Yi-6B-ruozhiba3

This model is a fine-tuned version of 01-ai/Yi-6B on the ruozhiba dataset. It achieves the following results on the evaluation set:

• Loss: 4.1874

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: 5e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss
2.7671	1.0	55	2.3123
2.0319	2.0	110	1.9679
1.7972	3.0	165	1.9426
1.5841	4.0	220	2.0110
1.2842	5.0	275	2.2671
0.9305	6.0	330	2.5263
0.6734	7.0	385	2.7798
0.4579	8.0	440	3.1052
0.3091	9.0	495	3.3409
0.2418	10.0	550	3.4999
0.1718	11.0	605	3.6688
0.1555	12.0	660	3.7819
0.1191	13.0	715	3.9108
0.1291	14.0	770	3.9953
0.1213	15.0	825	4.1020
0.1	16.0	880	4.1205
0.115	17.0	935	4.1606
0.1076	18.0	990	4.1839
0.0962	19.0	1045	4.1873
0.0917	20.0	1100	4.1874

Framework versions

• PEFT 0.7.1
• Transformers 4.36.2
• Pytorch 2.2.2+cu118
• Datasets 2.14.6
• Tokenizers 0.15.2