metadata

base_model: microsoft/mdeberta-v3-base
library_name: transformers
license: mit
metrics:
  - precision
  - recall
  - f1
  - accuracy
tags:
  - generated_from_trainer
model-index:
  - name: scenario-non-kd-pre-ner-full-mdeberta_data-univner_en66
    results: []

scenario-non-kd-pre-ner-full-mdeberta_data-univner_en66

This model is a fine-tuned version of microsoft/mdeberta-v3-base on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.1822
Precision: 0.7036
Recall: 0.7226
F1: 0.7130
Accuracy: 0.9771

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: 3e-05
train_batch_size: 32
eval_batch_size: 32
seed: 66
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.1469	1.2755	500	0.1352	0.4213	0.4658	0.4425	0.9588
0.066	2.5510	1000	0.1071	0.5530	0.6429	0.5945	0.9700
0.0349	3.8265	1500	0.1015	0.6313	0.7267	0.6756	0.9738
0.0222	5.1020	2000	0.1179	0.6442	0.7516	0.6937	0.9748
0.0137	6.3776	2500	0.1166	0.6736	0.7350	0.7030	0.9760
0.009	7.6531	3000	0.1178	0.6533	0.7236	0.6866	0.9753
0.0068	8.9286	3500	0.1312	0.6976	0.7091	0.7033	0.9763
0.0048	10.2041	4000	0.1438	0.6793	0.7215	0.6998	0.9754
0.0038	11.4796	4500	0.1479	0.6906	0.7371	0.7131	0.9761
0.0031	12.7551	5000	0.1503	0.6745	0.7422	0.7068	0.9760
0.0025	14.0306	5500	0.1489	0.6886	0.7143	0.7012	0.9768
0.0019	15.3061	6000	0.1565	0.6755	0.7350	0.7040	0.9757
0.0018	16.5816	6500	0.1564	0.7004	0.7453	0.7222	0.9771
0.0014	17.8571	7000	0.1661	0.7066	0.7381	0.7220	0.9771
0.0013	19.1327	7500	0.1724	0.6811	0.7164	0.6983	0.9757
0.0008	20.4082	8000	0.1760	0.6788	0.7329	0.7048	0.9763
0.0007	21.6837	8500	0.1777	0.6900	0.7350	0.7118	0.9768
0.0008	22.9592	9000	0.1821	0.6795	0.7464	0.7114	0.9761
0.0006	24.2347	9500	0.1888	0.6715	0.7660	0.7157	0.9763
0.0005	25.5102	10000	0.1830	0.6973	0.7298	0.7132	0.9767
0.0005	26.7857	10500	0.1805	0.7045	0.7257	0.7149	0.9769
0.0004	28.0612	11000	0.1812	0.6963	0.7215	0.7087	0.9765
0.0003	29.3367	11500	0.1822	0.7036	0.7226	0.7130	0.9771

Framework versions

Transformers 4.44.2
Pytorch 2.1.1+cu121
Datasets 2.14.5
Tokenizers 0.19.1