DevilishDaoSaint
commited on
Commit
•
b374edb
1
Parent(s):
fa22421
Update README.md
Browse files
README.md
CHANGED
@@ -1,121 +1,128 @@
|
|
1 |
-
---
|
2 |
-
license: apache-2.0
|
3 |
-
tags:
|
4 |
-
- light-novel
|
5 |
-
- translation
|
6 |
-
- Chinese
|
7 |
-
- English
|
8 |
-
- named-entity-recognition
|
9 |
-
- term-extraction
|
10 |
-
- unsloth
|
11 |
-
- trl
|
12 |
-
- sft
|
13 |
-
model_type: translation
|
14 |
-
language_bcp47:
|
15 |
-
- zh
|
16 |
-
- en
|
17 |
-
finetuned_from: unsloth/
|
18 |
-
description: 'Mythical Translator is a fine-tuned model specifically designed to translate
|
19 |
-
Chinese light novels into natural-sounding English suitable for the genre. It also
|
20 |
-
extracts terms and proper nouns for consistency in future translations.
|
21 |
-
|
22 |
-
'
|
23 |
-
---
|
24 |
-
|
25 |
-
# Model Card for Mythical Translator
|
26 |
-
|
27 |
-
## Model Details
|
28 |
-
|
29 |
-
### Model Description
|
30 |
-
|
31 |
-
Mythical Translator is a fine-tuned model specifically designed to translate Chinese light novels into natural-sounding English suitable for the genre. It also extracts terms and proper nouns for consistency in future translations. The model uses the unsloth fine-tuning technique and is trained on a dataset of high-quality, strictly translated Chinese light novels.
|
32 |
-
|
33 |
-
- **Developed by:** Arben Apura
|
34 |
-
- **Model type:** Translation Model
|
35 |
-
- **Language(s) (NLP):** Chinese to English
|
36 |
-
- **License:** Apache-2.0
|
37 |
-
- **Finetuned from model:** unsloth/
|
38 |
-
|
39 |
-
## Uses
|
40 |
-
|
41 |
-
### Direct Use
|
42 |
-
|
43 |
-
Translate a passage from a Chinese light novel into natural-sounding English, using the provided context as a dictionary. List the proper nouns and terminologies with their Chinese characters and English translations as they appear in the passage.
|
44 |
-
|
45 |
-
### Out-of-Scope Use
|
46 |
-
|
47 |
-
The model may not perform well with texts outside the light novel genre or with languages other than Chinese. It should not be used for tasks other than translating light novels or extracting terms.
|
48 |
-
|
49 |
-
## Bias, Risks, and Limitations
|
50 |
-
|
51 |
-
### Recommendations
|
52 |
-
|
53 |
-
Users should be aware of potential biases in translation related to genre-specific language and cultural nuances. Regular updates and reviews are recommended to maintain high translation quality.
|
54 |
-
|
55 |
-
## How to Get Started with the Model
|
56 |
-
|
57 |
-
Use the following template to start translating:
|
58 |
-
|
59 |
-
```markdown
|
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 |
-
|
111 |
-
|
112 |
-
|
113 |
-
|
114 |
-
|
115 |
-
|
116 |
-
|
117 |
-
|
118 |
-
|
119 |
-
|
120 |
-
|
121 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
tags:
|
4 |
+
- light-novel
|
5 |
+
- translation
|
6 |
+
- Chinese
|
7 |
+
- English
|
8 |
+
- named-entity-recognition
|
9 |
+
- term-extraction
|
10 |
+
- unsloth
|
11 |
+
- trl
|
12 |
+
- sft
|
13 |
+
model_type: translation
|
14 |
+
language_bcp47:
|
15 |
+
- zh
|
16 |
+
- en
|
17 |
+
finetuned_from: unsloth/OpenHermes-2.5-Mistral-7B-bnb-4bit
|
18 |
+
description: 'Mythical Translator is a fine-tuned model specifically designed to translate
|
19 |
+
Chinese light novels into natural-sounding English suitable for the genre. It also
|
20 |
+
extracts terms and proper nouns for consistency in future translations.
|
21 |
+
|
22 |
+
'
|
23 |
+
---
|
24 |
+
|
25 |
+
# Model Card for Mythical Translator
|
26 |
+
|
27 |
+
## Model Details
|
28 |
+
|
29 |
+
### Model Description
|
30 |
+
|
31 |
+
Mythical Translator is a fine-tuned model specifically designed to translate Chinese light novels into natural-sounding English suitable for the genre. It also extracts terms and proper nouns for consistency in future translations. The model uses the unsloth fine-tuning technique and is trained on a dataset of high-quality, strictly translated Chinese light novels.
|
32 |
+
|
33 |
+
- **Developed by:** Arben Apura
|
34 |
+
- **Model type:** Translation Model
|
35 |
+
- **Language(s) (NLP):** Chinese to English
|
36 |
+
- **License:** Apache-2.0
|
37 |
+
- **Finetuned from model:** unsloth/OpenHermes-2.5-Mistral-7B-bnb-4bit
|
38 |
+
|
39 |
+
## Uses
|
40 |
+
|
41 |
+
### Direct Use
|
42 |
+
|
43 |
+
Translate a passage from a Chinese light novel into natural-sounding English, using the provided context as a dictionary. List the proper nouns and terminologies with their Chinese characters and English translations as they appear in the passage.
|
44 |
+
|
45 |
+
### Out-of-Scope Use
|
46 |
+
|
47 |
+
The model may not perform well with texts outside the light novel genre or with languages other than Chinese. It should not be used for tasks other than translating light novels or extracting terms.
|
48 |
+
|
49 |
+
## Bias, Risks, and Limitations
|
50 |
+
|
51 |
+
### Recommendations
|
52 |
+
|
53 |
+
Users should be aware of potential biases in translation related to genre-specific language and cultural nuances. Regular updates and reviews are recommended to maintain high translation quality.
|
54 |
+
|
55 |
+
## How to Get Started with the Model
|
56 |
+
|
57 |
+
Use the following template to start translating:
|
58 |
+
|
59 |
+
```markdown
|
60 |
+
<|im_start|>system
|
61 |
+
Translate a passage from a Chinese light novel into natural-sounding English, using the provided context as a dictionary. List the proper nouns and terminologies with their Chinese characters and English translations as they appear in the passage.<|im_end|>
|
62 |
+
<|im_start|>user
|
63 |
+
<context>
|
64 |
+
...
|
65 |
+
</context>
|
66 |
+
|
67 |
+
<passage>
|
68 |
+
...
|
69 |
+
</passage>
|
70 |
+
|
71 |
+
<translation>
|
72 |
+
<|im_end|>
|
73 |
+
<|im_start|>assistant
|
74 |
+
|
75 |
+
```
|
76 |
+
|
77 |
+
## Sample Text Generation Output
|
78 |
+
|
79 |
+
```markdown
|
80 |
+
<|im_start|>system
|
81 |
+
Translate a passage from a Chinese light novel into natural-sounding English, using the provided context as a dictionary. List the proper nouns and terminologies with their Chinese characters and English translations as they appear in the passage.<|im_end|>
|
82 |
+
<|im_start|>user
|
83 |
+
<context>
|
84 |
+
剑 = sword
|
85 |
+
兵器 = weapon
|
86 |
+
师妹 = junior sister
|
87 |
+
孟师兄 = Senior Brother Meng
|
88 |
+
</context>
|
89 |
+
|
90 |
+
<passage>
|
91 |
+
孟师兄一握,赞许道,“不错,工艺精湛,比起一般的下品兵器要好上很多,只不过这名字,倒是有些……依我看,此剑灵光流动,表面灵光如水,不如便叫灵动剑更为合适,师妹你看如何?”
|
92 |
+
</passage>
|
93 |
+
|
94 |
+
<translation>
|
95 |
+
<|im_end|>
|
96 |
+
<|im_start|>assistant
|
97 |
+
Senior Brother Meng grasped the sword and, with approval, said, “Not bad, the craftsmanship is exquisite. It’s much better than an ordinary inferior-grade weapon. However, the name is a bit… In my opinion, since the sword's spiritual light flows and its surface spiritual light is like water, it would be more appropriate to call it the Spirit-Flow Sword. What do you think, junior sister?”
|
98 |
+
</translation>
|
99 |
+
|
100 |
+
<terms>
|
101 |
+
孟师兄 = Senior Brother Meng
|
102 |
+
剑 = sword
|
103 |
+
下品 = inferior-grade
|
104 |
+
兵器 = weapon
|
105 |
+
灵光 = spiritual light
|
106 |
+
灵动剑 = Spirit-Flow Sword
|
107 |
+
</terms><|im_end|>
|
108 |
+
```
|
109 |
+
|
110 |
+
## Training Details
|
111 |
+
|
112 |
+
### Training Data
|
113 |
+
|
114 |
+
The training data consists of multiple Chinese light novels, selected for their high-quality translations to ensure accuracy and fluency.
|
115 |
+
|
116 |
+
## Evaluation
|
117 |
+
|
118 |
+
#### Testing Data
|
119 |
+
|
120 |
+
The model's performance is evaluated using a dataset of light novel passages translated into natural-sounding English.
|
121 |
+
|
122 |
+
#### Metrics
|
123 |
+
|
124 |
+
Evaluation metrics include translation accuracy, fluency, and consistency of term extraction.
|
125 |
+
|
126 |
+
## More Information
|
127 |
+
|
128 |
+
For additional details, please contact Arben Apura at arbenapura.official@gmail.com.
|