search

Xilinx / Vitis-AI

Vitis AI is Xilinx’s development stack for AI inference on Xilinx hardware platforms, including both edge devices and Alveo cards.
https://www.xilinx.com/ai
Apache License 2.0
1.49k stars 633 forks source link

YOLOv5 quantization #1447

Open 60rw311 opened 4 months ago

60rw311 commented 4 months ago

Hi everyone,

I'm trying to quantize the YOLOv5n model from here. I'm using the Vitis-AI v3.0 docker with the following code:

import pytorch_nndct
import torch
from pytorch_nndct.apis import torch_quantizer, dump_xmodel

device = torch.device("cpu")

# Cargar modelo
model = torch.hub.load("ultralytics/yolov5", "yolov5n")

###### CUANTIZACION #######
input = torch.randn([1, 3, 640, 640])
quantizer = torch_quantizer('calib', model, (input), device=device)
quant_model = quantizer.quant_model
quantized_model = quantized_model.to(device)
###########################

quantized_model.eval()
results = quantized_model(rand_in)

# Resultados
quantizer.export_quant_config()

print('\nExprotar XMODEL\n')

quantizer.export_xmodel('./quantized/')

print('\nResultados cargados\n')

When I run the script, this is what I see in the terminal but finally there is no .xmodel in the directory. 

[VAIQ_NOTE]: Quant config file is empty, use default quant configuration

[VAIQ_NOTE]: Quantization calibration process start up...

[VAIQ_NOTE]: =>Quant Module is in 'cpu'.

[VAIQ_NOTE]: =>Parsing AutoShape...

[VAIQ_NOTE]: Start to trace model...
/home/vitis-ai-user/.cache/torch/hub/ultralytics_yolov5_master/models/yolo.py:100: 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.dynamic or self.grid[i].shape[2:4] != x[i].shape[2:4]:

[VAIQ_NOTE]: Finish tracing.

[VAIQ_NOTE]: Processing ops...
██████████████████████████████████████████████████| 258/258 [00:00<00:00, 2372.16it/s, OpInfo: name =

[VAIQ_WARN]: The quantizer recognize new op `aten::meshgrid` as a float operator by default.

[VAIQ_WARN]: The quantizer recognize new op `aten::silu_` as a float operator by default.

[VAIQ_WARN]: The quantizer recognize new op `aten::pow` as a float operator by default.

[VAIQ_WARN]: The quantizer recognize new op `stack` as a float operator by default.

[VAIQ_WARN]: The quantizer recognize new op `expand` as a float operator by default.

[VAIQ_WARN]: The quantizer recognize new op `aten::split_with_sizes` as a float operator by default.

[VAIQ_WARN]: The quantizer recognize new op `select` as a float operator by default.

[VAIQ_WARN]: The quantizer recognize new op `aten::type_as` as a float operator by default.

[VAIQ_NOTE]: =>Doing weights equalization...

[VAIQ_NOTE]: =>Quantizable module is generated.(quantize_result/AutoShape.py)

[VAIQ_NOTE]: =>Get module with quantization.

[VAIQ_NOTE]: =>Exporting quant config.(quantize_result/quant_info.json)

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12864 type: const

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12866 type: const

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12868 type: const

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12852 type: const

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12854 type: const

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12856 type: const

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12840 type: const

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12842 type: const

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12844 type: const

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12838 type: const

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12850 type: const

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12862 type: const

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12860 type: const

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12848 type: const

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12836 type: const

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::input_0 type: input

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Conv[model]/Conv[0]/Conv2d[conv]/input.3 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Conv[model]/Conv[1]/Conv2d[conv]/input.7 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[2]/Conv[cv1]/Conv2d[conv]/input.11 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[2]/Sequential[m]/Bottleneck[0]/Conv[cv1]/Conv2d[conv]/input.15 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[2]/Sequential[m]/Bottleneck[0]/Conv[cv2]/Conv2d[conv]/input.19 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[2]/Sequential[m]/Bottleneck[0]/10903 type: elemwise_add

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[2]/Conv[cv2]/Conv2d[conv]/input.21 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[2]/input.23 type: concat

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[2]/Conv[cv3]/Conv2d[conv]/input.25 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Conv[model]/Conv[3]/Conv2d[conv]/input.29 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[4]/Conv[cv1]/Conv2d[conv]/input.33 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[4]/Sequential[m]/Bottleneck[0]/Conv[cv1]/Conv2d[conv]/input.37 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[4]/Sequential[m]/Bottleneck[0]/Conv[cv2]/Conv2d[conv]/input.41 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[4]/Sequential[m]/Bottleneck[0]/input.43 type: elemwise_add

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[4]/Sequential[m]/Bottleneck[1]/Conv[cv1]/Conv2d[conv]/input.45 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[4]/Sequential[m]/Bottleneck[1]/Conv[cv2]/Conv2d[conv]/input.49 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[4]/Sequential[m]/Bottleneck[1]/11070 type: elemwise_add

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[4]/Conv[cv2]/Conv2d[conv]/input.51 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[4]/input.53 type: concat

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[4]/Conv[cv3]/Conv2d[conv]/input.55 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Conv[model]/Conv[5]/Conv2d[conv]/input.59 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[6]/Conv[cv1]/Conv2d[conv]/input.63 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[6]/Sequential[m]/Bottleneck[0]/Conv[cv1]/Conv2d[conv]/input.67 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[6]/Sequential[m]/Bottleneck[0]/Conv[cv2]/Conv2d[conv]/input.71 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[6]/Sequential[m]/Bottleneck[0]/input.73 type: elemwise_add

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[6]/Sequential[m]/Bottleneck[1]/Conv[cv1]/Conv2d[conv]/input.75 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[6]/Sequential[m]/Bottleneck[1]/Conv[cv2]/Conv2d[conv]/input.79 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[6]/Sequential[m]/Bottleneck[1]/input.81 type: elemwise_add

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[6]/Sequential[m]/Bottleneck[2]/Conv[cv1]/Conv2d[conv]/input.83 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[6]/Sequential[m]/Bottleneck[2]/Conv[cv2]/Conv2d[conv]/input.87 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[6]/Sequential[m]/Bottleneck[2]/11279 type: elemwise_add

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[6]/Conv[cv2]/Conv2d[conv]/input.89 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[6]/input.91 type: concat

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[6]/Conv[cv3]/Conv2d[conv]/input.93 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Conv[model]/Conv[7]/Conv2d[conv]/input.97 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[8]/Conv[cv1]/Conv2d[conv]/input.101 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[8]/Sequential[m]/Bottleneck[0]/Conv[cv1]/Conv2d[conv]/input.105 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[8]/Sequential[m]/Bottleneck[0]/Conv[cv2]/Conv2d[conv]/input.109 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[8]/Sequential[m]/Bottleneck[0]/11404 type: elemwise_add

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[8]/Conv[cv2]/Conv2d[conv]/input.111 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[8]/input.113 type: concat

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[8]/Conv[cv3]/Conv2d[conv]/input.115 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/SPPF[model]/SPPF[9]/Conv[cv1]/Conv2d[conv]/input.119 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/SPPF[model]/SPPF[9]/MaxPool2d[m]/11481 type: maxpool

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/SPPF[model]/SPPF[9]/MaxPool2d[m]/11495 type: maxpool

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/SPPF[model]/SPPF[9]/MaxPool2d[m]/11509 type: maxpool

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/SPPF[model]/SPPF[9]/input.121 type: concat

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/SPPF[model]/SPPF[9]/Conv[cv2]/Conv2d[conv]/input.123 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Conv[model]/Conv[10]/Conv2d[conv]/input.127 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Upsample[model]/Upsample[11]/11557 type: resize

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Concat[model]/Concat[12]/input.131 type: concat

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[13]/Conv[cv1]/Conv2d[conv]/input.133 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[13]/Sequential[m]/Bottleneck[0]/Conv[cv1]/Conv2d[conv]/input.137 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[13]/Sequential[m]/Bottleneck[0]/Conv[cv2]/Conv2d[conv]/input.141 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[13]/Conv[cv2]/Conv2d[conv]/input.143 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[13]/input.145 type: concat

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[13]/Conv[cv3]/Conv2d[conv]/input.147 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Conv[model]/Conv[14]/Conv2d[conv]/input.151 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Upsample[model]/Upsample[15]/11688 type: resize

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Concat[model]/Concat[16]/input.155 type: concat

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[17]/Conv[cv1]/Conv2d[conv]/input.157 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[17]/Sequential[m]/Bottleneck[0]/Conv[cv1]/Conv2d[conv]/input.161 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[17]/Sequential[m]/Bottleneck[0]/Conv[cv2]/Conv2d[conv]/input.165 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[17]/Conv[cv2]/Conv2d[conv]/input.167 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[17]/input.169 type: concat

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[17]/Conv[cv3]/Conv2d[conv]/input.171 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Conv[model]/Conv[18]/Conv2d[conv]/input.175 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Concat[model]/Concat[19]/input.177 type: concat

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[20]/Conv[cv1]/Conv2d[conv]/input.179 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[20]/Sequential[m]/Bottleneck[0]/Conv[cv1]/Conv2d[conv]/input.183 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[20]/Sequential[m]/Bottleneck[0]/Conv[cv2]/Conv2d[conv]/input.187 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[20]/Conv[cv2]/Conv2d[conv]/input.189 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[20]/input.191 type: concat

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[20]/Conv[cv3]/Conv2d[conv]/input.193 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Conv[model]/Conv[21]/Conv2d[conv]/input.197 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Concat[model]/Concat[22]/input.199 type: concat

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[23]/Conv[cv1]/Conv2d[conv]/input.201 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[23]/Sequential[m]/Bottleneck[0]/Conv[cv1]/Conv2d[conv]/input.205 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[23]/Sequential[m]/Bottleneck[0]/Conv[cv2]/Conv2d[conv]/input.209 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[23]/Conv[cv2]/Conv2d[conv]/input.211 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[23]/input.213 type: concat

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/C3[model]/C3[23]/Conv[cv3]/Conv2d[conv]/input.215 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Detect[model]/Detect[24]/Conv2d[m]/ModuleList[0]/12065 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12837 type: elementwise_sub

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12839 type: elemwise_mul

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12841 type: elemwise_mul

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Detect[model]/Detect[24]/12200 type: elemwise_add

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12843 type: elemwise_mul

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12845 type: elemwise_mul

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Detect[model]/Detect[24]/12209 type: elemwise_mul

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Detect[model]/Detect[24]/12212 type: concat

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Detect[model]/Detect[24]/Conv2d[m]/ModuleList[1]/12238 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12849 type: elementwise_sub

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12851 type: elemwise_mul

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12853 type: elemwise_mul

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Detect[model]/Detect[24]/12372 type: elemwise_add

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12855 type: elemwise_mul

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12857 type: elemwise_mul

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Detect[model]/Detect[24]/12381 type: elemwise_mul

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Detect[model]/Detect[24]/12384 type: concat

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Detect[model]/Detect[24]/Conv2d[m]/ModuleList[2]/12410 type: conv2d

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12861 type: elementwise_sub

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12863 type: elemwise_mul

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12865 type: elemwise_mul

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Detect[model]/Detect[24]/12544 type: elemwise_add

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12867 type: elemwise_mul

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::12869 type: elemwise_mul

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Detect[model]/Detect[24]/12553 type: elemwise_mul

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Detect[model]/Detect[24]/12556 type: concat

[VAIQ_WARN]: Node ouptut tensor is not quantized: AutoShape::AutoShape/DetectMultiBackend[model]/DetectionModel[model]/Detect[model]/Detect[24]/12566 type: concat

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.0.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.0.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.1.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.1.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.2.cv1.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.2.cv1.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.2.m.0.cv1.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.2.m.0.cv1.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.2.m.0.cv2.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.2.m.0.cv2.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.2.cv2.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.2.cv2.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.2.cv3.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.2.cv3.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.3.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.3.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.4.cv1.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.4.cv1.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.4.m.0.cv1.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.4.m.0.cv1.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.4.m.0.cv2.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.4.m.0.cv2.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.4.m.1.cv1.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.4.m.1.cv1.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.4.m.1.cv2.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.4.m.1.cv2.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.4.cv2.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.4.cv2.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.4.cv3.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.4.cv3.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.5.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.5.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.6.cv1.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.6.cv1.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.6.m.0.cv1.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.6.m.0.cv1.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.6.m.0.cv2.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.6.m.0.cv2.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.6.m.1.cv1.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.6.m.1.cv1.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.6.m.1.cv2.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.6.m.1.cv2.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.6.m.2.cv1.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.6.m.2.cv1.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.6.m.2.cv2.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.6.m.2.cv2.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.6.cv2.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.6.cv2.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.6.cv3.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.6.cv3.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.7.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.7.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.8.cv1.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.8.cv1.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.8.m.0.cv1.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.8.m.0.cv1.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.8.m.0.cv2.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.8.m.0.cv2.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.8.cv2.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.8.cv2.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.8.cv3.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.8.cv3.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.9.cv1.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.9.cv1.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.9.cv2.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.9.cv2.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.10.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.10.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.13.cv1.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.13.cv1.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.13.m.0.cv1.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.13.m.0.cv1.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.13.m.0.cv2.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.13.m.0.cv2.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.13.cv2.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.13.cv2.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.13.cv3.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.13.cv3.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.14.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.14.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.17.cv1.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.17.cv1.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.17.m.0.cv1.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.17.m.0.cv1.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.17.m.0.cv2.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.17.m.0.cv2.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.17.cv2.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.17.cv2.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.17.cv3.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.17.cv3.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.18.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.18.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.20.cv1.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.20.cv1.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.20.m.0.cv1.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.20.m.0.cv1.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.20.m.0.cv2.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.20.m.0.cv2.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.20.cv2.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.20.cv2.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.20.cv3.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.20.cv3.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.21.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.21.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.23.cv1.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.23.cv1.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.23.m.0.cv1.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.23.m.0.cv1.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.23.m.0.cv2.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.23.m.0.cv2.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.23.cv2.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.23.cv2.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.23.cv3.conv.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.23.cv3.conv.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.24.m.0.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.24.m.0.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.24.m.1.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.24.m.1.bias

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.24.m.2.weight

[VAIQ_WARN]: Parameter tensor is not quantized: AutoShape::model.model.model.24.m.2.bias

[VAIQ_WARN]: Quantization is not performed completely, check if model inference function is called!!!

Exprotar XMODEL

Resultados cargados

Anyone know why this happens? How can I fix this?

jacobdbrown4 commented 3 months ago

I suggest you look at this issue about quantizing a YOLOv5 model.