KevinMusgrave
Compatibility with latest pytorch-metric-learning library
Thank you very much for your work! It is really good. But currently it's not compatible with the latest version of pytorch-metric-learning 0.9.99
When running the example script:
!python run_example.py --experiment_name test3 --dataset~OVERRIDE~ {CUB200: {download: False}}
I get:
Traceback (most recent call last): File "/storage/groups/peng/projects/aidin_Kenjis_project/power_benchmark_HCS/run_example.py", line 33, in
r.run() File "/home/icb/aidin.biibosunov/miniconda2/envs/hcs_env/lib/python3.9/site-packages/powerful_benchmarker/runners/single_experiment_runner.py", line 18, in run return self.run_new_experiment_or_resume(self.YR) File "/home/icb/aidin.biibosunov/miniconda2/envs/hcs_env/lib/python3.9/site-packages/powerful_benchmarker/runners/single_experiment_runner.py", line 33, in run_new_experiment_or_resume return self.start_experiment(args) File "/home/icb/aidin.biibosunov/miniconda2/envs/hcs_env/lib/python3.9/site-packages/powerful_benchmarker/runners/single_experiment_runner.py", line 22, in start_experiment run_output = api_parser.run() File "/home/icb/aidin.biibosunov/miniconda2/envs/hcs_env/lib/python3.9/site-packages/powerful_benchmarker/api_parsers/base_api_parser.py", line 18, in run return self.run_train_or_eval() File "/home/icb/aidin.biibosunov/miniconda2/envs/hcs_env/lib/python3.9/site-packages/powerful_benchmarker/api_parsers/base_api_parser.py", line 28, in run_train_or_eval self.run_for_each_split_scheme() File "/home/icb/aidin.biibosunov/miniconda2/envs/hcs_env/lib/python3.9/site-packages/powerful_benchmarker/api_parsers/base_api_parser.py", line 49, in run_for_each_split_scheme self.set_models_optimizers_losses() File "/home/icb/aidin.biibosunov/miniconda2/envs/hcs_env/lib/python3.9/site-packages/powerful_benchmarker/api_parsers/getter_and_setter.py", line 52, in set_models_optimizers_losses self.tester = self.get_tester() File "/home/icb/aidin.biibosunov/miniconda2/envs/hcs_env/lib/python3.9/site-packages/powerful_benchmarker/api_parsers/getter_and_setter.py", line 168, in get_tester return self.factories["tester"].create(self.args.tester, **self.all_kwargs("tester", kwargs)) File "/home/icb/aidin.biibosunov/miniconda2/envs/hcs_env/lib/python3.9/site-packages/powerful_benchmarker/factories/base_factory.py", line 31, in create return self._create_general(specs, **kwargs) File "/home/icb/aidin.biibosunov/miniconda2/envs/hcs_env/lib/python3.9/site-packages/powerful_benchmarker/factories/tester_factory.py", line 21, in _create_general return tester(**tester_params) TypeError: init() got an unexpected keyword argument 'reference_set'
While with the version pytorch-metric-learning 0.9.92 it runs as expected.
Thanks for bringing this up. Unfortunately I'm currently busy with other projects, so I won't be able to fix this very soon.
