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: 1.8340

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
0.4276	1.0	1	1.9583
0.4276	2.0	2	1.9512
0.4107	3.0	3	1.9324
0.3762	4.0	4	1.9108
0.3493	5.0	5	1.8891
0.3269	6.0	6	1.8698
0.3086	7.0	7	1.8545
0.2947	8.0	8	1.8435
0.2844	9.0	9	1.8370
0.2775	10.0	10	1.8340

Framework versions

• PEFT 0.11.0
• Transformers 4.40.2
• Pytorch 2.1.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1