Update app.py

app.py

@@ -13,7 +13,7 @@ from huggingface_hub import HfApi, whoami
 from torch.jit import TracerWarning
 from transformers import AutoConfig, GenerationConfig
 
-# Suppress TorchScript tracer warnings in parent process
+# Suppress TorchScript tracer warnings in this process
 warnings.filterwarnings("ignore", category=TracerWarning)
 
 logging.basicConfig(level=logging.INFO)
@@ -100,50 +100,56 @@ class ModelConverter:
     def convert_model(self, input_model_id: str) -> Tuple[bool, Optional[str]]:
         """
         Convert the model to ONNX format, always exporting attention maps.
-        Relocate generation parameters and suppress warnings in subprocess.
-        Strip out static relocation and tracer warnings from stderr.
+        Relocate generation parameters, suppress tracer warnings, and
+        strip out both relocation and tracer warnings from stderr.
         """
         try:
-            # Prepare local model folder for config edits
+            # Prepare local directory for config edits
             model_dir = self.config.repo_path / "models" / input_model_id
             model_dir.mkdir(parents=True, exist_ok=True)
 
-            # Reload and relocate generation parameters
-            config = AutoConfig.from_pretrained(input_model_id)
-            gen_cfg = GenerationConfig.from_model_config(config)
-            for key in gen_cfg.to_dict().keys():
-                if hasattr(config, key):
-                    setattr(config, key, None)
-            config.save_pretrained(model_dir)
-            gen_cfg.save_pretrained(model_dir)
-
-            # Build command with warning suppression flag
+            # Load and relocate generation parameters
+            base_config = AutoConfig.from_pretrained(input_model_id)
+            gen_config = GenerationConfig.from_model_config(base_config)
+            # Remove generation params from base config
+            for key in gen_config.to_dict():
+                if hasattr(base_config, key):
+                    setattr(base_config, key, None)
+            base_config.save_pretrained(model_dir)
+            gen_config.save_pretrained(model_dir)
+
+            # Set up env to suppress tracer warnings in subprocess
+            env = os.environ.copy()
+            env["PYTHONWARNINGS"] = "ignore::torch.jit.TracerWarning"
+
+            # Build conversion command
             cmd = [
                 sys.executable,
-                "-W", "ignore::torch.jit.TracerWarning",
                 "-m", "scripts.convert",
                 "--quantize",
                 "--trust_remote_code",
                 "--model_id", input_model_id,
                 "--output_attentions",
             ]
+
+            # Execute conversion
             result = subprocess.run(
                 cmd,
                 cwd=self.config.repo_path,
                 capture_output=True,
                 text=True,
-                env=os.environ.copy(),
+                env=env,
             )
 
-            # Filter stderr lines
-            filtered = []
+            # Filter out relocation and tracer warnings
+            lines = []
             for ln in result.stderr.splitlines():
-                if ln.startswith("Moving the following attributes"):  # relocation warning
+                if ln.startswith("Moving the following attributes"):
                     continue
-                if "TracerWarning" in ln:  # any tracer warnings
+                if "TracerWarning" in ln:
                     continue
-                filtered.append(ln)
-            stderr = "\n".join(filtered)
+                lines.append(ln)
+            stderr = "\n".join(lines)
 
             if result.returncode != 0:
                 return False, stderr