Gemma-2-2B_task-2_120-samples_config-2_auto

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

• Loss: 0.6387

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: 16
• total_train_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: 50

Training results

Training Loss	Epoch	Step	Validation Loss
1.0219	0.9091	5	1.0249
0.9649	2.0	11	0.8980
0.7596	2.9091	16	0.8028
0.6873	4.0	22	0.7306
0.6361	4.9091	27	0.6907
0.5688	6.0	33	0.6625
0.5268	6.9091	38	0.6464
0.4821	8.0	44	0.6387
0.439	8.9091	49	0.6490
0.3766	10.0	55	0.6785
0.3045	10.9091	60	0.7300
0.2261	12.0	66	0.8326
0.1684	12.9091	71	0.9590
0.097	14.0	77	1.1236
0.0478	14.9091	82	1.2809

Framework versions

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