Medical-Finetune-BioMistral

This model is a fine-tuned version of BioMistral/BioMistral-7B on the generator dataset. It achieves the following results on the evaluation set:

• Loss: 2.9639
• F1: 0.0
• Accuracy: 0.0

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: 0.0002
• train_batch_size: 2
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.03
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Accuracy
0.7654	0.3436	100	2.8037	0.0	0.0
1.9543	0.6873	200	2.1952	0.0	0.0
1.9931	1.0309	300	2.1921	0.0143	0.0143
1.3308	1.3746	400	2.2393	0.0	0.0
1.4068	1.7182	500	2.2237	0.0	0.0
1.2879	2.0619	600	2.4974	0.0	0.0
0.7795	2.4055	700	2.3918	0.0	0.0
0.8464	2.7491	800	2.4622	0.0143	0.0143
0.7941	3.0928	900	2.6181	0.0	0.0
0.4868	3.4364	1000	2.6465	0.0	0.0
0.5224	3.7801	1100	2.6711	0.0	0.0
0.5074	4.1237	1200	2.8354	0.0	0.0
0.3309	4.4674	1300	2.8183	0.0	0.0
0.3761	4.8110	1400	2.8540	0.0	0.0
0.3173	5.1546	1500	2.9639	0.0	0.0

Framework versions

• PEFT 0.11.1
• Transformers 4.41.1
• Pytorch 2.1.2
• Datasets 2.19.1
• Tokenizers 0.19.1