Update app.py

app.py

@@ -7,6 +7,8 @@ import spaces
 import torch
 from transformers import AutoModelForCausalLM, GemmaTokenizerFast, TextIteratorStreamer
 
+huggingface_token = os.getenv("HUGGINGFACE_TOKEN")
+
 MAX_MAX_NEW_TOKENS = 2048
 DEFAULT_MAX_NEW_TOKENS = 1024
 MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))
@@ -14,11 +16,12 @@ MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))
 device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
 
 model_id = "google/gemma-2-9b-it"
-tokenizer = GemmaTokenizerFast.from_pretrained(model_id)
+tokenizer = GemmaTokenizerFast.from_pretrained(model_id, token = huggingface_token)
 model = AutoModelForCausalLM.from_pretrained(
     model_id,
     device_map="auto",
     torch_dtype=torch.bfloat16,
+    token = huggingface_token
 )
 model.config.sliding_window = 4096
 model.eval()