pierwszy commit
Browse files- a.py +8 -0
- app.py +72 -4
- requirements.txt +6 -0
a.py
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from gradio_client import Client
|
2 |
+
|
3 |
+
client = Client("wiklif/my-api")
|
4 |
+
result = client.predict(
|
5 |
+
prompt="Jakie są 3 największe kraje? Pisz po polsku.",
|
6 |
+
api_name="/chat"
|
7 |
+
)
|
8 |
+
print(result)
|
app.py
CHANGED
@@ -1,7 +1,75 @@
|
|
|
|
1 |
import gradio as gr
|
|
|
|
|
2 |
|
3 |
-
|
4 |
-
return "Hello " + name + "!!"
|
5 |
|
6 |
-
|
7 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import spaces
|
2 |
import gradio as gr
|
3 |
+
import transformers
|
4 |
+
import torch
|
5 |
|
6 |
+
model_id = "meta-llama/Meta-Llama-3.1-8B"
|
|
|
7 |
|
8 |
+
@spaces.GPU(duration=60)
|
9 |
+
def load_pipeline():
|
10 |
+
return transformers.pipeline(
|
11 |
+
"text-generation",
|
12 |
+
model=model_id,
|
13 |
+
model_kwargs={"torch_dtype": torch.bfloat16},
|
14 |
+
device_map="auto"
|
15 |
+
)
|
16 |
+
|
17 |
+
pipeline = load_pipeline()
|
18 |
+
|
19 |
+
def generate_response(chat, kwargs):
|
20 |
+
output = pipeline(chat, **kwargs)[0]['generated_text']
|
21 |
+
if output.endswith("</s>"):
|
22 |
+
output = output[:-4]
|
23 |
+
return output
|
24 |
+
|
25 |
+
def function(prompt, history=[]):
|
26 |
+
chat = "<s>"
|
27 |
+
for user_prompt, bot_response in history:
|
28 |
+
chat += f"[INST] {user_prompt} [/INST] {bot_response}</s> <s>"
|
29 |
+
chat += f"[INST] {prompt} [/INST]"
|
30 |
+
|
31 |
+
kwargs = dict(
|
32 |
+
max_new_tokens=4096,
|
33 |
+
do_sample=True,
|
34 |
+
temperature=0.5,
|
35 |
+
top_p=0.95,
|
36 |
+
repetition_penalty=1.0,
|
37 |
+
seed=1337
|
38 |
+
)
|
39 |
+
|
40 |
+
try:
|
41 |
+
output = generate_response(chat, kwargs)
|
42 |
+
return output
|
43 |
+
except:
|
44 |
+
return ''
|
45 |
+
|
46 |
+
# Interfejs Gradio
|
47 |
+
interface = gr.ChatInterface(
|
48 |
+
fn=function,
|
49 |
+
chatbot=gr.Chatbot(
|
50 |
+
avatar_images=None,
|
51 |
+
container=False,
|
52 |
+
show_copy_button=True,
|
53 |
+
layout='bubble',
|
54 |
+
render_markdown=True,
|
55 |
+
line_breaks=True
|
56 |
+
),
|
57 |
+
css='h1 {font-size:22px;} h2 {font-size:20px;} h3 {font-size:18px;} h4 {font-size:16px;}',
|
58 |
+
autofocus=True,
|
59 |
+
fill_height=True,
|
60 |
+
analytics_enabled=False,
|
61 |
+
submit_btn='Chat',
|
62 |
+
stop_btn=None,
|
63 |
+
retry_btn=None,
|
64 |
+
undo_btn=None,
|
65 |
+
clear_btn=None
|
66 |
+
)
|
67 |
+
|
68 |
+
# API endpoint
|
69 |
+
def api_predict(prompt):
|
70 |
+
return function(prompt)
|
71 |
+
|
72 |
+
interface.launch(show_api=True, share=True)
|
73 |
+
|
74 |
+
# Dodanie endpointu API
|
75 |
+
gr.Interface(fn=api_predict, inputs="text", outputs="text").launch(share=True)
|
requirements.txt
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
huggingface_hub
|
2 |
+
gradio
|
3 |
+
numpy<2
|
4 |
+
torch
|
5 |
+
transformers
|
6 |
+
bitsandbytes
|