Spaces:
Runtime error
Runtime error
File size: 5,681 Bytes
534e8dd 31e459d 534e8dd |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 |
import streamlit as st
import os
import io
from transformers import M2M100Tokenizer, M2M100ForConditionalGeneration
from transformers import AutoTokenizer, AutoModelForSequenceClassification
from languages import LANGUANGE_MAP
import time
import json
from typing import List
import torch
import random
import logging
if torch.cuda.is_available():
device = torch.device("cuda:0")
else:
device = torch.device("cpu")
logging.warning("GPU not found, using CPU, translation will be very slow.")
st.cache(suppress_st_warning=True, allow_output_mutation=True)
st.set_page_config(page_title="M2M100 Translator")
lang_id = {
"Afrikaans": "af",
"Amharic": "am",
"Arabic": "ar",
"Asturian": "ast",
"Azerbaijani": "az",
"Bashkir": "ba",
"Belarusian": "be",
"Bulgarian": "bg",
"Bengali": "bn",
"Breton": "br",
"Bosnian": "bs",
"Catalan": "ca",
"Cebuano": "ceb",
"Czech": "cs",
"Welsh": "cy",
"Danish": "da",
"German": "de",
"Greeek": "el",
"English": "en",
"Spanish": "es",
"Estonian": "et",
"Persian": "fa",
"Fulah": "ff",
"Finnish": "fi",
"French": "fr",
"Western Frisian": "fy",
"Irish": "ga",
"Gaelic": "gd",
"Galician": "gl",
"Gujarati": "gu",
"Hausa": "ha",
"Hebrew": "he",
"Hindi": "hi",
"Croatian": "hr",
"Haitian": "ht",
"Hungarian": "hu",
"Armenian": "hy",
"Indonesian": "id",
"Igbo": "ig",
"Iloko": "ilo",
"Icelandic": "is",
"Italian": "it",
"Japanese": "ja",
"Javanese": "jv",
"Georgian": "ka",
"Kazakh": "kk",
"Central Khmer": "km",
"Kannada": "kn",
"Korean": "ko",
"Luxembourgish": "lb",
"Ganda": "lg",
"Lingala": "ln",
"Lao": "lo",
"Lithuanian": "lt",
"Latvian": "lv",
"Malagasy": "mg",
"Macedonian": "mk",
"Malayalam": "ml",
"Mongolian": "mn",
"Marathi": "mr",
"Malay": "ms",
"Burmese": "my",
"Nepali": "ne",
"Dutch": "nl",
"Norwegian": "no",
"Northern Sotho": "ns",
"Occitan": "oc",
"Oriya": "or",
"Panjabi": "pa",
"Polish": "pl",
"Pushto": "ps",
"Portuguese": "pt",
"Romanian": "ro",
"Russian": "ru",
"Sindhi": "sd",
"Sinhala": "si",
"Slovak": "sk",
"Slovenian": "sl",
"Somali": "so",
"Albanian": "sq",
"Serbian": "sr",
"Swati": "ss",
"Sundanese": "su",
"Swedish": "sv",
"Swahili": "sw",
"Tamil": "ta",
"Thai": "th",
"Tagalog": "tl",
"Tswana": "tn",
"Turkish": "tr",
"Ukrainian": "uk",
"Urdu": "ur",
"Uzbek": "uz",
"Vietnamese": "vi",
"Wolof": "wo",
"Xhosa": "xh",
"Yiddish": "yi",
"Yoruba": "yo",
"Chinese": "zh",
"Zulu": "zu",
}
@st.cache(suppress_st_warning=True, allow_output_mutation=True)
def load_model(
pretrained_model: str = "facebook/m2m100_418M",
cache_dir: str = "models/",
):
tokenizer = M2M100Tokenizer.from_pretrained(pretrained_model, cache_dir=cache_dir)
model = M2M100ForConditionalGeneration.from_pretrained(
pretrained_model, cache_dir=cache_dir
).to(device)
model.eval()
return tokenizer, model
@st.cache(suppress_st_warning=True, allow_output_mutation=True)
def load_detection_model(
pretrained_model: str = "ivanlau/language-detection-fine-tuned-on-xlm-roberta-base",
cache_dir: str = "models/",
):
tokenizer = AutoTokenizer.from_pretrained(pretrained_model, cache_dir=cache_dir)
model = AutoModelForSequenceClassification.from_pretrained(pretrained_model, cache_dir=cache_dir).to(device)
model.eval()
return tokenizer, model
st.title("M2M100 Translator")
st.write("M2M100 is a multilingual encoder-decoder (seq-to-seq) model trained for Many-to-Many multilingual translation. It was introduced in this paper https://arxiv.org/abs/2010.11125 and first released in https://github.com/pytorch/fairseq/tree/master/examples/m2m_100 repository. The model that can directly translate between the 9,900 directions of 100 languages.\n")
st.write(" This demo uses the facebook/m2m100_1.2B model. For local inference see https://github.com/ikergarcia1996/Easy-Translate")
user_input: str = st.text_area(
"Input text",
height=200,
max_chars=5120,
)
target_lang = st.selectbox(label="Target language", options=list(lang_id.keys()))
if st.button("Run"):
time_start = time.time()
tokenizer, model = load_model()
de_tokenizer, de_model = load_detection_model()
with torch.no_grad():
tokenized_sentence = de_tokenizer(user_input, return_tensors='pt')
output = de_model(**tokenized_sentence)
de_predictions = torch.nn.functional.softmax(output.logits, dim=-1)
_, preds = torch.max(de_predictions, dim=-1)
lang_type = LANGUANGE_MAP[preds.item()]
if lang_type not in lang_id:
time_end = time.time()
st.success('Unsupported Language')
st.write(f"Computation time: {round((time_end-time_start),3)} segs")
else:
src_lang = lang_id[lang_type]
trg_lang = lang_id[target_lang]
tokenizer.src_lang = src_lang
encoded_input = tokenizer(user_input, return_tensors="pt").to(device)
generated_tokens = model.generate(
**encoded_input, forced_bos_token_id=tokenizer.get_lang_id(trg_lang)
)
translated_text = tokenizer.batch_decode(
generated_tokens, skip_special_tokens=True
)[0]
time_end = time.time()
st.success(translated_text)
st.write(f"Computation time: {round((time_end-time_start),3)} segs")
|