finetune_race_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: 3.2382

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
1.218	1.0	150	1.3450
0.8105	2.0	300	1.3977
0.5315	3.0	450	1.4943
0.6257	4.0	600	1.6812
0.1961	5.0	750	1.9107
0.141	6.0	900	2.1245
0.1549	7.0	1050	2.1764
0.0783	8.0	1200	2.4002
0.0846	9.0	1350	2.4001
0.0518	10.0	1500	2.5191
0.0568	11.0	1650	2.5257
0.0498	12.0	1800	2.5847
0.0494	13.0	1950	2.7678
0.0326	14.0	2100	2.8382
0.035	15.0	2250	2.9128
0.0389	16.0	2400	3.0079
0.0427	17.0	2550	3.0754
0.0353	18.0	2700	3.1909
0.0368	19.0	2850	3.2083
0.0316	20.0	3000	3.2382

Framework versions

• Transformers 4.36.0.dev0
• Pytorch 2.1.0+cu121
• Datasets 2.13.1
• Tokenizers 0.14.1