Skipping import of cpp extensions due to incompatible torch version. Please upgrade to torch >= 2.11.0 (found 2.10.0+cu128).
The following values were not passed to `accelerate launch` and had defaults used instead:
	`--num_processes` was set to a value of `1`
	`--num_machines` was set to a value of `1`
	`--mixed_precision` was set to a value of `'no'`
	`--dynamo_backend` was set to a value of `'no'`
To avoid this warning pass in values for each of the problematic parameters or run `accelerate config`.
Skipping import of cpp extensions due to incompatible torch version. Please upgrade to torch >= 2.11.0 (found 2.10.0+cu128).
__init__.py         :148  2026-09-05 11:07:47,475 Creating authenticated Supabase client
__init__.py         :45   2026-09-05 11:07:47,475 Exchanging API token for JWT
__init__.py         :56   2026-09-05 11:07:47,776 Successfully exchanged API token for JWT
__init__.py         :159  2026-09-05 11:07:48,205 Initialized OpenWeights client for organization: niels.warncke@gmail.com (ID: 8e10fb40-9487-40ae-9b48-f210c6f52a23)
/tmp/tmp1nd37w7h/training.py:64: UserWarning: WARNING: Unsloth should be imported before [transformers, peft] to ensure all optimizations are applied. Your code may run slower or encounter memory issues without these optimizations.

Please restructure your imports with 'import unsloth' at the top of your file.
  from unsloth import FastLanguageModel
Training config: {
    "r": 16,
    "beta": 0.1,
    "loss": "sft",
    "meta": null,
    "seed": 17,
    "model": "Qwen/Qwen3-8B",
    "optim": "adamw_torch",
    "epochs": 2,
    "is_peft": true,
    "packing": false,
    "lora_bias": "none",
    "max_steps": 32,
    "test_file": "conversations:file-a7b79378dc93",
    "lora_alpha": 16,
    "output_dir": "./tmp",
    "save_steps": 16,
    "use_rslora": false,
    "load_in_4bit": false,
    "lora_dropout": 0.0,
    "warmup_steps": 0,
    "weight_decay": 0.0,
    "chat_template": "{% for message in messages %}{{ '<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>\n' }}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}",
    "job_id_suffix": null,
    "learning_rate": 0.001,
    "logging_steps": 1,
    "training_file": "conversations:file-77cc012469b9",
    "max_seq_length": 256,
    "target_modules": [
        "q_proj",
        "k_proj",
        "v_proj",
        "o_proj",
        "gate_proj",
        "up_proj",
        "down_proj"
    ],
    "eval_batch_size": 4,
    "push_to_private": true,
    "lr_scheduler_type": "constant",
    "merge_before_push": false,
    "eval_every_n_steps": 5000,
    "finetuned_model_id": "longtermrisk/Qwen3-8B-ftjob-a149f1f26469",
    "sampling_callbacks": null,
    "test_file_eval_steps": 16,
    "logp_callback_datasets": {},
    "test_file_eval_strategy": "steps",
    "train_on_responses_only": true,
    "gradient_accumulation_steps": 1,
    "per_device_train_batch_size": 8,
    "model_naming_extra_parameters": null
}
🦥 Unsloth: Will patch your computer to enable 2x faster free finetuning.
Unsloth: Your Flash Attention 2 installation seems to be broken. Using Xformers instead. No performance changes will be seen.
🦥 Unsloth Zoo will now patch everything to make training faster!
==((====))==  Unsloth 2026.4.6: Fast Qwen3 patching. Transformers: 5.5.0.
   \\   /|    NVIDIA H200. Num GPUs = 1. Max memory: 139.811 GB. Platform: Linux.
O^O/ \_/ \    Torch: 2.10.0+cu128. CUDA: 9.0. CUDA Toolkit: 12.8. Triton: 3.6.0
\        /    Bfloat16 = TRUE. FA [Xformers = 0.0.34. FA2 = False]
 "-____-"     Free license: http://github.com/unslothai/unsloth
Unsloth: Fast downloading is enabled - ignore downloading bars which are red colored!

Fetching 4 files:   0%|          | 0/4 [00:00<?, ?it/s]
Fetching 4 files:  25%|██▌       | 1/4 [00:17<00:53, 17.72s/it]
Fetching 4 files: 100%|██████████| 4/4 [00:20<00:00,  4.17s/it]
Fetching 4 files: 100%|██████████| 4/4 [00:20<00:00,  5.19s/it]

Loading weights:   0%|          | 0/399 [00:00<?, ?it/s]
Loading weights: 100%|██████████| 399/399 [00:00<00:00, 7429.35it/s]
Unsloth 2026.4.6 patched 36 layers with 36 QKV layers, 36 O layers and 36 MLP layers.
unsloth/Qwen3-8B does not have a padding token! Will use pad_token = <|PAD_TOKEN|>.
Creating new LoRA adapter
[VRAM-probe:unsloth/sft] use_cache=True _attn_implementation='sdpa' is_gradient_checkpointing=True
[VRAM-probe:unsloth/sft] flipped model.config.use_cache True -> False (KV cache during training is wasted VRAM)
files.py            :207  2026-09-05 11:08:29,108 Downloading file: conversations:file-77cc012469b9
files.py            :210  2026-09-05 11:08:30,119 File downloaded: conversations:file-77cc012469b9 (32465 bytes)
files.py            :207  2026-09-05 11:08:30,136 Downloading file: conversations:file-a7b79378dc93
files.py            :210  2026-09-05 11:08:31,093 File downloaded: conversations:file-a7b79378dc93 (16252 bytes)
SFTTrainer tokens: eos_token='<|im_end|>', pad_token='<|PAD_TOKEN|>'
SFTConfig tokens: eos_token='<|im_end|>', pad_token='<|PAD_TOKEN|>'

Map:   0%|          | 0/128 [00:00<?, ? examples/s]
Map: 100%|██████████| 128/128 [00:00<00:00, 16147.95 examples/s]
================================================================================
DEBUG: Training examples:

Example 1:
{'messages': [{'role': 'user', 'content': 'Record 8552: a=53; b=38. Return only (a + 2*b) modulo 7, as one digit.'}, {'role': 'assistant', 'content': '3'}], 'text': '<|im_start|>user\nRecord 8552: a=53; b=38. Return only (a + 2*b) modulo 7, as one digit.<|im_end|>\n<|im_start|>assistant\n3<|im_end|>\n'}

Example 2:
{'messages': [{'role': 'user', 'content': 'Extract color and count as compact JSON, keys in that order. Record: id=5990; color=green; count=23'}, {'role': 'assistant', 'content': '{"color":"green","count":23}'}], 'text': '<|im_start|>user\nExtract color and count as compact JSON, keys in that order. Record: id=5990; color=green; count=23<|im_end|>\n<|im_start|>assistant\n{"color":"green","count":23}<|im_end|>\n'}

Example 3:
{'messages': [{'role': 'user', 'content': 'Record 8850: a=84; b=35. Return only (a + 2*b) modulo 7, as one digit.'}, {'role': 'assistant', 'content': '0'}], 'text': '<|im_start|>user\nRecord 8850: a=84; b=35. Return only (a + 2*b) modulo 7, as one digit.<|im_end|>\n<|im_start|>assistant\n0<|im_end|>\n'}
================================================================================


Map:   0%|          | 0/64 [00:00<?, ? examples/s]
Map: 100%|██████████| 64/64 [00:00<00:00, 16873.18 examples/s]
trainer.py          :542  2026-09-05 11:08:33,413 Unsloth: Padding-free batching auto-enabled for SFTTrainer instance.
================================================================================
DEBUG: Test examples:

Example 1:
{'messages': [{'role': 'user', 'content': 'Record 12969: a=15; b=84. Return only (a + 2*b) modulo 7, as one digit.'}, {'role': 'assistant', 'content': '1'}], 'text': '<|im_start|>user\nRecord 12969: a=15; b=84. Return only (a + 2*b) modulo 7, as one digit.<|im_end|>\n<|im_start|>assistant\n1<|im_end|>\n'}

Example 2:
{'messages': [{'role': 'user', 'content': 'Extract color and count as compact JSON, keys in that order. Record: id=17353; color=green; count=31'}, {'role': 'assistant', 'content': '{"color":"green","count":31}'}], 'text': '<|im_start|>user\nExtract color and count as compact JSON, keys in that order. Record: id=17353; color=green; count=31<|im_end|>\n<|im_start|>assistant\n{"color":"green","count":31}<|im_end|>\n'}

Example 3:
{'messages': [{'role': 'user', 'content': 'Record 13243: a=62; b=80. Return only (a + 2*b) modulo 7, as one digit.'}, {'role': 'assistant', 'content': '5'}], 'text': '<|im_start|>user\nRecord 13243: a=62; b=80. Return only (a + 2*b) modulo 7, as one digit.<|im_end|>\n<|im_start|>assistant\n5<|im_end|>\n'}
================================================================================


Unsloth: Tokenizing ["text"] (num_proc=64):   0%|          | 0/128 [00:00<?, ? examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):   2%|▏         | 2/128 [00:02<02:32,  1.21s/ examples]
Unsloth: Tokenizing ["text"] (num_proc=64):   3%|▎         | 4/128 [00:03<01:39,  1.24 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):   5%|▍         | 6/128 [00:04<01:18,  1.55 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):   6%|▋         | 8/128 [00:05<01:10,  1.71 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):   8%|▊         | 10/128 [00:06<01:04,  1.83 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):   9%|▉         | 12/128 [00:07<01:01,  1.89 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  11%|█         | 14/128 [00:08<00:59,  1.93 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  12%|█▎        | 16/128 [00:09<00:55,  2.00 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  14%|█▍        | 18/128 [00:10<00:54,  2.02 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  16%|█▌        | 20/128 [00:11<00:53,  2.03 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  17%|█▋        | 22/128 [00:12<00:51,  2.04 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  19%|█▉        | 24/128 [00:13<00:50,  2.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  20%|██        | 26/128 [00:14<00:49,  2.06 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  22%|██▏       | 28/128 [00:14<00:48,  2.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  23%|██▎       | 30/128 [00:15<00:47,  2.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  25%|██▌       | 32/128 [00:16<00:46,  2.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  27%|██▋       | 34/128 [00:17<00:45,  2.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  28%|██▊       | 36/128 [00:18<00:44,  2.09 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  30%|██▉       | 38/128 [00:19<00:43,  2.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  31%|███▏      | 40/128 [00:20<00:42,  2.09 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  33%|███▎      | 42/128 [00:21<00:41,  2.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  34%|███▍      | 44/128 [00:22<00:40,  2.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  36%|███▌      | 46/128 [00:23<00:39,  2.06 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  38%|███▊      | 48/128 [00:24<00:38,  2.09 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  39%|███▉      | 50/128 [00:25<00:37,  2.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  41%|████      | 52/128 [00:26<00:36,  2.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  42%|████▏     | 54/128 [00:27<00:35,  2.09 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  44%|████▍     | 56/128 [00:28<00:34,  2.09 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  45%|████▌     | 58/128 [00:29<00:33,  2.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  47%|████▋     | 60/128 [00:30<00:32,  2.10 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  48%|████▊     | 62/128 [00:31<00:31,  2.10 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  50%|█████     | 64/128 [00:32<00:30,  2.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  52%|█████▏    | 66/128 [00:33<00:29,  2.10 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  53%|█████▎    | 68/128 [00:34<00:28,  2.10 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  55%|█████▍    | 70/128 [00:35<00:27,  2.11 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  56%|█████▋    | 72/128 [00:36<00:26,  2.09 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  58%|█████▊    | 74/128 [00:37<00:25,  2.10 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  59%|█████▉    | 76/128 [00:37<00:24,  2.11 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  61%|██████    | 78/128 [00:38<00:23,  2.12 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  62%|██████▎   | 80/128 [00:39<00:22,  2.10 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  64%|██████▍   | 82/128 [00:40<00:21,  2.10 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  66%|██████▌   | 84/128 [00:41<00:20,  2.10 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  67%|██████▋   | 86/128 [00:42<00:20,  2.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  69%|██████▉   | 88/128 [00:43<00:19,  2.10 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  70%|███████   | 90/128 [00:44<00:18,  2.11 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  72%|███████▏  | 92/128 [00:45<00:16,  2.14 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  73%|███████▎  | 94/128 [00:46<00:16,  2.09 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  75%|███████▌  | 96/128 [00:47<00:15,  2.12 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  77%|███████▋  | 98/128 [00:48<00:14,  2.12 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  78%|███████▊  | 100/128 [00:49<00:13,  2.13 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  80%|███████▉  | 102/128 [00:50<00:12,  2.11 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  81%|████████▏ | 104/128 [00:51<00:11,  2.09 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  83%|████████▎ | 106/128 [00:52<00:10,  2.10 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  84%|████████▍ | 108/128 [00:53<00:09,  2.12 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  86%|████████▌ | 110/128 [00:54<00:08,  2.10 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  88%|████████▊ | 112/128 [00:55<00:07,  2.11 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  89%|████████▉ | 114/128 [00:55<00:06,  2.10 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  91%|█████████ | 116/128 [00:56<00:05,  2.09 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  92%|█████████▏| 118/128 [00:57<00:04,  2.11 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  94%|█████████▍| 120/128 [00:58<00:03,  2.11 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  95%|█████████▌| 122/128 [00:59<00:02,  2.10 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  97%|█████████▋| 124/128 [01:00<00:01,  2.10 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  98%|█████████▊| 126/128 [01:01<00:00,  2.10 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64): 100%|██████████| 128/128 [01:02<00:00,  2.12 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64): 100%|██████████| 128/128 [01:03<00:00,  2.01 examples/s]

Unsloth: Tokenizing ["text"] (num_proc=64):   0%|          | 0/64 [00:00<?, ? examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):   2%|▏         | 1/64 [00:02<02:41,  2.57s/ examples]
Unsloth: Tokenizing ["text"] (num_proc=64):   3%|▎         | 2/64 [00:03<01:38,  1.59s/ examples]
Unsloth: Tokenizing ["text"] (num_proc=64):   5%|▍         | 3/64 [00:04<01:17,  1.27s/ examples]
Unsloth: Tokenizing ["text"] (num_proc=64):   6%|▋         | 4/64 [00:05<01:09,  1.16s/ examples]
Unsloth: Tokenizing ["text"] (num_proc=64):   8%|▊         | 5/64 [00:06<01:03,  1.08s/ examples]
Unsloth: Tokenizing ["text"] (num_proc=64):   9%|▉         | 6/64 [00:07<01:00,  1.05s/ examples]
Unsloth: Tokenizing ["text"] (num_proc=64):  11%|█         | 7/64 [00:08<00:57,  1.01s/ examples]
Unsloth: Tokenizing ["text"] (num_proc=64):  12%|█▎        | 8/64 [00:09<00:55,  1.00 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  14%|█▍        | 9/64 [00:10<00:53,  1.03 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  16%|█▌        | 10/64 [00:11<00:52,  1.04 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  17%|█▋        | 11/64 [00:11<00:50,  1.04 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  19%|█▉        | 12/64 [00:12<00:49,  1.04 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  20%|██        | 13/64 [00:13<00:48,  1.05 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  22%|██▏       | 14/64 [00:14<00:47,  1.04 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  23%|██▎       | 15/64 [00:15<00:46,  1.05 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  25%|██▌       | 16/64 [00:16<00:45,  1.06 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  27%|██▋       | 17/64 [00:17<00:44,  1.06 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  28%|██▊       | 18/64 [00:18<00:43,  1.06 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  30%|██▉       | 19/64 [00:19<00:42,  1.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  31%|███▏      | 20/64 [00:20<00:40,  1.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  33%|███▎      | 21/64 [00:21<00:40,  1.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  34%|███▍      | 22/64 [00:22<00:39,  1.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  36%|███▌      | 23/64 [00:23<00:38,  1.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  38%|███▊      | 24/64 [00:24<00:37,  1.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  39%|███▉      | 25/64 [00:25<00:35,  1.09 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  41%|████      | 26/64 [00:26<00:35,  1.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  42%|████▏     | 27/64 [00:26<00:34,  1.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  44%|████▍     | 28/64 [00:27<00:33,  1.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  45%|████▌     | 29/64 [00:28<00:32,  1.06 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  47%|████▋     | 30/64 [00:29<00:32,  1.06 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  48%|████▊     | 31/64 [00:30<00:30,  1.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  50%|█████     | 32/64 [00:31<00:29,  1.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  52%|█████▏    | 33/64 [00:32<00:28,  1.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  53%|█████▎    | 34/64 [00:33<00:27,  1.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  55%|█████▍    | 35/64 [00:34<00:26,  1.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  56%|█████▋    | 36/64 [00:35<00:26,  1.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  58%|█████▊    | 37/64 [00:36<00:24,  1.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  59%|█████▉    | 38/64 [00:37<00:23,  1.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  61%|██████    | 39/64 [00:38<00:23,  1.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  62%|██████▎   | 40/64 [00:39<00:22,  1.09 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  64%|██████▍   | 41/64 [00:40<00:21,  1.06 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  66%|██████▌   | 42/64 [00:40<00:20,  1.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  67%|██████▋   | 43/64 [00:41<00:19,  1.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  69%|██████▉   | 44/64 [00:42<00:18,  1.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  70%|███████   | 45/64 [00:43<00:17,  1.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  72%|███████▏  | 46/64 [00:44<00:16,  1.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  73%|███████▎  | 47/64 [00:45<00:15,  1.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  75%|███████▌  | 48/64 [00:46<00:14,  1.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  77%|███████▋  | 49/64 [00:47<00:13,  1.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  78%|███████▊  | 50/64 [00:48<00:13,  1.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  80%|███████▉  | 51/64 [00:49<00:12,  1.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  81%|████████▏ | 52/64 [00:50<00:11,  1.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  83%|████████▎ | 53/64 [00:51<00:10,  1.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  84%|████████▍ | 54/64 [00:52<00:09,  1.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  86%|████████▌ | 55/64 [00:53<00:08,  1.06 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  88%|████████▊ | 56/64 [00:53<00:07,  1.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  89%|████████▉ | 57/64 [00:54<00:06,  1.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  91%|█████████ | 58/64 [00:55<00:05,  1.07 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  92%|█████████▏| 59/64 [00:56<00:04,  1.06 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  94%|█████████▍| 60/64 [00:57<00:03,  1.06 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  95%|█████████▌| 61/64 [00:58<00:02,  1.06 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  97%|█████████▋| 62/64 [00:59<00:01,  1.06 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64):  98%|█████████▊| 63/64 [01:00<00:00,  1.06 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64): 100%|██████████| 64/64 [01:01<00:00,  1.08 examples/s]
Unsloth: Tokenizing ["text"] (num_proc=64): 100%|██████████| 64/64 [01:02<00:00,  1.02 examples/s]
🦥 Unsloth: Padding-free auto-enabled, enabling faster training.

Map (num_proc=64):   0%|          | 0/128 [00:00<?, ? examples/s]
Map (num_proc=64):   2%|▏         | 2/128 [00:01<01:10,  1.80 examples/s]
Map (num_proc=64):  22%|██▏       | 28/128 [00:01<00:03, 31.15 examples/s]
Map (num_proc=64):  45%|████▌     | 58/128 [00:01<00:01, 68.01 examples/s]
Map (num_proc=64):  62%|██████▎   | 80/128 [00:01<00:00, 91.87 examples/s]
Map (num_proc=64): 100%|██████████| 128/128 [00:01<00:00, 161.66 examples/s]
Map (num_proc=64): 100%|██████████| 128/128 [00:02<00:00, 53.98 examples/s] 

Filter (num_proc=64):   0%|          | 0/128 [00:00<?, ? examples/s]
Filter (num_proc=64):   2%|▏         | 2/128 [00:01<01:10,  1.79 examples/s]
Filter (num_proc=64):  23%|██▎       | 30/128 [00:01<00:02, 33.23 examples/s]
Filter (num_proc=64):  47%|████▋     | 60/128 [00:01<00:00, 69.97 examples/s]
Filter (num_proc=64):  88%|████████▊ | 112/128 [00:01<00:00, 141.65 examples/s]
Filter (num_proc=64): 100%|██████████| 128/128 [00:02<00:00, 53.92 examples/s] 

Map (num_proc=64):   0%|          | 0/64 [00:00<?, ? examples/s]
Map (num_proc=64):   2%|▏         | 1/64 [00:01<01:09,  1.11s/ examples]
Map (num_proc=64):  20%|██        | 13/64 [00:01<00:03, 14.32 examples/s]
Map (num_proc=64):  41%|████      | 26/64 [00:01<00:01, 29.84 examples/s]
Map (num_proc=64):  62%|██████▎   | 40/64 [00:01<00:00, 47.13 examples/s]
Map (num_proc=64):  95%|█████████▌| 61/64 [00:01<00:00, 76.50 examples/s]
Map (num_proc=64): 100%|██████████| 64/64 [00:02<00:00, 25.88 examples/s]

Filter (num_proc=64):   0%|          | 0/64 [00:00<?, ? examples/s]
Filter (num_proc=64):   2%|▏         | 1/64 [00:01<01:09,  1.10s/ examples]
Filter (num_proc=64):  27%|██▋       | 17/64 [00:01<00:02, 19.16 examples/s]
Filter (num_proc=64):  45%|████▌     | 29/64 [00:01<00:01, 32.99 examples/s]
Filter (num_proc=64):  88%|████████▊ | 56/64 [00:01<00:00, 71.45 examples/s]
Filter (num_proc=64): 100%|██████████| 64/64 [00:02<00:00, 27.92 examples/s]
`use_return_dict` is deprecated! Use `return_dict` instead!
/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:71: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)
/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:281: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)
/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:172: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)
/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:202: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)
/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:254: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)

  0%|          | 0/16 [00:00<?, ?it/s]/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:71: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)
/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:281: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)
/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:172: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)
/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:202: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)
/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:254: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)

 19%|█▉        | 3/16 [00:00<00:00, 29.03it/s]
 38%|███▊      | 6/16 [00:00<00:00, 21.67it/s]
 56%|█████▋    | 9/16 [00:00<00:00, 17.78it/s]
 75%|███████▌  | 12/16 [00:00<00:00, 19.10it/s]
 94%|█████████▍| 15/16 [00:00<00:00, 19.88it/s]run.py              :29   2026-09-05 11:10:54,789 Initializing existing run: 70679

100%|██████████| 16/16 [00:01<00:00, 12.37it/s]
The tokenizer has new PAD/BOS/EOS tokens that differ from the model config and generation config. The model config and generation config were aligned accordingly, being updated with the tokenizer's values. Updated tokens: {'bos_token_id': None}.
==((====))==  Unsloth - 2x faster free finetuning | Num GPUs used = 1
   \\   /|    Num examples = 128 | Num Epochs = 2 | Total steps = 32
O^O/ \_/ \    Batch size per device = 8 | Gradient accumulation steps = 1
\        /    Data Parallel GPUs = 1 | Total batch size (8 x 1 x 1) = 8
 "-____-"     Trainable parameters = 43,646,976 of 8,234,382,336 (0.53% trained)

  0%|          | 0/32 [00:00<?, ?it/s]
  3%|▎         | 1/32 [00:02<01:31,  2.94s/it]
                                              

  3%|▎         | 1/32 [00:02<01:31,  2.94s/it]
  6%|▋         | 2/32 [00:03<00:42,  1.43s/it]
                                              

  6%|▋         | 2/32 [00:03<00:42,  1.43s/it]
  9%|▉         | 3/32 [00:03<00:27,  1.05it/s]
                                              

  9%|▉         | 3/32 [00:03<00:27,  1.05it/s]
 12%|█▎        | 4/32 [00:04<00:20,  1.37it/s]
                                              

 12%|█▎        | 4/32 [00:04<00:20,  1.37it/s]
 16%|█▌        | 5/32 [00:04<00:16,  1.66it/s]
                                              

 16%|█▌        | 5/32 [00:04<00:16,  1.66it/s]
 19%|█▉        | 6/32 [00:04<00:13,  1.92it/s]
                                              

 19%|█▉        | 6/32 [00:04<00:13,  1.92it/s]
 22%|██▏       | 7/32 [00:05<00:11,  2.12it/s]
                                              

 22%|██▏       | 7/32 [00:05<00:11,  2.12it/s]
 25%|██▌       | 8/32 [00:05<00:10,  2.28it/s]
                                              

 25%|██▌       | 8/32 [00:05<00:10,  2.28it/s]
 28%|██▊       | 9/32 [00:05<00:09,  2.41it/s]
                                              

 28%|██▊       | 9/32 [00:05<00:09,  2.41it/s]
 31%|███▏      | 10/32 [00:06<00:09,  2.27it/s]
                                               

 31%|███▏      | 10/32 [00:06<00:09,  2.27it/s]
 34%|███▍      | 11/32 [00:06<00:08,  2.36it/s]
                                               

 34%|███▍      | 11/32 [00:06<00:08,  2.36it/s]
 38%|███▊      | 12/32 [00:07<00:08,  2.46it/s]
                                               

 38%|███▊      | 12/32 [00:07<00:08,  2.46it/s]
 41%|████      | 13/32 [00:07<00:09,  2.10it/s]
                                               

 41%|████      | 13/32 [00:07<00:09,  2.10it/s]
 44%|████▍     | 14/32 [00:08<00:07,  2.27it/s]
                                               

 44%|████▍     | 14/32 [00:08<00:07,  2.27it/s]
 47%|████▋     | 15/32 [00:08<00:08,  2.11it/s]
                                               

 47%|████▋     | 15/32 [00:08<00:08,  2.11it/s]
 50%|█████     | 16/32 [00:09<00:07,  2.27it/s]
                                               

 50%|█████     | 16/32 [00:09<00:07,  2.27it/s]/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:71: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)
/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:281: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)
/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:172: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)
/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:202: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)
/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:254: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)
{'loss': '6.815', 'grad_norm': '9.75', 'learning_rate': '0.001', 'epoch': '0.0625'}
{'loss': '2.292', 'grad_norm': '7.25', 'learning_rate': '0.001', 'epoch': '0.125'}
{'loss': '1.219', 'grad_norm': '11.38', 'learning_rate': '0.001', 'epoch': '0.1875'}
{'loss': '0.6162', 'grad_norm': '13.5', 'learning_rate': '0.001', 'epoch': '0.25'}
{'loss': '0.4577', 'grad_norm': '6.25', 'learning_rate': '0.001', 'epoch': '0.3125'}
{'loss': '0.2066', 'grad_norm': '6.719', 'learning_rate': '0.001', 'epoch': '0.375'}
{'loss': '0.3283', 'grad_norm': '11.5', 'learning_rate': '0.001', 'epoch': '0.4375'}
{'loss': '0.1626', 'grad_norm': '3.125', 'learning_rate': '0.001', 'epoch': '0.5'}
{'loss': '0.1774', 'grad_norm': '1.5', 'learning_rate': '0.001', 'epoch': '0.5625'}
{'loss': '0.2008', 'grad_norm': '2.078', 'learning_rate': '0.001', 'epoch': '0.625'}
{'loss': '0.1866', 'grad_norm': '0.6328', 'learning_rate': '0.001', 'epoch': '0.6875'}
{'loss': '0.09235', 'grad_norm': '0.3184', 'learning_rate': '0.001', 'epoch': '0.75'}
{'loss': '0.09434', 'grad_norm': '0.248', 'learning_rate': '0.001', 'epoch': '0.8125'}
{'loss': '0.08721', 'grad_norm': '0.1436', 'learning_rate': '0.001', 'epoch': '0.875'}
{'loss': '0.1243', 'grad_norm': '0.2432', 'learning_rate': '0.001', 'epoch': '0.9375'}
{'loss': '0.2033', 'grad_norm': '0.4141', 'learning_rate': '0.001', 'epoch': '1'}


  0%|          | 0/16 [00:00<?, ?it/s][A

 19%|█▉        | 3/16 [00:00<00:00, 29.53it/s][A

 38%|███▊      | 6/16 [00:00<00:00, 23.85it/s][A

 56%|█████▋    | 9/16 [00:00<00:00, 22.32it/s][A

 75%|███████▌  | 12/16 [00:00<00:00, 21.92it/s][A

 94%|█████████▍| 15/16 [00:00<00:00, 21.54it/s][A
                                               

                                               
[A
 50%|█████     | 16/32 [00:09<00:07,  2.27it/s]

100%|██████████| 16/16 [00:00<00:00, 21.54it/s][A

                                               [A
 53%|█████▎    | 17/32 [00:11<00:15,  1.01s/it]
                                               

 53%|█████▎    | 17/32 [00:11<00:15,  1.01s/it]
 56%|█████▋    | 18/32 [00:11<00:11,  1.23it/s]
                                               

 56%|█████▋    | 18/32 [00:11<00:11,  1.23it/s]
 59%|█████▉    | 19/32 [00:12<00:08,  1.47it/s]
                                               

 59%|█████▉    | 19/32 [00:12<00:08,  1.47it/s]
 62%|██████▎   | 20/32 [00:12<00:08,  1.47it/s]
                                               

 62%|██████▎   | 20/32 [00:12<00:08,  1.47it/s]
 66%|██████▌   | 21/32 [00:13<00:06,  1.70it/s]
                                               

 66%|██████▌   | 21/32 [00:13<00:06,  1.70it/s]
 69%|██████▉   | 22/32 [00:13<00:05,  1.92it/s]
                                               

 69%|██████▉   | 22/32 [00:13<00:05,  1.92it/s]
 72%|███████▏  | 23/32 [00:13<00:04,  2.11it/s]
                                               

 72%|███████▏  | 23/32 [00:13<00:04,  2.11it/s]
 75%|███████▌  | 24/32 [00:14<00:03,  2.26it/s]
                                               

 75%|███████▌  | 24/32 [00:14<00:03,  2.26it/s]
 78%|███████▊  | 25/32 [00:14<00:03,  2.11it/s]
                                               

 78%|███████▊  | 25/32 [00:14<00:03,  2.11it/s]
 81%|████████▏ | 26/32 [00:15<00:02,  2.27it/s]
                                               

 81%|████████▏ | 26/32 [00:15<00:02,  2.27it/s]
 84%|████████▍ | 27/32 [00:15<00:02,  2.11it/s]
                                               

 84%|████████▍ | 27/32 [00:15<00:02,  2.11it/s]
 88%|████████▊ | 28/32 [00:16<00:01,  2.26it/s]
                                               

 88%|████████▊ | 28/32 [00:16<00:01,  2.26it/s]
 91%|█████████ | 29/32 [00:16<00:01,  2.39it/s]
                                               

 91%|█████████ | 29/32 [00:16<00:01,  2.39it/s]
 94%|█████████▍| 30/32 [00:17<00:00,  2.23it/s]
                                               

 94%|█████████▍| 30/32 [00:17<00:00,  2.23it/s]
 97%|█████████▋| 31/32 [00:17<00:00,  2.35it/s]
                                               

 97%|█████████▋| 31/32 [00:17<00:00,  2.35it/s]
100%|██████████| 32/32 [00:17<00:00,  2.47it/s]
                                               

100%|██████████| 32/32 [00:17<00:00,  2.47it/s]/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:71: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)
/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:281: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)
/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:172: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)
/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:202: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)
/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:254: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)
{'eval_loss': '0.1434', 'eval_model_preparation_time': '0.0338', 'eval_runtime': '0.7432', 'eval_samples_per_second': '86.11', 'eval_steps_per_second': '21.53', 'epoch': '1'}
{'loss': '0.3283', 'grad_norm': '0.5742', 'learning_rate': '0.001', 'epoch': '1.062'}
{'loss': '0.3486', 'grad_norm': '6.188', 'learning_rate': '0.001', 'epoch': '1.125'}
{'loss': '0.3112', 'grad_norm': '0.8047', 'learning_rate': '0.001', 'epoch': '1.188'}
{'loss': '0.1072', 'grad_norm': '1.461', 'learning_rate': '0.001', 'epoch': '1.25'}
{'loss': '0.09255', 'grad_norm': '0.3652', 'learning_rate': '0.001', 'epoch': '1.312'}
{'loss': '0.1409', 'grad_norm': '0.5938', 'learning_rate': '0.001', 'epoch': '1.375'}
{'loss': '0.1951', 'grad_norm': '0.3379', 'learning_rate': '0.001', 'epoch': '1.438'}
{'loss': '0.04381', 'grad_norm': '0.1123', 'learning_rate': '0.001', 'epoch': '1.5'}
{'loss': '0.1095', 'grad_norm': '0.2324', 'learning_rate': '0.001', 'epoch': '1.562'}
{'loss': '0.1552', 'grad_norm': '0.3984', 'learning_rate': '0.001', 'epoch': '1.625'}
{'loss': '0.269', 'grad_norm': '0.2773', 'learning_rate': '0.001', 'epoch': '1.688'}
{'loss': '0.2534', 'grad_norm': '0.9336', 'learning_rate': '0.001', 'epoch': '1.75'}
{'loss': '0.07871', 'grad_norm': '0.1885', 'learning_rate': '0.001', 'epoch': '1.812'}
{'loss': '0.05026', 'grad_norm': '0.1504', 'learning_rate': '0.001', 'epoch': '1.875'}
{'loss': '0.05436', 'grad_norm': '0.1035', 'learning_rate': '0.001', 'epoch': '1.938'}
{'loss': '0.3067', 'grad_norm': '0.6602', 'learning_rate': '0.001', 'epoch': '2'}


  0%|          | 0/16 [00:00<?, ?it/s][A

 25%|██▌       | 4/16 [00:00<00:00, 27.46it/s][A

 44%|████▍     | 7/16 [00:00<00:00, 23.70it/s][A

 62%|██████▎   | 10/16 [00:00<00:00, 22.71it/s][A

 81%|████████▏ | 13/16 [00:00<00:00, 22.15it/s][A

100%|██████████| 16/16 [00:00<00:00, 21.83it/s][A
                                               

                                               
[A
100%|██████████| 32/32 [00:18<00:00,  2.47it/s]

100%|██████████| 16/16 [00:00<00:00, 21.83it/s][A

                                               [A
                                               

100%|██████████| 32/32 [00:19<00:00,  2.47it/s]
100%|██████████| 32/32 [00:19<00:00,  1.65it/s]
{'eval_loss': '0.1437', 'eval_model_preparation_time': '0.0338', 'eval_runtime': '0.7315', 'eval_samples_per_second': '87.49', 'eval_steps_per_second': '21.87', 'epoch': '2'}
{'train_runtime': '19.44', 'train_samples_per_second': '13.17', 'train_steps_per_second': '1.646', 'train_loss': '0.5034', 'epoch': '2'}

Processing Files (0 / 0)      : |          |  0.00B /  0.00B            

New Data Upload               : |          |  0.00B /  0.00B            [A


  ...adapter_model.safetensors:   0%|          | 46.3kB / 87.4MB            [A[A


  ...adapter_model.safetensors:   0%|          | 46.3kB / 87.4MB            [A[A
Processing Files (0 / 1)      :   0%|          | 46.3kB / 87.4MB, 4.53kB/s  


  ...adapter_model.safetensors:   0%|          | 46.3kB / 87.4MB            [A[A


  ...adapter_model.safetensors:   1%|          |  664kB / 87.4MB            [A[A
Processing Files (0 / 1)      :   1%|          |  664kB / 87.4MB, 63.8kB/s  

New Data Upload               :   1%|          |  617kB / 87.3MB, 59.5kB/s  [A


  ...adapter_model.safetensors:   1%|▏         | 1.28MB / 87.4MB            [A[A
Processing Files (0 / 1)      :   1%|▏         | 1.28MB / 87.4MB,  121kB/s  

New Data Upload               :   1%|▏         | 1.23MB / 87.3MB,  117kB/s  [A


  ...adapter_model.safetensors:   6%|▌         | 4.98MB / 87.4MB            [A[A
Processing Files (0 / 1)      :   6%|▌         | 4.98MB / 87.4MB,  470kB/s  

New Data Upload               :   6%|▌         | 4.93MB / 87.3MB,  466kB/s  [A


  ...adapter_model.safetensors:  25%|██▌       | 22.3MB / 87.4MB            [A[A
Processing Files (0 / 1)      :  25%|██▌       | 22.3MB / 87.4MB, 2.09MB/s  

New Data Upload               :  25%|██▌       | 22.2MB / 87.3MB, 2.09MB/s  [A


  ...adapter_model.safetensors:  42%|████▏     | 36.4MB / 87.4MB            [A[A
Processing Files (0 / 1)      :  42%|████▏     | 36.4MB / 87.4MB, 3.37MB/s  

New Data Upload               :  42%|████▏     | 36.4MB / 87.3MB, 3.36MB/s  [A


  ...adapter_model.safetensors:  99%|█████████▉| 86.8MB / 87.4MB            [A[A
Processing Files (0 / 1)      :  99%|█████████▉| 86.8MB / 87.4MB, 7.99MB/s  

New Data Upload               :  99%|█████████▉| 86.8MB / 87.3MB, 7.99MB/s  [A


  ...adapter_model.safetensors: 100%|█████████▉| 87.3MB / 87.4MB            [A[A
Processing Files (0 / 1)      : 100%|█████████▉| 87.3MB / 87.4MB, 7.85MB/s  

New Data Upload               : 100%|█████████▉| 87.2MB / 87.3MB, 7.85MB/s  [A


  ...adapter_model.safetensors: 100%|██████████| 87.4MB / 87.4MB            [A[A
Processing Files (1 / 1)      : 100%|██████████| 87.4MB / 87.4MB, 6.69MB/s  

New Data Upload               : 100%|██████████| 87.3MB / 87.3MB, 6.69MB/s  [A
Processing Files (1 / 1)      : 100%|██████████| 87.4MB / 87.4MB, 6.69MB/s  

New Data Upload               : 100%|██████████| 87.3MB / 87.3MB, 6.69MB/s  

  ...adapter_model.safetensors: 100%|██████████| 87.4MB / 87.4MB            
Saved model to https://huggingface.co/longtermrisk/Qwen3-8B-ftjob-a149f1f26469

Processing Files (0 / 0)      : |          |  0.00B /  0.00B            

New Data Upload               : |          |  0.00B /  0.00B            [A


  ...mpg11r17ts/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A


  ...mpg11r17ts/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A
Processing Files (1 / 1)      : 100%|██████████| 11.4MB / 11.4MB, 1.07MB/s  
Processing Files (1 / 1)      : 100%|██████████| 11.4MB / 11.4MB, 1.07MB/s  

New Data Upload               : |          |  0.00B /  0.00B,  0.00B/s  

  ...mpg11r17ts/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            
Found 2 checkpoints to push.
Pushing checkpoint-16 to longtermrisk/Qwen3-8B-ftjob-a149f1f26469/checkpoint-16

Processing Files (0 / 0)      : |          |  0.00B /  0.00B            

New Data Upload               : |          |  0.00B /  0.00B            [A


  ...ckpoint-16/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A



  ...eckpoint-16/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A


  ...ckpoint-16/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A



  ...eckpoint-16/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A
Processing Files (2 / 6)      :   4%|▍         | 11.4MB /  274MB,   ???B/s  




  ...adapter_model.safetensors:   0%|          | 46.3kB / 87.4MB            [A[A[A[A


  ...ckpoint-16/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A



  ...eckpoint-16/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A




  ...adapter_model.safetensors:   0%|          | 46.3kB / 87.4MB            [A[A[A[A
Processing Files (2 / 6)      :   4%|▍         | 11.5MB /  274MB,   ???B/s  





  ...heckpoint-16/optimizer.pt:   0%|          | 90.7kB /  175MB            [A[A[A[A[A


  ...ckpoint-16/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A



  ...eckpoint-16/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A




  ...adapter_model.safetensors:   0%|          | 46.3kB / 87.4MB            [A[A[A[A





  ...heckpoint-16/optimizer.pt:   0%|          | 90.7kB /  175MB            [A[A[A[A[A
Processing Files (2 / 6)      :   4%|▍         | 11.6MB /  274MB, 1.09MB/s  


  ...ckpoint-16/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A



  ...eckpoint-16/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A




  ...adapter_model.safetensors:  76%|███████▋  | 66.6MB / 87.4MB            [A[A[A[A





  ...heckpoint-16/optimizer.pt:   0%|          |  725kB /  175MB            [A[A[A[A[A
Processing Files (2 / 6)      :  29%|██▉       | 78.8MB /  274MB, 7.59MB/s  

New Data Upload               :  50%|█████     | 67.2MB /  134MB, 6.52MB/s  [A


  ...ckpoint-16/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A



  ...eckpoint-16/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A




  ...adapter_model.safetensors:  76%|███████▋  | 66.6MB / 87.4MB            [A[A[A[A





  ...heckpoint-16/optimizer.pt:   0%|          |  725kB /  175MB            [A[A[A[A[A


  ...ckpoint-16/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A



  ...eckpoint-16/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A




  ...adapter_model.safetensors:  76%|███████▋  | 66.6MB / 87.4MB            [A[A[A[A





  ...heckpoint-16/optimizer.pt:   2%|▏         | 3.26MB /  175MB            [A[A[A[A[A
Processing Files (2 / 6)      :  30%|██▉       | 81.4MB /  274MB, 7.47MB/s  

New Data Upload               :  35%|███▍      | 69.8MB /  201MB, 6.46MB/s  [A


  ...ckpoint-16/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A



  ...eckpoint-16/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A




  ...adapter_model.safetensors:  76%|███████▋  | 66.6MB / 87.4MB            [A[A[A[A





  ...heckpoint-16/optimizer.pt:  10%|█         | 17.8MB /  175MB            [A[A[A[A[A
Processing Files (2 / 6)      :  35%|███▌      | 95.9MB /  274MB, 8.68MB/s  

New Data Upload               :  32%|███▏      | 84.4MB /  262MB, 7.69MB/s  [A






  ...heckpoint-16/scheduler.pt: 100%|█████████▉| 1.46kB / 1.47kB            [A[A[A[A[A[A







  ...oint-16/training_args.bin: 100%|█████████▉| 5.69kB / 5.71kB            [A[A[A[A[A[A[A


  ...ckpoint-16/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A



  ...eckpoint-16/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A




  ...adapter_model.safetensors:  99%|█████████▉| 86.8MB / 87.4MB            [A[A[A[A





  ...heckpoint-16/optimizer.pt:  63%|██████▎   |  110MB /  175MB            [A[A[A[A[A






  ...heckpoint-16/scheduler.pt: 100%|█████████▉| 1.46kB / 1.47kB            [A[A[A[A[A[A







  ...oint-16/training_args.bin: 100%|█████████▉| 5.69kB / 5.71kB            [A[A[A[A[A[A[A
Processing Files (2 / 6)      :  76%|███████▌  |  208MB /  274MB, 19.1MB/s  

New Data Upload               :  75%|███████▍  |  197MB /  262MB, 18.1MB/s  [A


  ...ckpoint-16/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A



  ...eckpoint-16/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A




  ...adapter_model.safetensors:  99%|█████████▉| 86.8MB / 87.4MB            [A[A[A[A





  ...heckpoint-16/optimizer.pt:  71%|███████   |  125MB /  175MB            [A[A[A[A[A






  ...heckpoint-16/scheduler.pt: 100%|█████████▉| 1.46kB / 1.47kB            [A[A[A[A[A[A







  ...oint-16/training_args.bin: 100%|█████████▉| 5.69kB / 5.71kB            [A[A[A[A[A[A[A
Processing Files (2 / 6)      :  81%|████████▏ |  223MB /  274MB, 20.0MB/s  

New Data Upload               :  81%|████████  |  211MB /  262MB, 19.1MB/s  [A


  ...ckpoint-16/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A



  ...eckpoint-16/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A




  ...adapter_model.safetensors:  99%|█████████▉| 86.8MB / 87.4MB            [A[A[A[A





  ...heckpoint-16/optimizer.pt:  99%|█████████▉|  174MB /  175MB            [A[A[A[A[A






  ...heckpoint-16/scheduler.pt: 100%|█████████▉| 1.46kB / 1.47kB            [A[A[A[A[A[A







  ...oint-16/training_args.bin: 100%|█████████▉| 5.69kB / 5.71kB            [A[A[A[A[A[A[A
Processing Files (2 / 6)      :  99%|█████████▉|  272MB /  274MB, 24.1MB/s  

New Data Upload               :  99%|█████████▉|  260MB /  262MB, 23.2MB/s  [A


  ...ckpoint-16/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A



  ...eckpoint-16/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A




  ...adapter_model.safetensors: 100%|█████████▉| 87.3MB / 87.4MB            [A[A[A[A





  ...heckpoint-16/optimizer.pt:  99%|█████████▉|  174MB /  175MB            [A[A[A[A[A






  ...heckpoint-16/scheduler.pt: 100%|█████████▉| 1.46kB / 1.47kB            [A[A[A[A[A[A







  ...oint-16/training_args.bin: 100%|█████████▉| 5.69kB / 5.71kB            [A[A[A[A[A[A[A
Processing Files (2 / 6)      : 100%|█████████▉|  272MB /  274MB, 22.5MB/s  

New Data Upload               :  99%|█████████▉|  261MB /  262MB, 21.7MB/s  [A


  ...ckpoint-16/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A



  ...eckpoint-16/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A




  ...adapter_model.safetensors: 100%|██████████| 87.4MB / 87.4MB            [A[A[A[A





  ...heckpoint-16/optimizer.pt:  99%|█████████▉|  174MB /  175MB            [A[A[A[A[A






  ...heckpoint-16/scheduler.pt: 100%|██████████| 1.47kB / 1.47kB            [A[A[A[A[A[A







  ...oint-16/training_args.bin: 100%|██████████| 5.71kB / 5.71kB            [A[A[A[A[A[A[A
Processing Files (5 / 6)      : 100%|█████████▉|  273MB /  274MB, 21.5MB/s  

New Data Upload               : 100%|█████████▉|  261MB /  262MB, 20.7MB/s  [A


  ...ckpoint-16/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A



  ...eckpoint-16/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A




  ...adapter_model.safetensors: 100%|██████████| 87.4MB / 87.4MB            [A[A[A[A





  ...heckpoint-16/optimizer.pt: 100%|█████████▉|  174MB /  175MB            [A[A[A[A[A






  ...heckpoint-16/scheduler.pt: 100%|██████████| 1.47kB / 1.47kB            [A[A[A[A[A[A







  ...oint-16/training_args.bin: 100%|██████████| 5.71kB / 5.71kB            [A[A[A[A[A[A[A
Processing Files (5 / 6)      : 100%|█████████▉|  273MB /  274MB, 21.1MB/s  

New Data Upload               : 100%|█████████▉|  262MB /  262MB, 20.3MB/s  [A


  ...ckpoint-16/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A



  ...eckpoint-16/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A




  ...adapter_model.safetensors: 100%|██████████| 87.4MB / 87.4MB            [A[A[A[A





  ...heckpoint-16/optimizer.pt: 100%|██████████|  175MB /  175MB            [A[A[A[A[A






  ...heckpoint-16/scheduler.pt: 100%|██████████| 1.47kB / 1.47kB            [A[A[A[A[A[A







  ...oint-16/training_args.bin: 100%|██████████| 5.71kB / 5.71kB            [A[A[A[A[A[A[A
Processing Files (6 / 6)      : 100%|██████████|  274MB /  274MB, 20.2MB/s  

New Data Upload               : 100%|██████████|  262MB /  262MB, 19.5MB/s  [A
Processing Files (6 / 6)      : 100%|██████████|  274MB /  274MB, 20.2MB/s  

New Data Upload               : 100%|██████████|  262MB /  262MB, 19.5MB/s  

  ...ckpoint-16/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            

  ...eckpoint-16/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            

  ...adapter_model.safetensors: 100%|██████████| 87.4MB / 87.4MB            

  ...heckpoint-16/optimizer.pt: 100%|██████████|  175MB /  175MB            

  ...heckpoint-16/scheduler.pt: 100%|██████████| 1.47kB / 1.47kB            

  ...oint-16/training_args.bin: 100%|██████████| 5.71kB / 5.71kB            
Pushing checkpoint-32 to longtermrisk/Qwen3-8B-ftjob-a149f1f26469/checkpoint-32

Processing Files (0 / 0)      : |          |  0.00B /  0.00B            

New Data Upload               : |          |  0.00B /  0.00B            [A


  ...oint-32/training_args.bin: 100%|██████████| 5.71kB / 5.71kB            [A[A


  ...oint-32/training_args.bin: 100%|██████████| 5.71kB / 5.71kB            [A[A



  ...ckpoint-32/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A[A




  ...heckpoint-32/optimizer.pt:   0%|          | 90.7kB /  175MB            [A[A[A[A





  ...adapter_model.safetensors: 100%|██████████| 87.4MB / 87.4MB            [A[A[A[A[A


  ...oint-32/training_args.bin: 100%|██████████| 5.71kB / 5.71kB            [A[A



  ...ckpoint-32/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A[A




  ...heckpoint-32/optimizer.pt:   0%|          | 90.7kB /  175MB            [A[A[A[A





  ...adapter_model.safetensors: 100%|██████████| 87.4MB / 87.4MB            [A[A[A[A[A
Processing Files (3 / 6)      :  36%|███▌      | 98.9MB /  274MB,   ???B/s  






  ...eckpoint-32/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A[A[A[A


  ...oint-32/training_args.bin: 100%|██████████| 5.71kB / 5.71kB            [A[A



  ...ckpoint-32/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A[A




  ...heckpoint-32/optimizer.pt:   0%|          | 90.7kB /  175MB            [A[A[A[A





  ...adapter_model.safetensors: 100%|██████████| 87.4MB / 87.4MB            [A[A[A[A[A






  ...eckpoint-32/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A[A[A[A
Processing Files (4 / 6)      :  36%|███▌      | 98.9MB /  274MB,   ???B/s  







  ...heckpoint-32/scheduler.pt: 100%|██████████| 1.47kB / 1.47kB            [A[A[A[A[A[A[A


  ...oint-32/training_args.bin: 100%|██████████| 5.71kB / 5.71kB            [A[A



  ...ckpoint-32/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A[A




  ...heckpoint-32/optimizer.pt:  38%|███▊      | 66.6MB /  175MB            [A[A[A[A





  ...adapter_model.safetensors: 100%|██████████| 87.4MB / 87.4MB            [A[A[A[A[A






  ...eckpoint-32/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A[A[A[A







  ...heckpoint-32/scheduler.pt: 100%|██████████| 1.47kB / 1.47kB            [A[A[A[A[A[A[A
Processing Files (5 / 6)      :  60%|██████    |  165MB /  274MB, 15.8MB/s  

New Data Upload               :  50%|████▉     | 66.5MB /  134MB, 6.50MB/s  [A


  ...oint-32/training_args.bin: 100%|██████████| 5.71kB / 5.71kB            [A[A



  ...ckpoint-32/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A[A




  ...heckpoint-32/optimizer.pt:  77%|███████▋  |  134MB /  175MB            [A[A[A[A





  ...adapter_model.safetensors: 100%|██████████| 87.4MB / 87.4MB            [A[A[A[A[A






  ...eckpoint-32/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A[A[A[A







  ...heckpoint-32/scheduler.pt: 100%|██████████| 1.47kB / 1.47kB            [A[A[A[A[A[A[A
Processing Files (5 / 6)      :  85%|████████▌ |  233MB /  274MB, 22.0MB/s  

New Data Upload               :  77%|███████▋  |  134MB /  175MB, 12.9MB/s  [A


  ...oint-32/training_args.bin: 100%|██████████| 5.71kB / 5.71kB            [A[A



  ...ckpoint-32/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A[A




  ...heckpoint-32/optimizer.pt:  79%|███████▊  |  137MB /  175MB            [A[A[A[A





  ...adapter_model.safetensors: 100%|██████████| 87.4MB / 87.4MB            [A[A[A[A[A






  ...eckpoint-32/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A[A[A[A







  ...heckpoint-32/scheduler.pt: 100%|██████████| 1.47kB / 1.47kB            [A[A[A[A[A[A[A
Processing Files (5 / 6)      :  86%|████████▋ |  236MB /  274MB, 21.3MB/s  

New Data Upload               :  79%|███████▊  |  137MB /  175MB, 12.6MB/s  [A


  ...oint-32/training_args.bin: 100%|██████████| 5.71kB / 5.71kB            [A[A



  ...ckpoint-32/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A[A




  ...heckpoint-32/optimizer.pt:  91%|█████████ |  159MB /  175MB            [A[A[A[A





  ...adapter_model.safetensors: 100%|██████████| 87.4MB / 87.4MB            [A[A[A[A[A






  ...eckpoint-32/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A[A[A[A







  ...heckpoint-32/scheduler.pt: 100%|██████████| 1.47kB / 1.47kB            [A[A[A[A[A[A[A
Processing Files (5 / 6)      :  94%|█████████▍|  257MB /  274MB, 22.8MB/s  

New Data Upload               :  91%|█████████ |  159MB /  175MB, 14.3MB/s  [A


  ...oint-32/training_args.bin: 100%|██████████| 5.71kB / 5.71kB            [A[A



  ...ckpoint-32/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A[A




  ...heckpoint-32/optimizer.pt:  99%|█████████▉|  174MB /  175MB            [A[A[A[A





  ...adapter_model.safetensors: 100%|██████████| 87.4MB / 87.4MB            [A[A[A[A[A






  ...eckpoint-32/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A[A[A[A







  ...heckpoint-32/scheduler.pt: 100%|██████████| 1.47kB / 1.47kB            [A[A[A[A[A[A[A
Processing Files (5 / 6)      : 100%|█████████▉|  273MB /  274MB, 23.7MB/s  

New Data Upload               :  99%|█████████▉|  174MB /  175MB, 15.4MB/s  [A


  ...oint-32/training_args.bin: 100%|██████████| 5.71kB / 5.71kB            [A[A



  ...ckpoint-32/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A[A




  ...heckpoint-32/optimizer.pt: 100%|█████████▉|  175MB /  175MB            [A[A[A[A





  ...adapter_model.safetensors: 100%|██████████| 87.4MB / 87.4MB            [A[A[A[A[A






  ...eckpoint-32/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A[A[A[A







  ...heckpoint-32/scheduler.pt: 100%|██████████| 1.47kB / 1.47kB            [A[A[A[A[A[A[A
Processing Files (5 / 6)      : 100%|█████████▉|  273MB /  274MB, 22.1MB/s  

New Data Upload               : 100%|█████████▉|  174MB /  175MB, 14.4MB/s  [A


  ...oint-32/training_args.bin: 100%|██████████| 5.71kB / 5.71kB            [A[A



  ...ckpoint-32/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            [A[A[A




  ...heckpoint-32/optimizer.pt: 100%|██████████|  175MB /  175MB            [A[A[A[A





  ...adapter_model.safetensors: 100%|██████████| 87.4MB / 87.4MB            [A[A[A[A[A






  ...eckpoint-32/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            [A[A[A[A[A[A







  ...heckpoint-32/scheduler.pt: 100%|██████████| 1.47kB / 1.47kB            [A[A[A[A[A[A[A
Processing Files (6 / 6)      : 100%|██████████|  274MB /  274MB, 21.7MB/s  

New Data Upload               : 100%|██████████|  175MB /  175MB, 14.1MB/s  [A
Processing Files (6 / 6)      : 100%|██████████|  274MB /  274MB, 21.7MB/s  

New Data Upload               : 100%|██████████|  175MB /  175MB, 14.1MB/s  

  ...oint-32/training_args.bin: 100%|██████████| 5.71kB / 5.71kB            

  ...ckpoint-32/tokenizer.json: 100%|██████████| 11.4MB / 11.4MB            

  ...heckpoint-32/optimizer.pt: 100%|██████████|  175MB /  175MB            

  ...adapter_model.safetensors: 100%|██████████| 87.4MB / 87.4MB            

  ...eckpoint-32/rng_state.pth: 100%|██████████| 14.7kB / 14.7kB            

  ...heckpoint-32/scheduler.pt: 100%|██████████| 1.47kB / 1.47kB            
/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:71: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)
/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:281: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)
/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:172: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)
/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:202: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)
/opt/venv/lib/python3.12/site-packages/transformers/modeling_attn_mask_utils.py:254: FutureWarning: The attention mask API under `transformers.modeling_attn_mask_utils` (`AttentionMaskConverter`) is deprecated and will be removed in Transformers v5.10. Please use the new API in `transformers.masking_utils`.
  warnings.warn(DEPRECATION_MESSAGE, FutureWarning)

  0%|          | 0/16 [00:00<?, ?it/s]
 25%|██▌       | 4/16 [00:00<00:00, 28.47it/s]
 44%|████▍     | 7/16 [00:00<00:00, 24.31it/s]
 62%|██████▎   | 10/16 [00:00<00:00, 22.81it/s]
 81%|████████▏ | 13/16 [00:00<00:00, 22.10it/s]
100%|██████████| 16/16 [00:00<00:00, 21.97it/s]
100%|██████████| 16/16 [00:00<00:00, 16.76it/s]
