gemma-7b-it-dolly-15k-japanese-brainstorming-ipo

This model is a fine-tuned version of google/gemma-7b-it on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.4864
• Rouge Scores: {'rouge1': 0.8349379511354852, 'rouge2': 0.7274477012465996, 'rougeL': 0.8017731965466872, 'rougeLsum': 0.8347274626949961}
• Bleu Scores: [0.8344549267092076, 0.7702670675251023, 0.7108107057859971, 0.6501643333897235]
• Gen Len: 241.8588

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: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge Scores	Bleu Scores	Gen Len
2.6263	1.0	398	2.3289	{'rouge1': 0.8429091284800581, 'rouge2': 0.6936436106699162, 'rougeL': 0.8148324021885509, 'rougeLsum': 0.8436115528203997}	[0.8182599825705432, 0.7573989130361584, 0.6993616529579798, 0.6397083137823109]	241.8588
1.4454	2.0	796	2.2673	{'rouge1': 0.8676191825821662, 'rouge2': 0.7714826591897748, 'rougeL': 0.8384600375456672, 'rougeLsum': 0.8676504211460437}	[0.838231259049653, 0.7749860292910674, 0.715811357776676, 0.6553787760578684]	241.8588
0.6829	3.0	1194	2.4864	{'rouge1': 0.8349379511354852, 'rouge2': 0.7274477012465996, 'rougeL': 0.8017731965466872, 'rougeLsum': 0.8347274626949961}	[0.8344549267092076, 0.7702670675251023, 0.7108107057859971, 0.6501643333897235]	241.8588

Framework versions

• PEFT 0.8.2
• Transformers 4.39.0.dev0
• Pytorch 2.1.0+cu121
• Datasets 2.17.2.dev0
• Tokenizers 0.15.2