a second workaround for deploying metrics to spaces

app.py

@@ -1,7 +1,40 @@
 import evaluate
-from evaluate.utils import launch_gradio_widget
+from evaluate.utils import infer_gradio_input_types, json_to_string_type, parse_readme
 from fixed_f1 import FixedF1
+# from evaluate.utils import launch_gradio_widget # using this directly is erroneous - lets fix this
 
+metric = FixedF1()
 
-module = FixedF1()
-launch_gradio_widget(module)
+if isinstance(metric.features, list):
+    (feature_names, feature_types) = zip(*metric.features[0].items())
+else:
+    (feature_names, feature_types) = zip(*metric.features.items())
+    gradio_input_types = infer_gradio_input_types(feature_types)
+
+gradio_input_types = infer_gradio_input_types(feature_types)
+
+def compute():
+    metric._compute()
+
+import gradio as gr
+
+space = gr.Interface(
+    fn=compute,
+    inputs=gr.Dataframe(
+        headers=feature_names,
+        col_count=len(feature_names),
+        row_count=1,
+        datatype=json_to_string_type(gradio_input_types),
+        ),
+    outputs=gr.Textbox(label=metric.name),
+    description=(
+        metric.info.description + "\nIf this is a text-based metric, make sure to wrap your input in double quotes."
+        " Alternatively you can use a JSON-formatted list as input."
+    ),
+    title=f"Metric: {metric.name}",
+    article=parse_readme("./README.md"),
+    # TODO: load test cases and use them to populate examples
+    # examples=[parse_test_cases(test_cases, feature_names, gradio_input_types)]
+    )
+
+space.launch()