The GPU architecture in your current machine is Pascal, which is not compatible with Paddle installation with arch: 70 75 80 86

[ynes99]

问题确认 Search before asking

• 我已经搜索过问题，但是没有找到解答。I have searched the question and found no related answer.

请提出你的问题 Please ask your question

Hello so i checked another issue with the same warning and the installation there had no problem, but mine had problems.

I did this test
!sh tests/install/check_predict.sh
and all I got was blurry and pixelated pictures instead of what was supposed to be predicted.

output/result/added_prediction

output/result/pseudo_color_prediction

the cuda version i'm using is 11.7 the installation is correct i checked with nvidia-smi and compiled some cuda code to be sure.
the only warning that might cause the error is this one but I installed the correct version and everything so I'm stuck on what to do

The GPU architecture in your current machine is Pascal, which is not compatible with Paddle installation with arch: 70 75 80 86

the CPU installation works just fine but the GPU doesn't for some reason.

2023-08-11 13:57:19 [WARNING] Add the num_classes in train_dataset and val_dataset config to model config. We suggest you manually set num_classes in model config.
2023-08-11 13:57:19 [INFO]

------------Environment Information-------------
platform: Linux-5.15.0-78-generic-x86_64-with-glibc2.35
Python: 3.10.12 (main, Jun 11 2023, 05:26:28) [GCC 11.4.0]
Paddle compiled with cuda: True
NVCC: Build cuda_11.7.r11.7/compiler.31442593_0
cudnn: 8.9
GPUs used: 1
CUDA_VISIBLE_DEVICES: None
GPU: ['GPU 0: NVIDIA TITAN']
GCC: gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0
PaddleSeg: 2.8.0
PaddlePaddle: 2.5.1
OpenCV: 4.5.5

2023-08-11 13:57:19 [INFO]
---------------Config Information---------------
batch_size: 4
iters: 1000
train_dataset:
dataset_root: data/optic_disc_seg
mode: train
num_classes: 2
train_path: data/optic_disc_seg/train_list.txt
transforms:

• max_scale_factor: 2.0
  min_scale_factor: 0.5
  scale_step_size: 0.25
  type: ResizeStepScaling
• crop_size:
  • 512
  • 512
    type: RandomPaddingCrop
• type: RandomHorizontalFlip
• brightness_range: 0.5
  contrast_range: 0.5
  saturation_range: 0.5
  type: RandomDistort
• type: Normalize
  type: Dataset
  val_dataset:
  dataset_root: data/optic_disc_seg
  mode: val
  num_classes: 2
  transforms:
• type: Normalize
  type: Dataset
  val_path: data/optic_disc_seg/val_list.txt
  optimizer:
  momentum: 0.9
  type: SGD
  weight_decay: 4.0e-05
  lr_scheduler:
  end_lr: 0
  learning_rate: 0.01
  power: 0.9
  type: PolynomialDecay
  loss:
  coef:
• 1
• 1
• 1
  types:
• type: CrossEntropyLoss
• type: CrossEntropyLoss
• type: CrossEntropyLoss
  model:
  backbone:
  pretrained: https://bj.bcebos.com/paddleseg/dygraph/PP_STDCNet2.tar.gz
  type: STDC2
  num_classes: 2
  type: PPLiteSeg

---

2023-08-11 13:57:19 [INFO] Set device: gpu
2023-08-11 13:57:19 [INFO] Use the following config to build model
model:
backbone:
pretrained: https://bj.bcebos.com/paddleseg/dygraph/PP_STDCNet2.tar.gz
type: STDC2
num_classes: 2
type: PPLiteSeg
W0811 13:57:19.390579 55924 gpu_resources.cc:96] The GPU architecture in your current machine is Pascal, which is not compatible with Paddle installation with arch: 70 75 80 86 , it is recommended to install the corresponding wheel package according to the installation information on the official Paddle website.
W0811 13:57:19.390600 55924 gpu_resources.cc:119] Please NOTE: device: 0, GPU Compute Capability: 6.1, Driver API Version: 12.2, Runtime API Version: 11.7
W0811 13:57:19.390609 55924 gpu_resources.cc:149] device: 0, cuDNN Version: 8.9.
2023-08-11 13:57:19 [INFO] Loading pretrained model from https://bj.bcebos.com/paddleseg/dygraph/PP_STDCNet2.tar.gz
2023-08-11 13:57:20 [INFO] There are 265/265 variables loaded into STDCNet.
2023-08-11 13:57:20 [INFO] The number of images: 1
2023-08-11 13:57:20 [INFO] Loading pretrained model from https://paddleseg.bj.bcebos.com/dygraph/optic_disc/pp_liteseg_optic_disc_512x512_1k/model.pdparams
2023-08-11 13:57:20 [WARNING] ppseg_head.arm_list.0._scale is not in pretrained model
2023-08-11 13:57:20 [WARNING] ppseg_head.arm_list.1._scale is not in pretrained model
2023-08-11 13:57:20 [WARNING] ppseg_head.arm_list.2._scale is not in pretrained model
2023-08-11 13:57:20 [INFO] There are 367/370 variables loaded into PPLiteSeg.
2023-08-11 13:57:20 [INFO] Start to predict...
1/1 [==============================] - 1s 562ms/step
2023-08-11 13:57:21 [INFO] Predicted images are saved in output/result/added_prediction and output/result/pseudo_color_prediction .

[Asthestarsfalll]

Plz report this to Paddle repo.
It seems that you have installed a version of CUDA that is too high.

[ToddBear]

以上回答已经充分解答了问题，如果有新的问题欢迎随时提交issue，或者在此条issue下继续回复～
我们开启了飞桨套件的ISSUE攻关活动，欢迎感兴趣的开发者参加：PaddlePaddle/PaddleOCR#10223