Update app.py

app.py

@@ -16,19 +16,19 @@ M5 = "consciousAI/question-generation-auto-hints-t5-v1-base-s-q-c"
 device = ['cuda' if torch.cuda.is_available() else 'cpu'][0]
 
 _m0 = AutoModelForSeq2SeqLM.from_pretrained(M0).to(device)
-_tk0 = AutoTokenizer.from_pretrained(M0, cache_dir="./cache")
+_tk0 = AutoTokenizer.from_pretrained(M0)
 
 _m1 = AutoModelForSeq2SeqLM.from_pretrained(M1).to(device)
-_tk1 = AutoTokenizer.from_pretrained(M1, cache_dir="./cache")
+_tk1 = AutoTokenizer.from_pretrained(M1)
 
 _m2 = AutoModelForSeq2SeqLM.from_pretrained(M2).to(device)
-_tk2 = AutoTokenizer.from_pretrained(M2, cache_dir="./cache")
+_tk2 = AutoTokenizer.from_pretrained(M2)
 
 _m4 = AutoModelForSeq2SeqLM.from_pretrained(M4).to(device)
-_tk4 = AutoTokenizer.from_pretrained(M4, cache_dir="./cache")
+_tk4 = AutoTokenizer.from_pretrained(M4)
 
 _m5 = AutoModelForSeq2SeqLM.from_pretrained(M5).to(device)
-_tk5 = AutoTokenizer.from_pretrained(M5, cache_dir="./cache")
+_tk5 = AutoTokenizer.from_pretrained(M5)
 
 def _formatQs(questions):
     _finalQs = ""
@@ -298,7 +298,7 @@ with gr.Blocks() as demo:
             topP = gr.Slider(0, 1, 0, label="Top P/Nucleus Sampling")
             temperature = gr.Slider(0.01, 1, 1, label="Temperature") 
         with gr.Row():
-            model = gr.Dropdown(["question-generation-auto-hints-t5-v1-base-s-q-c", "question-generation-auto-hints-t5-v1-base-s-q", "question-generation-auto-t5-v1-base-s-q-c", "question-generation-auto-t5-v1-base-s-q", "question-generation-auto-t5-v1-base-s", "All"], label="Model")    
+            model = gr.Dropdown(["question-generation-auto-hints-t5-v1-base-s-q-c", "question-generation-auto-hints-t5-v1-base-s-q", "question-generation-auto-t5-v1-base-s-q-c", "question-generation-auto-t5-v1-base-s-q", "question-generation-auto-t5-v1-base-s", "All"], label="Model", value="question-generation-auto-hints-t5-v1-base-s-q-c")    
          
     
     with gr.Accordion(variant='compact', label='Input Values'):