This model is a fine-tuned version of google/gemma-2-2b-it on the GaetanMichelet/chat-60_ft_task-3 and the GaetanMichelet/chat-120_ft_task-3 datasets. It achieves the following results on the evaluation set:

Loss: 0.4495

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
3.2814	0.9091	5	3.2207
2.3853	2.0	11	1.7735
1.0517	2.9091	16	0.9646
0.6575	4.0	22	0.6096
0.4492	4.9091	27	0.5386
0.3535	6.0	33	0.4848
0.3521	6.9091	38	0.4613
0.2821	8.0	44	0.4495
0.1878	8.9091	49	0.4496
0.1547	10.0	55	0.4928
0.1539	10.9091	60	0.4977
0.1194	12.0	66	0.6056
0.0872	12.9091	71	0.6938
0.0274	14.0	77	0.8379
0.0195	14.9091	82	0.8741

Framework versions

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

GaetanMichelet/Gemma-2-2B_task-3_120-samples_config-2

Gemma-2-2B_task-3_120-samples_config-2

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

GaetanMichelet/
Gemma-2-2B_task-3_120-samples_config-2