Model save
Browse files
README.md
ADDED
@@ -0,0 +1,85 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: mit
|
3 |
+
base_model: indolem/indobert-base-uncased
|
4 |
+
tags:
|
5 |
+
- generated_from_trainer
|
6 |
+
metrics:
|
7 |
+
- precision
|
8 |
+
- recall
|
9 |
+
- f1
|
10 |
+
- accuracy
|
11 |
+
model-index:
|
12 |
+
- name: nerugm-base-4
|
13 |
+
results: []
|
14 |
+
---
|
15 |
+
|
16 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
17 |
+
should probably proofread and complete it, then remove this comment. -->
|
18 |
+
|
19 |
+
# nerugm-base-4
|
20 |
+
|
21 |
+
This model is a fine-tuned version of [indolem/indobert-base-uncased](https://huggingface.co/indolem/indobert-base-uncased) on an unknown dataset.
|
22 |
+
It achieves the following results on the evaluation set:
|
23 |
+
- Loss: 0.2607
|
24 |
+
- Precision: 0.8198
|
25 |
+
- Recall: 0.8946
|
26 |
+
- F1: 0.8556
|
27 |
+
- Accuracy: 0.9651
|
28 |
+
|
29 |
+
## Model description
|
30 |
+
|
31 |
+
More information needed
|
32 |
+
|
33 |
+
## Intended uses & limitations
|
34 |
+
|
35 |
+
More information needed
|
36 |
+
|
37 |
+
## Training and evaluation data
|
38 |
+
|
39 |
+
More information needed
|
40 |
+
|
41 |
+
## Training procedure
|
42 |
+
|
43 |
+
### Training hyperparameters
|
44 |
+
|
45 |
+
The following hyperparameters were used during training:
|
46 |
+
- learning_rate: 5e-05
|
47 |
+
- train_batch_size: 16
|
48 |
+
- eval_batch_size: 64
|
49 |
+
- seed: 42
|
50 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
51 |
+
- lr_scheduler_type: linear
|
52 |
+
- num_epochs: 20.0
|
53 |
+
|
54 |
+
### Training results
|
55 |
+
|
56 |
+
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|
57 |
+
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
|
58 |
+
| 0.3182 | 1.0 | 106 | 0.1284 | 0.7463 | 0.8547 | 0.7968 | 0.9572 |
|
59 |
+
| 0.1137 | 2.0 | 212 | 0.1302 | 0.7230 | 0.8775 | 0.7928 | 0.9562 |
|
60 |
+
| 0.0683 | 3.0 | 318 | 0.1249 | 0.7833 | 0.8547 | 0.8174 | 0.9606 |
|
61 |
+
| 0.0454 | 4.0 | 424 | 0.1464 | 0.7711 | 0.8832 | 0.8234 | 0.9591 |
|
62 |
+
| 0.0325 | 5.0 | 530 | 0.1557 | 0.8010 | 0.8832 | 0.8401 | 0.9641 |
|
63 |
+
| 0.0211 | 6.0 | 636 | 0.2112 | 0.7915 | 0.8974 | 0.8411 | 0.9599 |
|
64 |
+
| 0.015 | 7.0 | 742 | 0.1944 | 0.7734 | 0.8946 | 0.8296 | 0.9606 |
|
65 |
+
| 0.0113 | 8.0 | 848 | 0.2151 | 0.8140 | 0.8974 | 0.8537 | 0.9665 |
|
66 |
+
| 0.0075 | 9.0 | 954 | 0.1996 | 0.8140 | 0.8974 | 0.8537 | 0.9685 |
|
67 |
+
| 0.0067 | 10.0 | 1060 | 0.2077 | 0.8470 | 0.8832 | 0.8647 | 0.9685 |
|
68 |
+
| 0.0039 | 11.0 | 1166 | 0.2609 | 0.7698 | 0.8860 | 0.8238 | 0.9579 |
|
69 |
+
| 0.0028 | 12.0 | 1272 | 0.2498 | 0.8263 | 0.8946 | 0.8591 | 0.9648 |
|
70 |
+
| 0.0035 | 13.0 | 1378 | 0.2407 | 0.8179 | 0.8832 | 0.8493 | 0.9643 |
|
71 |
+
| 0.003 | 14.0 | 1484 | 0.2475 | 0.7919 | 0.8889 | 0.8376 | 0.9631 |
|
72 |
+
| 0.0016 | 15.0 | 1590 | 0.2552 | 0.7975 | 0.8974 | 0.8445 | 0.9641 |
|
73 |
+
| 0.0016 | 16.0 | 1696 | 0.2463 | 0.8268 | 0.8974 | 0.8607 | 0.9665 |
|
74 |
+
| 0.0012 | 17.0 | 1802 | 0.2500 | 0.8324 | 0.8917 | 0.8611 | 0.9665 |
|
75 |
+
| 0.0009 | 18.0 | 1908 | 0.2629 | 0.8208 | 0.9003 | 0.8587 | 0.9653 |
|
76 |
+
| 0.0014 | 19.0 | 2014 | 0.2619 | 0.8182 | 0.8974 | 0.8560 | 0.9651 |
|
77 |
+
| 0.0006 | 20.0 | 2120 | 0.2607 | 0.8198 | 0.8946 | 0.8556 | 0.9651 |
|
78 |
+
|
79 |
+
|
80 |
+
### Framework versions
|
81 |
+
|
82 |
+
- Transformers 4.39.3
|
83 |
+
- Pytorch 2.3.0+cu121
|
84 |
+
- Datasets 2.19.1
|
85 |
+
- Tokenizers 0.15.2
|