gemma_FT_2

This model is a fine-tuned version of google/gemma-2b-it on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.6202

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: 0.0002
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 4
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 2
• training_steps: 12
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
3.0277	0.0005	1	3.5558
3.3686	0.0011	2	3.4906
3.285	0.0016	3	3.3311
3.025	0.0021	4	3.1640
3.2431	0.0026	5	3.0050
3.0431	0.0032	6	2.8713
2.3602	0.0037	7	2.7724
2.8696	0.0042	8	2.7056
2.6496	0.0047	9	2.6637
2.6983	0.0053	10	2.6401
2.2396	0.0058	11	2.6268
2.4519	0.0063	12	2.6202

Framework versions

• PEFT 0.8.2
• Transformers 4.41.1
• Pytorch 1.13.1+cu117
• Datasets 2.19.1
• Tokenizers 0.19.1