Mistral-7B_task-3_120-samples_config-1_full

This model is a fine-tuned version of mistralai/Mistral-7B-Instruct-v0.3 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: 0.8608

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
1.1682	1.0	11	1.1587
0.9812	2.0	22	1.0306
0.8834	3.0	33	0.9015
0.8826	4.0	44	0.8727
0.8207	5.0	55	0.8608
0.7611	6.0	66	0.8610
0.6922	7.0	77	0.8744
0.6264	8.0	88	0.9150
0.4961	9.0	99	0.9764
0.3722	10.0	110	1.0251
0.3418	11.0	121	1.1046
0.2556	12.0	132	1.1179

Framework versions

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