update

openi.py

@@ -47,7 +47,7 @@ def openi_multidataset_to_env(multi_data_url, data_dir):
         print("download_multidataset_input failed")
     return 
 
-def openi_ckpt_to_env(obs_ckpt_url, ckpt_url):
+def uploadfile_to_obs(obs_ckpt_url, ckpt_url):
     """
     copy ckpt file to training image or copy ckpt file to obs
     """
@@ -80,13 +80,13 @@ def env_to_openi(train_dir, train_url):
     device_num = int(os.getenv('RANK_SIZE'))
     local_rank=int(os.getenv('RANK_ID'))
     if device_num == 1:
-        upload_to_obs(train_dir, train_url)
+        uploadfolder_to_obs(train_dir, train_url)
     if device_num > 1:
         if local_rank%8==0:
-            upload_to_obs(train_dir, train_url)
+            uploadfolder_to_obs(train_dir, train_url)
     return
 
-def upload_to_obs(train_dir, obs_train_url):
+def uploadfolder_to_obs(train_dir, obs_train_url):
     """
     upload to obs
     """
@@ -136,4 +136,4 @@ class EnvToOpenIEpochEnd(Callback):
         self.train_dir = train_dir
         self.obs_train_url = obs_train_url
     def epoch_end(self,run_context):
-        upload_to_obs(self.train_dir,self.obs_train_url)      
+        uploadfolder_to_obs(self.train_dir,self.obs_train_url)      

train_continue.py

@@ -21,14 +21,12 @@ from mindspore import load_checkpoint, load_param_into_net
 from mindspore.train import Model
 from mindspore.nn.metrics import Accuracy
 from mindspore.communication.management import get_rank
-import mindspore.ops as ops
-import time
 
-from openi import openi_ckpt_to_env
+from openi import uploadfile_to_obs
 from openi import oponi_dataset_to_Env
 
 parser = argparse.ArgumentParser(description='MindSpore Lenet Example')
-parser.add_argument('--data_url',
+parser.add_argument('--multi_data_url',
                     help='path to training/inference dataset folder',
                     default= '/cache/data/')
 
@@ -71,13 +69,13 @@ if __name__ == "__main__":
     except Exception as e:
         print("path already exists")
 
-    oponi_dataset_to_Env(args.data_url, data_dir)
+    oponi_dataset_to_Env(args.multi_data_url, data_dir)
 
     device_num = int(os.getenv('RANK_SIZE'))
     if device_num == 1:
-        ds_train = create_dataset(os.path.join(data_dir, "train"),  cfg.batch_size)
+        ds_train = create_dataset(os.path.join(data_dir + "/MNISTData", "train"),  cfg.batch_size)
     if device_num > 1:
-        ds_train = create_dataset_parallel(os.path.join(data_dir, "train"),  cfg.batch_size)
+        ds_train = create_dataset_parallel(os.path.join(data_dir + "/MNISTData", "train"),  cfg.batch_size)
     if ds_train.get_dataset_size() == 0:
         raise ValueError("Please check dataset size > 0 and batch_size <= dataset size")
 
@@ -136,4 +134,4 @@ if __name__ == "__main__":
     for n in new_models:
         ckpt_url = base_path + "/" + n
         obs_ckpt_url = args.train_url + "/" + n
-        openi_ckpt_to_env(ckpt_url, obs_ckpt_url)
+        uploadfile_to_obs(ckpt_url, obs_ckpt_url)