Update download_models.py

download_models.py

@@ -4,7 +4,6 @@ import sys
 # ==========================
 # HUGGINGFACE AUTH
 # ==========================
-
 from huggingface_hub import login
 
 token = os.getenv("HF_TOKEN")
@@ -18,7 +17,7 @@ else:
 # ==========================
 # CACHE DIR
 # ==========================
-
+# Sinkronkan dengan CACHE_DIR di app.py (Canvas)
 CACHE_DIR = "/home/user/app/models_cache"
 os.makedirs(CACHE_DIR, exist_ok=True)
 
@@ -27,22 +26,22 @@ print("DOWNLOADING AI MODELS")
 print("================================")
 
 # ==========================
-# WHISPER MODEL
+# WHISPER MODEL (UPGRADED TO BASE)
 # ==========================
-
 try:
     from faster_whisper import WhisperModel
 
-    print("Downloading Whisper tiny model...")
+    # Mengubah "tiny" menjadi "base" sesuai kebutuhan app.py terbaru
+    print("Downloading Whisper BASE model (Better Accuracy)...")
 
     WhisperModel(
-        "tiny",
+        "base",
         device="cpu",
         compute_type="int8",
         download_root=CACHE_DIR
     )
 
-    print("Whisper download complete")
+    print("Whisper BASE download complete")
 
 except Exception as e:
     print("WHISPER ERROR:", e)
@@ -51,16 +50,13 @@ except Exception as e:
 # ==========================
 # DISTILBERT SENTIMENT
 # ==========================
-
 try:
-    from transformers import AutoTokenizer
-    from transformers import AutoModelForSequenceClassification
+    from transformers import AutoTokenizer, AutoModelForSequenceClassification
 
     model_name = "distilbert-base-uncased-finetuned-sst-2-english"
-
     save_path = os.path.join(CACHE_DIR, "distilbert")
 
-    print("Downloading DistilBERT...")
+    print(f"Downloading {model_name}...")
 
     tokenizer = AutoTokenizer.from_pretrained(model_name)
     model = AutoModelForSequenceClassification.from_pretrained(model_name)