Merge pull request #5719 from Masao-Someki/easy_to_ez

Modified easy to ez

espnetez/config.py

@@ -1,6 +1,6 @@
 import yaml
 
-from espnetez.task import get_easy_task
+from espnetez.task import get_ez_task
 
 
 def convert_none_to_None(dic):
@@ -14,7 +14,7 @@ def convert_none_to_None(dic):
 
 
 def from_yaml(task, path):
-    task_class = get_easy_task(task)
+    task_class = get_ez_task(task)
     with open(path, "r") as f:
         config = yaml.load(f, Loader=yaml.Loader)
 
@@ -30,7 +30,7 @@ def from_yaml(task, path):
 def update_finetune_config(task, pretrain_config, path):
     with open(path, "r") as f:
         finetune_config = yaml.load(f, Loader=yaml.Loader)
-    default_config = get_easy_task(task).get_default_config()
+    default_config = get_ez_task(task).get_default_config()
 
     # update pretrain_config with finetune_config
     # and update distributed related configs to the default.

espnetez/dataset.py

@@ -3,7 +3,7 @@
 from espnet2.train.dataset import AbsDataset
 
 
-class ESPnetEasyDataset(AbsDataset):
+class ESPnetEZDataset(AbsDataset):
     def __init__(self, dataset, data_info):
         self.dataset = dataset
         self.data_info = data_info

espnetez/task.py

@@ -1,4 +1,4 @@
-# ESPnet-Easy Task class
+# ESPnet-EZ Task class
 # This class is a wrapper for Task classes to support custom datasets.
 import argparse
 import logging
@@ -64,13 +64,13 @@
 )
 
 
-def get_easy_task(task_name: str, use_custom_dataset: bool = False) -> AbsTask:
+def get_ez_task(task_name: str, use_custom_dataset: bool = False) -> AbsTask:
     task_class = TASK_CLASSES[task_name]
 
     if use_custom_dataset:
-        return get_easy_task_with_dataset(task_name)
+        return get_ez_task_with_dataset(task_name)
 
-    class ESPnetEasyTask(task_class):
+    class ESPnetEZTask(task_class):
         build_model_fn = None
 
         @classmethod
@@ -80,13 +80,13 @@ def build_model(cls, args=None):
             else:
                 return task_class.build_model(args=args)
 
-    return ESPnetEasyTask
+    return ESPnetEZTask
 
 
-def get_easy_task_with_dataset(task_name: str) -> AbsTask:
+def get_ez_task_with_dataset(task_name: str) -> AbsTask:
     task_class = TASK_CLASSES[task_name]
 
-    class ESPnetEasyDataTask(task_class):
+    class ESPnetEZDataTask(task_class):
         build_model_fn = None
         train_dataset = None
         valid_dataset = None
@@ -317,4 +317,4 @@ def build_streaming_iterator(
                 **kwargs,
             )
 
-    return ESPnetEasyDataTask
+    return ESPnetEZDataTask

espnetez/trainer.py

@@ -2,7 +2,7 @@
 import os
 from argparse import Namespace
 
-from espnetez.task import get_easy_task
+from espnetez.task import get_ez_task
 
 
 def check_argument(
@@ -111,18 +111,18 @@ def __init__(
             )
 
         if train_dataset is not None and valid_dataset is not None:
-            self.task_class = get_easy_task(task, use_custom_dataset=True)
+            self.task_class = get_ez_task(task, use_custom_dataset=True)
             self.task_class.train_dataset = train_dataset
             self.task_class.valid_dataset = valid_dataset
         elif train_dataloader is not None and valid_dataloader is not None:
-            self.task_class = get_easy_task(task, use_custom_dataset=True)
+            self.task_class = get_ez_task(task, use_custom_dataset=True)
             self.task_class.train_dataloader = train_dataloader
             self.task_class.valid_dataloader = valid_dataloader
         else:
             assert data_info is not None, "data_info should be provided."
             assert train_dump_dir is not None, "Please provide train_dump_dir."
             assert valid_dump_dir is not None, "Please provide valid_dump_dir."
-            self.task_class = get_easy_task(task)
+            self.task_class = get_ez_task(task)
             train_dpnt = []
             valid_dpnt = []
             for k, v in data_info.items():

test/espnetez/test_ez.py

@@ -134,13 +134,13 @@ def test_join_dumps():
 
 @pytest.mark.parametrize("task_name,task_class", TASK_CLASSES)
 def test_task(task_name, task_class):
-    task = ez.task.get_easy_task(task_name)
+    task = ez.task.get_ez_task(task_name)
     assert issubclass(task, task_class)
 
 
 @pytest.mark.parametrize("task_name,task_class", TASK_CLASSES)
 def test_task_with_dataset(task_name, task_class):
-    task = ez.task.get_easy_task_with_dataset(task_name)
+    task = ez.task.get_ez_task_with_dataset(task_name)
     assert issubclass(task, task_class)
 
 
@@ -175,10 +175,10 @@ def test_load_config(task_name, task_class):
         config_path = Path(temp_dir) / "config.yaml"
         config_path.write_text("""task: {task_name}""")
         default_config = task_class.get_default_config()
-        easy_config = ez.config.from_yaml(task_name, config_path)
+        ez_config = ez.config.from_yaml(task_name, config_path)
 
         for k in default_config.keys():
-            assert default_config[k] == easy_config[k]
+            assert default_config[k] == ez_config[k]
 
 
 @pytest.mark.parametrize("task_name,task_class", TASK_CLASSES)
@@ -188,11 +188,11 @@ def test_update_finetune_config(task_name, task_class):
         config_path = Path(temp_dir) / "config.yaml"
         config_path.write_text("""use_lora: true""")
         pretrain_config = task_class.get_default_config()
-        easy_config = ez.config.update_finetune_config(
+        ez_config = ez.config.update_finetune_config(
             task_name, pretrain_config, config_path
         )
 
-        for k, v in easy_config.items():
+        for k, v in ez_config.items():
             if k != "use_lora":
                 assert v == pretrain_config[k]
             else: