File size: 1,317 Bytes
f810166
2b6b06d
 
 
 
 
f810166
 
 
2b6b06d
 
 
 
 
 
 
 
 
 
 
b24a61c
 
 
 
 
 
2b6b06d
b24a61c
 
 
 
 
 
 
 
 
 
2b6b06d
 
 
 
b24a61c
 
 
 
 
2b6b06d
b24a61c
2b6b06d
b24a61c
 
 
 
 
 
 
2b6b06d
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
from fastapi import FastAPI
from pydantic import BaseModel
import torch
from transformers import AutoTokenizer, T5ForConditionalGeneration

MODEL_NAME = "google/byt5-small"

app = FastAPI()

print("Loading model...")

tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
model = T5ForConditionalGeneration.from_pretrained(MODEL_NAME)
model.eval()

print("Model loaded.")

class TextRequest(BaseModel):
    text: str

def text_to_ipa(text: str) -> str:
    # Few-shot examples for better IPA predictions
    prompt = f"""
You are a Scottish Gaelic teacher.
Convert Scottish Gaelic text into the International Phonetic Alphabet (IPA).
Only return the IPA transcription.

Examples:
Text: halò
IPA: /haˈloː/

Text: uisge
IPA: /ˈɯʃkʲə/

Text: {text}
IPA:
"""

    inputs = tokenizer(prompt, return_tensors="pt", truncation=True, max_length=512)

    with torch.no_grad():
        outputs = model.generate(
            **inputs,
            max_new_tokens=64,
            do_sample=False  # deterministic output
        )

    # Decode and return only the IPA portion
    result = tokenizer.decode(outputs[0], skip_special_tokens=True)
    return result.split("IPA:")[-1].strip()


@app.post("/predict")
def predict(request: TextRequest):
    ipa_result = text_to_ipa(request.text)
    return {"ipa": ipa_result}