init

app.py

@@ -0,0 +1,46 @@
+import time
+import streamlit as st
+from transformers import AutoTokenizer, AutoModelWithLMHead, pipeline
+from googletrans import Translator
+
+from enums import MODEL_NAME, MESSAGES, DESCRIPTION
+
+def iso2lang(iso):
+    return MESSAGES["iso"][iso]
+
+def load_tokenizer():
+    return AutoTokenizer.from_pretrained(MODEL_NAME)
+
+@st.cache(allow_output_mutation=True)
+def load_model():
+    return AutoModelWithLMHead.from_pretrained(MODEL_NAME)
+
+def load_pipe():
+    model = load_model()
+    tokenizer = load_tokenizer()
+    return pipeline("text-generation", model=model, tokenizer=tokenizer)
+
+st.write(DESCRIPTION)
+
+lang = st.radio('Хэл / Language', ('mn', 'en'), format_func=iso2lang)
+translator = Translator()
+
+with st.spinner(MESSAGES["loading_text"][lang]):
+    pipe = load_pipe()
+st.success(MESSAGES["success_model_load"][lang])
+
+text = st.text_area(
+    MESSAGES["input_description"][lang], MESSAGES["input_default"][lang])
+
+with st.spinner(MESSAGES["loading_text"][lang]):
+    if lang == "mn":
+        result = pipe(text)[0]['generated_text']
+        st.write(result)
+    elif lang == "en":
+        text = translator.translate(text, src='en', dest='mn').text
+        result_mn = pipe(text)[0]['generated_text']
+        result_en = translator.translate(result_mn, src='mn', dest='en').text
+        st.write(f"*Translated:* {result_en}")
+        st.write(f"> *Original:* {result_mn}")
+
+    st.warning('Translation is done by [`googletrans`](https://github.com/ssut/py-googletrans). Please check out the usage. https://github.com/ssut/py-googletrans#note-on-library-usage')

enums.py

@@ -0,0 +1,33 @@
+MODEL_NAME = "bayartsogt/mongolian-gpt2"
+MESSAGES = {
+    "success_model_load": {
+        "mn": "Моделийг амжилттай уншлаа!!",
+        "en": "Model Loaded!!"
+    },
+    "loading_text": {
+        "mn": "Уншиж байна...",
+        "en": "Loading..."
+    },
+    "input_description": {
+        "mn": "Эхлэл хэсэг:",
+        "en": "Prompt:"
+    },
+    "input_default": {
+        "mn": "Хүний амьдрал гэдэг",
+        "en": "Life is"
+    },
+    "iso": {
+        'mn': 'Монгол / Mongolian',
+        'en': 'Англи / English'
+    }
+}
+
+DESCRIPTION = """
+## Mongolian GPT2
+* **Goal:** To create GPT2 model that is able write text in Mongolian during [HuggingFace Community Week #2](https://discuss.huggingface.co/t/open-to-the-community-community-week-using-jax-flax-for-nlp-cv/7104).
+* **Discussion:** https://discuss.huggingface.co/t/pretrain-gpt-2-from-scratch-in-mongolian/7879
+* **Creator:** Bayartsogt Yadamsuren 
+[[✉️ email](mailto:[email protected])]
+[[🤗 huggingface](https://huggingface.co/bayartsogt)]
+[[🤖 github](https://github.com/bayartsogt-ya)]
+"""

requirements.txt

@@ -0,0 +1,3 @@
+transformers
+streamlit
+googletrans==3.1.0a