README.md

metadata

license: apache-2.0
base_model: bert-base-cased
tags:
  - generated_from_trainer
datasets:
  - harem
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: bert-base-cased-finetuned-ner
    results:
      - task:
          name: Token Classification
          type: token-classification
        dataset:
          name: harem
          type: harem
          config: default
          split: validation
          args: default
        metrics:
          - name: Precision
            type: precision
            value: 0.3251366120218579
          - name: Recall
            type: recall
            value: 0.34097421203438394
          - name: F1
            type: f1
            value: 0.3328671328671328
          - name: Accuracy
            type: accuracy
            value: 0.8684278684278685

bert-base-cased-finetuned-ner

This model is a fine-tuned version of bert-base-cased on the harem dataset. It achieves the following results on the evaluation set:

• Loss: 0.5103
• Precision: 0.3251
• Recall: 0.3410
• F1: 0.3329
• Accuracy: 0.8684

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: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 40

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	4	1.1734	0.0	0.0	0.0	0.8083
No log	2.0	8	0.9781	0.0	0.0	0.0	0.8086
No log	3.0	12	0.8915	0.0	0.0	0.0	0.8086
No log	4.0	16	0.7901	0.0	0.0	0.0	0.8086
No log	5.0	20	0.7202	0.0	0.0	0.0	0.8086
No log	6.0	24	0.6846	0.4286	0.0344	0.0637	0.8130
No log	7.0	28	0.6596	0.2014	0.0802	0.1148	0.8306
No log	8.0	32	0.6355	0.1615	0.0745	0.1020	0.8324
No log	9.0	36	0.6193	0.1571	0.0946	0.1181	0.8345
No log	10.0	40	0.6106	0.1295	0.1032	0.1148	0.8335
No log	11.0	44	0.5919	0.1680	0.1232	0.1421	0.8350
No log	12.0	48	0.5789	0.2051	0.1375	0.1647	0.8384
No log	13.0	52	0.5827	0.1611	0.1375	0.1484	0.8355
No log	14.0	56	0.5638	0.2281	0.1862	0.2050	0.8433
No log	15.0	60	0.5576	0.1879	0.1691	0.1780	0.8420
No log	16.0	64	0.5485	0.2110	0.1862	0.1979	0.8456
No log	17.0	68	0.5479	0.2401	0.2264	0.2330	0.8500
No log	18.0	72	0.5460	0.2406	0.2378	0.2392	0.8503
No log	19.0	76	0.5374	0.2531	0.2350	0.2437	0.8542
No log	20.0	80	0.5365	0.2364	0.2493	0.2427	0.8539
No log	21.0	84	0.5284	0.2462	0.2350	0.2405	0.8552
No log	22.0	88	0.5306	0.2812	0.2837	0.2825	0.8601
No log	23.0	92	0.5262	0.2722	0.2722	0.2722	0.8573
No log	24.0	96	0.5306	0.2447	0.2665	0.2551	0.8555
No log	25.0	100	0.5249	0.2785	0.3009	0.2893	0.8594
No log	26.0	104	0.5201	0.2801	0.2865	0.2833	0.8586
No log	27.0	108	0.5213	0.2806	0.2894	0.2849	0.8604
No log	28.0	112	0.5207	0.2732	0.2951	0.2837	0.8612
No log	29.0	116	0.5144	0.3027	0.3209	0.3115	0.8630
No log	30.0	120	0.5135	0.3073	0.3381	0.3220	0.8648
No log	31.0	124	0.5147	0.2953	0.3266	0.3102	0.8651
No log	32.0	128	0.5121	0.2937	0.3181	0.3054	0.8645
No log	33.0	132	0.5092	0.3061	0.3324	0.3187	0.8645
No log	34.0	136	0.5064	0.3342	0.3696	0.3510	0.8677
No log	35.0	140	0.5056	0.3191	0.3438	0.3310	0.8674
No log	36.0	144	0.5091	0.3023	0.3352	0.3179	0.8661
No log	37.0	148	0.5104	0.3061	0.3324	0.3187	0.8658
No log	38.0	152	0.5100	0.3152	0.3324	0.3236	0.8677
No log	39.0	156	0.5102	0.3243	0.3410	0.3324	0.8684
No log	40.0	160	0.5103	0.3251	0.3410	0.3329	0.8684

Framework versions

• Transformers 4.32.1
• Pytorch 2.0.0
• Datasets 2.1.0
• Tokenizers 0.13.3