Upload det01_check_det_batch.py with huggingface_hub

det01_check_det_batch.py

@@ -0,0 +1,43 @@
+#!/usr/bin/env python3
+"""
+det01 — Inspect det_500m.onnx input/output shapes and test batch > 1.
+
+Answers: does the model support batched frame input, or is it fixed at batch=1?
+"""
+import argparse
+import numpy as np
+import onnxruntime as ort
+import onnx
+
+parser = argparse.ArgumentParser()
+parser.add_argument('--model', required=True, help='Path to det_500m.onnx')
+args = parser.parse_args()
+
+# ── Declared shapes from ONNX graph ──────────────────────────────────────────
+model = onnx.load(args.model)
+
+print('=== Declared input shapes ===')
+for inp in model.graph.input:
+    shape = [d.dim_param if d.HasField('dim_param') else d.dim_value
+             for d in inp.type.tensor_type.shape.dim]
+    print(f'  {inp.name}: {shape}')
+
+print('\n=== Declared output shapes ===')
+for out in model.graph.output:
+    shape = [d.dim_param if d.HasField('dim_param') else d.dim_value
+             for d in out.type.tensor_type.shape.dim]
+    print(f'  {out.name}: {shape}')
+
+# ── Runtime test with N=1 and N=2 ────────────────────────────────────────────
+print('\n=== Runtime batch test ===')
+sess = ort.InferenceSession(args.model, providers=['CPUExecutionProvider'])
+inp0 = sess.get_inputs()[0]
+out_names = [o.name for o in sess.get_outputs()]
+
+for N in (1, 2):
+    dummy = np.random.randint(0, 255, (N, 3, 640, 640)).astype(np.float32)
+    try:
+        outs = sess.run(out_names, {inp0.name: dummy})
+        print(f'  N={N}: OK  output shapes={[list(o.shape) for o in outs]}')
+    except Exception as e:
+        print(f'  N={N}: FAILED — {e}')