20250510_111525.log

[2025-05-10 11:15:25] Created output directory: train_results_pred_mask/google_gemma-3-1b-pt_ds100_upsample1000_predict_mask
[2025-05-10 11:15:25] Chat mode disabled
[2025-05-10 11:15:25] Model size is 3B or smaller (1 B). Using full fine-tuning.
[2025-05-10 11:15:25] No QA format data will be used
[2025-05-10 11:15:25] Limiting dataset size to: 100 samples
[2025-05-10 11:15:25] =======================================
[2025-05-10 11:15:25] Starting training for model: google/gemma-3-1b-pt
[2025-05-10 11:15:25] =======================================
[2025-05-10 11:15:25] CUDA_VISIBLE_DEVICES: 0,1,2,3
[2025-05-10 11:15:25] WANDB_PROJECT: wikidyk-ar
[2025-05-10 11:15:25] DATA_PATH: data/wikidyk2022-2025_01082025_gpt-4o_evalv2_pages_formatted_combined_v2.json
[2025-05-10 11:15:25] Global Batch Size: 128
[2025-05-10 11:15:25] Data Size: 100
[2025-05-10 11:15:25] Executing command: torchrun --nproc_per_node "4" --master-port 29506 src/train.py       --model_name_or_path "google/gemma-3-1b-pt"       --data_path "data/wikidyk2022-2025_01082025_gpt-4o_evalv2_pages_formatted_combined_v2.json"       --output_dir "train_results_pred_mask/google_gemma-3-1b-pt_ds100_upsample1000_predict_mask"       --num_upsample "1000"       --per_device_train_batch_size "32"       --gradient_accumulation_steps "1"       --learning_rate "2e-5"       --num_train_epochs "1"       --model_max_length "32768"       --report_to wandb --logging_steps 50 --save_strategy no       --bf16 True --use_flash_attention_2 True       --qa_data_ratio "-1"       --predict_mask "true"                            --ds_size 100
[2025-05-10 11:15:25] Training started at Sat May 10 11:15:25 UTC 2025
W0510 11:15:27.277000 361433 site-packages/torch/distributed/run.py:792] 
W0510 11:15:27.277000 361433 site-packages/torch/distributed/run.py:792] *****************************************
W0510 11:15:27.277000 361433 site-packages/torch/distributed/run.py:792] 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. 
W0510 11:15:27.277000 361433 site-packages/torch/distributed/run.py:792] *****************************************
WARNING:root:Output directory: train_results_pred_mask/google_gemma-3-1b-pt_ds100_upsample1000_predict_mask
The model was loaded with use_flash_attention_2=True, which is deprecated and may be removed in a future release. Please use `attn_implementation="flash_attention_2"` instead.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
WARNING:root:Output directory: train_results_pred_mask/google_gemma-3-1b-pt_ds100_upsample1000_predict_mask
WARNING:root:Output directory: train_results_pred_mask/google_gemma-3-1b-pt_ds100_upsample1000_predict_mask
The model was loaded with use_flash_attention_2=True, which is deprecated and may be removed in a future release. Please use `attn_implementation="flash_attention_2"` instead.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
WARNING:root:Output directory: train_results_pred_mask/google_gemma-3-1b-pt_ds100_upsample1000_predict_mask
The model was loaded with use_flash_attention_2=True, which is deprecated and may be removed in a future release. Please use `attn_implementation="flash_attention_2"` instead.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
The model was loaded with use_flash_attention_2=True, which is deprecated and may be removed in a future release. Please use `attn_implementation="flash_attention_2"` instead.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
WARNING:root:Loading data...
WARNING:root:Loading data...
WARNING:root:Loading data...
WARNING:root:Loading data...
WARNING:root:Dataset initialized with all QA data:
WARNING:root:  - 0 QA examples
WARNING:root:  - 100 fact examples with upsampling factor 1000
WARNING:root:  - Total examples: 100000
/root/WikiDYKEvalV2/src/train.py:119: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `Trainer.__init__`. Use `processing_class` instead.
  trainer = Trainer(model=model, tokenizer=tokenizer, args=training_args, **data_module)
WARNING:root:Dataset initialized with all QA data:
WARNING:root:  - 0 QA examples
WARNING:root:  - 100 fact examples with upsampling factor 1000
WARNING:root:  - Total examples: 100000
/root/WikiDYKEvalV2/src/train.py:119: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `Trainer.__init__`. Use `processing_class` instead.
  trainer = Trainer(model=model, tokenizer=tokenizer, args=training_args, **data_module)
WARNING:root:Dataset initialized with all QA data:
WARNING:root:  - 0 QA examples
WARNING:root:  - 100 fact examples with upsampling factor 1000
WARNING:root:  - Total examples: 100000
/root/WikiDYKEvalV2/src/train.py:119: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `Trainer.__init__`. Use `processing_class` instead.
  trainer = Trainer(model=model, tokenizer=tokenizer, args=training_args, **data_module)
WARNING:root:Dataset initialized with all QA data:
WARNING:root:  - 0 QA examples
WARNING:root:  - 100 fact examples with upsampling factor 1000
WARNING:root:  - Total examples: 100000
/root/WikiDYKEvalV2/src/train.py:119: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `Trainer.__init__`. Use `processing_class` instead.
  trainer = Trainer(model=model, tokenizer=tokenizer, args=training_args, **data_module)
wandb: WARNING The `run_name` is currently set to the same value as `TrainingArguments.output_dir`. If this was not intended, please specify a different run name by setting the `TrainingArguments.run_name` parameter.
It is strongly recommended to train Gemma3 models with the `eager` attention implementation instead of `flash_attention_2`. Use `eager` with `AutoModelForCausalLM.from_pretrained('<path-to-checkpoint>', attn_implementation='eager')`.
It is strongly recommended to train Gemma3 models with the `eager` attention implementation instead of `flash_attention_2`. Use `eager` with `AutoModelForCausalLM.from_pretrained('<path-to-checkpoint>', attn_implementation='eager')`.
It is strongly recommended to train Gemma3 models with the `eager` attention implementation instead of `flash_attention_2`. Use `eager` with `AutoModelForCausalLM.from_pretrained('<path-to-checkpoint>', attn_implementation='eager')`.
wandb: Currently logged in as: yuweiz to https://api.wandb.ai. Use `wandb login --relogin` to force relogin
wandb: Tracking run with wandb version 0.19.10
wandb: Run data is saved locally in /root/WikiDYKEvalV2/wandb/run-20250510_111541-7crva42l
wandb: Run `wandb offline` to turn off syncing.
wandb: Syncing run train_results_pred_mask/google_gemma-3-1b-pt_ds100_upsample1000_predict_mask
wandb: ⭐️ View project at https://wandb.ai/yuweiz/wikidyk-ar
wandb: 🚀 View run at https://wandb.ai/yuweiz/wikidyk-ar/runs/7crva42l

  0%|          | 0/782 [00:00<?, ?it/s]It is strongly recommended to train Gemma3 models with the `eager` attention implementation instead of `flash_attention_2`. Use `eager` with `AutoModelForCausalLM.from_pretrained('<path-to-checkpoint>', attn_implementation='eager')`.
[rank1]:[W510 11:15:42.657900696 reducer.cpp:1400] Warning: find_unused_parameters=True was specified in DDP constructor, but did not find any unused parameters in the forward pass. This flag results in an extra traversal of the autograd graph every iteration,  which can adversely affect performance. If your model indeed never has any unused parameters in the forward pass, consider turning this flag off. Note that this warning may be a false positive if your model has flow control causing later iterations to have unused parameters. (function operator())
[rank2]:[W510 11:15:42.662607711 reducer.cpp:1400] Warning: find_unused_parameters=True was specified in DDP constructor, but did not find any unused parameters in the forward pass. This flag results in an extra traversal of the autograd graph every iteration,  which can adversely affect performance. If your model indeed never has any unused parameters in the forward pass, consider turning this flag off. Note that this warning may be a false positive if your model has flow control causing later iterations to have unused parameters. (function operator())
[rank3]:[W510 11:15:42.675341163 reducer.cpp:1400] Warning: find_unused_parameters=True was specified in DDP constructor, but did not find any unused parameters in the forward pass. This flag results in an extra traversal of the autograd graph every iteration,  which can adversely affect performance. If your model indeed never has any unused parameters in the forward pass, consider turning this flag off. Note that this warning may be a false positive if your model has flow control causing later iterations to have unused parameters. (function operator())
[rank0]:[W510 11:15:42.736902974 reducer.cpp:1400] Warning: find_unused_parameters=True was specified in DDP constructor, but did not find any unused parameters in the forward pass. This flag results in an extra traversal of the autograd graph every iteration,  which can adversely affect performance. If your model indeed never has any unused parameters in the forward pass, consider turning this flag off. Note that this warning may be a false positive if your model has flow control causing later iterations to have unused parameters. (function operator())

  0%|          | 1/782 [00:01<17:43,  1.36s/it]
  0%|          | 2/782 [00:02<12:40,  1.03it/s]
  0%|          | 3/782 [00:02<11:14,  1.15it/s]
  1%|          | 4/782 [00:03<10:49,  1.20it/s]
  1%|          | 5/782 [00:04<10:32,  1.23it/s]
  1%|          | 6/782 [00:05<10:16,  1.26it/s]
  1%|          | 7/782 [00:05<10:06,  1.28it/s]
  1%|          | 8/782 [00:06<10:01,  1.29it/s]
  1%|          | 9/782 [00:07<09:53,  1.30it/s]
  1%|▏         | 10/782 [00:08<09:54,  1.30it/s]
  1%|▏         | 11/782 [00:08<09:53,  1.30it/s]
  2%|▏         | 12/782 [00:09<09:44,  1.32it/s]
  2%|▏         | 13/782 [00:10<09:49,  1.31it/s]
  2%|▏         | 14/782 [00:11<09:46,  1.31it/s]
  2%|▏         | 15/782 [00:11<09:44,  1.31it/s]
  2%|▏         | 16/782 [00:12<09:41,  1.32it/s]
  2%|▏         | 17/782 [00:13<09:42,  1.31it/s]
  2%|▏         | 18/782 [00:14<09:32,  1.34it/s]
  2%|▏         | 19/782 [00:14<09:38,  1.32it/s]
  3%|▎         | 20/782 [00:15<09:37,  1.32it/s]
  3%|▎         | 21/782 [00:16<09:24,  1.35it/s]
  3%|▎         | 22/782 [00:17<09:21,  1.35it/s]
  3%|▎         | 23/782 [00:17<09:26,  1.34it/s]
  3%|▎         | 24/782 [00:18<09:31,  1.33it/s]
  3%|▎         | 25/782 [00:19<10:07,  1.25it/s]
  3%|▎         | 26/782 [00:20<10:07,  1.24it/s]
  3%|▎         | 27/782 [00:21<10:01,  1.26it/s]
  4%|▎         | 28/782 [00:22<09:57,  1.26it/s]
  4%|▎         | 29/782 [00:22<09:55,  1.26it/s]
  4%|▍         | 30/782 [00:23<09:47,  1.28it/s]
  4%|▍         | 31/782 [00:24<10:49,  1.16it/s]
  4%|▍         | 32/782 [00:25<10:28,  1.19it/s]
  4%|▍         | 33/782 [00:26<10:11,  1.22it/s]
  4%|▍         | 34/782 [00:26<09:47,  1.27it/s]
  4%|▍         | 35/782 [00:27<09:43,  1.28it/s]
  5%|▍         | 36/782 [00:28<09:44,  1.28it/s]
  5%|▍         | 37/782 [00:29<09:43,  1.28it/s]
  5%|▍         | 38/782 [00:29<09:39,  1.28it/s]
  5%|▍         | 39/782 [00:30<09:34,  1.29it/s]
  5%|▌         | 40/782 [00:31<09:33,  1.29it/s]
  5%|▌         | 41/782 [00:32<09:25,  1.31it/s]
  5%|▌         | 42/782 [00:33<09:24,  1.31it/s]
  5%|▌         | 43/782 [00:33<09:25,  1.31it/s]
  6%|▌         | 44/782 [00:34<09:29,  1.30it/s]
  6%|▌         | 45/782 [00:35<09:32,  1.29it/s]
  6%|▌         | 46/782 [00:36<09:26,  1.30it/s]
  6%|▌         | 47/782 [00:36<09:31,  1.29it/s]
  6%|▌         | 48/782 [00:37<09:30,  1.29it/s]
  6%|▋         | 49/782 [00:38<09:21,  1.30it/s]
  6%|▋         | 50/782 [00:39<09:22,  1.30it/s]
                                                
{'loss': 3.1431, 'grad_norm': 8.1875, 'learning_rate': 1.874680306905371e-05, 'epoch': 0.06}

  6%|▋         | 50/782 [00:39<09:22,  1.30it/s]
  7%|▋         | 51/782 [00:39<09:18,  1.31it/s]
  7%|▋         | 52/782 [00:40<09:20,  1.30it/s]
  7%|▋         | 53/782 [00:41<09:23,  1.29it/s]
  7%|▋         | 54/782 [00:42<09:12,  1.32it/s]
  7%|▋         | 55/782 [00:43<09:15,  1.31it/s]
  7%|▋         | 56/782 [00:43<09:05,  1.33it/s]
  7%|▋         | 57/782 [00:44<09:06,  1.33it/s]
  7%|▋         | 58/782 [00:45<09:10,  1.32it/s]
  8%|▊         | 59/782 [00:46<09:10,  1.31it/s]
  8%|▊         | 60/782 [00:46<09:15,  1.30it/s]
  8%|▊         | 61/782 [00:47<09:18,  1.29it/s]
  8%|▊         | 62/782 [00:48<09:11,  1.31it/s]
  8%|▊         | 63/782 [00:49<09:04,  1.32it/s]
  8%|▊         | 64/782 [00:49<09:03,  1.32it/s]
  8%|▊         | 65/782 [00:50<09:12,  1.30it/s]
  8%|▊         | 66/782 [00:51<09:05,  1.31it/s]
  9%|▊         | 67/782 [00:52<09:06,  1.31it/s]
  9%|▊         | 68/782 [00:52<08:59,  1.32it/s]
  9%|▉         | 69/782 [00:53<08:52,  1.34it/s]
  9%|▉         | 70/782 [00:54<08:53,  1.34it/s]
  9%|▉         | 71/782 [00:55<08:55,  1.33it/s]
  9%|▉         | 72/782 [00:55<09:00,  1.31it/s]
  9%|▉         | 73/782 [00:56<09:03,  1.30it/s]
  9%|▉         | 74/782 [00:57<09:05,  1.30it/s]
 10%|▉         | 75/782 [00:58<08:57,  1.31it/s]
 10%|▉         | 76/782 [00:59<09:05,  1.29it/s]
 10%|▉         | 77/782 [00:59<09:03,  1.30it/s]
 10%|▉         | 78/782 [01:00<09:02,  1.30it/s]
 10%|█         | 79/782 [01:01<09:02,  1.30it/s]
 10%|█         | 80/782 [01:02<09:04,  1.29it/s]
 10%|█         | 81/782 [01:02<08:59,  1.30it/s]
 10%|█         | 82/782 [01:03<08:57,  1.30it/s]
 11%|█         | 83/782 [01:04<08:58,  1.30it/s]
 11%|█         | 84/782 [01:05<08:54,  1.31it/s]
 11%|█         | 85/782 [01:05<08:58,  1.29it/s]
 11%|█         | 86/782 [01:06<09:06,  1.27it/s]
 11%|█         | 87/782 [01:07<08:57,  1.29it/s]
 11%|█▏        | 88/782 [01:08<08:52,  1.30it/s]
 11%|█▏        | 89/782 [01:09<08:48,  1.31it/s]
 12%|█▏        | 90/782 [01:09<08:55,  1.29it/s]
 12%|█▏        | 91/782 [01:10<08:54,  1.29it/s]
 12%|█▏        | 92/782 [01:11<08:54,  1.29it/s]
 12%|█▏        | 93/782 [01:12<08:54,  1.29it/s]
 12%|█▏        | 94/782 [01:12<08:55,  1.29it/s]
 12%|█▏        | 95/782 [01:13<08:50,  1.30it/s]
 12%|█▏        | 96/782 [01:14<08:52,  1.29it/s]
 12%|█▏        | 97/782 [01:15<08:49,  1.29it/s]
 13%|█▎        | 98/782 [01:16<08:50,  1.29it/s]
 13%|█▎        | 99/782 [01:16<08:49,  1.29it/s]
 13%|█▎        | 100/782 [01:17<08:48,  1.29it/s]
                                                 

 13%|█▎        | 100/782 [01:17<08:48,  1.29it/s]{'loss': 0.1045, 'grad_norm': 3.90625, 'learning_rate': 1.7468030690537086e-05, 'epoch': 0.13}

 13%|█▎        | 101/782 [01:18<08:47,  1.29it/s]
 13%|█▎        | 102/782 [01:19<08:41,  1.30it/s]
 13%|█▎        | 103/782 [01:19<08:45,  1.29it/s]
 13%|█▎        | 104/782 [01:20<08:38,  1.31it/s]
 13%|█▎        | 105/782 [01:21<08:28,  1.33it/s]
 14%|█▎        | 106/782 [01:22<08:36,  1.31it/s]
 14%|█▎        | 107/782 [01:22<08:37,  1.30it/s]
 14%|█▍        | 108/782 [01:23<08:30,  1.32it/s]
 14%|█▍        | 109/782 [01:24<08:29,  1.32it/s]
 14%|█▍        | 110/782 [01:25<08:35,  1.30it/s]
 14%|█▍        | 111/782 [01:25<08:25,  1.33it/s]
 14%|█▍        | 112/782 [01:26<08:18,  1.35it/s]
 14%|█▍        | 113/782 [01:27<08:11,  1.36it/s]
 15%|█▍        | 114/782 [01:28<08:13,  1.35it/s]
 15%|█▍        | 115/782 [01:28<08:11,  1.36it/s]
 15%|█▍        | 116/782 [01:29<08:16,  1.34it/s]
 15%|█▍        | 117/782 [01:30<08:10,  1.36it/s]
 15%|█▌        | 118/782 [01:31<08:12,  1.35it/s]
 15%|█▌        | 119/782 [01:31<08:10,  1.35it/s]
 15%|█▌        | 120/782 [01:32<08:20,  1.32it/s]
 15%|█▌        | 121/782 [01:33<08:16,  1.33it/s]
 16%|█▌        | 122/782 [01:34<08:17,  1.33it/s]
 16%|█▌        | 123/782 [01:34<08:22,  1.31it/s]
 16%|█▌        | 124/782 [01:35<08:26,  1.30it/s]
 16%|█▌        | 125/782 [01:36<08:20,  1.31it/s]
 16%|█▌        | 126/782 [01:37<08:23,  1.30it/s]
 16%|█▌        | 127/782 [01:37<08:17,  1.32it/s]
 16%|█▋        | 128/782 [01:38<08:20,  1.31it/s]
 16%|█▋        | 129/782 [01:39<08:10,  1.33it/s]
 17%|█▋        | 130/782 [01:40<08:06,  1.34it/s]
 17%|█▋        | 131/782 [01:40<08:14,  1.32it/s]
 17%|█▋        | 132/782 [01:41<08:05,  1.34it/s]
 17%|█▋        | 133/782 [01:42<08:13,  1.32it/s]
 17%|█▋        | 134/782 [01:43<08:10,  1.32it/s]
 17%|█▋        | 135/782 [01:43<08:07,  1.33it/s]
 17%|█▋        | 136/782 [01:44<08:09,  1.32it/s]
 18%|█▊        | 137/782 [01:45<08:11,  1.31it/s]
 18%|█▊        | 138/782 [01:46<08:17,  1.29it/s]
 18%|█▊        | 139/782 [01:47<08:17,  1.29it/s]
 18%|█▊        | 140/782 [01:47<08:15,  1.30it/s]
 18%|█▊        | 141/782 [01:48<08:11,  1.30it/s]
 18%|█▊        | 142/782 [01:49<08:10,  1.30it/s]
 18%|█▊        | 143/782 [01:50<08:13,  1.30it/s]
 18%|█▊        | 144/782 [01:50<08:15,  1.29it/s]
 19%|█▊        | 145/782 [01:51<08:04,  1.32it/s]
 19%|█▊        | 146/782 [01:52<07:56,  1.33it/s]
 19%|█▉        | 147/782 [01:53<07:57,  1.33it/s]
 19%|█▉        | 148/782 [01:53<07:59,  1.32it/s]
 19%|█▉        | 149/782 [01:54<08:00,  1.32it/s]
 19%|█▉        | 150/782 [01:55<08:00,  1.31it/s]
                                                 
{'loss': 0.0137, 'grad_norm': 0.94921875, 'learning_rate': 1.6189258312020462e-05, 'epoch': 0.19}

 19%|█▉        | 150/782 [01:55<08:00,  1.31it/s]
 19%|█▉        | 151/782 [01:56<08:08,  1.29it/s]
 19%|█▉        | 152/782 [01:56<08:03,  1.30it/s]
 20%|█▉        | 153/782 [01:57<08:06,  1.29it/s]
 20%|█▉        | 154/782 [01:58<08:04,  1.30it/s]
 20%|█▉        | 155/782 [01:59<08:03,  1.30it/s]
 20%|█▉        | 156/782 [02:00<08:03,  1.29it/s]
 20%|██        | 157/782 [02:00<08:04,  1.29it/s]
 20%|██        | 158/782 [02:01<07:57,  1.31it/s]
 20%|██        | 159/782 [02:02<07:56,  1.31it/s]
 20%|██        | 160/782 [02:03<07:52,  1.32it/s]
 21%|██        | 161/782 [02:03<07:47,  1.33it/s]
 21%|██        | 162/782 [02:04<07:43,  1.34it/s]
 21%|██        | 163/782 [02:05<07:40,  1.34it/s]
 21%|██        | 164/782 [02:06<07:47,  1.32it/s]
 21%|██        | 165/782 [02:06<07:47,  1.32it/s]
 21%|██        | 166/782 [02:07<07:47,  1.32it/s]
 21%|██▏       | 167/782 [02:08<07:44,  1.32it/s]
 21%|██▏       | 168/782 [02:09<07:37,  1.34it/s]
 22%|██▏       | 169/782 [02:09<07:30,  1.36it/s]
 22%|██▏       | 170/782 [02:10<07:28,  1.37it/s]
 22%|██▏       | 171/782 [02:11<07:37,  1.33it/s]
 22%|██▏       | 172/782 [02:12<07:40,  1.32it/s]
 22%|██▏       | 173/782 [02:12<07:43,  1.31it/s]
 22%|██▏       | 174/782 [02:13<07:39,  1.32it/s]
 22%|██▏       | 175/782 [02:14<07:43,  1.31it/s]
 23%|██▎       | 176/782 [02:15<07:42,  1.31it/s]
 23%|██▎       | 177/782 [02:15<07:44,  1.30it/s]
 23%|██▎       | 178/782 [02:16<07:39,  1.31it/s]
 23%|██▎       | 179/782 [02:17<07:37,  1.32it/s]
 23%|██▎       | 180/782 [02:18<07:32,  1.33it/s]
 23%|██▎       | 181/782 [02:18<07:29,  1.34it/s]
 23%|██▎       | 182/782 [02:19<07:26,  1.34it/s]
 23%|██▎       | 183/782 [02:20<07:30,  1.33it/s]
 24%|██▎       | 184/782 [02:21<07:31,  1.32it/s]
 24%|██▎       | 185/782 [02:21<07:40,  1.30it/s]
 24%|██▍       | 186/782 [02:22<07:40,  1.29it/s]
 24%|██▍       | 187/782 [02:23<07:36,  1.30it/s]
 24%|██▍       | 188/782 [02:24<07:36,  1.30it/s]
 24%|██▍       | 189/782 [02:25<07:36,  1.30it/s]
 24%|██▍       | 190/782 [02:25<07:37,  1.29it/s]
 24%|██▍       | 191/782 [02:26<07:39,  1.29it/s]
 25%|██▍       | 192/782 [02:27<07:28,  1.32it/s]
 25%|██▍       | 193/782 [02:28<07:25,  1.32it/s]
 25%|██▍       | 194/782 [02:28<07:28,  1.31it/s]
 25%|██▍       | 195/782 [02:29<07:27,  1.31it/s]
 25%|██▌       | 196/782 [02:30<07:27,  1.31it/s]
 25%|██▌       | 197/782 [02:31<07:30,  1.30it/s]
 25%|██▌       | 198/782 [02:31<07:31,  1.29it/s]
 25%|██▌       | 199/782 [02:32<07:28,  1.30it/s]
 26%|██▌       | 200/782 [02:33<07:27,  1.30it/s]
                                                 
{'loss': 0.0047, 'grad_norm': 0.6484375, 'learning_rate': 1.4910485933503838e-05, 'epoch': 0.26}

 26%|██▌       | 200/782 [02:33<07:27,  1.30it/s]
 26%|██▌       | 201/782 [02:34<07:23,  1.31it/s]
 26%|██▌       | 202/782 [02:35<07:21,  1.31it/s]
 26%|██▌       | 203/782 [02:35<07:19,  1.32it/s]
 26%|██▌       | 204/782 [02:36<07:23,  1.30it/s]
 26%|██▌       | 205/782 [02:37<07:24,  1.30it/s]
 26%|██▋       | 206/782 [02:38<07:25,  1.29it/s]
 26%|██▋       | 207/782 [02:38<07:21,  1.30it/s]
 27%|██▋       | 208/782 [02:39<07:20,  1.30it/s]
 27%|██▋       | 209/782 [02:40<07:12,  1.33it/s]
 27%|██▋       | 210/782 [02:41<07:10,  1.33it/s]
 27%|██▋       | 211/782 [02:41<07:12,  1.32it/s]
 27%|██▋       | 212/782 [02:42<07:06,  1.34it/s]
 27%|██▋       | 213/782 [02:43<07:09,  1.32it/s]
 27%|██▋       | 214/782 [02:44<07:10,  1.32it/s]
 27%|██▋       | 215/782 [02:44<07:08,  1.32it/s]
 28%|██▊       | 216/782 [02:45<07:08,  1.32it/s]
 28%|██▊       | 217/782 [02:46<07:04,  1.33it/s]
 28%|██▊       | 218/782 [02:47<07:10,  1.31it/s]
 28%|██▊       | 219/782 [02:47<07:14,  1.30it/s]
 28%|██▊       | 220/782 [02:48<07:12,  1.30it/s]
 28%|██▊       | 221/782 [02:49<07:14,  1.29it/s]
 28%|██▊       | 222/782 [02:50<07:09,  1.30it/s]
 29%|██▊       | 223/782 [02:51<07:04,  1.32it/s]
 29%|██▊       | 224/782 [02:51<07:07,  1.30it/s]
 29%|██▉       | 225/782 [02:52<07:05,  1.31it/s]
 29%|██▉       | 226/782 [02:53<07:00,  1.32it/s]
 29%|██▉       | 227/782 [02:54<07:02,  1.31it/s]
 29%|██▉       | 228/782 [02:54<07:02,  1.31it/s]
 29%|██▉       | 229/782 [02:55<07:03,  1.31it/s]
 29%|██▉       | 230/782 [02:56<06:53,  1.33it/s]
 30%|██▉       | 231/782 [02:57<06:57,  1.32it/s]
 30%|██▉       | 232/782 [02:57<06:56,  1.32it/s]
 30%|██▉       | 233/782 [02:58<07:03,  1.30it/s]
 30%|██▉       | 234/782 [02:59<06:59,  1.31it/s]
 30%|███       | 235/782 [03:00<07:00,  1.30it/s]
 30%|███       | 236/782 [03:00<07:02,  1.29it/s]
 30%|███       | 237/782 [03:01<07:02,  1.29it/s]
 30%|███       | 238/782 [03:02<07:01,  1.29it/s]
 31%|███       | 239/782 [03:03<06:55,  1.31it/s]
 31%|███       | 240/782 [03:04<06:57,  1.30it/s]
 31%|███       | 241/782 [03:04<07:02,  1.28it/s]
 31%|███       | 242/782 [03:05<06:58,  1.29it/s]
 31%|███       | 243/782 [03:06<06:50,  1.31it/s]
 31%|███       | 244/782 [03:07<06:52,  1.30it/s]
 31%|███▏      | 245/782 [03:07<06:44,  1.33it/s]
 31%|███▏      | 246/782 [03:08<06:44,  1.32it/s]
 32%|███▏      | 247/782 [03:09<06:50,  1.30it/s]
 32%|███▏      | 248/782 [03:10<06:49,  1.30it/s]
 32%|███▏      | 249/782 [03:10<06:49,  1.30it/s]
 32%|███▏      | 250/782 [03:11<06:45,  1.31it/s]
                                                 
{'loss': 0.0028, 'grad_norm': 1.3984375, 'learning_rate': 1.3631713554987214e-05, 'epoch': 0.32}

 32%|███▏      | 250/782 [03:11<06:45,  1.31it/s]
 32%|███▏      | 251/782 [03:12<06:43,  1.32it/s]
 32%|███▏      | 252/782 [03:13<06:40,  1.32it/s]
 32%|███▏      | 253/782 [03:13<06:42,  1.31it/s]
 32%|███▏      | 254/782 [03:14<06:47,  1.30it/s]
 33%|███▎      | 255/782 [03:15<06:38,  1.32it/s]
 33%|███▎      | 256/782 [03:16<06:37,  1.32it/s]
 33%|███▎      | 257/782 [03:16<06:38,  1.32it/s]
 33%|███▎      | 258/782 [03:17<06:44,  1.30it/s]
 33%|███▎      | 259/782 [03:18<06:41,  1.30it/s]
 33%|███▎      | 260/782 [03:19<06:41,  1.30it/s]
 33%|███▎      | 261/782 [03:20<06:38,  1.31it/s]
 34%|███▎      | 262/782 [03:20<06:31,  1.33it/s]
 34%|███▎      | 263/782 [03:21<06:26,  1.34it/s]
 34%|███▍      | 264/782 [03:22<06:30,  1.33it/s]
 34%|███▍      | 265/782 [03:23<06:26,  1.34it/s]
 34%|███▍      | 266/782 [03:23<06:24,  1.34it/s]
 34%|███▍      | 267/782 [03:24<06:20,  1.35it/s]
 34%|███▍      | 268/782 [03:25<06:21,  1.35it/s]
 34%|███▍      | 269/782 [03:26<06:26,  1.33it/s]
 35%|███▍      | 270/782 [03:26<06:26,  1.33it/s]
 35%|███▍      | 271/782 [03:27<06:22,  1.34it/s]
 35%|███▍      | 272/782 [03:28<06:26,  1.32it/s]
 35%|███▍      | 273/782 [03:29<06:26,  1.32it/s]
 35%|███▌      | 274/782 [03:29<06:33,  1.29it/s]
 35%|███▌      | 275/782 [03:30<06:23,  1.32it/s]
 35%|███▌      | 276/782 [03:31<06:23,  1.32it/s]
 35%|███▌      | 277/782 [03:32<06:23,  1.32it/s]
 36%|███▌      | 278/782 [03:32<06:22,  1.32it/s]
 36%|███▌      | 279/782 [03:33<06:22,  1.32it/s]
 36%|███▌      | 280/782 [03:34<06:24,  1.31it/s]
 36%|███▌      | 281/782 [03:35<06:18,  1.32it/s]
 36%|███▌      | 282/782 [03:35<06:18,  1.32it/s]
 36%|███▌      | 283/782 [03:36<06:21,  1.31it/s]
 36%|███▋      | 284/782 [03:37<06:21,  1.30it/s]
 36%|███▋      | 285/782 [03:38<06:19,  1.31it/s]
 37%|███▋      | 286/782 [03:38<06:16,  1.32it/s]
 37%|███▋      | 287/782 [03:39<06:15,  1.32it/s]
 37%|███▋      | 288/782 [03:40<06:16,  1.31it/s]
 37%|███▋      | 289/782 [03:41<06:15,  1.31it/s]
 37%|███▋      | 290/782 [03:42<06:15,  1.31it/s]
 37%|███▋      | 291/782 [03:42<06:13,  1.31it/s]
 37%|███▋      | 292/782 [03:43<06:11,  1.32it/s]
 37%|███▋      | 293/782 [03:44<06:10,  1.32it/s]
 38%|███▊      | 294/782 [03:45<06:10,  1.32it/s]
 38%|███▊      | 295/782 [03:45<06:09,  1.32it/s]
 38%|███▊      | 296/782 [03:46<06:13,  1.30it/s]
 38%|███▊      | 297/782 [03:47<06:10,  1.31it/s]
 38%|███▊      | 298/782 [03:48<06:09,  1.31it/s]
 38%|███▊      | 299/782 [03:48<06:02,  1.33it/s]
 38%|███▊      | 300/782 [03:49<06:04,  1.32it/s]
                                                 
{'loss': 0.0013, 'grad_norm': 1.171875, 'learning_rate': 1.235294117647059e-05, 'epoch': 0.38}

 38%|███▊      | 300/782 [03:49<06:04,  1.32it/s]
 38%|███▊      | 301/782 [03:50<06:03,  1.32it/s]
 39%|███▊      | 302/782 [03:51<06:04,  1.32it/s]
 39%|███▊      | 303/782 [03:51<05:59,  1.33it/s]
 39%|███▉      | 304/782 [03:52<06:07,  1.30it/s]
 39%|███▉      | 305/782 [03:53<06:05,  1.30it/s]
 39%|███▉      | 306/782 [03:54<06:04,  1.30it/s]
 39%|███▉      | 307/782 [03:54<06:01,  1.32it/s]
 39%|███▉      | 308/782 [03:55<05:58,  1.32it/s]
 40%|███▉      | 309/782 [03:56<05:59,  1.32it/s]
 40%|███▉      | 310/782 [03:57<05:54,  1.33it/s]
 40%|███▉      | 311/782 [03:57<05:50,  1.34it/s]
 40%|███▉      | 312/782 [03:58<05:53,  1.33it/s]
 40%|████      | 313/782 [03:59<05:56,  1.31it/s]
 40%|████      | 314/782 [04:00<05:54,  1.32it/s]
 40%|████      | 315/782 [04:00<05:54,  1.32it/s]
 40%|████      | 316/782 [04:01<05:53,  1.32it/s]
 41%|████      | 317/782 [04:02<05:57,  1.30it/s]
 41%|████      | 318/782 [04:03<05:56,  1.30it/s]
 41%|████      | 319/782 [04:04<05:57,  1.29it/s]
 41%|████      | 320/782 [04:04<05:56,  1.30it/s]
 41%|████      | 321/782 [04:05<05:55,  1.30it/s]
 41%|████      | 322/782 [04:06<05:50,  1.31it/s]
 41%|████▏     | 323/782 [04:07<05:51,  1.30it/s]
 41%|████▏     | 324/782 [04:07<05:49,  1.31it/s]
 42%|████▏     | 325/782 [04:08<05:47,  1.31it/s]
 42%|████▏     | 326/782 [04:09<05:43,  1.33it/s]
 42%|████▏     | 327/782 [04:10<05:38,  1.34it/s]
 42%|████▏     | 328/782 [04:10<05:41,  1.33it/s]
 42%|████▏     | 329/782 [04:11<05:42,  1.32it/s]
 42%|████▏     | 330/782 [04:12<05:42,  1.32it/s]
 42%|████▏     | 331/782 [04:13<05:41,  1.32it/s]
 42%|████▏     | 332/782 [04:13<05:43,  1.31it/s]
 43%|████▎     | 333/782 [04:14<05:40,  1.32it/s]
 43%|████▎     | 334/782 [04:15<05:43,  1.30it/s]
 43%|████▎     | 335/782 [04:16<05:41,  1.31it/s]
 43%|████▎     | 336/782 [04:17<05:42,  1.30it/s]
 43%|████▎     | 337/782 [04:17<05:41,  1.30it/s]
 43%|████▎     | 338/782 [04:18<05:38,  1.31it/s]
 43%|████▎     | 339/782 [04:19<05:34,  1.32it/s]
 43%|████▎     | 340/782 [04:20<05:33,  1.32it/s]
 44%|████▎     | 341/782 [04:20<05:36,  1.31it/s]
 44%|████▎     | 342/782 [04:21<05:31,  1.33it/s]
 44%|████▍     | 343/782 [04:22<05:34,  1.31it/s]
 44%|████▍     | 344/782 [04:23<05:33,  1.31it/s]
 44%|████▍     | 345/782 [04:23<05:32,  1.31it/s]
 44%|████▍     | 346/782 [04:24<05:31,  1.31it/s]
 44%|████▍     | 347/782 [04:25<05:27,  1.33it/s]
 45%|████▍     | 348/782 [04:26<05:23,  1.34it/s]
 45%|████▍     | 349/782 [04:26<05:25,  1.33it/s]
 45%|████▍     | 350/782 [04:27<05:23,  1.34it/s]
                                                 
{'loss': 0.0007, 'grad_norm': 0.1298828125, 'learning_rate': 1.1074168797953967e-05, 'epoch': 0.45}

 45%|████▍     | 350/782 [04:27<05:23,  1.34it/s]
 45%|████▍     | 351/782 [04:28<05:30,  1.31it/s]
 45%|████▌     | 352/782 [04:29<05:29,  1.31it/s]
 45%|████▌     | 353/782 [04:29<05:28,  1.31it/s]
 45%|████▌     | 354/782 [04:30<05:28,  1.30it/s]
 45%|████▌     | 355/782 [04:31<05:22,  1.32it/s]
 46%|████▌     | 356/782 [04:32<05:21,  1.33it/s]
 46%|████▌     | 357/782 [04:32<05:20,  1.32it/s]
 46%|████▌     | 358/782 [04:33<05:17,  1.34it/s]
 46%|████▌     | 359/782 [04:34<05:17,  1.33it/s]
 46%|████▌     | 360/782 [04:35<05:16,  1.33it/s]
 46%|████▌     | 361/782 [04:35<05:15,  1.33it/s]
 46%|████▋     | 362/782 [04:36<05:14,  1.34it/s]
 46%|████▋     | 363/782 [04:37<05:12,  1.34it/s]
 47%|████▋     | 364/782 [04:38<05:09,  1.35it/s]
 47%|████▋     | 365/782 [04:38<05:14,  1.32it/s]
 47%|████▋     | 366/782 [04:39<05:14,  1.32it/s]
 47%|████▋     | 367/782 [04:40<05:15,  1.31it/s]
 47%|████▋     | 368/782 [04:41<05:15,  1.31it/s]
 47%|████▋     | 369/782 [04:41<05:17,  1.30it/s]
 47%|████▋     | 370/782 [04:42<05:13,  1.31it/s]
 47%|████▋     | 371/782 [04:43<05:13,  1.31it/s]
 48%|████▊     | 372/782 [04:44<05:14,  1.31it/s]
 48%|████▊     | 373/782 [04:45<05:13,  1.30it/s]
 48%|████▊     | 374/782 [04:45<05:13,  1.30it/s]
 48%|████▊     | 375/782 [04:46<05:13,  1.30it/s]
 48%|████▊     | 376/782 [04:47<05:11,  1.30it/s]
 48%|████▊     | 377/782 [04:48<05:11,  1.30it/s]
 48%|████▊     | 378/782 [04:48<05:07,  1.31it/s]
 48%|████▊     | 379/782 [04:49<05:09,  1.30it/s]
 49%|████▊     | 380/782 [04:50<05:03,  1.32it/s]
 49%|████▊     | 381/782 [04:51<05:05,  1.31it/s]
 49%|████▉     | 382/782 [04:51<05:06,  1.30it/s]
 49%|████▉     | 383/782 [04:52<05:03,  1.32it/s]
 49%|████▉     | 384/782 [04:53<05:04,  1.31it/s]
 49%|████▉     | 385/782 [04:54<05:04,  1.30it/s]
 49%|████▉     | 386/782 [04:55<05:07,  1.29it/s]
 49%|████▉     | 387/782 [04:55<05:08,  1.28it/s]
 50%|████▉     | 388/782 [04:56<05:04,  1.29it/s]
 50%|████▉     | 389/782 [04:57<05:03,  1.30it/s]
 50%|████▉     | 390/782 [04:58<05:00,  1.30it/s]
 50%|█████     | 391/782 [04:58<04:58,  1.31it/s]
 50%|█████     | 392/782 [04:59<04:59,  1.30it/s]
 50%|█████     | 393/782 [05:00<05:00,  1.29it/s]
 50%|█████     | 394/782 [05:01<04:57,  1.30it/s]
 51%|█████     | 395/782 [05:01<04:57,  1.30it/s]
 51%|█████     | 396/782 [05:02<04:53,  1.32it/s]
 51%|█████     | 397/782 [05:03<04:55,  1.30it/s]
 51%|█████     | 398/782 [05:04<04:55,  1.30it/s]
 51%|█████     | 399/782 [05:04<04:55,  1.30it/s]
 51%|█████     | 400/782 [05:05<04:57,  1.29it/s]
                                                 
{'loss': 0.0003, 'grad_norm': 0.037841796875, 'learning_rate': 9.795396419437341e-06, 'epoch': 0.51}

 51%|█████     | 400/782 [05:05<04:57,  1.29it/s]
 51%|█████▏    | 401/782 [05:06<04:52,  1.30it/s]
 51%|█████▏    | 402/782 [05:07<04:50,  1.31it/s]
 52%|█████▏    | 403/782 [05:08<04:52,  1.29it/s]
 52%|█████▏    | 404/782 [05:08<04:51,  1.30it/s]
 52%|█████▏    | 405/782 [05:09<04:51,  1.29it/s]
 52%|█████▏    | 406/782 [05:10<04:49,  1.30it/s]
 52%|█████▏    | 407/782 [05:11<04:48,  1.30it/s]
 52%|█████▏    | 408/782 [05:11<04:47,  1.30it/s]
 52%|█████▏    | 409/782 [05:12<04:35,  1.35it/s]
 52%|█████▏    | 410/782 [05:13<04:37,  1.34it/s]
 53%|█████▎    | 411/782 [05:14<04:41,  1.32it/s]
 53%|█████▎    | 412/782 [05:14<04:40,  1.32it/s]
 53%|█████▎    | 413/782 [05:15<04:38,  1.33it/s]
 53%|█████▎    | 414/782 [05:16<04:39,  1.32it/s]
 53%|█████▎    | 415/782 [05:17<04:37,  1.32it/s]
 53%|█████▎    | 416/782 [05:17<04:38,  1.31it/s]
 53%|█████▎    | 417/782 [05:18<04:37,  1.31it/s]
 53%|█████▎    | 418/782 [05:19<04:37,  1.31it/s]
 54%|█████▎    | 419/782 [05:20<04:39,  1.30it/s]
 54%|█████▎    | 420/782 [05:21<04:39,  1.30it/s]
 54%|█████▍    | 421/782 [05:21<04:35,  1.31it/s]
 54%|█████▍    | 422/782 [05:22<04:40,  1.29it/s]
 54%|█████▍    | 423/782 [05:23<04:35,  1.30it/s]
 54%|█████▍    | 424/782 [05:24<04:37,  1.29it/s]
 54%|█████▍    | 425/782 [05:24<04:36,  1.29it/s]
 54%|█████▍    | 426/782 [05:25<04:36,  1.29it/s]
 55%|█████▍    | 427/782 [05:26<04:34,  1.29it/s]
 55%|█████▍    | 428/782 [05:27<04:31,  1.30it/s]
 55%|█████▍    | 429/782 [05:27<04:29,  1.31it/s]
 55%|█████▍    | 430/782 [05:28<04:25,  1.32it/s]
 55%|█████▌    | 431/782 [05:29<04:24,  1.33it/s]
 55%|█████▌    | 432/782 [05:30<04:23,  1.33it/s]
 55%|█████▌    | 433/782 [05:30<04:24,  1.32it/s]
 55%|█████▌    | 434/782 [05:31<04:27,  1.30it/s]
 56%|█████▌    | 435/782 [05:32<04:26,  1.30it/s]
 56%|█████▌    | 436/782 [05:33<04:22,  1.32it/s]
 56%|█████▌    | 437/782 [05:34<04:20,  1.32it/s]
 56%|█████▌    | 438/782 [05:34<04:17,  1.34it/s]
 56%|█████▌    | 439/782 [05:35<04:17,  1.33it/s]
 56%|█████▋    | 440/782 [05:36<04:15,  1.34it/s]
 56%|█████▋    | 441/782 [05:37<04:18,  1.32it/s]
 57%|█████▋    | 442/782 [05:37<04:20,  1.31it/s]
 57%|█████▋    | 443/782 [05:38<04:18,  1.31it/s]
 57%|█████▋    | 444/782 [05:39<04:18,  1.31it/s]
 57%|█████▋    | 445/782 [05:40<04:20,  1.29it/s]
 57%|█████▋    | 446/782 [05:40<04:16,  1.31it/s]
 57%|█████▋    | 447/782 [05:41<04:17,  1.30it/s]
 57%|█████▋    | 448/782 [05:42<04:16,  1.30it/s]
 57%|█████▋    | 449/782 [05:43<04:11,  1.32it/s]
 58%|█████▊    | 450/782 [05:43<04:11,  1.32it/s]
                                                 
{'loss': 0.0002, 'grad_norm': 0.0400390625, 'learning_rate': 8.516624040920717e-06, 'epoch': 0.58}

 58%|█████▊    | 450/782 [05:43<04:11,  1.32it/s]
 58%|█████▊    | 451/782 [05:44<04:12,  1.31it/s]
 58%|█████▊    | 452/782 [05:45<04:11,  1.31it/s]
 58%|█████▊    | 453/782 [05:46<04:08,  1.32it/s]
 58%|█████▊    | 454/782 [05:46<04:07,  1.33it/s]
 58%|█████▊    | 455/782 [05:47<04:09,  1.31it/s]
 58%|█████▊    | 456/782 [05:48<04:08,  1.31it/s]
 58%|█████▊    | 457/782 [05:49<04:08,  1.31it/s]
 59%|█████▊    | 458/782 [05:49<04:02,  1.33it/s]
 59%|█████▊    | 459/782 [05:50<04:03,  1.33it/s]
 59%|█████▉    | 460/782 [05:51<04:05,  1.31it/s]
 59%|█████▉    | 461/782 [05:52<03:59,  1.34it/s]
 59%|█████▉    | 462/782 [05:52<04:02,  1.32it/s]
 59%|█████▉    | 463/782 [05:53<03:59,  1.33it/s]
 59%|█████▉    | 464/782 [05:54<04:00,  1.32it/s]
 59%|█████▉    | 465/782 [05:55<04:02,  1.31it/s]
 60%|█████▉    | 466/782 [05:56<04:00,  1.32it/s]
 60%|█████▉    | 467/782 [05:56<04:01,  1.31it/s]
 60%|█████▉    | 468/782 [05:57<03:57,  1.32it/s]
 60%|█████▉    | 469/782 [05:58<03:58,  1.31it/s]
 60%|██████    | 470/782 [05:59<03:57,  1.31it/s]
 60%|██████    | 471/782 [05:59<03:54,  1.33it/s]
 60%|██████    | 472/782 [06:00<03:52,  1.33it/s]
 60%|██████    | 473/782 [06:01<03:54,  1.32it/s]
 61%|██████    | 474/782 [06:02<03:53,  1.32it/s]
 61%|██████    | 475/782 [06:02<03:52,  1.32it/s]
 61%|██████    | 476/782 [06:03<03:48,  1.34it/s]
 61%|██████    | 477/782 [06:04<03:52,  1.31it/s]
 61%|██████    | 478/782 [06:05<03:54,  1.29it/s]
 61%|██████▏   | 479/782 [06:05<03:54,  1.29it/s]
 61%|██████▏   | 480/782 [06:06<03:55,  1.28it/s]
 62%|██████▏   | 481/782 [06:07<03:52,  1.30it/s]
 62%|██████▏   | 482/782 [06:08<03:52,  1.29it/s]
 62%|██████▏   | 483/782 [06:09<03:49,  1.30it/s]
 62%|██████▏   | 484/782 [06:09<03:49,  1.30it/s]
 62%|██████▏   | 485/782 [06:10<03:48,  1.30it/s]
 62%|██████▏   | 486/782 [06:11<03:48,  1.30it/s]
 62%|██████▏   | 487/782 [06:12<03:48,  1.29it/s]
 62%|██████▏   | 488/782 [06:12<03:43,  1.32it/s]
 63%|██████▎   | 489/782 [06:13<03:43,  1.31it/s]
 63%|██████▎   | 490/782 [06:14<03:45,  1.29it/s]
 63%|██████▎   | 491/782 [06:15<03:42,  1.31it/s]
 63%|██████▎   | 492/782 [06:15<03:41,  1.31it/s]
 63%|██████▎   | 493/782 [06:16<03:41,  1.30it/s]
 63%|██████▎   | 494/782 [06:17<03:40,  1.30it/s]
 63%|██████▎   | 495/782 [06:18<03:40,  1.30it/s]
 63%|██████▎   | 496/782 [06:19<03:42,  1.29it/s]
 64%|██████▎   | 497/782 [06:19<03:38,  1.30it/s]
 64%|██████▎   | 498/782 [06:20<03:39,  1.30it/s]
 64%|██████▍   | 499/782 [06:21<03:39,  1.29it/s]
 64%|██████▍   | 500/782 [06:22<03:38,  1.29it/s]
                                                 
{'loss': 0.0002, 'grad_norm': 0.050537109375, 'learning_rate': 7.237851662404093e-06, 'epoch': 0.64}

 64%|██████▍   | 500/782 [06:22<03:38,  1.29it/s]
 64%|██████▍   | 501/782 [06:22<03:35,  1.31it/s]
 64%|██████▍   | 502/782 [06:23<03:33,  1.31it/s]
 64%|██████▍   | 503/782 [06:24<03:32,  1.31it/s]
 64%|██████▍   | 504/782 [06:25<03:30,  1.32it/s]
 65%|██████▍   | 505/782 [06:25<03:31,  1.31it/s]
 65%|██████▍   | 506/782 [06:26<03:30,  1.31it/s]
 65%|██████▍   | 507/782 [06:27<03:30,  1.31it/s]
 65%|██████▍   | 508/782 [06:28<03:28,  1.31it/s]
 65%|██████▌   | 509/782 [06:28<03:27,  1.32it/s]
 65%|██████▌   | 510/782 [06:29<03:25,  1.32it/s]
 65%|██████▌   | 511/782 [06:30<03:26,  1.31it/s]
 65%|██████▌   | 512/782 [06:31<03:25,  1.31it/s]
 66%|██████▌   | 513/782 [06:31<03:24,  1.31it/s]
 66%|██████▌   | 514/782 [06:32<03:23,  1.31it/s]
 66%|██████▌   | 515/782 [06:33<03:22,  1.32it/s]
 66%|██████▌   | 516/782 [06:34<03:22,  1.31it/s]
 66%|██████▌   | 517/782 [06:35<03:21,  1.32it/s]
 66%|██████▌   | 518/782 [06:35<03:21,  1.31it/s]
 66%|██████▋   | 519/782 [06:36<03:20,  1.31it/s]
 66%|██████▋   | 520/782 [06:37<03:19,  1.31it/s]
 67%|██████▋   | 521/782 [06:38<03:21,  1.30it/s]
 67%|██████▋   | 522/782 [06:38<03:19,  1.31it/s]
 67%|██████▋   | 523/782 [06:39<03:17,  1.31it/s]
 67%|██████▋   | 524/782 [06:40<03:16,  1.31it/s]
 67%|██████▋   | 525/782 [06:41<03:12,  1.34it/s]
 67%|██████▋   | 526/782 [06:41<03:10,  1.34it/s]
 67%|██████▋   | 527/782 [06:42<03:12,  1.33it/s]
 68%|██████▊   | 528/782 [06:43<03:13,  1.31it/s]
 68%|██████▊   | 529/782 [06:44<03:12,  1.31it/s]
 68%|██████▊   | 530/782 [06:44<03:08,  1.34it/s]
 68%|██████▊   | 531/782 [06:45<03:09,  1.33it/s]
 68%|██████▊   | 532/782 [06:46<03:08,  1.33it/s]
 68%|██████▊   | 533/782 [06:47<03:08,  1.32it/s]
 68%|██████▊   | 534/782 [06:47<03:09,  1.31it/s]
 68%|██████▊   | 535/782 [06:48<03:09,  1.30it/s]
 69%|██████▊   | 536/782 [06:49<03:11,  1.29it/s]
 69%|██████▊   | 537/782 [06:50<03:07,  1.31it/s]
 69%|██████▉   | 538/782 [06:51<03:07,  1.30it/s]
 69%|██████▉   | 539/782 [06:51<03:06,  1.30it/s]
 69%|██████▉   | 540/782 [06:52<03:05,  1.30it/s]
 69%|██████▉   | 541/782 [06:53<03:05,  1.30it/s]
 69%|██████▉   | 542/782 [06:54<03:05,  1.29it/s]
 69%|██████▉   | 543/782 [06:54<03:04,  1.30it/s]
 70%|██████▉   | 544/782 [06:55<03:01,  1.31it/s]
 70%|██████▉   | 545/782 [06:56<02:59,  1.32it/s]
 70%|██████▉   | 546/782 [06:57<03:02,  1.29it/s]
 70%|██████▉   | 547/782 [06:57<03:03,  1.28it/s]
 70%|███████   | 548/782 [06:58<03:03,  1.28it/s]
 70%|███████   | 549/782 [06:59<03:02,  1.28it/s]
 70%|███████   | 550/782 [07:00<02:57,  1.31it/s]
                                                 
{'loss': 0.0002, 'grad_norm': 0.05517578125, 'learning_rate': 5.959079283887469e-06, 'epoch': 0.7}

 70%|███████   | 550/782 [07:00<02:57,  1.31it/s]
 70%|███████   | 551/782 [07:01<02:56,  1.31it/s]
 71%|███████   | 552/782 [07:01<02:54,  1.32it/s]
 71%|███████   | 553/782 [07:02<02:54,  1.31it/s]
 71%|███████   | 554/782 [07:03<02:55,  1.30it/s]
 71%|███████   | 555/782 [07:04<02:54,  1.30it/s]
 71%|███████   | 556/782 [07:04<02:53,  1.31it/s]
 71%|███████   | 557/782 [07:05<02:51,  1.31it/s]
 71%|███████▏  | 558/782 [07:06<02:52,  1.30it/s]
 71%|███████▏  | 559/782 [07:07<02:47,  1.33it/s]
 72%|███████▏  | 560/782 [07:07<02:47,  1.32it/s]
 72%|███████▏  | 561/782 [07:08<02:47,  1.32it/s]
 72%|███████▏  | 562/782 [07:09<02:46,  1.32it/s]
 72%|███████▏  | 563/782 [07:10<02:45,  1.32it/s]
 72%|███████▏  | 564/782 [07:10<02:44,  1.32it/s]
 72%|███████▏  | 565/782 [07:11<02:44,  1.32it/s]
 72%|███████▏  | 566/782 [07:12<02:43,  1.32it/s]
 73%|███████▎  | 567/782 [07:13<02:44,  1.31it/s]
 73%|███████▎  | 568/782 [07:13<02:44,  1.30it/s]
 73%|███████▎  | 569/782 [07:14<02:42,  1.31it/s]
 73%|███████▎  | 570/782 [07:15<02:39,  1.33it/s]
 73%|███████▎  | 571/782 [07:16<02:38,  1.33it/s]
 73%|███████▎  | 572/782 [07:16<02:40,  1.31it/s]
 73%|███████▎  | 573/782 [07:17<02:39,  1.31it/s]
 73%|███████▎  | 574/782 [07:18<02:36,  1.33it/s]
 74%|███████▎  | 575/782 [07:19<02:33,  1.35it/s]
 74%|███████▎  | 576/782 [07:19<02:34,  1.33it/s]
 74%|███████▍  | 577/782 [07:20<02:35,  1.32it/s]
 74%|███████▍  | 578/782 [07:21<02:31,  1.34it/s]
 74%|███████▍  | 579/782 [07:22<02:32,  1.33it/s]
 74%|███████▍  | 580/782 [07:23<02:35,  1.30it/s]
 74%|███████▍  | 581/782 [07:23<02:35,  1.29it/s]
 74%|███████▍  | 582/782 [07:24<02:32,  1.31it/s]
 75%|███████▍  | 583/782 [07:25<02:32,  1.31it/s]
 75%|███████▍  | 584/782 [07:26<02:30,  1.31it/s]
 75%|███████▍  | 585/782 [07:26<02:32,  1.29it/s]
 75%|███████▍  | 586/782 [07:27<02:31,  1.29it/s]
 75%|███████▌  | 587/782 [07:28<02:28,  1.31it/s]
 75%|███████▌  | 588/782 [07:29<02:29,  1.30it/s]
 75%|███████▌  | 589/782 [07:29<02:29,  1.29it/s]
 75%|███████▌  | 590/782 [07:30<02:25,  1.32it/s]
 76%|███████▌  | 591/782 [07:31<02:23,  1.33it/s]
 76%|███████▌  | 592/782 [07:32<02:23,  1.33it/s]
 76%|███████▌  | 593/782 [07:32<02:24,  1.31it/s]
 76%|███████▌  | 594/782 [07:33<02:23,  1.31it/s]
 76%|███████▌  | 595/782 [07:34<02:21,  1.33it/s]
 76%|███████▌  | 596/782 [07:35<02:21,  1.32it/s]
 76%|███████▋  | 597/782 [07:35<02:20,  1.32it/s]
 76%|███████▋  | 598/782 [07:36<02:20,  1.31it/s]
 77%|███████▋  | 599/782 [07:37<02:21,  1.29it/s]
 77%|███████▋  | 600/782 [07:38<02:17,  1.32it/s]
                                                 
{'loss': 0.0002, 'grad_norm': 0.0390625, 'learning_rate': 4.6803069053708444e-06, 'epoch': 0.77}

 77%|███████▋  | 600/782 [07:38<02:17,  1.32it/s]
 77%|███████▋  | 601/782 [07:39<02:16,  1.32it/s]
 77%|███████▋  | 602/782 [07:39<02:18,  1.30it/s]
 77%|███████▋  | 603/782 [07:40<02:18,  1.29it/s]
 77%|███████▋  | 604/782 [07:41<02:17,  1.29it/s]
 77%|███████▋  | 605/782 [07:42<02:14,  1.32it/s]
 77%|███████▋  | 606/782 [07:42<02:14,  1.31it/s]
 78%|███████▊  | 607/782 [07:43<02:13,  1.31it/s]
 78%|███████▊  | 608/782 [07:44<02:11,  1.32it/s]
 78%|███████▊  | 609/782 [07:45<02:12,  1.30it/s]
 78%|███████▊  | 610/782 [07:45<02:12,  1.29it/s]
 78%|███████▊  | 611/782 [07:46<02:11,  1.30it/s]
 78%|███████▊  | 612/782 [07:47<02:11,  1.30it/s]
 78%|███████▊  | 613/782 [07:48<02:10,  1.30it/s]
 79%|███████▊  | 614/782 [07:49<02:09,  1.30it/s]
 79%|███████▊  | 615/782 [07:49<02:06,  1.32it/s]
 79%|███████▉  | 616/782 [07:50<02:06,  1.31it/s]
 79%|███████▉  | 617/782 [07:51<02:06,  1.31it/s]
 79%|███████▉  | 618/782 [07:52<02:04,  1.31it/s]
 79%|███████▉  | 619/782 [07:52<02:03,  1.32it/s]
 79%|███████▉  | 620/782 [07:53<02:05,  1.29it/s]
 79%|███████▉  | 621/782 [07:54<02:03,  1.30it/s]
 80%|███████▉  | 622/782 [07:55<02:04,  1.29it/s]
 80%|███████▉  | 623/782 [07:55<02:04,  1.28it/s]
 80%|███████▉  | 624/782 [07:56<02:02,  1.29it/s]
 80%|███████▉  | 625/782 [07:57<02:01,  1.29it/s]
 80%|████████  | 626/782 [07:58<01:59,  1.30it/s]
 80%|████████  | 627/782 [07:58<01:58,  1.31it/s]
 80%|████████  | 628/782 [07:59<01:57,  1.31it/s]
 80%|████████  | 629/782 [08:00<01:55,  1.32it/s]
 81%|████████  | 630/782 [08:01<01:56,  1.31it/s]
 81%|████████  | 631/782 [08:02<01:55,  1.30it/s]
 81%|████████  | 632/782 [08:02<01:55,  1.30it/s]
 81%|████████  | 633/782 [08:03<01:54,  1.30it/s]
 81%|████████  | 634/782 [08:04<01:53,  1.30it/s]
 81%|████████  | 635/782 [08:05<01:51,  1.32it/s]
 81%|████████▏ | 636/782 [08:05<01:49,  1.33it/s]
 81%|████████▏ | 637/782 [08:06<01:48,  1.34it/s]
 82%|████████▏ | 638/782 [08:07<01:48,  1.33it/s]
 82%|████████▏ | 639/782 [08:08<01:46,  1.35it/s]
 82%|████████▏ | 640/782 [08:08<01:47,  1.32it/s]
 82%|████████▏ | 641/782 [08:09<01:47,  1.31it/s]
 82%|████████▏ | 642/782 [08:10<01:47,  1.31it/s]
 82%|████████▏ | 643/782 [08:11<01:48,  1.29it/s]
 82%|████████▏ | 644/782 [08:11<01:46,  1.30it/s]
 82%|████████▏ | 645/782 [08:12<01:45,  1.30it/s]
 83%|████████▎ | 646/782 [08:13<01:44,  1.30it/s]
 83%|████████▎ | 647/782 [08:14<01:44,  1.30it/s]
 83%|████████▎ | 648/782 [08:15<01:43,  1.30it/s]
 83%|████████▎ | 649/782 [08:15<01:40,  1.33it/s]
 83%|████████▎ | 650/782 [08:16<01:41,  1.30it/s]
                                                 
{'loss': 0.0002, 'grad_norm': 0.035888671875, 'learning_rate': 3.4015345268542205e-06, 'epoch': 0.83}

 83%|████████▎ | 650/782 [08:16<01:41,  1.30it/s]
 83%|████████▎ | 651/782 [08:17<01:40,  1.30it/s]
 83%|████████▎ | 652/782 [08:18<01:39,  1.31it/s]
 84%|████████▎ | 653/782 [08:18<01:38,  1.31it/s]
 84%|████████▎ | 654/782 [08:19<01:35,  1.34it/s]
 84%|████████▍ | 655/782 [08:20<01:35,  1.33it/s]
 84%|████████▍ | 656/782 [08:21<01:33,  1.34it/s]
 84%|████████▍ | 657/782 [08:21<01:34,  1.33it/s]
 84%|████████▍ | 658/782 [08:22<01:32,  1.35it/s]
 84%|████████▍ | 659/782 [08:23<01:32,  1.33it/s]
 84%|████████▍ | 660/782 [08:24<01:32,  1.33it/s]
 85%|████████▍ | 661/782 [08:24<01:31,  1.32it/s]
 85%|████████▍ | 662/782 [08:25<01:31,  1.31it/s]
 85%|████████▍ | 663/782 [08:26<01:31,  1.31it/s]
 85%|████████▍ | 664/782 [08:27<01:30,  1.31it/s]
 85%|████████▌ | 665/782 [08:27<01:30,  1.29it/s]
 85%|████████▌ | 666/782 [08:28<01:29,  1.30it/s]
 85%|████████▌ | 667/782 [08:29<01:27,  1.31it/s]
 85%|████████▌ | 668/782 [08:30<01:25,  1.34it/s]
 86%|████████▌ | 669/782 [08:30<01:24,  1.33it/s]
 86%|████████▌ | 670/782 [08:31<01:24,  1.33it/s]
 86%|████████▌ | 671/782 [08:32<01:22,  1.34it/s]
 86%|████████▌ | 672/782 [08:33<01:23,  1.32it/s]
 86%|████████▌ | 673/782 [08:33<01:22,  1.32it/s]
 86%|████████▌ | 674/782 [08:34<01:22,  1.32it/s]
 86%|████████▋ | 675/782 [08:35<01:20,  1.33it/s]
 86%|████████▋ | 676/782 [08:36<01:19,  1.33it/s]
 87%|████████▋ | 677/782 [08:36<01:20,  1.31it/s]
 87%|████████▋ | 678/782 [08:37<01:20,  1.29it/s]
 87%|████████▋ | 679/782 [08:38<01:20,  1.29it/s]
 87%|████████▋ | 680/782 [08:39<01:19,  1.28it/s]
 87%|████████▋ | 681/782 [08:40<01:17,  1.31it/s]
 87%|████████▋ | 682/782 [08:40<01:16,  1.31it/s]
 87%|████████▋ | 683/782 [08:41<01:15,  1.31it/s]
 87%|████████▋ | 684/782 [08:42<01:15,  1.29it/s]
 88%|████████▊ | 685/782 [08:43<01:14,  1.30it/s]
 88%|████████▊ | 686/782 [08:43<01:15,  1.28it/s]
 88%|████████▊ | 687/782 [08:44<01:13,  1.29it/s]
 88%|████████▊ | 688/782 [08:45<01:12,  1.29it/s]
 88%|████████▊ | 689/782 [08:46<01:10,  1.31it/s]
 88%|████████▊ | 690/782 [08:46<01:10,  1.31it/s]
 88%|████████▊ | 691/782 [08:47<01:10,  1.30it/s]
 88%|████████▊ | 692/782 [08:48<01:09,  1.30it/s]
 89%|████████▊ | 693/782 [08:49<01:08,  1.30it/s]
 89%|████████▊ | 694/782 [08:50<01:17,  1.13it/s]
 89%|████████▉ | 695/782 [08:51<01:13,  1.19it/s]
 89%|████████▉ | 696/782 [08:51<01:10,  1.22it/s]
 89%|████████▉ | 697/782 [08:52<01:08,  1.25it/s]
 89%|████████▉ | 698/782 [08:53<01:07,  1.25it/s]
 89%|████████▉ | 699/782 [08:54<01:04,  1.28it/s]
 90%|████████▉ | 700/782 [08:55<01:04,  1.28it/s]
                                                 
{'loss': 0.0002, 'grad_norm': 0.055908203125, 'learning_rate': 2.122762148337596e-06, 'epoch': 0.9}

 90%|████████▉ | 700/782 [08:55<01:04,  1.28it/s]
 90%|████████▉ | 701/782 [08:55<01:02,  1.29it/s]
 90%|████████▉ | 702/782 [08:56<01:00,  1.31it/s]
 90%|████████▉ | 703/782 [08:57<01:00,  1.31it/s]
 90%|█████████ | 704/782 [08:58<00:59,  1.31it/s]
 90%|█████████ | 705/782 [08:58<00:59,  1.29it/s]
 90%|█████████ | 706/782 [08:59<00:58,  1.30it/s]
 90%|█████████ | 707/782 [09:00<00:56,  1.32it/s]
 91%|█████████ | 708/782 [09:01<00:56,  1.30it/s]
 91%|█████████ | 709/782 [09:01<00:55,  1.31it/s]
 91%|█████████ | 710/782 [09:02<00:55,  1.30it/s]
 91%|█████████ | 711/782 [09:03<00:54,  1.29it/s]
 91%|█████████ | 712/782 [09:04<00:54,  1.29it/s]
 91%|█████████ | 713/782 [09:05<00:53,  1.30it/s]
 91%|█████████▏| 714/782 [09:05<00:51,  1.32it/s]
 91%|█████████▏| 715/782 [09:06<00:51,  1.31it/s]
 92%|█████████▏| 716/782 [09:07<00:49,  1.33it/s]
 92%|█████████▏| 717/782 [09:08<00:49,  1.32it/s]
 92%|█████████▏| 718/782 [09:08<00:48,  1.32it/s]
 92%|█████████▏| 719/782 [09:09<00:48,  1.31it/s]
 92%|█████████▏| 720/782 [09:10<00:47,  1.32it/s]
 92%|█████████▏| 721/782 [09:11<00:46,  1.32it/s]
 92%|█████████▏| 722/782 [09:11<00:45,  1.33it/s]
 92%|█████████▏| 723/782 [09:12<00:43,  1.34it/s]
 93%|█████████▎| 724/782 [09:13<00:44,  1.32it/s]
 93%|█████████▎| 725/782 [09:14<00:43,  1.30it/s]
 93%|█████████▎| 726/782 [09:14<00:42,  1.31it/s]
 93%|█████████▎| 727/782 [09:15<00:42,  1.30it/s]
 93%|█████████▎| 728/782 [09:16<00:41,  1.29it/s]
 93%|█████████▎| 729/782 [09:17<00:41,  1.28it/s]
 93%|█████████▎| 730/782 [09:17<00:39,  1.31it/s]
 93%|█████████▎| 731/782 [09:18<00:38,  1.31it/s]
 94%|█████████▎| 732/782 [09:19<00:37,  1.34it/s]
 94%|█████████▎| 733/782 [09:20<00:36,  1.33it/s]
 94%|█████████▍| 734/782 [09:20<00:36,  1.33it/s]
 94%|█████████▍| 735/782 [09:21<00:35,  1.34it/s]
 94%|█████████▍| 736/782 [09:22<00:34,  1.33it/s]
 94%|█████████▍| 737/782 [09:23<00:33,  1.34it/s]
 94%|█████████▍| 738/782 [09:23<00:33,  1.33it/s]
 95%|█████████▍| 739/782 [09:24<00:31,  1.35it/s]
 95%|█████████▍| 740/782 [09:25<00:31,  1.35it/s]
 95%|█████████▍| 741/782 [09:26<00:30,  1.35it/s]
 95%|█████████▍| 742/782 [09:26<00:29,  1.36it/s]
 95%|█████████▌| 743/782 [09:27<00:29,  1.33it/s]
 95%|█████████▌| 744/782 [09:28<00:28,  1.33it/s]
 95%|█████████▌| 745/782 [09:29<00:28,  1.31it/s]
 95%|█████████▌| 746/782 [09:29<00:27,  1.30it/s]
 96%|█████████▌| 747/782 [09:30<00:26,  1.30it/s]
 96%|█████████▌| 748/782 [09:31<00:25,  1.31it/s]
 96%|█████████▌| 749/782 [09:32<00:25,  1.32it/s]
 96%|█████████▌| 750/782 [09:32<00:24,  1.31it/s]
                                                 
{'loss': 0.0002, 'grad_norm': 0.04638671875, 'learning_rate': 8.439897698209719e-07, 'epoch': 0.96}

 96%|█████████▌| 750/782 [09:33<00:24,  1.31it/s]
 96%|█████████▌| 751/782 [09:33<00:23,  1.31it/s]
 96%|█████████▌| 752/782 [09:34<00:22,  1.31it/s]
 96%|█████████▋| 753/782 [09:35<00:21,  1.32it/s]
 96%|█████████▋| 754/782 [09:36<00:21,  1.31it/s]
 97%|█████████▋| 755/782 [09:36<00:20,  1.32it/s]
 97%|█████████▋| 756/782 [09:37<00:19,  1.31it/s]
 97%|█████████▋| 757/782 [09:38<00:19,  1.31it/s]
 97%|█████████▋| 758/782 [09:39<00:18,  1.31it/s]
 97%|█████████▋| 759/782 [09:39<00:17,  1.31it/s]
 97%|█████████▋| 760/782 [09:40<00:16,  1.31it/s]
 97%|█████████▋| 761/782 [09:41<00:15,  1.32it/s]
 97%|█████████▋| 762/782 [09:42<00:15,  1.30it/s]
 98%|█████████▊| 763/782 [09:42<00:14,  1.31it/s]
 98%|█████████▊| 764/782 [09:43<00:13,  1.30it/s]
 98%|█████████▊| 765/782 [09:44<00:13,  1.30it/s]
 98%|█████████▊| 766/782 [09:45<00:12,  1.30it/s]
 98%|█████████▊| 767/782 [09:45<00:11,  1.31it/s]
 98%|█████████▊| 768/782 [09:46<00:10,  1.29it/s]
 98%|█████████▊| 769/782 [09:47<00:10,  1.28it/s]
 98%|█████████▊| 770/782 [09:48<00:09,  1.28it/s]
 99%|█████████▊| 771/782 [09:49<00:08,  1.29it/s]
 99%|█████████▊| 772/782 [09:49<00:07,  1.29it/s]
 99%|█████████▉| 773/782 [09:50<00:06,  1.32it/s]
 99%|█████████▉| 774/782 [09:51<00:06,  1.30it/s]
 99%|█████████▉| 775/782 [09:52<00:05,  1.30it/s]
 99%|█████████▉| 776/782 [09:52<00:04,  1.31it/s]
 99%|█████████▉| 777/782 [09:53<00:03,  1.31it/s]
 99%|█████████▉| 778/782 [09:54<00:03,  1.30it/s]
100%|█████████▉| 779/782 [09:55<00:02,  1.30it/s]
100%|█████████▉| 780/782 [09:55<00:01,  1.31it/s]
100%|█████████▉| 781/782 [09:56<00:00,  1.30it/s]
100%|██████████| 782/782 [09:57<00:00,  1.31it/s]
                                                 
{'train_runtime': 598.6221, 'train_samples_per_second': 167.05, 'train_steps_per_second': 1.306, 'train_loss': 0.20923083342607027, 'epoch': 1.0}

100%|██████████| 782/782 [09:57<00:00,  1.31it/s]
100%|██████████| 782/782 [09:57<00:00,  1.31it/s]

model.safetensors:   0%|          | 0.00/2.00G [00:00<?, ?B/s]

tokenizer.model:   0%|          | 0.00/4.69M [00:00<?, ?B/s]


Upload 3 LFS files:   0%|          | 0/3 [00:00<?, ?it/s]



training_args.bin:   0%|          | 0.00/5.43k [00:00<?, ?B/s]
training_args.bin: 100%|██████████| 5.43k/5.43k [00:00<00:00, 69.7kB/s]

model.safetensors:   0%|          | 2.56M/2.00G [00:00<01:18, 25.5MB/s]

tokenizer.model:   6%|▌         | 279k/4.69M [00:00<00:01, 2.73MB/s]
model.safetensors:   1%|          | 14.6M/2.00G [00:00<00:24, 81.3MB/s]
tokenizer.model: 100%|██████████| 4.69M/4.69M [00:00<00:00, 14.0MB/s]

model.safetensors:   1%|          | 22.7M/2.00G [00:00<00:41, 48.1MB/s]
model.safetensors:   2%|▏         | 32.0M/2.00G [00:00<00:45, 43.1MB/s]
model.safetensors:   2%|▏         | 48.0M/2.00G [00:01<01:05, 29.8MB/s]
model.safetensors:   3%|▎         | 64.0M/2.00G [00:01<00:51, 37.3MB/s]
model.safetensors:   4%|▍         | 78.1M/2.00G [00:01<00:38, 49.5MB/s]
model.safetensors:   4%|▍         | 85.4M/2.00G [00:02<00:43, 44.5MB/s]
model.safetensors:   5%|▍         | 96.0M/2.00G [00:02<00:44, 43.2MB/s]
model.safetensors:   6%|▌         | 110M/2.00G [00:02<00:32, 57.5MB/s] 
model.safetensors:   6%|▌         | 118M/2.00G [00:02<00:38, 48.3MB/s]
model.safetensors:   6%|▋         | 128M/2.00G [00:02<00:42, 43.7MB/s]
model.safetensors:   7%|▋         | 142M/2.00G [00:03<00:32, 57.3MB/s]
model.safetensors:   7%|▋         | 150M/2.00G [00:03<00:54, 33.9MB/s]
model.safetensors:   8%|▊         | 160M/2.00G [00:03<00:53, 34.6MB/s]
model.safetensors:   9%|▊         | 174M/2.00G [00:03<00:38, 47.4MB/s]
model.safetensors:   9%|▉         | 181M/2.00G [00:04<00:40, 44.4MB/s]
model.safetensors:  10%|▉         | 192M/2.00G [00:04<00:40, 44.9MB/s]
model.safetensors:  10%|█         | 208M/2.00G [00:04<00:37, 47.8MB/s]
model.safetensors:  11%|█         | 222M/2.00G [00:04<00:29, 60.5MB/s]
model.safetensors:  11%|█▏        | 230M/2.00G [00:05<00:34, 51.4MB/s]
model.safetensors:  12%|█▏        | 240M/2.00G [00:05<00:38, 46.1MB/s]
model.safetensors:  13%|█▎        | 254M/2.00G [00:05<00:28, 61.2MB/s]
model.safetensors:  13%|█▎        | 263M/2.00G [00:05<00:33, 51.2MB/s]
model.safetensors:  14%|█▎        | 272M/2.00G [00:05<00:37, 45.8MB/s]
model.safetensors:  14%|█▍        | 285M/2.00G [00:06<00:28, 59.7MB/s]
model.safetensors:  15%|█▍        | 293M/2.00G [00:06<00:35, 47.7MB/s]
model.safetensors:  15%|█▌        | 304M/2.00G [00:06<00:35, 48.0MB/s]
model.safetensors:  16%|█▌        | 320M/2.00G [00:06<00:32, 52.3MB/s]
model.safetensors:  17%|█▋        | 334M/2.00G [00:06<00:25, 65.7MB/s]
model.safetensors:  17%|█▋        | 343M/2.00G [00:07<00:30, 53.7MB/s]
model.safetensors:  18%|█▊        | 352M/2.00G [00:07<00:33, 49.6MB/s]
model.safetensors:  18%|█▊        | 368M/2.00G [00:07<00:24, 67.4MB/s]
model.safetensors:  19%|█▉        | 377M/2.00G [00:07<00:31, 50.9MB/s]
model.safetensors:  19%|█▉        | 384M/2.00G [00:08<00:36, 44.4MB/s]
model.safetensors:  20%|██        | 400M/2.00G [00:08<00:32, 50.0MB/s]
model.safetensors:  21%|██        | 413M/2.00G [00:08<00:26, 60.7MB/s]
model.safetensors:  21%|██        | 421M/2.00G [00:08<00:32, 49.3MB/s]
model.safetensors:  22%|██▏       | 432M/2.00G [00:08<00:32, 48.2MB/s]
model.safetensors:  22%|██▏       | 447M/2.00G [00:09<00:23, 65.1MB/s]
model.safetensors:  23%|██▎       | 456M/2.00G [00:09<00:27, 55.5MB/s]
model.safetensors:  23%|██▎       | 464M/2.00G [00:09<00:31, 49.4MB/s]
model.safetensors:  24%|██▍       | 480M/2.00G [00:09<00:29, 51.3MB/s]
model.safetensors:  25%|██▍       | 496M/2.00G [00:10<00:28, 53.4MB/s]
model.safetensors:  26%|██▌       | 512M/2.00G [00:10<00:28, 52.0MB/s]
model.safetensors:  26%|██▋       | 528M/2.00G [00:10<00:26, 55.2MB/s]
model.safetensors:  27%|██▋       | 544M/2.00G [00:10<00:25, 57.0MB/s]
model.safetensors:  28%|██▊       | 560M/2.00G [00:11<00:33, 43.2MB/s]
model.safetensors:  29%|██▉       | 576M/2.00G [00:11<00:29, 47.6MB/s]
model.safetensors:  30%|██▉       | 592M/2.00G [00:11<00:27, 51.1MB/s]
model.safetensors:  30%|███       | 608M/2.00G [00:12<00:28, 48.2MB/s]
model.safetensors:  31%|███       | 624M/2.00G [00:12<00:27, 49.8MB/s]
model.safetensors:  32%|███▏      | 640M/2.00G [00:12<00:27, 49.1MB/s]
model.safetensors:  33%|███▎      | 655M/2.00G [00:13<00:22, 60.4MB/s]
model.safetensors:  33%|███▎      | 663M/2.00G [00:13<00:25, 52.1MB/s]
model.safetensors:  34%|███▎      | 672M/2.00G [00:13<00:26, 50.3MB/s]
model.safetensors:  34%|███▍      | 688M/2.00G [00:13<00:19, 66.6MB/s]
model.safetensors:  35%|███▍      | 697M/2.00G [00:13<00:23, 54.7MB/s]
model.safetensors:  35%|███▌      | 704M/2.00G [00:14<00:29, 43.9MB/s]
model.safetensors:  36%|███▌      | 720M/2.00G [00:14<00:27, 46.4MB/s]
model.safetensors:  37%|███▋      | 736M/2.00G [00:14<00:25, 48.8MB/s]
model.safetensors:  38%|███▊      | 752M/2.00G [00:15<00:26, 47.3MB/s]
model.safetensors:  38%|███▊      | 768M/2.00G [00:15<00:26, 47.0MB/s]
model.safetensors:  39%|███▉      | 784M/2.00G [00:15<00:24, 49.9MB/s]
model.safetensors:  40%|████      | 800M/2.00G [00:16<00:27, 43.9MB/s]
model.safetensors:  41%|████      | 816M/2.00G [00:16<00:24, 48.1MB/s]
model.safetensors:  42%|████▏     | 832M/2.00G [00:16<00:23, 49.4MB/s]
model.safetensors:  42%|████▏     | 848M/2.00G [00:17<00:21, 53.5MB/s]
model.safetensors:  43%|████▎     | 864M/2.00G [00:17<00:24, 46.9MB/s]
model.safetensors:  44%|████▍     | 880M/2.00G [00:17<00:22, 50.3MB/s]
model.safetensors:  45%|████▍     | 896M/2.00G [00:18<00:20, 52.9MB/s]
model.safetensors:  46%|████▌     | 912M/2.00G [00:18<00:19, 55.2MB/s]
model.safetensors:  46%|████▋     | 928M/2.00G [00:18<00:23, 46.0MB/s]
model.safetensors:  47%|████▋     | 944M/2.00G [00:19<00:21, 49.7MB/s]
model.safetensors:  48%|████▊     | 956M/2.00G [00:19<00:17, 58.2MB/s]
model.safetensors:  48%|████▊     | 964M/2.00G [00:19<00:19, 52.4MB/s]
model.safetensors:  49%|████▉     | 976M/2.00G [00:19<00:20, 49.7MB/s]
model.safetensors:  50%|████▉     | 992M/2.00G [00:19<00:20, 48.8MB/s]
model.safetensors:  50%|█████     | 1.01G/2.00G [00:20<00:19, 51.0MB/s]
model.safetensors:  51%|█████     | 1.02G/2.00G [00:20<00:18, 52.7MB/s]
model.safetensors:  52%|█████▏    | 1.04G/2.00G [00:20<00:17, 55.8MB/s]
model.safetensors:  53%|█████▎    | 1.06G/2.00G [00:21<00:16, 56.9MB/s]
model.safetensors:  54%|█████▎    | 1.07G/2.00G [00:21<00:13, 70.5MB/s]
model.safetensors:  54%|█████▍    | 1.08G/2.00G [00:21<00:16, 55.9MB/s]
model.safetensors:  54%|█████▍    | 1.09G/2.00G [00:21<00:18, 48.7MB/s]
model.safetensors:  55%|█████▌    | 1.10G/2.00G [00:21<00:15, 56.3MB/s]
model.safetensors:  56%|█████▌    | 1.12G/2.00G [00:22<00:14, 58.8MB/s]
model.safetensors:  57%|█████▋    | 1.14G/2.00G [00:22<00:15, 56.9MB/s]
model.safetensors:  58%|█████▊    | 1.15G/2.00G [00:22<00:12, 70.1MB/s]
model.safetensors:  58%|█████▊    | 1.16G/2.00G [00:22<00:14, 59.6MB/s]
model.safetensors:  58%|█████▊    | 1.17G/2.00G [00:23<00:16, 50.8MB/s]
model.safetensors:  59%|█████▉    | 1.18G/2.00G [00:23<00:15, 51.1MB/s]
model.safetensors:  60%|██████    | 1.20G/2.00G [00:23<00:14, 54.0MB/s]
model.safetensors:  61%|██████    | 1.21G/2.00G [00:23<00:11, 67.4MB/s]
model.safetensors:  61%|██████    | 1.22G/2.00G [00:23<00:13, 57.1MB/s]
model.safetensors:  62%|██████▏   | 1.23G/2.00G [00:24<00:14, 52.4MB/s]
model.safetensors:  62%|██████▏   | 1.25G/2.00G [00:24<00:13, 57.7MB/s]
model.safetensors:  63%|██████▎   | 1.26G/2.00G [00:24<00:12, 60.6MB/s]
model.safetensors:  64%|██████▍   | 1.28G/2.00G [00:24<00:09, 73.8MB/s]
model.safetensors:  64%|██████▍   | 1.29G/2.00G [00:24<00:11, 59.5MB/s]
model.safetensors:  65%|██████▍   | 1.30G/2.00G [00:25<00:13, 52.9MB/s]
model.safetensors:  66%|██████▌   | 1.31G/2.00G [00:25<00:09, 69.1MB/s]
model.safetensors:  66%|██████▌   | 1.32G/2.00G [00:25<00:12, 55.8MB/s]
model.safetensors:  66%|██████▋   | 1.33G/2.00G [00:25<00:14, 46.1MB/s]
model.safetensors:  67%|██████▋   | 1.34G/2.00G [00:26<00:12, 51.5MB/s]
model.safetensors:  68%|██████▊   | 1.36G/2.00G [00:26<00:11, 54.7MB/s]
model.safetensors:  69%|██████▉   | 1.38G/2.00G [00:26<00:11, 55.0MB/s]
model.safetensors:  70%|██████▉   | 1.39G/2.00G [00:26<00:10, 57.8MB/s]
model.safetensors:  70%|███████   | 1.41G/2.00G [00:27<00:10, 57.6MB/s]
model.safetensors:  71%|███████   | 1.42G/2.00G [00:27<00:10, 57.5MB/s]
model.safetensors:  72%|███████▏  | 1.44G/2.00G [00:27<00:09, 59.1MB/s]
model.safetensors:  73%|███████▎  | 1.46G/2.00G [00:27<00:08, 61.0MB/s]
model.safetensors:  74%|███████▎  | 1.47G/2.00G [00:28<00:08, 59.0MB/s]
model.safetensors:  74%|███████▍  | 1.49G/2.00G [00:28<00:08, 59.4MB/s]
model.safetensors:  75%|███████▌  | 1.50G/2.00G [00:28<00:08, 57.6MB/s]
model.safetensors:  76%|███████▌  | 1.52G/2.00G [00:29<00:08, 56.4MB/s]
model.safetensors:  77%|███████▋  | 1.54G/2.00G [00:29<00:07, 58.0MB/s]
model.safetensors:  78%|███████▊  | 1.55G/2.00G [00:29<00:07, 59.0MB/s]
model.safetensors:  78%|███████▊  | 1.57G/2.00G [00:29<00:07, 60.2MB/s]
model.safetensors:  79%|███████▉  | 1.58G/2.00G [00:30<00:07, 56.5MB/s]
model.safetensors:  80%|████████  | 1.60G/2.00G [00:30<00:06, 57.6MB/s]
model.safetensors:  81%|████████  | 1.62G/2.00G [00:30<00:06, 58.6MB/s]
model.safetensors:  82%|████████▏ | 1.63G/2.00G [00:30<00:06, 59.4MB/s]
model.safetensors:  82%|████████▏ | 1.65G/2.00G [00:31<00:05, 60.1MB/s]
model.safetensors:  83%|████████▎ | 1.66G/2.00G [00:31<00:05, 62.9MB/s]
model.safetensors:  84%|████████▍ | 1.68G/2.00G [00:31<00:05, 58.1MB/s]
model.safetensors:  85%|████████▍ | 1.70G/2.00G [00:31<00:04, 71.0MB/s]
model.safetensors:  85%|████████▌ | 1.70G/2.00G [00:32<00:05, 57.7MB/s]
model.safetensors:  86%|████████▌ | 1.71G/2.00G [00:32<00:06, 44.1MB/s]
model.safetensors:  86%|████████▋ | 1.73G/2.00G [00:32<00:05, 47.6MB/s]
model.safetensors:  87%|████████▋ | 1.74G/2.00G [00:33<00:05, 50.1MB/s]
model.safetensors:  88%|████████▊ | 1.76G/2.00G [00:33<00:04, 53.7MB/s]
model.safetensors:  89%|████████▉ | 1.78G/2.00G [00:34<00:05, 37.5MB/s]
model.safetensors:  90%|████████▉ | 1.79G/2.00G [00:34<00:04, 41.7MB/s]
model.safetensors:  90%|█████████ | 1.81G/2.00G [00:34<00:04, 47.7MB/s]
model.safetensors:  91%|█████████ | 1.82G/2.00G [00:34<00:03, 58.3MB/s]
model.safetensors:  92%|█████████▏| 1.83G/2.00G [00:34<00:03, 48.4MB/s]
model.safetensors:  92%|█████████▏| 1.84G/2.00G [00:35<00:03, 47.0MB/s]
model.safetensors:  93%|█████████▎| 1.86G/2.00G [00:35<00:02, 52.8MB/s]
model.safetensors:  94%|█████████▎| 1.87G/2.00G [00:35<00:02, 53.9MB/s]
model.safetensors:  94%|█████████▍| 1.89G/2.00G [00:35<00:01, 56.8MB/s]
model.safetensors:  95%|█████████▌| 1.90G/2.00G [00:36<00:01, 53.3MB/s]
model.safetensors:  96%|█████████▌| 1.92G/2.00G [00:36<00:01, 54.4MB/s]
model.safetensors:  97%|█████████▋| 1.94G/2.00G [00:36<00:01, 58.0MB/s]
model.safetensors:  98%|█████████▊| 1.95G/2.00G [00:37<00:00, 57.8MB/s]
model.safetensors:  98%|█████████▊| 1.97G/2.00G [00:37<00:00, 62.3MB/s]
model.safetensors:  99%|█████████▉| 1.98G/2.00G [00:37<00:00, 61.9MB/s]
model.safetensors: 100%|██████████| 2.00G/2.00G [00:37<00:00, 52.8MB/s]



Upload 3 LFS files:  33%|███▎      | 1/3 [00:38<01:16, 38.06s/it]
Upload 3 LFS files: 100%|██████████| 3/3 [00:38<00:00, 12.69s/it]