chore: regenerate_train_data accepts API key and https URL

scripts/regenerate_train_data.py

@@ -131,9 +131,22 @@ def parse_arguments():
         nargs="+",
         help="Server address and port for sglang model server",
     )
+    server_group.add_argument(
+        "--api-key",
+        type=str,
+        default="None",
+        help="API key for the endpoint",
+    )
     return parser.parse_args()
 
 
+def normalize_server_address(address: str) -> str:
+    """Add http:// prefix if no protocol is specified."""
+    if not address.startswith(("http://", "https://")):
+        return f"http://{address}"
+    return address
+
+
 def get_random_reasoning_effort() -> str:
     """Get a random reasoning effort level for the model with weighted probabilities."""
     # usage example: https://huggingface.co/openai/gpt-oss-20b/discussions/28
@@ -198,7 +211,7 @@ def call_sglang(
     max_tokens=None,
 ) -> str:
     """Send a batch of prompts to sglang /v1/completions."""
-    client = OpenAI(base_url=f"http://{server_address}/v1", api_key="None")
+    client = OpenAI(base_url=f"{server_address}/v1", api_key=args.api_key)
 
     messages = data["conversations"]
     regenerated_messages = []
@@ -248,6 +261,11 @@ def main():
     # Parse command line arguments
     args = parse_arguments()
 
+    # Normalize server addresses to ensure they have protocol prefix
+    args.server_address = [
+        normalize_server_address(addr) for addr in args.server_address
+    ]
+
     # Validate parameters
     if not (0.0 <= args.temperature <= 1.0):
         raise ValueError("Temperature must be between 0.0 and 1.0")
@@ -261,6 +279,7 @@ def main():
     print(f"  Concurrency: {args.concurrency}")
     print(f"  Temperature: {args.temperature}")
     print(f"  API URL: {args.server_address}")
+    print(f"  API key set: {args.api_key != "None"}")
     print(f"  Input file: {args.input_file_path}")
     print(f"  Output file: {args.output_file_path}")
     print(f"  Resume mode: {args.resume}")