Initial Commit
Browse files- app.py +36 -0
- requirements.txt +3 -0
app.py
ADDED
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from transformers import AutoTokenizer
|
2 |
+
import transformers
|
3 |
+
import torch
|
4 |
+
import streamlit as st
|
5 |
+
|
6 |
+
prompt = st.text_input('Prompt', 'Hello, How you doing ?')
|
7 |
+
|
8 |
+
model = "meta-llama/Llama-2-13b-chat-hf"
|
9 |
+
|
10 |
+
#Not Working
|
11 |
+
#tokenizer = AutoTokenizer.from_pretrained(model)
|
12 |
+
|
13 |
+
|
14 |
+
pipeline = transformers.pipeline(
|
15 |
+
"text-generation",
|
16 |
+
model=model,
|
17 |
+
torch_dtype=torch.float16,
|
18 |
+
device_map="auto",
|
19 |
+
do_sample=True,
|
20 |
+
)
|
21 |
+
|
22 |
+
def get_llama_response(prompt):
|
23 |
+
|
24 |
+
sequences = pipeline(
|
25 |
+
prompt,
|
26 |
+
do_sample=True,
|
27 |
+
top_k=10,
|
28 |
+
num_return_sequences=1,
|
29 |
+
max_length=256,
|
30 |
+
)
|
31 |
+
print(sequences[0]['generated_text'])
|
32 |
+
|
33 |
+
#prompt="Can you help me to write rest api endpoints in python ?"
|
34 |
+
response = get_llama_response(prompt)
|
35 |
+
|
36 |
+
st.write('Answer: ',response)
|
requirements.txt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
streamlit
|
2 |
+
transformers
|
3 |
+
torch
|