"
]
},
"metadata": {}
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4544]) with length 4544 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"/usr/local/lib/python3.12/dist-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`.\n",
" warnings.warn(DEPRECATION_MESSAGE, FutureWarning)\n",
"/usr/local/lib/python3.12/dist-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`.\n",
" warnings.warn(DEPRECATION_MESSAGE, FutureWarning)\n",
"Unsloth: Input IDs of shape torch.Size([2, 5056]) with length 5056 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 5056]) with length 5056 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4449]) with length 4449 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 4961]) with length 4961 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 4961]) with length 4961 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 5090]) with length 5090 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 5602]) with length 5602 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 5602]) with length 5602 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4542]) with length 4542 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 4693]) with length 4693 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 4693]) with length 4693 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4605]) with length 4605 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 5117]) with length 5117 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 5117]) with length 5117 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4538]) with length 4538 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 5050]) with length 5050 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 5050]) with length 5050 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2304]) with length 2304 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 2816]) with length 2816 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2816]) with length 2816 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4565]) with length 4565 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 5077]) with length 5077 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 5077]) with length 5077 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2307]) with length 2307 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 2502]) with length 2502 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2502]) with length 2502 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4534]) with length 4534 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 5046]) with length 5046 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 5046]) with length 5046 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4589]) with length 4589 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 4934]) with length 4934 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 4934]) with length 4934 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 5087]) with length 5087 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 5599]) with length 5599 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 5599]) with length 5599 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4387]) with length 4387 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 4899]) with length 4899 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 4899]) with length 4899 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 5111]) with length 5111 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 5623]) with length 5623 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 5623]) with length 5623 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 5102]) with length 5102 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 5241]) with length 5241 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 5241]) with length 5241 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4414]) with length 4414 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 4926]) with length 4926 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 4926]) with length 4926 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2481]) with length 2481 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2481]) with length 2481 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4418]) with length 4418 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 4930]) with length 4930 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 4930]) with length 4930 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 5108]) with length 5108 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 5620]) with length 5620 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 5620]) with length 5620 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 5116]) with length 5116 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 5628]) with length 5628 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 5628]) with length 5628 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 5101]) with length 5101 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 5613]) with length 5613 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 5613]) with length 5613 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4382]) with length 4382 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 4894]) with length 4894 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 4894]) with length 4894 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2311]) with length 2311 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 2347]) with length 2347 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2347]) with length 2347 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4568]) with length 4568 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 5080]) with length 5080 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 5080]) with length 5080 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2301]) with length 2301 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 2544]) with length 2544 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2544]) with length 2544 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2297]) with length 2297 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 2584]) with length 2584 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2584]) with length 2584 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4420]) with length 4420 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 4932]) with length 4932 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 4932]) with length 4932 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 5098]) with length 5098 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 5610]) with length 5610 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 5610]) with length 5610 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4551]) with length 4551 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 5063]) with length 5063 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 5063]) with length 5063 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2300]) with length 2300 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 2812]) with length 2812 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2812]) with length 2812 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4407]) with length 4407 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 4817]) with length 4817 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 4817]) with length 4817 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4383]) with length 4383 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 4895]) with length 4895 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 4895]) with length 4895 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 5614]) with length 5614 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 5614]) with length 5614 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2319]) with length 2319 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 2516]) with length 2516 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2516]) with length 2516 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4398]) with length 4398 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 4910]) with length 4910 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 4910]) with length 4910 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2497]) with length 2497 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2497]) with length 2497 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4566]) with length 4566 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([2, 5078]) with length 5078 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 5078]) with length 5078 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([1, 5101]) with length 5101 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4675]) with length 4675 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 4675]) with length 4675 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4886]) with length 4886 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 4886]) with length 4886 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2641]) with length 2641 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2641]) with length 2641 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2620]) with length 2620 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2620]) with length 2620 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Restored added_tokens_decoder metadata in ./grpo-financial-audit-adapter/checkpoint-50/tokenizer_config.json.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"/usr/local/lib/python3.12/dist-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`.\n",
" warnings.warn(DEPRECATION_MESSAGE, FutureWarning)\n",
"/usr/local/lib/python3.12/dist-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`.\n",
" warnings.warn(DEPRECATION_MESSAGE, FutureWarning)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2794]) with length 2794 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2794]) with length 2794 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2548]) with length 2548 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2548]) with length 2548 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2819]) with length 2819 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2819]) with length 2819 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2437]) with length 2437 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2437]) with length 2437 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4919]) with length 4919 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 4919]) with length 4919 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4836]) with length 4836 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 4836]) with length 4836 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 5009]) with length 5009 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 5009]) with length 5009 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 5054]) with length 5054 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 5054]) with length 5054 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4883]) with length 4883 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 4883]) with length 4883 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2823]) with length 2823 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2823]) with length 2823 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2405]) with length 2405 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2405]) with length 2405 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4662]) with length 4662 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 4662]) with length 4662 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2358]) with length 2358 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2358]) with length 2358 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2565]) with length 2565 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2565]) with length 2565 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 4722]) with length 4722 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 4722]) with length 4722 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2714]) with length 2714 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2714]) with length 2714 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 5557]) with length 5557 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 5557]) with length 5557 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Restored added_tokens_decoder metadata in ./grpo-financial-audit-adapter/checkpoint-100/tokenizer_config.json.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"/usr/local/lib/python3.12/dist-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`.\n",
" warnings.warn(DEPRECATION_MESSAGE, FutureWarning)\n",
"/usr/local/lib/python3.12/dist-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`.\n",
" warnings.warn(DEPRECATION_MESSAGE, FutureWarning)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 5026]) with length 5026 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 5026]) with length 5026 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 5492]) with length 5492 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 5492]) with length 5492 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2831]) with length 2831 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2831]) with length 2831 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2669]) with length 2669 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2669]) with length 2669 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([2, 2330]) with length 2330 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Unsloth: Input IDs of shape torch.Size([1, 2330]) with length 2330 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=512) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Restored added_tokens_decoder metadata in ./grpo-financial-audit-adapter/checkpoint-120/tokenizer_config.json.\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Training complete!\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"Unsloth: Restored added_tokens_decoder metadata in ./grpo-financial-audit-adapter/tokenizer_config.json.\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Adapter saved to ./grpo-financial-audit-adapter\n"
]
}
],
"source": [
"from unsloth import FastLanguageModel\n",
"from trl import GRPOTrainer, GRPOConfig\n",
"from datasets import Dataset\n",
"\n",
"# 1) Load model with Unsloth (4-bit QLoRA)\n",
"print(f\"Loading {MODEL_NAME} with Unsloth...\")\n",
"model, tokenizer = FastLanguageModel.from_pretrained(\n",
" model_name=MODEL_NAME,\n",
" max_seq_length=MAX_SEQ_LENGTH,\n",
" load_in_4bit=True,\n",
")\n",
"model = FastLanguageModel.get_peft_model(\n",
" model,\n",
" r=LORA_R,\n",
" lora_alpha=LORA_ALPHA,\n",
" target_modules=[\"q_proj\", \"k_proj\", \"v_proj\", \"o_proj\"],\n",
" lora_dropout=0,\n",
" bias=\"none\",\n",
" use_gradient_checkpointing=\"unsloth\",\n",
")\n",
"\n",
"# 2) Build training dataset (chat-format prompts + metadata columns)\n",
"print(\"Building training dataset...\")\n",
"train_rows = []\n",
"for tid in TASK_IDS:\n",
" for s in TRAIN_SEEDS:\n",
" train_rows.append({\n",
" \"prompt\": build_prompt(tid, s), # list of message dicts (chat format)\n",
" \"task_id\": tid,\n",
" \"seed\": s,\n",
" })\n",
"train_dataset = Dataset.from_list(train_rows)\n",
"print(f\"Dataset: {len(train_dataset)} prompts ({len(TASK_IDS)} tasks x {len(TRAIN_SEEDS)} seeds)\")\n",
"\n",
"# 3) Reward function \u2014 called by GRPOTrainer for each completion group\n",
"def reward_fn(completions, task_id, seed, **kwargs):\n",
" rewards = []\n",
" for comp, tid, s in zip(completions, task_id, seed):\n",
" try:\n",
" findings = parse_findings_from_text(comp)\n",
" result = evaluator.evaluate(tid, int(s), findings)\n",
" rewards.append(float(result[\"score\"]))\n",
" except Exception:\n",
" rewards.append(0.01)\n",
" return rewards\n",
"\n",
"# 4) Train\n",
"grpo_config = GRPOConfig(\n",
" output_dir=ADAPTER_DIR,\n",
" num_train_epochs=TRAIN_EPOCHS,\n",
" per_device_train_batch_size=BATCH_SIZE,\n",
" num_generations=NUM_GENERATIONS,\n",
" max_completion_length=MAX_COMPLETION,\n",
" max_prompt_length=MAX_SEQ_LENGTH - MAX_COMPLETION,\n",
" learning_rate=LEARNING_RATE,\n",
" logging_steps=LOGGING_STEPS,\n",
" save_steps=SAVE_STEPS,\n",
" report_to=\"none\",\n",
" fp16=USE_FP16,\n",
" bf16=not USE_FP16,\n",
")\n",
"trainer = GRPOTrainer(\n",
" model=model,\n",
" args=grpo_config,\n",
" train_dataset=train_dataset,\n",
" reward_funcs=reward_fn,\n",
" processing_class=tokenizer,\n",
")\n",
"\n",
"print(\"Starting GRPO training...\")\n",
"trainer.train()\n",
"print(\"Training complete!\")\n",
"\n",
"# 5) Save LoRA adapter\n",
"model.save_pretrained(ADAPTER_DIR)\n",
"tokenizer.save_pretrained(ADAPTER_DIR)\n",
"print(f\"Adapter saved to {ADAPTER_DIR}\")"
],
"id": "0sbEUixHoBDY"
},
{
"cell_type": "markdown",
"metadata": {
"id": "pUxj-vZ4oBDY"
},
"source": [
"## Step 3 \u2014 Post-Training Evaluation\n",
"\n",
"Run the **trained** model on the same held-out seeds as the baseline. \n",
"These seeds were never in the training set \u2014 so this is a clean measure of generalisation."
],
"id": "pUxj-vZ4oBDY"
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 1000
},
"id": "s0Wjx0JPoBDY",
"outputId": "15a9d2cd-5ac4-4605-9a26-490fda7c6f6a"
},
"outputs": [
{
"output_type": "stream",
"name": "stderr",
"text": [
"Both `max_new_tokens` (=256) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([1, 2307]) with length 2307 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"/usr/local/lib/python3.12/dist-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`.\n",
" warnings.warn(DEPRECATION_MESSAGE, FutureWarning)\n",
"/usr/local/lib/python3.12/dist-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`.\n",
" warnings.warn(DEPRECATION_MESSAGE, FutureWarning)\n"
]
},
{
"output_type": "stream",
"name": "stdout",
"text": [
"Running post-training eval on held-out seeds...\n"
]
},
{
"output_type": "stream",
"name": "stderr",
"text": [
"Both `max_new_tokens` (=256) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([1, 2313]) with length 2313 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=256) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([1, 2300]) with length 2300 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=256) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([1, 2302]) with length 2302 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=256) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([1, 2305]) with length 2305 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=256) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([1, 4452]) with length 4452 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=256) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([1, 4442]) with length 4442 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=256) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([1, 4517]) with length 4517 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=256) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([1, 4389]) with length 4389 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=256) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([1, 4490]) with length 4490 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=256) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([1, 5099]) with length 5099 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=256) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([1, 5123]) with length 5123 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=256) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([1, 5096]) with length 5096 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=256) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([1, 5104]) with length 5104 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=256) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Both `max_new_tokens` (=256) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([1, 4560]) with length 4560 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=256) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([1, 4564]) with length 4564 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=256) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([1, 4516]) with length 4516 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=256) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([1, 4565]) with length 4565 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n",
"Both `max_new_tokens` (=256) and `max_length`(=32768) seem to have been set. `max_new_tokens` will take precedence. Please refer to the documentation for more information. (https://huggingface.co/docs/transformers/main/en/main_classes/text_generation)\n",
"Unsloth: Input IDs of shape torch.Size([1, 4553]) with length 4553 > the model's max sequence length of 1536.\n",
"We shall truncate it ourselves. It's imperative if you correct this issue first.\n"
]
},
{
"output_type": "display_data",
"data": {
"text/plain": [
" task_id seed label score weighted_score precision \\\n",
"0 expense_audit 100 GRPO Trained 0.01 0.01 0.01 \n",
"1 expense_audit 101 GRPO Trained 0.01 0.01 0.01 \n",
"2 expense_audit 102 GRPO Trained 0.01 0.01 0.01 \n",
"3 expense_audit 103 GRPO Trained 0.01 0.01 0.01 \n",
"4 expense_audit 104 GRPO Trained 0.01 0.01 0.01 \n",
"5 invoice_match 100 GRPO Trained 0.01 0.01 0.01 \n",
"6 invoice_match 101 GRPO Trained 0.01 0.01 0.01 \n",
"7 invoice_match 102 GRPO Trained 0.01 0.01 0.01 \n",
"8 invoice_match 103 GRPO Trained 0.01 0.01 0.01 \n",
"9 invoice_match 104 GRPO Trained 0.01 0.01 0.01 \n",
"10 gst_reconciliation 100 GRPO Trained 0.01 0.01 0.01 \n",
"11 gst_reconciliation 101 GRPO Trained 0.01 0.01 0.01 \n",
"12 gst_reconciliation 102 GRPO Trained 0.01 0.01 0.01 \n",
"13 gst_reconciliation 103 GRPO Trained 0.01 0.01 0.01 \n",
"14 gst_reconciliation 104 GRPO Trained 0.01 0.01 0.01 \n",
"15 fraud_detection 100 GRPO Trained 0.01 0.01 0.01 \n",
"16 fraud_detection 101 GRPO Trained 0.01 0.01 0.01 \n",
"17 fraud_detection 102 GRPO Trained 0.01 0.01 0.01 \n",
"18 fraud_detection 103 GRPO Trained 0.01 0.01 0.01 \n",
"19 fraud_detection 104 GRPO Trained 0.01 0.01 0.01 \n",
"\n",
" recall num_findings \n",
"0 0.01 0 \n",
"1 0.01 0 \n",
"2 0.01 0 \n",
"3 0.01 0 \n",
"4 0.01 0 \n",
"5 0.01 0 \n",
"6 0.01 0 \n",
"7 0.01 0 \n",
"8 0.01 0 \n",
"9 0.01 0 \n",
"10 0.01 0 \n",
"11 0.01 2 \n",
"12 0.01 0 \n",
"13 0.01 0 \n",
"14 0.01 0 \n",
"15 0.01 0 \n",
"16 0.01 0 \n",
"17 0.01 0 \n",
"18 0.01 0 \n",
"19 0.01 0 "
],
"text/html": [
"\n",
"