update model card README.md
Browse files
README.md
ADDED
@@ -0,0 +1,95 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
tags:
|
3 |
+
- generated_from_trainer
|
4 |
+
datasets:
|
5 |
+
- nsmc
|
6 |
+
metrics:
|
7 |
+
- accuracy
|
8 |
+
- f1
|
9 |
+
- recall
|
10 |
+
- precision
|
11 |
+
model-index:
|
12 |
+
- name: kcbert-base-finetuned-nsmc
|
13 |
+
results:
|
14 |
+
- task:
|
15 |
+
name: Text Classification
|
16 |
+
type: text-classification
|
17 |
+
dataset:
|
18 |
+
name: nsmc
|
19 |
+
type: nsmc
|
20 |
+
args: default
|
21 |
+
metrics:
|
22 |
+
- name: Accuracy
|
23 |
+
type: accuracy
|
24 |
+
value: 0.90198
|
25 |
+
- name: F1
|
26 |
+
type: f1
|
27 |
+
value: 0.9033161705233671
|
28 |
+
- name: Recall
|
29 |
+
type: recall
|
30 |
+
value: 0.9095062169785088
|
31 |
+
- name: Precision
|
32 |
+
type: precision
|
33 |
+
value: 0.8972098126812446
|
34 |
+
---
|
35 |
+
|
36 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
37 |
+
should probably proofread and complete it, then remove this comment. -->
|
38 |
+
|
39 |
+
# kcbert-base-finetuned-nsmc
|
40 |
+
|
41 |
+
This model is a fine-tuned version of [beomi/kcbert-base](https://huggingface.co/beomi/kcbert-base) on the nsmc dataset.
|
42 |
+
It achieves the following results on the evaluation set:
|
43 |
+
- Loss: 0.4197
|
44 |
+
- Accuracy: 0.9020
|
45 |
+
- F1: 0.9033
|
46 |
+
- Recall: 0.9095
|
47 |
+
- Precision: 0.8972
|
48 |
+
|
49 |
+
## Model description
|
50 |
+
|
51 |
+
More information needed
|
52 |
+
|
53 |
+
## Intended uses & limitations
|
54 |
+
|
55 |
+
More information needed
|
56 |
+
|
57 |
+
## Training and evaluation data
|
58 |
+
|
59 |
+
More information needed
|
60 |
+
|
61 |
+
## Training procedure
|
62 |
+
|
63 |
+
### Training hyperparameters
|
64 |
+
|
65 |
+
The following hyperparameters were used during training:
|
66 |
+
- learning_rate: 2e-05
|
67 |
+
- train_batch_size: 16
|
68 |
+
- eval_batch_size: 16
|
69 |
+
- seed: 42
|
70 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
71 |
+
- lr_scheduler_type: linear
|
72 |
+
- num_epochs: 3
|
73 |
+
- mixed_precision_training: Native AMP
|
74 |
+
|
75 |
+
### Training results
|
76 |
+
|
77 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Recall | Precision |
|
78 |
+
|:-------------:|:-----:|:-----:|:---------------:|:--------:|:------:|:------:|:---------:|
|
79 |
+
| 0.3028 | 0.32 | 3000 | 0.2994 | 0.8769 | 0.8732 | 0.8422 | 0.9066 |
|
80 |
+
| 0.2833 | 0.64 | 6000 | 0.2766 | 0.8880 | 0.8844 | 0.8512 | 0.9203 |
|
81 |
+
| 0.2719 | 0.96 | 9000 | 0.2527 | 0.8980 | 0.8981 | 0.8933 | 0.9030 |
|
82 |
+
| 0.1938 | 1.28 | 12000 | 0.2934 | 0.8969 | 0.8965 | 0.8869 | 0.9062 |
|
83 |
+
| 0.1907 | 1.6 | 15000 | 0.3141 | 0.8992 | 0.8999 | 0.9003 | 0.8996 |
|
84 |
+
| 0.1824 | 1.92 | 18000 | 0.3537 | 0.8986 | 0.8964 | 0.8711 | 0.9232 |
|
85 |
+
| 0.1261 | 2.24 | 21000 | 0.4197 | 0.9020 | 0.9033 | 0.9095 | 0.8972 |
|
86 |
+
| 0.1237 | 2.56 | 24000 | 0.4170 | 0.8995 | 0.9017 | 0.9156 | 0.8882 |
|
87 |
+
| 0.1182 | 2.88 | 27000 | 0.4165 | 0.9020 | 0.9036 | 0.9130 | 0.8945 |
|
88 |
+
|
89 |
+
|
90 |
+
### Framework versions
|
91 |
+
|
92 |
+
- Transformers 4.11.3
|
93 |
+
- Pytorch 1.9.1
|
94 |
+
- Datasets 1.14.0
|
95 |
+
- Tokenizers 0.10.3
|