finetune_cs_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: 2.9803

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
1.7173	1.0	150	1.6256
1.5959	2.0	300	1.6775
0.5699	3.0	450	1.9236
0.4921	4.0	600	2.1318
0.3044	5.0	750	2.2606
0.2809	6.0	900	2.3494
0.2299	7.0	1050	2.4316
0.2073	8.0	1200	2.4576
0.1851	9.0	1350	2.4981
0.1906	10.0	1500	2.6060
0.1616	11.0	1650	2.6427
0.1529	12.0	1800	2.6856
0.1453	13.0	1950	2.7683
0.1507	14.0	2100	2.7889
0.1607	15.0	2250	2.8477
0.1545	16.0	2400	2.8710
0.1598	17.0	2550	2.8968
0.1593	18.0	2700	2.9389
0.1479	19.0	2850	2.9629
0.1298	20.0	3000	2.9803

Framework versions

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