gemma7b-summarize-gpt4o-4k

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

• Loss: 6.0322

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: 2
• seed: 42
• distributed_type: multi-GPU
• num_devices: 8
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• total_eval_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss
44.3098	1.0	7	13.7204
25.7366	2.0	14	8.6916
19.0375	3.0	21	7.6308
18.2973	4.0	28	7.1198
14.8387	5.0	35	6.8470
12.5684	6.0	42	6.8495
9.6308	7.0	49	6.6799
5.7187	8.0	56	6.0818
4.8487	9.0	63	6.0318
4.4303	10.0	70	6.0322

Framework versions

• PEFT 0.10.0
• Transformers 4.40.0
• Pytorch 2.1.2+cu121
• Datasets 2.18.0
• Tokenizers 0.19.1