Gemma-2-2B_task-2_180-samples_config-1_full_auto

This model is a fine-tuned version of google/gemma-2-2b-it on the GaetanMichelet/chat-60_ft_task-2_auto, the GaetanMichelet/chat-120_ft_task-2_auto and the GaetanMichelet/chat-180_ft_task-2_auto datasets. It achieves the following results on the evaluation set:

• Loss: 0.8559

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: 0.0001
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• distributed_type: multi-GPU
• gradient_accumulation_steps: 8
• total_train_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss
1.2757	1.0	17	1.2558
1.0993	2.0	34	1.0768
0.9179	3.0	51	0.9359
0.8135	4.0	68	0.8881
0.7851	5.0	85	0.8660
0.7461	6.0	102	0.8559
0.7244	7.0	119	0.8581
0.6484	8.0	136	0.8743
0.5865	9.0	153	0.9082
0.5719	10.0	170	0.9505
0.5453	11.0	187	1.0115
0.4079	12.0	204	1.1288
0.4321	13.0	221	1.1490

Framework versions

• PEFT 0.12.0
• Transformers 4.44.0
• Pytorch 2.1.2+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1