Upload data/lilac.yml with huggingface_hub

data/lilac.yml

@@ -11,6 +11,10 @@ datasets:
     embeddings:
       - path: premise
         embedding: gte-small
+      - path: premise
+        embedding: gte-base
+      - path: hypothesis
+        embedding: gte-small
     signals:
       - path: premise
         signal:
@@ -25,6 +29,7 @@ datasets:
       ui:
         media_paths:
           - premise
+        markdown_paths: []
   - namespace: local
     name: glue_ax
     source:
@@ -66,6 +71,7 @@ datasets:
       ui:
         media_paths:
           - hypothesis
+        markdown_paths: []
   - namespace: local
     name: imdb3
     source:
@@ -75,6 +81,7 @@ datasets:
       ui:
         media_paths:
           - text
+        markdown_paths: []
   - namespace: local
     name: imdb
     source:
@@ -94,6 +101,7 @@ datasets:
       ui:
         media_paths:
           - text
+        markdown_paths: []
   - namespace: local
     name: imdb2
     source:
@@ -103,6 +111,7 @@ datasets:
       ui:
         media_paths:
           - text
+        markdown_paths: []
   - namespace: lilac
     name: OpenOrca-100k
     source:
@@ -235,11 +244,15 @@ datasets:
       - path: response
         signal:
           signal_name: text_statistics
+      - path: system_prompt
+        signal:
+          signal_name: pii
     settings:
       ui:
         media_paths:
           - question
           - response
+        markdown_paths: []
   - namespace: local
     name: the_movies_dataset
     source:
@@ -251,6 +264,7 @@ datasets:
       ui:
         media_paths:
           - overview
+        markdown_paths: []
   - namespace: local
     name: glue_ax_parquet
     source:
@@ -261,6 +275,7 @@ datasets:
       ui:
         media_paths:
           - premise
+        markdown_paths: []
   - namespace: lilac
     name: mmlu_professional_law
     source:
@@ -425,6 +440,7 @@ datasets:
           - question
           - - choices
             - '*'
+        markdown_paths: []
       preferred_embedding: gte-small
   - namespace: local
     name: deepset-prompt-inj
@@ -438,6 +454,7 @@ datasets:
       ui:
         media_paths:
           - text
+        markdown_paths: []
   - namespace: local
     name: jasper-prompt-inj
     source:
@@ -450,6 +467,7 @@ datasets:
       ui:
         media_paths:
           - text
+        markdown_paths: []
   - namespace: local
     name: mosaic-chat-v2
     source:
@@ -550,6 +568,7 @@ datasets:
         media_paths:
           - prompt
           - response
+        markdown_paths: []
       preferred_embedding: gte-small
   - namespace: local
     name: databricks-dolly-15k-curated-en
@@ -564,6 +583,8 @@ datasets:
           - value
           - '*'
         embedding: gte-small
+      - path: original-instruction
+        embedding: gte-small
     signals:
       - path: original-instruction
         signal:
@@ -793,6 +814,9 @@ datasets:
           - '*'
         signal:
           signal_name: text_statistics
+      - path: original-instruction
+        signal:
+          signal_name: spacy_ner
     settings:
       ui:
         media_paths:
@@ -808,6 +832,7 @@ datasets:
           - - new-response
             - value
             - '*'
+        markdown_paths: []
       preferred_embedding: gte-small
   - namespace: local
     name: open-asssistant-conversations
@@ -888,6 +913,7 @@ datasets:
       ui:
         media_paths:
           - text
+        markdown_paths: []
       preferred_embedding: gte-small
   - namespace: local
     name: enron-emails
@@ -964,4 +990,5 @@ datasets:
       ui:
         media_paths:
           - text
+        markdown_paths: []
       preferred_embedding: gte-small