Open vnoblet-nanovare opened 2 years ago
After fixing this part, I have another error :
Traceback (most recent call last): File "train.py", line 622, in <module> train(hyp, opt, device, tb_writer) File "train.py", line 72, in train wandb_logger = WandbLogger(opt, Path(opt.save_dir).stem, run_id, data_dict) File "/mnt/batch/tasks/shared/LS_root/jobs/workspace_test/azureml/coral_panda_cc10fxj3z2/wd/azureml/coral_panda_cc10fxj3z2/utils/wandb_logging/wandb_utils.py", line 107, in __init__ self.wandb_run.config.data_dict = wandb_data_dict File "/azureml-envs/pytorch-1.9/lib/python3.7/site-packages/wandb/sdk/wandb_config.py", line 147, in __setitem__ self._raise_value_error_on_nested_artifact(val, nested=True) File "/azureml-envs/pytorch-1.9/lib/python3.7/site-packages/wandb/sdk/wandb_config.py", line 266, in _raise_value_error_on_nested_artifact "Instances of wandb.Artifact and wandb.apis.public.Artifact" ValueError: Instances of wandb.Artifact and wandb.apis.public.Artifact can only be top level keys in wandb.config wandb: Waiting for W&B process to finish... (failed 1). Press Control-C to abort syncing.
Maybe consider replacing:
nc, names = (1, ['item']) if single_cls else (int(data['nc']), data['names'])
by:nc, names = int(data['nc']), data['names']
in order to have the dataset tables conserve the original classOr else consider fixing the followinf code