_archive/inspect_models.py

"""Quick inspection of specific ONNX models."""
import onnx
from pathlib import Path

for idx in [11, 12, 22, 33]:
    matches = list(Path('oneocr_extracted/onnx_models').glob(f'model_{idx:02d}_*'))
    if not matches:
        continue
    m = onnx.load(str(matches[0]))
    print(f'\n=== model_{idx:02d} ({matches[0].name}) ===')
    print('Opsets:', [(o.domain, o.version) for o in m.opset_import])
    custom_ops = set()
    for n in m.graph.node:
        if n.domain:
            custom_ops.add((n.domain, n.op_type))
    print('Custom ops:', list(custom_ops))
    for i in m.graph.input:
        dims = []
        if i.type.tensor_type.HasField('shape'):
            for d in i.type.tensor_type.shape.dim:
                v = str(d.dim_value) if d.dim_value else (d.dim_param or '?')
                dims.append(v)
        etype = i.type.tensor_type.elem_type
        print(f'  Input:  {i.name}  shape=[{", ".join(dims)}]  dtype={etype}')
    for o in m.graph.output:
        dims = []
        if o.type.tensor_type.HasField('shape'):
            for d in o.type.tensor_type.shape.dim:
                v = str(d.dim_value) if d.dim_value else (d.dim_param or '?')
                dims.append(v)
        etype = o.type.tensor_type.elem_type
        print(f'  Output: {o.name}  shape=[{", ".join(dims)}]  dtype={etype}')
    cnt = {}
    for n in m.graph.node:
        key = f'{n.domain}::{n.op_type}' if n.domain else n.op_type
        cnt[key] = cnt.get(key, 0) + 1
    print('  Nodes:', dict(cnt))
    
    # Show attributes for custom ops
    for n in m.graph.node:
        if n.domain:
            attrs = {a.name: a for a in n.attribute}
            attr_summary = {}
            for k, a in attrs.items():
                if a.type == 2:  # FLOAT
                    attr_summary[k] = a.f
                elif a.type == 1:  # INT
                    attr_summary[k] = a.i
                elif a.type == 3:  # STRING
                    attr_summary[k] = a.s.decode()
            print(f'  CustomOp {n.op_type} attrs: {attr_summary}')
            # Show input/output names
            print(f'    inputs:  {list(n.input)}')
            print(f'    outputs: {list(n.output)}')
            break  # just show first custom op