v0.1: moved dataset to personal space and added access token. Avoids huggingface_hub.errors.HfHubHTTPError: 401 Client Error: Unauthorized for url: https://huggingface.co/api/repos/create

utils.py

@@ -17,7 +17,7 @@ from datasets import (
 
 
 # Hugging Face evaluation dataset
-HF_DATASET_NAME = "atlasia/Moroccan-STT-Eval-Dataset"
+HF_DATASET_NAME = "BounharAbdelaziz/Moroccan-STT-Eval-Dataset"
 
 # Models paths
 MODEL_PATHS = {
@@ -29,6 +29,9 @@ MODEL_PATHS = {
 # Access token to models
 STT_MODEL_TOKEN = os.environ.get("STT_MODEL_TOKEN")
 
+# Access token to dataset
+STT_EVAL_DATASET_TOKEN = os.environ.get("STT_EVAL_DATASET_TOKEN")
+
 # ---------------------------------------------------------------------------- #
 # ---------------------------------------------------------------------------- #
 
@@ -58,7 +61,7 @@ def create_html_image(image_path):
 
 def load_or_create_dataset():
     try:
-        dataset = load_dataset(HF_DATASET_NAME)
+        dataset = load_dataset(HF_DATASET_NAME, token=STT_EVAL_DATASET_TOKEN)
         return dataset
     except Exception as e:
         print(f"[INFO] Dataset not found or error loading: {e}. Creating a new one.")
@@ -85,7 +88,7 @@ def load_or_create_dataset():
 def save_to_hf_dataset(audio_signal, model_choice, transcription):
     print("[INFO] Loading dataset...")
     try:
-        dataset = load_dataset(HF_DATASET_NAME)
+        dataset = load_dataset(HF_DATASET_NAME, token=STT_EVAL_DATASET_TOKEN)
         print("[INFO] Dataset loaded successfully.")
     except Exception as e:
         print(f"[INFO] Dataset not found or error loading. Creating a new one.")
@@ -130,7 +133,7 @@ def save_to_hf_dataset(audio_signal, model_choice, transcription):
     dataset["train"] = updated_train_dataset
 
     print("[INFO] Pushing the updated dataset...")
-    dataset.push_to_hub(HF_DATASET_NAME)
+    dataset.push_to_hub(HF_DATASET_NAME, token=STT_EVAL_DATASET_TOKEN)
 
     print("[INFO] Dataset updated and pushed successfully.")