File size: 4,774 Bytes
2530af9
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
98
99
100
101
102
103
104
105
106
107
108
109
110
---
base_model: allenai/scibert_scivocab_uncased
tags:
- generated_from_trainer
model-index:
- name: CRAFT_SciBERT_NER
  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. -->

# CRAFT_SciBERT_NER

This model is a fine-tuned version of [allenai/scibert_scivocab_uncased](https://huggingface.co/allenai/scibert_scivocab_uncased) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1143
- Seqeval classification report:               precision    recall  f1-score   support

       CHEBI       0.74      0.70      0.72       457
          CL       0.82      0.75      0.78      1099
         GGP       0.92      0.93      0.93      2232
          GO       0.78      0.84      0.81      2508
          SO       0.83      0.81      0.82      1365
       Taxon       0.99      0.99      0.99     87655

   micro avg       0.98      0.98      0.98     95316
   macro avg       0.85      0.84      0.84     95316
weighted avg       0.98      0.98      0.98     95316


## 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: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3

### Training results

| Training Loss | Epoch | Step | Validation Loss | Seqeval classification report                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                  |
|:-------------:|:-----:|:----:|:---------------:|:--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------:|
| No log        | 1.0   | 347  | 0.1140          |               precision    recall  f1-score   support

       CHEBI       0.66      0.69      0.67       457
          CL       0.83      0.69      0.75      1099
         GGP       0.89      0.93      0.91      2232
          GO       0.76      0.85      0.80      2508
          SO       0.79      0.73      0.76      1365
       Taxon       0.99      0.99      0.99     87655

   micro avg       0.97      0.97      0.97     95316
   macro avg       0.82      0.81      0.81     95316
weighted avg       0.97      0.97      0.97     95316
 |
| 0.1263        | 2.0   | 695  | 0.1126          |               precision    recall  f1-score   support

       CHEBI       0.73      0.69      0.71       457
          CL       0.85      0.72      0.78      1099
         GGP       0.91      0.93      0.92      2232
          GO       0.74      0.87      0.80      2508
          SO       0.82      0.80      0.81      1365
       Taxon       0.99      0.99      0.99     87655

   micro avg       0.97      0.97      0.97     95316
   macro avg       0.84      0.83      0.83     95316
weighted avg       0.97      0.97      0.97     95316
 |
| 0.0326        | 3.0   | 1041 | 0.1143          |               precision    recall  f1-score   support

       CHEBI       0.74      0.70      0.72       457
          CL       0.82      0.75      0.78      1099
         GGP       0.92      0.93      0.93      2232
          GO       0.78      0.84      0.81      2508
          SO       0.83      0.81      0.82      1365
       Taxon       0.99      0.99      0.99     87655

   micro avg       0.98      0.98      0.98     95316
   macro avg       0.85      0.84      0.84     95316
weighted avg       0.98      0.98      0.98     95316
 |


### Framework versions

- Transformers 4.35.2
- Pytorch 2.1.0+cu121
- Datasets 2.15.0
- Tokenizers 0.15.0