mistral-7b-32k-billm-finetuned-token-classification

This model is a fine-tuned version of mistralai/Mistral-7B-Instruct-v0.2 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.6070
• Precision: 0.0
• Recall: 0.0
• F1: 0.0
• Accuracy: 0.7473

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: 0.0005
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 32
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	0.9957	73	1.0053	0.0	0.0	0.0	0.7275
No log	1.9915	146	0.8578	0.0	0.0	0.0	0.7426
No log	2.9872	219	0.7958	0.0	0.0	0.0	0.7195
No log	3.9966	293	0.6437	0.0	0.0	0.0	0.7144
No log	4.9923	366	0.6283	0.0	0.0	0.0	0.7318
No log	5.9881	439	0.6543	0.0	0.0	0.0	0.7553
0.9335	6.9974	513	0.6112	0.0	0.0	0.0	0.7037
0.9335	7.9932	586	0.6053	0.0	0.0	0.0	0.7373
0.9335	8.9889	659	0.6208	0.0	0.0	0.0	0.7518
0.9335	9.9574	730	0.6070	0.0	0.0	0.0	0.7473

Framework versions

• PEFT 0.10.0
• Transformers 4.40.1
• Pytorch 2.2.2
• Datasets 2.18.0
• Tokenizers 0.19.1