Update app.py

app.py

@@ -54,16 +54,16 @@ lyrics_pool = [
 if torch.cuda.is_available():
     #model_id = "Qwen/Qwen2.5-7B-Instruct"
     #model_id = "BenBranyon/sumbot7b-augmented"
-    model_id = "turboderp/Cat-Llama-3-70B-instruct"
-    model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto")
-    tokenizer = AutoTokenizer.from_pretrained(model_id, use_fast=True, model_max_length=MAX_INPUT_TOKEN_LENGTH, padding="longest", language="en")
-
-    #model_id = "BenBranyon/lora_sumbot_v2"
-    #model = AutoPeftModelForCausalLM.from_pretrained(
-    #    model_id, # YOUR MODEL YOU USED FOR TRAINING
-    #    load_in_4bit = True,
-    #)
-    #tokenizer = AutoTokenizer.from_pretrained(model_id, model_max_length=MAX_INPUT_TOKEN_LENGTH, langauge="en")
+    #model_id = "turboderp/Cat-Llama-3-70B-instruct"
+    #model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto")
+    #tokenizer = AutoTokenizer.from_pretrained(model_id, use_fast=True, model_max_length=MAX_INPUT_TOKEN_LENGTH, padding="longest", language="en")
+
+    model_id = "BenBranyon/lora_sumbot_v2"
+    model = AutoPeftModelForCausalLM.from_pretrained(
+        model_id, # YOUR MODEL YOU USED FOR TRAINING
+        load_in_4bit = True,
+    )
+    tokenizer = AutoTokenizer.from_pretrained(model_id, model_max_length=MAX_INPUT_TOKEN_LENGTH, langauge="en")
 
 #Transformers Code
 @spaces.GPU