formalmathatepfl/mistral-7B-v0.3-finetuned

mistral-lean-sft

This model is a fine-tuned version of mistralai/Mistral-7B-v0.3 on the lean_sft dataset. It achieves the following results on the evaluation set:

• Loss: 0.0542

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: 2
• eval_batch_size: 2
• seed: 42
• distributed_type: multi-GPU
• num_devices: 8
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• total_eval_batch_size: 16
• 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: 0.05
• num_epochs: 1.0

Training results

Training Loss	Epoch	Step	Validation Loss
0.0817	0.1633	1000	0.0849
0.0766	0.3266	2000	0.0762
0.0628	0.4900	3000	0.0654
0.0594	0.6533	4000	0.0602
0.0547	0.8166	5000	0.0559
0.0537	0.9799	6000	0.0542

Framework versions

• Transformers 4.57.3
• Pytorch 2.9.0+cu129
• Datasets 4.0.0
• Tokenizers 0.22.1