Update README.md
Browse files
README.md
CHANGED
@@ -1,60 +1,39 @@
|
|
1 |
---
|
2 |
license: apache-2.0
|
3 |
base_model: cross-encoder/nli-deberta-v3-base
|
4 |
-
tags:
|
5 |
-
- generated_from_trainer
|
6 |
-
metrics:
|
7 |
-
- accuracy
|
8 |
model-index:
|
9 |
-
- name:
|
10 |
results: []
|
11 |
---
|
12 |
|
13 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
14 |
should probably proofread and complete it, then remove this comment. -->
|
15 |
|
16 |
-
#
|
17 |
|
18 |
-
This model is a fine-tuned version of [cross-encoder/nli-deberta-v3-base](https://huggingface.co/cross-encoder/nli-deberta-v3-base) on
|
19 |
-
It achieves the following results on the evaluation set:
|
20 |
-
- Loss: 0.8729
|
21 |
-
- Accuracy: 0.6835
|
22 |
|
23 |
## Model description
|
24 |
|
25 |
-
|
26 |
|
27 |
-
## Intended uses & limitations
|
28 |
-
|
29 |
-
More information needed
|
30 |
-
|
31 |
-
## Training and evaluation data
|
32 |
-
|
33 |
-
More information needed
|
34 |
|
35 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
36 |
|
37 |
-
|
38 |
-
|
39 |
-
The following hyperparameters were used during training:
|
40 |
-
- learning_rate: 5e-05
|
41 |
-
- train_batch_size: 4
|
42 |
-
- eval_batch_size: 8
|
43 |
-
- seed: 42
|
44 |
-
- gradient_accumulation_steps: 2
|
45 |
-
- total_train_batch_size: 8
|
46 |
-
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
47 |
-
- lr_scheduler_type: linear
|
48 |
-
- num_epochs: 3.0
|
49 |
-
|
50 |
-
### Training results
|
51 |
-
|
52 |
-
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
53 |
-
|:-------------:|:------:|:----:|:---------------:|:--------:|
|
54 |
-
| No log | 0.9980 | 245 | 1.2681 | 0.5688 |
|
55 |
-
| No log | 2.0 | 491 | 1.0121 | 0.6651 |
|
56 |
-
| 1.1214 | 2.9939 | 735 | 0.8729 | 0.6835 |
|
57 |
|
|
|
58 |
|
59 |
### Framework versions
|
60 |
|
|
|
1 |
---
|
2 |
license: apache-2.0
|
3 |
base_model: cross-encoder/nli-deberta-v3-base
|
|
|
|
|
|
|
|
|
4 |
model-index:
|
5 |
+
- name: Komainu
|
6 |
results: []
|
7 |
---
|
8 |
|
9 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
10 |
should probably proofread and complete it, then remove this comment. -->
|
11 |
|
12 |
+
# Komainu - Content Moderator
|
13 |
|
14 |
+
This model is a fine-tuned version of [cross-encoder/nli-deberta-v3-base](https://huggingface.co/cross-encoder/nli-deberta-v3-base) on [Daijobu.ai's content moderation dataset](https://huggingface.co/datasets/DaijobuAI/moderation_dataset).
|
|
|
|
|
|
|
15 |
|
16 |
## Model description
|
17 |
|
18 |
+
This model can identify 9 types of contents to be moderated.
|
19 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
20 |
|
21 |
+
| Category | Output | Definition |
|
22 |
+
| -------- | ----- | ---------- |
|
23 |
+
| sexual | `0` | Content meant to arouse sexual excitement, such as the description of sexual activity, or that promotes sexual services (excluding sex education and wellness). |
|
24 |
+
| hate | `1` | Content that expresses, incites, or promotes hate based on race, gender, ethnicity, religion, nationality, sexual orientation, disability status, or caste. |
|
25 |
+
| violence | `2` | Content that promotes or glorifies violence or celebrates the suffering or humiliation of others. |
|
26 |
+
| harassment | `3` | Content that may be used to torment or annoy individuals in real life, or make harassment more likely to occur. |
|
27 |
+
| self-harm | `4` | Content that promotes, encourages, or depicts acts of self-harm, such as suicide, cutting, and eating disorders. |
|
28 |
+
| sexual/minors | `5` | Sexual content that includes an individual who is under 18 years old. |
|
29 |
+
| hate/threatening | `6` | Hateful content that also includes violence or serious harm towards the targeted group. |
|
30 |
+
| violence/graphic | `7` | Violent content that depicts death, violence, or serious physical injury in extreme graphic detail. |
|
31 |
+
| fake news | `8` | Fake news
|
32 |
+
| OK | `9` | OK
|
33 |
|
34 |
+
## Intended uses & limitations
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
35 |
|
36 |
+
English language only. Not intended for commercial or production use.
|
37 |
|
38 |
### Framework versions
|
39 |
|