Update app.py

app.py

@@ -3,9 +3,6 @@ import torch
 import random
 from transformers import T5Tokenizer, T5ForConditionalGeneration
 
-def load_model(model_path, dtype):
-    model = T5ForConditionalGeneration.from_pretrained(model_path, torch_dtype=dtype)
-    return model
 
 def generate(
     prompt,
@@ -21,7 +18,7 @@ def generate(
     dtype="fp16",
 ):
     tokenizer = T5Tokenizer.from_pretrained("google/flan-t5-small")
-    model = load_model(model_path, dtype)
+    model = T5ForConditionalGeneration.from_pretrained(model_path, torch_dtype=dtype)
 
     if torch.cuda.is_available():
         device = "cuda"
@@ -114,7 +111,7 @@ additional_inputs = [
     ),
     gr.Radio(
         choices=[("fp32", torch.float32), ("fp16", torch.float16)],
-        value="fp16",
+        value=torch.float16,
         label="Model Precision",
         info="fp32 is more precised, fp16 is faster and less memory consuming",
     ),