Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -1,11 +1,16 @@
|
|
| 1 |
import gradio as gr
|
| 2 |
from huggingface_hub import InferenceClient
|
| 3 |
|
| 4 |
-
""
|
| 5 |
-
For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
|
| 6 |
-
"""
|
| 7 |
-
client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
|
| 8 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
|
| 10 |
def respond(
|
| 11 |
message,
|
|
@@ -27,37 +32,37 @@ def respond(
|
|
| 27 |
|
| 28 |
response = ""
|
| 29 |
|
| 30 |
-
for message in client.
|
| 31 |
-
messages,
|
| 32 |
-
|
| 33 |
stream=True,
|
| 34 |
temperature=temperature,
|
| 35 |
top_p=top_p,
|
| 36 |
):
|
| 37 |
-
token = message.
|
| 38 |
|
| 39 |
response += token
|
| 40 |
yield response
|
| 41 |
|
| 42 |
-
|
| 43 |
-
|
| 44 |
-
"""
|
| 45 |
-
demo = gr.ChatInterface(
|
| 46 |
-
respond,
|
| 47 |
-
additional_inputs=[
|
| 48 |
-
gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
|
| 49 |
-
gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
|
| 50 |
-
gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
|
| 51 |
-
gr.Slider(
|
| 52 |
-
minimum=0.1,
|
| 53 |
-
maximum=1.0,
|
| 54 |
-
value=0.95,
|
| 55 |
-
step=0.05,
|
| 56 |
-
label="Top-p (nucleus sampling)",
|
| 57 |
-
),
|
| 58 |
-
],
|
| 59 |
-
)
|
| 60 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 61 |
|
| 62 |
if __name__ == "__main__":
|
| 63 |
demo.launch()
|
|
|
|
| 1 |
import gradio as gr
|
| 2 |
from huggingface_hub import InferenceClient
|
| 3 |
|
| 4 |
+
client = InferenceClient("microsoft/Phi-3.5-mini-instruct")
|
|
|
|
|
|
|
|
|
|
| 5 |
|
| 6 |
+
# Predefined prompts
|
| 7 |
+
prompts = [
|
| 8 |
+
"Tell me a joke about programming",
|
| 9 |
+
"Write a short story about a time-traveling robot",
|
| 10 |
+
"Explain quantum computing to a 5-year-old",
|
| 11 |
+
"Create a recipe for the most unusual pizza",
|
| 12 |
+
"Describe an alien civilization's first contact with Earth"
|
| 13 |
+
]
|
| 14 |
|
| 15 |
def respond(
|
| 16 |
message,
|
|
|
|
| 32 |
|
| 33 |
response = ""
|
| 34 |
|
| 35 |
+
for message in client.text_generation(
|
| 36 |
+
prompt="\n".join([m["content"] for m in messages]),
|
| 37 |
+
max_new_tokens=max_tokens,
|
| 38 |
stream=True,
|
| 39 |
temperature=temperature,
|
| 40 |
top_p=top_p,
|
| 41 |
):
|
| 42 |
+
token = message.token.text
|
| 43 |
|
| 44 |
response += token
|
| 45 |
yield response
|
| 46 |
|
| 47 |
+
def update_textbox(prompt):
|
| 48 |
+
return gr.Textbox.update(value=prompt)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 49 |
|
| 50 |
+
with gr.Blocks() as demo:
|
| 51 |
+
chatbot = gr.Chatbot()
|
| 52 |
+
msg = gr.Textbox()
|
| 53 |
+
clear = gr.ClearButton([msg, chatbot])
|
| 54 |
+
|
| 55 |
+
with gr.Accordion("Advanced options", open=False):
|
| 56 |
+
system = gr.Textbox(value="You are a friendly Chatbot.", label="System message")
|
| 57 |
+
max_tokens = gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens")
|
| 58 |
+
temperature = gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature")
|
| 59 |
+
top_p = gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)")
|
| 60 |
+
|
| 61 |
+
prompt_dropdown = gr.Dropdown(choices=prompts, label="Select a premade prompt")
|
| 62 |
+
prompt_dropdown.change(update_textbox, inputs=[prompt_dropdown], outputs=[msg])
|
| 63 |
+
|
| 64 |
+
msg.submit(respond, [msg, chatbot, system, max_tokens, temperature, top_p], chatbot)
|
| 65 |
+
clear.click(lambda: None, None, chatbot, queue=False)
|
| 66 |
|
| 67 |
if __name__ == "__main__":
|
| 68 |
demo.launch()
|