976', 'learning_rate': '4.055e-06', 'epoch': '0.7117'}

 71%|███████   | 7020/9864 [29:31:23<11:44:50, 14.87s/it]
 71%|███████   | 7021/9864 [29:31:39<12:00:26, 15.20s/it]
 71%|███████   | 7022/9864 [29:31:53<11:51:08, 15.01s/it]
 71%|███████   | 7023/9864 [29:32:07<11:36:36, 14.71s/it]
 71%|███████   | 7024/9864 [29:32:24<12:05:00, 15.32s/it]
 71%|███████   | 7025/9864 [29:32:38<11:42:59, 14.86s/it]
 71%|███████   | 7026/9864 [29:32:54<11:59:24, 15.21s/it]
 71%|███████   | 7027/9864 [29:33:08<11:42:13, 14.85s/it]
 71%|███████   | 7028/9864 [29:33:24<12:00:57, 15.25s/it]
 71%|███████▏  | 7029/9864 [29:33:39<11:52:04, 15.07s/it]
 71%|███████▏  | 7030/9864 [29:33:53<11:45:13, 14.93s/it]
                                                         
{'loss': '0.09371', 'grad_norm': '4.723', 'learning_rate': '4.029e-06', 'epoch': '0.7128'}

 71%|███████▏  | 7030/9864 [29:33:53<11:45:13, 14.93s/it]
 71%|███████▏  | 7031/9864 [29:34:09<12:01:58, 15.29s/it]
 71%|███████▏  | 7032/9864 [29:34:25<12:12:22, 15.52s/it]
 71%|███████▏  | 7033/9864 [29:34:42<12:29:49, 15.89s/it]
 71%|███████▏  | 7034/9864 [29:34:57<12:08:06, 15.44s/it]
 71%|███████▏  | 7035/9864 [29:35:11<11:47:39, 15.01s/it]
 71%|███████▏  | 7036/9864 [29:35:25<11:39:17, 14.84s/it]
 71%|███████▏  | 7037/9864 [29:35:41<11:51:45, 15.11s/it]
 71%|███████▏  | 7038/9864 [29:35:56<12:00:07, 15.29s/it]
 71%|███████▏  | 7039/9864 [29:36:12<12:06:45, 15.44s/it]
 71%|███████▏  | 7040/9864 [29:36:26<11:47:58, 15.04s/it]
                                                         
{'loss': '0.1051', 'grad_norm': '6.138', 'learning_rate': '4.002e-06', 'epoch': '0.7138'}

 71%|███████▏  | 7040/9864 [29:36:26<11:47:58, 15.04s/it]
 71%|███████▏  | 7041/9864 [29:36:42<11:59:19, 15.29s/it]
 71%|███████▏  | 7042/9864 [29:36:57<11:48:59, 15.07s/it]
 71%|███████▏  | 7043/9864 [29:37:11<11:38:25, 14.85s/it]
 71%|███████▏  | 7044/9864 [29:37:26<11:34:55, 14.79s/it]
 71%|███████▏  | 7045/9864 [29:37:42<11:55:27, 15.23s/it]
 71%|███████▏  | 7046/9864 [29:37:58<11:59:48, 15.33s/it]
 71%|███████▏  | 7047/9864 [29:38:13<12:00:23, 15.34s/it]
 71%|███████▏  | 7048/9864 [29:38:29<12:08:40, 15.53s/it]
 71%|███████▏  | 7049/9864 [29:38:43<11:47:24, 15.08s/it]
 71%|███████▏  | 7050/9864 [29:38:57<11:33:22, 14.78s/it]
                                                         
{'loss': '0.1077', 'grad_norm': '4.282', 'learning_rate': '3.976e-06', 'epoch': '0.7148'}

 71%|███████▏  | 7050/9864 [29:38:57<11:33:22, 14.78s/it]
 71%|███████▏  | 7051/9864 [29:39:11<11:18:13, 14.47s/it]
 71%|███████▏  | 7052/9864 [29:39:26<11:32:45, 14.78s/it]
 72%|███████▏  | 7053/9864 [29:39:40<11:20:25, 14.52s/it]
 72%|███████▏  | 7054/9864 [29:39:55<11:19:53, 14.52s/it]
 72%|███████▏  | 7055/9864 [29:40:13<12:14:31, 15.69s/it]
 72%|███████▏  | 7056/9864 [29:40:29<12:24:08, 15.90s/it]
 72%|███████▏  | 7057/9864 [29:40:44<12:04:42, 15.49s/it]
 72%|███████▏  | 7058/9864 [29:40:59<11:50:46, 15.20s/it]
 72%|███████▏  | 7059/9864 [29:41:13<11:40:12, 14.98s/it]
 72%|███████▏  | 7060/9864 [29:41:29<11:55:52, 15.32s/it]
                                                         
{'loss': '0.08503', 'grad_norm': '4.439', 'learning_rate': '3.95e-06', 'epoch': '0.7158'}

 72%|███████▏  | 7060/9864 [29:41:29<11:55:52, 15.32s/it]
 72%|███████▏  | 7061/9864 [29:41:44<11:42:26, 15.04s/it]
 72%|███████▏  | 7062/9864 [29:42:00<11:57:43, 15.37s/it]
 72%|███████▏  | 7063/9864 [29:42:15<11:59:39, 15.42s/it]
 72%|███████▏  | 7064/9864 [29:42:30<11:47:52, 15.17s/it]
 72%|███████▏  | 7065/9864 [29:42:45<11:54:29, 15.32s/it]
 72%|███████▏  | 7066/9864 [29:43:00<11:39:42, 15.00s/it]
 72%|███████▏  | 7067/9864 [29:43:15<11:45:12, 15.13s/it]
 72%|███████▏  | 7068/9864 [29:43:29<11:33:29, 14.88s/it]
 72%|███████▏  | 7069/9864 [29:43:44<11:23:06, 14.66s/it]
 72%|███████▏  | 7070/9864 [29:43:59<11:34:56, 14.92s/it]
                                                         
{'loss': '0.09553', 'grad_norm': '4.579', 'learning_rate': '3.924e-06', 'epoch': '0.7168'}

 72%|███████▏  | 7070/9864 [29:43:59<11:34:56, 14.92s/it]
 72%|███████▏  | 7071/9864 [29:44:13<11:19:20, 14.59s/it]
 72%|███████▏  | 7072/9864 [29:44:31<12:00:57, 15.49s/it]
 72%|███████▏  | 7073/9864 [29:44:45<11:46:32, 15.19s/it]
 72%|███████▏  | 7074/9864 [29:44:59<11:25:28, 14.74s/it]
 72%|███████▏  | 7075/9864 [29:45:15<11:42:12, 15.11s/it]
 72%|███████▏  | 7076/9864 [29:45:29<11:30:25, 14.86s/it]
 72%|███████▏  | 7077/9864 [29:45:43<11:23:42, 14.72s/it]
 72%|███████▏  | 7078/9864 [29:45:59<11:37:03, 15.01s/it]
 72%|███████▏  | 7079/9864 [29:46:16<12:10:21, 15.73s/it]
 72%|███████▏  | 7080/9864 [29:46:31<11:54:38, 15.40s/it]
                                                         
{'loss': '0.1109', 'grad_norm': '4.263', 'learning_rate': '3.898e-06', 'epoch': '0.7178'}

 72%|███████▏  | 7080/9864 [29:46:31<11:54:38, 15.40s/it]
 72%|███████▏  | 7081/9864 [29:46:45<11:38:03, 15.05s/it]
 72%|███████▏  | 7082/9864 [29:47:00<11:26:19, 14.80s/it]
 72%|███████▏  | 7083/9864 [29:47:14<11:19:33, 14.66s/it]
 72%|███████▏  | 7084/9864 [29:47:30<11:42:27, 15.16s/it]
 72%|███████▏  | 7085/9864 [29:47:45<11:32:03, 14.94s/it]
 72%|███████▏  | 7086/9864 [29:48:03<12:13:10, 15.84s/it]
 72%|███████▏  | 7087/9864 [29:48:19<12:26:30, 16.13s/it]
 72%|███████▏  | 7088/9864 [29:48:33<11:55:05, 15.46s/it]
 72%|███████▏  | 7089/9864 [29:48:48<11:38:54, 15.11s/it]
 72%|███████▏  | 7090/9864 [29:49:02<11:28:42, 14.90s/it]
                                                         
{'loss': '0.09271', 'grad_norm': '4.713', 'learning_rate': '3.872e-06', 'epoch': '0.7188'}

 72%|███████▏  | 7090/9864 [29:49:02<11:28:42, 14.90s/it]
 72%|███████▏  | 7091/9864 [29:49:16<11:17:07, 14.65s/it]
 72%|███████▏  | 7092/9864 [29:49:32<11:38:08, 15.11s/it]
 72%|███████▏  | 7093/9864 [29:49:46<11:23:18, 14.80s/it]
 72%|███████▏  | 7094/9864 [29:50:04<11:59:41, 15.59s/it]
 72%|███████▏  | 7095/9864 [29:50:20<12:07:08, 15.76s/it]
 72%|███████▏  | 7096/9864 [29:50:34<11:51:17, 15.42s/it]
 72%|███████▏  | 7097/9864 [29:50:48<11:27:12, 14.90s/it]
 72%|███████▏  | 7098/9864 [29:51:02<11:13:45, 14.62s/it]
 72%|███████▏  | 7099/9864 [29:51:16<11:06:29, 14.46s/it]
 72%|███████▏  | 7100/9864 [29:51:32<11:25:09, 14.87s/it]
                                                         
{'loss': '0.106', 'grad_norm': '3.521', 'learning_rate': '3.846e-06', 'epoch': '0.7198'}

 72%|███████▏  | 7100/9864 [29:51:32<11:25:09, 14.87s/it]
 72%|███████▏  | 7101/9864 [29:51:46<11:09:34, 14.54s/it]
 72%|███████▏  | 7102/9864 [29:52:03<11:46:56, 15.36s/it]
 72%|███████▏  | 7103/9864 [29:52:19<11:56:53, 15.58s/it]
 72%|███████▏  | 7104/9864 [29:52:33<11:30:35, 15.01s/it]
 72%|███████▏  | 7105/9864 [29:52:47<11:16:54, 14.72s/it]
 72%|███████▏  | 7106/9864 [29:53:01<11:05:08, 14.47s/it]
 72%|███████▏  | 7107/9864 [29:53:17<11:32:53, 15.08s/it]
 72%|███████▏  | 7108/9864 [29:53:32<11:23:07, 14.87s/it]
 72%|███████▏  | 7109/9864 [29:53:48<11:39:50, 15.24s/it]
 72%|███████▏  | 7110/9864 [29:54:03<11:45:13, 15.36s/it]
                                                         
{'loss': '0.1035', 'grad_norm': '3.51', 'learning_rate': '3.82e-06', 'epoch': '0.7209'}

 72%|███████▏  | 7110/9864 [29:54:03<11:45:13, 15.36s/it]
 72%|███████▏  | 7111/9864 [29:54:20<11:55:08, 15.59s/it]
 72%|███████▏  | 7112/9864 [29:54:34<11:39:12, 15.24s/it]
 72%|███████▏  | 7113/9864 [29:54:48<11:18:17, 14.79s/it]
 72%|███████▏  | 7114/9864 [29:55:02<11:04:36, 14.50s/it]
 72%|███████▏  | 7115/9864 [29:55:18<11:27:55, 15.01s/it]
 72%|███████▏  | 7116/9864 [29:55:32<11:18:59, 14.83s/it]
 72%|███████▏  | 7117/9864 [29:55:48<11:26:59, 15.01s/it]
 72%|███████▏  | 7118/9864 [29:56:05<12:00:51, 15.75s/it]
 72%|███████▏  | 7119/9864 [29:56:19<11:32:35, 15.14s/it]
 72%|███████▏  | 7120/9864 [29:56:33<11:21:57, 14.91s/it]
                                                         
{'loss': '0.08718', 'grad_norm': '3.582', 'learning_rate': '3.794e-06', 'epoch': '0.7219'}

 72%|███████▏  | 7120/9864 [29:56:33<11:21:57, 14.91s/it]
 72%|███████▏  | 7121/9864 [29:56:48<11:14:21, 14.75s/it]
 72%|███████▏  | 7122/9864 [29:57:02<11:09:17, 14.65s/it]
 72%|███████▏  | 7123/9864 [29:57:18<11:24:03, 14.97s/it]
 72%|███████▏  | 7124/9864 [29:57:32<11:17:27, 14.83s/it]
 72%|███████▏  | 7125/9864 [29:57:52<12:23:31, 16.29s/it]
 72%|███████▏  | 7126/9864 [29:58:06<11:56:01, 15.69s/it]
 72%|███████▏  | 7127/9864 [29:58:20<11:35:48, 15.25s/it]
 72%|███████▏  | 7128/9864 [29:58:34<11:14:45, 14.80s/it]
 72%|███████▏  | 7129/9864 [29:58:49<11:09:08, 14.68s/it]
 72%|███████▏  | 7130/9864 [29:59:03<11:02:35, 14.54s/it]
                                                         
{'loss': '0.08558', 'grad_norm': '3.278', 'learning_rate': '3.768e-06', 'epoch': '0.7229'}

 72%|███████▏  | 7130/9864 [29:59:03<11:02:35, 14.54s/it]
 72%|███████▏  | 7131/9864 [29:59:19<11:27:08, 15.09s/it]
 72%|███████▏  | 7132/9864 [29:59:36<11:45:59, 15.51s/it]
 72%|███████▏  | 7133/9864 [29:59:53<12:15:01, 16.15s/it]
 72%|███████▏  | 7134/9864 [30:00:08<11:53:25, 15.68s/it]
 72%|███████▏  | 7135/9864 [30:00:22<11:34:55, 15.28s/it]
 72%|███████▏  | 7136/9864 [30:00:37<11:21:59, 15.00s/it]
 72%|███████▏  | 7137/9864 [30:00:51<11:13:55, 14.83s/it]
 72%|███████▏  | 7138/9864 [30:01:05<11:09:53, 14.74s/it]
 72%|███████▏  | 7139/9864 [30:01:21<11:22:13, 15.02s/it]
 72%|███████▏  | 7140/9864 [30:01:37<11:35:41, 15.32s/it]
                                                         
{'loss': '0.111', 'grad_norm': '3.864', 'learning_rate': '3.743e-06', 'epoch': '0.7239'}

 72%|███████▏  | 7140/9864 [30:01:37<11:35:41, 15.32s/it]
 72%|███████▏  | 7141/9864 [30:01:54<12:01:09, 15.89s/it]
 72%|███████▏  | 7142/9864 [30:02:09<11:38:47, 15.40s/it]
 72%|███████▏  | 7143/9864 [30:02:23<11:24:21, 15.09s/it]
 72%|███████▏  | 7144/9864 [30:02:38<11:18:58, 14.98s/it]
 72%|███████▏  | 7145/9864 [30:02:52<11:06:28, 14.71s/it]
 72%|███████▏  | 7146/9864 [30:03:06<11:02:37, 14.63s/it]
 72%|███████▏  | 7147/9864 [30:03:20<10:55:15, 14.47s/it]
 72%|███████▏  | 7148/9864 [30:03:40<12:03:43, 15.99s/it]
 72%|███████▏  | 7149/9864 [30:03:54<11:41:38, 15.51s/it]
 72%|███████▏  | 7150/9864 [30:04:10<11:48:22, 15.66s/it]
                                                         
{'loss': '0.09854', 'grad_norm': '3.396', 'learning_rate': '3.717e-06', 'epoch': '0.7249'}

 72%|███████▏  | 7150/9864 [30:04:10<11:48:22, 15.66s/it]
 72%|███████▏  | 7151/9864 [30:04:25<11:30:26, 15.27s/it]
 73%|███████▎  | 7152/9864 [30:04:39<11:12:38, 14.88s/it]
 73%|███████▎  | 7153/9864 [30:04:53<11:05:49, 14.74s/it]
 73%|███████▎  | 7154/9864 [30:05:07<11:00:30, 14.62s/it]
 73%|███████▎  | 7155/9864 [30:05:22<10:55:31, 14.52s/it]
 73%|███████▎  | 7156/9864 [30:05:40<11:42:46, 15.57s/it]
 73%|███████▎  | 7157/9864 [30:05:56<11:55:13, 15.85s/it]
 73%|███████▎  | 7158/9864 [30:06:12<11:57:24, 15.91s/it]
 73%|███████▎  | 7159/9864 [30:06:26<11:33:32, 15.38s/it]
 73%|███████▎  | 7160/9864 [30:06:40<11:14:01, 14.96s/it]
                                                         
{'loss': '0.08558', 'grad_norm': '2.11', 'learning_rate': '3.692e-06', 'epoch': '0.7259'}

 73%|███████▎  | 7160/9864 [30:06:40<11:14:01, 14.96s/it]
 73%|███████▎  | 7161/9864 [30:06:55<11:07:41, 14.82s/it]
 73%|███████▎  | 7162/9864 [30:07:09<10:54:55, 14.54s/it]
 73%|███████▎  | 7163/9864 [30:07:27<11:40:48, 15.57s/it]
 73%|███████▎  | 7164/9864 [30:07:41<11:19:43, 15.11s/it]
 73%|███████▎  | 7165/9864 [30:07:57<11:31:36, 15.37s/it]
 73%|███████▎  | 7166/9864 [30:08:13<11:39:14, 15.55s/it]
 73%|███████▎  | 7167/9864 [30:08:28<11:30:35, 15.36s/it]
 73%|███████▎  | 7168/9864 [30:08:42<11:14:46, 15.02s/it]
 73%|███████▎  | 7169/9864 [30:08:56<11:03:17, 14.77s/it]
 73%|███████▎  | 7170/9864 [30:09:10<10:58:23, 14.66s/it]
                                                         
{'loss': '0.0875', 'grad_norm': '4.361', 'learning_rate': '3.666e-06', 'epoch': '0.7269'}

 73%|███████▎  | 7170/9864 [30:09:10<10:58:23, 14.66s/it]
 73%|███████▎  | 7171/9864 [30:09:28<11:36:57, 15.53s/it]
 73%|███████▎  | 7172/9864 [30:09:42<11:20:00, 15.16s/it]
 73%|███████▎  | 7173/9864 [30:09:58<11:32:51, 15.45s/it]
 73%|███████▎  | 7174/9864 [30:10:14<11:38:51, 15.59s/it]
 73%|███████▎  | 7175/9864 [30:10:28<11:18:00, 15.13s/it]
 73%|███████▎  | 7176/9864 [30:10:43<11:11:57, 15.00s/it]
 73%|███████▎  | 7177/9864 [30:10:57<10:58:07, 14.70s/it]
 73%|███████▎  | 7178/9864 [30:11:13<11:17:05, 15.12s/it]
 73%|███████▎  | 7179/9864 [30:11:29<11:21:46, 15.24s/it]
 73%|███████▎  | 7180/9864 [30:11:43<11:12:53, 15.04s/it]
                                                         
{'loss': '0.0953', 'grad_norm': '4.561', 'learning_rate': '3.641e-06', 'epoch': '0.728'}

 73%|███████▎  | 7180/9864 [30:11:43<11:12:53, 15.04s/it]
 73%|███████▎  | 7181/9864 [30:11:59<11:26:32, 15.35s/it]
 73%|███████▎  | 7182/9864 [30:12:16<11:37:35, 15.61s/it]
 73%|███████▎  | 7183/9864 [30:12:30<11:17:16, 15.16s/it]
 73%|███████▎  | 7184/9864 [30:12:44<11:09:18, 14.98s/it]
 73%|███████▎  | 7185/9864 [30:12:58<10:52:34, 14.62s/it]
 73%|███████▎  | 7186/9864 [30:13:16<11:41:33, 15.72s/it]
 73%|███████▎  | 7187/9864 [30:13:30<11:17:57, 15.20s/it]
 73%|███████▎  | 7188/9864 [30:13:44<10:59:09, 14.78s/it]
 73%|███████▎  | 7189/9864 [30:14:02<11:36:39, 15.63s/it]
 73%|███████▎  | 7190/9864 [30:14:16<11:14:23, 15.13s/it]
                                                         
{'loss': '0.08418', 'grad_norm': '4.598', 'learning_rate': '3.616e-06', 'epoch': '0.729'}

 73%|███████▎  | 7190/9864 [30:14:16<11:14:23, 15.13s/it]
 73%|███████▎  | 7191/9864 [30:14:30<11:09:42, 15.03s/it]
 73%|███████▎  | 7192/9864 [30:14:45<11:09:35, 15.04s/it]
 73%|███████▎  | 7193/9864 [30:15:01<11:09:42, 15.04s/it]
 73%|███████▎  | 7194/9864 [30:15:19<12:00:31, 16.19s/it]
 73%|███████▎  | 7195/9864 [30:15:35<11:46:42, 15.89s/it]
 73%|███████▎  | 7196/9864 [30:15:51<11:49:50, 15.96s/it]
 73%|███████▎  | 7197/9864 [30:16:07<11:52:43, 16.03s/it]
 73%|███████▎  | 7198/9864 [30:16:21<11:28:06, 15.49s/it]
 73%|███████▎  | 7199/9864 [30:16:35<11:08:18, 15.05s/it]
 73%|███████▎  | 7200/9864 [30:16:49<10:58:19, 14.83s/it]
                                                         
{'loss': '0.07899', 'grad_norm': '2.801', 'learning_rate': '3.59e-06', 'epoch': '0.73'}

 73%|███████▎  | 7200/9864 [30:16:49<10:58:19, 14.83s/it]
 73%|███████▎  | 7201/9864 [30:17:06<11:19:40, 15.31s/it]
 73%|███████▎  | 7202/9864 [30:17:22<11:24:10, 15.42s/it]
 73%|███████▎  | 7203/9864 [30:17:36<11:11:15, 15.14s/it]
 73%|███████▎  | 7204/9864 [30:17:53<11:35:52, 15.70s/it]
 73%|███████▎  | 7205/9864 [30:18:10<11:46:45, 15.95s/it]
 73%|███████▎  | 7206/9864 [30:18:24<11:25:05, 15.46s/it]
 73%|███████▎  | 7207/9864 [30:18:39<11:15:40, 15.26s/it]
 73%|███████▎  | 7208/9864 [30:18:53<11:03:52, 15.00s/it]
 73%|███████▎  | 7209/9864 [30:19:07<10:53:27, 14.77s/it]
 73%|███████▎  | 7210/9864 [30:19:25<11:33:36, 15.68s/it]
                                                         
{'loss': '0.08116', 'grad_norm': '4.978', 'learning_rate': '3.565e-06', 'epoch': '0.731'}

 73%|███████▎  | 7210/9864 [30:19:25<11:33:36, 15.68s/it]
 73%|███████▎  | 7211/9864 [30:19:40<11:22:01, 15.42s/it]
 73%|███████▎  | 7212/9864 [30:19:56<11:28:12, 15.57s/it]
 73%|███████▎  | 7213/9864 [30:20:12<11:37:07, 15.78s/it]
 73%|███████▎  | 7214/9864 [30:20:26<11:13:49, 15.26s/it]
 73%|███████▎  | 7215/9864 [30:20:40<10:53:33, 14.80s/it]
 73%|███████▎  | 7216/9864 [30:20:54<10:38:07, 14.46s/it]
 73%|███████▎  | 7217/9864 [30:21:10<11:06:50, 15.12s/it]
 73%|███████▎  | 7218/9864 [30:21:26<11:15:47, 15.32s/it]
 73%|███████▎  | 7219/9864 [30:21:40<10:57:07, 14.91s/it]
 73%|███████▎  | 7220/9864 [30:21:55<10:52:37, 14.81s/it]
                                                         
{'loss': '0.07174', 'grad_norm': '3.889', 'learning_rate': '3.54e-06', 'epoch': '0.732'}

 73%|███████▎  | 7220/9864 [30:21:55<10:52:37, 14.81s/it]
 73%|███████▎  | 7221/9864 [30:22:12<11:32:05, 15.71s/it]
 73%|███████▎  | 7222/9864 [30:22:27<11:13:49, 15.30s/it]
 73%|███████▎  | 7223/9864 [30:22:41<10:54:42, 14.87s/it]
 73%|███████▎  | 7224/9864 [30:22:57<11:13:19, 15.30s/it]
 73%|███████▎  | 7225/9864 [30:23:13<11:22:43, 15.52s/it]
 73%|███████▎  | 7226/9864 [30:23:27<11:00:07, 15.01s/it]
 73%|███████▎  | 7227/9864 [30:23:41<10:52:53, 14.86s/it]
 73%|███████▎  | 7228/9864 [30:23:55<10:44:00, 14.66s/it]
 73%|███████▎  | 7229/9864 [30:24:12<11:03:33, 15.11s/it]
 73%|███████▎  | 7230/9864 [30:24:26<10:50:34, 14.82s/it]
                                                         
{'loss': '0.09367', 'grad_norm': '7.115', 'learning_rate': '3.515e-06', 'epoch': '0.733'}

 73%|███████▎  | 7230/9864 [30:24:26<10:50:34, 14.82s/it]
 73%|███████▎  | 7231/9864 [30:24:40<10:40:37, 14.60s/it]
 73%|███████▎  | 7232/9864 [30:24:54<10:31:28, 14.40s/it]
 73%|███████▎  | 7233/9864 [30:25:09<10:48:07, 14.78s/it]
 73%|███████▎  | 7234/9864 [30:25:25<11:01:16, 15.09s/it]
 73%|███████▎  | 7235/9864 [30:25:39<10:45:51, 14.74s/it]
 73%|███████▎  | 7236/9864 [30:25:57<11:27:46, 15.70s/it]
 73%|███████▎  | 7237/9864 [30:26:11<11:05:24, 15.20s/it]
 73%|███████▎  | 7238/9864 [30:26:26<10:56:10, 14.99s/it]
 73%|███████▎  | 7239/9864 [30:26:40<10:47:42, 14.80s/it]
 73%|███████▎  | 7240/9864 [30:26:56<11:06:14, 15.23s/it]
                                                         
{'loss': '0.08274', 'grad_norm': '4.573', 'learning_rate': '3.49e-06', 'epoch': '0.734'}

 73%|███████▎  | 7240/9864 [30:26:56<11:06:14, 15.23s/it]
 73%|███████▎  | 7241/9864 [30:27:10<10:48:42, 14.84s/it]
 73%|███████▎  | 7242/9864 [30:27:26<11:03:22, 15.18s/it]
 73%|███████▎  | 7243/9864 [30:27:40<10:48:44, 14.85s/it]
 73%|███████▎  | 7244/9864 [30:27:58<11:24:05, 15.67s/it]
 73%|███████▎  | 7245/9864 [30:28:12<11:09:02, 15.33s/it]
 73%|███████▎  | 7246/9864 [30:28:27<11:03:12, 15.20s/it]
 73%|███████▎  | 7247/9864 [30:28:42<10:50:59, 14.93s/it]
 73%|███████▎  | 7248/9864 [30:28:58<11:06:51, 15.29s/it]
 73%|███████▎  | 7249/9864 [30:29:13<11:13:10, 15.45s/it]
 73%|███████▎  | 7250/9864 [30:29:28<11:00:24, 15.16s/it]
                                                         
{'loss': '0.0921', 'grad_norm': '4.599', 'learning_rate': '3.465e-06', 'epoch': '0.7351'}

 73%|███████▎  | 7250/9864 [30:29:28<11:00:24, 15.16s/it]
 74%|███████▎  | 7251/9864 [30:29:46<11:31:54, 15.89s/it]
 74%|███████▎  | 7252/9864 [30:30:00<11:17:50, 15.57s/it]
 74%|███████▎  | 7253/9864 [30:30:15<11:03:31, 15.25s/it]
 74%|███████▎  | 7254/9864 [30:30:30<10:55:19, 15.06s/it]
 74%|███████▎  | 7255/9864 [30:30:46<11:09:19, 15.39s/it]
 74%|███████▎  | 7256/9864 [30:31:00<10:49:47, 14.95s/it]
 74%|███████▎  | 7257/9864 [30:31:16<11:05:34, 15.32s/it]
 74%|███████▎  | 7258/9864 [30:31:30<10:56:34, 15.12s/it]
 74%|███████▎  | 7259/9864 [30:31:47<11:19:06, 15.64s/it]
 74%|███████▎  | 7260/9864 [30:32:03<11:24:37, 15.77s/it]
                                                         
{'loss': '0.07571', 'grad_norm': '3.138', 'learning_rate': '3.44e-06', 'epoch': '0.7361'}

 74%|███████▎  | 7260/9864 [30:32:03<11:24:37, 15.77s/it]
 74%|███████▎  | 7261/9864 [30:32:18<11:04:13, 15.31s/it]
 74%|███████▎  | 7262/9864 [30:32:34<11:18:42, 15.65s/it]
 74%|███████▎  | 7263/9864 [30:32:48<11:00:19, 15.23s/it]
 74%|███████▎  | 7264/9864 [30:33:05<11:13:22, 15.54s/it]
 74%|███████▎  | 7265/9864 [30:33:19<10:57:05, 15.17s/it]
 74%|███████▎  | 7266/9864 [30:33:34<11:02:02, 15.29s/it]
 74%|███████▎  | 7267/9864 [30:33:49<10:50:37, 15.03s/it]
 74%|███████▎  | 7268/9864 [30:34:05<11:05:24, 15.38s/it]
 74%|███████▎  | 7269/9864 [30:34:19<10:48:14, 14.99s/it]
 74%|███████▎  | 7270/9864 [30:34:35<10:56:31, 15.19s/it]
                                                         
{'loss': '0.08273', 'grad_norm': '2.777', 'learning_rate': '3.416e-06', 'epoch': '0.7371'}

 74%|███████▎  | 7270/9864 [30:34:35<10:56:31, 15.19s/it]
 74%|███████▎  | 7271/9864 [30:34:49<10:43:16, 14.89s/it]
 74%|███████▎  | 7272/9864 [30:35:03<10:34:26, 14.69s/it]
 74%|███████▎  | 7273/9864 [30:35:20<10:58:35, 15.25s/it]
 74%|███████▎  | 7274/9864 [30:35:36<11:06:08, 15.43s/it]
 74%|███████▍  | 7275/9864 [30:35:52<11:17:36, 15.70s/it]
 74%|███████▍  | 7276/9864 [30:36:06<11:01:29, 15.34s/it]
 74%|███████▍  | 7277/9864 [30:36:20<10:43:26, 14.92s/it]
 74%|███████▍  | 7278/9864 [30:36:37<11:00:14, 15.32s/it]
 74%|███████▍  | 7279/9864 [30:36:51<10:46:05, 15.00s/it]
 74%|███████▍  | 7280/9864 [30:37:07<10:59:45, 15.32s/it]
                                                         
{'loss': '0.09235', 'grad_norm': '3.59', 'learning_rate': '3.391e-06', 'epoch': '0.7381'}

 74%|███████▍  | 7280/9864 [30:37:07<10:59:45, 15.32s/it]
 74%|███████▍  | 7281/9864 [30:37:21<10:42:20, 14.92s/it]
 74%|███████▍  | 7282/9864 [30:37:37<10:55:52, 15.24s/it]
 74%|███████▍  | 7283/9864 [30:37:51<10:44:37, 14.99s/it]
 74%|███████▍  | 7284/9864 [30:38:08<11:01:40, 15.39s/it]
 74%|███████▍  | 7285/9864 [30:38:22<10:52:20, 15.18s/it]
 74%|███████▍  | 7286/9864 [30:38:39<11:14:07, 15.69s/it]
 74%|███████▍  | 7287/9864 [30:38:55<11:18:56, 15.81s/it]
 74%|███████▍  | 7288/9864 [30:39:09<10:50:30, 15.15s/it]
 74%|███████▍  | 7289/9864 [30:39:23<10:39:25, 14.90s/it]
 74%|███████▍  | 7290/9864 [30:39:37<10:27:08, 14.62s/it]
                                                         
{'loss': '0.07656', 'grad_norm': '3.834', 'learning_rate': '3.366e-06', 'epoch': '0.7391'}

 74%|███████▍  | 7290/9864 [30:39:37<10:27:08, 14.62s/it]
 74%|███████▍  | 7291/9864 [30:39:55<11:03:43, 15.48s/it]
 74%|███████▍  | 7292/9864 [30:40:09<10:53:12, 15.24s/it]
 74%|███████▍  | 7293/9864 [30:40:24<10:41:52, 14.98s/it]
 74%|███████▍  | 7294/9864 [30:40:40<10:59:05, 15.39s/it]
 74%|███████▍  | 7295/9864 [30:40:56<11:07:49, 15.60s/it]
 74%|███████▍  | 7296/9864 [30:41:11<10:55:01, 15.30s/it]
 74%|███████▍  | 7297/9864 [30:41:25<10:40:58, 14.98s/it]
 74%|███████▍  | 7298/9864 [30:41:41<10:56:16, 15.35s/it]
 74%|███████▍  | 7299/9864 [30:41:58<11:09:27, 15.66s/it]
 74%|███████▍  | 7300/9864 [30:42:12<10:56:37, 15.37s/it]
                                                         
{'loss': '0.09106', 'grad_norm': '3.979', 'learning_rate': '3.342e-06', 'epoch': '0.7401'}

 74%|███████▍  | 7300/9864 [30:42:12<10:56:37, 15.37s/it]
 74%|███████▍  | 7301/9864 [30:42:28<10:59:04, 15.43s/it]
 74%|███████▍  | 7302/9864 [30:42:44<11:06:47, 15.62s/it]
 74%|███████▍  | 7303/9864 [30:42:58<10:41:23, 15.03s/it]
 74%|███████▍  | 7304/9864 [30:43:12<10:32:28, 14.82s/it]
 74%|███████▍  | 7305/9864 [30:43:26<10:28:27, 14.74s/it]
 74%|███████▍  | 7306/9864 [30:43:44<11:08:48, 15.69s/it]
 74%|███████▍  | 7307/9864 [30:43:59<10:50:22, 15.26s/it]
 74%|███████▍  | 7308/9864 [30:44:13<10:33:45, 14.88s/it]
 74%|███████▍  | 7309/9864 [30:44:29<10:49:17, 15.25s/it]
 74%|███████▍  | 7310/9864 [30:44:45<11:04:17, 15.61s/it]
                                                         
{'loss': '0.1002', 'grad_norm': '4.302', 'learning_rate': '3.317e-06', 'epoch': '0.7411'}

 74%|███████▍  | 7310/9864 [30:44:45<11:04:17, 15.61s/it]
 74%|███████▍  | 7311/9864 [30:45:00<10:49:02, 15.25s/it]
 74%|███████▍  | 7312/9864 [30:45:14<10:37:11, 14.98s/it]
 74%|███████▍  | 7313/9864 [30:45:30<10:48:12, 15.25s/it]
 74%|███████▍  | 7314/9864 [30:45:44<10:34:17, 14.92s/it]
 74%|███████▍  | 7315/9864 [30:46:01<10:55:07, 15.42s/it]
 74%|███████▍  | 7316/9864 [30:46:15<10:45:31, 15.20s/it]
 74%|███████▍  | 7317/9864 [30:46:31<10:54:21, 15.41s/it]
 74%|███████▍  | 7318/9864 [30:46:47<10:57:26, 15.49s/it]
 74%|███████▍  | 7319/9864 [30:47:01<10:38:37, 15.06s/it]
 74%|███████▍  | 7320/9864 [30:47:16<10:33:42, 14.95s/it]
                                                         
{'loss': '0.09221', 'grad_norm': '5.071', 'learning_rate': '3.293e-06', 'epoch': '0.7422'}

 74%|███████▍  | 7320/9864 [30:47:16<10:33:42, 14.95s/it]
 74%|███████▍  | 7321/9864 [30:47:32<10:49:05, 15.31s/it]
 74%|███████▍  | 7322/9864 [30:47:46<10:36:44, 15.03s/it]
 74%|███████▍  | 7323/9864 [30:48:02<10:50:51, 15.37s/it]
 74%|███████▍  | 7324/9864 [30:48:17<10:41:58, 15.16s/it]
 74%|███████▍  | 7325/9864 [30:48:34<11:11:53, 15.88s/it]
 74%|███████▍  | 7326/9864 [30:48:48<10:46:08, 15.28s/it]
 74%|███████▍  | 7327/9864 [30:49:02<10:29:32, 14.89s/it]
 74%|███████▍  | 7328/9864 [30:49:17<10:20:29, 14.68s/it]
 74%|███████▍  | 7329/9864 [30:49:33<10:40:44, 15.17s/it]
 74%|███████▍  | 7330/9864 [30:49:48<10:34:23, 15.02s/it]
                                                         
{'loss': '0.09179', 'grad_norm': '3.581', 'learning_rate': '3.269e-06', 'epoch': '0.7432'}

 74%|███████▍  | 7330/9864 [30:49:48<10:34:23, 15.02s/it]
 74%|███████▍  | 7331/9864 [30:50:04<10:52:38, 15.46s/it]
 74%|███████▍  | 7332/9864 [30:50:20<10:58:15, 15.60s/it]
 74%|███████▍  | 7333/9864 [30:50:36<11:05:14, 15.77s/it]
 74%|███████▍  | 7334/9864 [30:50:51<10:50:01, 15.42s/it]
 74%|███████▍  | 7335/9864 [30:51:05<10:32:49, 15.01s/it]
 74%|███████▍  | 7336/9864 [30:51:19<10:21:18, 14.75s/it]
 74%|███████▍  | 7337/9864 [30:51:34<10:21:51, 14.77s/it]
 74%|███████▍  | 7338/9864 [30:51:51<10:57:25, 15.62s/it]
 74%|███████▍  | 7339/9864 [30:52:06<10:41:53, 15.25s/it]
 74%|███████▍  | 7340/9864 [30:52:20<10:28:02, 14.93s/it]
                                                         
{'loss': '0.09129', 'grad_norm': '3.775', 'learning_rate': '3.244e-06', 'epoch': '0.7442'}

 74%|███████▍  | 7340/9864 [30:52:20<10:28:02, 14.93s/it]
 74%|███████▍  | 7341/9864 [30:52:37<11:01:34, 15.73s/it]
 74%|███████▍  | 7342/9864 [30:52:52<10:42:23, 15.28s/it]
 74%|███████▍  | 7343/9864 [30:53:06<10:28:12, 14.95s/it]
 74%|███████▍  | 7344/9864 [30:53:20<10:15:21, 14.65s/it]
 74%|███████▍  | 7345/9864 [30:53:34<10:10:23, 14.54s/it]
 74%|███████▍  | 7346/9864 [30:53:52<10:56:10, 15.64s/it]
 74%|███████▍  | 7347/9864 [30:54:07<10:38:43, 15.23s/it]
 74%|███████▍  | 7348/9864 [30:54:25<11:13:32, 16.06s/it]
 75%|███████▍  | 7349/9864 [30:54:39<10:54:48, 15.62s/it]
 75%|███████▍  | 7350/9864 [30:54:54<10:38:43, 15.24s/it]
                                                         
{'loss': '0.08882', 'grad_norm': '2.725', 'learning_rate': '3.22e-06', 'epoch': '0.7452'}

 75%|███████▍  | 7350/9864 [30:54:54<10:38:43, 15.24s/it]
 75%|███████▍  | 7351/9864 [30:55:07<10:20:13, 14.81s/it]
 75%|███████▍  | 7352/9864 [30:55:21<10:08:13, 14.53s/it]
 75%|███████▍  | 7353/9864 [30:55:36<10:06:23, 14.49s/it]
 75%|███████▍  | 7354/9864 [30:55:52<10:34:08, 15.16s/it]
 75%|███████▍  | 7355/9864 [30:56:10<11:11:42, 16.06s/it]
 75%|███████▍  | 7356/9864 [30:56:25<10:49:45, 15.54s/it]
 75%|███████▍  | 7357/9864 [30:56:41<10:56:19, 15.71s/it]
 75%|███████▍  | 7358/9864 [30:56:55<10:39:34, 15.31s/it]
 75%|███████▍  | 7359/9864 [30:57:09<10:24:41, 14.96s/it]
 75%|███████▍  | 7360/9864 [30:57:24<10:19:56, 14.85s/it]
                                                         
{'loss': '0.08509', 'grad_norm': '3.757', 'learning_rate': '3.196e-06', 'epoch': '0.7462'}

 75%|███████▍  | 7360/9864 [30:57:24<10:19:56, 14.85s/it]
 75%|███████▍  | 7361/9864 [30:57:38<10:08:01, 14.58s/it]
 75%|███████▍  | 7362/9864 [30:57:55<10:38:33, 15.31s/it]
 75%|███████▍  | 7363/9864 [30:58:11<10:48:32, 15.56s/it]
 75%|███████▍  | 7364/9864 [30:58:25<10:28:08, 15.08s/it]
 75%|███████▍  | 7365/9864 [30:58:41<10:42:37, 15.43s/it]
 75%|███████▍  | 7366/9864 [30:58:56<10:28:39, 15.10s/it]
 75%|███████▍  | 7367/9864 [30:59:09<10:12:05, 14.71s/it]
 75%|███████▍  | 7368/9864 [30:59:23<10:02:55, 14.49s/it]
 75%|███████▍  | 7369/9864 [30:59:40<10:25:03, 15.03s/it]
 75%|███████▍  | 7370/9864 [30:59:56<10:35:18, 15.28s/it]
                                                         
{'loss': '0.09017', 'grad_norm': '5.803', 'learning_rate': '3.172e-06', 'epoch': '0.7472'}

 75%|███████▍  | 7370/9864 [30:59:56<10:35:18, 15.28s/it]
 75%|███████▍  | 7371/9864 [31:00:12<10:42:49, 15.47s/it]
 75%|███████▍  | 7372/9864 [31:00:28<10:48:42, 15.62s/it]
 75%|███████▍  | 7373/9864 [31:00:42<10:32:02, 15.22s/it]
 75%|███████▍  | 7374/9864 [31:00:56<10:19:09, 14.92s/it]
 75%|███████▍  | 7375/9864 [31:01:10<10:10:16, 14.71s/it]
 75%|███████▍  | 7376/9864 [31:01:25<10:09:20, 14.69s/it]
 75%|███████▍  | 7377/9864 [31:01:41<10:28:58, 15.17s/it]
 75%|███████▍  | 7378/9864 [31:01:59<11:02:24, 15.99s/it]
 75%|███████▍  | 7379/9864 [31:02:15<11:03:32, 16.02s/it]
 75%|███████▍  | 7380/9864 [31:02:29<10:38:30, 15.42s/it]
                                                         
{'loss': '0.08702', 'grad_norm': '5.927', 'learning_rate': '3.148e-06', 'epoch': '0.7482'}

 75%|███████▍  | 7380/9864 [31:02:29<10:38:30, 15.42s/it]
 75%|███████▍  | 7381/9864 [31:02:44<10:26:02, 15.13s/it]
 75%|███████▍  | 7382/9864 [31:02:58<10:14:13, 14.85s/it]
 75%|███████▍  | 7383/9864 [31:03:12<10:07:42, 14.70s/it]
 75%|███████▍  | 7384/9864 [31:03:26<10:01:12, 14.55s/it]
 75%|███████▍  | 7385/9864 [31:03:42<10:17:02, 14.93s/it]
 75%|███████▍  | 7386/9864 [31:04:00<10:50:03, 15.74s/it]
 75%|███████▍  | 7387/9864 [31:04:17<11:01:20, 16.02s/it]
 75%|███████▍  | 7388/9864 [31:04:31<10:38:15, 15.47s/it]
 75%|███████▍  | 7389/9864 [31:04:45<10:24:11, 15.13s/it]
 75%|███████▍  | 7390/9864 [31:04:59<10:07:57, 14.74s/it]
                                                         
{'loss': '0.09025', 'grad_norm': '3.08', 'learning_rate': '3.124e-06', 'epoch': '0.7493'}

 75%|███████▍  | 7390/9864 [31:04:59<10:07:57, 14.74s/it]
 75%|███████▍  | 7391/9864 [31:05:13<9:59:01, 14.53s/it] 
 75%|███████▍  | 7392/9864 [31:05:28<10:00:24, 14.57s/it]
 75%|███████▍  | 7393/9864 [31:05:45<10:37:50, 15.49s/it]
 75%|███████▍  | 7394/9864 [31:06:01<10:44:27, 15.65s/it]
 75%|███████▍  | 7395/9864 [31:06:17<10:47:27, 15.73s/it]
 75%|███████▍  | 7396/9864 [31:06:32<10:30:23, 15.33s/it]
 75%|███████▍  | 7397/9864 [31:06:46<10:16:19, 14.99s/it]
 75%|███████▌  | 7398/9864 [31:07:00<10:02:32, 14.66s/it]
 75%|███████▌  | 7399/9864 [31:07:14<10:04:05, 14.70s/it]
 75%|███████▌  | 7400/9864 [31:07:32<10:42:23, 15.64s/it]
                                                         
{'loss': '0.09013', 'grad_norm': '3.215', 'learning_rate': '3.1e-06', 'epoch': '0.7503'}

 75%|███████▌  | 7400/9864 [31:07:32<10:42:23, 15.64s/it]
 75%|███████▌  | 7401/9864 [31:07:47<10:24:44, 15.22s/it]
 75%|███████▌  | 7402/9864 [31:08:02<10:32:19, 15.41s/it]
 75%|███████▌  | 7403/9864 [31:08:18<10:37:10, 15.53s/it]
 75%|███████▌  | 7404/9864 [31:08:32<10:20:57, 15.15s/it]
 75%|███████▌  | 7405/9864 [31:08:46<10:06:17, 14.79s/it]
 75%|███████▌  | 7406/9864 [31:09:01<10:00:22, 14.66s/it]
 75%|███████▌  | 7407/9864 [31:09:15<9:54:11, 14.51s/it] 
 75%|███████▌  | 7408/9864 [31:09:31<10:15:28, 15.04s/it]
 75%|███████▌  | 7409/9864 [31:09:47<10:30:11, 15.40s/it]
 75%|███████▌  | 7410/9864 [31:10:04<10:38:48, 15.62s/it]
                                                         
{'loss': '0.08236', 'grad_norm': '4.195', 'learning_rate': '3.077e-06', 'epoch': '0.7513'}

 75%|███████▌  | 7410/9864 [31:10:04<10:38:48, 15.62s/it]
 75%|███████▌  | 7411/9864 [31:10:18<10:21:38, 15.21s/it]
 75%|███████▌  | 7412/9864 [31:10:34<10:28:12, 15.37s/it]
 75%|███████▌  | 7413/9864 [31:10:48<10:18:15, 15.13s/it]
 75%|███████▌  | 7414/9864 [31:11:02<10:06:55, 14.86s/it]
 75%|███████▌  | 7415/9864 [31:11:18<10:15:33, 15.08s/it]
 75%|███████▌  | 7416/9864 [31:11:34<10:30:50, 15.46s/it]
 75%|███████▌  | 7417/9864 [31:11:51<10:39:49, 15.69s/it]
 75%|███████▌  | 7418/9864 [31:12:05<10:21:52, 15.25s/it]
 75%|███████▌  | 7419/9864 [31:12:19<10:09:12, 14.95s/it]
 75%|███████▌  | 7420/9864 [31:12:35<10:21:23, 15.26s/it]
                                                         
{'loss': '0.08908', 'grad_norm': '4.608', 'learning_rate': '3.053e-06', 'epoch': '0.7523'}

 75%|███████▌  | 7420/9864 [31:12:35<10:21:23, 15.26s/it]
 75%|███████▌  | 7421/9864 [31:12:49<10:07:39, 14.92s/it]
 75%|███████▌  | 7422/9864 [31:13:03<9:52:43, 14.56s/it] 
 75%|███████▌  | 7423/9864 [31:13:19<10:10:32, 15.01s/it]
 75%|███████▌  | 7424/9864 [31:13:33<10:00:43, 14.77s/it]
 75%|███████▌  | 7425/9864 [31:13:49<10:14:53, 15.13s/it]
 75%|███████▌  | 7426/9864 [31:14:06<10:32:46, 15.57s/it]
 75%|███████▌  | 7427/9864 [31:14:20<10:18:59, 15.24s/it]
 75%|███████▌  | 7428/9864 [31:14:36<10:30:26, 15.53s/it]
 75%|███████▌  | 7429/9864 [31:14:51<10:13:46, 15.12s/it]
 75%|███████▌  | 7430/9864 [31:15:07<10:24:08, 15.39s/it]
                                                         
{'loss': '0.07984', 'grad_norm': '4.197', 'learning_rate': '3.029e-06', 'epoch': '0.7533'}

 75%|███████▌  | 7430/9864 [31:15:07<10:24:08, 15.39s/it]
 75%|███████▌  | 7431/9864 [31:15:21<10:09:52, 15.04s/it]
 75%|███████▌  | 7432/9864 [31:15:35<9:56:54, 14.73s/it] 
 75%|███████▌  | 7433/9864 [31:15:50<10:05:25, 14.94s/it]
 75%|███████▌  | 7434/9864 [31:16:07<10:22:17, 15.37s/it]
 75%|███████▌  | 7435/9864 [31:16:22<10:27:41, 15.50s/it]
 75%|███████▌  | 7436/9864 [31:16:37<10:14:54, 15.20s/it]
 75%|███████▌  | 7437/9864 [31:16:52<10:17:14, 15.26s/it]
 75%|███████▌  | 7438/9864 [31:17:07<10:08:58, 15.06s/it]
 75%|███████▌  | 7439/9864 [31:17:21<9:55:15, 14.73s/it] 
 75%|███████▌  | 7440/9864 [31:17:35<9:47:23, 14.54s/it]
                                                        
{'loss': '0.08561', 'grad_norm': '5.582', 'learning_rate': '3.006e-06', 'epoch': '0.7543'}

 75%|███████▌  | 7440/9864 [31:17:35<9:47:23, 14.54s/it]
 75%|███████▌  | 7441/9864 [31:17:51<10:07:35, 15.05s/it]
 75%|███████▌  | 7442/9864 [31:18:09<10:40:16, 15.86s/it]
 75%|███████▌  | 7443/9864 [31:18:24<10:26:57, 15.54s/it]
 75%|███████▌  | 7444/9864 [31:18:38<10:11:32, 15.16s/it]
 75%|███████▌  | 7445/9864 [31:18:52<10:00:13, 14.89s/it]
 75%|███████▌  | 7446/9864 [31:19:09<10:20:09, 15.39s/it]
 75%|███████▌  | 7447/9864 [31:19:23<10:01:45, 14.94s/it]
 76%|███████▌  | 7448/9864 [31:19:37<9:52:24, 14.71s/it] 
 76%|███████▌  | 7449/9864 [31:19:53<10:06:55, 15.08s/it]
 76%|███████▌  | 7450/9864 [31:20:11<10:39:35, 15.90s/it]
                                                         
{'loss': '0.09263', 'grad_norm': '4.249', 'learning_rate': '2.983e-06', 'epoch': '0.7553'}

 76%|███████▌  | 7450/9864 [31:20:11<10:39:35, 15.90s/it]
 76%|███████▌  | 7451/9864 [31:20:25<10:18:44, 15.39s/it]
 76%|███████▌  | 7452/9864 [31:20:39<10:06:52, 15.10s/it]
 76%|███████▌  | 7453/9864 [31:20:53<9:55:08, 14.81s/it] 
 76%|███████▌  | 7454/9864 [31:21:10<10:19:40, 15.43s/it]
 76%|███████▌  | 7455/9864 [31:21:24<10:05:11, 15.07s/it]
 76%|███████▌  | 7456/9864 [31:21:41<10:19:11, 15.43s/it]
 76%|███████▌  | 7457/9864 [31:21:57<10:27:45, 15.65s/it]
 76%|███████▌  | 7458/9864 [31:22:13<10:30:41, 15.73s/it]
 76%|███████▌  | 7459/9864 [31:22:27<10:13:59, 15.32s/it]
 76%|███████▌  | 7460/9864 [31:22:41<9:57:27, 14.91s/it] 
                                                        
{'loss': '0.07194', 'grad_norm': '2.088', 'learning_rate': '2.959e-06', 'epoch': '0.7563'}

 76%|███████▌  | 7460/9864 [31:22:41<9:57:27, 14.91s/it]
 76%|███████▌  | 7461/9864 [31:22:55<9:47:17, 14.66s/it]
 76%|███████▌  | 7462/9864 [31:23:11<10:04:43, 15.11s/it]
 76%|███████▌  | 7463/9864 [31:23:26<9:54:17, 14.85s/it] 
 76%|███████▌  | 7464/9864 [31:23:42<10:14:10, 15.35s/it]
 76%|███████▌  | 7465/9864 [31:23:59<10:37:50, 15.95s/it]
 76%|███████▌  | 7466/9864 [31:24:14<10:15:53, 15.41s/it]
 76%|███████▌  | 7467/9864 [31:24:28<9:57:58, 14.97s/it] 
 76%|███████▌  | 7468/9864 [31:24:41<9:45:18, 14.66s/it]
 76%|███████▌  | 7469/9864 [31:24:55<9:36:43, 14.45s/it]
 76%|███████▌  | 7470/9864 [31:25:12<9:58:22, 15.00s/it]
                                                        
{'loss': '0.08644', 'grad_norm': '3.588', 'learning_rate': '2.936e-06', 'epoch': '0.7574'}

 76%|███████▌  | 7470/9864 [31:25:12<9:58:22, 15.00s/it]
 76%|███████▌  | 7471/9864 [31:25:25<9:42:30, 14.61s/it]
 76%|███████▌  | 7472/9864 [31:25:43<10:19:22, 15.54s/it]
 76%|███████▌  | 7473/9864 [31:25:59<10:28:58, 15.78s/it]
 76%|███████▌  | 7474/9864 [31:26:14<10:09:58, 15.31s/it]
 76%|███████▌  | 7475/9864 [31:26:28<9:56:26, 14.98s/it] 
 76%|███████▌  | 7476/9864 [31:26:43<9:52:39, 14.89s/it]
 76%|███████▌  | 7477/9864 [31:26:57<9:45:03, 14.71s/it]
 76%|███████▌  | 7478/9864 [31:27:13<10:01:00, 15.11s/it]
 76%|███████▌  | 7479/9864 [31:27:29<10:16:30, 15.51s/it]
 76%|███████▌  | 7480/9864 [31:27:45<10:17:40, 15.55s/it]
                                                         
{'loss': '0.07534', 'grad_norm': '5.607', 'learning_rate': '2.913e-06', 'epoch': '0.7584'}

 76%|███████▌  | 7480/9864 [31:27:45<10:17:40, 15.55s/it]
 76%|███████▌  | 7481/9864 [31:27:59<9:59:22, 15.09s/it] 
 76%|███████▌  | 7482/9864 [31:28:16<10:16:39, 15.53s/it]
 76%|███████▌  | 7483/9864 [31:28:30<10:01:26, 15.16s/it]
 76%|███████▌  | 7484/9864 [31:28:44<9:50:25, 14.88s/it] 
 76%|███████▌  | 7485/9864 [31:29:00<10:02:29, 15.20s/it]
 76%|███████▌  | 7486/9864 [31:29:15<9:54:40, 15.00s/it] 
 76%|███████▌  | 7487/9864 [31:29:31<10:14:14, 15.50s/it]
 76%|███████▌  | 7488/9864 [31:29:46<10:04:39, 15.27s/it]
 76%|███████▌  | 7489/9864 [31:30:02<10:16:27, 15.57s/it]
 76%|███████▌  | 7490/9864 [31:30:18<10:22:00, 15.72s/it]
                                                         
{'loss': '0.09604', 'grad_norm': '3.01', 'learning_rate': '2.89e-06', 'epoch': '0.7594'}

 76%|███████▌  | 7490/9864 [31:30:18<10:22:00, 15.72s/it]
 76%|███████▌  | 7491/9864 [31:30:33<10:03:50, 15.27s/it]
 76%|███████▌  | 7492/9864 [31:30:47<9:52:54, 15.00s/it] 
 76%|███████▌  | 7493/9864 [31:31:03<10:05:44, 15.33s/it]
 76%|███████▌  | 7494/9864 [31:31:17<9:52:48, 15.01s/it] 
 76%|███████▌  | 7495/9864 [31:31:33<10:03:02, 15.27s/it]
 76%|███████▌  | 7496/9864 [31:31:49<10:05:26, 15.34s/it]
 76%|███████▌  | 7497/9864 [31:32:03<9:51:41, 15.00s/it] 
 76%|███████▌  | 7498/9864 [31:32:19<10:07:31, 15.41s/it]
 76%|███████▌  | 7499/9864 [31:32:34<9:55:04, 15.10s/it] 
 76%|███████▌  | 7500/9864 [31:32:49<10:02:29, 15.29s/it]
                                                         
{'loss': '0.07545', 'grad_norm': '3.652', 'learning_rate': '2.867e-06', 'epoch': '0.7604'}

 76%|███████▌  | 7500/9864 [31:32:49<10:02:29, 15.29s/it]
 76%|███████▌  | 7501/9864 [31:33:04<9:50:24, 14.99s/it] 
 76%|███████▌  | 7502/9864 [31:33:18<9:44:26, 14.85s/it]
 76%|███████▌  | 7503/9864 [31:33:35<10:08:02, 15.45s/it]
 76%|███████▌  | 7504/9864 [31:33:49<9:51:34, 15.04s/it] 
 76%|███████▌  | 7505/9864 [31:34:05<9:59:40, 15.25s/it]
 76%|███████▌  | 7506/9864 [31:34:21<10:13:50, 15.62s/it]
 76%|███████▌  | 7507/9864 [31:34:36<9:58:36, 15.24s/it] 
 76%|███████▌  | 7508/9864 [31:34:51<10:02:42, 15.35s/it]
 76%|███████▌  | 7509/9864 [31:35:05<9:46:50, 14.95s/it] 
 76%|███████▌  | 7510/9864 [31:35:21<9:56:46, 15.21s/it]
                                                        
{'loss': '0.0785', 'grad_norm': '3.233', 'learning_rate': '2.844e-06', 'epoch': '0.7614'}

 76%|███████▌  | 7510/9864 [31:35:21<9:56:46, 15.21s/it]
 76%|███████▌  | 7511/9864 [31:35:36<9:51:43, 15.09s/it]
 76%|███████▌  | 7512/9864 [31:35:50<9:36:31, 14.71s/it]
 76%|███████▌  | 7513/9864 [31:36:05<9:48:10, 15.01s/it]
 76%|███████▌  | 7514/9864 [31:36:21<9:53:13, 15.15s/it]
 76%|███████▌  | 7515/9864 [31:36:35<9:38:41, 14.78s/it]
 76%|███████▌  | 7516/9864 [31:36:50<9:48:37, 15.04s/it]
 76%|███████▌  | 7517/9864 [31:37:05<9:37:38, 14.77s/it]
 76%|███████▌  | 7518/9864 [31:37:21<9:54:12, 15.20s/it]
 76%|███████▌  | 7519/9864 [31:37:35<9:44:50, 14.96s/it]
 76%|███████▌  | 7520/9864 [31:37:51<9:58:52, 15.33s/it]
                                                        
{'loss': '0.09104', 'grad_norm': '4.151', 'learning_rate': '2.821e-06', 'epoch': '0.7624'}

 76%|███████▌  | 7520/9864 [31:37:51<9:58:52, 15.33s/it]
 76%|███████▌  | 7521/9864 [31:38:07<10:07:25, 15.55s/it]
 76%|███████▋  | 7522/9864 [31:38:22<9:57:56, 15.32s/it] 
 76%|███████▋  | 7523/9864 [31:38:38<10:06:47, 15.55s/it]
 76%|███████▋  | 7524/9864 [31:38:53<9:56:17, 15.29s/it] 
 76%|███████▋  | 7525/9864 [31:39:07<9:44:41, 15.00s/it]
 76%|███████▋  | 7526/9864 [31:39:24<9:58:07, 15.35s/it]
 76%|███████▋  | 7527/9864 [31:39:38<9:42:08, 14.95s/it]
 76%|███████▋  | 7528/9864 [31:39:54<9:54:45, 15.28s/it]
 76%|███████▋  | 7529/9864 [31:40:10<10:04:02, 15.52s/it]
 76%|███████▋  | 7530/9864 [31:40:24<9:47:55, 15.11s/it] 
                                                        
{'loss': '0.08288', 'grad_norm': '4.493', 'learning_rate': '2.798e-06', 'epoch': '0.7634'}

 76%|███████▋  | 7530/9864 [31:40:24<9:47:55, 15.11s/it]
 76%|███████▋  | 7531/9864 [31:40:39<9:43:13, 15.00s/it]
 76%|███████▋  | 7532/9864 [31:40:55<10:02:42, 15.51s/it]
 76%|███████▋  | 7533/9864 [31:41:10<9:51:19, 15.22s/it] 
 76%|███████▋  | 7534/9864 [31:41:27<10:10:05, 15.71s/it]
 76%|███████▋  | 7535/9864 [31:41:42<10:08:38, 15.68s/it]
 76%|███████▋  | 7536/9864 [31:41:59<10:17:23, 15.91s/it]
 76%|███████▋  | 7537/9864 [31:42:13<9:55:45, 15.36s/it] 
 76%|███████▋  | 7538/9864 [31:42:27<9:39:29, 14.95s/it]
 76%|███████▋  | 7539/9864 [31:42:42<9:44:34, 15.09s/it]
 76%|███████▋  | 7540/9864 [31:42:56<9:33:04, 14.80s/it]
                                                        
{'loss': '0.09334', 'grad_norm': '2.979', 'learning_rate': '2.775e-06', 'epoch': '0.7645'}

 76%|███████▋  | 7540/9864 [31:42:56<9:33:04, 14.80s/it]
 76%|███████▋  | 7541/9864 [31:43:12<9:46:06, 15.14s/it]
 76%|███████▋  | 7542/9864 [31:43:26<9:33:26, 14.82s/it]
 76%|███████▋  | 7543/9864 [31:43:43<9:50:16, 15.26s/it]
 76%|███████▋  | 7544/9864 [31:43:59<10:00:08, 15.52s/it]
 76%|███████▋  | 7545/9864 [31:44:13<9:47:18, 15.20s/it] 
 77%|███████▋  | 7546/9864 [31:44:29<9:59:49, 15.53s/it]
 77%|███████▋  | 7547/9864 [31:44:44<9:46:46, 15.19s/it]
 77%|███████▋  | 7548/9864 [31:44:59<9:40:15, 15.03s/it]
 77%|███████▋  | 7549/9864 [31:45:15<9:56:13, 15.45s/it]
 77%|███████▋  | 7550/9864 [31:45:30<9:48:30, 15.26s/it]
                                                        
{'loss': '0.08318', 'grad_norm': '4.355', 'learning_rate': '2.752e-06', 'epoch': '0.7655'}

 77%|███████▋  | 7550/9864 [31:45:30<9:48:30, 15.26s/it]
 77%|███████▋  | 7551/9864 [31:45:46<10:00:46, 15.58s/it]
 77%|███████▋  | 7552/9864 [31:46:02<10:01:36, 15.61s/it]
 77%|███████▋  | 7553/9864 [31:46:18<10:03:52, 15.68s/it]
 77%|███████▋  | 7554/9864 [31:46:32<9:47:55, 15.27s/it] 
 77%|███████▋  | 7555/9864 [31:46:46<9:35:03, 14.94s/it]
 77%|███████▋  | 7556/9864 [31:47:02<9:44:25, 15.19s/it]
 77%|███████▋  | 7557/9864 [31:47:15<9:25:01, 14.70s/it]
 77%|███████▋  | 7558/9864 [31:47:30<9:20:39, 14.59s/it]
 77%|███████▋  | 7559/9864 [31:47:46<9:41:51, 15.15s/it]
 77%|███████▋  | 7560/9864 [31:48:02<9:53:54, 15.47s/it]
                                                        
{'loss': '0.08051', 'grad_norm': '3.926', 'learning_rate': '2.73e-06', 'epoch': '0.7665'}

 77%|███████▋  | 7560/9864 [31:48:02<9:53:54, 15.47s/it]
 77%|███████▋  | 7561/9864 [31:48:19<10:02:33, 15.70s/it]
 77%|███████▋  | 7562/9864 [31:48:33<9:44:41, 15.24s/it] 
 77%|███████▋  | 7563/9864 [31:48:47<9:29:45, 14.86s/it]
 77%|███████▋  | 7564/9864 [31:49:03<9:46:01, 15.29s/it]
 77%|███████▋  | 7565/9864 [31:49:17<9:35:21, 15.02s/it]
 77%|███████▋  | 7566/9864 [31:49:33<9:42:16, 15.20s/it]
 77%|███████▋  | 7567/9864 [31:49:48<9:35:35, 15.03s/it]
 77%|███████▋  | 7568/9864 [31:50:04<9:44:00, 15.26s/it]
 77%|███████▋  | 7569/9864 [31:50:18<9:32:08, 14.96s/it]
 77%|███████▋  | 7570/9864 [31:50:34<9:45:42, 15.32s/it]
                                                        
{'loss': '0.08063', 'grad_norm': '4.017', 'learning_rate': '2.707e-06', 'epoch': '0.7675'}

 77%|███████▋  | 7570/9864 [31:50:34<9:45:42, 15.32s/it]
 77%|███████▋  | 7571/9864 [31:50:50<9:53:21, 15.53s/it]
 77%|███████▋  | 7572/9864 [31:51:04<9:34:43, 15.05s/it]
 77%|███████▋  | 7573/9864 [31:51:19<9:30:36, 14.94s/it]
 77%|███████▋  | 7574/9864 [31:51:34<9:41:02, 15.22s/it]
 77%|███████▋  | 7575/9864 [31:51:49<9:32:00, 14.99s/it]
 77%|███████▋  | 7576/9864 [31:52:05<9:46:15, 15.37s/it]
 77%|███████▋  | 7577/9864 [31:52:22<9:57:27, 15.67s/it]
 77%|███████▋  | 7578/9864 [31:52:36<9:42:05, 15.28s/it]
 77%|███████▋  | 7579/9864 [31:52:50<9:27:54, 14.91s/it]
 77%|███████▋  | 7580/9864 [31:53:06<9:35:46, 15.13s/it]
                                                        
{'loss': '0.08123', 'grad_norm': '4.33', 'learning_rate': '2.685e-06', 'epoch': '0.7685'}

 77%|███████▋  | 7580/9864 [31:53:06<9:35:46, 15.13s/it]
 77%|███████▋  | 7581/9864 [31:53:20<9:29:00, 14.95s/it]
 77%|███████▋  | 7582/9864 [31:53:34<9:20:11, 14.73s/it]
 77%|███████▋  | 7583/9864 [31:53:50<9:35:05, 15.13s/it]
 77%|███████▋  | 7584/9864 [31:54:06<9:44:06, 15.37s/it]
 77%|███████▋  | 7585/9864 [31:54:23<9:55:21, 15.67s/it]
 77%|███████▋  | 7586/9864 [31:54:37<9:36:30, 15.18s/it]
 77%|███████▋  | 7587/9864 [31:54:51<9:19:46, 14.75s/it]
 77%|███████▋  | 7588/9864 [31:55:05<9:14:39, 14.62s/it]
 77%|███████▋  | 7589/9864 [31:55:21<9:32:30, 15.10s/it]
 77%|███████▋  | 7590/9864 [31:55:37<9:39:11, 15.28s/it]
                                                        
{'loss': '0.08177', 'grad_norm': '2.741', 'learning_rate': '2.663e-06', 'epoch': '0.7695'}

 77%|███████▋  | 7590/9864 [31:55:37<9:39:11, 15.28s/it]
 77%|███████▋  | 7591/9864 [31:55:51<9:25:03, 14.92s/it]
 77%|███████▋  | 7592/9864 [31:56:07<9:40:20, 15.33s/it]
 77%|███████▋  | 7593/9864 [31:56:23<9:46:49, 15.50s/it]
 77%|███████▋  | 7594/9864 [31:56:38<9:36:47, 15.25s/it]
 77%|███████▋  | 7595/9864 [31:56:52<9:27:59, 15.02s/it]
 77%|███████▋  | 7596/9864 [31:57:08<9:37:46, 15.29s/it]
 77%|███████▋  | 7597/9864 [31:57:23<9:31:27, 15.12s/it]
 77%|███████▋  | 7598/9864 [31:57:39<9:48:08, 15.57s/it]
 77%|███████▋  | 7599/9864 [31:57:54<9:34:52, 15.23s/it]
 77%|███████▋  | 7600/9864 [31:58:11<10:01:36, 15.94s/it]
                                                         
{'loss': '0.08643', 'grad_norm': '3.343', 'learning_rate': '2.64e-06', 'epoch': '0.7705'}

 77%|███████▋  | 7600/9864 [31:58:11<10:01:36, 15.94s/it]
 77%|███████▋  | 7601/9864 [31:58:26<9:42:14, 15.44s/it] 
 77%|███████▋  | 7602/9864 [31:58:39<9:20:20, 14.86s/it]
 77%|███████▋  | 7603/9864 [31:58:55<9:33:39, 15.22s/it]
 77%|███████▋  | 7604/9864 [31:59:09<9:20:00, 14.87s/it]
 77%|███████▋  | 7605/9864 [31:59:24<9:11:56, 14.66s/it]
 77%|███████▋  | 7606/9864 [31:59:39<9:23:01, 14.96s/it]
 77%|███████▋  | 7607/9864 [31:59:55<9:33:58, 15.26s/it]
 77%|███████▋  | 7608/9864 [32:00:11<9:43:33, 15.52s/it]
 77%|███████▋  | 7609/9864 [32:00:26<9:31:36, 15.21s/it]
 77%|███████▋  | 7610/9864 [32:00:40<9:22:29, 14.97s/it]
                                                        
{'loss': '0.07937', 'grad_norm': '3.319', 'learning_rate': '2.618e-06', 'epoch': '0.7716'}

 77%|███████▋  | 7610/9864 [32:00:40<9:22:29, 14.97s/it]
 77%|███████▋  | 7611/9864 [32:00:56<9:27:19, 15.11s/it]
 77%|███████▋  | 7612/9864 [32:01:10<9:17:07, 14.84s/it]
 77%|███████▋  | 7613/9864 [32:01:24<9:04:23, 14.51s/it]
 77%|███████▋  | 7614/9864 [32:01:39<9:15:02, 14.80s/it]
 77%|███████▋  | 7615/9864 [32:01:57<9:51:43, 15.79s/it]
 77%|███████▋  | 7616/9864 [32:02:11<9:34:45, 15.34s/it]
 77%|███████▋  | 7617/9864 [32:02:26<9:27:34, 15.16s/it]
 77%|███████▋  | 7618/9864 [32:02:40<9:10:54, 14.72s/it]
 77%|███████▋  | 7619/9864 [32:02:56<9:21:08, 15.00s/it]
 77%|███████▋  | 7620/9864 [32:03:10<9:11:52, 14.76s/it]
                                                        
{'loss': '0.08316', 'grad_norm': '4.385', 'learning_rate': '2.596e-06', 'epoch': '0.7726'}

 77%|███████▋  | 7620/9864 [32:03:10<9:11:52, 14.76s/it]
 77%|███████▋  | 7621/9864 [32:03:25<9:20:30, 14.99s/it]
 77%|███████▋  | 7622/9864 [32:03:42<9:36:12, 15.42s/it]
 77%|███████▋  | 7623/9864 [32:03:58<9:42:24, 15.59s/it]
 77%|███████▋  | 7624/9864 [32:04:12<9:32:57, 15.35s/it]
 77%|███████▋  | 7625/9864 [32:04:27<9:21:54, 15.06s/it]
 77%|███████▋  | 7626/9864 [32:04:41<9:07:47, 14.69s/it]
 77%|███████▋  | 7627/9864 [32:04:57<9:23:10, 15.11s/it]
 77%|███████▋  | 7628/9864 [32:05:11<9:16:59, 14.95s/it]
 77%|███████▋  | 7629/9864 [32:05:28<9:33:58, 15.41s/it]
 77%|███████▋  | 7630/9864 [32:05:44<9:45:18, 15.72s/it]
                                                        
{'loss': '0.08202', 'grad_norm': '3.906', 'learning_rate': '2.574e-06', 'epoch': '0.7736'}

 77%|███████▋  | 7630/9864 [32:05:44<9:45:18, 15.72s/it]
 77%|███████▋  | 7631/9864 [32:06:01<9:52:32, 15.92s/it]
 77%|███████▋  | 7632/9864 [32:06:15<9:30:12, 15.33s/it]
 77%|███████▋  | 7633/9864 [32:06:29<9:14:47, 14.92s/it]
 77%|███████▋  | 7634/9864 [32:06:44<9:22:21, 15.13s/it]
 77%|███████▋  | 7635/9864 [32:06:58<9:11:23, 14.84s/it]
 77%|███████▋  | 7636/9864 [32:07:12<9:02:51, 14.62s/it]
 77%|███████▋  | 7637/9864 [32:07:27<8:56:39, 14.46s/it]
 77%|███████▋  | 7638/9864 [32:07:43<9:21:52, 15.15s/it]
 77%|███████▋  | 7639/9864 [32:08:00<9:34:07, 15.48s/it]
 77%|███████▋  | 7640/9864 [32:08:13<9:14:26, 14.96s/it]
                                                        
{'loss': '0.08914', 'grad_norm': '3.744', 'learning_rate': '2.552e-06', 'epoch': '0.7746'}

 77%|███████▋  | 7640/9864 [32:08:13<9:14:26, 14.96s/it]
 77%|███████▋  | 7641/9864 [32:08:27<9:02:09, 14.63s/it]
 77%|███████▋  | 7642/9864 [32:08:43<9:17:55, 15.07s/it]
 77%|███████▋  | 7643/9864 [32:08:57<9:04:27, 14.71s/it]
 77%|███████▋  | 7644/9864 [32:09:12<9:03:01, 14.68s/it]
 78%|███████▊  | 7645/9864 [32:09:30<9:46:12, 15.85s/it]
 78%|███████▊  | 7646/9864 [32:09:47<9:51:09, 15.99s/it]
 78%|███████▊  | 7647/9864 [32:10:01<9:34:32, 15.55s/it]
 78%|███████▊  | 7648/9864 [32:10:16<9:22:13, 15.22s/it]
 78%|███████▊  | 7649/9864 [32:10:30<9:10:38, 14.92s/it]
 78%|███████▊  | 7650/9864 [32:10:46<9:22:45, 15.25s/it]
                                                        
{'loss': '0.09046', 'grad_norm': '4.822', 'learning_rate': '2.53e-06', 'epoch': '0.7756'}

 78%|███████▊  | 7650/9864 [32:10:46<9:22:45, 15.25s/it]
 78%|███████▊  | 7651/9864 [32:11:00<9:10:32, 14.93s/it]
 78%|███████▊  | 7652/9864 [32:11:14<9:01:45, 14.70s/it]
 78%|███████▊  | 7653/9864 [32:11:34<9:58:47, 16.25s/it]
 78%|███████▊  | 7654/9864 [32:11:49<9:39:37, 15.74s/it]
 78%|███████▊  | 7655/9864 [32:12:03<9:20:40, 15.23s/it]
 78%|███████▊  | 7656/9864 [32:12:17<9:08:14, 14.90s/it]
 78%|███████▊  | 7657/9864 [32:12:31<8:57:19, 14.61s/it]
 78%|███████▊  | 7658/9864 [32:12:47<9:11:17, 14.99s/it]
 78%|███████▊  | 7659/9864 [32:13:01<9:05:17, 14.84s/it]
 78%|███████▊  | 7660/9864 [32:13:16<9:02:19, 14.76s/it]
                                                        
{'loss': '0.08539', 'grad_norm': '3.837', 'learning_rate': '2.508e-06', 'epoch': '0.7766'}

 78%|███████▊  | 7660/9864 [32:13:16<9:02:19, 14.76s/it]
 78%|███████▊  | 7661/9864 [32:13:35<9:49:54, 16.07s/it]
 78%|███████▊  | 7662/9864 [32:13:49<9:30:28, 15.54s/it]
 78%|███████▊  | 7663/9864 [32:14:04<9:23:14, 15.35s/it]
 78%|███████▊  | 7664/9864 [32:14:18<9:10:17, 15.01s/it]
 78%|███████▊  | 7665/9864 [32:14:33<9:02:51, 14.81s/it]
 78%|███████▊  | 7666/9864 [32:14:48<9:14:05, 15.13s/it]
 78%|███████▊  | 7667/9864 [32:15:02<9:02:21, 14.81s/it]
 78%|███████▊  | 7668/9864 [32:15:17<9:00:43, 14.77s/it]
 78%|███████▊  | 7669/9864 [32:15:37<9:56:19, 16.30s/it]
 78%|███████▊  | 7670/9864 [32:15:51<9:35:51, 15.75s/it]
                                                        
{'loss': '0.07872', 'grad_norm': '4.179', 'learning_rate': '2.487e-06', 'epoch': '0.7776'}

 78%|███████▊  | 7670/9864 [32:15:51<9:35:51, 15.75s/it]
 78%|███████▊  | 7671/9864 [32:16:06<9:23:32, 15.42s/it]
 78%|███████▊  | 7672/9864 [32:16:20<9:10:03, 15.06s/it]
 78%|███████▊  | 7673/9864 [32:16:36<9:21:52, 15.39s/it]
 78%|███████▊  | 7674/9864 [32:16:51<9:10:36, 15.09s/it]
 78%|███████▊  | 7675/9864 [32:17:05<9:02:36, 14.87s/it]
 78%|███████▊  | 7676/9864 [32:17:22<9:17:54, 15.30s/it]
 78%|███████▊  | 7677/9864 [32:17:37<9:23:06, 15.45s/it]
 78%|███████▊  | 7678/9864 [32:17:54<9:31:56, 15.70s/it]
 78%|███████▊  | 7679/9864 [32:18:08<9:15:32, 15.25s/it]
 78%|███████▊  | 7680/9864 [32:18:21<8:56:00, 14.73s/it]
                                                        
{'loss': '0.06599', 'grad_norm': '3.975', 'learning_rate': '2.465e-06', 'epoch': '0.7787'}

 78%|███████▊  | 7680/9864 [32:18:21<8:56:00, 14.73s/it]
 78%|███████▊  | 7681/9864 [32:18:37<9:06:33, 15.02s/it]
 78%|███████▊  | 7682/9864 [32:18:52<9:02:04, 14.91s/it]
 78%|███████▊  | 7683/9864 [32:19:06<8:53:18, 14.67s/it]
 78%|███████▊  | 7684/9864 [32:19:22<9:12:08, 15.20s/it]
 78%|███████▊  | 7685/9864 [32:19:38<9:19:55, 15.42s/it]
 78%|███████▊  | 7686/9864 [32:19:54<9:27:39, 15.64s/it]
 78%|███████▊  | 7687/9864 [32:20:09<9:17:44, 15.37s/it]
 78%|███████▊  | 7688/9864 [32:20:23<9:05:33, 15.04s/it]
 78%|███████▊  | 7689/9864 [32:20:39<9:15:04, 15.31s/it]
 78%|███████▊  | 7690/9864 [32:20:53<9:00:37, 14.92s/it]
                                                        
{'loss': '0.07803', 'grad_norm': '4.072', 'learning_rate': '2.443e-06', 'epoch': '0.7797'}

 78%|███████▊  | 7690/9864 [32:20:53<9:00:37, 14.92s/it]
 78%|███████▊  | 7691/9864 [32:21:07<8:50:18, 14.64s/it]
 78%|███████▊  | 7692/9864 [32:21:24<9:07:44, 15.13s/it]
 78%|███████▊  | 7693/9864 [32:21:40<9:22:49, 15.56s/it]
 78%|███████▊  | 7694/9864 [32:21:56<9:24:28, 15.61s/it]
 78%|███████▊  | 7695/9864 [32:22:11<9:14:22, 15.34s/it]
 78%|███████▊  | 7696/9864 [32:22:25<9:03:45, 15.05s/it]
 78%|███████▊  | 7697/9864 [32:22:39<8:56:05, 14.84s/it]
 78%|███████▊  | 7698/9864 [32:22:55<9:08:45, 15.20s/it]
 78%|███████▊  | 7699/9864 [32:23:09<8:57:50, 14.91s/it]
 78%|███████▊  | 7700/9864 [32:23:26<9:11:14, 15.28s/it]
                                                        
{'loss': '0.08231', 'grad_norm': '3.286', 'learning_rate': '2.422e-06', 'epoch': '0.7807'}

 78%|███████▊  | 7700/9864 [32:23:26<9:11:14, 15.28s/it]
 78%|███████▊  | 7701/9864 [32:23:42<9:22:37, 15.61s/it]
 78%|███████▊  | 7702/9864 [32:23:57<9:15:46, 15.42s/it]
 78%|███████▊  | 7703/9864 [32:24:14<9:27:35, 15.76s/it]
 78%|███████▊  | 7704/9864 [32:24:28<9:15:24, 15.43s/it]
 78%|███████▊  | 7705/9864 [32:24:43<9:03:52, 15.11s/it]
 78%|███████▊  | 7706/9864 [32:24:59<9:13:50, 15.40s/it]
 78%|███████▊  | 7707/9864 [32:25:13<9:03:15, 15.11s/it]
 78%|███████▊  | 7708/9864 [32:25:29<9:13:13, 15.40s/it]
 78%|███████▊  | 7709/9864 [32:25:45<9:21:32, 15.63s/it]
 78%|███████▊  | 7710/9864 [32:26:00<9:11:42, 15.37s/it]
                                                        
{'loss': '0.07258', 'grad_norm': '4.081', 'learning_rate': '2.401e-06', 'epoch': '0.7817'}

 78%|███████▊  | 7710/9864 [32:26:00<9:11:42, 15.37s/it]
 78%|███████▊  | 7711/9864 [32:26:16<9:17:05, 15.52s/it]
 78%|███████▊  | 7712/9864 [32:26:30<8:59:31, 15.04s/it]
 78%|███████▊  | 7713/9864 [32:26:44<8:51:32, 14.83s/it]
 78%|███████▊  | 7714/9864 [32:26:58<8:38:47, 14.48s/it]
 78%|███████▊  | 7715/9864 [32:27:14<8:55:13, 14.94s/it]
 78%|███████▊  | 7716/9864 [32:27:30<9:04:14, 15.20s/it]
 78%|███████▊  | 7717/9864 [32:27:44<8:58:41, 15.05s/it]
 78%|███████▊  | 7718/9864 [32:27:59<8:53:39, 14.92s/it]
 78%|███████▊  | 7719/9864 [32:28:15<9:04:45, 15.24s/it]
 78%|███████▊  | 7720/9864 [32:28:29<8:48:57, 14.80s/it]
                                                        
{'loss': '0.07584', 'grad_norm': '4.816', 'learning_rate': '2.379e-06', 'epoch': '0.7827'}

 78%|███████▊  | 7720/9864 [32:28:29<8:48:57, 14.80s/it]
 78%|███████▊  | 7721/9864 [32:28:43<8:46:17, 14.74s/it]
 78%|███████▊  | 7722/9864 [32:28:58<8:41:24, 14.61s/it]
 78%|███████▊  | 7723/9864 [32:29:15<9:13:24, 15.51s/it]
 78%|███████▊  | 7724/9864 [32:29:32<9:23:03, 15.79s/it]
 78%|███████▊  | 7725/9864 [32:29:46<9:08:10, 15.38s/it]
 78%|███████▊  | 7726/9864 [32:30:02<9:13:17, 15.53s/it]
 78%|███████▊  | 7727/9864 [32:30:16<9:01:25, 15.20s/it]
 78%|███████▊  | 7728/9864 [32:30:30<8:45:28, 14.76s/it]
 78%|███████▊  | 7729/9864 [32:30:45<8:42:00, 14.67s/it]
 78%|███████▊  | 7730/9864 [32:30:59<8:38:06, 14.57s/it]
                                                        
{'loss': '0.06389', 'grad_norm': '4.902', 'learning_rate': '2.358e-06', 'epoch': '0.7837'}

 78%|███████▊  | 7730/9864 [32:30:59<8:38:06, 14.57s/it]
 78%|███████▊  | 7731/9864 [32:31:17<9:09:19, 15.45s/it]
 78%|███████▊  | 7732/9864 [32:31:33<9:18:46, 15.73s/it]
 78%|███████▊  | 7733/9864 [32:31:47<9:05:22, 15.36s/it]
 78%|███████▊  | 7734/9864 [32:32:04<9:19:31, 15.76s/it]
 78%|███████▊  | 7735/9864 [32:32:18<9:02:40, 15.29s/it]
 78%|███████▊  | 7736/9864 [32:32:33<8:56:27, 15.13s/it]
 78%|███████▊  | 7737/9864 [32:32:47<8:48:59, 14.92s/it]
 78%|███████▊  | 7738/9864 [32:33:03<8:57:24, 15.17s/it]
 78%|███████▊  | 7739/9864 [32:33:19<9:05:51, 15.41s/it]
 78%|███████▊  | 7740/9864 [32:33:35<9:11:02, 15.57s/it]
                                                        
{'loss': '0.07262', 'grad_norm': '4.301', 'learning_rate': '2.337e-06', 'epoch': '0.7847'}

 78%|███████▊  | 7740/9864 [32:33:35<9:11:02, 15.57s/it]
 78%|███████▊  | 7741/9864 [32:33:50<9:04:01, 15.37s/it]
 78%|███████▊  | 7742/9864 [32:34:06<9:10:52, 15.58s/it]
 78%|███████▊  | 7743/9864 [32:34:21<8:58:57, 15.25s/it]
 79%|███████▊  | 7744/9864 [32:34:35<8:54:55, 15.14s/it]
 79%|███████▊  | 7745/9864 [32:34:50<8:46:02, 14.90s/it]
 79%|███████▊  | 7746/9864 [32:35:06<9:02:26, 15.37s/it]
 79%|███████▊  | 7747/9864 [32:35:22<9:10:07, 15.59s/it]
 79%|███████▊  | 7748/9864 [32:35:39<9:16:56, 15.79s/it]
 79%|███████▊  | 7749/9864 [32:35:53<9:01:20, 15.36s/it]
 79%|███████▊  | 7750/9864 [32:36:09<9:08:39, 15.57s/it]
                                                        
{'loss': '0.07638', 'grad_norm': '4.864', 'learning_rate': '2.316e-06', 'epoch': '0.7858'}

 79%|███████▊  | 7750/9864 [32:36:09<9:08:39, 15.57s/it]
 79%|███████▊  | 7751/9864 [32:36:23<8:52:35, 15.12s/it]
 79%|███████▊  | 7752/9864 [32:36:37<8:38:18, 14.72s/it]
 79%|███████▊  | 7753/9864 [32:36:53<8:50:06, 15.07s/it]
 79%|███████▊  | 7754/9864 [32:37:09<8:59:18, 15.34s/it]
 79%|███████▊  | 7755/9864 [32:37:25<9:07:55, 15.59s/it]
 79%|███████▊  | 7756/9864 [32:37:39<8:52:09, 15.15s/it]
 79%|███████▊  | 7757/9864 [32:37:54<8:49:35, 15.08s/it]
 79%|███████▊  | 7758/9864 [32:38:10<8:57:27, 15.31s/it]
 79%|███████▊  | 7759/9864 [32:38:25<8:50:47, 15.13s/it]
 79%|███████▊  | 7760/9864 [32:38:38<8:36:41, 14.73s/it]
                                                        
{'loss': '0.07641', 'grad_norm': '4.626', 'learning_rate': '2.295e-06', 'epoch': '0.7868'}

 79%|███████▊  | 7760/9864 [32:38:38<8:36:41, 14.73s/it]
 79%|███████▊  | 7761/9864 [32:38:55<8:53:08, 15.21s/it]
 79%|███████▊  | 7762/9864 [32:39:10<8:58:44, 15.38s/it]
 79%|███████▊  | 7763/9864 [32:39:27<9:10:29, 15.72s/it]
 79%|███████▊  | 7764/9864 [32:39:41<8:54:11, 15.26s/it]
 79%|███████▊  | 7765/9864 [32:39:56<8:46:32, 15.05s/it]
 79%|███████▊  | 7766/9864 [32:40:12<8:54:59, 15.30s/it]
 79%|███████▊  | 7767/9864 [32:40:26<8:45:15, 15.03s/it]
 79%|███████▉  | 7768/9864 [32:40:40<8:38:52, 14.85s/it]
 79%|███████▉  | 7769/9864 [32:40:57<8:55:34, 15.34s/it]
 79%|███████▉  | 7770/9864 [32:41:13<9:00:11, 15.48s/it]
                                                        
{'loss': '0.0687', 'grad_norm': '5.838', 'learning_rate': '2.274e-06', 'epoch': '0.7878'}

 79%|███████▉  | 7770/9864 [32:41:13<9:00:11, 15.48s/it]
 79%|███████▉  | 7771/9864 [32:41:29<9:12:42, 15.84s/it]
 79%|███████▉  | 7772/9864 [32:41:44<8:58:20, 15.44s/it]
 79%|███████▉  | 7773/9864 [32:41:58<8:47:46, 15.14s/it]
 79%|███████▉  | 7774/9864 [32:42:14<8:56:00, 15.39s/it]
 79%|███████▉  | 7775/9864 [32:42:29<8:49:07, 15.20s/it]
 79%|███████▉  | 7776/9864 [32:42:44<8:45:58, 15.11s/it]
 79%|███████▉  | 7777/9864 [32:43:00<8:58:48, 15.49s/it]
 79%|███████▉  | 7778/9864 [32:43:18<9:23:53, 16.22s/it]
 79%|███████▉  | 7779/9864 [32:43:32<9:00:02, 15.54s/it]
 79%|███████▉  | 7780/9864 [32:43:46<8:45:44, 15.14s/it]
                                                        
{'loss': '0.08299', 'grad_norm': '4.256', 'learning_rate': '2.253e-06', 'epoch': '0.7888'}

 79%|███████▉  | 7780/9864 [32:43:46<8:45:44, 15.14s/it]
 79%|███████▉  | 7781/9864 [32:44:01<8:37:41, 14.91s/it]
 79%|███████▉  | 7782/9864 [32:44:17<8:52:11, 15.34s/it]
 79%|███████▉  | 7783/9864 [32:44:32<8:44:46, 15.13s/it]
 79%|███████▉  | 7784/9864 [32:44:47<8:50:11, 15.29s/it]
 79%|███████▉  | 7785/9864 [32:45:02<8:41:34, 15.05s/it]
 79%|███████▉  | 7786/9864 [32:45:18<8:51:06, 15.34s/it]
 79%|███████▉  | 7787/9864 [32:45:34<8:59:15, 15.58s/it]
 79%|███████▉  | 7788/9864 [32:45:48<8:40:30, 15.04s/it]
 79%|███████▉  | 7789/9864 [32:46:02<8:30:20, 14.76s/it]
 79%|███████▉  | 7790/9864 [32:46:16<8:23:59, 14.58s/it]
                                                        
{'loss': '0.07475', 'grad_norm': '3.262', 'learning_rate': '2.233e-06', 'epoch': '0.7898'}

 79%|███████▉  | 7790/9864 [32:46:16<8:23:59, 14.58s/it]
 79%|███████▉  | 7791/9864 [32:46:34<8:57:40, 15.56s/it]
 79%|███████▉  | 7792/9864 [32:46:47<8:35:42, 14.93s/it]
 79%|███████▉  | 7793/9864 [32:47:04<8:48:53, 15.32s/it]
 79%|███████▉  | 7794/9864 [32:47:18<8:39:08, 15.05s/it]
 79%|███████▉  | 7795/9864 [32:47:34<8:48:45, 15.33s/it]
 79%|███████▉  | 7796/9864 [32:47:48<8:34:30, 14.93s/it]
 79%|███████▉  | 7797/9864 [32:48:02<8:23:02, 14.60s/it]
 79%|███████▉  | 7798/9864 [32:48:18<8:38:23, 15.05s/it]
 79%|███████▉  | 7799/9864 [32:48:34<8:49:57, 15.40s/it]
 79%|███████▉  | 7800/9864 [32:48:48<8:36:30, 15.01s/it]
                                                        
{'loss': '0.07325', 'grad_norm': '5.235', 'learning_rate': '2.212e-06', 'epoch': '0.7908'}

 79%|███████▉  | 7800/9864 [32:48:48<8:36:30, 15.01s/it]
 79%|███████▉  | 7801/9864 [32:49:04<8:43:39, 15.23s/it]
 79%|███████▉  | 7802/9864 [32:49:18<8:33:38, 14.95s/it]
 79%|███████▉  | 7803/9864 [32:49:34<8:40:58, 15.17s/it]
 79%|███████▉  | 7804/9864 [32:49:49<8:35:47, 15.02s/it]
 79%|███████▉  | 7805/9864 [32:50:03<8:27:07, 14.78s/it]
 79%|███████▉  | 7806/9864 [32:50:17<8:23:54, 14.69s/it]
 79%|███████▉  | 7807/9864 [32:50:34<8:38:21, 15.12s/it]
 79%|███████▉  | 7808/9864 [32:50:51<9:06:21, 15.94s/it]
 79%|███████▉  | 7809/9864 [32:51:06<8:48:02, 15.42s/it]
 79%|███████▉  | 7810/9864 [32:51:20<8:38:35, 15.15s/it]
                                                        
{'loss': '0.1027', 'grad_norm': '4.662', 'learning_rate': '2.191e-06', 'epoch': '0.7918'}

 79%|███████▉  | 7810/9864 [32:51:20<8:38:35, 15.15s/it]
 79%|███████▉  | 7811/9864 [32:51:36<8:44:00, 15.31s/it]
 79%|███████▉  | 7812/9864 [32:51:50<8:32:47, 14.99s/it]
 79%|███████▉  | 7813/9864 [32:52:04<8:24:17, 14.75s/it]
 79%|███████▉  | 7814/9864 [32:52:18<8:18:03, 14.58s/it]
 79%|███████▉  | 7815/9864 [32:52:36<8:51:47, 15.57s/it]
 79%|███████▉  | 7816/9864 [32:52:53<9:01:54, 15.88s/it]
 79%|███████▉  | 7817/9864 [32:53:08<8:50:09, 15.54s/it]
 79%|███████▉  | 7818/9864 [32:53:24<8:55:29, 15.70s/it]
 79%|███████▉  | 7819/9864 [32:53:38<8:44:04, 15.38s/it]
 79%|███████▉  | 7820/9864 [32:53:52<8:30:42, 14.99s/it]
                                                        
{'loss': '0.08586', 'grad_norm': '3.624', 'learning_rate': '2.171e-06', 'epoch': '0.7928'}

 79%|███████▉  | 7820/9864 [32:53:52<8:30:42, 14.99s/it]
 79%|███████▉  | 7821/9864 [32:54:06<8:17:15, 14.60s/it]
 79%|███████▉  | 7822/9864 [32:54:20<8:14:03, 14.52s/it]
 79%|███████▉  | 7823/9864 [32:54:38<8:45:52, 15.46s/it]
 79%|███████▉  | 7824/9864 [32:54:54<8:50:24, 15.60s/it]
 79%|███████▉  | 7825/9864 [32:55:08<8:35:52, 15.18s/it]
 79%|███████▉  | 7826/9864 [32:55:24<8:41:29, 15.35s/it]
 79%|███████▉  | 7827/9864 [32:55:38<8:31:53, 15.08s/it]
 79%|███████▉  | 7828/9864 [32:55:52<8:21:15, 14.77s/it]
 79%|███████▉  | 7829/9864 [32:56:07<8:18:44, 14.70s/it]
 79%|███████▉  | 7830/9864 [32:56:21<8:13:37, 14.56s/it]
                                                        
{'loss': '0.07612', 'grad_norm': '4.003', 'learning_rate': '2.15e-06', 'epoch': '0.7939'}

 79%|███████▉  | 7830/9864 [32:56:21<8:13:37, 14.56s/it]
 79%|███████▉  | 7831/9864 [32:56:39<8:47:21, 15.56s/it]
 79%|███████▉  | 7832/9864 [32:56:55<8:53:13, 15.74s/it]
 79%|███████▉  | 7833/9864 [32:57:10<8:38:34, 15.32s/it]
 79%|███████▉  | 7834/9864 [32:57:24<8:27:51, 15.01s/it]
 79%|███████▉  | 7835/9864 [32:57:40<8:35:08, 15.23s/it]
 79%|███████▉  | 7836/9864 [32:57:54<8:21:10, 14.83s/it]
 79%|███████▉  | 7837/9864 [32:58:08<8:16:40, 14.70s/it]
 79%|███████▉  | 7838/9864 [32:58:24<8:25:53, 14.98s/it]
 79%|███████▉  | 7839/9864 [32:58:39<8:33:44, 15.22s/it]
 79%|███████▉  | 7840/9864 [32:58:54<8:26:29, 15.01s/it]
                                                        
{'loss': '0.06859', 'grad_norm': '3.246', 'learning_rate': '2.13e-06', 'epoch': '0.7949'}

 79%|███████▉  | 7840/9864 [32:58:54<8:26:29, 15.01s/it]
 79%|███████▉  | 7841/9864 [32:59:10<8:31:47, 15.18s/it]
 80%|███████▉  | 7842/9864 [32:59:24<8:24:07, 14.96s/it]
 80%|███████▉  | 7843/9864 [32:59:38<8:19:00, 14.81s/it]
 80%|███████▉  | 7844/9864 [32:59:54<8:28:44, 15.11s/it]
 80%|███████▉  | 7845/9864 [33:00:08<8:17:27, 14.78s/it]
 80%|███████▉  | 7846/9864 [33:00:24<8:30:04, 15.17s/it]
 80%|███████▉  | 7847/9864 [33:00:41<8:42:21, 15.54s/it]
 80%|███████▉  | 7848/9864 [33:00:56<8:35:40, 15.35s/it]
 80%|███████▉  | 7849/9864 [33:01:12<8:45:29, 15.65s/it]
 80%|███████▉  | 7850/9864 [33:01:26<8:29:35, 15.18s/it]
                                                        
{'loss': '0.06127', 'grad_norm': '3.596', 'learning_rate': '2.11e-06', 'epoch': '0.7959'}

 80%|███████▉  | 7850/9864 [33:01:26<8:29:35, 15.18s/it]
 80%|███████▉  | 7851/9864 [33:01:40<8:20:43, 14.92s/it]
 80%|███████▉  | 7852/9864 [33:01:56<8:28:37, 15.17s/it]
 80%|███████▉  | 7853/9864 [33:02:12<8:35:51, 15.39s/it]
 80%|███████▉  | 7854/9864 [33:02:28<8:41:20, 15.56s/it]
 80%|███████▉  | 7855/9864 [33:02:42<8:27:04, 15.14s/it]
 80%|███████▉  | 7856/9864 [33:02:57<8:23:41, 15.05s/it]
 80%|███████▉  | 7857/9864 [33:03:13<8:34:34, 15.38s/it]
 80%|███████▉  | 7858/9864 [33:03:28<8:26:16, 15.14s/it]
 80%|███████▉  | 7859/9864 [33:03:42<8:15:46, 14.84s/it]
 80%|███████▉  | 7860/9864 [33:03:58<8:30:30, 15.28s/it]
                                                        
{'loss': '0.06579', 'grad_norm': '4.264', 'learning_rate': '2.09e-06', 'epoch': '0.7969'}

 80%|███████▉  | 7860/9864 [33:03:58<8:30:30, 15.28s/it]
 80%|███████▉  | 7861/9864 [33:04:13<8:29:06, 15.25s/it]
 80%|███████▉  | 7862/9864 [33:04:29<8:32:42, 15.37s/it]
 80%|███████▉  | 7863/9864 [33:04:43<8:19:36, 14.98s/it]
 80%|███████▉  | 7864/9864 [33:04:57<8:10:29, 14.71s/it]
 80%|███████▉  | 7865/9864 [33:05:13<8:21:22, 15.05s/it]
 80%|███████▉  | 7866/9864 [33:05:27<8:13:32, 14.82s/it]
 80%|███████▉  | 7867/9864 [33:05:43<8:24:05, 15.15s/it]
 80%|███████▉  | 7868/9864 [33:05:58<8:18:07, 14.97s/it]
 80%|███████▉  | 7869/9864 [33:06:13<8:22:49, 15.12s/it]
 80%|███████▉  | 7870/9864 [33:06:29<8:29:43, 15.34s/it]
                                                        
{'loss': '0.07336', 'grad_norm': '3.522', 'learning_rate': '2.07e-06', 'epoch': '0.7979'}

 80%|███████▉  | 7870/9864 [33:06:29<8:29:43, 15.34s/it]
 80%|███████▉  | 7871/9864 [33:06:43<8:19:56, 15.05s/it]
 80%|███████▉  | 7872/9864 [33:06:57<8:06:13, 14.65s/it]
 80%|███████▉  | 7873/9864 [33:07:11<7:58:26, 14.42s/it]
 80%|███████▉  | 7874/9864 [33:07:27<8:17:45, 15.01s/it]
 80%|███████▉  | 7875/9864 [33:07:42<8:09:37, 14.77s/it]
 80%|███████▉  | 7876/9864 [33:07:58<8:21:12, 15.13s/it]
 80%|███████▉  | 7877/9864 [33:08:13<8:26:46, 15.30s/it]
 80%|███████▉  | 7878/9864 [33:08:29<8:34:21, 15.54s/it]
 80%|███████▉  | 7879/9864 [33:08:43<8:16:26, 15.01s/it]
 80%|███████▉  | 7880/9864 [33:08:57<8:08:15, 14.77s/it]
                                                        
{'loss': '0.07477', 'grad_norm': '6.662', 'learning_rate': '2.05e-06', 'epoch': '0.7989'}

 80%|███████▉  | 7880/9864 [33:08:57<8:08:15, 14.77s/it]
 80%|███████▉  | 7881/9864 [33:09:12<8:09:39, 14.82s/it]
 80%|███████▉  | 7882/9864 [33:09:28<8:22:41, 15.22s/it]
 80%|███████▉  | 7883/9864 [33:09:44<8:27:55, 15.38s/it]
 80%|███████▉  | 7884/9864 [33:09:59<8:21:00, 15.18s/it]
 80%|███████▉  | 7885/9864 [33:10:16<8:41:49, 15.82s/it]
 80%|███████▉  | 7886/9864 [33:10:31<8:26:03, 15.35s/it]
 80%|███████▉  | 7887/9864 [33:10:45<8:17:52, 15.11s/it]
 80%|███████▉  | 7888/9864 [33:11:00<8:12:11, 14.94s/it]
 80%|███████▉  | 7889/9864 [33:11:14<8:07:25, 14.81s/it]
 80%|███████▉  | 7890/9864 [33:11:30<8:19:42, 15.19s/it]
                                                        
{'loss': '0.08412', 'grad_norm': '4.564', 'learning_rate': '2.03e-06', 'epoch': '0.7999'}

 80%|███████▉  | 7890/9864 [33:11:30<8:19:42, 15.19s/it]
 80%|███████▉  | 7891/9864 [33:11:47<8:31:45, 15.56s/it]
 80%|████████  | 7892/9864 [33:12:02<8:32:51, 15.60s/it]
 80%|████████  | 7893/9864 [33:12:19<8:43:38, 15.94s/it]
 80%|████████  | 7894/9864 [33:12:33<8:27:20, 15.45s/it]
 80%|████████  | 7895/9864 [33:12:47<8:11:06, 14.97s/it]
 80%|████████  | 7896/9864 [33:13:01<8:02:41, 14.72s/it]
 80%|████████  | 7897/9864 [33:13:17<8:13:52, 15.06s/it]
 80%|████████  | 7898/9864 [33:13:31<8:04:27, 14.79s/it]
 80%|████████  | 7899/9864 [33:13:48<8:19:00, 15.24s/it]
 80%|████████  | 7900/9864 [33:14:03<8:23:21, 15.38s/it]
                                                        
{'loss': '0.08011', 'grad_norm': '4.797', 'learning_rate': '2.01e-06', 'epoch': '0.801'}

 80%|████████  | 7900/9864 [33:14:03<8:23:21, 15.38s/it]
 80%|████████  | 7901/9864 [33:14:20<8:35:54, 15.77s/it]
 80%|████████  | 7902/9864 [33:14:34<8:19:01, 15.26s/it]
 80%|████████  | 7903/9864 [33:14:49<8:11:37, 15.04s/it]
 80%|████████  | 7904/9864 [33:15:03<8:03:50, 14.81s/it]
 80%|████████  | 7905/9864 [33:15:19<8:17:30, 15.24s/it]
 80%|████████  | 7906/9864 [33:15:35<8:25:58, 15.50s/it]
 80%|████████  | 7907/9864 [33:15:51<8:31:25, 15.68s/it]
 80%|████████  | 7908/9864 [33:16:07<8:30:25, 15.66s/it]
 80%|████████  | 7909/9864 [33:16:21<8:17:22, 15.26s/it]
 80%|████████  | 7910/9864 [33:16:36<8:09:51, 15.04s/it]
                                                        
{'loss': '0.07487', 'grad_norm': '5.494', 'learning_rate': '1.99e-06', 'epoch': '0.802'}

 80%|████████  | 7910/9864 [33:16:36<8:09:51, 15.04s/it]
 80%|████████  | 7911/9864 [33:16:50<8:03:23, 14.85s/it]
 80%|████████  | 7912/9864 [33:17:05<7:58:14, 14.70s/it]
 80%|████████  | 7913/9864 [33:17:21<8:10:55, 15.10s/it]
 80%|████████  | 7914/9864 [33:17:35<8:04:21, 14.90s/it]
 80%|████████  | 7915/9864 [33:17:51<8:12:23, 15.16s/it]
 80%|████████  | 7916/9864 [33:18:07<8:18:57, 15.37s/it]
 80%|████████  | 7917/9864 [33:18:21<8:08:08, 15.04s/it]
 80%|████████  | 7918/9864 [33:18:36<8:03:22, 14.90s/it]
 80%|████████  | 7919/9864 [33:18:50<7:58:22, 14.76s/it]
 80%|████████  | 7920/9864 [33:19:06<8:09:19, 15.10s/it]
                                                        
{'loss': '0.07077', 'grad_norm': '3.739', 'learning_rate': '1.971e-06', 'epoch': '0.803'}

 80%|████████  | 7920/9864 [33:19:06<8:09:19, 15.10s/it]
 80%|████████  | 7921/9864 [33:19:22<8:19:48, 15.43s/it]
 80%|████████  | 7922/9864 [33:19:37<8:13:26, 15.25s/it]
 80%|████████  | 7923/9864 [33:19:52<8:14:19, 15.28s/it]
 80%|████████  | 7924/9864 [33:20:09<8:24:45, 15.61s/it]
 80%|████████  | 7925/9864 [33:20:23<8:16:11, 15.35s/it]
 80%|████████  | 7926/9864 [33:20:38<8:05:33, 15.03s/it]
 80%|████████  | 7927/9864 [33:20:52<7:54:56, 14.71s/it]
 80%|████████  | 7928/9864 [33:21:07<7:57:36, 14.80s/it]
 80%|████████  | 7929/9864 [33:21:24<8:26:30, 15.71s/it]
 80%|████████  | 7930/9864 [33:21:40<8:29:08, 15.80s/it]
                                                        
{'loss': '0.06805', 'grad_norm': '3.871', 'learning_rate': '1.951e-06', 'epoch': '0.804'}

 80%|████████  | 7930/9864 [33:21:40<8:29:08, 15.80s/it]
 80%|████████  | 7931/9864 [33:21:55<8:19:49, 15.51s/it]
 80%|████████  | 7932/9864 [33:22:11<8:20:25, 15.54s/it]
 80%|████████  | 7933/9864 [33:22:25<8:08:10, 15.17s/it]
 80%|████████  | 7934/9864 [33:22:39<7:54:57, 14.77s/it]
 80%|████████  | 7935/9864 [33:22:53<7:50:38, 14.64s/it]
 80%|████████  | 7936/9864 [33:23:08<7:48:28, 14.58s/it]
 80%|████████  | 7937/9864 [33:23:23<7:58:12, 14.89s/it]
 80%|████████  | 7938/9864 [33:23:41<8:26:34, 15.78s/it]
 80%|████████  | 7939/9864 [33:23:57<8:28:32, 15.85s/it]
 80%|████████  | 7940/9864 [33:24:11<8:10:26, 15.29s/it]
                                                        
{'loss': '0.06446', 'grad_norm': '4.345', 'learning_rate': '1.932e-06', 'epoch': '0.805'}

 80%|████████  | 7940/9864 [33:24:11<8:10:26, 15.29s/it]
 81%|████████  | 7941/9864 [33:24:26<8:04:00, 15.10s/it]
 81%|████████  | 7942/9864 [33:24:41<7:59:10, 14.96s/it]
 81%|████████  | 7943/9864 [33:24:55<7:56:47, 14.89s/it]
 81%|████████  | 7944/9864 [33:25:10<7:51:43, 14.74s/it]
 81%|████████  | 7945/9864 [33:25:27<8:19:54, 15.63s/it]
 81%|████████  | 7946/9864 [33:25:43<8:22:43, 15.73s/it]
 81%|████████  | 7947/9864 [33:25:58<8:10:25, 15.35s/it]
 81%|████████  | 7948/9864 [33:26:14<8:16:25, 15.55s/it]
 81%|████████  | 7949/9864 [33:26:28<8:03:57, 15.16s/it]
 81%|████████  | 7950/9864 [33:26:42<7:52:00, 14.80s/it]
                                                        
{'loss': '0.08083', 'grad_norm': '5.749', 'learning_rate': '1.913e-06', 'epoch': '0.806'}

 81%|████████  | 7950/9864 [33:26:42<7:52:00, 14.80s/it]
 81%|████████  | 7951/9864 [33:26:56<7:48:34, 14.70s/it]
 81%|████████  | 7952/9864 [33:27:11<7:46:42, 14.65s/it]
 81%|████████  | 7953/9864 [33:27:30<8:31:08, 16.05s/it]
 81%|████████  | 7954/9864 [33:27:45<8:13:38, 15.51s/it]
 81%|████████  | 7955/9864 [33:27:59<8:04:53, 15.24s/it]
 81%|████████  | 7956/9864 [33:28:15<8:07:08, 15.32s/it]
 81%|████████  | 7957/9864 [33:28:29<7:57:48, 15.03s/it]
 81%|████████  | 7958/9864 [33:28:44<7:52:01, 14.86s/it]
 81%|████████  | 7959/9864 [33:28:57<7:42:42, 14.57s/it]
 81%|████████  | 7960/9864 [33:29:15<8:14:40, 15.59s/it]
                                                        
{'loss': '0.06737', 'grad_norm': '3.676', 'learning_rate': '1.893e-06', 'epoch': '0.807'}

 81%|████████  | 7960/9864 [33:29:15<8:14:40, 15.59s/it]
 81%|████████  | 7961/9864 [33:29:31<8:17:17, 15.68s/it]
 81%|████████  | 7962/9864 [33:29:45<7:58:03, 15.08s/it]
 81%|████████  | 7963/9864 [33:30:01<8:06:24, 15.35s/it]
 81%|████████  | 7964/9864 [33:30:15<7:56:40, 15.05s/it]
 81%|████████  | 7965/9864 [33:30:29<7:47:03, 14.76s/it]
 81%|████████  | 7966/9864 [33:30:44<7:45:08, 14.70s/it]
 81%|████████  | 7967/9864 [33:30:59<7:43:23, 14.66s/it]
 81%|████████  | 7968/9864 [33:31:16<8:13:24, 15.61s/it]
 81%|████████  | 7969/9864 [33:31:33<8:19:13, 15.81s/it]
 81%|████████  | 7970/9864 [33:31:46<7:59:04, 15.18s/it]
                                                        
{'loss': '0.07562', 'grad_norm': '3.81', 'learning_rate': '1.874e-06', 'epoch': '0.8081'}

 81%|████████  | 7970/9864 [33:31:46<7:59:04, 15.18s/it]
 81%|████████  | 7971/9864 [33:32:02<8:07:15, 15.44s/it]
 81%|████████  | 7972/9864 [33:32:17<7:55:57, 15.09s/it]
 81%|████████  | 7973/9864 [33:32:31<7:46:21, 14.80s/it]
 81%|████████  | 7974/9864 [33:32:45<7:37:22, 14.52s/it]
 81%|████████  | 7975/9864 [33:32:59<7:36:23, 14.50s/it]
 81%|████████  | 7976/9864 [33:33:19<8:25:29, 16.06s/it]
 81%|████████  | 7977/9864 [33:33:33<8:06:13, 15.46s/it]
 81%|████████  | 7978/9864 [33:33:49<8:13:33, 15.70s/it]
 81%|████████  | 7979/9864 [33:34:04<8:02:21, 15.35s/it]
 81%|████████  | 7980/9864 [33:34:18<7:51:06, 15.00s/it]
                                                        
{'loss': '0.08069', 'grad_norm': '5.799', 'learning_rate': '1.855e-06', 'epoch': '0.8091'}

 81%|████████  | 7980/9864 [33:34:18<7:51:06, 15.00s/it]
 81%|████████  | 7981/9864 [33:34:32<7:41:21, 14.70s/it]
 81%|████████  | 7982/9864 [33:34:46<7:39:36, 14.65s/it]
 81%|████████  | 7983/9864 [33:35:04<8:09:15, 15.61s/it]
 81%|████████  | 7984/9864 [33:35:20<8:11:08, 15.67s/it]
 81%|████████  | 7985/9864 [33:35:34<7:52:38, 15.09s/it]
 81%|████████  | 7986/9864 [33:35:48<7:43:54, 14.82s/it]
 81%|████████  | 7987/9864 [33:36:04<7:56:13, 15.22s/it]
 81%|████████  | 7988/9864 [33:36:19<7:50:47, 15.06s/it]
 81%|████████  | 7989/9864 [33:36:33<7:39:35, 14.71s/it]
 81%|████████  | 7990/9864 [33:36:47<7:36:49, 14.63s/it]
                                                        
{'loss': '0.06975', 'grad_norm': '3.42', 'learning_rate': '1.836e-06', 'epoch': '0.8101'}

 81%|████████  | 7990/9864 [33:36:47<7:36:49, 14.63s/it]
 81%|████████  | 7991/9864 [33:37:03<7:47:51, 14.99s/it]
 81%|████████  | 7992/9864 [33:37:19<7:55:48, 15.25s/it]
 81%|████████  | 7993/9864 [33:37:35<8:02:36, 15.48s/it]
 81%|████████  | 7994/9864 [33:37:49<7:50:47, 15.11s/it]
 81%|████████  | 7995/9864 [33:38:05<7:54:41, 15.24s/it]
 81%|████████  | 7996/9864 [33:38:19<7:45:55, 14.97s/it]
 81%|████████  | 7997/9864 [33:38:34<7:42:51, 14.87s/it]
 81%|████████  | 7998/9864 [33:38:50<7:58:28, 15.38s/it]
 81%|████████  | 7999/9864 [33:39:05<7:51:36, 15.17s/it]
 81%|████████  | 8000/9864 [33:39:21<7:59:53, 15.45s/it]
                                                        
{'loss': '0.08143', 'grad_norm': '3.317', 'learning_rate': '1.817e-06', 'epoch': '0.8111'}

 81%|████████  | 8000/9864 [33:39:21<7:59:53, 15.45s/it]

Writing model shards:   0%|          | 0/1 [00:00<?, ?it/s]

Writing model shards: 100%|██████████| 1/1 [00:10<00:00, 10.86s/it]
Writing model shards: 100%|██████████| 1/1 [00:10<00:00, 10.86s/it]
/home/dungpq6/Project/SpatialStack/src/qwen_vl/model/geometry_encoders/vggt_encoder.py:285: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(dtype=dtype):

 81%|████████  | 8001/9864 [33:40:40<17:53:37, 34.58s/it]
 81%|████████  | 8002/9864 [33:40:54<14:43:07, 28.46s/it]
 81%|████████  | 8003/9864 [33:41:10<12:43:18, 24.61s/it]
 81%|████████  | 8004/9864 [33:41:24<11:04:33, 21.44s/it]
 81%|████████  | 8005/9864 [33:41:40<10:12:02, 19.75s/it]
 81%|████████  | 8006/9864 [33:41:54<9:21:10, 18.12s/it] 
 81%|████████  | 8007/9864 [33:42:10<9:00:35, 17.47s/it]
 81%|████████  | 8008/9864 [33:42:26<8:42:37, 16.90s/it]
 81%|████████  | 8009/9864 [33:42:40<8:22:40, 16.26s/it]
 81%|████████  | 8010/9864 [33:42:54<8:00:44, 15.56s/it]
                                                        
{'loss': '0.07682', 'grad_norm': '4.209', 'learning_rate': '1.798e-06', 'epoch': '0.8121'}

 81%|████████  | 8010/9864 [33:42:54<8:00:44, 15.56s/it]
 81%|████████  | 8011/9864 [33:43:10<8:04:20, 15.68s/it]
 81%|████████  | 8012/9864 [33:43:27<8:12:36, 15.96s/it]
 81%|████████  | 8013/9864 [33:43:41<7:57:48, 15.49s/it]
 81%|████████  | 8014/9864 [33:43:55<7:43:27, 15.03s/it]
 81%|████████▏ | 8015/9864 [33:44:12<7:55:15, 15.42s/it]
 81%|████████▏ | 8016/9864 [33:44:28<8:00:34, 15.60s/it]
 81%|████████▏ | 8017/9864 [33:44:43<7:54:08, 15.40s/it]
 81%|████████▏ | 8018/9864 [33:44:58<7:55:34, 15.46s/it]
 81%|████████▏ | 8019/9864 [33:45:13<7:52:05, 15.35s/it]
 81%|████████▏ | 8020/9864 [33:45:30<8:02:02, 15.68s/it]
                                                        
{'loss': '0.08347', 'grad_norm': '4.193', 'learning_rate': '1.779e-06', 'epoch': '0.8131'}

 81%|████████▏ | 8020/9864 [33:45:30<8:02:02, 15.68s/it]
 81%|████████▏ | 8021/9864 [33:45:44<7:47:28, 15.22s/it]
 81%|████████▏ | 8022/9864 [33:45:59<7:45:29, 15.16s/it]
 81%|████████▏ | 8023/9864 [33:46:15<7:54:54, 15.48s/it]
 81%|████████▏ | 8024/9864 [33:46:30<7:53:04, 15.43s/it]
 81%|████████▏ | 8025/9864 [33:46:45<7:43:35, 15.13s/it]
 81%|████████▏ | 8026/9864 [33:47:00<7:47:23, 15.26s/it]
 81%|████████▏ | 8027/9864 [33:47:14<7:36:09, 14.90s/it]
 81%|████████▏ | 8028/9864 [33:47:29<7:29:48, 14.70s/it]
 81%|████████▏ | 8029/9864 [33:47:45<7:45:36, 15.22s/it]
 81%|████████▏ | 8030/9864 [33:47:59<7:36:12, 14.93s/it]
                                                        
{'loss': '0.0672', 'grad_norm': '3.155', 'learning_rate': '1.761e-06', 'epoch': '0.8141'}

 81%|████████▏ | 8030/9864 [33:47:59<7:36:12, 14.93s/it]
 81%|████████▏ | 8031/9864 [33:48:15<7:46:12, 15.26s/it]
 81%|████████▏ | 8032/9864 [33:48:31<7:52:46, 15.48s/it]
 81%|████████▏ | 8033/9864 [33:48:46<7:42:27, 15.15s/it]
 81%|████████▏ | 8034/9864 [33:49:02<7:51:00, 15.44s/it]
 81%|████████▏ | 8035/9864 [33:49:16<7:36:41, 14.98s/it]
 81%|████████▏ | 8036/9864 [33:49:30<7:25:37, 14.63s/it]
 81%|████████▏ | 8037/9864 [33:49:46<7:39:06, 15.08s/it]
 81%|████████▏ | 8038/9864 [33:50:00<7:33:33, 14.90s/it]
 81%|████████▏ | 8039/9864 [33:50:17<7:46:24, 15.33s/it]
 82%|████████▏ | 8040/9864 [33:50:31<7:38:49, 15.09s/it]
                                                        
{'loss': '0.08003', 'grad_norm': '4.334', 'learning_rate': '1.742e-06', 'epoch': '0.8152'}

 82%|████████▏ | 8040/9864 [33:50:31<7:38:49, 15.09s/it]
 82%|████████▏ | 8041/9864 [33:50:46<7:36:04, 15.01s/it]
 82%|████████▏ | 8042/9864 [33:51:02<7:42:56, 15.25s/it]
 82%|████████▏ | 8043/9864 [33:51:16<7:34:06, 14.96s/it]
 82%|████████▏ | 8044/9864 [33:51:30<7:23:51, 14.63s/it]
 82%|████████▏ | 8045/9864 [33:51:44<7:18:36, 14.47s/it]
 82%|████████▏ | 8046/9864 [33:52:02<7:51:51, 15.57s/it]
 82%|████████▏ | 8047/9864 [33:52:17<7:43:12, 15.30s/it]
 82%|████████▏ | 8048/9864 [33:52:31<7:34:36, 15.02s/it]
 82%|████████▏ | 8049/9864 [33:52:45<7:26:02, 14.75s/it]
 82%|████████▏ | 8050/9864 [33:53:01<7:38:49, 15.18s/it]
                                                        
{'loss': '0.0715', 'grad_norm': '4.826', 'learning_rate': '1.724e-06', 'epoch': '0.8162'}

 82%|████████▏ | 8050/9864 [33:53:01<7:38:49, 15.18s/it]
 82%|████████▏ | 8051/9864 [33:53:16<7:30:38, 14.91s/it]
 82%|████████▏ | 8052/9864 [33:53:30<7:24:41, 14.73s/it]
 82%|████████▏ | 8053/9864 [33:53:45<7:22:47, 14.67s/it]
 82%|████████▏ | 8054/9864 [33:54:02<7:45:51, 15.44s/it]
 82%|████████▏ | 8055/9864 [33:54:15<7:29:23, 14.91s/it]
 82%|████████▏ | 8056/9864 [33:54:29<7:21:03, 14.64s/it]
 82%|████████▏ | 8057/9864 [33:54:46<7:34:41, 15.10s/it]
 82%|████████▏ | 8058/9864 [33:55:00<7:30:47, 14.98s/it]
 82%|████████▏ | 8059/9864 [33:55:15<7:25:44, 14.82s/it]
 82%|████████▏ | 8060/9864 [33:55:29<7:21:25, 14.68s/it]
                                                        
{'loss': '0.06122', 'grad_norm': '4.349', 'learning_rate': '1.705e-06', 'epoch': '0.8172'}

 82%|████████▏ | 8060/9864 [33:55:29<7:21:25, 14.68s/it]
 82%|████████▏ | 8061/9864 [33:55:47<7:47:54, 15.57s/it]
 82%|████████▏ | 8062/9864 [33:56:03<7:55:29, 15.83s/it]
 82%|████████▏ | 8063/9864 [33:56:17<7:38:29, 15.27s/it]
 82%|████████▏ | 8064/9864 [33:56:32<7:32:06, 15.07s/it]
 82%|████████▏ | 8065/9864 [33:56:48<7:42:44, 15.43s/it]
 82%|████████▏ | 8066/9864 [33:57:02<7:32:01, 15.08s/it]
 82%|████████▏ | 8067/9864 [33:57:17<7:29:19, 15.00s/it]
 82%|████████▏ | 8068/9864 [33:57:32<7:25:03, 14.87s/it]
 82%|████████▏ | 8069/9864 [33:57:51<8:06:29, 16.26s/it]
 82%|████████▏ | 8070/9864 [33:58:06<7:54:11, 15.86s/it]
                                                        
{'loss': '0.07338', 'grad_norm': '4.145', 'learning_rate': '1.687e-06', 'epoch': '0.8182'}

 82%|████████▏ | 8070/9864 [33:58:06<7:54:11, 15.86s/it]
 82%|████████▏ | 8071/9864 [33:58:21<7:43:33, 15.51s/it]
 82%|████████▏ | 8072/9864 [33:58:35<7:33:37, 15.19s/it]
 82%|████████▏ | 8073/9864 [33:58:52<7:44:26, 15.56s/it]
 82%|████████▏ | 8074/9864 [33:59:06<7:32:33, 15.17s/it]
 82%|████████▏ | 8075/9864 [33:59:20<7:26:20, 14.97s/it]
 82%|████████▏ | 8076/9864 [33:59:35<7:21:43, 14.82s/it]
 82%|████████▏ | 8077/9864 [33:59:51<7:32:58, 15.21s/it]
 82%|████████▏ | 8078/9864 [34:00:06<7:34:32, 15.27s/it]
 82%|████████▏ | 8079/9864 [34:00:21<7:25:54, 14.99s/it]
 82%|████████▏ | 8080/9864 [34:00:37<7:36:58, 15.37s/it]
                                                        
{'loss': '0.06479', 'grad_norm': '3.825', 'learning_rate': '1.669e-06', 'epoch': '0.8192'}

 82%|████████▏ | 8080/9864 [34:00:37<7:36:58, 15.37s/it]
 82%|████████▏ | 8081/9864 [34:00:52<7:31:39, 15.20s/it]
 82%|████████▏ | 8082/9864 [34:01:06<7:21:44, 14.87s/it]
 82%|████████▏ | 8083/9864 [34:01:20<7:15:58, 14.69s/it]
 82%|████████▏ | 8084/9864 [34:01:37<7:31:32, 15.22s/it]
 82%|████████▏ | 8085/9864 [34:01:52<7:35:14, 15.35s/it]
 82%|████████▏ | 8086/9864 [34:02:08<7:39:54, 15.52s/it]
 82%|████████▏ | 8087/9864 [34:02:24<7:42:57, 15.63s/it]
 82%|████████▏ | 8088/9864 [34:02:39<7:32:38, 15.29s/it]
 82%|████████▏ | 8089/9864 [34:02:53<7:27:41, 15.13s/it]
 82%|████████▏ | 8090/9864 [34:03:08<7:20:24, 14.90s/it]
                                                        
{'loss': '0.07208', 'grad_norm': '4.097', 'learning_rate': '1.651e-06', 'epoch': '0.8202'}

 82%|████████▏ | 8090/9864 [34:03:08<7:20:24, 14.90s/it]
 82%|████████▏ | 8091/9864 [34:03:22<7:11:03, 14.59s/it]
 82%|████████▏ | 8092/9864 [34:03:36<7:11:57, 14.63s/it]
 82%|████████▏ | 8093/9864 [34:03:53<7:26:17, 15.12s/it]
 82%|████████▏ | 8094/9864 [34:04:08<7:31:51, 15.32s/it]
 82%|████████▏ | 8095/9864 [34:04:24<7:36:56, 15.50s/it]
 82%|████████▏ | 8096/9864 [34:04:39<7:29:16, 15.25s/it]
 82%|████████▏ | 8097/9864 [34:04:53<7:19:03, 14.91s/it]
 82%|████████▏ | 8098/9864 [34:05:08<7:16:36, 14.83s/it]
 82%|████████▏ | 8099/9864 [34:05:21<7:04:34, 14.43s/it]
 82%|████████▏ | 8100/9864 [34:05:35<7:00:42, 14.31s/it]
                                                        
{'loss': '0.07045', 'grad_norm': '4.492', 'learning_rate': '1.633e-06', 'epoch': '0.8212'}

 82%|████████▏ | 8100/9864 [34:05:35<7:00:42, 14.31s/it]
 82%|████████▏ | 8101/9864 [34:05:51<7:15:55, 14.84s/it]
 82%|████████▏ | 8102/9864 [34:06:07<7:25:06, 15.16s/it]
 82%|████████▏ | 8103/9864 [34:06:23<7:27:50, 15.26s/it]
 82%|████████▏ | 8104/9864 [34:06:37<7:16:37, 14.89s/it]
 82%|████████▏ | 8105/9864 [34:06:51<7:10:05, 14.67s/it]
 82%|████████▏ | 8106/9864 [34:07:05<7:04:45, 14.50s/it]
 82%|████████▏ | 8107/9864 [34:07:19<7:01:49, 14.40s/it]
 82%|████████▏ | 8108/9864 [34:07:35<7:15:21, 14.88s/it]
 82%|████████▏ | 8109/9864 [34:07:52<7:29:45, 15.38s/it]
 82%|████████▏ | 8110/9864 [34:08:09<7:47:48, 16.00s/it]
                                                        
{'loss': '0.06987', 'grad_norm': '3.946', 'learning_rate': '1.615e-06', 'epoch': '0.8222'}

 82%|████████▏ | 8110/9864 [34:08:09<7:47:48, 16.00s/it]
 82%|████████▏ | 8111/9864 [34:08:23<7:32:28, 15.49s/it]
 82%|████████▏ | 8112/9864 [34:08:38<7:19:32, 15.05s/it]
 82%|████████▏ | 8113/9864 [34:08:52<7:12:38, 14.82s/it]
 82%|████████▏ | 8114/9864 [34:09:06<7:06:15, 14.61s/it]
 82%|████████▏ | 8115/9864 [34:09:20<7:00:55, 14.44s/it]
 82%|████████▏ | 8116/9864 [34:09:36<7:12:15, 14.84s/it]
 82%|████████▏ | 8117/9864 [34:09:54<7:45:24, 15.98s/it]
 82%|████████▏ | 8118/9864 [34:10:09<7:32:00, 15.53s/it]
 82%|████████▏ | 8119/9864 [34:10:23<7:18:32, 15.08s/it]
 82%|████████▏ | 8120/9864 [34:10:37<7:11:48, 14.86s/it]
                                                        
{'loss': '0.07805', 'grad_norm': '4.635', 'learning_rate': '1.597e-06', 'epoch': '0.8233'}

 82%|████████▏ | 8120/9864 [34:10:37<7:11:48, 14.86s/it]
 82%|████████▏ | 8121/9864 [34:10:52<7:06:47, 14.69s/it]
 82%|████████▏ | 8122/9864 [34:11:05<6:58:32, 14.42s/it]
 82%|████████▏ | 8123/9864 [34:11:20<6:57:55, 14.40s/it]
 82%|████████▏ | 8124/9864 [34:11:40<7:50:37, 16.23s/it]
 82%|████████▏ | 8125/9864 [34:11:56<7:49:06, 16.19s/it]
 82%|████████▏ | 8126/9864 [34:12:10<7:29:19, 15.51s/it]
 82%|████████▏ | 8127/9864 [34:12:25<7:20:00, 15.20s/it]
 82%|████████▏ | 8128/9864 [34:12:39<7:11:26, 14.91s/it]
 82%|████████▏ | 8129/9864 [34:12:53<7:07:45, 14.79s/it]
 82%|████████▏ | 8130/9864 [34:13:08<7:01:45, 14.59s/it]
                                                        
{'loss': '0.05948', 'grad_norm': '2.844', 'learning_rate': '1.579e-06', 'epoch': '0.8243'}

 82%|████████▏ | 8130/9864 [34:13:08<7:01:45, 14.59s/it]
 82%|████████▏ | 8131/9864 [34:13:21<6:52:45, 14.29s/it]
 82%|████████▏ | 8132/9864 [34:13:41<7:37:53, 15.86s/it]
 82%|████████▏ | 8133/9864 [34:13:55<7:25:46, 15.45s/it]
 82%|████████▏ | 8134/9864 [34:14:10<7:17:37, 15.18s/it]
 82%|████████▏ | 8135/9864 [34:14:24<7:08:34, 14.87s/it]
 82%|████████▏ | 8136/9864 [34:14:38<7:00:35, 14.60s/it]
 82%|████████▏ | 8137/9864 [34:14:53<7:02:41, 14.69s/it]
 83%|████████▎ | 8138/9864 [34:15:07<6:58:19, 14.54s/it]
 83%|████████▎ | 8139/9864 [34:15:23<7:09:08, 14.93s/it]
 83%|████████▎ | 8140/9864 [34:15:40<7:31:43, 15.72s/it]
                                                        
{'loss': '0.06559', 'grad_norm': '4.833', 'learning_rate': '1.562e-06', 'epoch': '0.8253'}

 83%|████████▎ | 8140/9864 [34:15:40<7:31:43, 15.72s/it]
 83%|████████▎ | 8141/9864 [34:15:57<7:37:12, 15.92s/it]
 83%|████████▎ | 8142/9864 [34:16:11<7:26:49, 15.57s/it]
 83%|████████▎ | 8143/9864 [34:16:26<7:17:08, 15.24s/it]
 83%|████████▎ | 8144/9864 [34:16:40<7:06:50, 14.89s/it]
 83%|████████▎ | 8145/9864 [34:16:54<7:01:23, 14.71s/it]
 83%|████████▎ | 8146/9864 [34:17:08<6:52:24, 14.40s/it]
 83%|████████▎ | 8147/9864 [34:17:23<7:01:22, 14.73s/it]
 83%|████████▎ | 8148/9864 [34:17:41<7:24:38, 15.55s/it]
 83%|████████▎ | 8149/9864 [34:17:57<7:26:51, 15.63s/it]
 83%|████████▎ | 8150/9864 [34:18:11<7:18:05, 15.34s/it]
                                                        
{'loss': '0.06457', 'grad_norm': '4.647', 'learning_rate': '1.544e-06', 'epoch': '0.8263'}

 83%|████████▎ | 8150/9864 [34:18:11<7:18:05, 15.34s/it]
 83%|████████▎ | 8151/9864 [34:18:26<7:07:35, 14.98s/it]
 83%|████████▎ | 8152/9864 [34:18:40<6:59:24, 14.70s/it]
 83%|████████▎ | 8153/9864 [34:18:54<6:54:08, 14.52s/it]
 83%|████████▎ | 8154/9864 [34:19:08<6:49:58, 14.38s/it]
 83%|████████▎ | 8155/9864 [34:19:24<7:06:54, 14.99s/it]
 83%|████████▎ | 8156/9864 [34:19:40<7:14:31, 15.26s/it]
 83%|████████▎ | 8157/9864 [34:19:56<7:17:45, 15.39s/it]
 83%|████████▎ | 8158/9864 [34:20:11<7:18:29, 15.42s/it]
 83%|████████▎ | 8159/9864 [34:20:26<7:09:54, 15.13s/it]
 83%|████████▎ | 8160/9864 [34:20:40<7:02:27, 14.88s/it]
                                                        
{'loss': '0.07536', 'grad_norm': '4.633', 'learning_rate': '1.527e-06', 'epoch': '0.8273'}

 83%|████████▎ | 8160/9864 [34:20:40<7:02:27, 14.88s/it]
 83%|████████▎ | 8161/9864 [34:20:53<6:48:38, 14.40s/it]
 83%|████████▎ | 8162/9864 [34:21:08<6:48:24, 14.40s/it]
 83%|████████▎ | 8163/9864 [34:21:26<7:19:36, 15.51s/it]
 83%|████████▎ | 8164/9864 [34:21:41<7:17:04, 15.43s/it]
 83%|████████▎ | 8165/9864 [34:21:55<7:06:34, 15.06s/it]
 83%|████████▎ | 8166/9864 [34:22:11<7:16:24, 15.42s/it]
 83%|████████▎ | 8167/9864 [34:22:26<7:06:22, 15.08s/it]
 83%|████████▎ | 8168/9864 [34:22:40<6:58:41, 14.81s/it]
 83%|████████▎ | 8169/9864 [34:22:54<6:52:23, 14.60s/it]
 83%|████████▎ | 8170/9864 [34:23:09<6:54:19, 14.67s/it]
                                                        
{'loss': '0.07242', 'grad_norm': '4.729', 'learning_rate': '1.509e-06', 'epoch': '0.8283'}

 83%|████████▎ | 8170/9864 [34:23:09<6:54:19, 14.67s/it]
 83%|████████▎ | 8171/9864 [34:23:26<7:16:04, 15.45s/it]
 83%|████████▎ | 8172/9864 [34:23:42<7:21:30, 15.66s/it]
 83%|████████▎ | 8173/9864 [34:23:58<7:23:54, 15.75s/it]
 83%|████████▎ | 8174/9864 [34:24:12<7:09:36, 15.25s/it]
 83%|████████▎ | 8175/9864 [34:24:27<7:03:36, 15.05s/it]
 83%|████████▎ | 8176/9864 [34:24:41<6:55:41, 14.78s/it]
 83%|████████▎ | 8177/9864 [34:24:55<6:50:37, 14.60s/it]
 83%|████████▎ | 8178/9864 [34:25:11<7:00:45, 14.97s/it]
 83%|████████▎ | 8179/9864 [34:25:27<7:08:30, 15.26s/it]
 83%|████████▎ | 8180/9864 [34:25:43<7:17:34, 15.59s/it]
                                                        
{'loss': '0.0661', 'grad_norm': '4.674', 'learning_rate': '1.492e-06', 'epoch': '0.8293'}

 83%|████████▎ | 8180/9864 [34:25:43<7:17:34, 15.59s/it]
 83%|████████▎ | 8181/9864 [34:25:59<7:16:32, 15.56s/it]
 83%|████████▎ | 8182/9864 [34:26:13<7:05:49, 15.19s/it]
 83%|████████▎ | 8183/9864 [34:26:27<6:55:27, 14.83s/it]
 83%|████████▎ | 8184/9864 [34:26:41<6:46:44, 14.53s/it]
 83%|████████▎ | 8185/9864 [34:26:55<6:43:30, 14.42s/it]
 83%|████████▎ | 8186/9864 [34:27:12<7:01:11, 15.06s/it]
 83%|████████▎ | 8187/9864 [34:27:28<7:15:17, 15.57s/it]
 83%|████████▎ | 8188/9864 [34:27:45<7:22:26, 15.84s/it]
 83%|████████▎ | 8189/9864 [34:28:01<7:22:21, 15.85s/it]
 83%|████████▎ | 8190/9864 [34:28:15<7:09:12, 15.38s/it]
                                                        
{'loss': '0.0716', 'grad_norm': '4.032', 'learning_rate': '1.475e-06', 'epoch': '0.8304'}

 83%|████████▎ | 8190/9864 [34:28:15<7:09:12, 15.38s/it]
 83%|████████▎ | 8191/9864 [34:28:29<6:58:43, 15.02s/it]
 83%|████████▎ | 8192/9864 [34:28:44<6:53:41, 14.85s/it]
 83%|████████▎ | 8193/9864 [34:28:58<6:46:54, 14.61s/it]
 83%|████████▎ | 8194/9864 [34:29:16<7:15:47, 15.66s/it]
 83%|████████▎ | 8195/9864 [34:29:30<7:03:33, 15.23s/it]
 83%|████████▎ | 8196/9864 [34:29:46<7:11:33, 15.52s/it]
 83%|████████▎ | 8197/9864 [34:30:02<7:16:22, 15.71s/it]
 83%|████████▎ | 8198/9864 [34:30:16<7:00:42, 15.15s/it]
 83%|████████▎ | 8199/9864 [34:30:30<6:50:32, 14.79s/it]
 83%|████████▎ | 8200/9864 [34:30:44<6:41:36, 14.48s/it]
                                                        
{'loss': '0.0736', 'grad_norm': '3.634', 'learning_rate': '1.458e-06', 'epoch': '0.8314'}

 83%|████████▎ | 8200/9864 [34:30:44<6:41:36, 14.48s/it]
 83%|████████▎ | 8201/9864 [34:30:58<6:40:06, 14.44s/it]
 83%|████████▎ | 8202/9864 [34:31:13<6:41:09, 14.48s/it]
 83%|████████▎ | 8203/9864 [34:31:32<7:18:42, 15.85s/it]
 83%|████████▎ | 8204/9864 [34:31:48<7:20:46, 15.93s/it]
 83%|████████▎ | 8205/9864 [34:32:03<7:09:14, 15.52s/it]
 83%|████████▎ | 8206/9864 [34:32:17<6:56:19, 15.07s/it]
 83%|████████▎ | 8207/9864 [34:32:31<6:46:19, 14.71s/it]
 83%|████████▎ | 8208/9864 [34:32:44<6:39:08, 14.46s/it]
 83%|████████▎ | 8209/9864 [34:32:59<6:39:39, 14.49s/it]
 83%|████████▎ | 8210/9864 [34:33:15<6:48:31, 14.82s/it]
                                                        
{'loss': '0.066', 'grad_norm': '3.832', 'learning_rate': '1.44e-06', 'epoch': '0.8324'}

 83%|████████▎ | 8210/9864 [34:33:15<6:48:31, 14.82s/it]
 83%|████████▎ | 8211/9864 [34:33:33<7:13:56, 15.75s/it]
 83%|████████▎ | 8212/9864 [34:33:47<7:02:08, 15.33s/it]
 83%|████████▎ | 8213/9864 [34:34:02<7:03:55, 15.41s/it]
 83%|████████▎ | 8214/9864 [34:34:17<6:56:20, 15.14s/it]
 83%|████████▎ | 8215/9864 [34:34:31<6:46:52, 14.80s/it]
 83%|████████▎ | 8216/9864 [34:34:45<6:41:38, 14.62s/it]
 83%|████████▎ | 8217/9864 [34:35:00<6:40:04, 14.57s/it]
 83%|████████▎ | 8218/9864 [34:35:15<6:49:06, 14.91s/it]
 83%|████████▎ | 8219/9864 [34:35:33<7:11:00, 15.72s/it]
 83%|████████▎ | 8220/9864 [34:35:48<7:01:40, 15.39s/it]
                                                        
{'loss': '0.0597', 'grad_norm': '5.038', 'learning_rate': '1.424e-06', 'epoch': '0.8334'}

 83%|████████▎ | 8220/9864 [34:35:48<7:01:40, 15.39s/it]
 83%|████████▎ | 8221/9864 [34:36:04<7:10:15, 15.71s/it]
 83%|████████▎ | 8222/9864 [34:36:18<6:58:47, 15.30s/it]
 83%|████████▎ | 8223/9864 [34:36:33<6:49:32, 14.97s/it]
 83%|████████▎ | 8224/9864 [34:36:47<6:43:54, 14.78s/it]
 83%|████████▎ | 8225/9864 [34:37:01<6:39:22, 14.62s/it]
 83%|████████▎ | 8226/9864 [34:37:18<6:56:38, 15.26s/it]
 83%|████████▎ | 8227/9864 [34:37:34<7:02:53, 15.50s/it]
 83%|████████▎ | 8228/9864 [34:37:48<6:49:25, 15.02s/it]
 83%|████████▎ | 8229/9864 [34:38:04<7:01:39, 15.47s/it]
 83%|████████▎ | 8230/9864 [34:38:18<6:48:51, 15.01s/it]
                                                        
{'loss': '0.05684', 'grad_norm': '3.447', 'learning_rate': '1.407e-06', 'epoch': '0.8344'}

 83%|████████▎ | 8230/9864 [34:38:18<6:48:51, 15.01s/it]
 83%|████████▎ | 8231/9864 [34:38:32<6:41:30, 14.75s/it]
 83%|████████▎ | 8232/9864 [34:38:47<6:39:30, 14.69s/it]
 83%|████████▎ | 8233/9864 [34:39:01<6:35:17, 14.54s/it]
 83%|████████▎ | 8234/9864 [34:39:19<7:03:58, 15.61s/it]
 83%|████████▎ | 8235/9864 [34:39:35<7:07:09, 15.73s/it]
 83%|████████▎ | 8236/9864 [34:39:51<7:08:30, 15.79s/it]
 84%|████████▎ | 8237/9864 [34:40:06<6:56:31, 15.36s/it]
 84%|████████▎ | 8238/9864 [34:40:20<6:47:30, 15.04s/it]
 84%|████████▎ | 8239/9864 [34:40:34<6:41:23, 14.82s/it]
 84%|████████▎ | 8240/9864 [34:40:49<6:38:17, 14.72s/it]
                                                        
{'loss': '0.07096', 'grad_norm': '4.02', 'learning_rate': '1.39e-06', 'epoch': '0.8354'}

 84%|████████▎ | 8240/9864 [34:40:49<6:38:17, 14.72s/it]
 84%|████████▎ | 8241/9864 [34:41:03<6:36:01, 14.64s/it]
 84%|████████▎ | 8242/9864 [34:41:22<7:05:54, 15.76s/it]
 84%|████████▎ | 8243/9864 [34:41:39<7:23:40, 16.42s/it]
 84%|████████▎ | 8244/9864 [34:41:54<7:07:34, 15.84s/it]
 84%|████████▎ | 8245/9864 [34:42:09<6:57:23, 15.47s/it]
 84%|████████▎ | 8246/9864 [34:42:23<6:48:52, 15.16s/it]
 84%|████████▎ | 8247/9864 [34:42:37<6:40:13, 14.85s/it]
 84%|████████▎ | 8248/9864 [34:42:52<6:36:40, 14.73s/it]
 84%|████████▎ | 8249/9864 [34:43:08<6:47:36, 15.14s/it]
 84%|████████▎ | 8250/9864 [34:43:23<6:47:37, 15.15s/it]
                                                        
{'loss': '0.07128', 'grad_norm': '5.661', 'learning_rate': '1.373e-06', 'epoch': '0.8364'}

 84%|████████▎ | 8250/9864 [34:43:23<6:47:37, 15.15s/it]
 84%|████████▎ | 8251/9864 [34:43:39<6:51:22, 15.30s/it]
 84%|████████▎ | 8252/9864 [34:43:54<6:51:11, 15.31s/it]
 84%|████████▎ | 8253/9864 [34:44:08<6:43:37, 15.03s/it]
 84%|████████▎ | 8254/9864 [34:44:23<6:41:07, 14.95s/it]
 84%|████████▎ | 8255/9864 [34:44:37<6:34:54, 14.73s/it]
 84%|████████▎ | 8256/9864 [34:44:52<6:32:07, 14.63s/it]
 84%|████████▎ | 8257/9864 [34:45:08<6:42:16, 15.02s/it]
 84%|████████▎ | 8258/9864 [34:45:25<7:02:01, 15.77s/it]
 84%|████████▎ | 8259/9864 [34:45:41<7:04:22, 15.86s/it]
 84%|████████▎ | 8260/9864 [34:45:56<6:52:18, 15.42s/it]
                                                        
{'loss': '0.06653', 'grad_norm': '3.544', 'learning_rate': '1.357e-06', 'epoch': '0.8375'}

 84%|████████▎ | 8260/9864 [34:45:56<6:52:18, 15.42s/it]
 84%|████████▎ | 8261/9864 [34:46:09<6:40:04, 14.97s/it]
 84%|████████▍ | 8262/9864 [34:46:24<6:36:06, 14.84s/it]
 84%|████████▍ | 8263/9864 [34:46:38<6:30:45, 14.64s/it]
 84%|████████▍ | 8264/9864 [34:46:54<6:41:46, 15.07s/it]
 84%|████████▍ | 8265/9864 [34:47:10<6:49:15, 15.36s/it]
 84%|████████▍ | 8266/9864 [34:47:26<6:53:59, 15.54s/it]
 84%|████████▍ | 8267/9864 [34:47:42<6:55:31, 15.61s/it]
 84%|████████▍ | 8268/9864 [34:47:56<6:42:35, 15.13s/it]
 84%|████████▍ | 8269/9864 [34:48:11<6:37:42, 14.96s/it]
 84%|████████▍ | 8270/9864 [34:48:25<6:33:11, 14.80s/it]
                                                        
{'loss': '0.06454', 'grad_norm': '4.672', 'learning_rate': '1.34e-06', 'epoch': '0.8385'}

 84%|████████▍ | 8270/9864 [34:48:25<6:33:11, 14.80s/it]
 84%|████████▍ | 8271/9864 [34:48:39<6:23:27, 14.44s/it]
 84%|████████▍ | 8272/9864 [34:48:54<6:33:23, 14.83s/it]
 84%|████████▍ | 8273/9864 [34:49:12<6:58:47, 15.79s/it]
 84%|████████▍ | 8274/9864 [34:49:29<7:01:35, 15.91s/it]
 84%|████████▍ | 8275/9864 [34:49:43<6:49:27, 15.46s/it]
 84%|████████▍ | 8276/9864 [34:49:57<6:40:26, 15.13s/it]
 84%|████████▍ | 8277/9864 [34:50:12<6:35:41, 14.96s/it]
 84%|████████▍ | 8278/9864 [34:50:27<6:33:35, 14.89s/it]
 84%|████████▍ | 8279/9864 [34:50:41<6:27:53, 14.68s/it]
 84%|████████▍ | 8280/9864 [34:50:59<6:53:09, 15.65s/it]
                                                        
{'loss': '0.07064', 'grad_norm': '4.484', 'learning_rate': '1.324e-06', 'epoch': '0.8395'}

 84%|████████▍ | 8280/9864 [34:50:59<6:53:09, 15.65s/it]
 84%|████████▍ | 8281/9864 [34:51:15<6:59:25, 15.90s/it]
 84%|████████▍ | 8282/9864 [34:51:31<6:59:55, 15.93s/it]
 84%|████████▍ | 8283/9864 [34:51:45<6:42:11, 15.26s/it]
 84%|████████▍ | 8284/9864 [34:51:59<6:33:08, 14.93s/it]
 84%|████████▍ | 8285/9864 [34:52:13<6:27:54, 14.74s/it]
 84%|████████▍ | 8286/9864 [34:52:28<6:24:47, 14.63s/it]
 84%|████████▍ | 8287/9864 [34:52:42<6:22:56, 14.57s/it]
 84%|████████▍ | 8288/9864 [34:53:00<6:51:53, 15.68s/it]
 84%|████████▍ | 8289/9864 [34:53:16<6:52:40, 15.72s/it]
 84%|████████▍ | 8290/9864 [34:53:32<6:54:55, 15.82s/it]
                                                        
{'loss': '0.05529', 'grad_norm': '2.619', 'learning_rate': '1.308e-06', 'epoch': '0.8405'}

 84%|████████▍ | 8290/9864 [34:53:32<6:54:55, 15.82s/it]
 84%|████████▍ | 8291/9864 [34:53:46<6:40:00, 15.26s/it]
 84%|████████▍ | 8292/9864 [34:54:01<6:35:04, 15.08s/it]
 84%|████████▍ | 8293/9864 [34:54:16<6:31:41, 14.96s/it]
 84%|████████▍ | 8294/9864 [34:54:30<6:24:13, 14.68s/it]
 84%|████████▍ | 8295/9864 [34:54:44<6:22:49, 14.64s/it]
 84%|████████▍ | 8296/9864 [34:55:01<6:35:55, 15.15s/it]
 84%|████████▍ | 8297/9864 [34:55:18<6:52:49, 15.81s/it]
 84%|████████▍ | 8298/9864 [34:55:34<6:57:29, 16.00s/it]
 84%|████████▍ | 8299/9864 [34:55:49<6:45:55, 15.56s/it]
 84%|████████▍ | 8300/9864 [34:56:03<6:33:20, 15.09s/it]
                                                        
{'loss': '0.0597', 'grad_norm': '4.026', 'learning_rate': '1.291e-06', 'epoch': '0.8415'}

 84%|████████▍ | 8300/9864 [34:56:03<6:33:20, 15.09s/it]
 84%|████████▍ | 8301/9864 [34:56:17<6:27:55, 14.89s/it]
 84%|████████▍ | 8302/9864 [34:56:32<6:23:27, 14.73s/it]
 84%|████████▍ | 8303/9864 [34:56:47<6:31:42, 15.06s/it]
 84%|████████▍ | 8304/9864 [34:57:03<6:36:08, 15.24s/it]
 84%|████████▍ | 8305/9864 [34:57:21<6:56:06, 16.01s/it]
 84%|████████▍ | 8306/9864 [34:57:34<6:36:29, 15.27s/it]
 84%|████████▍ | 8307/9864 [34:57:48<6:26:23, 14.89s/it]
 84%|████████▍ | 8308/9864 [34:58:03<6:24:35, 14.83s/it]
 84%|████████▍ | 8309/9864 [34:58:18<6:21:11, 14.71s/it]
 84%|████████▍ | 8310/9864 [34:58:32<6:18:00, 14.60s/it]
                                                        
{'loss': '0.05888', 'grad_norm': '4.405', 'learning_rate': '1.275e-06', 'epoch': '0.8425'}

 84%|████████▍ | 8310/9864 [34:58:32<6:18:00, 14.60s/it]
 84%|████████▍ | 8311/9864 [34:58:46<6:15:06, 14.49s/it]
 84%|████████▍ | 8312/9864 [34:59:04<6:40:26, 15.48s/it]
 84%|████████▍ | 8313/9864 [34:59:18<6:32:15, 15.17s/it]
 84%|████████▍ | 8314/9864 [34:59:36<6:52:17, 15.96s/it]
 84%|████████▍ | 8315/9864 [34:59:50<6:39:09, 15.46s/it]
 84%|████████▍ | 8316/9864 [35:00:05<6:28:46, 15.07s/it]
 84%|████████▍ | 8317/9864 [35:00:19<6:23:42, 14.88s/it]
 84%|████████▍ | 8318/9864 [35:00:33<6:17:48, 14.66s/it]
 84%|████████▍ | 8319/9864 [35:00:49<6:25:23, 14.97s/it]
 84%|████████▍ | 8320/9864 [35:01:03<6:21:47, 14.84s/it]
                                                        
{'loss': '0.07349', 'grad_norm': '5.228', 'learning_rate': '1.259e-06', 'epoch': '0.8435'}

 84%|████████▍ | 8320/9864 [35:01:03<6:21:47, 14.84s/it]
 84%|████████▍ | 8321/9864 [35:01:23<6:56:39, 16.20s/it]
 84%|████████▍ | 8322/9864 [35:01:37<6:41:46, 15.63s/it]
 84%|████████▍ | 8323/9864 [35:01:52<6:32:28, 15.28s/it]
 84%|████████▍ | 8324/9864 [35:02:05<6:20:35, 14.83s/it]
 84%|████████▍ | 8325/9864 [35:02:19<6:12:27, 14.52s/it]
 84%|████████▍ | 8326/9864 [35:02:33<6:10:24, 14.45s/it]
 84%|████████▍ | 8327/9864 [35:02:49<6:22:08, 14.92s/it]
 84%|████████▍ | 8328/9864 [35:03:06<6:31:21, 15.29s/it]
 84%|████████▍ | 8329/9864 [35:03:22<6:38:36, 15.58s/it]
 84%|████████▍ | 8330/9864 [35:03:37<6:38:19, 15.58s/it]
                                                        
{'loss': '0.06507', 'grad_norm': '4.129', 'learning_rate': '1.243e-06', 'epoch': '0.8446'}

 84%|████████▍ | 8330/9864 [35:03:37<6:38:19, 15.58s/it]
 84%|████████▍ | 8331/9864 [35:03:52<6:27:40, 15.17s/it]
 84%|████████▍ | 8332/9864 [35:04:06<6:20:58, 14.92s/it]
 84%|████████▍ | 8333/9864 [35:04:21<6:20:38, 14.92s/it]
 84%|████████▍ | 8334/9864 [35:04:37<6:27:45, 15.21s/it]
 84%|████████▍ | 8335/9864 [35:04:53<6:33:09, 15.43s/it]
 85%|████████▍ | 8336/9864 [35:05:09<6:39:52, 15.70s/it]
 85%|████████▍ | 8337/9864 [35:05:23<6:27:43, 15.24s/it]
 85%|████████▍ | 8338/9864 [35:05:39<6:35:05, 15.53s/it]
 85%|████████▍ | 8339/9864 [35:05:54<6:28:06, 15.27s/it]
 85%|████████▍ | 8340/9864 [35:06:09<6:22:26, 15.06s/it]
                                                        
{'loss': '0.06139', 'grad_norm': '3.667', 'learning_rate': '1.228e-06', 'epoch': '0.8456'}

 85%|████████▍ | 8340/9864 [35:06:09<6:22:26, 15.06s/it]
 85%|████████▍ | 8341/9864 [35:06:23<6:19:06, 14.94s/it]
 85%|████████▍ | 8342/9864 [35:06:39<6:27:56, 15.29s/it]
 85%|████████▍ | 8343/9864 [35:06:54<6:19:02, 14.95s/it]
 85%|████████▍ | 8344/9864 [35:07:11<6:41:01, 15.83s/it]
 85%|████████▍ | 8345/9864 [35:07:26<6:29:15, 15.38s/it]
 85%|████████▍ | 8346/9864 [35:07:42<6:35:14, 15.62s/it]
 85%|████████▍ | 8347/9864 [35:07:57<6:27:08, 15.31s/it]
 85%|████████▍ | 8348/9864 [35:08:11<6:19:37, 15.02s/it]
 85%|████████▍ | 8349/9864 [35:08:27<6:27:17, 15.34s/it]
 85%|████████▍ | 8350/9864 [35:08:41<6:19:01, 15.02s/it]
                                                        
{'loss': '0.0581', 'grad_norm': '5.25', 'learning_rate': '1.212e-06', 'epoch': '0.8466'}

 85%|████████▍ | 8350/9864 [35:08:41<6:19:01, 15.02s/it]
 85%|████████▍ | 8351/9864 [35:08:59<6:41:12, 15.91s/it]
 85%|████████▍ | 8352/9864 [35:09:12<6:20:33, 15.10s/it]
 85%|████████▍ | 8353/9864 [35:09:28<6:25:50, 15.32s/it]
 85%|████████▍ | 8354/9864 [35:09:43<6:17:04, 14.98s/it]
 85%|████████▍ | 8355/9864 [35:09:57<6:14:15, 14.88s/it]
 85%|████████▍ | 8356/9864 [35:10:11<6:08:40, 14.67s/it]
 85%|████████▍ | 8357/9864 [35:10:28<6:19:56, 15.13s/it]
 85%|████████▍ | 8358/9864 [35:10:42<6:15:12, 14.95s/it]
 85%|████████▍ | 8359/9864 [35:11:00<6:34:51, 15.74s/it]
 85%|████████▍ | 8360/9864 [35:11:14<6:24:20, 15.33s/it]
                                                        
{'loss': '0.04978', 'grad_norm': '4.161', 'learning_rate': '1.196e-06', 'epoch': '0.8476'}

 85%|████████▍ | 8360/9864 [35:11:14<6:24:20, 15.33s/it]
 85%|████████▍ | 8361/9864 [35:11:30<6:27:59, 15.49s/it]
 85%|████████▍ | 8362/9864 [35:11:44<6:20:16, 15.19s/it]
 85%|████████▍ | 8363/9864 [35:11:58<6:10:07, 14.80s/it]
 85%|████████▍ | 8364/9864 [35:12:12<6:05:15, 14.61s/it]
 85%|████████▍ | 8365/9864 [35:12:27<6:04:07, 14.57s/it]
 85%|████████▍ | 8366/9864 [35:12:45<6:26:30, 15.48s/it]
 85%|████████▍ | 8367/9864 [35:12:59<6:17:11, 15.12s/it]
 85%|████████▍ | 8368/9864 [35:13:16<6:32:15, 15.73s/it]
 85%|████████▍ | 8369/9864 [35:13:31<6:25:52, 15.49s/it]
 85%|████████▍ | 8370/9864 [35:13:45<6:17:16, 15.15s/it]
                                                        
{'loss': '0.06248', 'grad_norm': '3.999', 'learning_rate': '1.181e-06', 'epoch': '0.8486'}

 85%|████████▍ | 8370/9864 [35:13:45<6:17:16, 15.15s/it]
 85%|████████▍ | 8371/9864 [35:13:59<6:08:27, 14.81s/it]
 85%|████████▍ | 8372/9864 [35:14:14<6:05:26, 14.70s/it]
 85%|████████▍ | 8373/9864 [35:14:28<6:03:47, 14.64s/it]
 85%|████████▍ | 8374/9864 [35:14:44<6:12:01, 14.98s/it]
 85%|████████▍ | 8375/9864 [35:15:00<6:17:52, 15.23s/it]
 85%|████████▍ | 8376/9864 [35:15:16<6:25:54, 15.56s/it]
 85%|████████▍ | 8377/9864 [35:15:32<6:28:03, 15.66s/it]
 85%|████████▍ | 8378/9864 [35:15:47<6:20:11, 15.35s/it]
 85%|████████▍ | 8379/9864 [35:16:01<6:13:00, 15.07s/it]
 85%|████████▍ | 8380/9864 [35:16:15<6:07:58, 14.88s/it]
                                                        
{'loss': '0.06529', 'grad_norm': '4.787', 'learning_rate': '1.165e-06', 'epoch': '0.8496'}

 85%|████████▍ | 8380/9864 [35:16:15<6:07:58, 14.88s/it]
 85%|████████▍ | 8381/9864 [35:16:32<6:16:27, 15.23s/it]
 85%|████████▍ | 8382/9864 [35:16:48<6:23:13, 15.52s/it]
 85%|████████▍ | 8383/9864 [35:17:02<6:10:57, 15.03s/it]
 85%|████████▍ | 8384/9864 [35:17:19<6:28:35, 15.75s/it]
 85%|████████▌ | 8385/9864 [35:17:33<6:15:01, 15.21s/it]
 85%|████████▌ | 8386/9864 [35:17:48<6:12:22, 15.12s/it]
 85%|████████▌ | 8387/9864 [35:18:03<6:08:31, 14.97s/it]
 85%|████████▌ | 8388/9864 [35:18:18<6:08:53, 15.00s/it]
 85%|████████▌ | 8389/9864 [35:18:32<6:03:23, 14.78s/it]
 85%|████████▌ | 8390/9864 [35:18:49<6:24:05, 15.63s/it]
                                                        
{'loss': '0.06691', 'grad_norm': '3.645', 'learning_rate': '1.15e-06', 'epoch': '0.8506'}

 85%|████████▌ | 8390/9864 [35:18:49<6:24:05, 15.63s/it]
 85%|████████▌ | 8391/9864 [35:19:04<6:13:41, 15.22s/it]
 85%|████████▌ | 8392/9864 [35:19:22<6:33:04, 16.02s/it]
 85%|████████▌ | 8393/9864 [35:19:36<6:17:06, 15.38s/it]
 85%|████████▌ | 8394/9864 [35:19:49<6:04:31, 14.88s/it]
 85%|████████▌ | 8395/9864 [35:20:03<5:56:21, 14.56s/it]
 85%|████████▌ | 8396/9864 [35:20:17<5:53:58, 14.47s/it]
 85%|████████▌ | 8397/9864 [35:20:33<6:03:42, 14.88s/it]
 85%|████████▌ | 8398/9864 [35:20:49<6:11:34, 15.21s/it]
 85%|████████▌ | 8399/9864 [35:21:04<6:05:56, 14.99s/it]
 85%|████████▌ | 8400/9864 [35:21:21<6:25:59, 15.82s/it]
                                                        
{'loss': '0.07206', 'grad_norm': '4.675', 'learning_rate': '1.135e-06', 'epoch': '0.8517'}

 85%|████████▌ | 8400/9864 [35:21:21<6:25:59, 15.82s/it]
 85%|████████▌ | 8401/9864 [35:21:36<6:15:43, 15.41s/it]
 85%|████████▌ | 8402/9864 [35:21:50<6:06:16, 15.03s/it]
 85%|████████▌ | 8403/9864 [35:22:04<5:55:51, 14.61s/it]
 85%|████████▌ | 8404/9864 [35:22:18<5:52:53, 14.50s/it]
 85%|████████▌ | 8405/9864 [35:22:34<6:04:30, 14.99s/it]
 85%|████████▌ | 8406/9864 [35:22:48<5:56:24, 14.67s/it]
 85%|████████▌ | 8407/9864 [35:23:04<6:07:47, 15.15s/it]
 85%|████████▌ | 8408/9864 [35:23:20<6:14:04, 15.42s/it]
 85%|████████▌ | 8409/9864 [35:23:36<6:17:17, 15.56s/it]
 85%|████████▌ | 8410/9864 [35:23:50<6:08:18, 15.20s/it]
                                                        
{'loss': '0.05702', 'grad_norm': '3.396', 'learning_rate': '1.12e-06', 'epoch': '0.8527'}

 85%|████████▌ | 8410/9864 [35:23:50<6:08:18, 15.20s/it]
 85%|████████▌ | 8411/9864 [35:24:05<6:02:29, 14.97s/it]
 85%|████████▌ | 8412/9864 [35:24:19<5:55:55, 14.71s/it]
 85%|████████▌ | 8413/9864 [35:24:35<6:03:01, 15.01s/it]
 85%|████████▌ | 8414/9864 [35:24:51<6:11:18, 15.36s/it]
 85%|████████▌ | 8415/9864 [35:25:07<6:14:27, 15.51s/it]
 85%|████████▌ | 8416/9864 [35:25:21<6:03:45, 15.07s/it]
 85%|████████▌ | 8417/9864 [35:25:37<6:09:50, 15.34s/it]
 85%|████████▌ | 8418/9864 [35:25:51<5:59:40, 14.92s/it]
 85%|████████▌ | 8419/9864 [35:26:05<5:58:07, 14.87s/it]
 85%|████████▌ | 8420/9864 [35:26:21<6:05:39, 15.19s/it]
                                                        
{'loss': '0.06529', 'grad_norm': '4.71', 'learning_rate': '1.105e-06', 'epoch': '0.8537'}

 85%|████████▌ | 8420/9864 [35:26:21<6:05:39, 15.19s/it]
 85%|████████▌ | 8421/9864 [35:26:36<6:01:26, 15.03s/it]
 85%|████████▌ | 8422/9864 [35:26:52<6:08:20, 15.33s/it]
 85%|████████▌ | 8423/9864 [35:27:09<6:17:00, 15.70s/it]
 85%|████████▌ | 8424/9864 [35:27:24<6:14:46, 15.62s/it]
 85%|████████▌ | 8425/9864 [35:27:38<6:03:43, 15.17s/it]
 85%|████████▌ | 8426/9864 [35:27:53<5:59:20, 14.99s/it]
 85%|████████▌ | 8427/9864 [35:28:08<6:02:58, 15.16s/it]
 85%|████████▌ | 8428/9864 [35:28:22<5:54:26, 14.81s/it]
 85%|████████▌ | 8429/9864 [35:28:39<6:06:35, 15.33s/it]
 85%|████████▌ | 8430/9864 [35:28:54<6:03:37, 15.21s/it]
                                                        
{'loss': '0.06598', 'grad_norm': '3.361', 'learning_rate': '1.09e-06', 'epoch': '0.8547'}

 85%|████████▌ | 8430/9864 [35:28:54<6:03:37, 15.21s/it]
 85%|████████▌ | 8431/9864 [35:29:10<6:09:47, 15.48s/it]
 85%|████████▌ | 8432/9864 [35:29:26<6:12:39, 15.61s/it]
 85%|████████▌ | 8433/9864 [35:29:40<6:03:23, 15.24s/it]
 86%|████████▌ | 8434/9864 [35:29:54<5:55:24, 14.91s/it]
 86%|████████▌ | 8435/9864 [35:30:11<6:04:38, 15.31s/it]
 86%|████████▌ | 8436/9864 [35:30:26<6:08:12, 15.47s/it]
 86%|████████▌ | 8437/9864 [35:30:41<6:03:15, 15.27s/it]
 86%|████████▌ | 8438/9864 [35:30:55<5:51:37, 14.79s/it]
 86%|████████▌ | 8439/9864 [35:31:11<6:03:23, 15.30s/it]
 86%|████████▌ | 8440/9864 [35:31:27<6:03:27, 15.31s/it]
                                                        
{'loss': '0.05544', 'grad_norm': '2.977', 'learning_rate': '1.075e-06', 'epoch': '0.8557'}

 86%|████████▌ | 8440/9864 [35:31:27<6:03:27, 15.31s/it]
 86%|████████▌ | 8441/9864 [35:31:41<5:54:28, 14.95s/it]
 86%|████████▌ | 8442/9864 [35:31:55<5:49:13, 14.74s/it]
 86%|████████▌ | 8443/9864 [35:32:10<5:53:05, 14.91s/it]
 86%|████████▌ | 8444/9864 [35:32:27<6:06:14, 15.47s/it]
 86%|████████▌ | 8445/9864 [35:32:41<5:53:10, 14.93s/it]
 86%|████████▌ | 8446/9864 [35:32:55<5:48:20, 14.74s/it]
 86%|████████▌ | 8447/9864 [35:33:09<5:44:04, 14.57s/it]
 86%|████████▌ | 8448/9864 [35:33:27<6:06:37, 15.53s/it]
 86%|████████▌ | 8449/9864 [35:33:41<5:56:07, 15.10s/it]
 86%|████████▌ | 8450/9864 [35:33:55<5:49:42, 14.84s/it]
                                                        
{'loss': '0.05986', 'grad_norm': '4.761', 'learning_rate': '1.06e-06', 'epoch': '0.8567'}

 86%|████████▌ | 8450/9864 [35:33:55<5:49:42, 14.84s/it]
 86%|████████▌ | 8451/9864 [35:34:12<5:59:42, 15.27s/it]
 86%|████████▌ | 8452/9864 [35:34:27<6:02:17, 15.40s/it]
 86%|████████▌ | 8453/9864 [35:34:42<6:00:22, 15.32s/it]
 86%|████████▌ | 8454/9864 [35:34:57<5:54:13, 15.07s/it]
 86%|████████▌ | 8455/9864 [35:35:13<5:59:03, 15.29s/it]
 86%|████████▌ | 8456/9864 [35:35:27<5:51:08, 14.96s/it]
 86%|████████▌ | 8457/9864 [35:35:43<5:59:16, 15.32s/it]
 86%|████████▌ | 8458/9864 [35:35:57<5:50:16, 14.95s/it]
 86%|████████▌ | 8459/9864 [35:36:15<6:12:26, 15.90s/it]
 86%|████████▌ | 8460/9864 [35:36:30<5:59:54, 15.38s/it]
                                                        
{'loss': '0.06995', 'grad_norm': '4.507', 'learning_rate': '1.045e-06', 'epoch': '0.8577'}

 86%|████████▌ | 8460/9864 [35:36:30<5:59:54, 15.38s/it]
 86%|████████▌ | 8461/9864 [35:36:44<5:52:20, 15.07s/it]
 86%|████████▌ | 8462/9864 [35:36:58<5:48:58, 14.93s/it]
 86%|████████▌ | 8463/9864 [35:37:15<6:01:36, 15.49s/it]
 86%|████████▌ | 8464/9864 [35:37:30<5:54:06, 15.18s/it]
 86%|████████▌ | 8465/9864 [35:37:46<6:00:10, 15.45s/it]
 86%|████████▌ | 8466/9864 [35:38:00<5:54:46, 15.23s/it]
 86%|████████▌ | 8467/9864 [35:38:19<6:15:29, 16.13s/it]
 86%|████████▌ | 8468/9864 [35:38:34<6:07:26, 15.79s/it]
 86%|████████▌ | 8469/9864 [35:38:48<5:55:18, 15.28s/it]
 86%|████████▌ | 8470/9864 [35:39:02<5:46:25, 14.91s/it]
                                                        
{'loss': '0.05569', 'grad_norm': '4.489', 'learning_rate': '1.031e-06', 'epoch': '0.8587'}

 86%|████████▌ | 8470/9864 [35:39:02<5:46:25, 14.91s/it]
 86%|████████▌ | 8471/9864 [35:39:17<5:49:04, 15.04s/it]
 86%|████████▌ | 8472/9864 [35:39:31<5:39:54, 14.65s/it]
 86%|████████▌ | 8473/9864 [35:39:45<5:33:51, 14.40s/it]
 86%|████████▌ | 8474/9864 [35:40:00<5:42:18, 14.78s/it]
 86%|████████▌ | 8475/9864 [35:40:19<6:07:10, 15.86s/it]
 86%|████████▌ | 8476/9864 [35:40:33<5:57:21, 15.45s/it]
 86%|████████▌ | 8477/9864 [35:40:47<5:46:16, 14.98s/it]
 86%|████████▌ | 8478/9864 [35:41:01<5:38:17, 14.64s/it]
 86%|████████▌ | 8479/9864 [35:41:15<5:36:17, 14.57s/it]
 86%|████████▌ | 8480/9864 [35:41:32<5:47:15, 15.05s/it]
                                                        
{'loss': '0.0584', 'grad_norm': '4.307', 'learning_rate': '1.016e-06', 'epoch': '0.8598'}

 86%|████████▌ | 8480/9864 [35:41:32<5:47:15, 15.05s/it]
 86%|████████▌ | 8481/9864 [35:41:48<5:55:52, 15.44s/it]
 86%|████████▌ | 8482/9864 [35:42:02<5:46:06, 15.03s/it]
 86%|████████▌ | 8483/9864 [35:42:20<6:09:16, 16.04s/it]
 86%|████████▌ | 8484/9864 [35:42:34<5:55:22, 15.45s/it]
 86%|████████▌ | 8485/9864 [35:42:48<5:44:51, 15.00s/it]
 86%|████████▌ | 8486/9864 [35:43:03<5:43:54, 14.97s/it]
 86%|████████▌ | 8487/9864 [35:43:17<5:37:22, 14.70s/it]
 86%|████████▌ | 8488/9864 [35:43:34<5:49:17, 15.23s/it]
 86%|████████▌ | 8489/9864 [35:43:50<5:53:20, 15.42s/it]
 86%|████████▌ | 8490/9864 [35:44:06<5:57:49, 15.63s/it]
                                                        
{'loss': '0.0574', 'grad_norm': '4.984', 'learning_rate': '1.002e-06', 'epoch': '0.8608'}

 86%|████████▌ | 8490/9864 [35:44:06<5:57:49, 15.63s/it]
 86%|████████▌ | 8491/9864 [35:44:22<6:01:42, 15.81s/it]
 86%|████████▌ | 8492/9864 [35:44:37<5:52:06, 15.40s/it]
 86%|████████▌ | 8493/9864 [35:44:51<5:42:19, 14.98s/it]
 86%|████████▌ | 8494/9864 [35:45:05<5:38:57, 14.85s/it]
 86%|████████▌ | 8495/9864 [35:45:20<5:36:55, 14.77s/it]
 86%|████████▌ | 8496/9864 [35:45:36<5:45:49, 15.17s/it]
 86%|████████▌ | 8497/9864 [35:45:54<6:06:11, 16.07s/it]
 86%|████████▌ | 8498/9864 [35:46:11<6:09:30, 16.23s/it]
 86%|████████▌ | 8499/9864 [35:46:25<5:53:50, 15.55s/it]
 86%|████████▌ | 8500/9864 [35:46:39<5:47:01, 15.27s/it]
                                                        
{'loss': '0.05534', 'grad_norm': '2.169', 'learning_rate': '9.877e-07', 'epoch': '0.8618'}

 86%|████████▌ | 8500/9864 [35:46:39<5:47:01, 15.27s/it]
 86%|████████▌ | 8501/9864 [35:46:54<5:42:43, 15.09s/it]
 86%|████████▌ | 8502/9864 [35:47:08<5:36:49, 14.84s/it]
 86%|████████▌ | 8503/9864 [35:47:24<5:46:55, 15.29s/it]
 86%|████████▌ | 8504/9864 [35:47:40<5:49:29, 15.42s/it]
 86%|████████▌ | 8505/9864 [35:47:55<5:44:19, 15.20s/it]
 86%|████████▌ | 8506/9864 [35:48:13<6:02:03, 16.00s/it]
 86%|████████▌ | 8507/9864 [35:48:27<5:48:41, 15.42s/it]
 86%|████████▋ | 8508/9864 [35:48:41<5:39:21, 15.02s/it]
 86%|████████▋ | 8509/9864 [35:48:55<5:33:36, 14.77s/it]
 86%|████████▋ | 8510/9864 [35:49:09<5:29:39, 14.61s/it]
                                                        
{'loss': '0.05038', 'grad_norm': '4.86', 'learning_rate': '9.735e-07', 'epoch': '0.8628'}

 86%|████████▋ | 8510/9864 [35:49:09<5:29:39, 14.61s/it]
 86%|████████▋ | 8511/9864 [35:49:26<5:43:01, 15.21s/it]
 86%|████████▋ | 8512/9864 [35:49:42<5:48:31, 15.47s/it]
 86%|████████▋ | 8513/9864 [35:49:58<5:53:57, 15.72s/it]
 86%|████████▋ | 8514/9864 [35:50:14<5:55:14, 15.79s/it]
 86%|████████▋ | 8515/9864 [35:50:28<5:42:21, 15.23s/it]
 86%|████████▋ | 8516/9864 [35:50:42<5:35:25, 14.93s/it]
 86%|████████▋ | 8517/9864 [35:50:57<5:30:40, 14.73s/it]
 86%|████████▋ | 8518/9864 [35:51:10<5:24:21, 14.46s/it]
 86%|████████▋ | 8519/9864 [35:51:27<5:37:25, 15.05s/it]
 86%|████████▋ | 8520/9864 [35:51:43<5:45:59, 15.45s/it]
                                                        
{'loss': '0.0529', 'grad_norm': '5.295', 'learning_rate': '9.594e-07', 'epoch': '0.8638'}

 86%|████████▋ | 8520/9864 [35:51:43<5:45:59, 15.45s/it]
 86%|████████▋ | 8521/9864 [35:52:01<5:58:59, 16.04s/it]
 86%|████████▋ | 8522/9864 [35:52:15<5:49:47, 15.64s/it]
 86%|████████▋ | 8523/9864 [35:52:29<5:39:22, 15.18s/it]
 86%|████████▋ | 8524/9864 [35:52:44<5:32:13, 14.88s/it]
 86%|████████▋ | 8525/9864 [35:52:58<5:25:41, 14.59s/it]
 86%|████████▋ | 8526/9864 [35:53:12<5:25:55, 14.62s/it]
 86%|████████▋ | 8527/9864 [35:53:28<5:35:20, 15.05s/it]
 86%|████████▋ | 8528/9864 [35:53:44<5:42:04, 15.36s/it]
 86%|████████▋ | 8529/9864 [35:54:01<5:53:27, 15.89s/it]
 86%|████████▋ | 8530/9864 [35:54:16<5:43:04, 15.43s/it]
                                                        
{'loss': '0.05444', 'grad_norm': '5.217', 'learning_rate': '9.454e-07', 'epoch': '0.8648'}

 86%|████████▋ | 8530/9864 [35:54:16<5:43:04, 15.43s/it]
 86%|████████▋ | 8531/9864 [35:54:30<5:37:12, 15.18s/it]
 86%|████████▋ | 8532/9864 [35:54:44<5:25:30, 14.66s/it]
 87%|████████▋ | 8533/9864 [35:54:58<5:20:10, 14.43s/it]
 87%|████████▋ | 8534/9864 [35:55:12<5:18:25, 14.37s/it]
 87%|████████▋ | 8535/9864 [35:55:30<5:42:33, 15.47s/it]
 87%|████████▋ | 8536/9864 [35:55:44<5:34:44, 15.12s/it]
 87%|████████▋ | 8537/9864 [35:56:01<5:42:49, 15.50s/it]
 87%|████████▋ | 8538/9864 [35:56:17<5:47:01, 15.70s/it]
 87%|████████▋ | 8539/9864 [35:56:31<5:38:10, 15.31s/it]
 87%|████████▋ | 8540/9864 [35:56:46<5:31:08, 15.01s/it]
                                                        
{'loss': '0.05202', 'grad_norm': '4.552', 'learning_rate': '9.315e-07', 'epoch': '0.8658'}

 87%|████████▋ | 8540/9864 [35:56:46<5:31:08, 15.01s/it]
 87%|████████▋ | 8541/9864 [35:57:00<5:28:12, 14.88s/it]
 87%|████████▋ | 8542/9864 [35:57:16<5:33:31, 15.14s/it]
 87%|████████▋ | 8543/9864 [35:57:32<5:40:04, 15.45s/it]
 87%|████████▋ | 8544/9864 [35:57:47<5:34:38, 15.21s/it]
 87%|████████▋ | 8545/9864 [35:58:04<5:48:03, 15.83s/it]
 87%|████████▋ | 8546/9864 [35:58:18<5:38:01, 15.39s/it]
 87%|████████▋ | 8547/9864 [35:58:32<5:28:25, 14.96s/it]
 87%|████████▋ | 8548/9864 [35:58:47<5:24:25, 14.79s/it]
 87%|████████▋ | 8549/9864 [35:59:03<5:32:32, 15.17s/it]
 87%|████████▋ | 8550/9864 [35:59:17<5:25:45, 14.88s/it]
                                                        
{'loss': '0.05595', 'grad_norm': '4.904', 'learning_rate': '9.177e-07', 'epoch': '0.8669'}

 87%|████████▋ | 8550/9864 [35:59:17<5:25:45, 14.88s/it]
 87%|████████▋ | 8551/9864 [35:59:31<5:22:09, 14.72s/it]
 87%|████████▋ | 8552/9864 [35:59:47<5:31:00, 15.14s/it]
 87%|████████▋ | 8553/9864 [36:00:03<5:31:19, 15.16s/it]
 87%|████████▋ | 8554/9864 [36:00:18<5:34:12, 15.31s/it]
 87%|████████▋ | 8555/9864 [36:00:33<5:29:01, 15.08s/it]
 87%|████████▋ | 8556/9864 [36:00:47<5:24:10, 14.87s/it]
 87%|████████▋ | 8557/9864 [36:01:03<5:31:49, 15.23s/it]
 87%|████████▋ | 8558/9864 [36:01:18<5:25:36, 14.96s/it]
 87%|████████▋ | 8559/9864 [36:01:34<5:31:15, 15.23s/it]
 87%|████████▋ | 8560/9864 [36:01:48<5:28:13, 15.10s/it]
                                                        
{'loss': '0.06602', 'grad_norm': '6.881', 'learning_rate': '9.041e-07', 'epoch': '0.8679'}

 87%|████████▋ | 8560/9864 [36:01:48<5:28:13, 15.10s/it]
 87%|████████▋ | 8561/9864 [36:02:04<5:34:04, 15.38s/it]
 87%|████████▋ | 8562/9864 [36:02:20<5:37:39, 15.56s/it]
 87%|████████▋ | 8563/9864 [36:02:35<5:28:52, 15.17s/it]
 87%|████████▋ | 8564/9864 [36:02:51<5:34:58, 15.46s/it]
 87%|████████▋ | 8565/9864 [36:03:05<5:26:10, 15.07s/it]
 87%|████████▋ | 8566/9864 [36:03:20<5:24:36, 15.00s/it]
 87%|████████▋ | 8567/9864 [36:03:36<5:29:39, 15.25s/it]
 87%|████████▋ | 8568/9864 [36:03:51<5:31:36, 15.35s/it]
 87%|████████▋ | 8569/9864 [36:04:07<5:35:50, 15.56s/it]
 87%|████████▋ | 8570/9864 [36:04:22<5:27:38, 15.19s/it]
                                                        
{'loss': '0.05389', 'grad_norm': '5.516', 'learning_rate': '8.905e-07', 'epoch': '0.8689'}

 87%|████████▋ | 8570/9864 [36:04:22<5:27:38, 15.19s/it]
 87%|████████▋ | 8571/9864 [36:04:37<5:31:04, 15.36s/it]
 87%|████████▋ | 8572/9864 [36:04:51<5:21:27, 14.93s/it]
 87%|████████▋ | 8573/9864 [36:05:05<5:15:38, 14.67s/it]
 87%|████████▋ | 8574/9864 [36:05:20<5:15:05, 14.66s/it]
 87%|████████▋ | 8575/9864 [36:05:36<5:24:10, 15.09s/it]
 87%|████████▋ | 8576/9864 [36:05:53<5:36:04, 15.66s/it]
 87%|████████▋ | 8577/9864 [36:06:09<5:38:27, 15.78s/it]
 87%|████████▋ | 8578/9864 [36:06:23<5:24:19, 15.13s/it]
 87%|████████▋ | 8579/9864 [36:06:38<5:27:59, 15.31s/it]
 87%|████████▋ | 8580/9864 [36:06:53<5:24:10, 15.15s/it]
                                                        
{'loss': '0.05996', 'grad_norm': '4.694', 'learning_rate': '8.77e-07', 'epoch': '0.8699'}

 87%|████████▋ | 8580/9864 [36:06:53<5:24:10, 15.15s/it]
 87%|████████▋ | 8581/9864 [36:07:08<5:19:27, 14.94s/it]
 87%|████████▋ | 8582/9864 [36:07:22<5:17:43, 14.87s/it]
 87%|████████▋ | 8583/9864 [36:07:40<5:37:36, 15.81s/it]
 87%|████████▋ | 8584/9864 [36:07:55<5:30:59, 15.52s/it]
 87%|████████▋ | 8585/9864 [36:08:12<5:36:41, 15.80s/it]
 87%|████████▋ | 8586/9864 [36:08:26<5:26:52, 15.35s/it]
 87%|████████▋ | 8587/9864 [36:08:40<5:19:18, 15.00s/it]
 87%|████████▋ | 8588/9864 [36:08:56<5:25:39, 15.31s/it]
 87%|████████▋ | 8589/9864 [36:09:10<5:18:52, 15.01s/it]
 87%|████████▋ | 8590/9864 [36:09:27<5:28:05, 15.45s/it]
                                                        
{'loss': '0.06062', 'grad_norm': '4.845', 'learning_rate': '8.636e-07', 'epoch': '0.8709'}

 87%|████████▋ | 8590/9864 [36:09:27<5:28:05, 15.45s/it]
 87%|████████▋ | 8591/9864 [36:09:43<5:32:18, 15.66s/it]
 87%|████████▋ | 8592/9864 [36:09:59<5:34:51, 15.80s/it]
 87%|████████▋ | 8593/9864 [36:10:14<5:26:50, 15.43s/it]
 87%|████████▋ | 8594/9864 [36:10:28<5:19:11, 15.08s/it]
 87%|████████▋ | 8595/9864 [36:10:42<5:14:16, 14.86s/it]
 87%|████████▋ | 8596/9864 [36:10:59<5:24:28, 15.35s/it]
 87%|████████▋ | 8597/9864 [36:11:14<5:24:26, 15.36s/it]
 87%|████████▋ | 8598/9864 [36:11:29<5:16:57, 15.02s/it]
 87%|████████▋ | 8599/9864 [36:11:46<5:31:35, 15.73s/it]
 87%|████████▋ | 8600/9864 [36:12:00<5:23:23, 15.35s/it]
                                                        
{'loss': '0.04872', 'grad_norm': '2.581', 'learning_rate': '8.503e-07', 'epoch': '0.8719'}

 87%|████████▋ | 8600/9864 [36:12:00<5:23:23, 15.35s/it]
 87%|████████▋ | 8601/9864 [36:12:14<5:10:51, 14.77s/it]
 87%|████████▋ | 8602/9864 [36:12:28<5:07:56, 14.64s/it]
 87%|████████▋ | 8603/9864 [36:12:44<5:15:19, 15.00s/it]
 87%|████████▋ | 8604/9864 [36:12:58<5:11:07, 14.82s/it]
 87%|████████▋ | 8605/9864 [36:13:12<5:04:48, 14.53s/it]
 87%|████████▋ | 8606/9864 [36:13:29<5:16:07, 15.08s/it]
 87%|████████▋ | 8607/9864 [36:13:47<5:34:00, 15.94s/it]
 87%|████████▋ | 8608/9864 [36:14:00<5:20:24, 15.31s/it]
 87%|████████▋ | 8609/9864 [36:14:15<5:15:47, 15.10s/it]
 87%|████████▋ | 8610/9864 [36:14:29<5:09:00, 14.79s/it]
                                                        
{'loss': '0.05746', 'grad_norm': '4.11', 'learning_rate': '8.371e-07', 'epoch': '0.8729'}

 87%|████████▋ | 8610/9864 [36:14:29<5:09:00, 14.79s/it]
 87%|████████▋ | 8611/9864 [36:14:45<5:18:33, 15.25s/it]
 87%|████████▋ | 8612/9864 [36:15:00<5:13:00, 15.00s/it]
 87%|████████▋ | 8613/9864 [36:15:16<5:18:50, 15.29s/it]
 87%|████████▋ | 8614/9864 [36:15:32<5:21:53, 15.45s/it]
 87%|████████▋ | 8615/9864 [36:15:48<5:25:27, 15.63s/it]
 87%|████████▋ | 8616/9864 [36:16:02<5:17:25, 15.26s/it]
 87%|████████▋ | 8617/9864 [36:16:17<5:15:27, 15.18s/it]
 87%|████████▋ | 8618/9864 [36:16:31<5:10:07, 14.93s/it]
 87%|████████▋ | 8619/9864 [36:16:46<5:05:03, 14.70s/it]
 87%|████████▋ | 8620/9864 [36:17:02<5:14:43, 15.18s/it]
                                                        
{'loss': '0.05633', 'grad_norm': '4.186', 'learning_rate': '8.24e-07', 'epoch': '0.874'}

 87%|████████▋ | 8620/9864 [36:17:02<5:14:43, 15.18s/it]
 87%|████████▋ | 8621/9864 [36:17:20<5:30:59, 15.98s/it]
 87%|████████▋ | 8622/9864 [36:17:34<5:18:32, 15.39s/it]
 87%|████████▋ | 8623/9864 [36:17:49<5:21:03, 15.52s/it]
 87%|████████▋ | 8624/9864 [36:18:04<5:13:12, 15.16s/it]
 87%|████████▋ | 8625/9864 [36:18:18<5:07:52, 14.91s/it]
 87%|████████▋ | 8626/9864 [36:18:32<5:00:38, 14.57s/it]
 87%|████████▋ | 8627/9864 [36:18:46<4:57:43, 14.44s/it]
 87%|████████▋ | 8628/9864 [36:19:00<4:51:43, 14.16s/it]
 87%|████████▋ | 8629/9864 [36:19:17<5:12:27, 15.18s/it]
 87%|████████▋ | 8630/9864 [36:19:32<5:11:37, 15.15s/it]
                                                        
{'loss': '0.0556', 'grad_norm': '3.718', 'learning_rate': '8.11e-07', 'epoch': '0.875'}

 87%|████████▋ | 8630/9864 [36:19:32<5:11:37, 15.15s/it]
 88%|████████▊ | 8631/9864 [36:19:48<5:13:39, 15.26s/it]
 88%|████████▊ | 8632/9864 [36:20:02<5:07:45, 14.99s/it]
 88%|████████▊ | 8633/9864 [36:20:16<5:01:23, 14.69s/it]
 88%|████████▊ | 8634/9864 [36:20:31<5:00:10, 14.64s/it]
 88%|████████▊ | 8635/9864 [36:20:45<4:56:58, 14.50s/it]
 88%|████████▊ | 8636/9864 [36:21:01<5:06:30, 14.98s/it]
 88%|████████▊ | 8637/9864 [36:21:16<5:09:11, 15.12s/it]
 88%|████████▊ | 8638/9864 [36:21:33<5:18:22, 15.58s/it]
 88%|████████▊ | 8639/9864 [36:21:49<5:22:38, 15.80s/it]
 88%|████████▊ | 8640/9864 [36:22:04<5:13:22, 15.36s/it]
                                                        
{'loss': '0.06206', 'grad_norm': '5.364', 'learning_rate': '7.981e-07', 'epoch': '0.876'}

 88%|████████▊ | 8640/9864 [36:22:04<5:13:22, 15.36s/it]
 88%|████████▊ | 8641/9864 [36:22:18<5:06:44, 15.05s/it]
 88%|████████▊ | 8642/9864 [36:22:32<5:00:57, 14.78s/it]
 88%|████████▊ | 8643/9864 [36:22:46<4:58:11, 14.65s/it]
 88%|████████▊ | 8644/9864 [36:23:01<4:55:37, 14.54s/it]
 88%|████████▊ | 8645/9864 [36:23:19<5:15:47, 15.54s/it]
 88%|████████▊ | 8646/9864 [36:23:35<5:18:10, 15.67s/it]
 88%|████████▊ | 8647/9864 [36:23:51<5:22:16, 15.89s/it]
 88%|████████▊ | 8648/9864 [36:24:06<5:14:43, 15.53s/it]
 88%|████████▊ | 8649/9864 [36:24:20<5:07:26, 15.18s/it]
 88%|████████▊ | 8650/9864 [36:24:34<4:57:27, 14.70s/it]
                                                        
{'loss': '0.04831', 'grad_norm': '3.568', 'learning_rate': '7.853e-07', 'epoch': '0.877'}

 88%|████████▊ | 8650/9864 [36:24:34<4:57:27, 14.70s/it]
 88%|████████▊ | 8651/9864 [36:24:48<4:54:18, 14.56s/it]
 88%|████████▊ | 8652/9864 [36:25:03<4:55:32, 14.63s/it]
 88%|████████▊ | 8653/9864 [36:25:21<5:16:20, 15.67s/it]
 88%|████████▊ | 8654/9864 [36:25:37<5:18:53, 15.81s/it]
 88%|████████▊ | 8655/9864 [36:25:51<5:08:15, 15.30s/it]
 88%|████████▊ | 8656/9864 [36:26:06<5:07:55, 15.29s/it]
 88%|████████▊ | 8657/9864 [36:26:20<5:00:34, 14.94s/it]
 88%|████████▊ | 8658/9864 [36:26:34<4:55:04, 14.68s/it]
 88%|████████▊ | 8659/9864 [36:26:48<4:50:39, 14.47s/it]
 88%|████████▊ | 8660/9864 [36:27:05<5:02:58, 15.10s/it]
                                                        
{'loss': '0.04539', 'grad_norm': '4.096', 'learning_rate': '7.726e-07', 'epoch': '0.878'}

 88%|████████▊ | 8660/9864 [36:27:05<5:02:58, 15.10s/it]
 88%|████████▊ | 8661/9864 [36:27:21<5:08:08, 15.37s/it]
 88%|████████▊ | 8662/9864 [36:27:36<5:03:07, 15.13s/it]
 88%|████████▊ | 8663/9864 [36:27:53<5:18:37, 15.92s/it]
 88%|████████▊ | 8664/9864 [36:28:08<5:12:06, 15.61s/it]
 88%|████████▊ | 8665/9864 [36:28:22<5:01:55, 15.11s/it]
 88%|████████▊ | 8666/9864 [36:28:37<4:58:11, 14.93s/it]
 88%|████████▊ | 8667/9864 [36:28:54<5:09:20, 15.51s/it]
 88%|████████▊ | 8668/9864 [36:29:10<5:13:18, 15.72s/it]
 88%|████████▊ | 8669/9864 [36:29:24<5:03:01, 15.21s/it]
 88%|████████▊ | 8670/9864 [36:29:40<5:07:37, 15.46s/it]
                                                        
{'loss': '0.05647', 'grad_norm': '3.464', 'learning_rate': '7.599e-07', 'epoch': '0.879'}

 88%|████████▊ | 8670/9864 [36:29:40<5:07:37, 15.46s/it]
 88%|████████▊ | 8671/9864 [36:29:56<5:10:46, 15.63s/it]
 88%|████████▊ | 8672/9864 [36:30:10<5:04:06, 15.31s/it]
 88%|████████▊ | 8673/9864 [36:30:25<4:58:40, 15.05s/it]
 88%|████████▊ | 8674/9864 [36:30:39<4:54:03, 14.83s/it]
 88%|████████▊ | 8675/9864 [36:30:57<5:10:37, 15.67s/it]
 88%|████████▊ | 8676/9864 [36:31:11<5:02:35, 15.28s/it]
 88%|████████▊ | 8677/9864 [36:31:25<4:56:37, 14.99s/it]
 88%|████████▊ | 8678/9864 [36:31:39<4:49:30, 14.65s/it]
 88%|████████▊ | 8679/9864 [36:31:57<5:09:16, 15.66s/it]
 88%|████████▊ | 8680/9864 [36:32:12<5:00:17, 15.22s/it]
                                                        
{'loss': '0.05772', 'grad_norm': '4.052', 'learning_rate': '7.474e-07', 'epoch': '0.88'}

 88%|████████▊ | 8680/9864 [36:32:12<5:00:17, 15.22s/it]
 88%|████████▊ | 8681/9864 [36:32:26<4:53:20, 14.88s/it]
 88%|████████▊ | 8682/9864 [36:32:40<4:49:07, 14.68s/it]
 88%|████████▊ | 8683/9864 [36:32:55<4:54:41, 14.97s/it]
 88%|████████▊ | 8684/9864 [36:33:12<5:00:54, 15.30s/it]
 88%|████████▊ | 8685/9864 [36:33:26<4:54:25, 14.98s/it]
 88%|████████▊ | 8686/9864 [36:33:40<4:49:05, 14.72s/it]
 88%|████████▊ | 8687/9864 [36:33:59<5:13:13, 15.97s/it]
 88%|████████▊ | 8688/9864 [36:34:13<5:01:59, 15.41s/it]
 88%|████████▊ | 8689/9864 [36:34:27<4:55:21, 15.08s/it]
 88%|████████▊ | 8690/9864 [36:34:42<4:51:47, 14.91s/it]
                                                        
{'loss': '0.05174', 'grad_norm': '3.914', 'learning_rate': '7.35e-07', 'epoch': '0.8811'}

 88%|████████▊ | 8690/9864 [36:34:42<4:51:47, 14.91s/it]
 88%|████████▊ | 8691/9864 [36:34:58<5:00:59, 15.40s/it]
 88%|████████▊ | 8692/9864 [36:35:15<5:06:04, 15.67s/it]
 88%|████████▊ | 8693/9864 [36:35:30<5:01:40, 15.46s/it]
 88%|████████▊ | 8694/9864 [36:35:46<5:06:46, 15.73s/it]
 88%|████████▊ | 8695/9864 [36:36:02<5:10:41, 15.95s/it]
 88%|████████▊ | 8696/9864 [36:36:17<5:00:52, 15.46s/it]
 88%|████████▊ | 8697/9864 [36:36:31<4:54:14, 15.13s/it]
 88%|████████▊ | 8698/9864 [36:36:47<4:59:03, 15.39s/it]
 88%|████████▊ | 8699/9864 [36:37:01<4:50:07, 14.94s/it]
 88%|████████▊ | 8700/9864 [36:37:16<4:48:32, 14.87s/it]
                                                        
{'loss': '0.06176', 'grad_norm': '3.885', 'learning_rate': '7.227e-07', 'epoch': '0.8821'}

 88%|████████▊ | 8700/9864 [36:37:16<4:48:32, 14.87s/it]
 88%|████████▊ | 8701/9864 [36:37:33<5:02:54, 15.63s/it]
 88%|████████▊ | 8702/9864 [36:37:47<4:55:12, 15.24s/it]
 88%|████████▊ | 8703/9864 [36:38:03<4:58:03, 15.40s/it]
 88%|████████▊ | 8704/9864 [36:38:17<4:47:39, 14.88s/it]
 88%|████████▊ | 8705/9864 [36:38:32<4:51:05, 15.07s/it]
 88%|████████▊ | 8706/9864 [36:38:47<4:47:13, 14.88s/it]
 88%|████████▊ | 8707/9864 [36:39:01<4:45:45, 14.82s/it]
 88%|████████▊ | 8708/9864 [36:39:17<4:51:02, 15.11s/it]
 88%|████████▊ | 8709/9864 [36:39:33<4:56:40, 15.41s/it]
 88%|████████▊ | 8710/9864 [36:39:47<4:49:06, 15.03s/it]
                                                        
{'loss': '0.04546', 'grad_norm': '3.534', 'learning_rate': '7.105e-07', 'epoch': '0.8831'}

 88%|████████▊ | 8710/9864 [36:39:47<4:49:06, 15.03s/it]
 88%|████████▊ | 8711/9864 [36:40:02<4:46:22, 14.90s/it]
 88%|████████▊ | 8712/9864 [36:40:18<4:53:12, 15.27s/it]
 88%|████████▊ | 8713/9864 [36:40:32<4:47:14, 14.97s/it]
 88%|████████▊ | 8714/9864 [36:40:49<4:54:36, 15.37s/it]
 88%|████████▊ | 8715/9864 [36:41:05<4:57:28, 15.53s/it]
 88%|████████▊ | 8716/9864 [36:41:21<4:59:10, 15.64s/it]
 88%|████████▊ | 8717/9864 [36:41:35<4:49:46, 15.16s/it]
 88%|████████▊ | 8718/9864 [36:41:49<4:47:32, 15.05s/it]
 88%|████████▊ | 8719/9864 [36:42:04<4:45:21, 14.95s/it]
 88%|████████▊ | 8720/9864 [36:42:20<4:49:54, 15.20s/it]
                                                        
{'loss': '0.0562', 'grad_norm': '3.974', 'learning_rate': '6.984e-07', 'epoch': '0.8841'}

 88%|████████▊ | 8720/9864 [36:42:20<4:49:54, 15.20s/it]
 88%|████████▊ | 8721/9864 [36:42:34<4:42:29, 14.83s/it]
 88%|████████▊ | 8722/9864 [36:42:50<4:48:23, 15.15s/it]
 88%|████████▊ | 8723/9864 [36:43:06<4:53:11, 15.42s/it]
 88%|████████▊ | 8724/9864 [36:43:22<4:58:50, 15.73s/it]
 88%|████████▊ | 8725/9864 [36:43:36<4:47:26, 15.14s/it]
 88%|████████▊ | 8726/9864 [36:43:50<4:41:36, 14.85s/it]
 88%|████████▊ | 8727/9864 [36:44:05<4:41:29, 14.85s/it]
 88%|████████▊ | 8728/9864 [36:44:22<4:50:41, 15.35s/it]
 88%|████████▊ | 8729/9864 [36:44:37<4:49:32, 15.31s/it]
 89%|████████▊ | 8730/9864 [36:44:55<5:03:19, 16.05s/it]
                                                        
{'loss': '0.03583', 'grad_norm': '2.808', 'learning_rate': '6.864e-07', 'epoch': '0.8851'}

 89%|████████▊ | 8730/9864 [36:44:55<5:03:19, 16.05s/it]
 89%|████████▊ | 8731/9864 [36:45:09<4:54:45, 15.61s/it]
 89%|████████▊ | 8732/9864 [36:45:25<4:55:56, 15.69s/it]
 89%|████████▊ | 8733/9864 [36:45:39<4:46:39, 15.21s/it]
 89%|████████▊ | 8734/9864 [36:45:53<4:41:38, 14.95s/it]
 89%|████████▊ | 8735/9864 [36:46:08<4:36:27, 14.69s/it]
 89%|████████▊ | 8736/9864 [36:46:24<4:43:30, 15.08s/it]
 89%|████████▊ | 8737/9864 [36:46:38<4:38:21, 14.82s/it]
 89%|████████▊ | 8738/9864 [36:46:56<4:56:50, 15.82s/it]
 89%|████████▊ | 8739/9864 [36:47:10<4:48:47, 15.40s/it]
 89%|████████▊ | 8740/9864 [36:47:26<4:51:10, 15.54s/it]
                                                        
{'loss': '0.06085', 'grad_norm': '3.992', 'learning_rate': '6.745e-07', 'epoch': '0.8861'}

 89%|████████▊ | 8740/9864 [36:47:26<4:51:10, 15.54s/it]
 89%|████████▊ | 8741/9864 [36:47:40<4:43:44, 15.16s/it]
 89%|████████▊ | 8742/9864 [36:47:54<4:36:27, 14.78s/it]
 89%|████████▊ | 8743/9864 [36:48:08<4:32:17, 14.57s/it]
 89%|████████▊ | 8744/9864 [36:48:23<4:31:45, 14.56s/it]
 89%|████████▊ | 8745/9864 [36:48:41<4:48:13, 15.45s/it]
 89%|████████▊ | 8746/9864 [36:48:57<4:52:03, 15.67s/it]
 89%|████████▊ | 8747/9864 [36:49:13<4:58:01, 16.01s/it]
 89%|████████▊ | 8748/9864 [36:49:27<4:45:45, 15.36s/it]
 89%|████████▊ | 8749/9864 [36:49:41<4:37:36, 14.94s/it]
 89%|████████▊ | 8750/9864 [36:49:56<4:34:11, 14.77s/it]
                                                        
{'loss': '0.06207', 'grad_norm': '5.402', 'learning_rate': '6.627e-07', 'epoch': '0.8871'}

 89%|████████▊ | 8750/9864 [36:49:56<4:34:11, 14.77s/it]
 89%|████████▊ | 8751/9864 [36:50:10<4:31:36, 14.64s/it]
 89%|████████▊ | 8752/9864 [36:50:26<4:40:16, 15.12s/it]
 89%|████████▊ | 8753/9864 [36:50:44<4:53:15, 15.84s/it]
 89%|████████▊ | 8754/9864 [36:50:58<4:44:43, 15.39s/it]
 89%|████████▉ | 8755/9864 [36:51:14<4:48:59, 15.64s/it]
 89%|████████▉ | 8756/9864 [36:51:29<4:42:56, 15.32s/it]
 89%|████████▉ | 8757/9864 [36:51:43<4:35:38, 14.94s/it]
 89%|████████▉ | 8758/9864 [36:51:57<4:32:52, 14.80s/it]
 89%|████████▉ | 8759/9864 [36:52:14<4:42:32, 15.34s/it]
 89%|████████▉ | 8760/9864 [36:52:28<4:36:03, 15.00s/it]
                                                        
{'loss': '0.062', 'grad_norm': '3.962', 'learning_rate': '6.51e-07', 'epoch': '0.8882'}

 89%|████████▉ | 8760/9864 [36:52:28<4:36:03, 15.00s/it]
 89%|████████▉ | 8761/9864 [36:52:46<4:51:17, 15.85s/it]
 89%|████████▉ | 8762/9864 [36:53:01<4:44:13, 15.48s/it]
 89%|████████▉ | 8763/9864 [36:53:17<4:48:22, 15.72s/it]
 89%|████████▉ | 8764/9864 [36:53:31<4:39:21, 15.24s/it]
 89%|████████▉ | 8765/9864 [36:53:45<4:33:46, 14.95s/it]
 89%|████████▉ | 8766/9864 [36:54:02<4:41:09, 15.36s/it]
 89%|████████▉ | 8767/9864 [36:54:16<4:36:56, 15.15s/it]
 89%|████████▉ | 8768/9864 [36:54:32<4:39:58, 15.33s/it]
 89%|████████▉ | 8769/9864 [36:54:46<4:30:34, 14.83s/it]
 89%|████████▉ | 8770/9864 [36:55:03<4:45:30, 15.66s/it]
                                                        
{'loss': '0.04756', 'grad_norm': '8.907', 'learning_rate': '6.394e-07', 'epoch': '0.8892'}

 89%|████████▉ | 8770/9864 [36:55:03<4:45:30, 15.66s/it]
 89%|████████▉ | 8771/9864 [36:55:17<4:37:05, 15.21s/it]
 89%|████████▉ | 8772/9864 [36:55:32<4:33:04, 15.00s/it]
 89%|████████▉ | 8773/9864 [36:55:48<4:39:37, 15.38s/it]
 89%|████████▉ | 8774/9864 [36:56:03<4:36:37, 15.23s/it]
 89%|████████▉ | 8775/9864 [36:56:19<4:41:20, 15.50s/it]
 89%|████████▉ | 8776/9864 [36:56:33<4:33:39, 15.09s/it]
 89%|████████▉ | 8777/9864 [36:56:49<4:38:47, 15.39s/it]
 89%|████████▉ | 8778/9864 [36:57:05<4:41:27, 15.55s/it]
 89%|████████▉ | 8779/9864 [36:57:20<4:33:13, 15.11s/it]
 89%|████████▉ | 8780/9864 [36:57:34<4:27:51, 14.83s/it]
                                                        
{'loss': '0.05746', 'grad_norm': '3.591', 'learning_rate': '6.279e-07', 'epoch': '0.8902'}

 89%|████████▉ | 8780/9864 [36:57:34<4:27:51, 14.83s/it]
 89%|████████▉ | 8781/9864 [36:57:49<4:32:44, 15.11s/it]
 89%|████████▉ | 8782/9864 [36:58:06<4:39:33, 15.50s/it]
 89%|████████▉ | 8783/9864 [36:58:20<4:29:32, 14.96s/it]
 89%|████████▉ | 8784/9864 [36:58:34<4:23:50, 14.66s/it]
 89%|████████▉ | 8785/9864 [36:58:50<4:30:50, 15.06s/it]
 89%|████████▉ | 8786/9864 [36:59:06<4:37:10, 15.43s/it]
 89%|████████▉ | 8787/9864 [36:59:20<4:30:40, 15.08s/it]
 89%|████████▉ | 8788/9864 [36:59:34<4:22:45, 14.65s/it]
 89%|████████▉ | 8789/9864 [36:59:48<4:22:10, 14.63s/it]
 89%|████████▉ | 8790/9864 [37:00:05<4:31:43, 15.18s/it]
                                                        
{'loss': '0.07091', 'grad_norm': '6.292', 'learning_rate': '6.165e-07', 'epoch': '0.8912'}

 89%|████████▉ | 8790/9864 [37:00:05<4:31:43, 15.18s/it]
 89%|████████▉ | 8791/9864 [37:00:21<4:38:04, 15.55s/it]
 89%|████████▉ | 8792/9864 [37:00:37<4:41:55, 15.78s/it]
 89%|████████▉ | 8793/9864 [37:00:51<4:32:07, 15.24s/it]
 89%|████████▉ | 8794/9864 [37:01:08<4:37:19, 15.55s/it]
 89%|████████▉ | 8795/9864 [37:01:22<4:30:07, 15.16s/it]
 89%|████████▉ | 8796/9864 [37:01:36<4:25:43, 14.93s/it]
 89%|████████▉ | 8797/9864 [37:01:51<4:23:03, 14.79s/it]
 89%|████████▉ | 8798/9864 [37:02:05<4:21:41, 14.73s/it]
 89%|████████▉ | 8799/9864 [37:02:23<4:37:07, 15.61s/it]
 89%|████████▉ | 8800/9864 [37:02:39<4:37:08, 15.63s/it]
                                                        
{'loss': '0.05637', 'grad_norm': '3.954', 'learning_rate': '6.052e-07', 'epoch': '0.8922'}

 89%|████████▉ | 8800/9864 [37:02:39<4:37:08, 15.63s/it]
 89%|████████▉ | 8801/9864 [37:02:53<4:31:12, 15.31s/it]
 89%|████████▉ | 8802/9864 [37:03:10<4:37:32, 15.68s/it]
 89%|████████▉ | 8803/9864 [37:03:24<4:30:23, 15.29s/it]
 89%|████████▉ | 8804/9864 [37:03:39<4:28:05, 15.17s/it]
 89%|████████▉ | 8805/9864 [37:03:53<4:22:57, 14.90s/it]
 89%|████████▉ | 8806/9864 [37:04:10<4:29:03, 15.26s/it]
 89%|████████▉ | 8807/9864 [37:04:28<4:43:21, 16.08s/it]
 89%|████████▉ | 8808/9864 [37:04:41<4:30:57, 15.40s/it]
 89%|████████▉ | 8809/9864 [37:04:56<4:26:18, 15.15s/it]
 89%|████████▉ | 8810/9864 [37:05:10<4:22:55, 14.97s/it]
                                                        
{'loss': '0.06218', 'grad_norm': '5.968', 'learning_rate': '5.94e-07', 'epoch': '0.8932'}

 89%|████████▉ | 8810/9864 [37:05:10<4:22:55, 14.97s/it]
 89%|████████▉ | 8811/9864 [37:05:27<4:30:30, 15.41s/it]
 89%|████████▉ | 8812/9864 [37:05:42<4:27:34, 15.26s/it]
 89%|████████▉ | 8813/9864 [37:05:58<4:30:50, 15.46s/it]
 89%|████████▉ | 8814/9864 [37:06:14<4:35:10, 15.72s/it]
 89%|████████▉ | 8815/9864 [37:06:28<4:23:19, 15.06s/it]
 89%|████████▉ | 8816/9864 [37:06:44<4:27:33, 15.32s/it]
 89%|████████▉ | 8817/9864 [37:06:58<4:22:26, 15.04s/it]
 89%|████████▉ | 8818/9864 [37:07:12<4:18:34, 14.83s/it]
 89%|████████▉ | 8819/9864 [37:07:26<4:14:35, 14.62s/it]
 89%|████████▉ | 8820/9864 [37:07:42<4:20:49, 14.99s/it]
                                                        
{'loss': '0.05582', 'grad_norm': '2.185', 'learning_rate': '5.829e-07', 'epoch': '0.8942'}

 89%|████████▉ | 8820/9864 [37:07:42<4:20:49, 14.99s/it]
 89%|████████▉ | 8821/9864 [37:08:00<4:35:03, 15.82s/it]
 89%|████████▉ | 8822/9864 [37:08:14<4:27:04, 15.38s/it]
 89%|████████▉ | 8823/9864 [37:08:30<4:30:07, 15.57s/it]
 89%|████████▉ | 8824/9864 [37:08:45<4:25:15, 15.30s/it]
 89%|████████▉ | 8825/9864 [37:09:00<4:21:26, 15.10s/it]
 89%|████████▉ | 8826/9864 [37:09:14<4:17:51, 14.91s/it]
 89%|████████▉ | 8827/9864 [37:09:29<4:16:41, 14.85s/it]
 89%|████████▉ | 8828/9864 [37:09:47<4:33:25, 15.84s/it]
 90%|████████▉ | 8829/9864 [37:10:03<4:35:37, 15.98s/it]
 90%|████████▉ | 8830/9864 [37:10:17<4:25:06, 15.38s/it]
                                                        
{'loss': '0.04762', 'grad_norm': '4.049', 'learning_rate': '5.719e-07', 'epoch': '0.8952'}

 90%|████████▉ | 8830/9864 [37:10:17<4:25:06, 15.38s/it]
 90%|████████▉ | 8831/9864 [37:10:34<4:29:42, 15.67s/it]
 90%|████████▉ | 8832/9864 [37:10:47<4:20:09, 15.13s/it]
 90%|████████▉ | 8833/9864 [37:11:02<4:17:14, 14.97s/it]
 90%|████████▉ | 8834/9864 [37:11:16<4:13:36, 14.77s/it]
 90%|████████▉ | 8835/9864 [37:11:32<4:18:29, 15.07s/it]
 90%|████████▉ | 8836/9864 [37:11:48<4:22:34, 15.33s/it]
 90%|████████▉ | 8837/9864 [37:12:04<4:27:02, 15.60s/it]
 90%|████████▉ | 8838/9864 [37:12:19<4:19:38, 15.18s/it]
 90%|████████▉ | 8839/9864 [37:12:35<4:24:13, 15.47s/it]
 90%|████████▉ | 8840/9864 [37:12:49<4:15:50, 14.99s/it]
                                                        
{'loss': '0.05344', 'grad_norm': '4.621', 'learning_rate': '5.61e-07', 'epoch': '0.8963'}

 90%|████████▉ | 8840/9864 [37:12:49<4:15:50, 14.99s/it]
 90%|████████▉ | 8841/9864 [37:13:03<4:13:42, 14.88s/it]
 90%|████████▉ | 8842/9864 [37:13:18<4:12:33, 14.83s/it]
 90%|████████▉ | 8843/9864 [37:13:36<4:28:27, 15.78s/it]
 90%|████████▉ | 8844/9864 [37:13:50<4:20:54, 15.35s/it]
 90%|████████▉ | 8845/9864 [37:14:06<4:24:23, 15.57s/it]
 90%|████████▉ | 8846/9864 [37:14:21<4:19:59, 15.32s/it]
 90%|████████▉ | 8847/9864 [37:14:37<4:21:56, 15.45s/it]
 90%|████████▉ | 8848/9864 [37:14:51<4:17:19, 15.20s/it]
 90%|████████▉ | 8849/9864 [37:15:06<4:11:38, 14.88s/it]
 90%|████████▉ | 8850/9864 [37:15:20<4:08:36, 14.71s/it]
                                                        
{'loss': '0.05039', 'grad_norm': '4.959', 'learning_rate': '5.502e-07', 'epoch': '0.8973'}

 90%|████████▉ | 8850/9864 [37:15:20<4:08:36, 14.71s/it]
 90%|████████▉ | 8851/9864 [37:15:38<4:25:01, 15.70s/it]
 90%|████████▉ | 8852/9864 [37:15:53<4:19:38, 15.39s/it]
 90%|████████▉ | 8853/9864 [37:16:09<4:22:25, 15.57s/it]
 90%|████████▉ | 8854/9864 [37:16:22<4:13:50, 15.08s/it]
 90%|████████▉ | 8855/9864 [37:16:38<4:16:24, 15.25s/it]
 90%|████████▉ | 8856/9864 [37:16:53<4:12:38, 15.04s/it]
 90%|████████▉ | 8857/9864 [37:17:07<4:09:45, 14.88s/it]
 90%|████████▉ | 8858/9864 [37:17:22<4:07:56, 14.79s/it]
 90%|████████▉ | 8859/9864 [37:17:39<4:19:03, 15.47s/it]
 90%|████████▉ | 8860/9864 [37:17:53<4:10:47, 14.99s/it]
                                                        
{'loss': '0.04536', 'grad_norm': '4.164', 'learning_rate': '5.395e-07', 'epoch': '0.8983'}

 90%|████████▉ | 8860/9864 [37:17:53<4:10:47, 14.99s/it]
 90%|████████▉ | 8861/9864 [37:18:08<4:14:51, 15.25s/it]
 90%|████████▉ | 8862/9864 [37:18:23<4:08:29, 14.88s/it]
 90%|████████▉ | 8863/9864 [37:18:38<4:13:00, 15.16s/it]
 90%|████████▉ | 8864/9864 [37:18:53<4:08:20, 14.90s/it]
 90%|████████▉ | 8865/9864 [37:19:07<4:05:25, 14.74s/it]
 90%|████████▉ | 8866/9864 [37:19:24<4:14:33, 15.30s/it]
 90%|████████▉ | 8867/9864 [37:19:40<4:17:34, 15.50s/it]
 90%|████████▉ | 8868/9864 [37:19:54<4:13:37, 15.28s/it]
 90%|████████▉ | 8869/9864 [37:20:10<4:16:35, 15.47s/it]
 90%|████████▉ | 8870/9864 [37:20:24<4:09:54, 15.09s/it]
                                                        
{'loss': '0.06126', 'grad_norm': '3.667', 'learning_rate': '5.289e-07', 'epoch': '0.8993'}

 90%|████████▉ | 8870/9864 [37:20:24<4:09:54, 15.09s/it]
 90%|████████▉ | 8871/9864 [37:20:40<4:11:24, 15.19s/it]
 90%|████████▉ | 8872/9864 [37:20:55<4:08:24, 15.03s/it]
 90%|████████▉ | 8873/9864 [37:21:10<4:12:44, 15.30s/it]
 90%|████████▉ | 8874/9864 [37:21:27<4:16:44, 15.56s/it]
 90%|████████▉ | 8875/9864 [37:21:41<4:12:04, 15.29s/it]
 90%|████████▉ | 8876/9864 [37:21:58<4:17:34, 15.64s/it]
 90%|████████▉ | 8877/9864 [37:22:12<4:11:40, 15.30s/it]
 90%|█████████ | 8878/9864 [37:22:29<4:17:34, 15.67s/it]
 90%|█████████ | 8879/9864 [37:22:44<4:13:24, 15.44s/it]
 90%|█████████ | 8880/9864 [37:22:58<4:07:35, 15.10s/it]
                                                        
{'loss': '0.06087', 'grad_norm': '4.082', 'learning_rate': '5.185e-07', 'epoch': '0.9003'}

 90%|█████████ | 8880/9864 [37:22:58<4:07:35, 15.10s/it]
 90%|█████████ | 8881/9864 [37:23:14<4:12:09, 15.39s/it]
 90%|█████████ | 8882/9864 [37:23:30<4:12:53, 15.45s/it]
 90%|█████████ | 8883/9864 [37:23:46<4:17:57, 15.78s/it]
 90%|█████████ | 8884/9864 [37:24:01<4:10:48, 15.36s/it]
 90%|█████████ | 8885/9864 [37:24:16<4:08:36, 15.24s/it]
 90%|█████████ | 8886/9864 [37:24:31<4:10:25, 15.36s/it]
 90%|█████████ | 8887/9864 [37:24:46<4:05:58, 15.11s/it]
 90%|█████████ | 8888/9864 [37:25:01<4:05:33, 15.10s/it]
 90%|█████████ | 8889/9864 [37:25:14<3:57:52, 14.64s/it]
 90%|█████████ | 8890/9864 [37:25:30<4:02:20, 14.93s/it]
                                                        
{'loss': '0.04824', 'grad_norm': '4.892', 'learning_rate': '5.081e-07', 'epoch': '0.9013'}

 90%|█████████ | 8890/9864 [37:25:30<4:02:20, 14.93s/it]
 90%|█████████ | 8891/9864 [37:25:46<4:08:28, 15.32s/it]
 90%|█████████ | 8892/9864 [37:26:00<4:02:19, 14.96s/it]
 90%|█████████ | 8893/9864 [37:26:16<4:04:02, 15.08s/it]
 90%|█████████ | 8894/9864 [37:26:30<4:01:08, 14.92s/it]
 90%|█████████ | 8895/9864 [37:26:47<4:08:29, 15.39s/it]
 90%|█████████ | 8896/9864 [37:27:01<4:02:39, 15.04s/it]
 90%|█████████ | 8897/9864 [37:27:15<4:00:01, 14.89s/it]
 90%|█████████ | 8898/9864 [37:27:33<4:13:37, 15.75s/it]
 90%|█████████ | 8899/9864 [37:27:48<4:07:35, 15.39s/it]
 90%|█████████ | 8900/9864 [37:28:02<4:00:26, 14.97s/it]
                                                        
{'loss': '0.05189', 'grad_norm': '3.656', 'learning_rate': '4.978e-07', 'epoch': '0.9023'}

 90%|█████████ | 8900/9864 [37:28:02<4:00:26, 14.97s/it]
 90%|█████████ | 8901/9864 [37:28:16<3:57:22, 14.79s/it]
 90%|█████████ | 8902/9864 [37:28:33<4:06:11, 15.35s/it]
 90%|█████████ | 8903/9864 [37:28:47<4:00:04, 14.99s/it]
 90%|█████████ | 8904/9864 [37:29:03<4:03:41, 15.23s/it]
 90%|█████████ | 8905/9864 [37:29:17<3:56:40, 14.81s/it]
 90%|█████████ | 8906/9864 [37:29:34<4:10:15, 15.67s/it]
 90%|█████████ | 8907/9864 [37:29:48<4:01:53, 15.17s/it]
 90%|█████████ | 8908/9864 [37:30:02<3:54:51, 14.74s/it]
 90%|█████████ | 8909/9864 [37:30:16<3:53:28, 14.67s/it]
 90%|█████████ | 8910/9864 [37:30:32<3:57:39, 14.95s/it]
                                                        
{'loss': '0.0557', 'grad_norm': '3.654', 'learning_rate': '4.876e-07', 'epoch': '0.9034'}

 90%|█████████ | 8910/9864 [37:30:32<3:57:39, 14.95s/it]
 90%|█████████ | 8911/9864 [37:30:46<3:54:09, 14.74s/it]
 90%|█████████ | 8912/9864 [37:31:03<4:02:13, 15.27s/it]
 90%|█████████ | 8913/9864 [37:31:17<3:56:25, 14.92s/it]
 90%|█████████ | 8914/9864 [37:31:33<4:00:02, 15.16s/it]
 90%|█████████ | 8915/9864 [37:31:49<4:05:43, 15.54s/it]
 90%|█████████ | 8916/9864 [37:32:03<3:59:45, 15.17s/it]
 90%|█████████ | 8917/9864 [37:32:18<3:54:33, 14.86s/it]
 90%|█████████ | 8918/9864 [37:32:34<4:02:56, 15.41s/it]
 90%|█████████ | 8919/9864 [37:32:50<4:04:04, 15.50s/it]
 90%|█████████ | 8920/9864 [37:33:05<4:00:45, 15.30s/it]
                                                        
{'loss': '0.04953', 'grad_norm': '4.415', 'learning_rate': '4.775e-07', 'epoch': '0.9044'}

 90%|█████████ | 8920/9864 [37:33:05<4:00:45, 15.30s/it]
 90%|█████████ | 8921/9864 [37:33:19<3:55:25, 14.98s/it]
 90%|█████████ | 8922/9864 [37:33:35<4:00:25, 15.31s/it]
 90%|█████████ | 8923/9864 [37:33:51<4:01:28, 15.40s/it]
 90%|█████████ | 8924/9864 [37:34:05<3:55:58, 15.06s/it]
 90%|█████████ | 8925/9864 [37:34:21<4:01:15, 15.42s/it]
 90%|█████████ | 8926/9864 [37:34:35<3:55:02, 15.03s/it]
 91%|█████████ | 8927/9864 [37:34:50<3:53:49, 14.97s/it]
 91%|█████████ | 8928/9864 [37:35:05<3:51:41, 14.85s/it]
 91%|█████████ | 8929/9864 [37:35:21<3:57:25, 15.24s/it]
 91%|█████████ | 8930/9864 [37:35:35<3:51:22, 14.86s/it]
                                                        
{'loss': '0.0513', 'grad_norm': '3.363', 'learning_rate': '4.676e-07', 'epoch': '0.9054'}

 91%|█████████ | 8930/9864 [37:35:35<3:51:22, 14.86s/it]
 91%|█████████ | 8931/9864 [37:35:50<3:54:37, 15.09s/it]
 91%|█████████ | 8932/9864 [37:36:05<3:51:16, 14.89s/it]
 91%|█████████ | 8933/9864 [37:36:19<3:46:11, 14.58s/it]
 91%|█████████ | 8934/9864 [37:36:35<3:52:32, 15.00s/it]
 91%|█████████ | 8935/9864 [37:36:51<3:57:31, 15.34s/it]
 91%|█████████ | 8936/9864 [37:37:07<3:59:52, 15.51s/it]
 91%|█████████ | 8937/9864 [37:37:21<3:55:28, 15.24s/it]
 91%|█████████ | 8938/9864 [37:37:35<3:48:33, 14.81s/it]
 91%|█████████ | 8939/9864 [37:37:49<3:45:55, 14.65s/it]
 91%|█████████ | 8940/9864 [37:38:06<3:52:03, 15.07s/it]
                                                        
{'loss': '0.06936', 'grad_norm': '4.058', 'learning_rate': '4.577e-07', 'epoch': '0.9064'}

 91%|█████████ | 8940/9864 [37:38:06<3:52:03, 15.07s/it]
 91%|█████████ | 8941/9864 [37:38:20<3:49:14, 14.90s/it]
 91%|█████████ | 8942/9864 [37:38:37<3:56:31, 15.39s/it]
 91%|█████████ | 8943/9864 [37:38:52<3:58:36, 15.54s/it]
 91%|█████████ | 8944/9864 [37:39:06<3:50:49, 15.05s/it]
 91%|█████████ | 8945/9864 [37:39:23<3:56:38, 15.45s/it]
 91%|█████████ | 8946/9864 [37:39:37<3:52:19, 15.18s/it]
 91%|█████████ | 8947/9864 [37:39:52<3:47:56, 14.91s/it]
 91%|█████████ | 8948/9864 [37:40:07<3:52:04, 15.20s/it]
 91%|█████████ | 8949/9864 [37:40:22<3:47:06, 14.89s/it]
 91%|█████████ | 8950/9864 [37:40:36<3:45:56, 14.83s/it]
                                                        
{'loss': '0.04557', 'grad_norm': '3.23', 'learning_rate': '4.479e-07', 'epoch': '0.9074'}

 91%|█████████ | 8950/9864 [37:40:36<3:45:56, 14.83s/it]
 91%|█████████ | 8951/9864 [37:40:54<3:58:58, 15.71s/it]
 91%|█████████ | 8952/9864 [37:41:10<3:58:45, 15.71s/it]
 91%|█████████ | 8953/9864 [37:41:24<3:53:46, 15.40s/it]
 91%|█████████ | 8954/9864 [37:41:39<3:48:48, 15.09s/it]
 91%|█████████ | 8955/9864 [37:41:54<3:49:42, 15.16s/it]
 91%|█████████ | 8956/9864 [37:42:08<3:44:37, 14.84s/it]
 91%|█████████ | 8957/9864 [37:42:23<3:42:08, 14.70s/it]
 91%|█████████ | 8958/9864 [37:42:37<3:42:20, 14.72s/it]
 91%|█████████ | 8959/9864 [37:42:55<3:56:31, 15.68s/it]
 91%|█████████ | 8960/9864 [37:43:11<3:58:04, 15.80s/it]
                                                        
{'loss': '0.05044', 'grad_norm': '5.398', 'learning_rate': '4.383e-07', 'epoch': '0.9084'}

 91%|█████████ | 8960/9864 [37:43:11<3:58:04, 15.80s/it]
 91%|█████████ | 8961/9864 [37:43:26<3:50:48, 15.34s/it]
 91%|█████████ | 8962/9864 [37:43:40<3:46:29, 15.07s/it]
 91%|█████████ | 8963/9864 [37:43:56<3:52:05, 15.46s/it]
 91%|█████████ | 8964/9864 [37:44:11<3:47:59, 15.20s/it]
 91%|█████████ | 8965/9864 [37:44:26<3:44:27, 14.98s/it]
 91%|█████████ | 8966/9864 [37:44:42<3:49:16, 15.32s/it]
 91%|█████████ | 8967/9864 [37:44:56<3:44:42, 15.03s/it]
 91%|█████████ | 8968/9864 [37:45:14<3:57:58, 15.94s/it]
 91%|█████████ | 8969/9864 [37:45:29<3:53:27, 15.65s/it]
 91%|█████████ | 8970/9864 [37:45:43<3:47:39, 15.28s/it]
                                                        
{'loss': '0.05273', 'grad_norm': '3.867', 'learning_rate': '4.287e-07', 'epoch': '0.9094'}

 91%|█████████ | 8970/9864 [37:45:43<3:47:39, 15.28s/it]
 91%|█████████ | 8971/9864 [37:46:00<3:51:17, 15.54s/it]
 91%|█████████ | 8972/9864 [37:46:13<3:43:49, 15.06s/it]
 91%|█████████ | 8973/9864 [37:46:28<3:40:36, 14.86s/it]
 91%|█████████ | 8974/9864 [37:46:44<3:45:13, 15.18s/it]
 91%|█████████ | 8975/9864 [37:47:00<3:48:00, 15.39s/it]
 91%|█████████ | 8976/9864 [37:47:16<3:53:16, 15.76s/it]
 91%|█████████ | 8977/9864 [37:47:30<3:45:49, 15.28s/it]
 91%|█████████ | 8978/9864 [37:47:45<3:41:56, 15.03s/it]
 91%|█████████ | 8979/9864 [37:48:01<3:46:17, 15.34s/it]
 91%|█████████ | 8980/9864 [37:48:15<3:40:47, 14.99s/it]
                                                        
{'loss': '0.04683', 'grad_norm': '4.496', 'learning_rate': '4.192e-07', 'epoch': '0.9105'}

 91%|█████████ | 8980/9864 [37:48:15<3:40:47, 14.99s/it]
 91%|█████████ | 8981/9864 [37:48:29<3:37:26, 14.78s/it]
 91%|█████████ | 8982/9864 [37:48:47<3:49:06, 15.59s/it]
 91%|█████████ | 8983/9864 [37:49:01<3:44:10, 15.27s/it]
 91%|█████████ | 8984/9864 [37:49:18<3:49:05, 15.62s/it]
 91%|█████████ | 8985/9864 [37:49:32<3:43:58, 15.29s/it]
 91%|█████████ | 8986/9864 [37:49:47<3:40:51, 15.09s/it]
 91%|█████████ | 8987/9864 [37:50:03<3:45:13, 15.41s/it]
 91%|█████████ | 8988/9864 [37:50:18<3:41:07, 15.15s/it]
 91%|█████████ | 8989/9864 [37:50:35<3:50:38, 15.82s/it]
 91%|█████████ | 8990/9864 [37:50:50<3:45:08, 15.46s/it]
                                                        
{'loss': '0.05358', 'grad_norm': '4.012', 'learning_rate': '4.099e-07', 'epoch': '0.9115'}

 91%|█████████ | 8990/9864 [37:50:50<3:45:08, 15.46s/it]
 91%|█████████ | 8991/9864 [37:51:06<3:47:10, 15.61s/it]
 91%|█████████ | 8992/9864 [37:51:20<3:41:06, 15.21s/it]
 91%|█████████ | 8993/9864 [37:51:34<3:35:07, 14.82s/it]
 91%|█████████ | 8994/9864 [37:51:49<3:34:31, 14.79s/it]
 91%|█████████ | 8995/9864 [37:52:05<3:39:33, 15.16s/it]
 91%|█████████ | 8996/9864 [37:52:20<3:41:19, 15.30s/it]
 91%|█████████ | 8997/9864 [37:52:34<3:34:34, 14.85s/it]
 91%|█████████ | 8998/9864 [37:52:52<3:45:49, 15.65s/it]
 91%|█████████ | 8999/9864 [37:53:06<3:39:29, 15.22s/it]
 91%|█████████ | 9000/9864 [37:53:20<3:36:39, 15.05s/it]
                                                        
{'loss': '0.05165', 'grad_norm': '5.345', 'learning_rate': '4.006e-07', 'epoch': '0.9125'}

 91%|█████████ | 9000/9864 [37:53:20<3:36:39, 15.05s/it]

Writing model shards:   0%|          | 0/1 [00:00<?, ?it/s]

Writing model shards: 100%|██████████| 1/1 [00:08<00:00,  8.68s/it]
Writing model shards: 100%|██████████| 1/1 [00:08<00:00,  8.68s/it]
/home/dungpq6/Project/SpatialStack/src/qwen_vl/model/geometry_encoders/vggt_encoder.py:285: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with torch.cuda.amp.autocast(dtype=dtype):

 91%|█████████▏| 9001/9864 [37:54:38<8:05:13, 33.74s/it]
 91%|█████████▏| 9002/9864 [37:54:54<6:48:35, 28.44s/it]
 91%|█████████▏| 9003/9864 [37:55:08<5:45:38, 24.09s/it]
 91%|█████████▏| 9004/9864 [37:55:23<5:07:51, 21.48s/it]
 91%|█████████▏| 9005/9864 [37:55:39<4:42:39, 19.74s/it]
 91%|█████████▏| 9006/9864 [37:55:54<4:24:30, 18.50s/it]
 91%|█████████▏| 9007/9864 [37:56:09<4:06:48, 17.28s/it]
 91%|█████████▏| 9008/9864 [37:56:24<3:55:33, 16.51s/it]
 91%|█████████▏| 9009/9864 [37:56:38<3:46:06, 15.87s/it]
 91%|█████████▏| 9010/9864 [37:56:54<3:46:50, 15.94s/it]
                                                        
{'loss': '0.04002', 'grad_norm': '4.44', 'learning_rate': '3.915e-07', 'epoch': '0.9135'}

 91%|█████████▏| 9010/9864 [37:56:54<3:46:50, 15.94s/it]
 91%|█████████▏| 9011/9864 [37:57:08<3:39:52, 15.47s/it]
 91%|█████████▏| 9012/9864 [37:57:25<3:42:50, 15.69s/it]
 91%|█████████▏| 9013/9864 [37:57:42<3:51:19, 16.31s/it]
 91%|█████████▏| 9014/9864 [37:57:57<3:41:51, 15.66s/it]
 91%|█████████▏| 9015/9864 [37:58:10<3:33:51, 15.11s/it]
 91%|█████████▏| 9016/9864 [37:58:25<3:30:40, 14.91s/it]
 91%|█████████▏| 9017/9864 [37:58:40<3:33:21, 15.11s/it]
 91%|█████████▏| 9018/9864 [37:58:55<3:29:39, 14.87s/it]
 91%|█████████▏| 9019/9864 [37:59:09<3:27:19, 14.72s/it]
 91%|█████████▏| 9020/9864 [37:59:25<3:33:04, 15.15s/it]
                                                        
{'loss': '0.0555', 'grad_norm': '5.279', 'learning_rate': '3.824e-07', 'epoch': '0.9145'}

 91%|█████████▏| 9020/9864 [37:59:25<3:33:04, 15.15s/it]
 91%|█████████▏| 9021/9864 [37:59:43<3:45:43, 16.07s/it]
 91%|█████████▏| 9022/9864 [37:59:58<3:37:46, 15.52s/it]
 91%|█████████▏| 9023/9864 [38:00:12<3:31:27, 15.09s/it]
 91%|█████████▏| 9024/9864 [38:00:26<3:25:49, 14.70s/it]
 91%|█████████▏| 9025/9864 [38:00:40<3:22:58, 14.52s/it]
 92%|█████████▏| 9026/9864 [38:00:56<3:28:31, 14.93s/it]
 92%|█████████▏| 9027/9864 [38:01:12<3:33:05, 15.28s/it]
 92%|█████████▏| 9028/9864 [38:01:26<3:27:51, 14.92s/it]
 92%|█████████▏| 9029/9864 [38:01:43<3:36:54, 15.59s/it]
 92%|█████████▏| 9030/9864 [38:01:57<3:30:42, 15.16s/it]
                                                        
{'loss': '0.04289', 'grad_norm': '3.165', 'learning_rate': '3.735e-07', 'epoch': '0.9155'}

 92%|█████████▏| 9030/9864 [38:01:57<3:30:42, 15.16s/it]
 92%|█████████▏| 9031/9864 [38:02:12<3:28:17, 15.00s/it]
 92%|█████████▏| 9032/9864 [38:02:26<3:24:59, 14.78s/it]
 92%|█████████▏| 9033/9864 [38:02:42<3:30:06, 15.17s/it]
 92%|█████████▏| 9034/9864 [38:02:56<3:25:14, 14.84s/it]
 92%|█████████▏| 9035/9864 [38:03:12<3:28:57, 15.12s/it]
 92%|█████████▏| 9036/9864 [38:03:26<3:25:17, 14.88s/it]
 92%|█████████▏| 9037/9864 [38:03:42<3:29:50, 15.22s/it]
 92%|█████████▏| 9038/9864 [38:03:58<3:31:30, 15.36s/it]
 92%|█████████▏| 9039/9864 [38:04:12<3:26:00, 14.98s/it]
 92%|█████████▏| 9040/9864 [38:04:28<3:29:19, 15.24s/it]
                                                        
{'loss': '0.05799', 'grad_norm': '3.103', 'learning_rate': '3.647e-07', 'epoch': '0.9165'}

 92%|█████████▏| 9040/9864 [38:04:28<3:29:19, 15.24s/it]
 92%|█████████▏| 9041/9864 [38:04:42<3:25:48, 15.00s/it]
 92%|█████████▏| 9042/9864 [38:04:58<3:27:23, 15.14s/it]
 92%|█████████▏| 9043/9864 [38:05:12<3:23:29, 14.87s/it]
 92%|█████████▏| 9044/9864 [38:05:27<3:22:50, 14.84s/it]
 92%|█████████▏| 9045/9864 [38:05:45<3:36:32, 15.86s/it]
 92%|█████████▏| 9046/9864 [38:06:00<3:30:55, 15.47s/it]
 92%|█████████▏| 9047/9864 [38:06:14<3:26:25, 15.16s/it]
 92%|█████████▏| 9048/9864 [38:06:30<3:30:39, 15.49s/it]
 92%|█████████▏| 9049/9864 [38:06:47<3:36:02, 15.90s/it]
 92%|█████████▏| 9050/9864 [38:07:01<3:29:25, 15.44s/it]
                                                        
{'loss': '0.05274', 'grad_norm': '4.803', 'learning_rate': '3.559e-07', 'epoch': '0.9176'}

 92%|█████████▏| 9050/9864 [38:07:01<3:29:25, 15.44s/it]
 92%|█████████▏| 9051/9864 [38:07:16<3:24:13, 15.07s/it]
 92%|█████████▏| 9052/9864 [38:07:34<3:36:03, 15.97s/it]
 92%|█████████▏| 9053/9864 [38:07:48<3:27:15, 15.33s/it]
 92%|█████████▏| 9054/9864 [38:08:02<3:22:02, 14.97s/it]
 92%|█████████▏| 9055/9864 [38:08:16<3:20:28, 14.87s/it]
 92%|█████████▏| 9056/9864 [38:08:32<3:25:25, 15.25s/it]
 92%|█████████▏| 9057/9864 [38:08:49<3:30:07, 15.62s/it]
 92%|█████████▏| 9058/9864 [38:09:04<3:25:41, 15.31s/it]
 92%|█████████▏| 9059/9864 [38:09:19<3:26:02, 15.36s/it]
 92%|█████████▏| 9060/9864 [38:09:33<3:21:53, 15.07s/it]
                                                        
{'loss': '0.05006', 'grad_norm': '5.094', 'learning_rate': '3.473e-07', 'epoch': '0.9186'}

 92%|█████████▏| 9060/9864 [38:09:33<3:21:53, 15.07s/it]
 92%|█████████▏| 9061/9864 [38:09:49<3:25:00, 15.32s/it]
 92%|█████████▏| 9062/9864 [38:10:03<3:19:29, 14.92s/it]
 92%|█████████▏| 9063/9864 [38:10:19<3:23:04, 15.21s/it]
 92%|█████████▏| 9064/9864 [38:10:35<3:25:41, 15.43s/it]
 92%|█████████▏| 9065/9864 [38:10:49<3:19:17, 14.97s/it]
 92%|█████████▏| 9066/9864 [38:11:03<3:16:01, 14.74s/it]
 92%|█████████▏| 9067/9864 [38:11:19<3:19:16, 15.00s/it]
 92%|█████████▏| 9068/9864 [38:11:33<3:15:29, 14.74s/it]
 92%|█████████▏| 9069/9864 [38:11:47<3:13:19, 14.59s/it]
 92%|█████████▏| 9070/9864 [38:12:06<3:28:13, 15.74s/it]
                                                        
{'loss': '0.06136', 'grad_norm': '3.971', 'learning_rate': '3.388e-07', 'epoch': '0.9196'}

 92%|█████████▏| 9070/9864 [38:12:06<3:28:13, 15.74s/it]
 92%|█████████▏| 9071/9864 [38:12:22<3:29:52, 15.88s/it]
 92%|█████████▏| 9072/9864 [38:12:37<3:24:53, 15.52s/it]
 92%|█████████▏| 9073/9864 [38:12:51<3:20:32, 15.21s/it]
 92%|█████████▏| 9074/9864 [38:13:07<3:21:39, 15.32s/it]
 92%|█████████▏| 9075/9864 [38:13:21<3:18:22, 15.09s/it]
 92%|█████████▏| 9076/9864 [38:13:35<3:14:25, 14.80s/it]
 92%|█████████▏| 9077/9864 [38:13:51<3:18:55, 15.17s/it]
 92%|█████████▏| 9078/9864 [38:14:07<3:20:40, 15.32s/it]
 92%|█████████▏| 9079/9864 [38:14:23<3:22:27, 15.47s/it]
 92%|█████████▏| 9080/9864 [38:14:37<3:17:09, 15.09s/it]
                                                        
{'loss': '0.05447', 'grad_norm': '2.886', 'learning_rate': '3.303e-07', 'epoch': '0.9206'}

 92%|█████████▏| 9080/9864 [38:14:37<3:17:09, 15.09s/it]
 92%|█████████▏| 9081/9864 [38:14:53<3:19:58, 15.32s/it]
 92%|█████████▏| 9082/9864 [38:15:07<3:15:04, 14.97s/it]
 92%|█████████▏| 9083/9864 [38:15:21<3:12:52, 14.82s/it]
 92%|█████████▏| 9084/9864 [38:15:37<3:16:59, 15.15s/it]
 92%|█████████▏| 9085/9864 [38:15:52<3:13:29, 14.90s/it]
 92%|█████████▏| 9086/9864 [38:16:08<3:17:08, 15.20s/it]
 92%|█████████▏| 9087/9864 [38:16:24<3:20:12, 15.46s/it]
 92%|█████████▏| 9088/9864 [38:16:38<3:14:43, 15.06s/it]
 92%|█████████▏| 9089/9864 [38:16:54<3:19:14, 15.43s/it]
 92%|█████████▏| 9090/9864 [38:17:08<3:14:55, 15.11s/it]
                                                        
{'loss': '0.05104', 'grad_norm': '7.217', 'learning_rate': '3.22e-07', 'epoch': '0.9216'}

 92%|█████████▏| 9090/9864 [38:17:08<3:14:55, 15.11s/it]
 92%|█████████▏| 9091/9864 [38:17:23<3:13:38, 15.03s/it]
 92%|█████████▏| 9092/9864 [38:17:37<3:10:06, 14.77s/it]
 92%|█████████▏| 9093/9864 [38:17:53<3:14:03, 15.10s/it]
 92%|█████████▏| 9094/9864 [38:18:11<3:25:05, 15.98s/it]
 92%|█████████▏| 9095/9864 [38:18:26<3:18:23, 15.48s/it]
 92%|█████████▏| 9096/9864 [38:18:40<3:13:11, 15.09s/it]
 92%|█████████▏| 9097/9864 [38:18:55<3:14:35, 15.22s/it]
 92%|█████████▏| 9098/9864 [38:19:10<3:10:22, 14.91s/it]
 92%|█████████▏| 9099/9864 [38:19:24<3:07:41, 14.72s/it]
 92%|█████████▏| 9100/9864 [38:19:41<3:15:25, 15.35s/it]
                                                        
{'loss': '0.05923', 'grad_norm': '7.049', 'learning_rate': '3.138e-07', 'epoch': '0.9226'}

 92%|█████████▏| 9100/9864 [38:19:41<3:15:25, 15.35s/it]
 92%|█████████▏| 9101/9864 [38:19:57<3:17:30, 15.53s/it]
 92%|█████████▏| 9102/9864 [38:20:13<3:19:30, 15.71s/it]
 92%|█████████▏| 9103/9864 [38:20:27<3:14:51, 15.36s/it]
 92%|█████████▏| 9104/9864 [38:20:42<3:10:39, 15.05s/it]
 92%|█████████▏| 9105/9864 [38:20:58<3:14:00, 15.34s/it]
 92%|█████████▏| 9106/9864 [38:21:12<3:09:04, 14.97s/it]
 92%|█████████▏| 9107/9864 [38:21:26<3:04:22, 14.61s/it]
 92%|█████████▏| 9108/9864 [38:21:43<3:13:15, 15.34s/it]
 92%|█████████▏| 9109/9864 [38:21:58<3:13:44, 15.40s/it]
 92%|█████████▏| 9110/9864 [38:22:13<3:09:53, 15.11s/it]
                                                        
{'loss': '0.04466', 'grad_norm': '3.521', 'learning_rate': '3.057e-07', 'epoch': '0.9236'}

 92%|█████████▏| 9110/9864 [38:22:13<3:09:53, 15.11s/it]
 92%|█████████▏| 9111/9864 [38:22:27<3:06:28, 14.86s/it]
 92%|█████████▏| 9112/9864 [38:22:43<3:09:35, 15.13s/it]
 92%|█████████▏| 9113/9864 [38:22:57<3:06:45, 14.92s/it]
 92%|█████████▏| 9114/9864 [38:23:11<3:03:54, 14.71s/it]
 92%|█████████▏| 9115/9864 [38:23:25<3:01:15, 14.52s/it]
 92%|█████████▏| 9116/9864 [38:23:43<3:13:29, 15.52s/it]
 92%|█████████▏| 9117/9864 [38:23:59<3:15:38, 15.71s/it]
 92%|█████████▏| 9118/9864 [38:24:13<3:09:27, 15.24s/it]
 92%|█████████▏| 9119/9864 [38:24:28<3:06:14, 15.00s/it]
 92%|█████████▏| 9120/9864 [38:24:44<3:08:50, 15.23s/it]
                                                        
{'loss': '0.0474', 'grad_norm': '2.305', 'learning_rate': '2.977e-07', 'epoch': '0.9247'}

 92%|█████████▏| 9120/9864 [38:24:44<3:08:50, 15.23s/it]
 92%|█████████▏| 9121/9864 [38:24:58<3:03:56, 14.85s/it]
 92%|█████████▏| 9122/9864 [38:25:12<3:02:29, 14.76s/it]
 92%|█████████▏| 9123/9864 [38:25:26<3:00:33, 14.62s/it]
 92%|█████████▏| 9124/9864 [38:25:41<3:00:36, 14.64s/it]
 93%|█████████▎| 9125/9864 [38:26:00<3:16:40, 15.97s/it]
 93%|█████████▎| 9126/9864 [38:26:15<3:10:49, 15.51s/it]
 93%|█████████▎| 9127/9864 [38:26:29<3:05:27, 15.10s/it]
 93%|█████████▎| 9128/9864 [38:26:43<3:02:54, 14.91s/it]
 93%|█████████▎| 9129/9864 [38:26:59<3:06:32, 15.23s/it]
 93%|█████████▎| 9130/9864 [38:27:13<3:01:56, 14.87s/it]
                                                        
{'loss': '0.04467', 'grad_norm': '1.989', 'learning_rate': '2.898e-07', 'epoch': '0.9257'}

 93%|█████████▎| 9130/9864 [38:27:13<3:01:56, 14.87s/it]
 93%|█████████▎| 9131/9864 [38:27:28<3:02:14, 14.92s/it]
 93%|█████████▎| 9132/9864 [38:27:45<3:06:44, 15.31s/it]
 93%|█████████▎| 9133/9864 [38:28:01<3:09:33, 15.56s/it]
 93%|█████████▎| 9134/9864 [38:28:17<3:11:22, 15.73s/it]
 93%|█████████▎| 9135/9864 [38:28:31<3:06:11, 15.32s/it]
 93%|█████████▎| 9136/9864 [38:28:47<3:07:41, 15.47s/it]
 93%|█████████▎| 9137/9864 [38:29:02<3:04:51, 15.26s/it]
 93%|█████████▎| 9138/9864 [38:29:16<3:00:16, 14.90s/it]
 93%|█████████▎| 9139/9864 [38:29:30<2:57:58, 14.73s/it]
 93%|█████████▎| 9140/9864 [38:29:48<3:09:23, 15.70s/it]
                                                        
{'loss': '0.04986', 'grad_norm': '5.533', 'learning_rate': '2.82e-07', 'epoch': '0.9267'}

 93%|█████████▎| 9140/9864 [38:29:48<3:09:23, 15.70s/it]
 93%|█████████▎| 9141/9864 [38:30:02<3:04:03, 15.28s/it]
 93%|█████████▎| 9142/9864 [38:30:18<3:05:39, 15.43s/it]
 93%|█████████▎| 9143/9864 [38:30:34<3:08:40, 15.70s/it]
 93%|█████████▎| 9144/9864 [38:30:49<3:03:12, 15.27s/it]
 93%|█████████▎| 9145/9864 [38:31:04<3:01:52, 15.18s/it]
 93%|█████████▎| 9146/9864 [38:31:18<2:58:23, 14.91s/it]
 93%|█████████▎| 9147/9864 [38:31:34<3:02:08, 15.24s/it]
 93%|█████████▎| 9148/9864 [38:31:49<3:02:43, 15.31s/it]
 93%|█████████▎| 9149/9864 [38:32:04<2:58:37, 14.99s/it]
 93%|█████████▎| 9150/9864 [38:32:18<2:55:07, 14.72s/it]
                                                        
{'loss': '0.04426', 'grad_norm': '4.431', 'learning_rate': '2.743e-07', 'epoch': '0.9277'}

 93%|█████████▎| 9150/9864 [38:32:18<2:55:07, 14.72s/it]
 93%|█████████▎| 9151/9864 [38:32:36<3:06:37, 15.71s/it]
 93%|█████████▎| 9152/9864 [38:32:51<3:02:52, 15.41s/it]
 93%|█████████▎| 9153/9864 [38:33:05<2:59:09, 15.12s/it]
 93%|█████████▎| 9154/9864 [38:33:19<2:55:55, 14.87s/it]
 93%|█████████▎| 9155/9864 [38:33:35<2:59:39, 15.20s/it]
 93%|█████████▎| 9156/9864 [38:33:50<2:56:26, 14.95s/it]
 93%|█████████▎| 9157/9864 [38:34:03<2:49:33, 14.39s/it]
 93%|█████████▎| 9158/9864 [38:34:17<2:49:18, 14.39s/it]
 93%|█████████▎| 9159/9864 [38:34:35<3:00:09, 15.33s/it]
 93%|█████████▎| 9160/9864 [38:34:49<2:56:11, 15.02s/it]
                                                        
{'loss': '0.04049', 'grad_norm': '3.724', 'learning_rate': '2.667e-07', 'epoch': '0.9287'}

 93%|█████████▎| 9160/9864 [38:34:49<2:56:11, 15.02s/it]
 93%|█████████▎| 9161/9864 [38:35:03<2:51:09, 14.61s/it]
 93%|█████████▎| 9162/9864 [38:35:20<3:01:58, 15.55s/it]
 93%|█████████▎| 9163/9864 [38:35:35<2:57:27, 15.19s/it]
 93%|█████████▎| 9164/9864 [38:35:49<2:55:04, 15.01s/it]
 93%|█████████▎| 9165/9864 [38:36:04<2:52:22, 14.80s/it]
 93%|█████████▎| 9166/9864 [38:36:20<2:56:58, 15.21s/it]
 93%|█████████▎| 9167/9864 [38:36:36<2:58:53, 15.40s/it]
 93%|█████████▎| 9168/9864 [38:36:50<2:54:17, 15.03s/it]
 93%|█████████▎| 9169/9864 [38:37:06<2:58:56, 15.45s/it]
 93%|█████████▎| 9170/9864 [38:37:23<3:02:05, 15.74s/it]
                                                        
{'loss': '0.05243', 'grad_norm': '5.459', 'learning_rate': '2.592e-07', 'epoch': '0.9297'}

 93%|█████████▎| 9170/9864 [38:37:23<3:02:05, 15.74s/it]
 93%|█████████▎| 9171/9864 [38:37:37<2:56:02, 15.24s/it]
 93%|█████████▎| 9172/9864 [38:37:51<2:51:48, 14.90s/it]
 93%|█████████▎| 9173/9864 [38:38:05<2:50:02, 14.76s/it]
 93%|█████████▎| 9174/9864 [38:38:21<2:53:26, 15.08s/it]
 93%|█████████▎| 9175/9864 [38:38:37<2:54:54, 15.23s/it]
 93%|█████████▎| 9176/9864 [38:38:53<2:58:41, 15.58s/it]
 93%|█████████▎| 9177/9864 [38:39:09<3:01:17, 15.83s/it]
 93%|█████████▎| 9178/9864 [38:39:24<2:57:37, 15.54s/it]
 93%|█████████▎| 9179/9864 [38:39:39<2:53:32, 15.20s/it]
 93%|█████████▎| 9180/9864 [38:39:53<2:49:04, 14.83s/it]
                                                        
{'loss': '0.04873', 'grad_norm': '5.18', 'learning_rate': '2.519e-07', 'epoch': '0.9307'}

 93%|█████████▎| 9180/9864 [38:39:53<2:49:04, 14.83s/it]
 93%|█████████▎| 9181/9864 [38:40:09<2:52:25, 15.15s/it]
 93%|█████████▎| 9182/9864 [38:40:23<2:50:29, 15.00s/it]
 93%|█████████▎| 9183/9864 [38:40:39<2:51:58, 15.15s/it]
 93%|█████████▎| 9184/9864 [38:40:55<2:56:01, 15.53s/it]
 93%|█████████▎| 9185/9864 [38:41:11<2:57:57, 15.73s/it]
 93%|█████████▎| 9186/9864 [38:41:25<2:49:26, 15.00s/it]
 93%|█████████▎| 9187/9864 [38:41:39<2:46:30, 14.76s/it]
 93%|█████████▎| 9188/9864 [38:41:53<2:44:57, 14.64s/it]
 93%|█████████▎| 9189/9864 [38:42:07<2:43:13, 14.51s/it]
 93%|█████████▎| 9190/9864 [38:42:25<2:53:40, 15.46s/it]
                                                        
{'loss': '0.04065', 'grad_norm': '3.813', 'learning_rate': '2.446e-07', 'epoch': '0.9317'}

 93%|█████████▎| 9190/9864 [38:42:25<2:53:40, 15.46s/it]
 93%|█████████▎| 9191/9864 [38:42:40<2:50:16, 15.18s/it]
 93%|█████████▎| 9192/9864 [38:42:56<2:54:03, 15.54s/it]
 93%|█████████▎| 9193/9864 [38:43:12<2:56:55, 15.82s/it]
 93%|█████████▎| 9194/9864 [38:43:26<2:50:30, 15.27s/it]
 93%|█████████▎| 9195/9864 [38:43:41<2:47:10, 14.99s/it]
 93%|█████████▎| 9196/9864 [38:43:54<2:42:03, 14.56s/it]
 93%|█████████▎| 9197/9864 [38:44:08<2:39:16, 14.33s/it]
 93%|█████████▎| 9198/9864 [38:44:26<2:51:42, 15.47s/it]
 93%|█████████▎| 9199/9864 [38:44:40<2:47:24, 15.10s/it]
 93%|█████████▎| 9200/9864 [38:44:56<2:50:14, 15.38s/it]
                                                        
{'loss': '0.05763', 'grad_norm': '5.037', 'learning_rate': '2.374e-07', 'epoch': '0.9328'}

 93%|█████████▎| 9200/9864 [38:44:56<2:50:14, 15.38s/it]
 93%|█████████▎| 9201/9864 [38:45:13<2:52:34, 15.62s/it]
 93%|█████████▎| 9202/9864 [38:45:27<2:49:28, 15.36s/it]
 93%|█████████▎| 9203/9864 [38:45:42<2:45:27, 15.02s/it]
 93%|█████████▎| 9204/9864 [38:45:56<2:42:44, 14.79s/it]
 93%|█████████▎| 9205/9864 [38:46:12<2:46:00, 15.12s/it]
 93%|█████████▎| 9206/9864 [38:46:28<2:48:16, 15.34s/it]
 93%|█████████▎| 9207/9864 [38:46:44<2:50:58, 15.61s/it]
 93%|█████████▎| 9208/9864 [38:46:59<2:47:36, 15.33s/it]
 93%|█████████▎| 9209/9864 [38:47:15<2:50:08, 15.58s/it]
 93%|█████████▎| 9210/9864 [38:47:29<2:46:14, 15.25s/it]
                                                        
{'loss': '0.03936', 'grad_norm': '3.235', 'learning_rate': '2.304e-07', 'epoch': '0.9338'}

 93%|█████████▎| 9210/9864 [38:47:29<2:46:14, 15.25s/it]
 93%|█████████▎| 9211/9864 [38:47:43<2:42:48, 14.96s/it]
 93%|█████████▎| 9212/9864 [38:47:58<2:39:39, 14.69s/it]
 93%|█████████▎| 9213/9864 [38:48:14<2:44:57, 15.20s/it]
 93%|█████████▎| 9214/9864 [38:48:31<2:49:13, 15.62s/it]
 93%|█████████▎| 9215/9864 [38:48:47<2:51:36, 15.86s/it]
 93%|█████████▎| 9216/9864 [38:49:01<2:44:37, 15.24s/it]
 93%|█████████▎| 9217/9864 [38:49:17<2:47:52, 15.57s/it]
 93%|█████████▎| 9218/9864 [38:49:32<2:43:55, 15.23s/it]
 93%|█████████▎| 9219/9864 [38:49:46<2:40:55, 14.97s/it]
 93%|█████████▎| 9220/9864 [38:50:00<2:39:20, 14.85s/it]
                                                        
{'loss': '0.04394', 'grad_norm': '4.98', 'learning_rate': '2.234e-07', 'epoch': '0.9348'}

 93%|█████████▎| 9220/9864 [38:50:00<2:39:20, 14.85s/it]
 93%|█████████▎| 9221/9864 [38:50:16<2:41:10, 15.04s/it]
 93%|█████████▎| 9222/9864 [38:50:32<2:44:42, 15.39s/it]
 94%|█████████▎| 9223/9864 [38:50:47<2:42:10, 15.18s/it]
 94%|█████████▎| 9224/9864 [38:51:03<2:44:39, 15.44s/it]
 94%|█████████▎| 9225/9864 [38:51:17<2:39:41, 14.99s/it]
 94%|█████████▎| 9226/9864 [38:51:31<2:35:54, 14.66s/it]
 94%|█████████▎| 9227/9864 [38:51:45<2:33:03, 14.42s/it]
 94%|█████████▎| 9228/9864 [38:51:59<2:33:52, 14.52s/it]
 94%|█████████▎| 9229/9864 [38:52:15<2:38:06, 14.94s/it]
 94%|█████████▎| 9230/9864 [38:52:31<2:40:30, 15.19s/it]
                                                        
{'loss': '0.05029', 'grad_norm': '6.495', 'learning_rate': '2.166e-07', 'epoch': '0.9358'}

 94%|█████████▎| 9230/9864 [38:52:31<2:40:30, 15.19s/it]
 94%|█████████▎| 9231/9864 [38:52:47<2:42:14, 15.38s/it]
 94%|█████████▎| 9232/9864 [38:53:03<2:43:33, 15.53s/it]
 94%|█████████▎| 9233/9864 [38:53:17<2:40:14, 15.24s/it]
 94%|█████████▎| 9234/9864 [38:53:32<2:37:33, 15.01s/it]
 94%|█████████▎| 9235/9864 [38:53:46<2:34:07, 14.70s/it]
 94%|█████████▎| 9236/9864 [38:54:00<2:33:31, 14.67s/it]
 94%|█████████▎| 9237/9864 [38:54:16<2:37:00, 15.02s/it]
 94%|█████████▎| 9238/9864 [38:54:32<2:38:47, 15.22s/it]
 94%|█████████▎| 9239/9864 [38:54:48<2:41:22, 15.49s/it]
 94%|█████████▎| 9240/9864 [38:55:04<2:42:31, 15.63s/it]
                                                        
{'loss': '0.03884', 'grad_norm': '3.797', 'learning_rate': '2.098e-07', 'epoch': '0.9368'}

 94%|█████████▎| 9240/9864 [38:55:04<2:42:31, 15.63s/it]
 94%|█████████▎| 9241/9864 [38:55:18<2:38:41, 15.28s/it]
 94%|█████████▎| 9242/9864 [38:55:32<2:33:27, 14.80s/it]
 94%|█████████▎| 9243/9864 [38:55:47<2:32:22, 14.72s/it]
 94%|█████████▎| 9244/9864 [38:56:03<2:35:56, 15.09s/it]
 94%|█████████▎| 9245/9864 [38:56:17<2:33:11, 14.85s/it]
 94%|█████████▎| 9246/9864 [38:56:33<2:35:58, 15.14s/it]
 94%|█████████▎| 9247/9864 [38:56:49<2:37:48, 15.35s/it]
 94%|█████████▍| 9248/9864 [38:57:04<2:36:49, 15.27s/it]
 94%|█████████▍| 9249/9864 [38:57:20<2:38:32, 15.47s/it]
 94%|█████████▍| 9250/9864 [38:57:34<2:34:51, 15.13s/it]
                                                        
{'loss': '0.03739', 'grad_norm': '3.003', 'learning_rate': '2.032e-07', 'epoch': '0.9378'}

 94%|█████████▍| 9250/9864 [38:57:34<2:34:51, 15.13s/it]
 94%|█████████▍| 9251/9864 [38:57:49<2:33:29, 15.02s/it]
 94%|█████████▍| 9252/9864 [38:58:05<2:38:07, 15.50s/it]
 94%|█████████▍| 9253/9864 [38:58:19<2:33:32, 15.08s/it]
 94%|█████████▍| 9254/9864 [38:58:37<2:40:26, 15.78s/it]
 94%|█████████▍| 9255/9864 [38:58:51<2:36:06, 15.38s/it]
 94%|█████████▍| 9256/9864 [38:59:07<2:38:29, 15.64s/it]
 94%|█████████▍| 9257/9864 [38:59:21<2:32:55, 15.12s/it]
 94%|█████████▍| 9258/9864 [38:59:35<2:29:16, 14.78s/it]
 94%|█████████▍| 9259/9864 [38:59:52<2:33:09, 15.19s/it]
 94%|█████████▍| 9260/9864 [39:00:06<2:31:53, 15.09s/it]
                                                        
{'loss': '0.04809', 'grad_norm': '3.992', 'learning_rate': '1.967e-07', 'epoch': '0.9388'}

 94%|█████████▍| 9260/9864 [39:00:06<2:31:53, 15.09s/it]
 94%|█████████▍| 9261/9864 [39:00:22<2:34:29, 15.37s/it]
 94%|█████████▍| 9262/9864 [39:00:38<2:35:15, 15.47s/it]
 94%|█████████▍| 9263/9864 [39:00:55<2:38:10, 15.79s/it]
 94%|█████████▍| 9264/9864 [39:01:09<2:34:08, 15.41s/it]
 94%|█████████▍| 9265/9864 [39:01:23<2:30:10, 15.04s/it]
 94%|█████████▍| 9266/9864 [39:01:38<2:28:05, 14.86s/it]
 94%|█████████▍| 9267/9864 [39:01:52<2:25:04, 14.58s/it]
 94%|█████████▍| 9268/9864 [39:02:08<2:29:19, 15.03s/it]
 94%|█████████▍| 9269/9864 [39:02:23<2:28:30, 14.98s/it]
 94%|█████████▍| 9270/9864 [39:02:41<2:36:56, 15.85s/it]
                                                        
{'loss': '0.05404', 'grad_norm': '3.62', 'learning_rate': '1.902e-07', 'epoch': '0.9399'}

 94%|█████████▍| 9270/9864 [39:02:41<2:36:56, 15.85s/it]
 94%|█████████▍| 9271/9864 [39:02:57<2:37:45, 15.96s/it]
 94%|█████████▍| 9272/9864 [39:03:11<2:32:53, 15.50s/it]
 94%|█████████▍| 9273/9864 [39:03:25<2:28:00, 15.03s/it]
 94%|█████████▍| 9274/9864 [39:03:39<2:24:33, 14.70s/it]
 94%|█████████▍| 9275/9864 [39:03:55<2:26:39, 14.94s/it]
 94%|█████████▍| 9276/9864 [39:04:09<2:25:07, 14.81s/it]
 94%|█████████▍| 9277/9864 [39:04:25<2:28:12, 15.15s/it]
 94%|█████████▍| 9278/9864 [39:04:42<2:31:57, 15.56s/it]
 94%|█████████▍| 9279/9864 [39:04:58<2:33:43, 15.77s/it]
 94%|█████████▍| 9280/9864 [39:05:12<2:28:45, 15.28s/it]
                                                        
{'loss': '0.04313', 'grad_norm': '4.166', 'learning_rate': '1.839e-07', 'epoch': '0.9409'}

 94%|█████████▍| 9280/9864 [39:05:12<2:28:45, 15.28s/it]
 94%|█████████▍| 9281/9864 [39:05:26<2:25:42, 15.00s/it]
 94%|█████████▍| 9282/9864 [39:05:40<2:22:11, 14.66s/it]
 94%|█████████▍| 9283/9864 [39:05:55<2:21:15, 14.59s/it]
 94%|█████████▍| 9284/9864 [39:06:10<2:24:46, 14.98s/it]
 94%|█████████▍| 9285/9864 [39:06:26<2:25:44, 15.10s/it]
 94%|█████████▍| 9286/9864 [39:06:43<2:32:49, 15.86s/it]
 94%|█████████▍| 9287/9864 [39:06:58<2:27:38, 15.35s/it]
 94%|█████████▍| 9288/9864 [39:07:12<2:23:38, 14.96s/it]
 94%|█████████▍| 9289/9864 [39:07:26<2:20:24, 14.65s/it]
 94%|█████████▍| 9290/9864 [39:07:40<2:18:39, 14.49s/it]
                                                        
{'loss': '0.04455', 'grad_norm': '5.065', 'learning_rate': '1.777e-07', 'epoch': '0.9419'}

 94%|█████████▍| 9290/9864 [39:07:40<2:18:39, 14.49s/it]
 94%|█████████▍| 9291/9864 [39:07:56<2:22:23, 14.91s/it]
 94%|█████████▍| 9292/9864 [39:08:12<2:26:36, 15.38s/it]
 94%|█████████▍| 9293/9864 [39:08:28<2:29:15, 15.68s/it]
 94%|█████████▍| 9294/9864 [39:08:45<2:30:50, 15.88s/it]
 94%|█████████▍| 9295/9864 [39:09:00<2:27:33, 15.56s/it]
 94%|█████████▍| 9296/9864 [39:09:14<2:23:25, 15.15s/it]
 94%|█████████▍| 9297/9864 [39:09:28<2:20:28, 14.87s/it]
 94%|█████████▍| 9298/9864 [39:09:42<2:17:55, 14.62s/it]
 94%|█████████▍| 9299/9864 [39:10:00<2:27:34, 15.67s/it]
 94%|█████████▍| 9300/9864 [39:10:16<2:27:40, 15.71s/it]
                                                        
{'loss': '0.04352', 'grad_norm': '3.829', 'learning_rate': '1.716e-07', 'epoch': '0.9429'}

 94%|█████████▍| 9300/9864 [39:10:16<2:27:40, 15.71s/it]
 94%|█████████▍| 9301/9864 [39:10:32<2:28:58, 15.88s/it]
 94%|█████████▍| 9302/9864 [39:10:47<2:25:19, 15.51s/it]
 94%|█████████▍| 9303/9864 [39:11:01<2:21:44, 15.16s/it]
 94%|█████████▍| 9304/9864 [39:11:16<2:19:08, 14.91s/it]
 94%|█████████▍| 9305/9864 [39:11:30<2:17:46, 14.79s/it]
 94%|█████████▍| 9306/9864 [39:11:46<2:20:16, 15.08s/it]
 94%|█████████▍| 9307/9864 [39:12:03<2:25:04, 15.63s/it]
 94%|█████████▍| 9308/9864 [39:12:19<2:26:01, 15.76s/it]
 94%|█████████▍| 9309/9864 [39:12:35<2:25:56, 15.78s/it]
 94%|█████████▍| 9310/9864 [39:12:49<2:22:17, 15.41s/it]
                                                        
{'loss': '0.05158', 'grad_norm': '5.572', 'learning_rate': '1.656e-07', 'epoch': '0.9439'}

 94%|█████████▍| 9310/9864 [39:12:49<2:22:17, 15.41s/it]
 94%|█████████▍| 9311/9864 [39:13:03<2:18:00, 14.97s/it]
 94%|█████████▍| 9312/9864 [39:13:18<2:17:08, 14.91s/it]
 94%|█████████▍| 9313/9864 [39:13:32<2:15:29, 14.75s/it]
 94%|█████████▍| 9314/9864 [39:13:49<2:20:02, 15.28s/it]
 94%|█████████▍| 9315/9864 [39:14:03<2:17:12, 14.99s/it]
 94%|█████████▍| 9316/9864 [39:14:19<2:19:58, 15.33s/it]
 94%|█████████▍| 9317/9864 [39:14:36<2:22:35, 15.64s/it]
 94%|█████████▍| 9318/9864 [39:14:49<2:17:05, 15.07s/it]
 94%|█████████▍| 9319/9864 [39:15:03<2:14:12, 14.77s/it]
 94%|█████████▍| 9320/9864 [39:15:17<2:11:39, 14.52s/it]
                                                        
{'loss': '0.05127', 'grad_norm': '4.339', 'learning_rate': '1.597e-07', 'epoch': '0.9449'}

 94%|█████████▍| 9320/9864 [39:15:17<2:11:39, 14.52s/it]
 94%|█████████▍| 9321/9864 [39:15:32<2:10:50, 14.46s/it]
 95%|█████████▍| 9322/9864 [39:15:48<2:15:17, 14.98s/it]
 95%|█████████▍| 9323/9864 [39:16:04<2:18:15, 15.33s/it]
 95%|█████████▍| 9324/9864 [39:16:22<2:24:52, 16.10s/it]
 95%|█████████▍| 9325/9864 [39:16:36<2:20:18, 15.62s/it]
 95%|█████████▍| 9326/9864 [39:16:51<2:17:58, 15.39s/it]
 95%|█████████▍| 9327/9864 [39:17:05<2:12:57, 14.86s/it]
 95%|█████████▍| 9328/9864 [39:17:19<2:11:50, 14.76s/it]
 95%|█████████▍| 9329/9864 [39:17:36<2:15:29, 15.20s/it]
 95%|█████████▍| 9330/9864 [39:17:52<2:17:12, 15.42s/it]
                                                        
{'loss': '0.03968', 'grad_norm': '5.511', 'learning_rate': '1.539e-07', 'epoch': '0.9459'}

 95%|█████████▍| 9330/9864 [39:17:52<2:17:12, 15.42s/it]
 95%|█████████▍| 9331/9864 [39:18:06<2:13:48, 15.06s/it]
 95%|█████████▍| 9332/9864 [39:18:22<2:16:55, 15.44s/it]
 95%|█████████▍| 9333/9864 [39:18:38<2:18:53, 15.69s/it]
 95%|█████████▍| 9334/9864 [39:18:53<2:15:54, 15.39s/it]
 95%|█████████▍| 9335/9864 [39:19:07<2:12:24, 15.02s/it]
 95%|█████████▍| 9336/9864 [39:19:21<2:09:37, 14.73s/it]
 95%|█████████▍| 9337/9864 [39:19:38<2:13:55, 15.25s/it]
 95%|█████████▍| 9338/9864 [39:19:54<2:16:20, 15.55s/it]
 95%|█████████▍| 9339/9864 [39:20:08<2:12:48, 15.18s/it]
 95%|█████████▍| 9340/9864 [39:20:24<2:13:41, 15.31s/it]
                                                        
{'loss': '0.05014', 'grad_norm': '2.55', 'learning_rate': '1.482e-07', 'epoch': '0.947'}

 95%|█████████▍| 9340/9864 [39:20:24<2:13:41, 15.31s/it]
 95%|█████████▍| 9341/9864 [39:20:40<2:14:20, 15.41s/it]
 95%|█████████▍| 9342/9864 [39:20:54<2:11:36, 15.13s/it]
 95%|█████████▍| 9343/9864 [39:21:08<2:08:11, 14.76s/it]
 95%|█████████▍| 9344/9864 [39:21:24<2:11:51, 15.22s/it]
 95%|█████████▍| 9345/9864 [39:21:40<2:13:18, 15.41s/it]
 95%|█████████▍| 9346/9864 [39:21:54<2:10:27, 15.11s/it]
 95%|█████████▍| 9347/9864 [39:22:09<2:08:25, 14.90s/it]
 95%|█████████▍| 9348/9864 [39:22:25<2:10:31, 15.18s/it]
 95%|█████████▍| 9349/9864 [39:22:41<2:12:53, 15.48s/it]
 95%|█████████▍| 9350/9864 [39:22:55<2:08:58, 15.05s/it]
                                                        
{'loss': '0.04042', 'grad_norm': '3.433', 'learning_rate': '1.426e-07', 'epoch': '0.948'}

 95%|█████████▍| 9350/9864 [39:22:55<2:08:58, 15.05s/it]
 95%|█████████▍| 9351/9864 [39:23:09<2:06:21, 14.78s/it]
 95%|█████████▍| 9352/9864 [39:23:26<2:11:00, 15.35s/it]
 95%|█████████▍| 9353/9864 [39:23:42<2:13:10, 15.64s/it]
 95%|█████████▍| 9354/9864 [39:23:57<2:10:01, 15.30s/it]
 95%|█████████▍| 9355/9864 [39:24:11<2:07:34, 15.04s/it]
 95%|█████████▍| 9356/9864 [39:24:29<2:13:48, 15.80s/it]
 95%|█████████▍| 9357/9864 [39:24:43<2:10:13, 15.41s/it]
 95%|█████████▍| 9358/9864 [39:24:57<2:07:02, 15.06s/it]
 95%|█████████▍| 9359/9864 [39:25:14<2:10:12, 15.47s/it]
 95%|█████████▍| 9360/9864 [39:25:28<2:06:56, 15.11s/it]
                                                        
{'loss': '0.04711', 'grad_norm': '5.588', 'learning_rate': '1.372e-07', 'epoch': '0.949'}

 95%|█████████▍| 9360/9864 [39:25:28<2:06:56, 15.11s/it]
 95%|█████████▍| 9361/9864 [39:25:44<2:09:25, 15.44s/it]
 95%|█████████▍| 9362/9864 [39:25:59<2:06:40, 15.14s/it]
 95%|█████████▍| 9363/9864 [39:26:13<2:04:57, 14.97s/it]
 95%|█████████▍| 9364/9864 [39:26:31<2:10:49, 15.70s/it]
 95%|█████████▍| 9365/9864 [39:26:45<2:07:15, 15.30s/it]
 95%|█████████▍| 9366/9864 [39:27:01<2:08:03, 15.43s/it]
 95%|█████████▍| 9367/9864 [39:27:15<2:04:03, 14.98s/it]
 95%|█████████▍| 9368/9864 [39:27:31<2:06:51, 15.35s/it]
 95%|█████████▍| 9369/9864 [39:27:45<2:03:59, 15.03s/it]
 95%|█████████▍| 9370/9864 [39:28:00<2:02:01, 14.82s/it]
                                                        
{'loss': '0.04957', 'grad_norm': '5.029', 'learning_rate': '1.318e-07', 'epoch': '0.95'}

 95%|█████████▍| 9370/9864 [39:28:00<2:02:01, 14.82s/it]
 95%|█████████▌| 9371/9864 [39:28:16<2:05:20, 15.25s/it]
 95%|█████████▌| 9372/9864 [39:28:32<2:08:03, 15.62s/it]
 95%|█████████▌| 9373/9864 [39:28:47<2:04:41, 15.24s/it]
 95%|█████████▌| 9374/9864 [39:29:03<2:07:01, 15.55s/it]
 95%|█████████▌| 9375/9864 [39:29:19<2:06:55, 15.57s/it]
 95%|█████████▌| 9376/9864 [39:29:33<2:02:55, 15.11s/it]
 95%|█████████▌| 9377/9864 [39:29:46<1:59:34, 14.73s/it]
 95%|█████████▌| 9378/9864 [39:30:01<1:58:03, 14.58s/it]
 95%|█████████▌| 9379/9864 [39:30:19<2:06:20, 15.63s/it]
 95%|█████████▌| 9380/9864 [39:30:33<2:03:05, 15.26s/it]
                                                        
{'loss': '0.05071', 'grad_norm': '4.196', 'learning_rate': '1.265e-07', 'epoch': '0.951'}

 95%|█████████▌| 9380/9864 [39:30:33<2:03:05, 15.26s/it]
 95%|█████████▌| 9381/9864 [39:30:48<2:01:02, 15.04s/it]
 95%|█████████▌| 9382/9864 [39:31:04<2:02:50, 15.29s/it]
 95%|█████████▌| 9383/9864 [39:31:17<1:59:03, 14.85s/it]
 95%|█████████▌| 9384/9864 [39:31:33<2:01:27, 15.18s/it]
 95%|█████████▌| 9385/9864 [39:31:47<1:57:49, 14.76s/it]
 95%|█████████▌| 9386/9864 [39:32:02<1:56:56, 14.68s/it]
 95%|█████████▌| 9387/9864 [39:32:18<2:00:04, 15.10s/it]
 95%|█████████▌| 9388/9864 [39:32:34<2:02:16, 15.41s/it]
 95%|█████████▌| 9389/9864 [39:32:50<2:04:23, 15.71s/it]
 95%|█████████▌| 9390/9864 [39:33:04<1:59:46, 15.16s/it]
                                                        
{'loss': '0.05167', 'grad_norm': '6.555', 'learning_rate': '1.214e-07', 'epoch': '0.952'}

 95%|█████████▌| 9390/9864 [39:33:04<1:59:46, 15.16s/it]
 95%|█████████▌| 9391/9864 [39:33:19<1:58:13, 15.00s/it]
 95%|█████████▌| 9392/9864 [39:33:35<2:00:49, 15.36s/it]
 95%|█████████▌| 9393/9864 [39:33:49<1:56:52, 14.89s/it]
 95%|█████████▌| 9394/9864 [39:34:05<1:59:12, 15.22s/it]
 95%|█████████▌| 9395/9864 [39:34:19<1:56:18, 14.88s/it]
 95%|█████████▌| 9396/9864 [39:34:35<1:59:54, 15.37s/it]
 95%|█████████▌| 9397/9864 [39:34:51<2:01:22, 15.59s/it]
 95%|█████████▌| 9398/9864 [39:35:06<1:58:10, 15.22s/it]
 95%|█████████▌| 9399/9864 [39:35:22<2:00:30, 15.55s/it]
 95%|█████████▌| 9400/9864 [39:35:36<1:56:56, 15.12s/it]
                                                        
{'loss': '0.03981', 'grad_norm': '4.599', 'learning_rate': '1.163e-07', 'epoch': '0.953'}

 95%|█████████▌| 9400/9864 [39:35:36<1:56:56, 15.12s/it]
 95%|█████████▌| 9401/9864 [39:35:51<1:56:03, 15.04s/it]
 95%|█████████▌| 9402/9864 [39:36:05<1:53:54, 14.79s/it]
 95%|█████████▌| 9403/9864 [39:36:22<1:58:49, 15.47s/it]
 95%|█████████▌| 9404/9864 [39:36:36<1:55:28, 15.06s/it]
 95%|█████████▌| 9405/9864 [39:36:53<1:57:42, 15.39s/it]
 95%|█████████▌| 9406/9864 [39:37:06<1:53:01, 14.81s/it]
 95%|█████████▌| 9407/9864 [39:37:20<1:50:31, 14.51s/it]
 95%|█████████▌| 9408/9864 [39:37:36<1:53:03, 14.88s/it]
 95%|█████████▌| 9409/9864 [39:37:50<1:51:34, 14.71s/it]
 95%|█████████▌| 9410/9864 [39:38:06<1:53:50, 15.04s/it]
                                                        
{'loss': '0.04556', 'grad_norm': '5.103', 'learning_rate': '1.114e-07', 'epoch': '0.9541'}

 95%|█████████▌| 9410/9864 [39:38:06<1:53:50, 15.04s/it]
 95%|█████████▌| 9411/9864 [39:38:22<1:55:19, 15.28s/it]
 95%|█████████▌| 9412/9864 [39:38:35<1:51:20, 14.78s/it]
 95%|█████████▌| 9413/9864 [39:38:51<1:53:11, 15.06s/it]
 95%|█████████▌| 9414/9864 [39:39:05<1:51:05, 14.81s/it]
 95%|█████████▌| 9415/9864 [39:39:21<1:52:59, 15.10s/it]
 95%|█████████▌| 9416/9864 [39:39:35<1:51:17, 14.91s/it]
 95%|█████████▌| 9417/9864 [39:39:50<1:49:35, 14.71s/it]
 95%|█████████▌| 9418/9864 [39:40:08<1:57:00, 15.74s/it]
 95%|█████████▌| 9419/9864 [39:40:22<1:54:38, 15.46s/it]
 95%|█████████▌| 9420/9864 [39:40:38<1:55:26, 15.60s/it]
                                                        
{'loss': '0.04639', 'grad_norm': '6.243', 'learning_rate': '1.066e-07', 'epoch': '0.9551'}

 95%|█████████▌| 9420/9864 [39:40:38<1:55:26, 15.60s/it]
 96%|█████████▌| 9421/9864 [39:40:53<1:52:24, 15.22s/it]
 96%|█████████▌| 9422/9864 [39:41:09<1:54:20, 15.52s/it]
 96%|█████████▌| 9423/9864 [39:41:23<1:50:06, 14.98s/it]
 96%|█████████▌| 9424/9864 [39:41:37<1:47:48, 14.70s/it]
 96%|█████████▌| 9425/9864 [39:41:53<1:50:23, 15.09s/it]
 96%|█████████▌| 9426/9864 [39:42:08<1:51:24, 15.26s/it]
 96%|█████████▌| 9427/9864 [39:42:24<1:52:33, 15.45s/it]
 96%|█████████▌| 9428/9864 [39:42:39<1:49:37, 15.09s/it]
 96%|█████████▌| 9429/9864 [39:42:54<1:51:06, 15.33s/it]
 96%|█████████▌| 9430/9864 [39:43:09<1:48:20, 14.98s/it]
                                                        
{'loss': '0.04723', 'grad_norm': '3.635', 'learning_rate': '1.018e-07', 'epoch': '0.9561'}

 96%|█████████▌| 9430/9864 [39:43:09<1:48:20, 14.98s/it]
 96%|█████████▌| 9431/9864 [39:43:23<1:46:50, 14.80s/it]
 96%|█████████▌| 9432/9864 [39:43:39<1:49:48, 15.25s/it]
 96%|█████████▌| 9433/9864 [39:43:53<1:46:59, 14.89s/it]
 96%|█████████▌| 9434/9864 [39:44:10<1:51:27, 15.55s/it]
 96%|█████████▌| 9435/9864 [39:44:24<1:46:55, 14.95s/it]
 96%|█████████▌| 9436/9864 [39:44:39<1:47:48, 15.11s/it]
 96%|█████████▌| 9437/9864 [39:44:54<1:45:34, 14.83s/it]
 96%|█████████▌| 9438/9864 [39:45:09<1:45:38, 14.88s/it]
 96%|█████████▌| 9439/9864 [39:45:24<1:47:25, 15.17s/it]
 96%|█████████▌| 9440/9864 [39:45:39<1:45:51, 14.98s/it]
                                                        
{'loss': '0.03827', 'grad_norm': '4.091', 'learning_rate': '9.721e-08', 'epoch': '0.9571'}

 96%|█████████▌| 9440/9864 [39:45:39<1:45:51, 14.98s/it]
 96%|█████████▌| 9441/9864 [39:45:56<1:48:59, 15.46s/it]
 96%|█████████▌| 9442/9864 [39:46:12<1:49:40, 15.59s/it]
 96%|█████████▌| 9443/9864 [39:46:26<1:46:53, 15.23s/it]
 96%|█████████▌| 9444/9864 [39:46:42<1:48:55, 15.56s/it]
 96%|█████████▌| 9445/9864 [39:46:56<1:45:35, 15.12s/it]
 96%|█████████▌| 9446/9864 [39:47:13<1:47:49, 15.48s/it]
 96%|█████████▌| 9447/9864 [39:47:27<1:44:24, 15.02s/it]
 96%|█████████▌| 9448/9864 [39:47:41<1:43:24, 14.91s/it]
 96%|█████████▌| 9449/9864 [39:47:57<1:45:25, 15.24s/it]
 96%|█████████▌| 9450/9864 [39:48:14<1:48:06, 15.67s/it]
                                                        
{'loss': '0.04955', 'grad_norm': '4.707', 'learning_rate': '9.269e-08', 'epoch': '0.9581'}

 96%|█████████▌| 9450/9864 [39:48:14<1:48:06, 15.67s/it]
 96%|█████████▌| 9451/9864 [39:48:28<1:44:35, 15.19s/it]
 96%|█████████▌| 9452/9864 [39:48:44<1:46:20, 15.49s/it]
 96%|█████████▌| 9453/9864 [39:48:59<1:43:54, 15.17s/it]
 96%|█████████▌| 9454/9864 [39:49:13<1:41:28, 14.85s/it]
 96%|█████████▌| 9455/9864 [39:49:28<1:42:30, 15.04s/it]
 96%|█████████▌| 9456/9864 [39:49:43<1:41:05, 14.87s/it]
 96%|█████████▌| 9457/9864 [39:49:57<1:40:32, 14.82s/it]
 96%|█████████▌| 9458/9864 [39:50:15<1:46:04, 15.68s/it]
 96%|█████████▌| 9459/9864 [39:50:30<1:43:52, 15.39s/it]
 96%|█████████▌| 9460/9864 [39:50:44<1:41:38, 15.10s/it]
                                                        
{'loss': '0.04361', 'grad_norm': '4.932', 'learning_rate': '8.829e-08', 'epoch': '0.9591'}

 96%|█████████▌| 9460/9864 [39:50:44<1:41:38, 15.10s/it]
 96%|█████████▌| 9461/9864 [39:51:00<1:42:52, 15.32s/it]
 96%|█████████▌| 9462/9864 [39:51:16<1:43:39, 15.47s/it]
 96%|█████████▌| 9463/9864 [39:51:30<1:41:19, 15.16s/it]
 96%|█████████▌| 9464/9864 [39:51:44<1:39:03, 14.86s/it]
 96%|█████████▌| 9465/9864 [39:51:58<1:36:45, 14.55s/it]
 96%|█████████▌| 9466/9864 [39:52:14<1:39:20, 14.98s/it]
 96%|█████████▌| 9467/9864 [39:52:29<1:38:06, 14.83s/it]
 96%|█████████▌| 9468/9864 [39:52:45<1:41:27, 15.37s/it]
 96%|█████████▌| 9469/9864 [39:53:00<1:39:17, 15.08s/it]
 96%|█████████▌| 9470/9864 [39:53:16<1:41:15, 15.42s/it]
                                                        
{'loss': '0.03834', 'grad_norm': '4.274', 'learning_rate': '8.399e-08', 'epoch': '0.9601'}

 96%|█████████▌| 9470/9864 [39:53:16<1:41:15, 15.42s/it]
 96%|█████████▌| 9471/9864 [39:53:30<1:39:02, 15.12s/it]
 96%|█████████▌| 9472/9864 [39:53:45<1:37:22, 14.91s/it]
 96%|█████████▌| 9473/9864 [39:53:59<1:35:41, 14.68s/it]
 96%|█████████▌| 9474/9864 [39:54:13<1:34:21, 14.52s/it]
 96%|█████████▌| 9475/9864 [39:54:31<1:40:55, 15.57s/it]
 96%|█████████▌| 9476/9864 [39:54:47<1:41:21, 15.67s/it]
 96%|█████████▌| 9477/9864 [39:55:02<1:38:57, 15.34s/it]
 96%|█████████▌| 9478/9864 [39:55:17<1:39:19, 15.44s/it]
 96%|█████████▌| 9479/9864 [39:55:31<1:36:26, 15.03s/it]
 96%|█████████▌| 9480/9864 [39:55:46<1:35:39, 14.95s/it]
                                                        
{'loss': '0.04353', 'grad_norm': '6.45', 'learning_rate': '7.979e-08', 'epoch': '0.9611'}

 96%|█████████▌| 9480/9864 [39:55:46<1:35:39, 14.95s/it]
 96%|█████████▌| 9481/9864 [39:56:01<1:34:23, 14.79s/it]
 96%|█████████▌| 9482/9864 [39:56:16<1:36:17, 15.12s/it]
 96%|█████████▌| 9483/9864 [39:56:33<1:38:12, 15.47s/it]
 96%|█████████▌| 9484/9864 [39:56:48<1:37:50, 15.45s/it]
 96%|█████████▌| 9485/9864 [39:57:03<1:35:57, 15.19s/it]
 96%|█████████▌| 9486/9864 [39:57:19<1:37:18, 15.45s/it]
 96%|█████████▌| 9487/9864 [39:57:33<1:34:40, 15.07s/it]
 96%|█████████▌| 9488/9864 [39:57:47<1:32:16, 14.72s/it]
 96%|█████████▌| 9489/9864 [39:58:01<1:31:14, 14.60s/it]
 96%|█████████▌| 9490/9864 [39:58:17<1:33:21, 14.98s/it]
                                                        
{'loss': '0.0412', 'grad_norm': '4.735', 'learning_rate': '7.571e-08', 'epoch': '0.9622'}

 96%|█████████▌| 9490/9864 [39:58:17<1:33:21, 14.98s/it]
 96%|█████████▌| 9491/9864 [39:58:33<1:35:54, 15.43s/it]
 96%|█████████▌| 9492/9864 [39:58:50<1:36:57, 15.64s/it]
 96%|█████████▌| 9493/9864 [39:59:04<1:33:58, 15.20s/it]
 96%|█████████▌| 9494/9864 [39:59:18<1:32:08, 14.94s/it]
 96%|█████████▋| 9495/9864 [39:59:34<1:34:27, 15.36s/it]
 96%|█████████▋| 9496/9864 [39:59:49<1:33:02, 15.17s/it]
 96%|█████████▋| 9497/9864 [40:00:03<1:31:04, 14.89s/it]
 96%|█████████▋| 9498/9864 [40:00:19<1:32:29, 15.16s/it]
 96%|█████████▋| 9499/9864 [40:00:35<1:33:35, 15.38s/it]
 96%|█████████▋| 9500/9864 [40:00:51<1:34:01, 15.50s/it]
                                                        
{'loss': '0.04541', 'grad_norm': '4.298', 'learning_rate': '7.173e-08', 'epoch': '0.9632'}

 96%|█████████▋| 9500/9864 [40:00:51<1:34:01, 15.50s/it]
 96%|█████████▋| 9501/9864 [40:01:05<1:31:28, 15.12s/it]
 96%|█████████▋| 9502/9864 [40:01:19<1:29:26, 14.83s/it]
 96%|█████████▋| 9503/9864 [40:01:35<1:31:28, 15.20s/it]
 96%|█████████▋| 9504/9864 [40:01:49<1:28:39, 14.78s/it]
 96%|█████████▋| 9505/9864 [40:02:05<1:29:59, 15.04s/it]
 96%|█████████▋| 9506/9864 [40:02:20<1:29:15, 14.96s/it]
 96%|█████████▋| 9507/9864 [40:02:35<1:30:34, 15.22s/it]
 96%|█████████▋| 9508/9864 [40:02:52<1:32:30, 15.59s/it]
 96%|█████████▋| 9509/9864 [40:03:07<1:30:40, 15.33s/it]
 96%|█████████▋| 9510/9864 [40:03:22<1:31:14, 15.46s/it]
                                                        
{'loss': '0.05139', 'grad_norm': '6.561', 'learning_rate': '6.786e-08', 'epoch': '0.9642'}

 96%|█████████▋| 9510/9864 [40:03:22<1:31:14, 15.46s/it]
 96%|█████████▋| 9511/9864 [40:03:37<1:29:33, 15.22s/it]
 96%|█████████▋| 9512/9864 [40:03:51<1:27:27, 14.91s/it]
 96%|█████████▋| 9513/9864 [40:04:05<1:25:48, 14.67s/it]
 96%|█████████▋| 9514/9864 [40:04:22<1:29:19, 15.31s/it]
 96%|█████████▋| 9515/9864 [40:04:39<1:31:57, 15.81s/it]
 96%|█████████▋| 9516/9864 [40:04:53<1:28:34, 15.27s/it]
 96%|█████████▋| 9517/9864 [40:05:08<1:26:52, 15.02s/it]
 96%|█████████▋| 9518/9864 [40:05:22<1:25:49, 14.88s/it]
 97%|█████████▋| 9519/9864 [40:05:39<1:28:17, 15.36s/it]
 97%|█████████▋| 9520/9864 [40:05:53<1:26:20, 15.06s/it]
                                                        
{'loss': '0.04657', 'grad_norm': '3.282', 'learning_rate': '6.409e-08', 'epoch': '0.9652'}

 97%|█████████▋| 9520/9864 [40:05:53<1:26:20, 15.06s/it]
 97%|█████████▋| 9521/9864 [40:06:09<1:28:15, 15.44s/it]
 97%|█████████▋| 9522/9864 [40:06:25<1:29:05, 15.63s/it]
 97%|█████████▋| 9523/9864 [40:06:39<1:25:57, 15.12s/it]
 97%|█████████▋| 9524/9864 [40:06:55<1:26:27, 15.26s/it]
 97%|█████████▋| 9525/9864 [40:07:09<1:24:16, 14.91s/it]
 97%|█████████▋| 9526/9864 [40:07:23<1:23:03, 14.75s/it]
 97%|█████████▋| 9527/9864 [40:07:40<1:25:40, 15.25s/it]
 97%|█████████▋| 9528/9864 [40:07:54<1:23:51, 14.97s/it]
 97%|█████████▋| 9529/9864 [40:08:12<1:28:00, 15.76s/it]
 97%|█████████▋| 9530/9864 [40:08:26<1:24:56, 15.26s/it]
                                                        
{'loss': '0.05289', 'grad_norm': '4.403', 'learning_rate': '6.043e-08', 'epoch': '0.9662'}

 97%|█████████▋| 9530/9864 [40:08:26<1:24:56, 15.26s/it]
 97%|█████████▋| 9531/9864 [40:08:40<1:23:22, 15.02s/it]
 97%|█████████▋| 9532/9864 [40:08:56<1:24:38, 15.30s/it]
 97%|█████████▋| 9533/9864 [40:09:11<1:23:14, 15.09s/it]
 97%|█████████▋| 9534/9864 [40:09:27<1:25:12, 15.49s/it]
 97%|█████████▋| 9535/9864 [40:09:41<1:22:25, 15.03s/it]
 97%|█████████▋| 9536/9864 [40:09:57<1:24:00, 15.37s/it]
 97%|█████████▋| 9537/9864 [40:10:13<1:24:15, 15.46s/it]
 97%|█████████▋| 9538/9864 [40:10:27<1:22:18, 15.15s/it]
 97%|█████████▋| 9539/9864 [40:10:44<1:23:59, 15.51s/it]
 97%|█████████▋| 9540/9864 [40:10:58<1:21:41, 15.13s/it]
                                                        
{'loss': '0.04766', 'grad_norm': '3.794', 'learning_rate': '5.688e-08', 'epoch': '0.9672'}

 97%|█████████▋| 9540/9864 [40:10:58<1:21:41, 15.13s/it]
 97%|█████████▋| 9541/9864 [40:11:13<1:20:45, 15.00s/it]
 97%|█████████▋| 9542/9864 [40:11:27<1:19:46, 14.86s/it]
 97%|█████████▋| 9543/9864 [40:11:44<1:21:58, 15.32s/it]
 97%|█████████▋| 9544/9864 [40:12:00<1:23:33, 15.67s/it]
 97%|█████████▋| 9545/9864 [40:12:14<1:21:08, 15.26s/it]
 97%|█████████▋| 9546/9864 [40:12:28<1:18:32, 14.82s/it]
 97%|█████████▋| 9547/9864 [40:12:44<1:20:06, 15.16s/it]
 97%|█████████▋| 9548/9864 [40:12:59<1:18:52, 14.98s/it]
 97%|█████████▋| 9549/9864 [40:13:13<1:17:17, 14.72s/it]
 97%|█████████▋| 9550/9864 [40:13:29<1:18:50, 15.07s/it]
                                                        
{'loss': '0.04344', 'grad_norm': '3.586', 'learning_rate': '5.344e-08', 'epoch': '0.9682'}

 97%|█████████▋| 9550/9864 [40:13:29<1:18:50, 15.07s/it]
 97%|█████████▋| 9551/9864 [40:13:43<1:17:23, 14.83s/it]
 97%|█████████▋| 9552/9864 [40:14:01<1:22:35, 15.88s/it]
 97%|█████████▋| 9553/9864 [40:14:16<1:19:49, 15.40s/it]
 97%|█████████▋| 9554/9864 [40:14:29<1:16:27, 14.80s/it]
 97%|█████████▋| 9555/9864 [40:14:43<1:15:07, 14.59s/it]
 97%|█████████▋| 9556/9864 [40:14:59<1:16:21, 14.88s/it]
 97%|█████████▋| 9557/9864 [40:15:14<1:17:37, 15.17s/it]
 97%|█████████▋| 9558/9864 [40:15:28<1:15:17, 14.76s/it]
 97%|█████████▋| 9559/9864 [40:15:43<1:14:15, 14.61s/it]
 97%|█████████▋| 9560/9864 [40:15:59<1:16:26, 15.09s/it]
                                                        
{'loss': '0.05159', 'grad_norm': '3.474', 'learning_rate': '5.01e-08', 'epoch': '0.9693'}

 97%|█████████▋| 9560/9864 [40:15:59<1:16:26, 15.09s/it]
 97%|█████████▋| 9561/9864 [40:16:15<1:17:23, 15.32s/it]
 97%|█████████▋| 9562/9864 [40:16:29<1:15:15, 14.95s/it]
 97%|█████████▋| 9563/9864 [40:16:43<1:14:07, 14.78s/it]
 97%|█████████▋| 9564/9864 [40:17:00<1:17:39, 15.53s/it]
 97%|█████████▋| 9565/9864 [40:17:15<1:16:12, 15.29s/it]
 97%|█████████▋| 9566/9864 [40:17:30<1:14:54, 15.08s/it]
 97%|█████████▋| 9567/9864 [40:17:44<1:13:18, 14.81s/it]
 97%|█████████▋| 9568/9864 [40:17:58<1:11:51, 14.57s/it]
 97%|█████████▋| 9569/9864 [40:18:14<1:13:53, 15.03s/it]
 97%|█████████▋| 9570/9864 [40:18:30<1:15:06, 15.33s/it]
                                                        
{'loss': '0.04581', 'grad_norm': '2.751', 'learning_rate': '4.687e-08', 'epoch': '0.9703'}

 97%|█████████▋| 9570/9864 [40:18:30<1:15:06, 15.33s/it]
 97%|█████████▋| 9571/9864 [40:18:46<1:15:18, 15.42s/it]
 97%|█████████▋| 9572/9864 [40:19:01<1:15:05, 15.43s/it]
 97%|█████████▋| 9573/9864 [40:19:15<1:13:07, 15.08s/it]
 97%|█████████▋| 9574/9864 [40:19:30<1:11:57, 14.89s/it]
 97%|█████████▋| 9575/9864 [40:19:44<1:10:51, 14.71s/it]
 97%|█████████▋| 9576/9864 [40:19:58<1:09:54, 14.56s/it]
 97%|█████████▋| 9577/9864 [40:20:14<1:11:49, 15.01s/it]
 97%|█████████▋| 9578/9864 [40:20:30<1:12:39, 15.24s/it]
 97%|█████████▋| 9579/9864 [40:20:48<1:15:39, 15.93s/it]
 97%|█████████▋| 9580/9864 [40:21:02<1:13:17, 15.48s/it]
                                                        
{'loss': '0.04315', 'grad_norm': '3.369', 'learning_rate': '4.375e-08', 'epoch': '0.9713'}

 97%|█████████▋| 9580/9864 [40:21:02<1:13:17, 15.48s/it]
 97%|█████████▋| 9581/9864 [40:21:16<1:11:12, 15.10s/it]
 97%|█████████▋| 9582/9864 [40:21:31<1:09:57, 14.88s/it]
 97%|█████████▋| 9583/9864 [40:21:45<1:08:31, 14.63s/it]
 97%|█████████▋| 9584/9864 [40:21:59<1:07:54, 14.55s/it]
 97%|█████████▋| 9585/9864 [40:22:15<1:09:49, 15.02s/it]
 97%|█████████▋| 9586/9864 [40:22:32<1:12:40, 15.69s/it]
 97%|█████████▋| 9587/9864 [40:22:49<1:12:59, 15.81s/it]
 97%|█████████▋| 9588/9864 [40:23:03<1:11:05, 15.45s/it]
 97%|█████████▋| 9589/9864 [40:23:18<1:09:19, 15.12s/it]
 97%|█████████▋| 9590/9864 [40:23:32<1:07:59, 14.89s/it]
                                                        
{'loss': '0.0496', 'grad_norm': '3.135', 'learning_rate': '4.074e-08', 'epoch': '0.9723'}

 97%|█████████▋| 9590/9864 [40:23:32<1:07:59, 14.89s/it]
 97%|█████████▋| 9591/9864 [40:23:46<1:06:25, 14.60s/it]
 97%|█████████▋| 9592/9864 [40:24:00<1:05:30, 14.45s/it]
 97%|█████████▋| 9593/9864 [40:24:18<1:09:40, 15.43s/it]
 97%|█████████▋| 9594/9864 [40:24:31<1:07:18, 14.96s/it]
 97%|█████████▋| 9595/9864 [40:24:50<1:11:26, 15.94s/it]
 97%|█████████▋| 9596/9864 [40:25:04<1:09:24, 15.54s/it]
 97%|█████████▋| 9597/9864 [40:25:19<1:07:24, 15.15s/it]
 97%|█████████▋| 9598/9864 [40:25:33<1:06:20, 14.96s/it]
 97%|█████████▋| 9599/9864 [40:25:48<1:05:55, 14.93s/it]
 97%|█████████▋| 9600/9864 [40:26:04<1:06:59, 15.23s/it]
                                                        
{'loss': '0.04831', 'grad_norm': '4.546', 'learning_rate': '3.783e-08', 'epoch': '0.9733'}

 97%|█████████▋| 9600/9864 [40:26:04<1:06:59, 15.23s/it]
 97%|█████████▋| 9601/9864 [40:26:20<1:07:39, 15.44s/it]
 97%|█████████▋| 9602/9864 [40:26:34<1:06:20, 15.19s/it]
 97%|█████████▋| 9603/9864 [40:26:51<1:07:28, 15.51s/it]
 97%|█████████▋| 9604/9864 [40:27:07<1:08:25, 15.79s/it]
 97%|█████████▋| 9605/9864 [40:27:22<1:06:26, 15.39s/it]
 97%|█████████▋| 9606/9864 [40:27:36<1:04:19, 14.96s/it]
 97%|█████████▋| 9607/9864 [40:27:52<1:05:40, 15.33s/it]
 97%|█████████▋| 9608/9864 [40:28:06<1:04:13, 15.05s/it]
 97%|█████████▋| 9609/9864 [40:28:22<1:05:26, 15.40s/it]
 97%|█████████▋| 9610/9864 [40:28:37<1:03:42, 15.05s/it]
                                                        
{'loss': '0.04672', 'grad_norm': '3.516', 'learning_rate': '3.503e-08', 'epoch': '0.9743'}

 97%|█████████▋| 9610/9864 [40:28:37<1:03:42, 15.05s/it]
 97%|█████████▋| 9611/9864 [40:28:51<1:02:13, 14.76s/it]
 97%|█████████▋| 9612/9864 [40:29:09<1:06:09, 15.75s/it]
 97%|█████████▋| 9613/9864 [40:29:23<1:04:04, 15.32s/it]
 97%|█████████▋| 9614/9864 [40:29:37<1:02:36, 15.02s/it]
 97%|█████████▋| 9615/9864 [40:29:51<1:00:35, 14.60s/it]
 97%|█████████▋| 9616/9864 [40:30:09<1:04:42, 15.66s/it]
 97%|█████████▋| 9617/9864 [40:30:23<1:02:18, 15.13s/it]
 98%|█████████▊| 9618/9864 [40:30:37<1:00:45, 14.82s/it]
 98%|█████████▊| 9619/9864 [40:30:53<1:02:13, 15.24s/it]
 98%|█████████▊| 9620/9864 [40:31:09<1:02:48, 15.44s/it]
                                                        
{'loss': '0.05592', 'grad_norm': '4.503', 'learning_rate': '3.234e-08', 'epoch': '0.9753'}

 98%|█████████▊| 9620/9864 [40:31:09<1:02:48, 15.44s/it]
 98%|█████████▊| 9621/9864 [40:31:24<1:01:52, 15.28s/it]
 98%|█████████▊| 9622/9864 [40:31:39<1:01:01, 15.13s/it]
 98%|█████████▊| 9623/9864 [40:31:55<1:01:56, 15.42s/it]
 98%|█████████▊| 9624/9864 [40:32:11<1:02:23, 15.60s/it]
 98%|█████████▊| 9625/9864 [40:32:25<1:00:32, 15.20s/it]
 98%|█████████▊| 9626/9864 [40:32:42<1:01:36, 15.53s/it]
 98%|█████████▊| 9627/9864 [40:32:56<1:00:05, 15.21s/it]
 98%|█████████▊| 9628/9864 [40:33:11<58:58, 14.99s/it]  
 98%|█████████▊| 9629/9864 [40:33:26<59:41, 15.24s/it]
 98%|█████████▊| 9630/9864 [40:33:42<59:49, 15.34s/it]
                                                      
{'loss': '0.0556', 'grad_norm': '2.475', 'learning_rate': '2.975e-08', 'epoch': '0.9764'}

 98%|█████████▊| 9630/9864 [40:33:42<59:49, 15.34s/it]
 98%|█████████▊| 9631/9864 [40:33:56<58:11, 14.98s/it]
 98%|█████████▊| 9632/9864 [40:34:12<59:18, 15.34s/it]
 98%|█████████▊| 9633/9864 [40:34:28<59:57, 15.57s/it]
 98%|█████████▊| 9634/9864 [40:34:43<58:59, 15.39s/it]
 98%|█████████▊| 9635/9864 [40:34:57<57:14, 15.00s/it]
 98%|█████████▊| 9636/9864 [40:35:13<58:02, 15.27s/it]
 98%|█████████▊| 9637/9864 [40:35:29<58:31, 15.47s/it]
 98%|█████████▊| 9638/9864 [40:35:43<56:48, 15.08s/it]
 98%|█████████▊| 9639/9864 [40:35:58<55:31, 14.81s/it]
 98%|█████████▊| 9640/9864 [40:36:14<56:46, 15.21s/it]
                                                      
{'loss': '0.04049', 'grad_norm': '4.623', 'learning_rate': '2.728e-08', 'epoch': '0.9774'}

 98%|█████████▊| 9640/9864 [40:36:14<56:46, 15.21s/it]
 98%|█████████▊| 9641/9864 [40:36:30<57:10, 15.38s/it]
 98%|█████████▊| 9642/9864 [40:36:44<55:56, 15.12s/it]
 98%|█████████▊| 9643/9864 [40:36:58<54:40, 14.84s/it]
 98%|█████████▊| 9644/9864 [40:37:14<55:20, 15.09s/it]
 98%|█████████▊| 9645/9864 [40:37:30<56:08, 15.38s/it]
 98%|█████████▊| 9646/9864 [40:37:44<54:39, 15.04s/it]
 98%|█████████▊| 9647/9864 [40:37:59<53:51, 14.89s/it]
 98%|█████████▊| 9648/9864 [40:38:16<56:11, 15.61s/it]
 98%|█████████▊| 9649/9864 [40:38:30<53:39, 14.97s/it]
 98%|█████████▊| 9650/9864 [40:38:44<52:52, 14.83s/it]
                                                      
{'loss': '0.04694', 'grad_norm': '6.182', 'learning_rate': '2.491e-08', 'epoch': '0.9784'}

 98%|█████████▊| 9650/9864 [40:38:44<52:52, 14.83s/it]
 98%|█████████▊| 9651/9864 [40:39:00<53:49, 15.16s/it]
 98%|█████████▊| 9652/9864 [40:39:17<55:05, 15.59s/it]
 98%|█████████▊| 9653/9864 [40:39:31<53:29, 15.21s/it]
 98%|█████████▊| 9654/9864 [40:39:45<52:08, 14.90s/it]
 98%|█████████▊| 9655/9864 [40:39:59<51:10, 14.69s/it]
 98%|█████████▊| 9656/9864 [40:40:16<53:30, 15.43s/it]
 98%|█████████▊| 9657/9864 [40:40:31<52:25, 15.19s/it]
 98%|█████████▊| 9658/9864 [40:40:47<52:56, 15.42s/it]
 98%|█████████▊| 9659/9864 [40:41:01<51:25, 15.05s/it]
 98%|█████████▊| 9660/9864 [40:41:17<52:16, 15.38s/it]
                                                      
{'loss': '0.047', 'grad_norm': '6.662', 'learning_rate': '2.264e-08', 'epoch': '0.9794'}

 98%|█████████▊| 9660/9864 [40:41:17<52:16, 15.38s/it]
 98%|█████████▊| 9661/9864 [40:41:32<51:22, 15.19s/it]
 98%|█████████▊| 9662/9864 [40:41:46<49:59, 14.85s/it]
 98%|█████████▊| 9663/9864 [40:42:00<48:52, 14.59s/it]
 98%|█████████▊| 9664/9864 [40:42:18<52:21, 15.71s/it]
 98%|█████████▊| 9665/9864 [40:42:33<50:35, 15.26s/it]
 98%|█████████▊| 9666/9864 [40:42:48<50:49, 15.40s/it]
 98%|█████████▊| 9667/9864 [40:43:03<49:39, 15.12s/it]
 98%|█████████▊| 9668/9864 [40:43:17<48:40, 14.90s/it]
 98%|█████████▊| 9669/9864 [40:43:34<49:54, 15.36s/it]
 98%|█████████▊| 9670/9864 [40:43:48<48:47, 15.09s/it]
                                                      
{'loss': '0.03271', 'grad_norm': '5.01', 'learning_rate': '2.049e-08', 'epoch': '0.9804'}

 98%|█████████▊| 9670/9864 [40:43:48<48:47, 15.09s/it]
 98%|█████████▊| 9671/9864 [40:44:04<49:04, 15.26s/it]
 98%|█████████▊| 9672/9864 [40:44:20<49:36, 15.50s/it]
 98%|█████████▊| 9673/9864 [40:44:33<47:36, 14.96s/it]
 98%|█████████▊| 9674/9864 [40:44:50<48:24, 15.29s/it]
 98%|█████████▊| 9675/9864 [40:45:04<47:14, 15.00s/it]
 98%|█████████▊| 9676/9864 [40:45:18<46:15, 14.76s/it]
 98%|█████████▊| 9677/9864 [40:45:34<46:53, 15.04s/it]
 98%|█████████▊| 9678/9864 [40:45:48<46:08, 14.88s/it]
 98%|█████████▊| 9679/9864 [40:46:04<46:28, 15.07s/it]
 98%|█████████▊| 9680/9864 [40:46:19<46:39, 15.21s/it]
                                                      
{'loss': '0.05145', 'grad_norm': '4.372', 'learning_rate': '1.844e-08', 'epoch': '0.9814'}

 98%|█████████▊| 9680/9864 [40:46:19<46:39, 15.21s/it]
 98%|█████████▊| 9681/9864 [40:46:34<45:33, 14.94s/it]
 98%|█████████▊| 9682/9864 [40:46:50<46:24, 15.30s/it]
 98%|█████████▊| 9683/9864 [40:47:04<45:13, 14.99s/it]
 98%|█████████▊| 9684/9864 [40:47:19<44:42, 14.90s/it]
 98%|█████████▊| 9685/9864 [40:47:35<45:24, 15.22s/it]
 98%|█████████▊| 9686/9864 [40:47:48<43:35, 14.70s/it]
 98%|█████████▊| 9687/9864 [40:48:05<44:50, 15.20s/it]
 98%|█████████▊| 9688/9864 [40:48:21<45:28, 15.50s/it]
 98%|█████████▊| 9689/9864 [40:48:37<45:43, 15.67s/it]
 98%|█████████▊| 9690/9864 [40:48:51<44:28, 15.34s/it]
                                                      
{'loss': '0.05608', 'grad_norm': '4.276', 'learning_rate': '1.65e-08', 'epoch': '0.9824'}

 98%|█████████▊| 9690/9864 [40:48:51<44:28, 15.34s/it]
 98%|█████████▊| 9691/9864 [40:49:06<43:37, 15.13s/it]
 98%|█████████▊| 9692/9864 [40:49:22<44:16, 15.45s/it]
 98%|█████████▊| 9693/9864 [40:49:35<42:00, 14.74s/it]
 98%|█████████▊| 9694/9864 [40:49:50<41:34, 14.68s/it]
 98%|█████████▊| 9695/9864 [40:50:08<44:15, 15.71s/it]
 98%|█████████▊| 9696/9864 [40:50:22<42:27, 15.16s/it]
 98%|█████████▊| 9697/9864 [40:50:38<42:59, 15.45s/it]
 98%|█████████▊| 9698/9864 [40:50:52<41:38, 15.05s/it]
 98%|█████████▊| 9699/9864 [40:51:06<40:27, 14.71s/it]
 98%|█████████▊| 9700/9864 [40:51:22<41:07, 15.05s/it]
                                                      
{'loss': '0.03496', 'grad_norm': '2.899', 'learning_rate': '1.467e-08', 'epoch': '0.9835'}

 98%|█████████▊| 9700/9864 [40:51:22<41:07, 15.05s/it]
 98%|█████████▊| 9701/9864 [40:51:36<40:18, 14.84s/it]
 98%|█████████▊| 9702/9864 [40:51:52<40:54, 15.15s/it]
 98%|█████████▊| 9703/9864 [40:52:06<40:02, 14.92s/it]
 98%|█████████▊| 9704/9864 [40:52:23<41:06, 15.42s/it]
 98%|█████████▊| 9705/9864 [40:52:39<41:33, 15.68s/it]
 98%|█████████▊| 9706/9864 [40:52:54<40:08, 15.24s/it]
 98%|█████████▊| 9707/9864 [40:53:08<39:17, 15.01s/it]
 98%|█████████▊| 9708/9864 [40:53:24<39:33, 15.22s/it]
 98%|█████████▊| 9709/9864 [40:53:38<38:53, 15.06s/it]
 98%|█████████▊| 9710/9864 [40:53:53<38:13, 14.89s/it]
                                                      
{'loss': '0.04402', 'grad_norm': '5.253', 'learning_rate': '1.295e-08', 'epoch': '0.9845'}

 98%|█████████▊| 9710/9864 [40:53:53<38:13, 14.89s/it]
 98%|█████████▊| 9711/9864 [40:54:09<39:01, 15.30s/it]
 98%|█████████▊| 9712/9864 [40:54:25<39:22, 15.54s/it]
 98%|█████████▊| 9713/9864 [40:54:41<39:33, 15.72s/it]
 98%|█████████▊| 9714/9864 [40:54:55<38:02, 15.21s/it]
 98%|█████████▊| 9715/9864 [40:55:10<37:01, 14.91s/it]
 98%|█████████▊| 9716/9864 [40:55:25<37:14, 15.10s/it]
 99%|█████████▊| 9717/9864 [40:55:40<36:38, 14.96s/it]
 99%|█████████▊| 9718/9864 [40:55:54<35:58, 14.79s/it]
 99%|█████████▊| 9719/9864 [40:56:10<36:48, 15.23s/it]
 99%|█████████▊| 9720/9864 [40:56:27<37:24, 15.59s/it]
                                                      
{'loss': '0.04356', 'grad_norm': '4.484', 'learning_rate': '1.133e-08', 'epoch': '0.9855'}

 99%|█████████▊| 9720/9864 [40:56:27<37:24, 15.59s/it]
 99%|█████████▊| 9721/9864 [40:56:43<37:45, 15.84s/it]
 99%|█████████▊| 9722/9864 [40:56:58<36:40, 15.50s/it]
 99%|█████████▊| 9723/9864 [40:57:12<35:37, 15.16s/it]
 99%|█████████▊| 9724/9864 [40:57:29<36:17, 15.55s/it]
 99%|█████████▊| 9725/9864 [40:57:43<35:05, 15.15s/it]
 99%|█████████▊| 9726/9864 [40:57:57<33:59, 14.78s/it]
 99%|█████████▊| 9727/9864 [40:58:13<34:24, 15.07s/it]
 99%|█████████▊| 9728/9864 [40:58:29<34:40, 15.30s/it]
 99%|█████████▊| 9729/9864 [40:58:43<33:59, 15.11s/it]
 99%|█████████▊| 9730/9864 [40:58:59<34:00, 15.23s/it]
                                                      
{'loss': '0.04691', 'grad_norm': '2.456', 'learning_rate': '9.823e-09', 'epoch': '0.9865'}

 99%|█████████▊| 9730/9864 [40:58:59<34:00, 15.23s/it]
 99%|█████████▊| 9731/9864 [40:59:13<33:00, 14.89s/it]
 99%|█████████▊| 9732/9864 [40:59:29<33:45, 15.34s/it]
 99%|█████████▊| 9733/9864 [40:59:44<32:47, 15.02s/it]
 99%|█████████▊| 9734/9864 [41:00:00<33:22, 15.40s/it]
 99%|█████████▊| 9735/9864 [41:00:16<33:29, 15.58s/it]
 99%|█████████▊| 9736/9864 [41:00:30<32:29, 15.23s/it]
 99%|█████████▊| 9737/9864 [41:00:45<31:48, 15.03s/it]
 99%|█████████▊| 9738/9864 [41:01:01<32:19, 15.39s/it]
 99%|█████████▊| 9739/9864 [41:01:15<31:10, 14.97s/it]
 99%|█████████▊| 9740/9864 [41:01:31<31:37, 15.30s/it]
                                                      
{'loss': '0.05805', 'grad_norm': '4.724', 'learning_rate': '8.421e-09', 'epoch': '0.9875'}

 99%|█████████▊| 9740/9864 [41:01:31<31:37, 15.30s/it]
 99%|█████████▉| 9741/9864 [41:01:47<31:42, 15.47s/it]
 99%|█████████▉| 9742/9864 [41:02:01<30:44, 15.12s/it]
 99%|█████████▉| 9743/9864 [41:02:18<31:19, 15.53s/it]
 99%|█████████▉| 9744/9864 [41:02:32<30:37, 15.31s/it]
 99%|█████████▉| 9745/9864 [41:02:49<31:05, 15.68s/it]
 99%|█████████▉| 9746/9864 [41:03:03<29:43, 15.12s/it]
 99%|█████████▉| 9747/9864 [41:03:17<29:03, 14.90s/it]
 99%|█████████▉| 9748/9864 [41:03:33<29:17, 15.15s/it]
 99%|█████████▉| 9749/9864 [41:03:49<29:32, 15.41s/it]
 99%|█████████▉| 9750/9864 [41:04:05<29:36, 15.58s/it]
                                                      
{'loss': '0.04559', 'grad_norm': '4.236', 'learning_rate': '7.128e-09', 'epoch': '0.9885'}

 99%|█████████▉| 9750/9864 [41:04:05<29:36, 15.58s/it]
 99%|█████████▉| 9751/9864 [41:04:19<28:24, 15.08s/it]
 99%|█████████▉| 9752/9864 [41:04:32<27:13, 14.58s/it]
 99%|█████████▉| 9753/9864 [41:04:49<27:56, 15.10s/it]
 99%|█████████▉| 9754/9864 [41:05:03<27:02, 14.75s/it]
 99%|█████████▉| 9755/9864 [41:05:17<26:35, 14.63s/it]
 99%|█████████▉| 9756/9864 [41:05:34<27:26, 15.24s/it]
 99%|█████████▉| 9757/9864 [41:05:48<26:51, 15.07s/it]
 99%|█████████▉| 9758/9864 [41:06:07<28:21, 16.06s/it]
 99%|█████████▉| 9759/9864 [41:06:21<27:16, 15.59s/it]
 99%|█████████▉| 9760/9864 [41:06:37<27:09, 15.67s/it]
                                                      
{'loss': '0.04975', 'grad_norm': '6.025', 'learning_rate': '5.942e-09', 'epoch': '0.9895'}

 99%|█████████▉| 9760/9864 [41:06:37<27:09, 15.67s/it]
 99%|█████████▉| 9761/9864 [41:06:51<26:19, 15.33s/it]
 99%|█████████▉| 9762/9864 [41:07:06<25:35, 15.06s/it]
 99%|█████████▉| 9763/9864 [41:07:20<24:57, 14.82s/it]
 99%|█████████▉| 9764/9864 [41:07:36<25:19, 15.20s/it]
 99%|█████████▉| 9765/9864 [41:07:52<25:26, 15.42s/it]
 99%|█████████▉| 9766/9864 [41:08:08<25:24, 15.56s/it]
 99%|█████████▉| 9767/9864 [41:08:22<24:34, 15.20s/it]
 99%|█████████▉| 9768/9864 [41:08:38<24:39, 15.41s/it]
 99%|█████████▉| 9769/9864 [41:08:53<23:52, 15.08s/it]
 99%|█████████▉| 9770/9864 [41:09:07<23:13, 14.83s/it]
                                                      
{'loss': '0.03931', 'grad_norm': '3.763', 'learning_rate': '4.865e-09', 'epoch': '0.9906'}

 99%|█████████▉| 9770/9864 [41:09:07<23:13, 14.83s/it]
 99%|█████████▉| 9771/9864 [41:09:23<23:39, 15.26s/it]
 99%|█████████▉| 9772/9864 [41:09:39<23:45, 15.50s/it]
 99%|█████████▉| 9773/9864 [41:09:55<23:49, 15.71s/it]
 99%|█████████▉| 9774/9864 [41:10:10<22:58, 15.32s/it]
 99%|█████████▉| 9775/9864 [41:10:24<22:23, 15.10s/it]
 99%|█████████▉| 9776/9864 [41:10:40<22:32, 15.37s/it]
 99%|█████████▉| 9777/9864 [41:10:55<21:55, 15.12s/it]
 99%|█████████▉| 9778/9864 [41:11:08<20:55, 14.60s/it]
 99%|█████████▉| 9779/9864 [41:11:24<21:02, 14.86s/it]
 99%|█████████▉| 9780/9864 [41:11:40<21:21, 15.25s/it]
                                                      
{'loss': '0.03773', 'grad_norm': '3.815', 'learning_rate': '3.894e-09', 'epoch': '0.9916'}

 99%|█████████▉| 9780/9864 [41:11:40<21:21, 15.25s/it]
 99%|█████████▉| 9781/9864 [41:11:56<21:28, 15.52s/it]
 99%|█████████▉| 9782/9864 [41:12:10<20:40, 15.13s/it]
 99%|█████████▉| 9783/9864 [41:12:25<20:15, 15.01s/it]
 99%|█████████▉| 9784/9864 [41:12:41<20:31, 15.40s/it]
 99%|█████████▉| 9785/9864 [41:12:55<19:36, 14.89s/it]
 99%|█████████▉| 9786/9864 [41:13:11<19:42, 15.16s/it]
 99%|█████████▉| 9787/9864 [41:13:26<19:16, 15.02s/it]
 99%|█████████▉| 9788/9864 [41:13:43<19:59, 15.78s/it]
 99%|█████████▉| 9789/9864 [41:13:58<19:19, 15.46s/it]
 99%|█████████▉| 9790/9864 [41:14:12<18:45, 15.21s/it]
                                                      
{'loss': '0.05126', 'grad_norm': '4.096', 'learning_rate': '3.032e-09', 'epoch': '0.9926'}

 99%|█████████▉| 9790/9864 [41:14:12<18:45, 15.21s/it]
 99%|█████████▉| 9791/9864 [41:14:27<18:18, 15.05s/it]
 99%|█████████▉| 9792/9864 [41:14:43<18:13, 15.19s/it]
 99%|█████████▉| 9793/9864 [41:14:57<17:40, 14.94s/it]
 99%|█████████▉| 9794/9864 [41:15:11<17:09, 14.71s/it]
 99%|█████████▉| 9795/9864 [41:15:27<17:12, 14.96s/it]
 99%|█████████▉| 9796/9864 [41:15:45<18:02, 15.91s/it]
 99%|█████████▉| 9797/9864 [41:15:59<17:05, 15.30s/it]
 99%|█████████▉| 9798/9864 [41:16:13<16:34, 15.06s/it]
 99%|█████████▉| 9799/9864 [41:16:29<16:37, 15.35s/it]
 99%|█████████▉| 9800/9864 [41:16:44<16:11, 15.17s/it]
                                                      
{'loss': '0.03783', 'grad_norm': '5.047', 'learning_rate': '2.277e-09', 'epoch': '0.9936'}

 99%|█████████▉| 9800/9864 [41:16:44<16:11, 15.17s/it]
 99%|█████████▉| 9801/9864 [41:16:58<15:36, 14.87s/it]
 99%|█████████▉| 9802/9864 [41:17:14<15:42, 15.20s/it]
 99%|█████████▉| 9803/9864 [41:17:32<16:09, 15.89s/it]
 99%|█████████▉| 9804/9864 [41:17:46<15:19, 15.33s/it]
 99%|█████████▉| 9805/9864 [41:18:00<14:39, 14.90s/it]
 99%|█████████▉| 9806/9864 [41:18:16<14:43, 15.22s/it]
 99%|█████████▉| 9807/9864 [41:18:30<14:08, 14.88s/it]
 99%|█████████▉| 9808/9864 [41:18:44<13:46, 14.76s/it]
 99%|█████████▉| 9809/9864 [41:19:00<13:48, 15.07s/it]
 99%|█████████▉| 9810/9864 [41:19:16<13:49, 15.36s/it]
                                                      
{'loss': '0.03901', 'grad_norm': '3.583', 'learning_rate': '1.631e-09', 'epoch': '0.9946'}

 99%|█████████▉| 9810/9864 [41:19:16<13:49, 15.36s/it]
 99%|█████████▉| 9811/9864 [41:19:32<13:43, 15.54s/it]
 99%|█████████▉| 9812/9864 [41:19:46<13:04, 15.09s/it]
 99%|█████████▉| 9813/9864 [41:20:00<12:37, 14.86s/it]
 99%|█████████▉| 9814/9864 [41:20:16<12:39, 15.20s/it]
100%|█████████▉| 9815/9864 [41:20:30<12:07, 14.85s/it]
100%|█████████▉| 9816/9864 [41:20:44<11:41, 14.62s/it]
100%|█████████▉| 9817/9864 [41:21:00<11:41, 14.92s/it]
100%|█████████▉| 9818/9864 [41:21:18<12:10, 15.88s/it]
100%|█████████▉| 9819/9864 [41:21:32<11:24, 15.21s/it]
100%|█████████▉| 9820/9864 [41:21:46<10:52, 14.84s/it]
                                                      
{'loss': '0.04941', 'grad_norm': '3.852', 'learning_rate': '1.092e-09', 'epoch': '0.9956'}

100%|█████████▉| 9820/9864 [41:21:46<10:52, 14.84s/it]
100%|█████████▉| 9821/9864 [41:22:00<10:29, 14.64s/it]
100%|█████████▉| 9822/9864 [41:22:14<10:08, 14.49s/it]
100%|█████████▉| 9823/9864 [41:22:30<10:10, 14.88s/it]
100%|█████████▉| 9824/9864 [41:22:45<09:52, 14.81s/it]
100%|█████████▉| 9825/9864 [41:23:02<10:10, 15.67s/it]
100%|█████████▉| 9826/9864 [41:23:19<10:03, 15.89s/it]
100%|█████████▉| 9827/9864 [41:23:33<09:29, 15.39s/it]
100%|█████████▉| 9828/9864 [41:23:48<09:10, 15.31s/it]
100%|█████████▉| 9829/9864 [41:24:02<08:46, 15.04s/it]
100%|█████████▉| 9830/9864 [41:24:17<08:22, 14.79s/it]
                                                      
{'loss': '0.05181', 'grad_norm': '5.322', 'learning_rate': '6.603e-10', 'epoch': '0.9966'}

100%|█████████▉| 9830/9864 [41:24:17<08:22, 14.79s/it]
100%|█████████▉| 9831/9864 [41:24:33<08:20, 15.16s/it]
100%|█████████▉| 9832/9864 [41:24:48<08:09, 15.31s/it]
100%|█████████▉| 9833/9864 [41:25:02<07:41, 14.87s/it]
100%|█████████▉| 9834/9864 [41:25:20<07:51, 15.71s/it]
100%|█████████▉| 9835/9864 [41:25:34<07:24, 15.32s/it]
100%|█████████▉| 9836/9864 [41:25:48<07:00, 15.02s/it]
100%|█████████▉| 9837/9864 [41:26:02<06:34, 14.63s/it]
100%|█████████▉| 9838/9864 [41:26:17<06:21, 14.68s/it]
100%|█████████▉| 9839/9864 [41:26:33<06:14, 14.99s/it]
100%|█████████▉| 9840/9864 [41:26:47<05:52, 14.68s/it]
                                                      
{'loss': '0.04244', 'grad_norm': '3.192', 'learning_rate': '3.369e-10', 'epoch': '0.9976'}

100%|█████████▉| 9840/9864 [41:26:47<05:52, 14.68s/it]
100%|█████████▉| 9841/9864 [41:27:03<05:46, 15.08s/it]
100%|█████████▉| 9842/9864 [41:27:21<05:50, 15.92s/it]
100%|█████████▉| 9843/9864 [41:27:35<05:23, 15.39s/it]
100%|█████████▉| 9844/9864 [41:27:49<05:01, 15.06s/it]
100%|█████████▉| 9845/9864 [41:28:03<04:42, 14.86s/it]
100%|█████████▉| 9846/9864 [41:28:18<04:24, 14.69s/it]
100%|█████████▉| 9847/9864 [41:28:32<04:05, 14.44s/it]
100%|█████████▉| 9848/9864 [41:28:48<03:59, 14.98s/it]
100%|█████████▉| 9849/9864 [41:29:06<04:00, 16.01s/it]
100%|█████████▉| 9850/9864 [41:29:20<03:36, 15.45s/it]
                                                      
{'loss': '0.04076', 'grad_norm': '3.234', 'learning_rate': '1.213e-10', 'epoch': '0.9987'}

100%|█████████▉| 9850/9864 [41:29:20<03:36, 15.45s/it]
100%|█████████▉| 9851/9864 [41:29:37<03:25, 15.83s/it]
100%|█████████▉| 9852/9864 [41:29:51<03:05, 15.42s/it]
100%|█████████▉| 9853/9864 [41:30:06<02:45, 15.04s/it]
100%|█████████▉| 9854/9864 [41:30:20<02:29, 14.92s/it]
100%|█████████▉| 9855/9864 [41:30:35<02:12, 14.75s/it]
100%|█████████▉| 9856/9864 [41:30:52<02:04, 15.56s/it]
100%|█████████▉| 9857/9864 [41:31:08<01:49, 15.63s/it]
100%|█████████▉| 9858/9864 [41:31:22<01:31, 15.26s/it]
100%|█████████▉| 9859/9864 [41:31:38<01:17, 15.47s/it]
100%|█████████▉| 9860/9864 [41:31:52<01:00, 15.01s/it]
                                                      
{'loss': '0.03721', 'grad_norm': '5.038', 'learning_rate': '1.348e-11', 'epoch': '0.9997'}

100%|█████████▉| 9860/9864 [41:31:52<01:00, 15.01s/it]
100%|█████████▉| 9861/9864 [41:32:06<00:43, 14.59s/it]
100%|█████████▉| 9862/9864 [41:32:20<00:28, 14.38s/it]
100%|█████████▉| 9863/9864 [41:32:35<00:14, 14.58s/it]
100%|██████████| 9864/9864 [41:32:39<00:00, 11.41s/it]

Writing model shards:   0%|          | 0/1 [00:00<?, ?it/s]

Writing model shards: 100%|██████████| 1/1 [00:08<00:00,  8.89s/it]
Writing model shards: 100%|██████████| 1/1 [00:08<00:00,  8.89s/it]

                                                      
{'train_runtime': '1.496e+05', 'train_samples_per_second': '4.219', 'train_steps_per_second': '0.066', 'train_loss': '0.1259', 'epoch': '1'}

100%|██████████| 9864/9864 [41:33:31<00:00, 11.41s/it]
100%|██████████| 9864/9864 [41:33:31<00:00, 15.17s/it]

Writing model shards:   0%|          | 0/1 [00:00<?, ?it/s]
Writing model shards: 100%|██████████| 1/1 [00:09<00:00,  9.85s/it]
Writing model shards: 100%|██████████| 1/1 [00:09<00:00,  9.85s/it]
[rank0]:[W624 22:57:46.670833342 ProcessGroupNCCL.cpp:1476] Warning: WARNING: destroy_process_group() was not called before program exit, which can leak resources. For more info, please see https://pytorch.org/docs/stable/distributed.html#shutdown (function operator())