Experimenting with few-shot classification.

app.py

@@ -49,17 +49,12 @@ file_name = "onnx/model_quantized.onnx"
 tokenizer_name = "cross-encoder/nli-deberta-v3-small"
 model = ORTModelForSequenceClassification.from_pretrained(model_name, file_name=file_name)
 tokenizer = AutoTokenizer.from_pretrained(tokenizer_name, model_max_length=512)
+classifier = pipeline(task="zero-shot-classification", model=model, tokenizer=tokenizer)
 
-few_shot_model_name = "moshew/bge-small-en-v1.5_setfit-sst2-english"
-few_shot_model = setFitModel.from_pretrained(few_shot_model_name)
 few_shot_tokenizer = AutoTokenizer.from_pretrained('bge_auto_opt_04', model_max_length=512)
 ort_model = ORTModelForFeatureExtraction.from_pretrained('bge_auto_opt_O4')
-onnx_few_shot_model = OnnxSetFitModel(ort_model, tokenizer, model.model_head)
-
-# file = cached_download("https://huggingface.co/" + model_name + "")
-# sess = InferenceSession(file)
-
-classifier = pipeline(task="zero-shot-classification", model=model, tokenizer=tokenizer)
+few_shot_model = SetFitModel.from_pretrained("moshew/bge-small-en-v1.5_setfit-sst2-english")
+onnx_few_shot_model = OnnxSetFitModel(ort_model, few_shot_tokenizer, few_shot_model.model_head)
 
 def classify(data_string, request: gradio.Request):
     if request: