metadata

library_name: peft
tags:
  - generated_from_trainer
base_model: mistralai/Mistral-7B-v0.1
datasets:
  - conll2003
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: billm-mistral-7b-conll03-ner-maxlen-256
    results: []

billm-mistral-7b-conll03-ner-maxlen-256

This model is a fine-tuned version of mistralai/Mistral-7B-v0.1 on the conll2003 dataset. It achieves the following results on the evaluation set:

Loss: 0.2232
Precision: 0.9277
Recall: 0.9363
F1: 0.9320
Accuracy: 0.9863

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: 2
eval_batch_size: 2
seed: 42
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
0.0329	1.0	7021	0.1599	0.9256	0.9357	0.9306	0.9856
0.0145	2.0	14042	0.1789	0.9312	0.9340	0.9326	0.9860
0.0106	3.0	21063	0.1931	0.9288	0.9359	0.9324	0.9864
0.0065	4.0	28084	0.2161	0.9277	0.9361	0.9319	0.9863
0.0043	5.0	35105	0.2168	0.9276	0.9363	0.9319	0.9864
0.002	6.0	42126	0.2250	0.9274	0.9359	0.9316	0.9863
0.0027	7.0	49147	0.2246	0.9269	0.9356	0.9312	0.9862
0.0023	8.0	56168	0.2235	0.9277	0.9364	0.9321	0.9863
0.0024	9.0	63189	0.2232	0.9276	0.9364	0.9320	0.9863
0.0016	10.0	70210	0.2232	0.9277	0.9363	0.9320	0.9863

Framework versions

PEFT 0.10.0
Transformers 4.41.2
Pytorch 2.3.0+cu121
Datasets 2.19.1
Tokenizers 0.19.1