lmind_hotpot_train8000_eval7405_v1_doc_qa_meta-llama_Llama-2-7b-hf_3e-5_lora2

This model is a fine-tuned version of meta-llama/Llama-2-7b-hf on the tyzhu/lmind_hotpot_train8000_eval7405_v1_doc_qa dataset. It achieves the following results on the evaluation set:

• Loss: 2.8310
• Accuracy: 0.5804

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: 3e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• total_eval_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: constant
• lr_scheduler_warmup_ratio: 0.05
• num_epochs: 20.0

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.2166	1.0	1089	1.8581	0.5926
1.1949	2.0	2178	1.8345	0.5961
1.1542	3.0	3267	1.8206	0.5964
1.1236	4.0	4357	1.8230	0.6004
1.0989	5.0	5446	1.8371	0.6036
1.0543	6.0	6535	1.8655	0.6022
1.0139	7.0	7624	1.9280	0.5929
0.9764	8.0	8714	1.9914	0.5913
0.9351	9.0	9803	2.0565	0.5909
0.9177	10.0	10892	2.1248	0.5892
0.8872	11.0	11981	2.2182	0.5875
0.8458	12.0	13071	2.2863	0.5863
0.8148	13.0	14160	2.3525	0.5842
0.7955	14.0	15249	2.4442	0.5834
0.7765	15.0	16338	2.4962	0.5836
0.7412	16.0	17428	2.5740	0.5818
0.7257	17.0	18517	2.6537	0.5817
0.7085	18.0	19606	2.6944	0.5813
0.6756	19.0	20695	2.7615	0.5795
0.6493	20.0	21780	2.8310	0.5804

Framework versions

• Transformers 4.34.0
• Pytorch 2.1.0+cu121
• Datasets 2.18.0
• Tokenizers 0.14.1