Create app.py

app.py

@@ -0,0 +1,55 @@
+import gradio as gr
+import torch
+import transformers
+
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+
+
+
+class Model(torch.nn.Module):
+    def __init__(self):
+        super().__init__()
+        self.MODEL_NAME = "distilbert-base-uncased-finetuned-sst-2-english"
+        self.bert = transformers.AutoModelForSequenceClassification.from_pretrained(self.MODEL_NAME)
+        self.tokenizer = transformers.AutoTokenizer.from_pretrained(self.MODEL_NAME)
+        
+    def forward(self, input_text):
+        encoding = self.tokenizer.encode_plus(
+            input_text,
+            add_special_tokens = True, 
+            pad_to_max_length = True,
+            return_token_type_ids = False,
+            return_attention_mask = True,
+            return_tensors = 'pt' 
+        )
+
+        input_ids = encoding['input_ids'].to(device)
+        attention_mask = encoding['attention_mask'].to(device)
+
+        output = self.bert(
+            input_ids = input_ids,
+            attention_mask = attention_mask)
+
+        return output
+
+
+def predict(input_text):
+    model = Model()
+    model.load_state_dict(torch.load("ayse/distilbert-english-finetuned", map_location=device), strict=False)
+    model.eval()
+    outputs = model(input_text)
+    logits = outputs.logits
+    prediction = torch.argmax(logits, dim=-1)
+    if prediction.item() == 0:
+        return "NEGATIVE"
+    if prediction.item() == 1:
+        return "POSITIVE"
+
+
+iface = gr.Interface(predict,
+inputs="text",
+outputs="text",
+title="Bert Base Sentiment Analysis",
+description="This is a bert based sentiment classifier that is trained with tinder application reviews (EN).",
+allow_flagging="never")
+iface.launch(inbrowser=True)