neelimapreeti297
commited on
Commit
•
e3a8c8f
1
Parent(s):
1313172
Update README.md
Browse files
README.md
CHANGED
@@ -20,6 +20,10 @@ This model translates german language to english language. It used Sequence to S
|
|
20 |
- **License:** MIT
|
21 |
- **Contact:** monjusha.2017@juniv.edu
|
22 |
|
|
|
|
|
|
|
|
|
23 |
# Data Processing
|
24 |
|
25 |
Defining source and target languages and then Tokenization. Tokenizers for German and English are initialized using spaCy (spacy library).
|
@@ -99,13 +103,13 @@ These utility functions and transformations handle the preprocessing of text dat
|
|
99 |
|
100 |
Then the model is trained with Seq2SeqTransformer and evaluated with function evaluate(model).
|
101 |
|
102 |
-
|
103 |
greedy_decode() - this function takes
|
104 |
-
|
105 |
-
|
106 |
-
|
107 |
-
|
108 |
-
|
109 |
|
110 |
as parameter and returns the generated target sequence tensor ys, which contains the complete translation.
|
111 |
|
@@ -129,7 +133,7 @@ def translate(src_sentence: str):
|
|
129 |
This function first loads the saved model. Then it tokenizes and implements greedy_decode for getting the translated output. Then returns the
|
130 |
output.
|
131 |
|
132 |
-
|
133 |
|
134 |
For creating interface gradio and torch as well as Seq2SeqTransformer, translate and greedy_decode function from the germantoenglish.py file was loaded.
|
135 |
```bash
|
@@ -153,8 +157,11 @@ if __name__ == "__main__":
|
|
153 |
iface.launch(share=True)
|
154 |
|
155 |
```
|
|
|
|
|
|
|
156 |
|
157 |
-
|
158 |
```bash
|
159 |
|---Readme.md
|
160 |
|
|
@@ -170,7 +177,7 @@ iface.launch(share=True)
|
|
170 |
|
171 |
```
|
172 |
|
173 |
-
|
174 |
|
175 |
```bash
|
176 |
|
@@ -184,11 +191,11 @@ python app.py
|
|
184 |
```
|
185 |
|
186 |
|
187 |
-
|
188 |
This project is licensed under the MIT License.
|
189 |
|
190 |
-
|
191 |
Neelima Monjusha Preeti - monjusha.stu2017@juniv.edu
|
192 |
|
193 |
-
link: https://huggingface.co/spaces/neelimapreeti297/GermanToEnglish
|
194 |
|
|
|
20 |
- **License:** MIT
|
21 |
- **Contact:** monjusha.2017@juniv.edu
|
22 |
|
23 |
+
# Task Description
|
24 |
+
This app translates German to English. First the language is tokenized, passed through encoder, decoder and trained with Seq2SeqTransformer.
|
25 |
+
Then as output the language is english.
|
26 |
+
|
27 |
# Data Processing
|
28 |
|
29 |
Defining source and target languages and then Tokenization. Tokenizers for German and English are initialized using spaCy (spacy library).
|
|
|
103 |
|
104 |
Then the model is trained with Seq2SeqTransformer and evaluated with function evaluate(model).
|
105 |
|
106 |
+
# Result Analysis
|
107 |
greedy_decode() - this function takes
|
108 |
+
### model: The sequence-to-sequence transformer model.
|
109 |
+
### src: The source sequence tensor.
|
110 |
+
### src_mask: The mask for the source sequence.
|
111 |
+
### max_len: The maximum length of the output sequence.
|
112 |
+
### start_symbol: The index of the start symbol in the target vocabulary
|
113 |
|
114 |
as parameter and returns the generated target sequence tensor ys, which contains the complete translation.
|
115 |
|
|
|
133 |
This function first loads the saved model. Then it tokenizes and implements greedy_decode for getting the translated output. Then returns the
|
134 |
output.
|
135 |
|
136 |
+
# Hugging Face Interface:
|
137 |
|
138 |
For creating interface gradio and torch as well as Seq2SeqTransformer, translate and greedy_decode function from the germantoenglish.py file was loaded.
|
139 |
```bash
|
|
|
157 |
iface.launch(share=True)
|
158 |
|
159 |
```
|
160 |
+
The app interface looks like this:
|
161 |
+
|
162 |
+
![image/png](https://cdn-uploads.huggingface.co/production/uploads/65b2665fee3f66b2b0f7b765/J_Q4eqXiN7cNuhOM3NbjR.png)
|
163 |
|
164 |
+
# Project Structure
|
165 |
```bash
|
166 |
|---Readme.md
|
167 |
|
|
|
|
177 |
|
178 |
```
|
179 |
|
180 |
+
# How to Run
|
181 |
|
182 |
```bash
|
183 |
|
|
|
191 |
```
|
192 |
|
193 |
|
194 |
+
# License
|
195 |
This project is licensed under the MIT License.
|
196 |
|
197 |
+
# Contributor
|
198 |
Neelima Monjusha Preeti - monjusha.stu2017@juniv.edu
|
199 |
|
200 |
+
App link: https://huggingface.co/spaces/neelimapreeti297/GermanToEnglish
|
201 |
|