Saiga_timelist_task20steps

This model is a fine-tuned version of TheBloke/Llama-2-7B-fp16 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.0584

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: 0.0003
• train_batch_size: 2
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 10
• total_train_batch_size: 20
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• training_steps: 20

Training results

Training Loss	Epoch	Step	Validation Loss
2.2298	0.37	2	2.2031
2.0996	0.74	4	2.1519
2.0299	1.11	6	2.1202
2.0007	1.48	8	2.0978
1.9777	1.85	10	2.0817
1.9089	2.22	12	2.0715
1.9379	2.59	14	2.0650
1.9515	2.96	16	2.0610
1.9178	3.33	18	2.0589
1.8801	3.7	20	2.0584

Framework versions

• PEFT 0.10.0
• Transformers 4.39.3
• Pytorch 2.2.2+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2