update model card README.md
Browse files
README.md
ADDED
@@ -0,0 +1,101 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: mit
|
3 |
+
tags:
|
4 |
+
- generated_from_trainer
|
5 |
+
datasets:
|
6 |
+
- tweets_hate_speech_detection
|
7 |
+
metrics:
|
8 |
+
- accuracy
|
9 |
+
- f1
|
10 |
+
model-index:
|
11 |
+
- name: FirstTry
|
12 |
+
results:
|
13 |
+
- task:
|
14 |
+
name: Text Classification
|
15 |
+
type: text-classification
|
16 |
+
dataset:
|
17 |
+
name: tweets_hate_speech_detection
|
18 |
+
type: tweets_hate_speech_detection
|
19 |
+
config: default
|
20 |
+
split: train
|
21 |
+
args: default
|
22 |
+
metrics:
|
23 |
+
- name: Accuracy
|
24 |
+
type: accuracy
|
25 |
+
value: 0.9759098967567004
|
26 |
+
- name: F1
|
27 |
+
type: f1
|
28 |
+
value: 0.8034042553191489
|
29 |
+
---
|
30 |
+
|
31 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
32 |
+
should probably proofread and complete it, then remove this comment. -->
|
33 |
+
|
34 |
+
# FirstTry
|
35 |
+
|
36 |
+
This model is a fine-tuned version of [roberta-base](https://huggingface.co/roberta-base) on the tweets_hate_speech_detection dataset.
|
37 |
+
It achieves the following results on the evaluation set:
|
38 |
+
- Loss: 0.0977
|
39 |
+
- Accuracy: 0.9759
|
40 |
+
- F1: 0.8034
|
41 |
+
|
42 |
+
## Model description
|
43 |
+
|
44 |
+
More information needed
|
45 |
+
|
46 |
+
## Intended uses & limitations
|
47 |
+
|
48 |
+
More information needed
|
49 |
+
|
50 |
+
## Training and evaluation data
|
51 |
+
|
52 |
+
More information needed
|
53 |
+
|
54 |
+
## Training procedure
|
55 |
+
|
56 |
+
### Training hyperparameters
|
57 |
+
|
58 |
+
The following hyperparameters were used during training:
|
59 |
+
- learning_rate: 2e-05
|
60 |
+
- train_batch_size: 16
|
61 |
+
- eval_batch_size: 16
|
62 |
+
- seed: 42
|
63 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
64 |
+
- lr_scheduler_type: linear
|
65 |
+
- num_epochs: 15
|
66 |
+
|
67 |
+
### Training results
|
68 |
+
|
69 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|
70 |
+
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
|
71 |
+
| No log | 0.04 | 50 | 0.2125 | 0.9337 | 0.0 |
|
72 |
+
| No log | 0.07 | 100 | 0.2210 | 0.9341 | 0.0125 |
|
73 |
+
| No log | 0.11 | 150 | 0.1832 | 0.9554 | 0.5103 |
|
74 |
+
| No log | 0.14 | 200 | 0.1539 | 0.9583 | 0.6377 |
|
75 |
+
| No log | 0.18 | 250 | 0.2435 | 0.9523 | 0.4434 |
|
76 |
+
| No log | 0.21 | 300 | 0.1818 | 0.9589 | 0.5736 |
|
77 |
+
| No log | 0.25 | 350 | 0.1138 | 0.9618 | 0.7136 |
|
78 |
+
| No log | 0.29 | 400 | 0.1045 | 0.9667 | 0.7243 |
|
79 |
+
| No log | 0.32 | 450 | 0.0958 | 0.9676 | 0.7330 |
|
80 |
+
| 0.1788 | 0.36 | 500 | 0.0935 | 0.9695 | 0.7306 |
|
81 |
+
| 0.1788 | 0.39 | 550 | 0.1289 | 0.9666 | 0.7178 |
|
82 |
+
| 0.1788 | 0.43 | 600 | 0.1039 | 0.9648 | 0.7507 |
|
83 |
+
| 0.1788 | 0.46 | 650 | 0.1234 | 0.9646 | 0.6435 |
|
84 |
+
| 0.1788 | 0.5 | 700 | 0.0984 | 0.9703 | 0.7725 |
|
85 |
+
| 0.1788 | 0.54 | 750 | 0.1364 | 0.9702 | 0.7185 |
|
86 |
+
| 0.1788 | 0.57 | 800 | 0.1004 | 0.9739 | 0.7792 |
|
87 |
+
| 0.1788 | 0.61 | 850 | 0.0998 | 0.9684 | 0.7616 |
|
88 |
+
| 0.1788 | 0.64 | 900 | 0.1068 | 0.9738 | 0.7857 |
|
89 |
+
| 0.1788 | 0.68 | 950 | 0.1206 | 0.9732 | 0.7644 |
|
90 |
+
| 0.1198 | 0.71 | 1000 | 0.0977 | 0.9759 | 0.8034 |
|
91 |
+
| 0.1198 | 0.75 | 1050 | 0.0864 | 0.9742 | 0.7916 |
|
92 |
+
| 0.1198 | 0.79 | 1100 | 0.1297 | 0.9727 | 0.7849 |
|
93 |
+
| 0.1198 | 0.82 | 1150 | 0.0969 | 0.9751 | 0.8026 |
|
94 |
+
|
95 |
+
|
96 |
+
### Framework versions
|
97 |
+
|
98 |
+
- Transformers 4.26.1
|
99 |
+
- Pytorch 2.0.1
|
100 |
+
- Datasets 2.10.1
|
101 |
+
- Tokenizers 0.13.3
|