Update app.py

app.py

@@ -104,7 +104,7 @@ def imageMode(image, question):
     imgToTextEmb = img_proj_head(image_embedding)
     print('-------text embedding from projection obtained-----------')
     question = "Question: " + question + "Answer: "
-    Qtokens = torch.tensor(tokenizer_text.encode(question, add_special_tokens=True))
+    Qtokens = torch.tensor(tokenizer_text.encode(question, add_special_tokens=True)).unsqueeze(0)
     Qtoken_embeddings = phi2_finetuned.get_submodule('model.embed_tokens')(Qtokens)
     print('-------question embedding from phi2 obtained-----------')
     inputs = torch.concat((imgToTextEmb, Qtoken_embeddings), axis=-2)