Update app.py

app.py

@@ -40,8 +40,8 @@ os.environ["TRANSFORMERS_CACHE"] = "/data/models"
 MAX_FILE_SIZE_MB = 20
 MAX_FILE_SIZE_BYTES = MAX_FILE_SIZE_MB * 1024 * 1024
 
-MODEL_ID = "unsloth/Llama-3.2-3B"#meta-llama/Meta-Llama-3-8B" #"unsloth/Llama-4-Scout-17B-16E-Instruct-GGUF"# unsloth/Qwen2.5-1.5B" #unsloth/Llama-3.2-3B" #unsloth/Llama-3.2-1B"
-#glotoken=Tokentest
+MODEL_ID = "meta-llama/Meta-Llama-3-8B"# unsloth/Llama-3.2-3B"#meta-llama/Meta-Llama-3-8B" #"unsloth/Llama-4-Scout-17B-16E-Instruct-GGUF"# unsloth/Qwen2.5-1.5B" #unsloth/Llama-3.2-3B" #unsloth/Llama-3.2-1B"
+glotoken=Tokentest
 # Global logging system - CRITICAL FIX #1
 logs = []
 
@@ -82,6 +82,7 @@ def initialize_model():
         tokenizer = AutoTokenizer.from_pretrained(
             MODEL_ID,
             cache_dir="/data/models",
+            token=glotoken,
             trust_remote_code=True,
             use_fast=False
         )
@@ -96,6 +97,7 @@ def initialize_model():
             torch_dtype=torch.float16,
             cache_dir="/data/models",
             trust_remote_code=True,
+            token=glotoken,
             device_map={"": 0},  # <- force GPU:0
             low_cpu_mem_usage=True
         )