bloomz-560m_lora_lr5e-05_bs4_epoch20_wd0.01

This model is a fine-tuned version of bigscience/bloomz-560m on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 4.8047

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: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss
4.2484	1.0	157	3.6028
3.5565	2.0	314	3.3295
3.3894	3.0	471	3.2916
3.2189	4.0	628	3.2876
3.1204	5.0	785	3.3139
2.9755	6.0	942	3.3707
2.919	7.0	1099	3.4452
2.7196	8.0	1256	3.5097
2.6831	9.0	1413	3.6234
2.5105	10.0	1570	3.7513
2.3998	11.0	1727	3.8193
2.314	12.0	1884	3.9362
2.1941	13.0	2041	4.1743
2.1129	14.0	2198	4.2149
2.0442	15.0	2355	4.3023
1.8967	16.0	2512	4.4912
1.9345	17.0	2669	4.5690
1.8908	18.0	2826	4.6751
1.8891	19.0	2983	4.7869
1.7753	20.0	3140	4.8047

Framework versions

• PEFT 0.7.1
• Transformers 4.36.2
• Pytorch 2.0.1
• Datasets 2.16.1
• Tokenizers 0.15.0