Were you ablle to convert custom yolov5 and yolov8 models to rknn and build inference using the python api?
I am running the steps for conversion and it is working with the default models given in the rk libraries. While converting my custom models from onnx to rknn, I am getting version mismatch errors between the toolkit libraries and the onnx and torch libraries used to build my model.
On updating the libraries to the compatible ones, my toolkit convert code iss failing with
/home/dt/Projects/Work/greenox/rknn/venv/lib/python3.8/site-packages/rknn/api/rknn.py(57)init()
-> self.rknn_base = RKNNBase(cur_path, verbose)
(Pdb) c
File "../../../../../common/rknn_converter/rknn_convert.py", line 19, in convert
rknn = cp.convert()
File "/home/dt/Projects/Work/greenox/rknn/rknn_model_zoo/common/rknn_converter/phase.py", line 96, in convert
rknn = RKNN(verbose=model_config_dict['verbose'], verbose_file=model_config_dict['verbose_file'])
File "/home/dt/Projects/Work/greenox/rknn/venv/lib/python3.8/site-packages/rknn/api/rknn.py", line 57, in init
self.rknn_base = RKNNBase(cur_path, verbose)
File "rknn/api/rknn_base.py", line 75, in rknn.api.rknn_base.RKNNBase.init
File "/home/dt/Projects/Work/greenox/rknn/venv/lib/python3.8/site-packages/pkg_resources/init.py", line 526, in get_distribution
dist = get_provider(dist)
File "/home/dt/Projects/Work/greenox/rknn/venv/lib/python3.8/site-packages/pkg_resources/init.py", line 398, in get_provider
return working_set.find(moduleOrReq) or require(str(moduleOrReq))[0]
File "/home/dt/Projects/Work/greenox/rknn/venv/lib/python3.8/site-packages/pkg_resources/init.py", line 695, in find
if dist is not None and dist not in req:
File "/home/dt/Projects/Work/greenox/rknn/venv/lib/python3.8/site-packages/pkg_resources/init.py", line 3204, in contains
return self.specifier.contains(item, prereleases=True)
File "/home/dt/Projects/Work/greenox/rknn/venv/lib/python3.8/site-packages/pkg_resources/_vendor/packaging/specifiers.py", line 902, in contains
item = Version(item)
File "/home/dt/Projects/Work/greenox/rknn/venv/lib/python3.8/site-packages/pkg_resources/_vendor/packaging/version.py", line 197, in init
raise InvalidVersion(f"Invalid version: '{version}'")
pkg_resources.extern.packaging.version.InvalidVersion: Invalid version: '1.4.0-22dcfef4'
Do you have a requirements file and the versions of yolov5,v7 and v8 which the rknntoolit2 is compatible with?
Were you ablle to convert custom yolov5 and yolov8 models to rknn and build inference using the python api?
I am running the steps for conversion and it is working with the default models given in the rk libraries. While converting my custom models from onnx to rknn, I am getting version mismatch errors between the toolkit libraries and the onnx and torch libraries used to build my model.
On updating the libraries to the compatible ones, my toolkit convert code iss failing with
export_pre_compile_path: ./model_cvt/RK3588/yolov5s.torchscript_RK3588_i8_precompile.rknn ---> Create RKNN object