File size: 5,812 Bytes
7505319
 
 
 
 
 
55acd9b
7505319
 
 
 
 
 
 
 
8c20f44
7505319
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
8c20f44
 
7505319
 
 
 
 
 
 
 
 
 
 
 
 
 
55acd9b
 
 
 
 
 
 
 
 
 
 
 
 
7505319
 
 
55acd9b
7505319
 
 
 
 
55acd9b
 
 
 
 
 
 
 
 
 
7505319
 
 
 
 
 
 
 
 
 
 
 
 
 
 
d5b3766
7505319
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
d5b3766
55acd9b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
import os

os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'

import logging
import io
import pickle
from pathlib import Path
from fastapi import FastAPI, File, UploadFile, HTTPException
from fastapi.staticfiles import StaticFiles
from fastapi.responses import FileResponse
from pydantic import BaseModel
from PIL import Image
import torch
import torch.nn.functional as F
from transformers import BertTokenizer, BertForSequenceClassification
import tensorflow as tf
import numpy as np
from huggingface_hub import hf_hub_download

# --- Configuration ---
logging.basicConfig(level=logging.INFO)
STATIC_DIR = Path("static")

# --- Device Configuration ---
device = torch.device('cpu')
try:
    tf.config.set_visible_devices([], 'GPU')
    logging.info("TensorFlow GPU disabled. Using CPU.")
except (RuntimeError, ValueError) as e:
    logging.warning(f"Could not disable GPU for TensorFlow: {e}")

# --- Model Loading ---
def load_models():
    logging.info("Loading all models from the Hub...")
    try:
        tokenizer = BertTokenizer.from_pretrained("muhalwan/sental")
        sentiment_model = BertForSequenceClassification.from_pretrained("muhalwan/sental")
        sentiment_model.to(device)
        logging.info("Sentiment analysis model loaded successfully.")
    except Exception as e:
        tokenizer, sentiment_model = None, None
        logging.error(f"Error loading sentiment model: {e}")

    try:
        model_path = hf_hub_download(repo_id="muhalwan/catndog", filename="catdog_best.keras")
        cat_dog_model = tf.keras.models.load_model(model_path, compile=False)
        logging.info("Cat & Dog classifier model loaded successfully.")
    except Exception as e:
        cat_dog_model = None
        logging.error(f"Error loading cat & dog model: {e}")

    try:
        xgb_model_path = hf_hub_download(repo_id="muhalwan/california_housing_price_predictor", filename="xgb_model.pkl")
        with open(xgb_model_path, "rb") as f:
            housing_model = pickle.load(f)
        scaler_path = hf_hub_download(repo_id="muhalwan/california_housing_price_predictor", filename="scaler.pkl")
        with open(scaler_path, "rb") as f:
            housing_scaler = pickle.load(f)
        logging.info("Housing price model and scaler loaded successfully.")
    except Exception as e:
        housing_model, housing_scaler = None, None
        logging.error(f"Error loading housing price model: {e}")

    return tokenizer, sentiment_model, cat_dog_model, housing_model, housing_scaler

# --- FastAPI App Initialization ---
app = FastAPI()
tokenizer, sentiment_model, cat_dog_model, housing_model, housing_scaler = load_models()
app.mount("/static", StaticFiles(directory=STATIC_DIR), name="static")

class SentimentRequest(BaseModel):
    text: str

class HousingRequest(BaseModel):
    MedInc: float
    HouseAge: float
    AveRooms: float
    AveBedrms: float
    Population: float
    AveOccup: float
    Latitude: float
    Longitude: float

# --- API Endpoints ---
@app.get("/")
async def read_root():
    return FileResponse('index.html')

@app.post("/predict/sentiment")
async def predict_sentiment(request: SentimentRequest):
    if not tokenizer or not sentiment_model:
        raise HTTPException(status_code=503, detail="Sentiment model is not available.")
    try:
        inputs = tokenizer(request.text, return_tensors='pt', truncation=True, max_length=512)
        inputs = {k: v.to(device) for k, v in inputs.items()}
        with torch.no_grad():
            outputs = sentiment_model(**inputs)
        probabilities = F.softmax(outputs.logits, dim=-1).squeeze()
        labels = ['Bearish', 'Bullish']
        prediction = labels[torch.argmax(probabilities).item()]
        return {"prediction": prediction}
    except Exception as e:
        logging.error(f"Sentiment prediction error: {e}")
        raise HTTPException(status_code=500, detail="An error occurred during sentiment analysis.")

@app.post("/predict/catdog")
async def predict_catdog(file: UploadFile = File(...)):
    if not cat_dog_model:
        raise HTTPException(status_code=503, detail="Cat & Dog model is not available.")
    try:
        contents = await file.read()
        image = Image.open(io.BytesIO(contents))
        _, height, width, _ = cat_dog_model.input_shape
        img_resized = image.resize((width, height))
        if img_resized.mode == 'RGBA':
            img_resized = img_resized.convert('RGB')
        img_array = tf.keras.utils.img_to_array(img_resized)
        img_array = tf.keras.applications.efficientnet.preprocess_input(img_array)
        img_array = np.expand_dims(img_array, axis=0)
        prob = cat_dog_model.predict(img_array, verbose=0)[0, 0]
        label = "Dog" if prob >= 0.5 else "Cat"
        return {"prediction": label}
    except Exception as e:
        logging.error(f"Cat/Dog prediction error: {e}")
        raise HTTPException(status_code=500, detail="An error occurred during image classification.")

@app.post("/predict/housing")
async def predict_housing(request: HousingRequest):
    if not housing_model or not housing_scaler:
        raise HTTPException(status_code=503, detail="Housing model is not available.")
    try:
        input_data = np.array([[
            request.MedInc, request.HouseAge, request.AveRooms, request.AveBedrms,
            request.Population, request.AveOccup, request.Latitude, request.Longitude
        ]])
        data_scaled = housing_scaler.transform(input_data)
        raw_prediction = housing_model.predict(data_scaled)[0]
        final_prediction = raw_prediction * 100000
        return {"prediction": f"${final_prediction:,.2f}"}
    except Exception as e:
        logging.error(f"Housing prediction error: {e}")
        raise HTTPException(status_code=500, detail="An error occurred during housing price prediction.")