Spaces:
Sleeping
Sleeping
cdborinstein commited on
Commit ·
bb9fe67
1
Parent(s): 69f8f8e
Fix server startup and backup script
Browse files- backup_to_hub.py +12 -6
- start.sh +8 -4
backup_to_hub.py
CHANGED
|
@@ -3,24 +3,30 @@ Automatic backup of Argilla annotations to HF Dataset.
|
|
| 3 |
Runs as background process, syncs every 10 minutes.
|
| 4 |
"""
|
| 5 |
import os
|
|
|
|
| 6 |
import json
|
| 7 |
import time
|
| 8 |
from pathlib import Path
|
| 9 |
from datetime import datetime
|
| 10 |
-
from huggingface_hub import CommitScheduler, login
|
| 11 |
-
import argilla as rg
|
| 12 |
|
| 13 |
# Configuration
|
| 14 |
EXPORT_DIR = Path("/exports")
|
| 15 |
-
EXPORT_DIR.mkdir(exist_ok=True)
|
| 16 |
BACKUP_REPO = os.environ.get("BACKUP_DATASET_REPO", "argus-systems/vuln-preferences-data")
|
| 17 |
SYNC_INTERVAL = int(os.environ.get("SYNC_INTERVAL_MINUTES", "10"))
|
| 18 |
ARGILLA_API_URL = "http://localhost:6900"
|
| 19 |
|
| 20 |
-
#
|
| 21 |
hf_token = os.environ.get("HF_TOKEN")
|
| 22 |
-
if hf_token:
|
| 23 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 24 |
|
| 25 |
# Setup CommitScheduler
|
| 26 |
scheduler = CommitScheduler(
|
|
|
|
| 3 |
Runs as background process, syncs every 10 minutes.
|
| 4 |
"""
|
| 5 |
import os
|
| 6 |
+
import sys
|
| 7 |
import json
|
| 8 |
import time
|
| 9 |
from pathlib import Path
|
| 10 |
from datetime import datetime
|
|
|
|
|
|
|
| 11 |
|
| 12 |
# Configuration
|
| 13 |
EXPORT_DIR = Path("/exports")
|
|
|
|
| 14 |
BACKUP_REPO = os.environ.get("BACKUP_DATASET_REPO", "argus-systems/vuln-preferences-data")
|
| 15 |
SYNC_INTERVAL = int(os.environ.get("SYNC_INTERVAL_MINUTES", "10"))
|
| 16 |
ARGILLA_API_URL = "http://localhost:6900"
|
| 17 |
|
| 18 |
+
# Check for HF_TOKEN before importing heavy dependencies
|
| 19 |
hf_token = os.environ.get("HF_TOKEN")
|
| 20 |
+
if not hf_token:
|
| 21 |
+
print("ERROR: HF_TOKEN not set, backup script exiting")
|
| 22 |
+
sys.exit(1)
|
| 23 |
+
|
| 24 |
+
from huggingface_hub import CommitScheduler, login
|
| 25 |
+
import argilla as rg
|
| 26 |
+
|
| 27 |
+
# Setup
|
| 28 |
+
EXPORT_DIR.mkdir(exist_ok=True)
|
| 29 |
+
login(token=hf_token)
|
| 30 |
|
| 31 |
# Setup CommitScheduler
|
| 32 |
scheduler = CommitScheduler(
|
start.sh
CHANGED
|
@@ -1,7 +1,11 @@
|
|
| 1 |
#!/bin/bash
|
| 2 |
|
| 3 |
-
# Start backup script in background
|
| 4 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
|
| 6 |
-
# Start Argilla server
|
| 7 |
-
|
|
|
|
| 1 |
#!/bin/bash
|
| 2 |
|
| 3 |
+
# Start backup script in background (only if HF_TOKEN is set)
|
| 4 |
+
if [ -n "$HF_TOKEN" ]; then
|
| 5 |
+
python /app/backup_to_hub.py &
|
| 6 |
+
else
|
| 7 |
+
echo "WARNING: HF_TOKEN not set, backup disabled"
|
| 8 |
+
fi
|
| 9 |
|
| 10 |
+
# Start Argilla server using the CLI from argilla-server package
|
| 11 |
+
exec argilla server start --host 0.0.0.0 --port 6900
|