Spaces:
Configuration error
Configuration error
File size: 7,184 Bytes
87705b0 ea103f3 87705b0 5e4885c ea103f3 5e4885c ea103f3 00b43a2 fe32675 ea103f3 fe32675 00b43a2 ea103f3 4ab1ac8 c51dbd9 4ab1ac8 ea103f3 4ab1ac8 0c02318 4ab1ac8 00b43a2 ea103f3 00b43a2 ea103f3 00b43a2 87705b0 ea103f3 5e4885c 87705b0 ea103f3 5e4885c fe32675 87705b0 50bf3a0 87705b0 50bf3a0 87705b0 ea103f3 fe32675 87705b0 ea103f3 00b43a2 5e4885c 87705b0 5e4885c 00b43a2 5e4885c 00b43a2 5e4885c 00b43a2 5e4885c ea103f3 dd26f06 29fec18 5e4885c ea103f3 5e4885c ea103f3 5e4885c ea103f3 5e4885c 00b43a2 ea103f3 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 |
from fastapi import FastAPI, HTTPException
from fastapi.responses import RedirectResponse
from pydantic import BaseModel
from contextlib import asynccontextmanager
import pandas as pd
import time
import json
import os
from datetime import datetime
from pathlib import Path
from huggingface_hub import hf_hub_download
from src.model.model import load_model
class ClientID(BaseModel):
sk_id_curr: int
def get_features_by_id(sk_id_curr: int) -> pd.DataFrame:
# CAS TEST (pytest injecte app.state.features)
if hasattr(app.state, "features") and app.state.features is not None:
df = app.state.features
else:
# CAS LOCAL
local_path = Path(__file__).resolve().parents[2] / "Data" / "features_clients.csv"
if local_path.exists():
df = pd.read_csv(local_path)
else:
# CAS HUGGINGFACE
path = hf_hub_download(
repo_id="PCelia/credit-scoring-model",
filename="features_clients.csv",
token=os.environ.get("HF_TOKEN")
)
df = pd.read_csv(path)
if "SK_ID_CURR" not in df.columns:
raise KeyError("Dataset invalide")
row = df[df["SK_ID_CURR"] == sk_id_curr]
if row.empty:
raise KeyError("Client not found")
return row.drop(columns=["SK_ID_CURR"])
def load_features() -> pd.DataFrame:
# CAS LOCAL
local_path = Path(__file__).resolve().parents[2] / "Data" / "features_clients.csv"
if local_path.exists():
return pd.read_csv(local_path)
# CAS HUGGINGFACE
path = hf_hub_download(
repo_id="PCelia/credit-scoring-model",
filename="features_clients.csv",
token=os.environ.get("HF_TOKEN")
)
return pd.read_csv(path)
@asynccontextmanager
async def lifespan(app: FastAPI):
app.state.model = load_model()
# En test (pytest), on ne charge pas les features ici
if os.environ.get("PYTEST_CURRENT_TEST") is None:
app.state.features = load_features()
yield
# @asynccontextmanager
# async def lifespan(app: FastAPI):
# app.state.model = load_model()
# app.state.features = load_features()
# yield
app = FastAPI(lifespan=lifespan)
@app.get("/")
def root():
return RedirectResponse(url="/docs")
@app.post("/predict_by_id")
def predict_by_id(payload: ClientID):
start_total = time.perf_counter()
try:
start_features = time.perf_counter()
X = get_features_by_id(payload.sk_id_curr)
features_time = time.perf_counter() - start_features
except KeyError:
raise HTTPException(status_code=404, detail="Client not found")
start_infer = time.perf_counter()
score = float(app.state.model.predict_proba(X)[:, 1][0])
infer_time = time.perf_counter() - start_infer
total_time = time.perf_counter() - start_total
print(
f"features={features_time:.3f}s | "
f"infer={infer_time:.3f}s | "
f"total={total_time:.3f}s"
)
if not X.empty:
inputs_dict = X.to_dict(orient="records")[0]
else:
inputs_dict = {}
log_entry = {
"timestamp": datetime.utcnow().isoformat(),
"endpoint": "/predict_by_id",
"sk_id_curr": payload.sk_id_curr,
"score": score,
"features_time": features_time,
"inference_time": infer_time,
"total_time": total_time,
"inputs": inputs_dict,
}
with open("api_logs.jsonl", "a") as f:
f.write(json.dumps(log_entry) + "\n")
return {
"score": score,
"features_time": features_time,
"inference_time": infer_time,
"total_time": total_time,
}
# from fastapi import FastAPI, HTTPException
# from pydantic import BaseModel
# from contextlib import asynccontextmanager
# import pandas as pd
# import os
# import time
# import json
# from datetime import datetime
# from huggingface_hub import hf_hub_download
# from src.model.model import load_model
# from fastapi.responses import RedirectResponse
# class ClientID(BaseModel):
# sk_id_curr: int
# from pathlib import Path
# import os
# def get_features_by_id(sk_id_curr: int) -> pd.DataFrame:
# # CAS TEST : features injectées par pytest
# if hasattr(app.state, "features") and app.state.features is not None:
# df = app.state.features
# else:
# # CAS LOCAL
# local_path = Path(__file__).resolve().parents[2] / "Data" / "features_clients.csv"
# if local_path.exists():
# df = pd.read_csv(local_path)
# else:
# # CAS HF : repo MODELE
# path = hf_hub_download(
# repo_id="PCelia/credit-scoring-model",
# filename="features_clients.csv",
# token=os.environ.get("HF_TOKEN")
# )
# df = pd.read_csv(path)
# row = df[df["SK_ID_CURR"] == sk_id_curr]
# if row.empty:
# raise KeyError("Client not found")
# return row.drop(columns=["SK_ID_CURR"])
# def load_features() -> pd.DataFrame:
# # CAS LOCAL
# local_path = Path(__file__).resolve().parents[2] / "Data" / "features_clients.csv"
# if local_path.exists():
# return pd.read_csv(local_path)
# # CAS HF
# path = hf_hub_download(
# repo_id="PCelia/credit-scoring-model",
# filename="features_clients.csv",
# token=os.environ.get("HF_TOKEN")
# )
# return pd.read_csv(path)
# @asynccontextmanager
# async def lifespan(app: FastAPI):
# app.state.model = load_model()
# # chargement unique des features
# app.state.features = load_features()
# yield
# app = FastAPI(lifespan=lifespan)
# @app.get("/")
# def root():
# return RedirectResponse(url="/docs")
# @app.post("/predict_by_id")
# def predict_by_id(payload: ClientID):
# start_total = time.perf_counter()
# try:
# start_features = time.perf_counter()
# X = get_features_by_id(payload.sk_id_curr)
# features_time = time.perf_counter() - start_features
# except KeyError:
# raise HTTPException(status_code=404, detail="Client not found")
# start_infer = time.perf_counter()
# score = float(app.state.model.predict_proba(X)[:, 1][0])
# infer_time = time.perf_counter() - start_infer
# total_time = time.perf_counter() - start_total
# print(
# f"features={features_time:.3f}s | "
# f"infer={infer_time:.3f}s | "
# f"total={total_time:.3f}s"
# )
# inputs_dict = X.to_dict(orient="records")[0]
# log_entry = {
# "timestamp": datetime.utcnow().isoformat(),
# "endpoint": "/predict_by_id",
# "sk_id_curr": payload.sk_id_curr,
# "score": score,
# "features_time": features_time,
# "inference_time": infer_time,
# "total_time": total_time,
# "inputs": inputs_dict
# }
# import os
# print("LOG PATH:", os.path.abspath("api_logs.jsonl"))
# with open("api_logs.jsonl", "a") as f:
# f.write(json.dumps(log_entry) + "\n")
# return {
# "score": score,
# "features_time": features_time,
# "inference_time": infer_time,
# "total_time": total_time
# }
|