add support for running without argilla

src/distilabel_dataset_generator/apps/eval.py

@@ -39,9 +39,9 @@ from src.distilabel_dataset_generator.utils import (
     extract_column_names,
     get_argilla_client,
     get_org_dropdown,
+    pad_or_truncate_list,
     process_columns,
     swap_visibility,
-    pad_or_truncate_list,
 )
 
 
@@ -334,8 +334,10 @@ def push_dataset(
     push_dataset_to_hub(dataframe, org_name, repo_name, oauth_token, private)
     try:
         progress(0.1, desc="Setting up user and workspace")
-        client = get_argilla_client()
         hf_user = HfApi().whoami(token=oauth_token.token)["name"]
+        client = get_argilla_client()
+        if client is None:
+            return ""
         if eval_type == "ultrafeedback":
             num_generations = len((dataframe["generations"][0]))
             fields = [
@@ -580,6 +582,7 @@ def push_dataset(
 def show_pipeline_code_visibility():
     return {pipeline_code_ui: gr.Accordion(visible=True)}
 
+
 def hide_pipeline_code_visibility():
     return {pipeline_code_ui: gr.Accordion(visible=False)}
 
@@ -708,15 +711,15 @@ with gr.Blocks() as app:
                     visible=False,
                 ) as pipeline_code_ui:
                     code = generate_pipeline_code(
-                            repo_id=search_in.value,
-                            aspects=aspects_instruction_response.value,
-                            instruction_column=instruction_instruction_response,
-                            response_columns=response_instruction_response,
-                            prompt_template=prompt_template.value,
-                            structured_output=structured_output.value,
-                            num_rows=num_rows.value,
-                            eval_type=eval_type.value,
-                        )
+                        repo_id=search_in.value,
+                        aspects=aspects_instruction_response.value,
+                        instruction_column=instruction_instruction_response,
+                        response_columns=response_instruction_response,
+                        prompt_template=prompt_template.value,
+                        structured_output=structured_output.value,
+                        num_rows=num_rows.value,
+                        eval_type=eval_type.value,
+                    )
                     pipeline_code = gr.Code(
                         value=code,
                         language="python",

src/distilabel_dataset_generator/apps/sft.py

@@ -220,8 +220,10 @@ def push_dataset(
     push_dataset_to_hub(dataframe, org_name, repo_name, oauth_token, private)
     try:
         progress(0.1, desc="Setting up user and workspace")
-        client = get_argilla_client()
         hf_user = HfApi().whoami(token=oauth_token.token)["name"]
+        client = get_argilla_client()
+        if client is None:
+            return ""
         if "messages" in dataframe.columns:
             settings = rg.Settings(
                 fields=[

src/distilabel_dataset_generator/apps/textcat.py

@@ -58,7 +58,10 @@ def generate_system_prompt(dataset_description, temperature, progress=gr.Progres
     labels = data["labels"]
     return system_prompt, labels
 
-def generate_sample_dataset(system_prompt, difficulty, clarity, labels, num_labels, progress=gr.Progress()):
+
+def generate_sample_dataset(
+    system_prompt, difficulty, clarity, labels, num_labels, progress=gr.Progress()
+):
     dataframe = generate_dataset(
         system_prompt=system_prompt,
         difficulty=difficulty,
@@ -138,11 +141,7 @@ def generate_dataset(
     # create final dataset
     distiset_results = []
     for result in labeller_results:
-        record = {
-            key: result[key]
-            for key in ["labels", "text"]
-            if key in result
-        }
+        record = {key: result[key] for key in ["labels", "text"] if key in result}
         distiset_results.append(record)
 
     dataframe = pd.DataFrame(distiset_results)
@@ -212,13 +211,16 @@ def push_dataset(
     push_dataset_to_hub(
         dataframe, org_name, repo_name, num_labels, labels, oauth_token, private
     )
+
     dataframe = dataframe[
         (dataframe["text"].str.strip() != "") & (dataframe["text"].notna())
     ]
     try:
         progress(0.1, desc="Setting up user and workspace")
-        client = get_argilla_client()
         hf_user = HfApi().whoami(token=oauth_token.token)["name"]
+        client = get_argilla_client()
+        if client is None:
+            return ""
         labels = get_preprocess_labels(labels)
         settings = rg.Settings(
             fields=[