Mistral-7B-v0.1_cola_original2

This model is a fine-tuned version of mistralai/Mistral-7B-v0.1 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3773
• Accuracy: {'accuracy': 0.8539719626168224}
• Matthews Correlation: 0.6454

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: 1e-05
• train_batch_size: 64
• eval_batch_size: 64
• seed: 2
• distributed_type: multi-GPU
• num_devices: 6
• total_train_batch_size: 384
• total_eval_batch_size: 384
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• training_steps: 750

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Matthews Correlation
0.4986	2.38	50	0.5042	{'accuracy': 0.7938638542665388}	0.4836
0.3193	4.76	100	0.4002	{'accuracy': 0.8264621284755513}	0.6029
0.2489	7.14	150	0.3795	{'accuracy': 0.8389261744966443}	0.6123
0.1258	9.52	200	0.4322	{'accuracy': 0.8418024928092043}	0.6284
0.0625	11.9	250	0.5921	{'accuracy': 0.8427612655800575}	0.6258
0.0251	14.29	300	0.8451	{'accuracy': 0.8446788111217641}	0.6248

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.1+cu121
• Datasets 2.15.0
• Tokenizers 0.15.0