您好，请问动态导出后，batch变为了-1，为什么在多张图片输入进行多batch推理时仍然会报错呢？
错误信息是：onnxruntime.capi.onnxruntime_pybind11_state.Fail: [ONNXRuntimeError] : 1 : FAIL : Non-zero status code returned while running Split node. Name:'Split_9' Status Message: Cannot split using values in 'split' attribute. Axis=0 Input shape={6,3,800,800} NumOutputs=1 Num entries in 'split' (must equal number of outputs) was 1 Sum of sizes in 'split' (must equal size of selected axis) was 1
我的输入时
pil_img = [transform(self.pils_img[i]).unsqueeze(0).cpu().numpy() for i in range(len(self.pils_img))]
pil_img = np.concatenate(pil_img, axis=0)
ort_inputs = {"inputs": pil_img}
scores,boxs = self.session.run(None,ort_inputs)
pil_img shape是 ndarray(6,3,800,800)
您好，请问动态导出后，batch变为了-1，为什么在多张图片输入进行多batch推理时仍然会报错呢？
错误信息是：onnxruntime.capi.onnxruntime_pybind11_state.Fail: [ONNXRuntimeError] : 1 : FAIL : Non-zero status code returned while running Split node. Name:'Split_9' Status Message: Cannot split using values in 'split' attribute. Axis=0 Input shape={6,3,800,800} NumOutputs=1 Num entries in 'split' (must equal number of outputs) was 1 Sum of sizes in 'split' (must equal size of selected axis) was 1
我的输入时
pil_img = [transform(self.pils_img[i]).unsqueeze(0).cpu().numpy() for i in range(len(self.pils_img))]
pil_img = np.concatenate(pil_img, axis=0)
ort_inputs = {"inputs": pil_img}
scores,boxs = self.session.run(None,ort_inputs)
pil_img shape是 ndarray(6,3,800,800)