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

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: 0.4998
• Precision: 0.0
• Recall: 0.0
• F1: 0.0
• Accuracy: 0.7829

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: 0.001
• 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	0.9784	34	0.9557	0.0	0.0	0.0	0.7596
No log	1.9856	69	0.7691	0.0	0.0	0.0	0.7707
No log	2.9928	104	0.7086	0.0	0.0	0.0	0.7794
No log	4.0	139	0.5693	0.0	0.0	0.0	0.7697
No log	4.9784	173	0.5449	0.0	0.0	0.0	0.7758
No log	5.9856	208	0.5168	0.0	0.0	0.0	0.7805
No log	6.9928	243	0.5379	0.0	0.0	0.0	0.7838
No log	8.0	278	0.5301	0.0	0.0	0.0	0.7847
No log	8.9784	312	0.5007	0.0	0.0	0.0	0.7829
No log	9.7842	340	0.4998	0.0	0.0	0.0	0.7829

Framework versions

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