add more languages

README.md

@@ -1,5 +1,5 @@
 ---
-title: Vietnamese - English Translation 
+title: Text Translation 
 emoji: 🐠
 colorFrom: yellow
 colorTo: gray
@@ -8,30 +8,27 @@ app_file: app.py
 pinned: false
 ---
 
-# Configuration
+## Info
 
-`title`: _string_  
-Display title for the Space
 
-`emoji`: _string_  
-Space emoji (emoji-only character allowed)
+Using facebook/m2m100_1.2B pre-trained model
 
-`colorFrom`: _string_  
-Color for Thumbnail gradient (red, yellow, green, blue, indigo, purple, pink, gray)
+facebook/m2m100_1.2B supports 100 languages.
 
-`colorTo`: _string_  
-Color for Thumbnail gradient (red, yellow, green, blue, indigo, purple, pink, gray)
+Here, this app uses/tests these languages only.
 
-`sdk`: _string_  
-Can be either `gradio` or `streamlit`
+```
+Chinese(zh)
+English(en)
+Hindi(hi)
+Japanese(ja)
+Sinhalese(si)
+Thai(th)
+Vietnamese(vi)
 
-`sdk_version` : _string_  
-Only applicable for `streamlit` SDK.  
-See [doc](https://hf.co/docs/hub/spaces) for more info on supported versions.
+```
 
-`app_file`: _string_  
-Path to your main application file (which contains either `gradio` or `streamlit` Python code).  
-Path is relative to the root of the repository.
 
-`pinned`: _boolean_  
-Whether the Space stays on top of your list.
+## Read more:
+
+https://huggingface.co/facebook/m2m100_1.2B

app.py

@@ -3,30 +3,72 @@
 
 import gradio as gr
 from transformers import M2M100ForConditionalGeneration, M2M100Tokenizer
+
 model = M2M100ForConditionalGeneration.from_pretrained("facebook/m2m100_1.2B")
+
 tokenizer = M2M100Tokenizer.from_pretrained("facebook/m2m100_1.2B")
 
-# from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
-# tokenizer = AutoTokenizer.from_pretrained("facebook/m2m100_1.2B")
-# model = AutoModelForSeq2SeqLM.from_pretrained("facebook/m2m100_1.2B")
+
+this_description = '''
+Using facebook/m2m100_1.2B pre-trained model. Language code:
+Chinese(zh)
+English(en)
+Hindi(hi)
+Japanese(ja)
+Sinhalese(si)
+Thai(th)
+Vietnamese(vi)
+'''
 
 
-def vi_en(Vietnamese_Text):
-    tokenizer.src_lang = "vi"
-    encoded_vi = tokenizer(Vietnamese_Text, return_tensors="pt")
+def m2m_translate(Input_Text, from_lang, to_lang):
+    tokenizer.src_lang = from_lang
+
+    encoded_from_lang = tokenizer(Input_Text, return_tensors="pt")
+
     generated_tokens = model.generate(
-        **encoded_vi, forced_bos_token_id=tokenizer.get_lang_id("en"))
+        **encoded_from_lang, forced_bos_token_id=tokenizer.get_lang_id(to_lang))
+
     res = tokenizer.batch_decode(generated_tokens, skip_special_tokens=True)
+
     return res[0]
     # if res:
-      # return '\n'.join(res)
+    # return '\n'.join(res)
 
 
 iface = gr.Interface(
-    fn=vi_en,
-    title="Vietnamese to English",
-    description="Using/testing facebook/m2m100_1.2B pre-trained model",
-    inputs=gr.inputs.Textbox(lines=5, placeholder="Enter text in Vietnamese"),
+    fn=m2m_translate,
+
+    title="M2M100 Translation",
+    description=this_description,
+
+    inputs=[
+        gr.inputs.Textbox(lines=5, placeholder="Enter text"),
+
+        gr.inputs.Radio(
+            choices=[
+                'zh',
+                'en',
+                'hi',
+                'ja',
+                'si',
+                'th',
+                'vi'],
+            default='vi',
+            label='From language'),
+
+        gr.inputs.Radio(
+            choices=[
+                'zh',
+                'en',
+                'hi',
+                'ja',
+                'si',
+                'th',
+                'vi'],
+            default='en',
+            label='To language'),
+    ],
     outputs="text")
 
 iface.launch()