Spaces:
Sleeping
Sleeping
Waseem7711
commited on
Commit
•
822b0d9
1
Parent(s):
b51e4ba
Update app.py
Browse files
app.py
CHANGED
@@ -9,11 +9,8 @@ from dotenv import load_dotenv
|
|
9 |
# Load environment variables
|
10 |
load_dotenv()
|
11 |
|
12 |
-
|
13 |
-
|
14 |
-
|
15 |
-
# Retrieve Hugging Face API token from environment variables (if accessing private models)
|
16 |
-
HF_API_TOKEN = os.getenv("HF_API_TOKEN") # Ensure you set this in Hugging Face Secrets
|
17 |
|
18 |
# Streamlit app setup
|
19 |
st.title('Llama2 Chatbot Deployment on Hugging Face Spaces')
|
@@ -27,13 +24,13 @@ def load_model():
|
|
27 |
"""
|
28 |
tokenizer = AutoTokenizer.from_pretrained(
|
29 |
"meta-llama/Llama-2-7b-chat-hf",
|
30 |
-
use_auth_token=
|
31 |
)
|
32 |
model = AutoModelForCausalLM.from_pretrained(
|
33 |
"meta-llama/Llama-2-7b-chat-hf",
|
34 |
torch_dtype=torch.float16, # Use float16 for reduced memory usage
|
35 |
device_map="auto",
|
36 |
-
use_auth_token=HF_API_TOKEN #
|
37 |
)
|
38 |
return tokenizer, model
|
39 |
|
|
|
9 |
# Load environment variables
|
10 |
load_dotenv()
|
11 |
|
12 |
+
# Retrieve Hugging Face API token from environment variables
|
13 |
+
HF_API_TOKEN = os.getenv("HF_API_TOKEN")
|
|
|
|
|
|
|
14 |
|
15 |
# Streamlit app setup
|
16 |
st.title('Llama2 Chatbot Deployment on Hugging Face Spaces')
|
|
|
24 |
"""
|
25 |
tokenizer = AutoTokenizer.from_pretrained(
|
26 |
"meta-llama/Llama-2-7b-chat-hf",
|
27 |
+
use_auth_token=HF_API_TOKEN # Use the secret token
|
28 |
)
|
29 |
model = AutoModelForCausalLM.from_pretrained(
|
30 |
"meta-llama/Llama-2-7b-chat-hf",
|
31 |
torch_dtype=torch.float16, # Use float16 for reduced memory usage
|
32 |
device_map="auto",
|
33 |
+
use_auth_token=HF_API_TOKEN # Use the secret token
|
34 |
)
|
35 |
return tokenizer, model
|
36 |
|