epidemiology_sft_10000_mcq

This model is a fine-tuned version of mistralai/Mistral-Nemo-Instruct-2407 on the epidemiology_10000_mcq dataset. It achieves the following results on the evaluation set:

• Loss: 0.0037

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: 12
• eval_batch_size: 12
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• total_train_batch_size: 48
• total_eval_batch_size: 48
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss
0.0059	0.1596	30	0.0058
0.0058	0.3191	60	0.0059
0.0058	0.4787	90	0.0058
0.0058	0.6383	120	0.0058
0.0058	0.7979	150	0.0058
0.0057	0.9574	180	0.0057
0.0057	1.1170	210	0.0057
0.0056	1.2766	240	0.0059
0.0052	1.4362	270	0.0046
0.004	1.5957	300	0.0038
0.0041	1.7553	330	0.0038
0.0042	1.9149	360	0.0037

Framework versions

• PEFT 0.12.0
• Transformers 4.46.0
• Pytorch 2.4.0+cu121
• Datasets 2.21.0
• Tokenizers 0.20.1