Spaces:
Running
Running
| import streamlit as st | |
| import pandas as pd | |
| from streamlit_pandas_profiling import st_profile_report | |
| from pathlib import Path | |
| from transformers import MBartForConditionalGeneration, MBart50TokenizerFast | |
| st.set_page_config(page_title="Francesco Daimon Fernicola", page_icon=":milky_way:", layout="wide") | |
| with st.container(): | |
| st.subheader("Hello, and welcome to my official webpage! I am Daimon :alien:") | |
| st.title("PhD Candidate in Machine Translation / Translator / Mountain enthusiast") | |
| st.write("I am passionate about finding new ways to effectively use and understand Machine Translation and effectively evaluating its quality.") | |
| st.write(""" | |
| [Github](https://github.com/FrancescoFernicola) | |
| [Unibo](https://www.unibo.it/sitoweb/francesco.fernicola2) | |
| [LinkedIn](https://www.linkedin.com/in/francesco-fernicola-69a0771b7/?locale=en_US) | |
| [Twitter](https://twitter.com/FrancescoDaimon) | |
| """) | |
| st.title('Upload your data') | |
| st.subheader('Input TSV/CSV') | |
| uploaded_file = st.file_uploader("Choose a file") | |
| with st.spinner("Loading..."): | |
| if uploaded_file is not None: | |
| if uploaded_file.name.endswith('.tsv'): | |
| data = pd.read_csv(uploaded_file, sep="\t") | |
| else: | |
| data = pd.read_csv(uploaded_file) | |
| st.subheader("DataFrame") | |
| st.write(data) | |
| st.write(data.describe()) | |
| else: | |
| st.info("☝️ Upload a TSV/CSV file") | |
| st.subheader("MBART-50 Translator") | |
| source = "In the beginning the Universe was created. This has made a lot of people very angry and been widely regarded as a bad move." | |
| target = "" | |
| model = MBartForConditionalGeneration.from_pretrained("facebook/mbart-large-50-many-to-many-mmt") | |
| tokenizer = MBart50TokenizerFast.from_pretrained("facebook/mbart-large-50-many-to-many-mmt") | |
| def get_translation(src_code, trg_code, src): | |
| tokenizer.src_lang = src_code | |
| encoded = tokenizer(src, return_tensors="pt") | |
| generated_tokens = model.generate( | |
| **encoded, | |
| forced_bos_token_id=tokenizer.lang_code_to_id[trg_code] | |
| ) | |
| trg = tokenizer.batch_decode(generated_tokens, skip_special_tokens=True) | |
| return trg | |
| valid_languages = ['en_XX', 'fr_XX', 'de_DE', 'it_IT', 'es_XX'] | |
| with st.form("my_form"): | |
| left_c, right_c = st.columns(2) | |
| with left_c: | |
| src_lang = st.selectbox( | |
| 'Source language', | |
| ('en_XX', 'fr_XX', 'de_DE', 'it_IT', 'es_XX'), | |
| ) | |
| with right_c: | |
| trg_lang = st.selectbox( | |
| 'Target language', | |
| ('fr_XX', 'en_XX', 'de_DE', 'it_IT', 'es_XX') | |
| ) | |
| source = st.text_area("Source", value=source, height=130, placeholder="Enter the source text...") | |
| submitted = st.form_submit_button("Translate") | |
| if submitted: | |
| if len(source) > 0 and src_lang in valid_languages and trg_lang in valid_languages: | |
| with st.spinner("Translating..."): | |
| try: | |
| target = get_translation(src_lang, trg_lang, source)[0] | |
| st.subheader("Translation done!") | |
| target = st.text_area("Target", value=target, height=130) | |
| except: | |
| st.subheader("Translation failed :sad:") | |
| else: | |
| st.write("Please enter the source text, source language and target language.") | |
| # ---- CONTACT ---- | |
| with st.container(): | |
| st.write("---") | |
| st.header("Get in Touch With Me!") | |
| st.write("##") | |
| contact_form = """ | |
| <form action="https://formsubmit.co/[email protected]" method="POST"> | |
| <input type="hidden" name="_captcha" value="false"> | |
| <input type="text" name="name" placeholder="Your name" required> | |
| <input type="email" name="email" placeholder="Your email" required> | |
| <textarea name="message" placeholder="Your message here" required></textarea> | |
| <button type="submit">Send</button> | |
| </form> | |
| """ | |
| left_column, right_column = st.columns(2) | |
| with left_column: | |
| st.markdown(contact_form, unsafe_allow_html=True) | |
| with right_column: | |
| st.empty() |