GSK-2844-to-2841-fix-for-model-url-add-doc-and-UI (#120)

- add doc seo; fix model id link; fix leaderboard; add loading bar (c88b4981105a1857d96638bb7ccc7f7c99d80ad4)

app_leaderboard.py

@@ -90,7 +90,7 @@ def get_demo(leaderboard_tab):
     column_names = records.columns.tolist()
     issue_columns = column_names[:11]
     info_columns = column_names[15:]
-    default_columns = ["dataset_id", "total_issues", "report_link"]
+    default_columns = ["model_id", "dataset_id", "total_issues", "report_link"]
     default_df = records[default_columns]  # extract columns selected
     types = get_types(default_df)
     display_df = get_display_df(default_df)  # the styled dataframe to display

app_text_classification.py

@@ -59,7 +59,7 @@ def get_demo():
     with gr.Row():
         first_line_ds = gr.DataFrame(label="Dataset Preview", visible=False)
     with gr.Row():
-        loading_status = gr.HTML(visible=True)
+        loading_dataset_info = gr.HTML(visible=True)
     with gr.Row():
         example_btn = gr.Button(
             "Validate Model & Dataset",
@@ -67,9 +67,10 @@ def get_demo():
             variant="primary",
             interactive=False,
         )
-
     with gr.Row():
-        example_input = gr.HTML(visible=False)
+        loading_validation = gr.HTML(visible=True)
+    with gr.Row():
+        validation_result = gr.HTML(visible=False)
     with gr.Row():
         example_prediction = gr.Label(label="Model Sample Prediction", visible=False)
 
@@ -153,7 +154,7 @@ def get_demo():
         triggers=[dataset_id_input.input, dataset_id_input.select],
         fn=check_dataset,
         inputs=[dataset_id_input],
-        outputs=[dataset_config_input, dataset_split_input, loading_status]
+        outputs=[dataset_config_input, dataset_split_input, loading_dataset_info]
     )
 
     dataset_config_input.change(fn=get_dataset_splits, inputs=[dataset_id_input, dataset_config_input], outputs=[dataset_split_input])
@@ -197,7 +198,12 @@ def get_demo():
             dataset_config_input,
             dataset_split_input,
         ],
-        outputs=[example_btn, first_line_ds, loading_status],
+        outputs=[
+            example_btn, 
+            first_line_ds,
+            validation_result,
+            example_prediction,
+            column_mapping_accordion,],
     )
 
     gr.on(
@@ -215,11 +221,11 @@ def get_demo():
             inference_token,
         ],
         outputs=[
-            example_input,
+            validation_result,
             example_prediction,
             column_mapping_accordion,
             run_btn,
-            loading_status,
+            loading_validation,
             *column_mappings,
         ],
     )

text_classification.py

@@ -380,7 +380,7 @@ def text_classification_fix_column_mapping(column_mapping, ppl, d_id, config, sp
 
 def strip_model_id_from_url(model_id):
     if model_id.startswith("https://huggingface.co/"):
-        return "/".join(model_id.split("/")[-2])
+        return "/".join(model_id.split("/")[-2:])
     return model_id
 
 def check_hf_token_validity(hf_token):
@@ -393,21 +393,4 @@ def check_hf_token_validity(hf_token):
     response = requests.get(AUTH_CHECK_URL, headers=headers)
     if response.status_code != 200:
         return False
-    return True
-
-def get_dataset_info_from_server(dataset_id):
-    url = "https://datasets-server.huggingface.co/splits?dataset=" + dataset_id
-    response = requests.get(url)
-    if response.status_code != 200:
-        return None
-    return response.json()
-
-def get_dataset_splits(dataset_id, dataset_config):
-    dataset_info = get_dataset_info_from_server(dataset_id)
-    if dataset_info is None:
-        return None
-    try:
-        splits = dataset_info["splits"]
-        return [split["split"] for split in splits if split["config"] == dataset_config]
-    except Exception:
-        return None
+    return True

text_classification_ui_helpers.py

@@ -179,29 +179,57 @@ def precheck_model_ds_enable_example_btn(
     model_task = check_model_task(model_id)
     preload_hf_inference_api(model_id)
 
-    if model_task is None or model_task != "text-classification":
-        gr.Warning(NOT_TEXT_CLASSIFICATION_MODEL_RAW)
-        return (gr.update(interactive=False), gr.update(visible=False),"")
-
     if dataset_config is None or dataset_split is None or len(dataset_config) == 0:
-        return (gr.update(interactive=False), gr.update(visible=False), "")
-    
+        return (
+            gr.update(interactive=False), 
+            gr.update(visible=False),
+            gr.update(visible=False),
+            gr.update(visible=False),
+            gr.update(visible=False),
+        )
+
     try:
         ds = datasets.load_dataset(dataset_id, dataset_config, trust_remote_code=True)
         df: pd.DataFrame = ds[dataset_split].to_pandas().head(5)
         ds_labels, ds_features = get_labels_and_features_from_dataset(ds[dataset_split])
-        if model_id == "" or model_id is None:
-          return (gr.update(interactive=False), gr.update(value=df, visible=True), "")
+
+        if model_task is None or model_task != "text-classification":
+          gr.Warning(NOT_TEXT_CLASSIFICATION_MODEL_RAW)
+          return (
+              gr.update(interactive=False), 
+              gr.update(value=df, visible=True),
+              gr.update(visible=False), 
+              gr.update(visible=False),
+              gr.update(visible=False),
+          )
 
         if not isinstance(ds_labels, list) or not isinstance(ds_features, list):
             gr.Warning(CHECK_CONFIG_OR_SPLIT_RAW)
-            return (gr.update(interactive=False), gr.update(value=df, visible=True), "")
+            return (
+                gr.update(interactive=False), 
+                gr.update(value=df, visible=True),
+                gr.update(visible=False),
+                gr.update(visible=False),
+                gr.update(visible=False),
+            )
 
-        return (gr.update(interactive=True), gr.update(value=df, visible=True), "")
+        return (
+                gr.update(interactive=True), 
+                gr.update(value=df, visible=True),
+                gr.update(visible=False),
+                gr.update(visible=False),
+                gr.update(visible=False),
+            )
     except Exception as e:
         # Config or split wrong
         logger.warn(f"Check your dataset {dataset_id} and config {dataset_config} on split {dataset_split}: {e}")
-        return (gr.update(interactive=False), gr.update(visible=False), "")
+        return (
+            gr.update(interactive=False), 
+            gr.update(visible=False),
+            gr.update(visible=False),
+            gr.update(visible=False),
+            gr.update(visible=False),
+        )
 
 
 def align_columns_and_show_prediction(

wordings.py

@@ -2,7 +2,8 @@ INTRODUCTION_MD = """
                 <h1 style="text-align: center;">
                 🐢Giskard Evaluator - Text Classification
                 </h1>
-                Welcome to the Giskard Evaluator Space! Get a model vulnerability report immediately by simply sharing your model and dataset id below.
+                Welcome to the Giskard Evaluator Space! Get a model vulnerability report immediately by simply sharing your model and dataset id below. 
+                You can also checkout our library documentation <a href="https://docs.giskard.ai/en/latest/getting_started/quickstart/index.html">here</a>.
                 """
 CONFIRM_MAPPING_DETAILS_MD = """
                             <h1 style="text-align: center;">