Spaces:
Runtime error
Runtime error
Update app.py
Browse files
app.py
CHANGED
@@ -3,7 +3,6 @@ from threading import Thread
|
|
3 |
from typing import Iterator
|
4 |
|
5 |
import gradio as gr
|
6 |
-
import spaces
|
7 |
import torch
|
8 |
from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamer
|
9 |
DESCRIPTION = """\
|
@@ -14,6 +13,8 @@ This is a demo of [`Qwen/Qwen2-0.5B-Instruct`](https://huggingface.co/Qwen/Qwen2
|
|
14 |
This space allows you to input text and have the AI complete it. Simply type your text in the input box, click "Complete", and watch as the AI generates a continuation of your text.
|
15 |
|
16 |
You can adjust various parameters such as temperature and top-p sampling to control the generation process.
|
|
|
|
|
17 |
"""
|
18 |
|
19 |
MAX_MAX_NEW_TOKENS = 2048
|
@@ -30,7 +31,6 @@ model = AutoModelForCausalLM.from_pretrained(
|
|
30 |
torch_dtype=torch.bfloat16,
|
31 |
)
|
32 |
model.eval()
|
33 |
-
@spaces.GPU(duration=90)
|
34 |
def generate(
|
35 |
message: str,
|
36 |
max_new_tokens: int = 1024,
|
@@ -64,7 +64,8 @@ def generate(
|
|
64 |
for text in streamer:
|
65 |
full_message += text
|
66 |
yield full_message
|
67 |
-
|
|
|
68 |
gr.Markdown(DESCRIPTION)
|
69 |
gr.DuplicateButton(value="Duplicate Space for private use", elem_id="duplicate-button")
|
70 |
|
@@ -143,4 +144,7 @@ def generate(
|
|
143 |
)
|
144 |
|
145 |
if __name__ == "__main__":
|
|
|
|
|
|
|
146 |
demo.queue(max_size=20).launch()
|
|
|
3 |
from typing import Iterator
|
4 |
|
5 |
import gradio as gr
|
|
|
6 |
import torch
|
7 |
from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamer
|
8 |
DESCRIPTION = """\
|
|
|
13 |
This space allows you to input text and have the AI complete it. Simply type your text in the input box, click "Complete", and watch as the AI generates a continuation of your text.
|
14 |
|
15 |
You can adjust various parameters such as temperature and top-p sampling to control the generation process.
|
16 |
+
|
17 |
+
Note: You may see a warning about bitsandbytes being compiled without GPU support. This is expected in environments without GPU and does not affect the basic functionality of the demo.
|
18 |
"""
|
19 |
|
20 |
MAX_MAX_NEW_TOKENS = 2048
|
|
|
31 |
torch_dtype=torch.bfloat16,
|
32 |
)
|
33 |
model.eval()
|
|
|
34 |
def generate(
|
35 |
message: str,
|
36 |
max_new_tokens: int = 1024,
|
|
|
64 |
for text in streamer:
|
65 |
full_message += text
|
66 |
yield full_message
|
67 |
+
|
68 |
+
with gr.Blocks(css="style.css", fill_height=True) as demo:
|
69 |
gr.Markdown(DESCRIPTION)
|
70 |
gr.DuplicateButton(value="Duplicate Space for private use", elem_id="duplicate-button")
|
71 |
|
|
|
144 |
)
|
145 |
|
146 |
if __name__ == "__main__":
|
147 |
+
|
148 |
+
demo = gr.Blocks(css="style.css", fill_height=True)
|
149 |
+
|
150 |
demo.queue(max_size=20).launch()
|