format code

llama_lora/lib/finetune.py

@@ -87,6 +87,7 @@ def train(
     #     os.environ["WANDB_PROJECT"] = wandb_project
     # if wandb_run_name:
     #     os.environ["WANDB_RUN_NAME"] = wandb_run_name
+
     if wandb_watch:
         os.environ["WANDB_WATCH"] = wandb_watch
     if wandb_log_model:

llama_lora/ui/finetune_ui.py

@@ -79,9 +79,6 @@ def load_sample_dataset_to_text_input(format):
         return gr.Code.update(value=sample_plain_text_value)
 
 
-
-
-
 def get_data_from_input(load_dataset_from, dataset_text, dataset_text_format,
                         dataset_plain_text_input_variables_separator,
                         dataset_plain_text_input_and_output_separator,
@@ -153,7 +150,8 @@ def refresh_preview(
             prompter=prompter
         )
 
-        train_data = prompter.get_train_data_from_dataset(data, max_preview_count)
+        train_data = prompter.get_train_data_from_dataset(
+            data, max_preview_count)
 
         data_count = len(data)
 
@@ -172,18 +170,6 @@ def refresh_preview(
             ]
             preview_data = [d + v for d, v in zip(preview_data, variables)]
 
-
-        # if preview_show_actual_prompt:
-        #     headers = headers + ["Prompt (actual input)"]
-        #     rendered = [prompter.generate_prompt(
-        #         item['variables']) for item in data[:max_preview_count]]
-        #     preview_data = result = [d + [i]
-        #                              for d, i in zip(preview_data, rendered)]
-
-        # headers = headers + ["Completion (output)"]
-        # preview_data = result = [pd + [d['output']]
-        #                          for pd, d in zip(preview_data, data[:max_preview_count])]
-
         preview_info_message = f"The dataset has about {data_count} item(s)."
         if data_count > max_preview_count:
             preview_info_message += f" Previewing the first {max_preview_count}."

llama_lora/utils/prompter.py

@@ -139,18 +139,21 @@ class Prompter(object):
         if self.template_name == "None":
             return ["prompt"]
         elif "variables" in self.template:
-            return self.template.get("variables")
+            return self.template['variables']
         else:
             return ["instruction", "input"]
 
     def get_train_data_from_dataset(self, data, only_first_n_items=None):
         if self.template_module:
-            if hasattr(self.template_module, "get_train_data_list_from_dataset"):
+            if hasattr(self.template_module,
+                       "get_train_data_list_from_dataset"):
                 data = self.template_module.get_train_data_list_from_dataset(
                     data)
             if only_first_n_items:
                 data = data[:only_first_n_items]
-            return list(itertools.chain(*list(map(self.template_module.get_train_data, data))))
+            return list(itertools.chain(*list(
+                map(self.template_module.get_train_data, data)
+            )))
 
         if only_first_n_items:
             data = data[:only_first_n_items]
@@ -179,7 +182,11 @@ def get_prompt_name(variables, variable_names):
 
 
 def variables_to_dict(variables, variable_names):
-    return {key: (variables[i] if i < len(variables) and variables[i] is not None else '') for i, key in enumerate(variable_names)}
+    return {
+        key: (variables[i] if i < len(variables)
+              and variables[i] is not None else '')
+        for i, key in enumerate(variable_names)
+    }
 
 
 def process_json_dataset(data):