File size: 3,979 Bytes
e84b5e6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
be33810
e84b5e6
 
 
 
 
 
be33810
e84b5e6
 
 
 
 
 
be33810
e84b5e6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
'''
        Created By Lewis Kamau Kimaru 
        Sema fastapi backend
        August 2023
'''

from fastapi import FastAPI, HTTPException, Request
from fastapi.middleware.cors import CORSMiddleware
from fastapi.responses import HTMLResponse
import gradio as gr
import ctranslate2
import sentencepiece as spm
import fasttext
import uvicorn
from pyngrok import ngrok
import nest_asyncio
import os

app = FastAPI()

# Set your ngrok authtoken
#ngrok.set_auth_token("2UAhCqf5zP0cCgJzeadNANkbIqx_7ZJvhkDSNWccqMX2hyxXP")
#ngrok.set_auth_token("2S6xeFEoSVFWr2egtDRcqgeUtSx_2juefHFkEW6nGbpRHS37W")
#ngrok.set_auth_token("2UAmdjHdAFV9x84TdyEknIfNhYk_4Ye8n4YK7ZhfCMob3yPBh")
#ngrok.set_auth_token("2UAqm26HuWiWvQjzK58xYufSGpy_6tStKSyLLyR9f7pcezh6R")
ngrok.set_auth_token("2UGQqzZoI3bx7SSk8H4wuFC3iaC_2WniWyNAsW5fd2rFyKVq1")


fasttext.FastText.eprint = lambda x: None

# Load the model and tokenizer ..... only once!
beam_size = 1  # change to a smaller value for faster inference
device = "cpu"  # or "cuda"

# Language Prediction model
print("\nimporting Language Prediction model")
lang_model_file = "modules/lid218e.bin"
lang_model_full_path = os.path.join(os.path.dirname(__file__), lang_model_file)
lang_model = fasttext.load_model(lang_model_full_path)


# Load the source SentencePiece model
print("\nimporting SentencePiece model")
sp_model_file = "modules/spm.model"
sp_model_full_path = os.path.join(os.path.dirname(__file__), sp_model_file)
sp = spm.SentencePieceProcessor()
sp.load(sp_model_full_path)

# Import The Translator model
print("\nimporting Translator model")
ct_model_file = "modules/sematrans-3.3B"
ct_model_full_path = os.path.join(os.path.dirname(__file__), ct_model_file)
translator = ctranslate2.Translator(ct_model_full_path, device)

print('\nDone importing models\n')

    
def translate_text(userinput: str, target_lang: str):
    source_sents = [userinput]
    source_sents = [sent.strip() for sent in source_sents]
    target_prefix = [[target_lang]] * len(source_sents)

    # Predict the source language
    predictions = lang_model.predict(source_sents[0], k=1)
    source_lang = predictions[0][0].replace('__label__', '')

    # Subword the source sentences
    source_sents_subworded = sp.encode(source_sents, out_type=str)
    source_sents_subworded = [[source_lang] + sent + ["</s>"] for sent in source_sents_subworded]

    # Translate the source sentences
    translations = translator.translate_batch(
        source_sents_subworded,
        batch_type="tokens",
        max_batch_size=2024,
        beam_size=beam_size,
        target_prefix=target_prefix,
    )
    translations = [translation[0]['tokens'] for translation in translations]

    # Desubword the target sentences
    translations_desubword = sp.decode(translations)
    translations_desubword = [sent[len(target_lang):] for sent in translations_desubword]

    # Return the source language and the translated text
    return source_lang, translations_desubword


@app.get("/")
def read_root():
    return {"message": "Welcome to the Sema Translation API! \nThis API was created by Lewsi Kamau Kimaru"}


@app.post("/translate/")
async def translate_endpoint(request: Request):
    data = await request.json()
    userinput = data.get("userinput")
    target_lang = data.get("target_lang")
    print(f"\n Target Language; {target_lang}, User Input: {userinput}\n")

    if not userinput or not target_lang:
        raise HTTPException(status_code=422, detail="Both 'userinput' and 'target_lang' are required.")

    source_lang, translated_text = translate_text(userinput, target_lang)
    print(f"\nsource_language: {source_lang}, Translated Text: {translated_text}\n\n")
    return {
        "source_language": source_lang,
        "translated_text": translated_text[0],
    }

ngrok_tunnel = ngrok.connect(7860)
public_url = ngrok_tunnel.public_url
print('\nPublic URL✅:', public_url)
nest_asyncio.apply()

print("\nAPI starting .......\n")
#uvicorn.run(app, port=7860)