Update app.py

app.py

@@ -8,17 +8,7 @@ from threading import Thread
 
 
 
-# Loading the tokenizer and model from Hugging Face's model hub.
-tokenizer = AutoTokenizer.from_pretrained(
-    "macadeliccc/SOLAR-math-2x10.7b",
-    trust_remote_code=True
-)
-model = AutoModelForCausalLM.from_pretrained(
-    "macadeliccc/SOLAR-math-2x10.7b",
-    torch_dtype="auto",
-    load_in_8bit=True,
-    trust_remote_code=True
-)
+
 
 # Defining a custom stopping criteria class for the model's text generation.
 class StopOnTokens(StoppingCriteria):
@@ -34,7 +24,18 @@ class StopOnTokens(StoppingCriteria):
 @spaces.GPU
 def predict(message, history):
     torch.set_default_device("cuda")
-    
+
+    # Loading the tokenizer and model from Hugging Face's model hub.
+    tokenizer = AutoTokenizer.from_pretrained(
+        "macadeliccc/SOLAR-math-2x10.7b",
+        trust_remote_code=True
+    )
+    model = AutoModelForCausalLM.from_pretrained(
+        "macadeliccc/SOLAR-math-2x10.7b",
+        torch_dtype="auto",
+        load_in_8bit=True,
+        trust_remote_code=True
+    )
     history_transformer_format = history + [[message, ""]]
     stop = StopOnTokens()