fix some issues

Dockerfile

@@ -1,11 +1,24 @@
-FROM python:3.10
+FROM python:3.10-slim
 
+# Set working directory
 WORKDIR /app
 
+# Copy all files
 COPY . .
 
+# Install Python dependencies
 RUN pip install --no-cache-dir -r requirements.txt
 
+# Make a writable cache directory for model downloads
+RUN mkdir -p /app/cache
+
+# Set environment variable to use this cache path (optional but good practice)
+ENV HF_HOME=/app/cache
+ENV TRANSFORMERS_CACHE=/app/cache
+ENV HF_HUB_CACHE=/app/cache
+
+# Expose FastAPI port
 EXPOSE 7860
 
+# Run the FastAPI app
 CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]

config.py

@@ -1,18 +1,15 @@
-import psutil
 import os
-#from dotenv import load_dotenv
-
+import psutil
 
-#load_dotenv()
-WHISPER_MODEL_NAME = "tiny"
+WHISPER_MODEL_NAME = "Systran/faster-whisper-tiny"
 WHISPER_DEVICE = "cpu"
 WHISPER_COMPUTE_TYPE = "int8"
+
 PYANNOTE_AUTH_TOKEN = os.getenv("HUGGINGFACE_API_KEY")
+
 SUMMARIZER_MODEL = "facebook/bart-large-cnn"
 SUMMARIZER_MAX_LENGTH = 150
 SUMMARIZER_MIN_LENGTH = 50
+
 SPACY_MODEL = "en_core_web_sm"
 CPU_THREADS = max(1, psutil.cpu_count(logical=False))
-
-if not PYANNOTE_AUTH_TOKEN:
-    raise ValueError("HUGGINGFACE_API_KEY not set in environment variables")

models.py

@@ -1,59 +1,34 @@
+from huggingface_hub import snapshot_download
+import os
 import logging
 from faster_whisper import WhisperModel
-import spacy
-from transformers import pipeline
-import os
 
 logger = logging.getLogger(__name__)
 
-
 def load_whisper(config):
     logger.info("Loading Whisper model...")
+
     try:
+        # Use /app/cache or /tmp — both are writable on HF Spaces
+        cache_dir = "/app/cache"
+        os.makedirs(cache_dir, exist_ok=True)
+
+        model_dir = snapshot_download(
+            repo_id=config.WHISPER_MODEL_NAME,
+            cache_dir=cache_dir,
+            token=os.getenv("HUGGINGFACE_API_KEY")
+        )
+
         model = WhisperModel(
-            config.WHISPER_MODEL_NAME,
+            model_dir,
             device=config.WHISPER_DEVICE,
             compute_type=config.WHISPER_COMPUTE_TYPE,
             cpu_threads=config.CPU_THREADS
         )
-        logger.info(f"Whisper model '{config.WHISPER_MODEL_NAME}' loaded on {config.WHISPER_DEVICE}.")
+
+        logger.info(f"Whisper model '{config.WHISPER_MODEL_NAME}' loaded from {model_dir} on {config.WHISPER_DEVICE}.")
         return model
-    except Exception as e:
-        logger.error(f"Failed to load Whisper model: {e}", exc_info=True)
-        return None
 
-def load_summarizer(config):
-    logger.info("Loading Summarization pipeline...")
-    try:
-        summarizer = pipeline("summarization", model=config.SUMMARIZER_MODEL)
-        logger.info("Summarization pipeline loaded.")
-        return summarizer
     except Exception as e:
-        logger.error(f"Failed to load Summarization pipeline: {e}", exc_info=True)
+        logger.error(f"Failed to load Whisper model: {e}", exc_info=True)
         return None
-
-def load_spacy(config):
-    logger.info("Loading spaCy model...")
-
-    try:
-        nlp = spacy.load("en_core_web_sm")
-        logger.info("spaCy model 'en_core_web_sm' loaded.")
-        
-        return nlp
-
-    except OSError:
-        logger.warning("spaCy model 'en_core_web_sm' not found. Trying to download...")
-
-        try:
-            spacy.cli.download("en_core_web_sm")
-            nlp = spacy.load("en_core_web_sm")
-            logger.info("spaCy model 'en_core_web_sm' downloaded and loaded.")
-            return nlp
-        
-        except Exception as download_e:
-            logger.error(f"Failed to download or load spaCy model 'en_core_web_sm': {download_e}")
-            return None
-
-    except Exception as e:
-        logger.error(f"Failed to load spaCy model: {e}")
-        return None