Update app.py
Browse files
app.py
CHANGED
@@ -1,23 +1,21 @@
|
|
1 |
-
import gradio as gr
|
2 |
from huggingface_hub import InferenceClient
|
|
|
|
|
|
|
|
|
|
|
3 |
|
4 |
-
# Initialize the client with your desired model
|
5 |
-
client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
|
6 |
|
7 |
-
# Define the system prompt as an AI Dermatologist
|
8 |
def format_prompt(message, history):
|
9 |
-
|
10 |
-
|
11 |
-
prompt += "[INST]
|
12 |
-
|
13 |
-
|
14 |
-
|
15 |
-
prompt += f"[INST] {message} [/INST]"
|
16 |
-
return prompt
|
17 |
|
18 |
-
# Function to generate responses with the AI Dermatologist context
|
19 |
def generate(
|
20 |
-
prompt, history, temperature=0.9, max_new_tokens=256, top_p=0.95, repetition_penalty=1.0
|
21 |
):
|
22 |
temperature = float(temperature)
|
23 |
if temperature < 1e-2:
|
@@ -35,9 +33,7 @@ def generate(
|
|
35 |
|
36 |
formatted_prompt = format_prompt(prompt, history)
|
37 |
|
38 |
-
stream = client.text_generation(
|
39 |
-
formatted_prompt, **generate_kwargs, stream=True, details=True, return_full_text=False
|
40 |
-
)
|
41 |
output = ""
|
42 |
|
43 |
for response in stream:
|
@@ -45,8 +41,8 @@ def generate(
|
|
45 |
yield output
|
46 |
return output
|
47 |
|
48 |
-
|
49 |
-
additional_inputs
|
50 |
gr.Slider(
|
51 |
label="Temperature",
|
52 |
value=0.9,
|
@@ -85,10 +81,10 @@ additional_inputs = [
|
|
85 |
)
|
86 |
]
|
87 |
|
88 |
-
|
89 |
gr.ChatInterface(
|
90 |
fn=generate,
|
91 |
chatbot=gr.Chatbot(show_label=False, show_share_button=False, show_copy_button=True, likeable=True, layout="panel"),
|
92 |
additional_inputs=additional_inputs,
|
93 |
-
title="
|
94 |
).launch(show_api=False)
|
|
|
|
|
1 |
from huggingface_hub import InferenceClient
|
2 |
+
import gradio as gr
|
3 |
+
|
4 |
+
client = InferenceClient(
|
5 |
+
"mistralai/Mistral-7B-Instruct-v0.3"
|
6 |
+
)
|
7 |
|
|
|
|
|
8 |
|
|
|
9 |
def format_prompt(message, history):
|
10 |
+
prompt = "<s>"
|
11 |
+
for user_prompt, bot_response in history:
|
12 |
+
prompt += f"[INST] {user_prompt} [/INST]"
|
13 |
+
prompt += f" {bot_response}</s> "
|
14 |
+
prompt += f"[INST] {message} [/INST]"
|
15 |
+
return prompt
|
|
|
|
|
16 |
|
|
|
17 |
def generate(
|
18 |
+
prompt, history, temperature=0.9, max_new_tokens=256, top_p=0.95, repetition_penalty=1.0,
|
19 |
):
|
20 |
temperature = float(temperature)
|
21 |
if temperature < 1e-2:
|
|
|
33 |
|
34 |
formatted_prompt = format_prompt(prompt, history)
|
35 |
|
36 |
+
stream = client.text_generation(formatted_prompt, **generate_kwargs, stream=True, details=True, return_full_text=False)
|
|
|
|
|
37 |
output = ""
|
38 |
|
39 |
for response in stream:
|
|
|
41 |
yield output
|
42 |
return output
|
43 |
|
44 |
+
|
45 |
+
additional_inputs=[
|
46 |
gr.Slider(
|
47 |
label="Temperature",
|
48 |
value=0.9,
|
|
|
81 |
)
|
82 |
]
|
83 |
|
84 |
+
|
85 |
gr.ChatInterface(
|
86 |
fn=generate,
|
87 |
chatbot=gr.Chatbot(show_label=False, show_share_button=False, show_copy_button=True, likeable=True, layout="panel"),
|
88 |
additional_inputs=additional_inputs,
|
89 |
+
title="""Chatbot"""
|
90 |
).launch(show_api=False)
|