README.md

metadata

license: apache-2.0
base_model: distilbert-base-uncased
tags:
  - generated_from_trainer
metrics:
  - f1
  - precision
  - recall
  - accuracy
model-index:
  - name: distil-bert-imeoocap
    results: []

distil-bert-imeoocap

This model is a fine-tuned version of distilbert-base-uncased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.5520
• F1: 0.6416
• Precision: 0.6437
• Recall: 0.6442
• Accuracy: 0.6442

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: 2e-05
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Precision	Recall	Accuracy
0.4777	1.0	74	1.1554	0.6498	0.6489	0.6538	0.6538
0.395	2.0	148	1.2060	0.6062	0.6109	0.6135	0.6135
0.364	3.0	222	1.2625	0.6329	0.6436	0.6423	0.6423
0.3402	4.0	296	1.3512	0.6247	0.6330	0.6269	0.6269
0.3135	5.0	370	1.3587	0.6472	0.6442	0.6519	0.6519
0.307	6.0	444	1.4376	0.6258	0.6334	0.6288	0.6288
0.2903	7.0	518	1.3565	0.6502	0.6550	0.65	0.65
0.2931	8.0	592	1.4059	0.6310	0.6273	0.6365	0.6365
0.2805	9.0	666	1.3972	0.6357	0.6370	0.6365	0.6365
0.2772	10.0	740	1.4938	0.6205	0.6204	0.6308	0.6308
0.2767	11.0	814	1.4324	0.6256	0.6324	0.6269	0.6269
0.2634	12.0	888	1.5399	0.6457	0.6487	0.65	0.65
0.2829	13.0	962	1.4857	0.6369	0.6363	0.6385	0.6385
0.2444	14.0	1036	1.4879	0.6314	0.6385	0.6308	0.6308
0.2424	15.0	1110	1.5049	0.6357	0.6399	0.6365	0.6365
0.2332	16.0	1184	1.5277	0.6233	0.6268	0.625	0.625
0.2215	17.0	1258	1.5550	0.6390	0.6379	0.6423	0.6423
0.2235	18.0	1332	1.5593	0.6434	0.6449	0.6481	0.6481
0.213	19.0	1406	1.5631	0.6337	0.6303	0.6385	0.6385
0.2098	20.0	1480	1.5520	0.6416	0.6437	0.6442	0.6442

Framework versions

• Transformers 4.39.3
• Pytorch 2.1.2
• Datasets 2.18.0
• Tokenizers 0.15.2