mistral-finetunedv2

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

• Loss: 0.3008

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: 1
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 8
• 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
• lr_scheduler_warmup_steps: 50
• training_steps: 500

Training results

Training Loss	Epoch	Step	Validation Loss
0.4742	0.4545	50	0.4152
0.3541	0.9091	100	0.3482
0.329	1.3636	150	0.3183
0.3116	1.8182	200	0.3054
0.2993	2.2727	250	0.3028
0.3058	2.7273	300	0.3032
0.2978	3.1818	350	0.3020
0.2952	3.6364	400	0.3019
0.294	4.0909	450	0.3008
0.2946	4.5455	500	0.3008

Framework versions

• PEFT 0.17.1
• Transformers 4.57.1
• Pytorch 2.5.1+cu121
• Datasets 4.3.0
• Tokenizers 0.22.1