Deep Patel commited on
Commit
3364379
·
unverified ·
1 Parent(s): d895a7f

Suppress ONNX export trace warning (#5437)

Browse files

Checking for `onnx_dynamic` first should suppress the warning:

```log
TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
if self.grid[i].shape[2:4] != x[i].shape[2:4] or self.onnx_dynamic
```

Files changed (1) hide show
  1. models/yolo.py +1 -1
models/yolo.py CHANGED
@@ -55,7 +55,7 @@ class Detect(nn.Module):
55
  x[i] = x[i].view(bs, self.na, self.no, ny, nx).permute(0, 1, 3, 4, 2).contiguous()
56
 
57
  if not self.training: # inference
58
- if self.grid[i].shape[2:4] != x[i].shape[2:4] or self.onnx_dynamic:
59
  self.grid[i], self.anchor_grid[i] = self._make_grid(nx, ny, i)
60
 
61
  y = x[i].sigmoid()
 
55
  x[i] = x[i].view(bs, self.na, self.no, ny, nx).permute(0, 1, 3, 4, 2).contiguous()
56
 
57
  if not self.training: # inference
58
+ if self.onnx_dynamic or self.grid[i].shape[2:4] != x[i].shape[2:4]:
59
  self.grid[i], self.anchor_grid[i] = self._make_grid(nx, ny, i)
60
 
61
  y = x[i].sigmoid()