Update project_model.py

project_model.py

@@ -16,20 +16,29 @@ from ultralytics import YOLO
 from transformers import pipeline, DPTFeatureExtractor, DPTForDepthEstimation
 from TTS.api import TTS
 
+
+from huggingface_hub import login
+import os
+
+# Login using token stored in environment variable
+login(token=os.getenv("gemma3_token"))
+
 # Load models
 
-device = "cuda" if torch.cuda.is_available() else "cpu"
-yolo_model = YOLO("yolov8n.pt")
-depth_model = DPTForDepthEstimation.from_pretrained("Intel/dpt-large").to(device).eval()
+device = "cuda" if torch.cuda.is_available() else "cpu" # Enable GPU
+
+yolo_model = YOLO("yolov9n.pt") # Load YOLOv9
+depth_model = DPTForDepthEstimation.from_pretrained("Intel/dpt-large").to(device).eval() # Load MiDaS
 depth_feat = DPTFeatureExtractor.from_pretrained("Intel/dpt-large")
-whisper_pipe = pipeline("automatic-speech-recognition", model="openai/whisper-small", device=0 if torch.cuda.is_available() else -1)
+whisper_pipe = pipeline("automatic-speech-recognition", model="openai/whisper-small", device=0 if torch.cuda.is_available() else -1) # Load Whisper
+# Load Gemma-3-4B
 gemma_pipe = pipeline(
     "image-text-to-text",
     model="google/gemma-3-4b-it",
     device=0 if torch.cuda.is_available() else -1,
     torch_dtype=torch.bfloat16 if torch.cuda.is_available() else torch.float32
 )
-tts = TTS(model_name="tts_models/en/ljspeech/tacotron2-DDC")
+tts = TTS(model_name="tts_models/en/ljspeech/tacotron2-DDC") # Load Text-to-Speech (TTS)
 
 # Function to process image and audio
 def process_inputs(image: Image.Image, audio_path: str):