mediconcenoss-v1

This model is a fine-tuned version of openai/gpt-oss-20b on the mediconcen_finetune_train dataset. It achieves the following results on the evaluation set:

• Loss: 0.3587

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: 0.0001
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 4
• 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: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 2.0

Training results

Training Loss	Epoch	Step	Validation Loss
0.5792	0.1379	100	0.5646
0.51	0.2759	200	0.4403
0.479	0.4138	300	0.4121
0.4308	0.5517	400	0.3971
0.4446	0.6897	500	0.3854
0.4635	0.8276	600	0.3771
0.4238	0.9655	700	0.3751
0.4187	1.1034	800	0.3719
0.3868	1.2414	900	0.3683
0.3985	1.3793	1000	0.3648
0.3834	1.5172	1100	0.3614
0.4284	1.6552	1200	0.3590
0.4227	1.7931	1300	0.3588
0.3957	1.9310	1400	0.3581

Framework versions

• PEFT 0.17.1
• Transformers 4.56.1
• Pytorch 2.8.0+cu128
• Datasets 3.2.0
• Tokenizers 0.22.1