update tags

.gitattributes

@@ -26,3 +26,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zstandard filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
 *.json filter=lfs diff=lfs merge=lfs -text
+metadata_942418870f35246ce32d98caf3f230024a58428b.json filter=lfs diff=lfs merge=lfs -text

metadata_526dba9dce6f983ebcacd65bef35a868a5735df6.json → metadata_942418870f35246ce32d98caf3f230024a58428b.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5232c816fea8dbdca0878f9520848496cacb64e5d28e375d46471cbdcc81fb11
-size 30119915
+oid sha256:7052da9dcb564799fdfb47ac8f50232fd3022e8d14023d76c280c51eb613e405
+size 30721389

requirements.txt

@@ -1,4 +1,4 @@
 pyyaml
 streamlit>=0.88.0
 langcodes[data]
-datasets~=2.1
+git+https://github.com/huggingface/datasets@7e514c3#egg=datasets

tagging_app.py

@@ -76,10 +76,10 @@ def multiselect(
     return w.multiselect(markdown, valid_set, default=valid_values, format_func=format_func, key=title)
 
 
-def validate_dict(w: st.delta_generator.DeltaGenerator, state_dict: Dict):
+def validate_yaml(w: st.delta_generator.DeltaGenerator, yamlblock: str):
     try:
-        DatasetMetadata(**state_dict)
-        if not state_dict.get("pretty_name"):
+        DatasetMetadata.from_yaml_string(yamlblock)
+        if "pretty_name: " not in yamlblock or "pretty_name: ''" in yamlblock:
             raise ValueError("Please specify a non-empty Dataset name.")
         w.markdown("✅ This is a valid tagset! 🤗")
     except Exception as e:
@@ -168,10 +168,11 @@ if leftbtn.button("pre-load"):
     state = initial_state or new_state()
     st.experimental_set_query_params(preload_dataset=preloaded_id)
 if not is_state_empty(state):
-    if rightbtn.button("flush state"):
+    if rightbtn.button("reset"):
         state = new_state()
         initial_state = None
         preloaded_id = None
+        initial_infos = None
         st.experimental_set_query_params()
 
 if preloaded_id is not None and initial_state is not None:
@@ -181,7 +182,7 @@ if preloaded_id is not None and initial_state is not None:
 The current base tagset is [`{preloaded_id}`](https://huggingface.co/datasets/{preloaded_id})
 """
     )
-    validate_dict(st.sidebar, initial_state)
+    validate_yaml(st.sidebar, yaml.dump(initial_state))
     st.sidebar.markdown(
         f"""
 Here is the matching yaml block:
@@ -382,7 +383,7 @@ rightcol.markdown(
 if is_state_empty(state):
     rightcol.markdown("❌ This is an invalid tagset: it's empty!")
 else:
-    validate_dict(rightcol, state)
+    validate_yaml(rightcol, yaml.dump(state))
 
 
 rightcol.markdown(
@@ -400,5 +401,4 @@ This is a standalone tool, it is useful to check for errors on an existing tagse
 
 yamlblock = rightcol.text_area("Input your yaml here")
 if yamlblock.strip() != "":
-    inputdict = yaml.safe_load(yamlblock)
-    validate_dict(rightcol, inputdict)
+    validate_yaml(rightcol, yamlblock)