finetune_arc_20

This model is a fine-tuned version of lmsys/vicuna-7b-v1.5 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.7783

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: 0.0001
• train_batch_size: 4
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 5
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
1.3635	1.0	150	1.2844
0.8018	2.0	300	1.3583
0.4765	3.0	450	1.5943
0.2923	4.0	600	1.8834
0.1973	5.0	750	1.9693
0.1986	6.0	900	2.0187
0.1769	7.0	1050	2.1674
0.1359	8.0	1200	2.1402
0.1778	9.0	1350	2.3226
0.1353	10.0	1500	2.3321
0.1426	11.0	1650	2.4006
0.1412	12.0	1800	2.5354
0.164	13.0	1950	2.5339
0.1034	14.0	2100	2.5972
0.1087	15.0	2250	2.6059
0.0878	16.0	2400	2.6054
0.0985	17.0	2550	2.6881
0.1018	18.0	2700	2.7388
0.1091	19.0	2850	2.7657
0.0846	20.0	3000	2.7783

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu121
• Datasets 2.15.0
• Tokenizers 0.15.0