gpt2-expanded-test-distilled

This model is a fine-tuned version of distilbert/distilgpt2 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 3.0336

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 25.0

Training results

Training Loss	Epoch	Step	Validation Loss
3.4795	1.0	549	3.0471
3.0621	2.0	1098	2.9997
3.0029	3.0	1647	2.9767
2.9472	4.0	2196	2.9579
2.9013	5.0	2745	2.9493
2.8283	6.0	3294	2.9476
2.8005	7.0	3843	2.9485
2.7523	8.0	4392	2.9494
2.7257	9.0	4941	2.9516
2.6927	10.0	5490	2.9594
2.6327	11.0	6039	2.9631
2.6106	12.0	6588	2.9706
2.5743	13.0	7137	2.9760
2.5676	14.0	7686	2.9845
2.5504	15.0	8235	2.9869
2.5242	16.0	8784	2.9978
2.5097	17.0	9333	3.0007
2.4938	18.0	9882	3.0114
2.4908	19.0	10431	3.0178
2.4731	20.0	10980	3.0204
2.4503	21.0	11529	3.0228
2.4441	22.0	12078	3.0254
2.4378	23.0	12627	3.0287
2.4286	24.0	13176	3.0312
2.4216	25.0	13725	3.0336

Framework versions

• Transformers 4.57.1
• Pytorch 2.8.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.1