现在单用onnx推理有结果，用ultralytics推理没有结果
wxsdiscover
- Members
- Joined 2026年2月26日
-
并且直接使用model.export(format='onnx'),也不行，你们提供的镜像里是
root@367d9cb74f5f:/predict# pip list | grep 'onnx'
onnx 1.12.0
onnxoptimizer 0.3.13
onnxruntime-gpu 1.12.0+maca3.1.0.5
onnxsim 0.4.36在调用model.export(format='onnx')，ultralytics 框架会校验onnxruntime和onnxslim，会从后台下载相关依赖，覆盖掉你们的onnx框架，你们有没有解决方法
-
我之前的demo onnx的脚本不对，我的基准模型是yolo26s，我按照v8去解析的，我重新编写了demo onnx的脚本，是可以直接运行onnx的模型，
root@367d9cb74f5f:/predict# python onnx_demo.py
✅ 模型加载成功 (E2E 格式)
📦 配置类别数: 1
📷 图片尺寸 (原图): 1918 x 1000 | 缩放比例: 0.3337
🔍 检测到 2 个有效目标:
类别 置信度 像素坐标 (x1, y1, x2, y2)
drone 0.9307 [698, 11, 1883, 504]
drone 0.9175 [91, 590, 857, 912]
但是使用ultralytics 框架 仍然不行
-
[root@standalone-amd64-mxc500 ~]# mx-smi
mx-smi version: 2.2.9=================== MetaX System Management Interface Log ===================
Timestamp : Wed Mar 25 11:39:24 2026Attached GPUs : 16
+---------------------------------------------------------------------------------+
| MX-SMI 2.2.9 Kernel Mode Driver Version: 3.3.12 |
| MACA Version: 3.2.1.10 BIOS Version: 1.29.1.0 |
|------------------+-----------------+---------------------+----------------------|
| Board Name | GPU Persist-M | Bus-id | GPU-Util sGPU-M |
| Pwr:Usage/Cap | Temp Perf | Memory-Usage | GPU-State |
|==================+=================+=====================+======================|
| 0 MetaX C500 | 0 Off | 0000:05:00.0 | 0% Disabled |
| 68W / 350W | 49C P0 | 859/65536 MiB | Available |
+------------------+-----------------+---------------------+----------------------+
| 1 MetaX C500 | 1 Off | 0000:06:00.0 | 0% Disabled |
| 62W / 350W | 48C P0 | 859/65536 MiB | Available |
+------------------+-----------------+---------------------+----------------------+
| 2 MetaX C500 | 2 Off | 0000:09:00.0 | 0% Disabled |
| 59W / 350W | 48C P0 | 859/65536 MiB | Available |
+------------------+-----------------+---------------------+----------------------+
| 3 MetaX C500 | 3 Off | 0000:0f:00.0 | 0% Disabled |
| 56W / 350W | 46C P0 | 859/65536 MiB | Available |
+------------------+-----------------+---------------------+----------------------+
| 4 MetaX C500 | 4 Off | 0000:27:00.0 | 0% Disabled |
| 60W / 350W | 49C P0 | 859/65536 MiB | Available |
+------------------+-----------------+---------------------+----------------------+
| 5 MetaX C500 | 5 Off | 0000:28:00.0 | 0% Disabled |
| 64W / 350W | 51C P0 | 859/65536 MiB | Available |
+------------------+-----------------+---------------------+----------------------+
| 6 MetaX C500 | 6 Off | 0000:2b:00.0 | 0% Disabled |
| 59W / 350W | 50C P0 | 859/65536 MiB | Available |
+------------------+-----------------+---------------------+----------------------+
| 7 MetaX C500 | 7 Off | 0000:31:00.0 | 0% Disabled |
| 63W / 350W | 52C P0 | 859/65536 MiB | Available |
+------------------+-----------------+---------------------+----------------------+
| 8 MetaX C500 | 8 Off | 0000:85:00.0 | 0% Disabled |
| 60W / 350W | 48C P0 | 859/65536 MiB | Available |
+------------------+-----------------+---------------------+----------------------+
| 9 MetaX C500 | 9 Off | 0000:86:00.0 | 0% Disabled |
| 62W / 350W | 48C P0 | 859/65536 MiB | Available |
+------------------+-----------------+---------------------+----------------------+
| 10 MetaX C500 | 10 Off | 0000:89:00.0 | 0% Disabled |
| 62W / 350W | 47C P0 | 859/65536 MiB | Available |
+------------------+-----------------+---------------------+----------------------+
| 11 MetaX C500 | 11 Off | 0000:8f:00.0 | 0% Disabled |
| 59W / 350W | 47C P0 | 859/65536 MiB | Available |
+------------------+-----------------+---------------------+----------------------+
| 12 MetaX C500 | 12 Off | 0000:a5:00.0 | 0% Disabled |
| 64W / 350W | 52C P0 | 859/65536 MiB | Available |
+------------------+-----------------+---------------------+----------------------+
| 13 MetaX C500 | 13 Off | 0000:a6:00.0 | 0% Disabled |
| 65W / 350W | 51C P0 | 859/65536 MiB | Available |
+------------------+-----------------+---------------------+----------------------+
| 14 MetaX C500 | 14 Off | 0000:a9:00.0 | 0% Disabled |
| 63W / 350W | 51C P0 | 859/65536 MiB | Available |
+------------------+-----------------+---------------------+----------------------+
| 15 MetaX C500 | 15 Off | 0000:af:00.0 | 0% Disabled |
| 62W / 350W | 51C P0 | 859/65536 MiB | Available |
+------------------+-----------------+---------------------+----------------------++---------------------------------------------------------------------------------+
| Process: |
| GPU PID Process Name GPU Memory |
| Usage(MiB) |
|=================================================================================|
| no process found |
+---------------------------------------------------------------------------------+End of Log
[root@standalone-amd64-mxc500 ~]# -
试过了，一样的没有结果，这个onnx是我从yolo.export出来的，没有加nms
-
代码启动，直接使用YOLO读取onnx模型文件，然后识别， self.identify_model = YOLO(
identify_model_path,
task='detect'
) detect_results = self.identify_model(
source=image_path,
conf=self.conf_threshold,
iou=self.iou_threshold,
device=self.gpu_device,
verbose=False
) -
root@a6fee036f708:/predict# pip show ultralytics
Name: ultralytics
Version: 8.4.26
Summary: Ultralytics YOLO 🚀 for SOTA object detection, multi-object tracking, instance segmentation, pose estimation and image classification.
Home-page: ultralytics.com
Author:
Author-email: Glenn Jocher glenn.jocher@ultralytics.com, Jing Qiu jing.qiu@ultralytics.com
License: AGPL-3.0
Location: /opt/conda/lib/python3.10/site-packages
Requires: matplotlib, numpy, opencv-python, pillow, polars, psutil, pyyaml, requests, scipy, torch, torchvision, ultralytics-thop
Required-by:
root@a6fee036f708:/predict# cd /opt/conda/lib/python3.10/site-packages/ultralytics/nn/backends/
root@a6fee036f708:/opt/conda/lib/python3.10/site-packages/ultralytics/nn/backends# vim
root@a6fee036f708:/opt/conda/lib/python3.10/site-packages/ultralytics/nn/backends# vim onnx.py添加macaexecutionprovider
-
使用官方镜像
cr.metax-tech.com/public-ai-release/maca/onnxruntime maca.ai3.1.0.5-torch2.4-py310-ubuntu22.04-amd64 2fac80e009f7 6 months ago 28.4GB安装ultralytics,修改了源码使其在调用onnx格式yolo模型，
2026-03-24 19:08:43,090 - root - INFO - -----准备加载目标识别模型：/file_manage/models/train_weights/49b34b4427214d2bba0dd27dd7b01490.onnx
2026-03-24 19:08:43,093 - root - INFO - -----加载目标识别模型成功
Loading /file_manage/models/train_weights/49b34b4427214d2bba0dd27dd7b01490.onnx for ONNX Runtime inference...
Using ONNX Runtime 1.12.0+git-a811f5a3 with MACAExecutionProvider使用CPU可以推理结果，同环境下-使用GPU推理的推理不出来
2026-03-24 19:08:43,090 - root - INFO - -----准备加载目标识别模型：/file_manage/models/train_weights/49b34b4427214d2bba0dd27dd7b01490.onnx
2026-03-24 19:08:43,093 - root - INFO - -----加载目标识别模型成功
Loading /file_manage/models/train_weights/49b34b4427214d2bba0dd27dd7b01490.onnx for ONNX Runtime inference...
Using ONNX Runtime 1.12.0+git-a811f5a3 with MACAExecutionProvider
2026-03-24 19:08:55,020 - root - INFO - 第0张图片：未检测到任何目标
2026-03-24 19:08:55,048 - werkzeug - INFO - 172.18.0.5 - - [24/Mar/2026 19:08:55] "POST /api/predict HTTP/1.1" 200 -使用5060也能推理出来
配置如下
root@a6fee036f708:/predict# mx-smi
mx-smi version: 2.2.8=================== MetaX System Management Interface Log ===================
Timestamp : Tue Mar 24 19:13:48 2026Attached GPUs : 16
+---------------------------------------------------------------------------------+
| MX-SMI 2.2.8 Kernel Mode Driver Version: 3.3.12 |
| MACA Version: 3.1.0.14 BIOS Version: 1.29.1.0 |
|------------------------------------+---------------------+----------------------+
| GPU NAME Persistence-M | Bus-id | GPU-Util sGPU-M |
| Temp Pwr:Usage/Cap Perf | Memory-Usage | GPU-State |
|====================================+=====================+======================|
| 0 MetaX C500 Off | 0000:05:00.0 | 0% Native |
| 47C 67W / 350W P0 | 858/65536 MiB | Available |
+------------------------------------+---------------------+----------------------+
| 1 MetaX C500 Off | 0000:06:00.0 | 0% Native |
| 47C 62W / 350W P0 | 858/65536 MiB | Available |
+------------------------------------+---------------------+----------------------+
| 2 MetaX C500 Off | 0000:09:00.0 | 0% Native |
| 47C 58W / 350W P0 | 858/65536 MiB | Available |
+------------------------------------+---------------------+----------------------+
| 3 MetaX C500 Off | 0000:0f:00.0 | 0% Native |
| 45C 56W / 350W P0 | 858/65536 MiB | Available |
+------------------------------------+---------------------+----------------------+
| 4 MetaX C500 Off | 0000:27:00.0 | 0% Native |
| 48C 60W / 350W P0 | 858/65536 MiB | Available |
+------------------------------------+---------------------+----------------------+
| 5 MetaX C500 Off | 0000:28:00.0 | 0% Native |
| 50C 64W / 350W P0 | 858/65536 MiB | Available |
+------------------------------------+---------------------+----------------------+
| 6 MetaX C500 Off | 0000:2b:00.0 | 0% Native |
| 49C 58W / 350W P0 | 858/65536 MiB | Available |
+------------------------------------+---------------------+----------------------+
| 7 MetaX C500 Off | 0000:31:00.0 | 0% Native |
| 51C 62W / 350W P0 | 858/65536 MiB | Available |
+------------------------------------+---------------------+----------------------+
| 8 MetaX C500 Off | 0000:85:00.0 | 0% Native |
| 47C 59W / 350W P0 | 858/65536 MiB | Available |
+------------------------------------+---------------------+----------------------+
| 9 MetaX C500 Off | 0000:86:00.0 | 0% Native |
| 47C 62W / 350W P0 | 858/65536 MiB | Available |
+------------------------------------+---------------------+----------------------+
| 10 MetaX C500 Off | 0000:89:00.0 | 0% Native |
| 47C 62W / 350W P0 | 858/65536 MiB | Available |
+------------------------------------+---------------------+----------------------+
| 11 MetaX C500 Off | 0000:8f:00.0 | 0% Native |
| 46C 58W / 350W P0 | 858/65536 MiB | Available |
+------------------------------------+---------------------+----------------------+
| 12 MetaX C500 Off | 0000:a5:00.0 | 0% Native |
| 50C 64W / 350W P0 | 858/65536 MiB | Available |
+------------------------------------+---------------------+----------------------+
| 13 MetaX C500 Off | 0000:a6:00.0 | 0% Native |
| 50C 65W / 350W P0 | 858/65536 MiB | Available |
+------------------------------------+---------------------+----------------------+
| 14 MetaX C500 Off | 0000:a9:00.0 | 0% Native |
| 49C 63W / 350W P0 | 858/65536 MiB | Available |
+------------------------------------+---------------------+----------------------+
| 15 MetaX C500 Off | 0000:af:00.0 | 0% Native |
| 50C 62W / 350W P0 | 858/65536 MiB | Available |
+------------------------------------+---------------------+----------------------++---------------------------------------------------------------------------------+
| Process: |
| GPU PID Process Name GPU Memory |
| Usage(MiB) |
|=================================================================================|
| no process found |
+---------------------------------------------------------------------------------+End of Log
-
-
-
这里我只按装了torch相关的3个包，其他的也有安装吗，-rw-r--r-- 1 1002 1002 13619423 10月 15 00:01 apex-0.1+metax3.2.1.3-cp312-cp312-linux_x86_64.whl
-rw-r--r-- 1 1002 1002 11850047 10月 15 01:08 causal_conv1d-1.5.0.post8+metax3.2.1.3torch2.4-cp312-cp312-linux_x86_64.whl
-rw-r--r-- 1 1002 1002 415231211 10月 15 00:55 dropout_layer_norm-0.1+metax3.2.1.3torch2.4-cp312-cp312-linux_x86_64.whl
-rw-r--r-- 1 1002 1002 1292684192 10月 15 00:35 flash_attn-2.6.3+metax3.2.1.3torch2.4-cp312-cp312-linux_x86_64.whl
-rw-r--r-- 1 1002 1002 15244749 10月 15 00:59 flashinfer-0.2.6+metax3.2.1.3torch2.4-cp312-cp312-linux_x86_64.whl
-rw-r--r-- 1 1002 1002 13701829 10月 15 00:40 flash_mla-1.0.1+metax3.2.1.3torch2.4-cp312-cp312-linux_x86_64.whl
-rw-r--r-- 1 1002 1002 90097 10月 15 00:44 fused_dense_lib-2.6.3+metax3.2.1.3torch2.4-cp312-cp312-linux_x86_64.whl
-rw-r--r-- 1 1002 1002 11925324 10月 15 01:04 mamba_ssm-2.2.4+metax3.2.1.3torch2.4-cp312-cp312-linux_x86_64.whl
-rw-r--r-- 1 1002 1002 115219 10月 15 01:09 mctlassEx-0.1.1+metax3.2.1.3torch2.4-cp312-cp312-linux_x86_64.whl
-rw-r--r-- 1 1002 1002 4275579 10月 15 00:56 rotary_emb-0.1+metax3.2.1.3torch2.4-cp312-cp312-linux_x86_64.whl
-rw-r--r-- 1 1002 1002 1066199 10月 15 00:43 sageattention-2.0.1+metax3.2.1.3torch2.4-cp312-cp312-linux_x86_64.whl
-rw-r--r-- 1 1002 1002 2847558 10月 15 01:33 spconv-2.1.0+metax3.2.1.3torch2.4-cp312-cp312-linux_x86_64.whl
-rw-r--r-- 1 1002 1002 645582726 10月 14 23:24 torch-2.4.0+metax3.2.1.3-cp312-cp312-linux_x86_64.whl
-rw-r--r-- 1 1002 1002 1987325 10月 15 00:03 torchaudio-2.4.1+metax3.2.1.3-cp312-cp312-linux_x86_64.whl
-rw-r--r-- 1 1002 1002 1816945 10月 14 23:47 torchvision-0.15.1+metax3.2.1.3-cp312-cp312-linux_x86_64.whl
-rw-r--r-- 1 1002 1002 239538789 10月 14 23:46 triton-3.0.0+metax3.2.1.3-cp312-cp312-linux_x86_64.whl
-rw-r--r-- 1 1002 1002 150442 10月 15 00:57 xentropy_cuda_lib-0.1+metax3.2.1.3torch2.4-cp312-cp312-linux_x86_64.whl
-rw-r--r-- 1 1002 1002 601895420 10月 15 01:31 xformers-0.0.22+metax3.2.1.3torch2.4-cp312-cp312-linux_x86_64.whl -
-