File size: 5,276 Bytes
36d59d5
413be94
 
 
 
 
36d59d5
1914537
413be94
1914537
413be94
 
 
 
 
 
 
 
 
 
 
 
 
36d59d5
413be94
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
36d59d5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
413be94
 
1914537
413be94
 
 
 
 
 
1914537
 
413be94
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1914537
 
413be94
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
36d59d5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
from fastapi import FastAPI, HTTPException, WebSocket
from fastapi.middleware.cors import CORSMiddleware
from pydantic import BaseModel
from typing import List, Dict
from app.bpe_tokenizer import BPETokenizer
import os
import asyncio
import logging

logging.basicConfig(level=logging.ERROR, format="%(asctime)s - %(levelname)s - %(message)s")
app = FastAPI(title="Hindi BPE Tokenizer API")

# Configure CORS
app.add_middleware(
    CORSMiddleware,
    allow_origins=["*"],
    allow_credentials=True,
    allow_methods=["*"],
    allow_headers=["*"],
)

# Initialize tokenizer and load model
tokenizer = BPETokenizer()
model_path = "bpe_model_latest.json"
if os.path.exists(model_path):
    tokenizer.load_model(model_path)
else:
    print(f"Warning: Model file {model_path} not found. Starting with empty model.")


class TokenizeRequest(BaseModel):
    text: str


class TokenStats(BaseModel):
    original_chars: int
    token_count: int
    compression_ratio: float
    unique_tokens: int


class TokenizeResponse(BaseModel):
    original_text: str
    original_tokens: List[str]
    bpe_tokens: List[str]
    stats: Dict
    token_details: List[Dict]
    merge_history: List[Dict]


# Add WebSocket manager
class ConnectionManager:
    def __init__(self):
        self.active_connections = []

    async def connect(self, websocket: WebSocket):
        await websocket.accept()
        self.active_connections.append(websocket)

    def disconnect(self, websocket: WebSocket):
        self.active_connections.remove(websocket)

    async def broadcast(self, message: dict):
        for connection in self.active_connections:
            await connection.send_json(message)


manager = ConnectionManager()


# Add WebSocket endpoint
@app.websocket("/ws")
async def websocket_endpoint(websocket: WebSocket):
    await manager.connect(websocket)
    try:
        while True:
            await websocket.receive_text()
    except:
        manager.disconnect(websocket)


@app.post("/tokenize")
async def tokenize_text(request: TokenizeRequest):
    # try:
        result = tokenizer.tokenize_with_details(request.text)
        # Ensure all required fields are present
        return {
            "original_text": result["original_text"],
            "original_tokens": result["original_tokens"],
            "bpe_tokens": result["bpe_tokens"],
            "original_encoded_tokens": result["original_encoded_tokens"],
            "token_numbers": result["token_numbers"],
            "stats": {
                "original_chars": result["stats"]["original_chars"],
                "token_count": len(result["bpe_tokens"]),
                "compression_ratio": result["stats"]["compression_ratio"],
                "unique_tokens": result["stats"]["unique_tokens"],
            },
            "token_details": [
                {
                    "token": token,
                    "type": tokenizer._get_token_type(token),
                    "length": len(token),
                }
                for token in result["bpe_tokens"]
            ],
            "merge_history": tokenizer.merge_history[-10:]
            if tokenizer.merge_history
            else [],
        }
    # except Exception as e:
    #     raise HTTPException(status_code=500, detail=str(e))


@app.get("/vocabulary-stats")
async def get_vocab_stats():
    return {
        "vocab_size": len(tokenizer.vocab),
        "most_frequent_tokens": tokenizer.token_usage.most_common(20),
        "most_frequent_pairs": dict(
            sorted(
                tokenizer.pair_frequencies.items(), key=lambda x: x[1], reverse=True
            )[:20]
        ),
    }


@app.get("/training-progress")
async def get_training_progress():
    if hasattr(tokenizer, "training_progress"):
        return tokenizer.training_progress
    return {"message": "No training progress available"}


@app.get("/training-stats")
async def get_training_stats():
    """Get detailed training statistics"""
    if hasattr(tokenizer, "vocab_growth"):
        return {
            "vocab_growth": tokenizer.vocab_growth,
            "base_vocab_stats": tokenizer.base_vocab_stats,
            "current_stats": {
                "total_vocab": len(tokenizer.vocab),
                "learned_vocab": len(tokenizer.learned_vocab),
                "base_vocab": len(tokenizer.BASE_VOCAB),
            },
        }
    return {"message": "No training statistics available"}


@app.post("/resume-training")
async def resume_training(checkpoint_file: str):
    """Resume training from checkpoint"""
    try:
        tokenizer.learn_bpe(text, resume_from=checkpoint_file)
        return {"message": "Training resumed successfully"}
    except Exception as e:
        raise HTTPException(status_code=500, detail=str(e))


@app.post("/start-training")
async def start_training(request: dict):
    try:
        text = load_sample_data(
            "data/hindi_wiki_corpus.txt",
            max_sentences=request.get("max_sentences", 10000),
        )
        tokenizer = await train_and_save_bpe(
            text, vocab_size=request.get("vocab_size", 10000), manager=manager
        )
        return {"message": "Training completed successfully"}
    except Exception as e:
        raise HTTPException(status_code=500, detail=str(e))