finetune_cs_20_cot

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: 2.9873

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
1.7119	1.0	150	1.6237
1.5767	2.0	300	1.6768
0.5982	3.0	450	1.9207
0.5098	4.0	600	2.1488
0.3085	5.0	750	2.2398
0.2595	6.0	900	2.2685
0.2118	7.0	1050	2.4754
0.2221	8.0	1200	2.4747
0.1796	9.0	1350	2.5642
0.1706	10.0	1500	2.5938
0.161	11.0	1650	2.6814
0.1484	12.0	1800	2.7009
0.1408	13.0	1950	2.7635
0.1472	14.0	2100	2.7593
0.154	15.0	2250	2.8357
0.1455	16.0	2400	2.8705
0.164	17.0	2550	2.9052
0.1613	18.0	2700	2.9437
0.1397	19.0	2850	2.9716
0.1177	20.0	3000	2.9873

Framework versions

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