shawgpt-ft

This model is a fine-tuned version of TheBloke/Mistral-7B-Instruct-v0.2-GPTQ on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.6063

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: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 2
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.6351	1.0	1	1.1246
0.6319	2.0	2	1.0575
0.6034	3.0	3	0.9250
0.5414	4.0	4	0.8252
0.4959	5.0	5	0.7540
0.5698	6.0	6	0.7035
0.4303	7.0	7	0.6648
0.4095	8.0	8	0.6361
0.3955	9.0	9	0.6165
0.3857	10.0	10	0.6063

Framework versions

• PEFT 0.10.0
• Transformers 4.38.2
• Pytorch 2.1.0+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2