Initial swap to baseline bert-base-uncased

app.py

@@ -14,8 +14,10 @@ categories = [
 ]
 logging.debug("Categories to classify: " + repr(categories))
 
-# classifier = pipeline("zero-shot-classification",
-#                       model="facebook/bart-large-mnli")
+classifier = pipeline("zero-shot-classification",
+                      model="google-bert/bert-base-uncased")
+                      # model="Xylor/bert-case-uncased-single-sampled")
+                      # model="facebook/bart-large-mnli")
 
 # sequence_to_classify = "one day I will see the world"
 # candidate_labels = ['travel', 'cooking', 'dancing']
@@ -56,12 +58,13 @@ def clf_text(txt: str | list[str]):
 
 from transformers import AutoModel
 # comment out the flash_attention_2 line if you don't have a compatible GPU
-model = AutoModel.from_pretrained(
-    'jinaai/jina-reranker-m0',
-    torch_dtype="auto",
-    trust_remote_code=True,
-    # attn_implementation="flash_attention_2"
-)
+# model = AutoModel.from_pretrained(
+#     # 'jinaai/jina-reranker-m0',
+#     "Xylor/bert-case-uncased-single-sampled",
+#     torch_dtype="auto",
+#     trust_remote_code=True,
+#     # attn_implementation="flash_attention_2"
+# )
 
 def clf_jina(txt: str | list[str]):
     # construct sentence pairs
@@ -81,9 +84,8 @@ def my_inference_function(name):
   return "Hello " + name + "!"
 
 gradio_interface = gradio.Interface(
-  # fn = my_inference_function,
-  # fn = clf_text,
-  clf_jina,
+  fn = clf_text,
+  # clf_jina,
   inputs = "text",
   outputs = gradio.JSON()
 )