lc-7b-sft-lora

This model is a fine-tuned version of mistralai/Mistral-7B-v0.1 on the generator dataset. It achieves the following results on the evaluation set:

• Loss: 1.4782

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: 2e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
1.6469	1.0	16	1.6182
1.5733	2.0	32	1.5646
1.5088	3.0	48	1.5251
1.4873	4.0	64	1.4960
1.4409	5.0	80	1.4758
1.4061	6.0	96	1.4655
1.4014	7.0	112	1.4588
1.3723	8.0	128	1.4573
1.346	9.0	144	1.4604
1.306	10.0	160	1.4605
1.307	11.0	176	1.4639
1.3115	12.0	192	1.4663
1.3093	13.0	208	1.4683
1.2644	14.0	224	1.4705
1.2641	15.0	240	1.4748
1.2578	16.0	256	1.4754
1.2538	17.0	272	1.4771
1.2611	18.0	288	1.4779
1.2609	19.0	304	1.4783
1.2805	20.0	320	1.4782

Framework versions

• PEFT 0.11.1
• Transformers 4.41.0
• Pytorch 2.0.1+cu118
• Datasets 2.19.1
• Tokenizers 0.19.1