llama_lora_20k_alpaca_sencond_tiny_learingRate

This model is a fine-tuned version of mifeng09/my_final_llama_model_v2_add_wiki_fix_resume on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 3.8042

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: 8e-05
• train_batch_size: 4
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• 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_ratio: 0.1
• lr_scheduler_warmup_steps: 67
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss
3.8201	0.2227	50	4.0046
4.0256	0.4454	100	3.9221
3.6268	0.6682	150	3.8848
3.8767	0.8909	200	3.8590
3.8853	1.1114	250	3.8412
3.8609	1.3341	300	3.8283
3.8614	1.5568	350	3.8205
3.8693	1.7795	400	3.8144
3.7834	2.0	450	3.8096
3.9981	2.2227	500	3.8062
3.7208	2.4454	550	3.8050
3.9458	2.6682	600	3.8044
3.7216	2.8909	650	3.8042

Framework versions

• PEFT 0.18.0
• Transformers 4.57.3
• Pytorch 2.9.0+cpu
• Datasets 4.0.0
• Tokenizers 0.22.1