repo-type + rss

rss.xml

@@ -0,0 +1,38 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<rss version="2.0"
+     xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd"
+     xmlns:media="http://search.yahoo.com/mrss/">
+
+  <channel>
+    <title>Daily Papers – Hugging Face</title>
+    <link>https://huggingface.co/papers</link>
+    <description>Listen to an AI-generated conversation about the most upvoted research paper on Hugging Face each day.</description>
+    <language>en-us</language>
+    <itunes:author>Hugging Face</itunes:author>
+    <itunes:summary>Each day, this podcast dives into the top trending ML paper on Hugging Face. Discussions are AI-generated — verify facts before citing.</itunes:summary>
+    <itunes:explicit>false</itunes:explicit>
+    <itunes:category text="Technology"/>
+    <itunes:owner>
+      <itunes:name>HF</itunes:name>
+      <itunes:email>[email protected]</itunes:email> <!-- Replace with your email -->
+    </itunes:owner>
+    <itunes:image href="https://huggingface.co/spaces/fdaudens/podcast-jobs/resolve/main/images/cover.png"/>
+    <image>
+      <url>https://huggingface.co/spaces/fdaudens/podcast-jobs/resolve/main/images/cover.png</url>
+      <title>Daily Papers – Hugging Face</title>
+      <link>https://huggingface.co/papers</link>
+    </image>
+    <lastBuildDate>Tue, 13 May 2025 10:00:00 +0000</lastBuildDate>
+
+    <!-- Example Episode -->
+    <item>
+      <title>Ep 1 – “Title of the Most Upvoted Paper”</title>
+      <description>Today’s top paper on Hugging Face is: “XYZ.” Listen to an AI discuss the ideas, findings, and questions it raises. (Warning: AI may hallucinate)</description>
+      <pubDate>Tue, 13 May 2025 10:00:00 +0000</pubDate>
+      <enclosure url="https://yourpodcastwebsite.com/audio/episode1.mp3" length="12345678" type="audio/mpeg"/>
+      <guid>https://yourpodcastwebsite.com/audio/episode1.mp3</guid>
+      <itunes:explicit>false</itunes:explicit>
+    </item>
+
+  </channel>
+</rss>

run_job.py

@@ -8,21 +8,7 @@ import requests
 import json
 from datetime import datetime
 import os
-
-# Scan for existing podcasts in the space
-def scan_podcasts():
-    podcast_dir = "podcasts"
-    
-    if not os.path.exists(podcast_dir):
-        os.makedirs(podcast_dir, exist_ok=True)
-        return []
-    
-    # Get all WAV files in the podcasts directory
-    podcasts = sorted([f.replace(".wav", "") 
-                      for f in os.listdir(podcast_dir) 
-                      if f.endswith(".wav")], reverse=True)
-    
-    return podcasts
+import tempfile
 
 def submit_job(
     inference_provider: str,
@@ -107,27 +93,44 @@ def main():
     if audio_segments:
         full_audio = np.concatenate(audio_segments)
         
-        # 4. Save as WAV file in the Space's file system
-        # Create podcasts directory if it doesn't exist
-        podcast_dir = "podcasts"
-        os.makedirs(podcast_dir, exist_ok=True)
+        # Create a temporary file
+        with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as temp_file:
+            temp_path = temp_file.name
+            sf.write(temp_path, full_audio, sr)
+        
+        # Get API token from environment
+        hf_token = os.environ.get("HF_TOKEN") or os.environ.get("HF_API_KEY")
+        if hf_token is None:
+            print("No Hugging Face token found in environment. Cannot upload to Space.")
+            return
+            
+        # Initialize the Hugging Face API
+        api = HfApi(token=hf_token)
         
-        # Generate filename with base name and date
+        # Set up Space path info
+        space_id = "fdaudens/podcast-jobs"  # Your Space ID
         today = datetime.now().strftime("%Y-%m-%d")
         base_name = args.name
         podcast_filename = f"{base_name}-{today}.wav"
-        podcast_path = os.path.join(podcast_dir, podcast_filename)
         
-        # Save the file
-        sf.write(podcast_path, full_audio, sr)
-        print(f"Podcast audio saved to {podcast_path}")
+        # Path in the Space repository
+        space_path = f"podcasts/{podcast_filename}"
+        
+        # Upload directly to the Space (crucial: repo_type="space")
+        print(f"Uploading podcast to Space {space_id} at path {space_path}...")
+        api.upload_file(
+            path_or_fileobj=temp_path,
+            path_in_repo=space_path,
+            repo_id=space_id,
+            repo_type="space", 
+            token=hf_token
+        )
         
-        # Provide the access URL
-        print(f"Access URL: https://huggingface.co/spaces/fdaudens/podcast-jobs/blob/main/{podcast_path}")
+        # Clean up temporary file
+        os.unlink(temp_path)
         
-        # Also save a temporary local copy for debugging
-        sf.write("podcast.wav", full_audio, sr)
-        print("Temporary copy saved as podcast.wav")
+        print(f"Podcast audio uploaded to Space at {space_path}")
+        print(f"Access URL: https://huggingface.co/spaces/{space_id}/blob/main/{space_path}")
     else:
         print("No audio generated.")