app setup

.gitattributes

@@ -25,3 +25,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zstandard filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+models_with_readmes.csv filter=lfs diff=lfs merge=lfs -text

app.py

@@ -0,0 +1,26 @@
+import pandas as pd
+import streamlit as st
+import math
+
+models_df = pd.read_csv("models_with_readmes.csv").dropna(subset=["readme"])
+
+st.title("Huggingface model explorer")
+st.text(f"search {len(models_df)} models by name or readme")
+st.text("note that there are many more models but here we only show those with readme")
+
+query = st.text_input("model name query", value="")
+readme_query = st.text_input("readme query", value="")
+
+selected_models_df = models_df[
+    models_df["readme"].str.contains(readme_query)
+    & models_df["modelId"].str.contains(query)
+]
+
+n_per_page = 10
+slider = st.number_input("page", 0, math.ceil(len(selected_models_df) / 10))
+selected_models_df_subset = selected_models_df.iloc[
+    slider * n_per_page : (slider + 1) * n_per_page
+]
+st.write(f"matched {len(selected_models_df)} models")
+for i, (model_name, readme) in enumerate(selected_models_df_subset[["modelId", "readme"]].itertuples(index=False)):
+    st.write(readme)