File size: 4,454 Bytes
f80157a
ca0a4fd
f80157a
 
 
 
 
 
 
ca0a4fd
 
f80157a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
---
base_model: haryoaw/scenario-TCR-NER_data-univner_half
library_name: transformers
license: mit
metrics:
- precision
- recall
- f1
- accuracy
tags:
- generated_from_trainer
model-index:
- name: scenario-non-kd-pre-ner-full-xlmr_data-univner_half55
  results: []
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# scenario-non-kd-pre-ner-full-xlmr_data-univner_half55

This model is a fine-tuned version of [haryoaw/scenario-TCR-NER_data-univner_half](https://huggingface.co/haryoaw/scenario-TCR-NER_data-univner_half) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1450
- Precision: 0.8516
- Recall: 0.8357
- F1: 0.8436
- Accuracy: 0.9832

## 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: 55
- 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 |
|:-------------:|:-------:|:-----:|:---------------:|:---------:|:------:|:------:|:--------:|
| 0.0106        | 0.5828  | 500   | 0.0804          | 0.8385    | 0.8489 | 0.8437 | 0.9838   |
| 0.0101        | 1.1655  | 1000  | 0.0863          | 0.8549    | 0.8432 | 0.8490 | 0.9838   |
| 0.009         | 1.7483  | 1500  | 0.0960          | 0.8282    | 0.8557 | 0.8418 | 0.9829   |
| 0.0077        | 2.3310  | 2000  | 0.1040          | 0.8320    | 0.8498 | 0.8408 | 0.9827   |
| 0.0077        | 2.9138  | 2500  | 0.0931          | 0.8461    | 0.8478 | 0.8469 | 0.9835   |
| 0.0054        | 3.4965  | 3000  | 0.0982          | 0.8482    | 0.8523 | 0.8502 | 0.9843   |
| 0.0061        | 4.0793  | 3500  | 0.1075          | 0.8450    | 0.8355 | 0.8402 | 0.9832   |
| 0.0053        | 4.6620  | 4000  | 0.1068          | 0.8409    | 0.8536 | 0.8472 | 0.9839   |
| 0.0047        | 5.2448  | 4500  | 0.1073          | 0.8372    | 0.8592 | 0.8480 | 0.9837   |
| 0.0047        | 5.8275  | 5000  | 0.1100          | 0.8451    | 0.8585 | 0.8517 | 0.9839   |
| 0.0034        | 6.4103  | 5500  | 0.1167          | 0.8350    | 0.8497 | 0.8422 | 0.9832   |
| 0.0034        | 6.9930  | 6000  | 0.1122          | 0.8405    | 0.8478 | 0.8441 | 0.9836   |
| 0.0035        | 7.5758  | 6500  | 0.1125          | 0.8424    | 0.8419 | 0.8421 | 0.9834   |
| 0.0032        | 8.1585  | 7000  | 0.1145          | 0.8454    | 0.8504 | 0.8479 | 0.9836   |
| 0.0035        | 8.7413  | 7500  | 0.1075          | 0.8499    | 0.8407 | 0.8453 | 0.9838   |
| 0.0027        | 9.3240  | 8000  | 0.1213          | 0.8493    | 0.8384 | 0.8438 | 0.9837   |
| 0.0031        | 9.9068  | 8500  | 0.1083          | 0.8551    | 0.8440 | 0.8495 | 0.9842   |
| 0.0027        | 10.4895 | 9000  | 0.1273          | 0.8329    | 0.8639 | 0.8482 | 0.9835   |
| 0.0024        | 11.0723 | 9500  | 0.1247          | 0.8478    | 0.8411 | 0.8444 | 0.9834   |
| 0.0021        | 11.6550 | 10000 | 0.1161          | 0.8487    | 0.8378 | 0.8432 | 0.9838   |
| 0.0019        | 12.2378 | 10500 | 0.1284          | 0.8316    | 0.8556 | 0.8434 | 0.9830   |
| 0.0021        | 12.8205 | 11000 | 0.1208          | 0.8492    | 0.8510 | 0.8501 | 0.9840   |
| 0.0015        | 13.4033 | 11500 | 0.1266          | 0.8374    | 0.8499 | 0.8436 | 0.9830   |
| 0.002         | 13.9860 | 12000 | 0.1236          | 0.8403    | 0.8530 | 0.8466 | 0.9832   |
| 0.0016        | 14.5688 | 12500 | 0.1313          | 0.8453    | 0.8409 | 0.8430 | 0.9833   |
| 0.0013        | 15.1515 | 13000 | 0.1362          | 0.8460    | 0.8482 | 0.8471 | 0.9835   |
| 0.0015        | 15.7343 | 13500 | 0.1246          | 0.8480    | 0.8511 | 0.8496 | 0.9840   |
| 0.0012        | 16.3170 | 14000 | 0.1335          | 0.8549    | 0.8423 | 0.8485 | 0.9837   |
| 0.0014        | 16.8998 | 14500 | 0.1265          | 0.8445    | 0.8433 | 0.8439 | 0.9833   |
| 0.0009        | 17.4825 | 15000 | 0.1450          | 0.8516    | 0.8357 | 0.8436 | 0.9832   |


### Framework versions

- Transformers 4.44.2
- Pytorch 2.1.1+cu121
- Datasets 2.14.5
- Tokenizers 0.19.1