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

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

• Loss: 0.6713

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
1.0825	0.9091	5	1.0682
0.9795	2.0	11	0.9158
0.7555	2.9091	16	0.8181
0.7129	4.0	22	0.7546
0.6824	4.9091	27	0.7213
0.6146	6.0	33	0.6980
0.5689	6.9091	38	0.6854
0.5226	8.0	44	0.6731
0.4847	8.9091	49	0.6713
0.4181	10.0	55	0.6911
0.3633	10.9091	60	0.7228
0.2845	12.0	66	0.7945
0.2149	12.9091	71	0.9006
0.1446	14.0	77	1.0794
0.0753	14.9091	82	1.2625
0.0483	16.0	88	1.4373

Framework versions

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