nohup: ignoring input
Starting run with config: mario/grid_search_path_gemma.yaml
Rodando
🦥 Unsloth: Will patch your computer to enable 2x faster free finetuning.
🦥 Unsloth Zoo will now patch everything to make training faster!
Loaded 17071 levels from evaluation dataset
Starting grid search...

===== Processing Split: str_window_vertical_newline =====

===== Processing Subset: path_snake =====

Generating str_window split:   0%|          | 0/17071 [00:00<?, ? examples/s]
Generating str_window split: 100%|██████████| 17071/17071 [00:00<00:00, 350502.32 examples/s]

Generating str_window_bar split:   0%|          | 0/17071 [00:00<?, ? examples/s]
Generating str_window_bar split: 100%|██████████| 17071/17071 [00:00<00:00, 529107.65 examples/s]

Generating str_window_newline split:   0%|          | 0/17071 [00:00<?, ? examples/s]
Generating str_window_newline split: 100%|██████████| 17071/17071 [00:00<00:00, 541201.54 examples/s]

Generating str_window_vertical split:   0%|          | 0/17071 [00:00<?, ? examples/s]
Generating str_window_vertical split: 100%|██████████| 17071/17071 [00:00<00:00, 648447.85 examples/s]

Generating str_window_vertical_bar split:   0%|          | 0/17071 [00:00<?, ? examples/s]
Generating str_window_vertical_bar split: 100%|██████████| 17071/17071 [00:00<00:00, 532151.35 examples/s]

Generating str_window_vertical_newline split:   0%|          | 0/17071 [00:00<?, ? examples/s]
Generating str_window_vertical_newline split: 100%|██████████| 17071/17071 [00:00<00:00, 557301.36 examples/s]

Generating str_window_single split:   0%|          | 0/17071 [00:00<?, ? examples/s]
Generating str_window_single split:  94%|█████████▎| 16000/17071 [00:00<00:00, 154160.91 examples/s]
Generating str_window_single split: 100%|██████████| 17071/17071 [00:00<00:00, 160098.75 examples/s]

Generating str_window_single_vertical split:   0%|          | 0/17071 [00:00<?, ? examples/s]
Generating str_window_single_vertical split: 100%|██████████| 17071/17071 [00:00<00:00, 188130.01 examples/s]

Generating str_window_single_newline split:   0%|          | 0/17071 [00:00<?, ? examples/s]
Generating str_window_single_newline split: 100%|██████████| 17071/17071 [00:00<00:00, 262379.35 examples/s]

Generating str_window_single_vertical_newline split:   0%|          | 0/17071 [00:00<?, ? examples/s]
Generating str_window_single_vertical_newline split: 100%|██████████| 17071/17071 [00:00<00:00, 391369.03 examples/s]
Dataset Loaded for str_window_vertical_newline - Subset: path_snake

Unsloth: Standardizing formats (num_proc=224):   0%|          | 0/17071 [00:00<?, ? examples/s]
Unsloth: Standardizing formats (num_proc=224):   0%|          | 77/17071 [00:00<00:57, 293.34 examples/s]
Unsloth: Standardizing formats (num_proc=224):  27%|██▋       | 4682/17071 [00:00<00:00, 16403.60 examples/s]
Unsloth: Standardizing formats (num_proc=224):  56%|█████▌    | 9547/17071 [00:00<00:00, 27148.90 examples/s]
Unsloth: Standardizing formats (num_proc=224):  78%|███████▊  | 13271/17071 [00:00<00:00, 27819.40 examples/s]
Unsloth: Standardizing formats (num_proc=224):  98%|█████████▊| 16691/17071 [00:00<00:00, 26843.17 examples/s]
Unsloth: Standardizing formats (num_proc=224): 100%|██████████| 17071/17071 [00:00<00:00, 19670.93 examples/s]

--- Processing Model Family: gemma3 ---

- Model: unsloth/gemma-3-12b-it-bnb-4bit
- LoRA Pair: r=128, alpha=128
- Other Params: {'num_train_epochs': 1, 'train_on_responses_only': True}
Run Name: mario_sft_dataset_numbers_v2.1_str_window_vertical_newline-gemma-3-12b-it--resp-r128-a128-e1-1353
==((====))==  Unsloth 2025.4.7: Fast Gemma3 patching. Transformers: 4.51.3.
   \\   /|    NVIDIA H100 80GB HBM3. Num GPUs = 1. Max memory: 79.189 GB. Platform: Linux.
O^O/ \_/ \    Torch: 2.6.0+cu124. CUDA: 9.0. CUDA Toolkit: 12.4. Triton: 3.2.0
\        /    Bfloat16 = TRUE. FA [Xformers = 0.0.29.post3. FA2 = False]
 "-____-"     Free license: http://github.com/unslothai/unsloth
Unsloth: Fast downloading is enabled - ignore downloading bars which are red colored!

Fetching 2 files:   0%|          | 0/2 [00:00<?, ?it/s]
Fetching 2 files:  50%|█████     | 1/2 [02:55<02:55, 175.30s/it]
Fetching 2 files: 100%|██████████| 2/2 [02:55<00:00, 87.65s/it] 

Loading checkpoint shards:   0%|          | 0/2 [00:00<?, ?it/s]
Loading checkpoint shards:  50%|█████     | 1/2 [00:02<00:02,  2.90s/it]
Loading checkpoint shards: 100%|██████████| 2/2 [00:04<00:00,  1.90s/it]
Loading checkpoint shards: 100%|██████████| 2/2 [00:04<00:00,  2.05s/it]
Using a slow image processor as `use_fast` is unset and a slow processor was saved with this model. `use_fast=True` will be the default behavior in v4.52, even if the model was saved with a slow processor. This will result in minor differences in outputs. You'll still be able to use a slow processor with `use_fast=False`.
Unsloth: Making `model.base_model.model.language_model.model` require gradients

Unsloth: Standardizing formats (num_proc=224):   0%|          | 0/17071 [00:00<?, ? examples/s]
Unsloth: Standardizing formats (num_proc=224):   0%|          | 77/17071 [00:00<01:12, 235.60 examples/s]
Unsloth: Standardizing formats (num_proc=224):  27%|██▋       | 4605/17071 [00:00<00:00, 13885.25 examples/s]
Unsloth: Standardizing formats (num_proc=224):  46%|████▌     | 7798/17071 [00:00<00:00, 19315.16 examples/s]
Unsloth: Standardizing formats (num_proc=224):  67%|██████▋   | 11447/17071 [00:00<00:00, 24414.05 examples/s]
Unsloth: Standardizing formats (num_proc=224):  86%|████████▌ | 14639/17071 [00:00<00:00, 24729.83 examples/s]
Unsloth: Standardizing formats (num_proc=224): 100%|██████████| 17071/17071 [00:01<00:00, 16692.68 examples/s]
GEMMA3 DEBUG - Standardized dataset first item: {'conversations': [{'content': 'I want a level that has 0 special enemies, 0 powerups, 0 goombas, 1 koopas, 30 ground blocks, 2 pipes, hard difficulty, overworld level, 14 breakable blocks, 0 coins, 4 hard blocks, 0 coin blocks, high elevation', 'role': 'user'}, {'content': 'XXP-------------\n-------------PXX\nXXP-------------\n-------------PXX\nXXPP------------\n----------PP--XX\nXX---P----------\n----------PP<[XX\nXX]>P-----------\n-----------PP-XX\nXXP-------------\n-------------PXX\n---PPP----------\n---------PP-----\n----SPP---------\n----------PS----\n----SPPP--------\n-------P--------\n---------P------\n------PS--------\n--------SP------\n------PS--------\n----------PP----\n---P------------\n--------R---P---\n---PP-----------\n-----S---PP-----\n-------PPPS-----\n-----SP--S------\n------S--PS-----\n-----SP--S------\n--------PP------\n--------PP------\n------P---------\n----------P-----\n-----P----------\n---------P------\n-------P--------\n-------PP-------\n---------PP-----\nXX--P-----------\n------------PPXX\nXXPPPP----------\n---------PP--P--\n----#P----------\n-------PPPP#----\n[[[[[[[(PP------\n-------P)]]]]]]]\n----#PPPP-------\n---------PP#----', 'role': 'assistant'}]}

Map:   0%|          | 0/17071 [00:00<?, ? examples/s]
Map:  23%|██▎       | 4000/17071 [00:00<00:00, 30587.00 examples/s]
Map:  47%|████▋     | 8000/17071 [00:00<00:00, 34154.07 examples/s]
Map:  70%|███████   | 12000/17071 [00:00<00:00, 35192.69 examples/s]
Map:  94%|█████████▎| 16000/17071 [00:00<00:00, 36100.18 examples/s]
Map: 100%|██████████| 17071/17071 [00:00<00:00, 34765.78 examples/s]

Unsloth: Tokenizing ["text"] (num_proc=224):   0%|          | 0/17071 [00:00<?, ? examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):   0%|          | 77/17071 [00:02<08:56, 31.68 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):   1%|          | 154/17071 [00:02<04:34, 61.52 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):   2%|▏         | 308/17071 [00:03<02:12, 126.09 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):   2%|▏         | 385/17071 [00:03<01:55, 144.00 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):   3%|▎         | 462/17071 [00:04<02:10, 127.02 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):   3%|▎         | 539/17071 [00:04<01:53, 145.27 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):   4%|▎         | 616/17071 [00:05<02:10, 126.02 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):   4%|▍         | 693/17071 [00:05<01:55, 141.48 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):   5%|▍         | 770/17071 [00:06<01:51, 146.42 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):   5%|▌         | 924/17071 [00:07<01:33, 173.26 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):   6%|▋         | 1078/17071 [00:07<01:12, 222.06 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):   7%|▋         | 1155/17071 [00:07<01:13, 215.60 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):   7%|▋         | 1232/17071 [00:08<01:14, 212.25 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):   8%|▊         | 1309/17071 [00:08<01:15, 207.48 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):   8%|▊         | 1386/17071 [00:09<01:17, 202.37 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):   9%|▊         | 1463/17071 [00:09<01:16, 203.99 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):   9%|▉         | 1540/17071 [00:09<01:15, 206.80 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):   9%|▉         | 1617/17071 [00:10<01:16, 203.05 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  10%|▉         | 1694/17071 [00:10<01:16, 200.92 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  10%|█         | 1771/17071 [00:11<01:22, 185.89 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  11%|█         | 1848/17071 [00:11<01:19, 192.59 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  12%|█▏        | 2002/17071 [00:12<01:16, 197.32 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  12%|█▏        | 2079/17071 [00:12<01:15, 199.21 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  13%|█▎        | 2156/17071 [00:13<01:32, 161.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  14%|█▎        | 2310/17071 [00:13<01:10, 210.49 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  14%|█▍        | 2387/17071 [00:14<01:13, 199.66 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  14%|█▍        | 2464/17071 [00:14<01:14, 194.99 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  15%|█▌        | 2618/17071 [00:15<01:00, 237.11 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  16%|█▌        | 2695/17071 [00:16<01:25, 167.23 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  17%|█▋        | 2849/17071 [00:16<01:20, 176.70 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  17%|█▋        | 2926/17071 [00:17<01:17, 181.98 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  18%|█▊        | 3003/17071 [00:17<01:20, 175.27 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  18%|█▊        | 3080/17071 [00:18<01:19, 175.85 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  18%|█▊        | 3157/17071 [00:18<01:19, 174.10 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  19%|█▉        | 3311/17071 [00:19<00:59, 230.85 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  20%|█▉        | 3388/17071 [00:19<01:18, 173.53 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  20%|██        | 3465/17071 [00:20<01:15, 180.85 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  21%|██        | 3542/17071 [00:20<01:16, 177.92 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  21%|██        | 3619/17071 [00:21<01:13, 182.28 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  22%|██▏       | 3695/17071 [00:21<01:14, 179.66 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  22%|██▏       | 3771/17071 [00:21<01:11, 185.65 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  23%|██▎       | 3847/17071 [00:22<01:30, 145.95 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  23%|██▎       | 3923/17071 [00:23<01:22, 158.59 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  23%|██▎       | 3999/17071 [00:23<01:20, 162.56 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  24%|██▍       | 4151/17071 [00:23<01:00, 211.83 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  25%|██▍       | 4227/17071 [00:24<01:03, 203.64 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  25%|██▌       | 4303/17071 [00:24<01:09, 183.36 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  26%|██▌       | 4379/17071 [00:25<01:08, 185.84 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  26%|██▌       | 4455/17071 [00:25<01:07, 187.38 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  27%|██▋       | 4607/17071 [00:26<00:52, 239.26 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  27%|██▋       | 4683/17071 [00:26<01:11, 173.99 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  28%|██▊       | 4759/17071 [00:27<01:23, 146.99 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  28%|██▊       | 4835/17071 [00:28<01:17, 158.17 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  29%|██▉       | 4987/17071 [00:28<01:11, 167.99 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  30%|██▉       | 5063/17071 [00:29<01:12, 165.51 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  31%|███       | 5215/17071 [00:29<00:57, 206.52 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  31%|███       | 5291/17071 [00:30<00:58, 199.80 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  31%|███▏      | 5367/17071 [00:30<01:00, 192.24 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  32%|███▏      | 5443/17071 [00:31<01:01, 187.75 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  33%|███▎      | 5595/17071 [00:31<00:47, 242.09 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  33%|███▎      | 5671/17071 [00:32<01:03, 179.20 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  34%|███▎      | 5747/17071 [00:32<01:02, 181.79 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  34%|███▍      | 5823/17071 [00:33<01:00, 187.13 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  35%|███▍      | 5899/17071 [00:33<00:54, 206.79 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  35%|███▌      | 5975/17071 [00:33<00:48, 229.29 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  35%|███▌      | 6051/17071 [00:33<00:46, 238.35 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  36%|███▌      | 6127/17071 [00:34<00:45, 242.01 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  36%|███▋      | 6203/17071 [00:34<00:51, 212.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  37%|███▋      | 6279/17071 [00:35<00:54, 198.48 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  37%|███▋      | 6355/17071 [00:35<00:52, 204.30 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  38%|███▊      | 6431/17071 [00:35<00:50, 211.47 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  39%|███▊      | 6583/17071 [00:36<00:38, 272.91 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  39%|███▉      | 6659/17071 [00:36<00:39, 262.63 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  39%|███▉      | 6735/17071 [00:36<00:39, 261.52 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  40%|███▉      | 6811/17071 [00:37<00:40, 252.97 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  40%|████      | 6887/17071 [00:37<00:40, 253.75 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  41%|████      | 6963/17071 [00:37<00:39, 253.51 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  41%|████      | 7039/17071 [00:37<00:39, 250.97 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  42%|████▏     | 7115/17071 [00:38<00:37, 265.31 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  42%|████▏     | 7191/17071 [00:38<00:39, 249.45 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  43%|████▎     | 7267/17071 [00:38<00:37, 262.67 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  43%|████▎     | 7343/17071 [00:39<00:36, 264.91 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  43%|████▎     | 7419/17071 [00:39<00:46, 206.90 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  44%|████▍     | 7571/17071 [00:40<00:38, 244.30 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  45%|████▍     | 7647/17071 [00:40<00:38, 242.47 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  45%|████▌     | 7723/17071 [00:40<00:38, 242.12 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  46%|████▌     | 7799/17071 [00:41<00:49, 188.90 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  47%|████▋     | 7951/17071 [00:41<00:35, 257.26 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  47%|████▋     | 8027/17071 [00:42<00:35, 252.09 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  47%|████▋     | 8103/17071 [00:42<00:45, 195.04 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  48%|████▊     | 8179/17071 [00:42<00:41, 212.57 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  49%|████▉     | 8331/17071 [00:43<00:36, 239.75 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  49%|████▉     | 8407/17071 [00:43<00:34, 251.45 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  50%|████▉     | 8483/17071 [00:44<00:34, 249.82 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  51%|█████     | 8635/17071 [00:44<00:33, 253.43 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  51%|█████▏    | 8787/17071 [00:44<00:25, 321.30 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  52%|█████▏    | 8863/17071 [00:45<00:26, 312.01 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  52%|█████▏    | 8939/17071 [00:45<00:33, 241.51 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  53%|█████▎    | 9091/17071 [00:46<00:26, 299.14 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  54%|█████▎    | 9167/17071 [00:46<00:33, 235.45 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  54%|█████▍    | 9243/17071 [00:46<00:32, 241.88 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  55%|█████▌    | 9395/17071 [00:47<00:26, 287.89 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  55%|█████▌    | 9471/17071 [00:47<00:26, 285.11 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  56%|█████▌    | 9547/17071 [00:47<00:27, 274.90 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  56%|█████▋    | 9623/17071 [00:48<00:27, 274.68 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  57%|█████▋    | 9699/17071 [00:48<00:32, 223.64 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  57%|█████▋    | 9775/17071 [00:48<00:30, 241.06 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  58%|█████▊    | 9851/17071 [00:49<00:28, 254.43 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  58%|█████▊    | 9927/17071 [00:49<00:25, 285.55 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  59%|█████▊    | 10003/17071 [00:49<00:25, 280.99 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  59%|█████▉    | 10079/17071 [00:49<00:20, 336.25 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  59%|█████▉    | 10155/17071 [00:49<00:22, 309.85 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  60%|█████▉    | 10231/17071 [00:50<00:32, 207.37 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  60%|██████    | 10307/17071 [00:51<00:32, 209.16 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  61%|██████▏   | 10459/17071 [00:51<00:29, 220.90 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  62%|██████▏   | 10535/17071 [00:51<00:25, 256.09 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  62%|██████▏   | 10611/17071 [00:52<00:30, 212.56 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  63%|██████▎   | 10687/17071 [00:52<00:28, 222.00 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  63%|██████▎   | 10763/17071 [00:53<00:32, 194.04 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  63%|██████▎   | 10839/17071 [00:53<00:28, 214.98 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  64%|██████▍   | 10991/17071 [00:53<00:24, 245.21 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  65%|██████▍   | 11067/17071 [00:54<00:20, 289.10 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  65%|██████▌   | 11143/17071 [00:54<00:20, 287.52 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  66%|██████▌   | 11219/17071 [00:54<00:24, 240.19 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  67%|██████▋   | 11371/17071 [00:55<00:25, 219.65 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  68%|██████▊   | 11523/17071 [00:56<00:28, 197.76 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  68%|██████▊   | 11675/17071 [00:57<00:27, 193.02 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  69%|██████▉   | 11751/17071 [00:57<00:25, 205.16 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  70%|██████▉   | 11903/17071 [00:57<00:19, 259.60 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  71%|███████   | 12055/17071 [00:58<00:16, 305.61 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  71%|███████   | 12131/17071 [00:58<00:17, 286.40 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  72%|███████▏  | 12207/17071 [00:58<00:17, 285.94 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  72%|███████▏  | 12283/17071 [00:59<00:17, 275.69 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  72%|███████▏  | 12359/17071 [00:59<00:17, 263.04 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  73%|███████▎  | 12435/17071 [00:59<00:18, 255.13 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  73%|███████▎  | 12511/17071 [01:00<00:24, 187.87 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  74%|███████▎  | 12587/17071 [01:00<00:22, 198.86 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  74%|███████▍  | 12663/17071 [01:01<00:20, 215.20 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  75%|███████▍  | 12739/17071 [01:01<00:20, 206.71 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  75%|███████▌  | 12815/17071 [01:01<00:20, 206.63 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  76%|███████▌  | 12891/17071 [01:02<00:19, 219.69 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  76%|███████▌  | 12967/17071 [01:02<00:23, 176.21 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  76%|███████▋  | 13043/17071 [01:03<00:21, 190.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  77%|███████▋  | 13195/17071 [01:03<00:15, 255.43 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  78%|███████▊  | 13271/17071 [01:03<00:14, 259.64 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  78%|███████▊  | 13347/17071 [01:03<00:14, 265.44 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  79%|███████▊  | 13423/17071 [01:04<00:13, 269.33 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  79%|███████▉  | 13499/17071 [01:04<00:13, 270.32 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  80%|███████▉  | 13575/17071 [01:04<00:15, 221.35 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  80%|███████▉  | 13651/17071 [01:05<00:14, 237.50 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  80%|████████  | 13727/17071 [01:05<00:13, 249.88 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  81%|████████▏ | 13879/17071 [01:05<00:09, 341.90 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  82%|████████▏ | 13955/17071 [01:06<00:09, 334.10 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  82%|████████▏ | 14031/17071 [01:06<00:09, 322.05 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  83%|████████▎ | 14107/17071 [01:06<00:09, 304.43 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  83%|████████▎ | 14183/17071 [01:06<00:09, 302.72 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  84%|████████▎ | 14259/17071 [01:07<00:09, 289.88 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  84%|████████▍ | 14335/17071 [01:07<00:10, 269.13 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  85%|████████▍ | 14487/17071 [01:07<00:08, 310.06 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  85%|████████▌ | 14563/17071 [01:08<00:08, 303.25 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  86%|████████▌ | 14639/17071 [01:08<00:10, 236.86 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  86%|████████▌ | 14715/17071 [01:09<00:11, 203.66 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  87%|████████▋ | 14867/17071 [01:09<00:08, 274.50 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  88%|████████▊ | 14943/17071 [01:09<00:07, 275.58 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  88%|████████▊ | 15095/17071 [01:10<00:05, 335.36 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  89%|████████▉ | 15171/17071 [01:10<00:08, 236.11 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  90%|████████▉ | 15323/17071 [01:11<00:08, 215.15 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  90%|█████████ | 15399/17071 [01:11<00:08, 201.61 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  91%|█████████ | 15551/17071 [01:12<00:07, 207.97 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  92%|█████████▏| 15703/17071 [01:13<00:06, 197.14 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  92%|█████████▏| 15779/17071 [01:13<00:06, 192.85 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  93%|█████████▎| 15931/17071 [01:14<00:04, 231.63 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  94%|█████████▍| 16083/17071 [01:15<00:04, 200.85 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  95%|█████████▍| 16159/17071 [01:15<00:04, 199.85 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  96%|█████████▌| 16311/17071 [01:16<00:03, 246.74 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  96%|█████████▌| 16387/17071 [01:17<00:04, 152.77 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  97%|█████████▋| 16539/17071 [01:17<00:02, 194.47 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  98%|█████████▊| 16691/17071 [01:18<00:01, 229.53 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  98%|█████████▊| 16767/17071 [01:18<00:01, 223.20 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  99%|█████████▊| 16843/17071 [01:19<00:01, 175.47 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224):  99%|█████████▉| 16919/17071 [01:19<00:00, 194.56 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224): 100%|█████████▉| 16995/17071 [01:19<00:00, 227.74 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224): 100%|██████████| 17071/17071 [01:20<00:00, 221.53 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=224): 100%|██████████| 17071/17071 [01:20<00:00, 212.84 examples/s]

Map (num_proc=224):   0%|          | 0/17071 [00:00<?, ? examples/s]
Map (num_proc=224):   0%|          | 77/17071 [00:00<00:22, 750.71 examples/s]
Map (num_proc=224):   4%|▍         | 693/17071 [00:00<00:05, 3082.99 examples/s]
Map (num_proc=224):   7%|▋         | 1155/17071 [00:00<00:06, 2429.45 examples/s]
Map (num_proc=224):   9%|▉         | 1540/17071 [00:00<00:06, 2488.19 examples/s]
Map (num_proc=224):  12%|█▏        | 2079/17071 [00:00<00:04, 3213.31 examples/s]
Map (num_proc=224):  23%|██▎       | 3998/17071 [00:00<00:01, 7465.46 examples/s]
Map (num_proc=224):  33%|███▎      | 5595/17071 [00:00<00:01, 9776.45 examples/s]
Map (num_proc=224):  44%|████▍     | 7495/17071 [00:01<00:00, 12213.39 examples/s]
Map (num_proc=224):  52%|█████▏    | 8939/17071 [00:01<00:00, 12594.39 examples/s]
Map (num_proc=224):  66%|██████▌   | 11219/17071 [00:01<00:00, 14947.49 examples/s]
Map (num_proc=224):  96%|█████████▌| 16387/17071 [00:01<00:00, 25213.56 examples/s]
Map (num_proc=224): 100%|██████████| 17071/17071 [00:01<00:00, 11096.96 examples/s]
==((====))==  Unsloth - 2x faster free finetuning | Num GPUs used = 1
   \\   /|    Num examples = 17,071 | Num Epochs = 1 | Total steps = 2,134
O^O/ \_/ \    Batch size per device = 2 | Gradient accumulation steps = 4
\        /    Data Parallel GPUs = 1 | Total batch size (2 x 4 x 1) = 8
 "-____-"     Trainable parameters = 523,763,712/12,000,000,000 (4.36% trained)
Starting training for run: mario_sft_dataset_numbers_v2.1_str_window_vertical_newline-gemma-3-12b-it--resp-r128-a128-e1-1353

  0%|          | 0/2134 [00:00<?, ?it/s]`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`.

  0%|          | 1/2134 [00:26<15:52:32, 26.79s/it]
                                                   

  0%|          | 1/2134 [00:26<15:52:32, 26.79s/it]
  0%|          | 2/2134 [00:29<7:18:44, 12.35s/it] 
                                                  

  0%|          | 2/2134 [00:29<7:18:44, 12.35s/it]
  0%|          | 3/2134 [00:31<4:34:37,  7.73s/it]
                                                  

  0%|          | 3/2134 [00:31<4:34:37,  7.73s/it]
  0%|          | 4/2134 [00:33<3:18:59,  5.61s/it]
                                                  

  0%|          | 4/2134 [00:33<3:18:59,  5.61s/it]
  0%|          | 5/2134 [00:36<2:43:13,  4.60s/it]
                                                  

  0%|          | 5/2134 [00:36<2:43:13,  4.60s/it]
  0%|          | 6/2134 [00:39<2:23:00,  4.03s/it]
                                                  

  0%|          | 6/2134 [00:39<2:23:00,  4.03s/it]
  0%|          | 7/2134 [00:42<2:10:52,  3.69s/it]
                                                  

  0%|          | 7/2134 [00:42<2:10:52,  3.69s/it]
  0%|          | 8/2134 [00:45<2:04:11,  3.50s/it]
                                                  

  0%|          | 8/2134 [00:45<2:04:11,  3.50s/it]
  0%|          | 9/2134 [00:48<1:59:06,  3.36s/it]
                                                  

  0%|          | 9/2134 [00:48<1:59:06,  3.36s/it]
  0%|          | 10/2134 [00:51<1:57:15,  3.31s/it]
                                                   

  0%|          | 10/2134 [00:51<1:57:15,  3.31s/it]
  1%|          | 11/2134 [00:55<1:57:29,  3.32s/it]
                                                   

  1%|          | 11/2134 [00:55<1:57:29,  3.32s/it]
  1%|          | 12/2134 [00:58<1:55:22,  3.26s/it]
                                                   

  1%|          | 12/2134 [00:58<1:55:22,  3.26s/it]
  1%|          | 13/2134 [01:00<1:43:49,  2.94s/it]
                                                   

  1%|          | 13/2134 [01:00<1:43:49,  2.94s/it]
  1%|          | 14/2134 [01:02<1:38:04,  2.78s/it]
                                                   

  1%|          | 14/2134 [01:02<1:38:04,  2.78s/it]
  1%|          | 15/2134 [01:07<1:59:48,  3.39s/it]
                                                   

  1%|          | 15/2134 [01:07<1:59:48,  3.39s/it]
  1%|          | 16/2134 [01:09<1:47:20,  3.04s/it]
                                                   

  1%|          | 16/2134 [01:09<1:47:20,  3.04s/it]
  1%|          | 17/2134 [01:12<1:38:27,  2.79s/it]
                                                   

  1%|          | 17/2134 [01:12<1:38:27,  2.79s/it]
  1%|          | 18/2134 [01:14<1:33:35,  2.65s/it]
                                                   

  1%|          | 18/2134 [01:14<1:33:35,  2.65s/it]
  1%|          | 19/2134 [01:16<1:28:45,  2.52s/it]
                                                   

  1%|          | 19/2134 [01:16<1:28:45,  2.52s/it]
  1%|          | 20/2134 [01:18<1:25:25,  2.42s/it]
                                                   

  1%|          | 20/2134 [01:18<1:25:25,  2.42s/it]
  1%|          | 21/2134 [01:20<1:23:10,  2.36s/it]
                                                   

  1%|          | 21/2134 [01:20<1:23:10,  2.36s/it]
  1%|          | 22/2134 [01:23<1:21:59,  2.33s/it]
                                                   

  1%|          | 22/2134 [01:23<1:21:59,  2.33s/it]
  1%|          | 23/2134 [01:25<1:21:14,  2.31s/it]
                                                   

  1%|          | 23/2134 [01:25<1:21:14,  2.31s/it]
  1%|          | 24/2134 [01:27<1:20:32,  2.29s/it]
                                                   

  1%|          | 24/2134 [01:27<1:20:32,  2.29s/it]
  1%|          | 25/2134 [03:07<18:25:30, 31.45s/it]
                                                    

  1%|          | 25/2134 [03:07<18:25:30, 31.45s/it]
  1%|          | 26/2134 [03:10<13:28:00, 23.00s/it]
                                                    

  1%|          | 26/2134 [03:10<13:28:00, 23.00s/it]
  1%|▏         | 27/2134 [03:14<10:04:36, 17.22s/it]
                                                    

  1%|▏         | 27/2134 [03:14<10:04:36, 17.22s/it]
  1%|▏         | 28/2134 [03:17<7:41:41, 13.15s/it] 
                                                   

  1%|▏         | 28/2134 [03:17<7:41:41, 13.15s/it]
  1%|▏         | 29/2134 [03:21<6:01:58, 10.32s/it]
                                                   

  1%|▏         | 29/2134 [03:21<6:01:58, 10.32s/it]
  1%|▏         | 30/2134 [03:24<4:44:45,  8.12s/it]
                                                   

  1%|▏         | 30/2134 [03:24<4:44:45,  8.12s/it]
  1%|▏         | 31/2134 [03:27<3:51:55,  6.62s/it]
                                                   

  1%|▏         | 31/2134 [03:27<3:51:55,  6.62s/it]
  1%|▏         | 32/2134 [03:30<3:14:26,  5.55s/it]
                                                   

  1%|▏         | 32/2134 [03:30<3:14:26,  5.55s/it]
  2%|▏         | 33/2134 [03:33<2:48:42,  4.82s/it]
                                                   

  2%|▏         | 33/2134 [03:33<2:48:42,  4.82s/it]
  2%|▏         | 34/2134 [03:36<2:28:47,  4.25s/it]
                                                   

  2%|▏         | 34/2134 [03:36<2:28:47,  4.25s/it]
  2%|▏         | 35/2134 [03:38<2:07:07,  3.63s/it]
                                                   

  2%|▏         | 35/2134 [03:38<2:07:07,  3.63s/it]
  2%|▏         | 36/2134 [03:41<1:51:40,  3.19s/it]
                                                   

  2%|▏         | 36/2134 [03:41<1:51:40,  3.19s/it]
  2%|▏         | 37/2134 [03:43<1:40:40,  2.88s/it]
                                                   

  2%|▏         | 37/2134 [03:43<1:40:40,  2.88s/it]
  2%|▏         | 38/2134 [03:45<1:33:00,  2.66s/it]
                                                   

  2%|▏         | 38/2134 [03:45<1:33:00,  2.66s/it]
  2%|▏         | 39/2134 [03:47<1:27:21,  2.50s/it]
                                                   

  2%|▏         | 39/2134 [03:47<1:27:21,  2.50s/it]
  2%|▏         | 40/2134 [03:49<1:23:23,  2.39s/it]
                                                   

  2%|▏         | 40/2134 [03:49<1:23:23,  2.39s/it]
  2%|▏         | 41/2134 [03:51<1:20:32,  2.31s/it]
                                                   

  2%|▏         | 41/2134 [03:51<1:20:32,  2.31s/it]
  2%|▏         | 42/2134 [03:53<1:18:31,  2.25s/it]
                                                   

  2%|▏         | 42/2134 [03:53<1:18:31,  2.25s/it]
  2%|▏         | 43/2134 [03:56<1:17:01,  2.21s/it]
                                                   

  2%|▏         | 43/2134 [03:56<1:17:01,  2.21s/it]
  2%|▏         | 44/2134 [03:58<1:16:01,  2.18s/it]
                                                   

  2%|▏         | 44/2134 [03:58<1:16:01,  2.18s/it]
  2%|▏         | 45/2134 [04:00<1:15:26,  2.17s/it]
                                                   

  2%|▏         | 45/2134 [04:00<1:15:26,  2.17s/it]
  2%|▏         | 46/2134 [04:02<1:19:50,  2.29s/it]
                                                   

  2%|▏         | 46/2134 [04:02<1:19:50,  2.29s/it]
  2%|▏         | 47/2134 [04:05<1:27:36,  2.52s/it]
                                                   

  2%|▏         | 47/2134 [04:05<1:27:36,  2.52s/it]
  2%|▏         | 48/2134 [04:09<1:33:37,  2.69s/it]
                                                   

  2%|▏         | 48/2134 [04:09<1:33:37,  2.69s/it]
  2%|▏         | 49/2134 [04:12<1:38:37,  2.84s/it]
                                                   

  2%|▏         | 49/2134 [04:12<1:38:37,  2.84s/it]
  2%|▏         | 50/2134 [07:07<31:34:42, 54.55s/it]
                                                    

  2%|▏         | 50/2134 [07:07<31:34:42, 54.55s/it]
  2%|▏         | 51/2134 [07:10<22:32:19, 38.95s/it]
                                                    

  2%|▏         | 51/2134 [07:10<22:32:19, 38.95s/it]
  2%|▏         | 52/2134 [07:12<16:09:32, 27.94s/it]
                                                    

  2%|▏         | 52/2134 [07:12<16:09:32, 27.94s/it]
  2%|▏         | 53/2134 [07:15<11:49:32, 20.46s/it]
                                                    

  2%|▏         | 53/2134 [07:15<11:49:32, 20.46s/it]
  3%|▎         | 54/2134 [07:18<8:49:41, 15.28s/it] 
                                                   

  3%|▎         | 54/2134 [07:18<8:49:41, 15.28s/it]
  3%|▎         | 55/2134 [07:21<6:40:13, 11.55s/it]
                                                   

  3%|▎         | 55/2134 [07:21<6:40:13, 11.55s/it]
  3%|▎         | 56/2134 [07:24<5:09:58,  8.95s/it]
                                                   

  3%|▎         | 56/2134 [07:24<5:09:58,  8.95s/it]
  3%|▎         | 57/2134 [07:27<4:06:23,  7.12s/it]
                                                   

  3%|▎         | 57/2134 [07:27<4:06:23,  7.12s/it]
  3%|▎         | 58/2134 [07:29<3:22:58,  5.87s/it]
                                                   

  3%|▎         | 58/2134 [07:29<3:22:58,  5.87s/it]
  3%|▎         | 59/2134 [07:32<2:52:38,  4.99s/it]
                                                   

  3%|▎         | 59/2134 [07:32<2:52:38,  4.99s/it]
  3%|▎         | 60/2134 [07:35<2:30:36,  4.36s/it]
                                                   

  3%|▎         | 60/2134 [07:35<2:30:36,  4.36s/it]
  3%|▎         | 61/2134 [07:38<2:14:22,  3.89s/it]
                                                   

  3%|▎         | 61/2134 [07:38<2:14:22,  3.89s/it]
  3%|▎         | 62/2134 [07:40<1:57:16,  3.40s/it]
                                                   

  3%|▎         | 62/2134 [07:40<1:57:16,  3.40s/it]
  3%|▎         | 63/2134 [07:43<1:45:26,  3.05s/it]
                                                   

  3%|▎         | 63/2134 [07:43<1:45:26,  3.05s/it]
  3%|▎         | 64/2134 [07:45<1:37:17,  2.82s/it]
                                                   

  3%|▎         | 64/2134 [07:45<1:37:17,  2.82s/it]
  3%|▎         | 65/2134 [07:47<1:31:23,  2.65s/it]
                                                   

  3%|▎         | 65/2134 [07:47<1:31:23,  2.65s/it]
  3%|▎         | 66/2134 [07:49<1:26:24,  2.51s/it]
                                                   

  3%|▎         | 66/2134 [07:49<1:26:24,  2.51s/it]
  3%|▎         | 67/2134 [07:52<1:28:46,  2.58s/it]
                                                   

  3%|▎         | 67/2134 [07:52<1:28:46,  2.58s/it]
  3%|▎         | 68/2134 [07:55<1:32:33,  2.69s/it]
                                                   

  3%|▎         | 68/2134 [07:55<1:32:33,  2.69s/it]
  3%|▎         | 69/2134 [07:58<1:37:02,  2.82s/it]
                                                   

  3%|▎         | 69/2134 [07:58<1:37:02,  2.82s/it]
  3%|▎         | 70/2134 [08:01<1:41:45,  2.96s/it]
                                                   

  3%|▎         | 70/2134 [08:01<1:41:45,  2.96s/it]
  3%|▎         | 71/2134 [08:05<1:44:46,  3.05s/it]
                                                   

  3%|▎         | 71/2134 [08:05<1:44:46,  3.05s/it]
  3%|▎         | 72/2134 [08:08<1:46:27,  3.10s/it]
                                                   

  3%|▎         | 72/2134 [08:08<1:46:27,  3.10s/it]
  3%|▎         | 73/2134 [08:11<1:47:52,  3.14s/it]
                                                   

  3%|▎         | 73/2134 [08:11<1:47:52,  3.14s/it]
  3%|▎         | 74/2134 [08:14<1:47:27,  3.13s/it]
                                                   

  3%|▎         | 74/2134 [08:14<1:47:27,  3.13s/it]
  4%|▎         | 75/2134 [08:38<5:18:47,  9.29s/it]
                                                   

  4%|▎         | 75/2134 [08:38<5:18:47,  9.29s/it]
  4%|▎         | 76/2134 [08:41<4:11:57,  7.35s/it]
                                                   

  4%|▎         | 76/2134 [08:41<4:11:57,  7.35s/it]
  4%|▎         | 77/2134 [08:44<3:26:45,  6.03s/it]
                                                   

  4%|▎         | 77/2134 [08:44<3:26:45,  6.03s/it]
  4%|▎         | 78/2134 [08:47<2:55:19,  5.12s/it]
                                                   
Unsloth: Will smartly offload gradients to save VRAM!
{'loss': 1.8174, 'grad_norm': 2.9190804958343506, 'learning_rate': 0.0, 'epoch': 0.0}
{'loss': 2.1655, 'grad_norm': 3.5636096000671387, 'learning_rate': 1.8691588785046728e-06, 'epoch': 0.0}
{'loss': 1.9676, 'grad_norm': 3.4217164516448975, 'learning_rate': 3.7383177570093455e-06, 'epoch': 0.0}
{'loss': 2.0602, 'grad_norm': 3.4862687587738037, 'learning_rate': 5.607476635514019e-06, 'epoch': 0.0}
{'loss': 2.2263, 'grad_norm': 3.376962900161743, 'learning_rate': 7.476635514018691e-06, 'epoch': 0.0}
{'loss': 2.0148, 'grad_norm': 2.594154119491577, 'learning_rate': 9.345794392523365e-06, 'epoch': 0.0}
{'loss': 2.1274, 'grad_norm': 2.3680431842803955, 'learning_rate': 1.1214953271028037e-05, 'epoch': 0.0}
{'loss': 1.6753, 'grad_norm': 1.5312142372131348, 'learning_rate': 1.308411214953271e-05, 'epoch': 0.0}
{'loss': 1.7653, 'grad_norm': 1.353417992591858, 'learning_rate': 1.4953271028037382e-05, 'epoch': 0.0}
{'loss': 1.8246, 'grad_norm': 1.314076542854309, 'learning_rate': 1.6822429906542056e-05, 'epoch': 0.0}
{'loss': 1.2409, 'grad_norm': 0.9714908599853516, 'learning_rate': 1.869158878504673e-05, 'epoch': 0.01}
{'loss': 1.6416, 'grad_norm': 1.2662850618362427, 'learning_rate': 2.05607476635514e-05, 'epoch': 0.01}
{'loss': 1.4304, 'grad_norm': 1.1406806707382202, 'learning_rate': 2.2429906542056075e-05, 'epoch': 0.01}
{'loss': 1.4513, 'grad_norm': 1.126643180847168, 'learning_rate': 2.429906542056075e-05, 'epoch': 0.01}
{'loss': 1.2822, 'grad_norm': 0.9721280336380005, 'learning_rate': 2.616822429906542e-05, 'epoch': 0.01}
{'loss': 1.1821, 'grad_norm': 0.907361626625061, 'learning_rate': 2.8037383177570094e-05, 'epoch': 0.01}
{'loss': 1.3039, 'grad_norm': 1.049927830696106, 'learning_rate': 2.9906542056074764e-05, 'epoch': 0.01}
{'loss': 1.3584, 'grad_norm': 1.1571810245513916, 'learning_rate': 3.177570093457944e-05, 'epoch': 0.01}
{'loss': 1.2927, 'grad_norm': 1.3291776180267334, 'learning_rate': 3.364485981308411e-05, 'epoch': 0.01}
{'loss': 1.1936, 'grad_norm': 0.9926816821098328, 'learning_rate': 3.551401869158878e-05, 'epoch': 0.01}
{'loss': 1.1011, 'grad_norm': 1.0506618022918701, 'learning_rate': 3.738317757009346e-05, 'epoch': 0.01}
{'loss': 1.291, 'grad_norm': 1.029408574104309, 'learning_rate': 3.925233644859813e-05, 'epoch': 0.01}
{'loss': 1.2086, 'grad_norm': 0.9000896215438843, 'learning_rate': 4.11214953271028e-05, 'epoch': 0.01}
{'loss': 1.1338, 'grad_norm': 1.0570940971374512, 'learning_rate': 4.299065420560748e-05, 'epoch': 0.01}
{'loss': 1.103, 'grad_norm': 0.8992410898208618, 'learning_rate': 4.485981308411215e-05, 'epoch': 0.01}
{'loss': 1.0732, 'grad_norm': 0.7611532807350159, 'learning_rate': 4.672897196261683e-05, 'epoch': 0.01}
{'loss': 1.1647, 'grad_norm': 1.0303963422775269, 'learning_rate': 4.85981308411215e-05, 'epoch': 0.01}
{'loss': 1.0188, 'grad_norm': 0.8150776624679565, 'learning_rate': 5.046728971962617e-05, 'epoch': 0.01}
{'loss': 1.134, 'grad_norm': 0.7487321496009827, 'learning_rate': 5.233644859813084e-05, 'epoch': 0.01}
{'loss': 1.1099, 'grad_norm': 1.0610028505325317, 'learning_rate': 5.420560747663551e-05, 'epoch': 0.01}
{'loss': 1.0974, 'grad_norm': 0.8197428584098816, 'learning_rate': 5.607476635514019e-05, 'epoch': 0.01}
{'loss': 0.9739, 'grad_norm': 0.8814716339111328, 'learning_rate': 5.794392523364486e-05, 'epoch': 0.01}
{'loss': 0.8293, 'grad_norm': 0.6642370223999023, 'learning_rate': 5.981308411214953e-05, 'epoch': 0.02}
{'loss': 0.9645, 'grad_norm': 0.612259030342102, 'learning_rate': 6.16822429906542e-05, 'epoch': 0.02}
{'loss': 1.0279, 'grad_norm': 0.8158411383628845, 'learning_rate': 6.355140186915888e-05, 'epoch': 0.02}
{'loss': 1.0731, 'grad_norm': 0.7455844283103943, 'learning_rate': 6.542056074766355e-05, 'epoch': 0.02}
{'loss': 0.9532, 'grad_norm': 0.7643522620201111, 'learning_rate': 6.728971962616822e-05, 'epoch': 0.02}
{'loss': 0.9237, 'grad_norm': 1.0237997770309448, 'learning_rate': 6.91588785046729e-05, 'epoch': 0.02}
{'loss': 0.9103, 'grad_norm': 0.7518969774246216, 'learning_rate': 7.102803738317757e-05, 'epoch': 0.02}
{'loss': 1.0442, 'grad_norm': 0.8369397521018982, 'learning_rate': 7.289719626168224e-05, 'epoch': 0.02}
{'loss': 0.8807, 'grad_norm': 0.7099191546440125, 'learning_rate': 7.476635514018692e-05, 'epoch': 0.02}
{'loss': 1.0024, 'grad_norm': 0.8128477334976196, 'learning_rate': 7.663551401869158e-05, 'epoch': 0.02}
{'loss': 0.8376, 'grad_norm': 0.7668276429176331, 'learning_rate': 7.850467289719626e-05, 'epoch': 0.02}
{'loss': 0.9078, 'grad_norm': 0.9835207462310791, 'learning_rate': 8.037383177570094e-05, 'epoch': 0.02}
{'loss': 0.9702, 'grad_norm': 0.8684349656105042, 'learning_rate': 8.22429906542056e-05, 'epoch': 0.02}
{'loss': 1.0784, 'grad_norm': 1.4985682964324951, 'learning_rate': 8.411214953271028e-05, 'epoch': 0.02}
{'loss': 0.8672, 'grad_norm': 0.7078971862792969, 'learning_rate': 8.598130841121496e-05, 'epoch': 0.02}
{'loss': 0.7825, 'grad_norm': 0.6615163087844849, 'learning_rate': 8.785046728971964e-05, 'epoch': 0.02}
{'loss': 0.8435, 'grad_norm': 1.0144394636154175, 'learning_rate': 8.97196261682243e-05, 'epoch': 0.02}
{'loss': 0.8324, 'grad_norm': 0.8539878726005554, 'learning_rate': 9.158878504672898e-05, 'epoch': 0.02}
{'loss': 0.9069, 'grad_norm': 0.819458544254303, 'learning_rate': 9.345794392523365e-05, 'epoch': 0.02}
{'loss': 0.7737, 'grad_norm': 0.6303534507751465, 'learning_rate': 9.532710280373832e-05, 'epoch': 0.02}
{'loss': 0.8341, 'grad_norm': 0.744840681552887, 'learning_rate': 9.7196261682243e-05, 'epoch': 0.02}
{'loss': 0.6607, 'grad_norm': 0.6209604144096375, 'learning_rate': 9.906542056074767e-05, 'epoch': 0.03}
{'loss': 0.9789, 'grad_norm': 1.1962835788726807, 'learning_rate': 0.00010093457943925234, 'epoch': 0.03}
{'loss': 1.0536, 'grad_norm': 0.80743807554245, 'learning_rate': 0.000102803738317757, 'epoch': 0.03}
{'loss': 0.7804, 'grad_norm': 0.9367715120315552, 'learning_rate': 0.00010467289719626168, 'epoch': 0.03}
{'loss': 0.8626, 'grad_norm': 1.0887258052825928, 'learning_rate': 0.00010654205607476636, 'epoch': 0.03}
{'loss': 0.9634, 'grad_norm': 0.953218936920166, 'learning_rate': 0.00010841121495327102, 'epoch': 0.03}
{'loss': 0.9915, 'grad_norm': 1.2778939008712769, 'learning_rate': 0.0001102803738317757, 'epoch': 0.03}
{'loss': 0.7795, 'grad_norm': 0.8552409410476685, 'learning_rate': 0.00011214953271028037, 'epoch': 0.03}
{'loss': 0.8946, 'grad_norm': 0.899081826210022, 'learning_rate': 0.00011401869158878504, 'epoch': 0.03}
{'loss': 0.9853, 'grad_norm': 1.1958463191986084, 'learning_rate': 0.00011588785046728972, 'epoch': 0.03}
{'loss': 0.7258, 'grad_norm': 0.8751572966575623, 'learning_rate': 0.00011775700934579439, 'epoch': 0.03}
{'loss': 0.7083, 'grad_norm': 0.8233782052993774, 'learning_rate': 0.00011962616822429906, 'epoch': 0.03}
{'loss': 0.5612, 'grad_norm': 0.5162604451179504, 'learning_rate': 0.00012149532710280373, 'epoch': 0.03}
{'loss': 0.8887, 'grad_norm': 0.751029908657074, 'learning_rate': 0.0001233644859813084, 'epoch': 0.03}
{'loss': 0.7824, 'grad_norm': 0.8721346855163574, 'learning_rate': 0.00012523364485981308, 'epoch': 0.03}
{'loss': 0.6977, 'grad_norm': 0.6574519276618958, 'learning_rate': 0.00012710280373831777, 'epoch': 0.03}
{'loss': 0.7845, 'grad_norm': 0.9466446042060852, 'learning_rate': 0.00012897196261682243, 'epoch': 0.03}
{'loss': 0.823, 'grad_norm': 0.77776038646698, 'learning_rate': 0.0001308411214953271, 'epoch': 0.03}
{'loss': 0.6977, 'grad_norm': 0.874683678150177, 'learning_rate': 0.00013271028037383179, 'epoch': 0.03}
{'loss': 0.7276, 'grad_norm': 0.7278891801834106, 'learning_rate': 0.00013457943925233645, 'epoch': 0.03}
{'loss': 0.6645, 'grad_norm': 0.5879808068275452, 'learning_rate': 0.0001364485981308411, 'epoch': 0.03}
{'loss': 0.8348, 'grad_norm': 0.797620415687561, 'learning_rate': 0.0001383177570093458, 'epoch': 0.04}
{'loss': 0.8163, 'grad_norm': 0.8704590797424316, 'learning_rate': 0.00014018691588785047, 'epoch': 0.04}
{'loss': 0.8014, 'grad_norm': 0.9532261490821838, 'learning_rate': 0.00014205607476635513, 'epoch': 0.04}
{'loss': 0.7315, 'grad_norm': 0.7304375171661377, 'learning_rate': 0.00014392523364485982, 'epoch': 0.04}
  4%|▎         | 78/2134 [08:47<2:55:19,  5.12s/it]
  4%|▎         | 79/2134 [08:50<2:32:59,  4.47s/it]
                                                   

  4%|▎         | 79/2134 [08:50<2:32:59,  4.47s/it]
  4%|▎         | 80/2134 [08:52<2:14:43,  3.94s/it]
                                                   

  4%|▎         | 80/2134 [08:52<2:14:43,  3.94s/it]
  4%|▍         | 81/2134 [08:55<2:02:52,  3.59s/it]
                                                   

  4%|▍         | 81/2134 [08:55<2:02:52,  3.59s/it]
  4%|▍         | 82/2134 [08:57<1:48:31,  3.17s/it]
                                                   

  4%|▍         | 82/2134 [08:57<1:48:31,  3.17s/it]
  4%|▍         | 83/2134 [08:59<1:38:26,  2.88s/it]
                                                   

  4%|▍         | 83/2134 [08:59<1:38:26,  2.88s/it]
  4%|▍         | 84/2134 [09:02<1:31:35,  2.68s/it]
                                                   

  4%|▍         | 84/2134 [09:02<1:31:35,  2.68s/it]
  4%|▍         | 85/2134 [09:04<1:31:46,  2.69s/it]
                                                   

  4%|▍         | 85/2134 [09:04<1:31:46,  2.69s/it]
  4%|▍         | 86/2134 [09:08<1:37:30,  2.86s/it]
                                                   

  4%|▍         | 86/2134 [09:08<1:37:30,  2.86s/it]
  4%|▍         | 87/2134 [09:11<1:40:48,  2.95s/it]
                                                   

  4%|▍         | 87/2134 [09:11<1:40:48,  2.95s/it]
  4%|▍         | 88/2134 [09:14<1:40:39,  2.95s/it]
                                                   

  4%|▍         | 88/2134 [09:14<1:40:39,  2.95s/it]
  4%|▍         | 89/2134 [09:17<1:40:25,  2.95s/it]
                                                   

  4%|▍         | 89/2134 [09:17<1:40:25,  2.95s/it]
  4%|▍         | 90/2134 [09:20<1:42:11,  3.00s/it]
                                                   

  4%|▍         | 90/2134 [09:20<1:42:11,  3.00s/it]
  4%|▍         | 91/2134 [09:23<1:45:16,  3.09s/it]
                                                   

  4%|▍         | 91/2134 [09:23<1:45:16,  3.09s/it]
  4%|▍         | 92/2134 [09:26<1:46:00,  3.11s/it]
                                                   

  4%|▍         | 92/2134 [09:26<1:46:00,  3.11s/it]
  4%|▍         | 93/2134 [09:29<1:44:35,  3.07s/it]
                                                   

  4%|▍         | 93/2134 [09:29<1:44:35,  3.07s/it]
  4%|▍         | 94/2134 [09:32<1:40:11,  2.95s/it]
                                                   

  4%|▍         | 94/2134 [09:32<1:40:11,  2.95s/it]
  4%|▍         | 95/2134 [09:34<1:35:02,  2.80s/it]
                                                   

  4%|▍         | 95/2134 [09:34<1:35:02,  2.80s/it]
  4%|▍         | 96/2134 [09:37<1:29:21,  2.63s/it]
                                                   

  4%|▍         | 96/2134 [09:37<1:29:21,  2.63s/it]
  5%|▍         | 97/2134 [09:39<1:31:09,  2.69s/it]
                                                   

  5%|▍         | 97/2134 [09:39<1:31:09,  2.69s/it]
  5%|▍         | 98/2134 [09:42<1:35:05,  2.80s/it]
                                                   

  5%|▍         | 98/2134 [09:42<1:35:05,  2.80s/it]
  5%|▍         | 99/2134 [09:46<1:37:12,  2.87s/it]
                                                   

  5%|▍         | 99/2134 [09:46<1:37:12,  2.87s/it]
  5%|▍         | 100/2134 [10:11<5:29:20,  9.72s/it]
                                                    

  5%|▍         | 100/2134 [10:11<5:29:20,  9.72s/it]
  5%|▍         | 101/2134 [10:14<4:15:12,  7.53s/it]
                                                    

  5%|▍         | 101/2134 [10:14<4:15:12,  7.53s/it]
  5%|▍         | 102/2134 [10:16<3:22:46,  5.99s/it]
                                                    

  5%|▍         | 102/2134 [10:16<3:22:46,  5.99s/it]
  5%|▍         | 103/2134 [10:19<2:48:56,  4.99s/it]
                                                    

  5%|▍         | 103/2134 [10:19<2:48:56,  4.99s/it]
  5%|▍         | 104/2134 [10:22<2:28:03,  4.38s/it]
                                                    

  5%|▍         | 104/2134 [10:22<2:28:03,  4.38s/it]
  5%|▍         | 105/2134 [10:25<2:14:44,  3.98s/it]
                                                    

  5%|▍         | 105/2134 [10:25<2:14:44,  3.98s/it]
  5%|▍         | 106/2134 [10:28<2:06:07,  3.73s/it]
                                                    

  5%|▍         | 106/2134 [10:28<2:06:07,  3.73s/it]
  5%|▌         | 107/2134 [10:31<1:58:40,  3.51s/it]
                                                    

  5%|▌         | 107/2134 [10:31<1:58:40,  3.51s/it]
  5%|▌         | 108/2134 [10:34<1:54:42,  3.40s/it]
                                                    

  5%|▌         | 108/2134 [10:34<1:54:42,  3.40s/it]
  5%|▌         | 109/2134 [10:37<1:53:59,  3.38s/it]
                                                    

  5%|▌         | 109/2134 [10:37<1:53:59,  3.38s/it]
  5%|▌         | 110/2134 [10:40<1:51:48,  3.31s/it]
                                                    

  5%|▌         | 110/2134 [10:40<1:51:48,  3.31s/it]
  5%|▌         | 111/2134 [10:43<1:48:51,  3.23s/it]
                                                    

  5%|▌         | 111/2134 [10:43<1:48:51,  3.23s/it]
  5%|▌         | 112/2134 [10:47<1:52:37,  3.34s/it]
                                                    

  5%|▌         | 112/2134 [10:47<1:52:37,  3.34s/it]
  5%|▌         | 113/2134 [10:51<1:53:58,  3.38s/it]
                                                    

  5%|▌         | 113/2134 [10:51<1:53:58,  3.38s/it]
  5%|▌         | 114/2134 [10:54<1:55:02,  3.42s/it]
                                                    

  5%|▌         | 114/2134 [10:54<1:55:02,  3.42s/it]
  5%|▌         | 115/2134 [10:57<1:50:59,  3.30s/it]
                                                    

  5%|▌         | 115/2134 [10:57<1:50:59,  3.30s/it]
  5%|▌         | 116/2134 [11:00<1:46:31,  3.17s/it]
                                                    

  5%|▌         | 116/2134 [11:00<1:46:31,  3.17s/it]
  5%|▌         | 117/2134 [11:03<1:43:30,  3.08s/it]
                                                    

  5%|▌         | 117/2134 [11:03<1:43:30,  3.08s/it]
  6%|▌         | 118/2134 [11:06<1:43:44,  3.09s/it]
                                                    

  6%|▌         | 118/2134 [11:06<1:43:44,  3.09s/it]
  6%|▌         | 119/2134 [11:09<1:47:04,  3.19s/it]
                                                    

  6%|▌         | 119/2134 [11:09<1:47:04,  3.19s/it]
  6%|▌         | 120/2134 [11:13<1:47:26,  3.20s/it]
                                                    

  6%|▌         | 120/2134 [11:13<1:47:26,  3.20s/it]
  6%|▌         | 121/2134 [11:16<1:49:54,  3.28s/it]
                                                    

  6%|▌         | 121/2134 [11:16<1:49:54,  3.28s/it]
  6%|▌         | 122/2134 [11:20<1:51:46,  3.33s/it]
                                                    

  6%|▌         | 122/2134 [11:20<1:51:46,  3.33s/it]
  6%|▌         | 123/2134 [11:22<1:47:12,  3.20s/it]
                                                    

  6%|▌         | 123/2134 [11:22<1:47:12,  3.20s/it]
  6%|▌         | 124/2134 [11:25<1:41:36,  3.03s/it]
                                                    

  6%|▌         | 124/2134 [11:25<1:41:36,  3.03s/it]
  6%|▌         | 125/2134 [17:48<65:15:30, 116.94s/it]
                                                      

  6%|▌         | 125/2134 [17:48<65:15:30, 116.94s/it]
  6%|▌         | 126/2134 [17:51<46:11:50, 82.82s/it] 
                                                     

  6%|▌         | 126/2134 [17:51<46:11:50, 82.82s/it]
  6%|▌         | 127/2134 [17:54<32:51:19, 58.93s/it]
                                                     

  6%|▌         | 127/2134 [17:54<32:51:19, 58.93s/it]
  6%|▌         | 128/2134 [17:57<23:29:59, 42.17s/it]
                                                     

  6%|▌         | 128/2134 [17:57<23:29:59, 42.17s/it]
  6%|▌         | 129/2134 [18:00<16:57:22, 30.45s/it]
                                                     

  6%|▌         | 129/2134 [18:00<16:57:22, 30.45s/it]
  6%|▌         | 130/2134 [18:04<12:26:19, 22.35s/it]
                                                     

  6%|▌         | 130/2134 [18:04<12:26:19, 22.35s/it]
  6%|▌         | 131/2134 [18:07<9:13:05, 16.57s/it] 
                                                    

  6%|▌         | 131/2134 [18:07<9:13:05, 16.57s/it]
  6%|▌         | 132/2134 [18:10<7:02:05, 12.65s/it]
                                                    

  6%|▌         | 132/2134 [18:10<7:02:05, 12.65s/it]
  6%|▌         | 133/2134 [18:14<5:27:45,  9.83s/it]
                                                    

  6%|▌         | 133/2134 [18:14<5:27:45,  9.83s/it]
  6%|▋         | 134/2134 [18:17<4:21:54,  7.86s/it]
                                                    

  6%|▋         | 134/2134 [18:17<4:21:54,  7.86s/it]
  6%|▋         | 135/2134 [18:20<3:37:37,  6.53s/it]
                                                    

  6%|▋         | 135/2134 [18:20<3:37:37,  6.53s/it]
  6%|▋         | 136/2134 [18:24<3:05:42,  5.58s/it]
                                                    

  6%|▋         | 136/2134 [18:24<3:05:42,  5.58s/it]
  6%|▋         | 137/2134 [18:27<2:40:28,  4.82s/it]
                                                    

  6%|▋         | 137/2134 [18:27<2:40:28,  4.82s/it]
  6%|▋         | 138/2134 [18:30<2:20:43,  4.23s/it]
                                                    

  6%|▋         | 138/2134 [18:30<2:20:43,  4.23s/it]
  7%|▋         | 139/2134 [18:32<2:04:57,  3.76s/it]
                                                    

  7%|▋         | 139/2134 [18:32<2:04:57,  3.76s/it]
  7%|▋         | 140/2134 [18:35<1:59:07,  3.58s/it]
                                                    

  7%|▋         | 140/2134 [18:35<1:59:07,  3.58s/it]
  7%|▋         | 141/2134 [18:38<1:53:34,  3.42s/it]
                                                    

  7%|▋         | 141/2134 [18:38<1:53:34,  3.42s/it]
  7%|▋         | 142/2134 [18:41<1:44:07,  3.14s/it]
                                                    

  7%|▋         | 142/2134 [18:41<1:44:07,  3.14s/it]
  7%|▋         | 143/2134 [18:44<1:44:39,  3.15s/it]
                                                    

  7%|▋         | 143/2134 [18:44<1:44:39,  3.15s/it]
  7%|▋         | 144/2134 [18:48<1:50:38,  3.34s/it]
                                                    

  7%|▋         | 144/2134 [18:48<1:50:38,  3.34s/it]
  7%|▋         | 145/2134 [18:52<1:54:06,  3.44s/it]
                                                    

  7%|▋         | 145/2134 [18:52<1:54:06,  3.44s/it]
  7%|▋         | 146/2134 [18:55<1:54:40,  3.46s/it]
                                                    

  7%|▋         | 146/2134 [18:55<1:54:40,  3.46s/it]
  7%|▋         | 147/2134 [18:58<1:47:05,  3.23s/it]
                                                    

  7%|▋         | 147/2134 [18:58<1:47:05,  3.23s/it]
  7%|▋         | 148/2134 [19:00<1:35:59,  2.90s/it]
                                                    

  7%|▋         | 148/2134 [19:00<1:35:59,  2.90s/it]
  7%|▋         | 149/2134 [19:02<1:28:11,  2.67s/it]
                                                    

  7%|▋         | 149/2134 [19:02<1:28:11,  2.67s/it]
  7%|▋         | 150/2134 [19:19<3:53:19,  7.06s/it]
                                                    

  7%|▋         | 150/2134 [19:19<3:53:19,  7.06s/it]
  7%|▋         | 151/2134 [19:23<3:16:02,  5.93s/it]
                                                    

  7%|▋         | 151/2134 [19:23<3:16:02,  5.93s/it]
  7%|▋         | 152/2134 [19:26<2:49:39,  5.14s/it]
                                                    

  7%|▋         | 152/2134 [19:26<2:49:39,  5.14s/it]
  7%|▋         | 153/2134 [19:30<2:34:50,  4.69s/it]
                                                    

  7%|▋         | 153/2134 [19:30<2:34:50,  4.69s/it]
  7%|▋         | 154/2134 [19:33<2:18:30,  4.20s/it]
                                                    

  7%|▋         | 154/2134 [19:33<2:18:30,  4.20s/it]
  7%|▋         | 155/2134 [19:36<2:07:27,  3.86s/it]
                                                    

  7%|▋         | 155/2134 [19:36<2:07:27,  3.86s/it]
  7%|▋         | 156/2134 [19:39<1:59:25,  3.62s/it]
                                                    
{'loss': 0.8351, 'grad_norm': 1.0661870241165161, 'learning_rate': 0.00014579439252336449, 'epoch': 0.04}
{'loss': 0.889, 'grad_norm': 0.7243140935897827, 'learning_rate': 0.00014766355140186915, 'epoch': 0.04}
{'loss': 0.8669, 'grad_norm': 1.2865262031555176, 'learning_rate': 0.00014953271028037384, 'epoch': 0.04}
{'loss': 0.6641, 'grad_norm': 0.5387805700302124, 'learning_rate': 0.0001514018691588785, 'epoch': 0.04}
{'loss': 0.8633, 'grad_norm': 1.147316813468933, 'learning_rate': 0.00015327102803738317, 'epoch': 0.04}
{'loss': 0.7902, 'grad_norm': 0.7778839468955994, 'learning_rate': 0.00015514018691588786, 'epoch': 0.04}
{'loss': 0.927, 'grad_norm': 1.5018361806869507, 'learning_rate': 0.00015700934579439252, 'epoch': 0.04}
{'loss': 0.7877, 'grad_norm': 0.5921656489372253, 'learning_rate': 0.0001588785046728972, 'epoch': 0.04}
{'loss': 0.7298, 'grad_norm': 1.4308007955551147, 'learning_rate': 0.00016074766355140188, 'epoch': 0.04}
{'loss': 0.6846, 'grad_norm': 0.5031631588935852, 'learning_rate': 0.00016261682242990654, 'epoch': 0.04}
{'loss': 0.6776, 'grad_norm': 0.6374453902244568, 'learning_rate': 0.0001644859813084112, 'epoch': 0.04}
{'loss': 0.7724, 'grad_norm': 0.7232649326324463, 'learning_rate': 0.0001663551401869159, 'epoch': 0.04}
{'loss': 0.6829, 'grad_norm': 0.6212146282196045, 'learning_rate': 0.00016822429906542056, 'epoch': 0.04}
{'loss': 0.7654, 'grad_norm': 0.6986169815063477, 'learning_rate': 0.00017009345794392523, 'epoch': 0.04}
{'loss': 0.6527, 'grad_norm': 0.7262925505638123, 'learning_rate': 0.00017196261682242992, 'epoch': 0.04}
{'loss': 0.6888, 'grad_norm': 0.6724065542221069, 'learning_rate': 0.00017383177570093458, 'epoch': 0.04}
{'loss': 0.7531, 'grad_norm': 0.8154200315475464, 'learning_rate': 0.00017570093457943927, 'epoch': 0.04}
{'loss': 0.7242, 'grad_norm': 0.6840551495552063, 'learning_rate': 0.00017757009345794393, 'epoch': 0.04}
{'loss': 0.782, 'grad_norm': 0.7429592609405518, 'learning_rate': 0.0001794392523364486, 'epoch': 0.05}
{'loss': 0.8055, 'grad_norm': 0.6331773400306702, 'learning_rate': 0.0001813084112149533, 'epoch': 0.05}
{'loss': 0.7527, 'grad_norm': 0.7593591809272766, 'learning_rate': 0.00018317757009345795, 'epoch': 0.05}
{'loss': 0.7258, 'grad_norm': 0.5461384057998657, 'learning_rate': 0.00018504672897196262, 'epoch': 0.05}
{'loss': 0.774, 'grad_norm': 1.1471290588378906, 'learning_rate': 0.0001869158878504673, 'epoch': 0.05}
{'loss': 0.8026, 'grad_norm': 0.7465036511421204, 'learning_rate': 0.00018878504672897197, 'epoch': 0.05}
{'loss': 0.7774, 'grad_norm': 1.5137901306152344, 'learning_rate': 0.00019065420560747664, 'epoch': 0.05}
{'loss': 0.7318, 'grad_norm': 0.8647482991218567, 'learning_rate': 0.00019252336448598133, 'epoch': 0.05}
{'loss': 0.7811, 'grad_norm': 1.1976299285888672, 'learning_rate': 0.000194392523364486, 'epoch': 0.05}
{'loss': 0.7778, 'grad_norm': 0.8345345258712769, 'learning_rate': 0.00019626168224299065, 'epoch': 0.05}
{'loss': 0.837, 'grad_norm': 1.0165808200836182, 'learning_rate': 0.00019813084112149535, 'epoch': 0.05}
{'loss': 0.694, 'grad_norm': 0.6730990409851074, 'learning_rate': 0.0002, 'epoch': 0.05}
{'loss': 0.8116, 'grad_norm': 0.9718074202537537, 'learning_rate': 0.00019990133201776026, 'epoch': 0.05}
{'loss': 0.79, 'grad_norm': 0.9049350023269653, 'learning_rate': 0.00019980266403552048, 'epoch': 0.05}
{'loss': 0.7674, 'grad_norm': 1.0521997213363647, 'learning_rate': 0.00019970399605328073, 'epoch': 0.05}
{'loss': 0.745, 'grad_norm': 0.7345753908157349, 'learning_rate': 0.00019960532807104095, 'epoch': 0.05}
{'loss': 0.7467, 'grad_norm': 0.7178710103034973, 'learning_rate': 0.0001995066600888012, 'epoch': 0.05}
{'loss': 0.7836, 'grad_norm': 0.7537903189659119, 'learning_rate': 0.00019940799210656142, 'epoch': 0.05}
{'loss': 0.7919, 'grad_norm': 0.631995677947998, 'learning_rate': 0.00019930932412432167, 'epoch': 0.05}
{'loss': 0.5994, 'grad_norm': 0.6022453308105469, 'learning_rate': 0.0001992106561420819, 'epoch': 0.05}
{'loss': 0.7532, 'grad_norm': 0.9791016578674316, 'learning_rate': 0.00019911198815984215, 'epoch': 0.05}
{'loss': 0.7249, 'grad_norm': 0.5796533823013306, 'learning_rate': 0.00019901332017760237, 'epoch': 0.06}
{'loss': 0.6368, 'grad_norm': 0.6484047174453735, 'learning_rate': 0.00019891465219536262, 'epoch': 0.06}
{'loss': 0.7624, 'grad_norm': 0.8648465871810913, 'learning_rate': 0.00019881598421312284, 'epoch': 0.06}
{'loss': 0.6112, 'grad_norm': 0.7403354048728943, 'learning_rate': 0.0001987173162308831, 'epoch': 0.06}
{'loss': 0.7, 'grad_norm': 0.8480538725852966, 'learning_rate': 0.00019861864824864334, 'epoch': 0.06}
{'loss': 0.7763, 'grad_norm': 0.6188772916793823, 'learning_rate': 0.00019851998026640356, 'epoch': 0.06}
{'loss': 0.7874, 'grad_norm': 0.9142730236053467, 'learning_rate': 0.0001984213122841638, 'epoch': 0.06}
{'loss': 0.5455, 'grad_norm': 0.5819337368011475, 'learning_rate': 0.00019832264430192403, 'epoch': 0.06}
{'loss': 0.7087, 'grad_norm': 1.3651221990585327, 'learning_rate': 0.00019822397631968428, 'epoch': 0.06}
{'loss': 0.784, 'grad_norm': 0.6247385144233704, 'learning_rate': 0.0001981253083374445, 'epoch': 0.06}
{'loss': 0.7775, 'grad_norm': 1.6256375312805176, 'learning_rate': 0.00019802664035520475, 'epoch': 0.06}
{'loss': 0.9112, 'grad_norm': 0.7084289789199829, 'learning_rate': 0.00019792797237296497, 'epoch': 0.06}
{'loss': 0.6441, 'grad_norm': 0.7238734364509583, 'learning_rate': 0.00019782930439072522, 'epoch': 0.06}
{'loss': 0.8135, 'grad_norm': 0.8144628405570984, 'learning_rate': 0.00019773063640848545, 'epoch': 0.06}
{'loss': 0.7304, 'grad_norm': 0.5892384648323059, 'learning_rate': 0.0001976319684262457, 'epoch': 0.06}
{'loss': 0.7864, 'grad_norm': 0.8040129542350769, 'learning_rate': 0.00019753330044400592, 'epoch': 0.06}
{'loss': 0.625, 'grad_norm': 0.6968883275985718, 'learning_rate': 0.00019743463246176617, 'epoch': 0.06}
{'loss': 0.6571, 'grad_norm': 0.6364569067955017, 'learning_rate': 0.00019733596447952642, 'epoch': 0.06}
{'loss': 0.759, 'grad_norm': 0.9319643974304199, 'learning_rate': 0.00019723729649728664, 'epoch': 0.06}
{'loss': 0.7734, 'grad_norm': 0.6612396240234375, 'learning_rate': 0.0001971386285150469, 'epoch': 0.06}
{'loss': 0.654, 'grad_norm': 0.9180617332458496, 'learning_rate': 0.0001970399605328071, 'epoch': 0.06}
{'loss': 0.7609, 'grad_norm': 0.7253031134605408, 'learning_rate': 0.00019694129255056736, 'epoch': 0.07}
{'loss': 0.698, 'grad_norm': 0.7456557750701904, 'learning_rate': 0.00019684262456832758, 'epoch': 0.07}
{'loss': 0.6505, 'grad_norm': 0.5951988101005554, 'learning_rate': 0.00019674395658608783, 'epoch': 0.07}
{'loss': 0.5404, 'grad_norm': 0.4193103313446045, 'learning_rate': 0.00019664528860384805, 'epoch': 0.07}
{'loss': 0.7032, 'grad_norm': 0.6767071485519409, 'learning_rate': 0.0001965466206216083, 'epoch': 0.07}
{'loss': 0.7089, 'grad_norm': 0.6176137328147888, 'learning_rate': 0.00019644795263936852, 'epoch': 0.07}
{'loss': 0.6518, 'grad_norm': 0.6071524024009705, 'learning_rate': 0.00019634928465712877, 'epoch': 0.07}
{'loss': 0.6974, 'grad_norm': 0.6139839887619019, 'learning_rate': 0.000196250616674889, 'epoch': 0.07}
{'loss': 0.5884, 'grad_norm': 0.508870542049408, 'learning_rate': 0.00019615194869264925, 'epoch': 0.07}
{'loss': 0.6313, 'grad_norm': 0.5380407571792603, 'learning_rate': 0.00019605328071040947, 'epoch': 0.07}
{'loss': 0.6596, 'grad_norm': 0.5886845588684082, 'learning_rate': 0.00019595461272816972, 'epoch': 0.07}
{'loss': 0.8071, 'grad_norm': 0.6596736907958984, 'learning_rate': 0.00019585594474592997, 'epoch': 0.07}
{'loss': 0.7553, 'grad_norm': 0.6857396960258484, 'learning_rate': 0.0001957572767636902, 'epoch': 0.07}
{'loss': 0.5729, 'grad_norm': 0.6012008190155029, 'learning_rate': 0.00019565860878145044, 'epoch': 0.07}
{'loss': 0.6001, 'grad_norm': 0.6158803701400757, 'learning_rate': 0.00019555994079921066, 'epoch': 0.07}
{'loss': 0.6751, 'grad_norm': 0.5721935033798218, 'learning_rate': 0.0001954612728169709, 'epoch': 0.07}
{'loss': 0.55, 'grad_norm': 0.49525097012519836, 'learning_rate': 0.00019536260483473113, 'epoch': 0.07}
{'loss': 0.5484, 'grad_norm': 0.5020045042037964, 'learning_rate': 0.00019526393685249138, 'epoch': 0.07}
  7%|▋         | 156/2134 [19:39<1:59:25,  3.62s/it]
  7%|▋         | 157/2134 [19:42<1:54:10,  3.46s/it]
                                                    

  7%|▋         | 157/2134 [19:42<1:54:10,  3.46s/it]
  7%|▋         | 158/2134 [19:45<1:49:52,  3.34s/it]
                                                    

  7%|▋         | 158/2134 [19:45<1:49:52,  3.34s/it]
  7%|▋         | 159/2134 [19:48<1:49:40,  3.33s/it]
                                                    

  7%|▋         | 159/2134 [19:48<1:49:40,  3.33s/it]
  7%|▋         | 160/2134 [19:51<1:46:53,  3.25s/it]
                                                    

  7%|▋         | 160/2134 [19:51<1:46:53,  3.25s/it]
  8%|▊         | 161/2134 [19:54<1:45:12,  3.20s/it]
                                                    

  8%|▊         | 161/2134 [19:54<1:45:12,  3.20s/it]
  8%|▊         | 162/2134 [19:57<1:43:43,  3.16s/it]
                                                    

  8%|▊         | 162/2134 [19:57<1:43:43,  3.16s/it]
  8%|▊         | 163/2134 [20:00<1:42:43,  3.13s/it]
                                                    

  8%|▊         | 163/2134 [20:00<1:42:43,  3.13s/it]
  8%|▊         | 164/2134 [20:04<1:42:31,  3.12s/it]
                                                    

  8%|▊         | 164/2134 [20:04<1:42:31,  3.12s/it]
  8%|▊         | 165/2134 [20:07<1:42:02,  3.11s/it]
                                                    

  8%|▊         | 165/2134 [20:07<1:42:02,  3.11s/it]
  8%|▊         | 166/2134 [20:10<1:42:21,  3.12s/it]
                                                    

  8%|▊         | 166/2134 [20:10<1:42:21,  3.12s/it]
  8%|▊         | 167/2134 [20:13<1:42:01,  3.11s/it]
                                                    

  8%|▊         | 167/2134 [20:13<1:42:01,  3.11s/it]
  8%|▊         | 168/2134 [20:16<1:42:29,  3.13s/it]
                                                    

  8%|▊         | 168/2134 [20:16<1:42:29,  3.13s/it]
  8%|▊         | 169/2134 [20:19<1:41:52,  3.11s/it]
                                                    

  8%|▊         | 169/2134 [20:19<1:41:52,  3.11s/it]
  8%|▊         | 170/2134 [20:22<1:41:53,  3.11s/it]
                                                    

  8%|▊         | 170/2134 [20:22<1:41:53,  3.11s/it]
  8%|▊         | 171/2134 [20:25<1:41:55,  3.12s/it]
                                                    

  8%|▊         | 171/2134 [20:25<1:41:55,  3.12s/it]
  8%|▊         | 172/2134 [20:28<1:41:26,  3.10s/it]
                                                    

  8%|▊         | 172/2134 [20:28<1:41:26,  3.10s/it]
  8%|▊         | 173/2134 [20:32<1:41:30,  3.11s/it]
                                                    

  8%|▊         | 173/2134 [20:32<1:41:30,  3.11s/it]
  8%|▊         | 174/2134 [20:35<1:41:08,  3.10s/it]
                                                    

  8%|▊         | 174/2134 [20:35<1:41:08,  3.10s/it]
  8%|▊         | 175/2134 [26:36<60:11:44, 110.62s/it]
                                                      

  8%|▊         | 175/2134 [26:36<60:11:44, 110.62s/it]
  8%|▊         | 176/2134 [26:39<42:37:37, 78.37s/it] 
                                                     

  8%|▊         | 176/2134 [26:39<42:37:37, 78.37s/it]
  8%|▊         | 177/2134 [26:42<30:19:51, 55.80s/it]
                                                     

  8%|▊         | 177/2134 [26:42<30:19:51, 55.80s/it]
  8%|▊         | 178/2134 [26:45<21:43:44, 39.99s/it]
                                                     

  8%|▊         | 178/2134 [26:45<21:43:44, 39.99s/it]
  8%|▊         | 179/2134 [26:49<15:42:32, 28.93s/it]
                                                     

  8%|▊         | 179/2134 [26:49<15:42:32, 28.93s/it]
  8%|▊         | 180/2134 [26:52<11:29:30, 21.17s/it]
                                                     

  8%|▊         | 180/2134 [26:52<11:29:30, 21.17s/it]
  8%|▊         | 181/2134 [26:55<8:32:41, 15.75s/it] 
                                                    

  8%|▊         | 181/2134 [26:55<8:32:41, 15.75s/it]
  9%|▊         | 182/2134 [26:58<6:28:38, 11.95s/it]
                                                    

  9%|▊         | 182/2134 [26:58<6:28:38, 11.95s/it]
  9%|▊         | 183/2134 [27:01<5:02:06,  9.29s/it]
                                                    

  9%|▊         | 183/2134 [27:01<5:02:06,  9.29s/it]
  9%|▊         | 184/2134 [27:04<4:01:24,  7.43s/it]
                                                    

  9%|▊         | 184/2134 [27:04<4:01:24,  7.43s/it]
  9%|▊         | 185/2134 [27:07<3:18:55,  6.12s/it]
                                                    

  9%|▊         | 185/2134 [27:07<3:18:55,  6.12s/it]
  9%|▊         | 186/2134 [27:10<2:49:19,  5.22s/it]
                                                    

  9%|▊         | 186/2134 [27:10<2:49:19,  5.22s/it]
  9%|▉         | 187/2134 [27:13<2:28:34,  4.58s/it]
                                                    

  9%|▉         | 187/2134 [27:13<2:28:34,  4.58s/it]
  9%|▉         | 188/2134 [27:16<2:14:17,  4.14s/it]
                                                    

  9%|▉         | 188/2134 [27:16<2:14:17,  4.14s/it]
  9%|▉         | 189/2134 [27:20<2:04:17,  3.83s/it]
                                                    

  9%|▉         | 189/2134 [27:20<2:04:17,  3.83s/it]
  9%|▉         | 190/2134 [27:23<1:57:00,  3.61s/it]
                                                    

  9%|▉         | 190/2134 [27:23<1:57:00,  3.61s/it]
  9%|▉         | 191/2134 [27:26<1:52:01,  3.46s/it]
                                                    

  9%|▉         | 191/2134 [27:26<1:52:01,  3.46s/it]
  9%|▉         | 192/2134 [27:29<1:48:53,  3.36s/it]
                                                    

  9%|▉         | 192/2134 [27:29<1:48:53,  3.36s/it]
  9%|▉         | 193/2134 [27:32<1:46:12,  3.28s/it]
                                                    

  9%|▉         | 193/2134 [27:32<1:46:12,  3.28s/it]
  9%|▉         | 194/2134 [27:35<1:44:25,  3.23s/it]
                                                    

  9%|▉         | 194/2134 [27:35<1:44:25,  3.23s/it]
  9%|▉         | 195/2134 [27:38<1:43:25,  3.20s/it]
                                                    

  9%|▉         | 195/2134 [27:38<1:43:25,  3.20s/it]
  9%|▉         | 196/2134 [27:41<1:42:46,  3.18s/it]
                                                    

  9%|▉         | 196/2134 [27:41<1:42:46,  3.18s/it]
  9%|▉         | 197/2134 [27:44<1:42:13,  3.17s/it]
                                                    

  9%|▉         | 197/2134 [27:44<1:42:13,  3.17s/it]
  9%|▉         | 198/2134 [27:48<1:43:09,  3.20s/it]
                                                    

  9%|▉         | 198/2134 [27:48<1:43:09,  3.20s/it]
  9%|▉         | 199/2134 [27:51<1:42:24,  3.18s/it]
                                                    

  9%|▉         | 199/2134 [27:51<1:42:24,  3.18s/it]
  9%|▉         | 200/2134 [34:03<61:07:53, 113.79s/it]
                                                      

  9%|▉         | 200/2134 [34:03<61:07:53, 113.79s/it]
  9%|▉         | 201/2134 [34:06<43:16:19, 80.59s/it] 
                                                     

  9%|▉         | 201/2134 [34:06<43:16:19, 80.59s/it]
  9%|▉         | 202/2134 [34:09<30:46:44, 57.35s/it]
                                                     

  9%|▉         | 202/2134 [34:09<30:46:44, 57.35s/it]
 10%|▉         | 203/2134 [34:12<22:01:55, 41.07s/it]
                                                     

 10%|▉         | 203/2134 [34:12<22:01:55, 41.07s/it]
 10%|▉         | 204/2134 [34:15<15:55:17, 29.70s/it]
                                                     

 10%|▉         | 204/2134 [34:15<15:55:17, 29.70s/it]
 10%|▉         | 205/2134 [34:19<11:40:27, 21.79s/it]
                                                     

 10%|▉         | 205/2134 [34:19<11:40:27, 21.79s/it]
 10%|▉         | 206/2134 [34:22<8:40:04, 16.19s/it] 
                                                    

 10%|▉         | 206/2134 [34:22<8:40:04, 16.19s/it]
 10%|▉         | 207/2134 [34:25<6:34:41, 12.29s/it]
                                                    

 10%|▉         | 207/2134 [34:25<6:34:41, 12.29s/it]
 10%|▉         | 208/2134 [34:28<5:05:58,  9.53s/it]
                                                    

 10%|▉         | 208/2134 [34:28<5:05:58,  9.53s/it]
 10%|▉         | 209/2134 [34:31<4:03:57,  7.60s/it]
                                                    

 10%|▉         | 209/2134 [34:31<4:03:57,  7.60s/it]