Messing with configuration.

app.py

@@ -5,7 +5,7 @@ import torch
 import onnxruntime
 from optimum.onnxruntime import ORTModelForSequenceClassification
 from transformers import AutoTokenizer
-from optimum.pipelines import pipeline
+from transformers import pipeline
 from fastapi import FastAPI
 from fastapi.middleware.cors import CORSMiddleware
 
@@ -30,9 +30,9 @@ print(f"CUDA device: {torch.cuda.get_device_name(torch.cuda.current_device())}")
 # "Xenova/deBERTa-v3-base-mnli" "MoritzLaurer/DeBERTa-v3-base-mnli" Still a bit slow and not great answers
 # "xenova/nli-deberta-v3-small" "cross-encoder/nli-deberta-v3-small" Was using this for a good while and it was...okay
 
-# model_name = "MoritzLaurer/deberta-v3-base-zeroshot-v2.0"
+model_name = "MoritzLaurer/deberta-v3-base-zeroshot-v2.0"
 # file_name = "onnx/model.onnx"
-# tokenizer_name = "MoritzLaurer/deberta-v3-base-zeroshot-v2.0"
+tokenizer_name = "MoritzLaurer/deberta-v3-base-zeroshot-v2.0"
 
 # model = ORTModelForSequenceClassification.from_pretrained(model_name, export=True, provider="CUDAExecutionProvider")
 # tokenizer = AutoTokenizer.from_pretrained(tokenizer_name, model_max_length=512)
@@ -40,20 +40,21 @@ print(f"CUDA device: {torch.cuda.get_device_name(torch.cuda.current_device())}")
 session_options = onnxruntime.SessionOptions()
 session_options.log_severity_level = 0
 
-print(f"ORTModelForSequenceClassification.from_pretrained")
-model = ORTModelForSequenceClassification.from_pretrained(
-    "distilbert-base-uncased-finetuned-sst-2-english",
-    export=True,
-    provider="CUDAExecutionProvider",
-    session_options=session_options
-)
-print(f"AutoTokenizer.from_pretrained")
-tokenizer = AutoTokenizer.from_pretrained("distilbert-base-uncased-finetuned-sst-2-english")
+# print(f"ORTModelForSequenceClassification.from_pretrained")
+# model = ORTModelForSequenceClassification.from_pretrained(
+#     "distilbert-base-uncased-finetuned-sst-2-english",
+#     export=True,
+#     provider="CUDAExecutionProvider",
+#     session_options=session_options
+# )
+# print(f"AutoTokenizer.from_pretrained")
+# tokenizer = AutoTokenizer.from_pretrained("distilbert-base-uncased-finetuned-sst-2-english")
 
-# classifier = pipeline(task="zero-shot-classification", model=model, tokenizer=tokenizer, device="cuda:0")
+print(f"pipeline")
+classifier = pipeline(task="zero-shot-classification", model=model_name, tokenizer=tokenizer_name, device="cuda:0")
 
 print(f"Testing 1")
-@spaces.GPU()
+@spaces.GPU
 def classify(data_string, request: gradio.Request):
     if request:
         if request.headers["origin"] not in ["https://statosphere-3704059fdd7e.c5v4v4jx6pq5.win", "https://crunchatize-77a78ffcc6a6.c5v4v4jx6pq5.win", "https://crunchatize-2-2b4f5b1479a6.c5v4v4jx6pq5.win", "https://tamabotchi-2dba63df3bf1.c5v4v4jx6pq5.win", "https://ravenok-statosphere-backend.hf.space", "https://lord-raven.github.io"]:
@@ -66,8 +67,7 @@ def classify(data_string, request: gradio.Request):
 
 print(f"Testing 2")
 def zero_shot_classification(data):
-    results = []
-    # classifier(data['sequence'], candidate_labels=data['candidate_labels'], hypothesis_template=data['hypothesis_template'], multi_label=data['multi_label'])
+    results = classifier(data['sequence'], candidate_labels=data['candidate_labels'], hypothesis_template=data['hypothesis_template'], multi_label=data['multi_label'])
     response_string = json.dumps(results)
     return response_string