FastAPI-Backend-Models / services /label_model_manage.py
Yassine Mhirsi
Enhance KpaModelManager to load fine-tuned weights from Hugging Face and update requirements to include huggingface_hub
d289997
raw
history blame
4.68 kB
"""Model manager for keypoint–argument matching model"""
import os
import torch
from transformers import AutoTokenizer, AutoModelForSequenceClassification
from huggingface_hub import hf_hub_download
import logging
logger = logging.getLogger(__name__)
class KpaModelManager:
"""Manages loading and inference for keypoint matching model"""
def __init__(self):
self.model = None
self.tokenizer = None
self.device = None
self.model_loaded = False
self.max_length = 256
self.model_id = None
def load_model(self, model_id: str, api_key: str = None):
"""Load model with weights from Hugging Face repository"""
if self.model_loaded:
logger.info("KPA model already loaded")
return
try:
logger.info(f"Loading KPA model from Hugging Face: {model_id}")
# Determine device
self.device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
logger.info(f"Using device: {self.device}")
# Store model ID
self.model_id = model_id
# Prepare token for authentication if API key is provided
token = api_key if api_key else None
# Load base tokenizer (distilbert-base-uncased)
base_model_name = "distilbert-base-uncased"
logger.info(f"Loading tokenizer from {base_model_name}...")
self.tokenizer = AutoTokenizer.from_pretrained(base_model_name)
# Load base model architecture
logger.info(f"Loading base model architecture from {base_model_name}...")
self.model = AutoModelForSequenceClassification.from_pretrained(
base_model_name,
num_labels=2
)
# Download and load fine-tuned weights from Hugging Face
logger.info(f"Downloading fine-tuned weights from {model_id}...")
weights_path = hf_hub_download(
repo_id=model_id,
filename="modele_appariement_rapide.pth",
token=token
)
logger.info(f"Loading fine-tuned weights from {weights_path}...")
checkpoint = torch.load(weights_path, map_location=self.device)
# Load state dict
if "model_state_dict" in checkpoint:
self.model.load_state_dict(checkpoint["model_state_dict"])
else:
self.model.load_state_dict(checkpoint)
self.model.to(self.device)
self.model.eval()
self.model_loaded = True
logger.info("✓ KPA model loaded successfully from Hugging Face!")
except Exception as e:
logger.error(f"Error loading KPA model: {str(e)}")
raise RuntimeError(f"Failed to load KPA model: {str(e)}")
def predict(self, argument: str, key_point: str) -> dict:
"""Run a prediction for (argument, key_point)"""
if not self.model_loaded:
raise RuntimeError("KPA model not loaded")
try:
# Tokenize input
encoding = self.tokenizer(
argument,
key_point,
truncation=True,
padding="max_length",
max_length=self.max_length,
return_tensors="pt"
).to(self.device)
# Forward pass
with torch.no_grad():
outputs = self.model(**encoding)
logits = outputs.logits
probabilities = torch.softmax(logits, dim=-1)
predicted_class = torch.argmax(probabilities, dim=-1).item()
confidence = probabilities[0][predicted_class].item()
return {
"prediction": predicted_class,
"confidence": confidence,
"label": "apparie" if predicted_class == 1 else "non_apparie",
"probabilities": {
"non_apparie": probabilities[0][0].item(),
"apparie": probabilities[0][1].item(),
},
}
except Exception as e:
logger.error(f"Error during prediction: {str(e)}")
raise RuntimeError(f"KPA prediction failed: {str(e)}")
def get_model_info(self):
return {
"model_name": self.model_id,
"device": str(self.device),
"max_length": self.max_length,
"num_labels": 2,
"loaded": self.model_loaded
}
kpa_model_manager = KpaModelManager()