gemma-2b-dolly-qa

This model is a fine-tuned version of google/gemma-2b on the generator dataset. It achieves the following results on the evaluation set:

• Loss: 2.0211

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: 1e-05
• train_batch_size: 2
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.05
• training_steps: 1480

Training results

Training Loss	Epoch	Step	Validation Loss
2.9203	1.6393	100	2.5631
2.4253	3.2787	200	2.2695
2.2443	4.9180	300	2.1581
2.1512	6.5574	400	2.1002
2.1033	8.1967	500	2.0723
2.0876	9.8361	600	2.0565
2.0668	11.4754	700	2.0460
2.0478	13.1148	800	2.0387
2.0403	14.7541	900	2.0328
2.0366	16.3934	1000	2.0286
2.0238	18.0328	1100	2.0255
2.0231	19.6721	1200	2.0233
2.0126	21.3115	1300	2.0220
2.0164	22.9508	1400	2.0211

Framework versions

• PEFT 0.10.0
• Transformers 4.40.1
• Pytorch 2.1.0.post0+cxx11.abi
• Datasets 2.19.0
• Tokenizers 0.19.1