猫狗识别模型
-
转换模型失败,有大佬可以看看吗,求帮助
m5stack中Failed to convert model (ckpt -> onnx)details:
[TASK] 2023-06-26 09:55:56,515 - [line:323] - INFO: Task e170ea835ed77e9a started, user email: 1731323217@qq.com
[TASK] 2023-06-26 09:55:56,515 - [line:288] - INFO: Downloading 695c696793837fb3054a520b180584eb.zip
[TASK] 2023-06-26 09:55:56,772 - [line:375] - INFO: Preparing dataset.
[TASK] 2023-06-26 09:55:56,865 - [line:266] - INFO: num of image: [total: 38, discard: 0, rest: 38]
[TASK] 2023-06-26 09:55:57,256 - [line:440] - INFO: Split dataset: Saved 36 entries in /root/NanoDet/dataset/train_ann.json and 4 in /root/NanoDet/dataset/val_ann.json[TASK] 2023-06-26 09:55:57,282 - [line:480] - INFO: Start nanodet_m training.
[TASK] 2023-06-26 09:55:57,293 - [line:472] - INFO: Training timeout watcher started.
[TASK] 2023-06-26 09:56:01,418 - [line:576] - INFO: Converting model.
[TASK] 2023-06-26 09:56:03,285 - [line:685] - ERROR: Failed to convert model (ckpt -> onnx)
Traceback (most recent call last):
File "framework.py", line 592, in convertModel
out = subprocess.check_output(['python3', converter_path, '--cfg_path', DATA['train_config_path'],
File "/opt/conda/lib/python3.8/subprocess.py", line 415, in check_output
return run(*popenargs, stdout=PIPE, timeout=timeout, check=True,
File "/opt/conda/lib/python3.8/subprocess.py", line 516, in run
raise CalledProcessError(retcode, process.args,
subprocess.CalledProcessError: Command '['python3', '/root/NanoDet/nanodet/tools/export_onnx.py', '--cfg_path', '/root/NanoDet/dataset/nanodet-m-config.yml', '--model_path', '/root/NanoDet/output/model_best/model_best.ckpt', '--out_path', '/root/NanoDet/output/nanodet.onnx']' returned non-zero exit status 1.During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "framework.py", line 679, in main
convertModel()
File "framework.py", line 596, in convertModel
raise Exception('Failed to convert model (ckpt -> onnx)')
Exception: Failed to convert model (ckpt -> onnx)
Hello! It looks like you're interested in this conversation, but you don't have an account yet.
Getting fed up of having to scroll through the same posts each visit? When you register for an account, you'll always come back to exactly where you were before, and choose to be notified of new replies (either via email, or push notification). You'll also be able to save bookmarks and upvote posts to show your appreciation to other community members.
With your input, this post could be even better 💗
Register Login