|
import os |
|
import time |
|
import spaces |
|
import torch |
|
from transformers import AutoModelForCausalLM, AutoTokenizer |
|
import gradio as gr |
|
from threading import Thread |
|
|
|
HF_TOKEN = os.environ.get("HF_TOKEN", None) |
|
MODEL = "evabyte/EvaByte-SFT" |
|
MODEL_BASE = "evabyte/EvaByte" |
|
|
|
TITLE = "<h1><center>EvaByte</center></h1>" |
|
|
|
PLACEHOLDER = """ |
|
<center> |
|
<p>Hi! How can I help you today?</p> |
|
</center> |
|
""" |
|
|
|
|
|
CSS = """ |
|
.duplicate-button { |
|
margin: auto !important; |
|
color: white !important; |
|
background: black !important; |
|
border-radius: 100vh !important; |
|
} |
|
h3 { |
|
text-align: center; |
|
} |
|
""" |
|
|
|
device = "cuda" |
|
|
|
tokenizer = AutoTokenizer.from_pretrained(MODEL, trust_remote_code=True) |
|
model = AutoModelForCausalLM.from_pretrained( |
|
MODEL, |
|
torch_dtype=torch.bfloat16, |
|
device_map="auto", |
|
trust_remote_code=True).eval().to(device) |
|
|
|
@spaces.GPU() |
|
def stream_chat( |
|
message: str, |
|
history: list, |
|
system_prompt: str, |
|
temperature: float = 0.8, |
|
max_new_tokens: int = 512, |
|
top_p: float = 1.0, |
|
): |
|
print(f'message: {message}') |
|
print(f'history: {history}') |
|
|
|
conversation = [ |
|
{"role": "system", "content": system_prompt} |
|
] |
|
for prompt, answer in history: |
|
conversation.extend([ |
|
{"role": "user", "content": prompt}, |
|
{"role": "assistant", "content": answer}, |
|
]) |
|
|
|
conversation.append({"role": "user", "content": message}) |
|
|
|
input_ids = tokenizer.apply_chat_template(conversation, add_generation_prompt=True, return_tensors="pt").to(device) |
|
|
|
gen_out = model.multi_byte_generate( |
|
input_ids=input_ids, |
|
max_new_tokens = max_new_tokens, |
|
do_sample = False if temperature == 0 else True, |
|
top_p = top_p, |
|
temperature = temperature, |
|
) |
|
|
|
response = tokenizer.decode( |
|
gen_out[0][input_ids.shape[1]:], |
|
skip_special_tokens=False, |
|
clean_up_tokenization_spaces=False |
|
) |
|
|
|
for i in range(len(response)): |
|
time.sleep(0.02) |
|
yield response[: i + 1] |
|
|
|
chatbot = gr.Chatbot(height=600, placeholder=PLACEHOLDER) |
|
|
|
with gr.Blocks(css=CSS, theme="soft") as demo: |
|
gr.HTML(TITLE) |
|
gr.DuplicateButton(value="Duplicate Space for private use", elem_classes="duplicate-button") |
|
gr.ChatInterface( |
|
fn=stream_chat, |
|
chatbot=chatbot, |
|
fill_height=True, |
|
additional_inputs_accordion=gr.Accordion(label="⚙️ Parameters", open=False, render=False), |
|
additional_inputs=[ |
|
gr.Textbox( |
|
value="You are a helpful assistant.", |
|
label="System Prompt", |
|
lines=5, |
|
render=False, |
|
), |
|
gr.Slider( |
|
minimum=0, |
|
maximum=1, |
|
step=0.1, |
|
value=0.8, |
|
label="Temperature", |
|
render=False, |
|
), |
|
gr.Slider( |
|
minimum=128, |
|
maximum=8192, |
|
step=1, |
|
value= 512, |
|
label="Max new tokens", |
|
render=False, |
|
), |
|
gr.Slider( |
|
minimum=0.0, |
|
maximum=1.0, |
|
step=0.1, |
|
value=1.0, |
|
label="top_p", |
|
render=False, |
|
), |
|
], |
|
examples=[ |
|
["Help me study vocabulary: write a sentence for me to fill in the blank, and I'll try to pick the correct option."], |
|
["What are 5 creative things I could do with my kids' art? I don't want to throw them away, but it's also so much clutter."], |
|
["Tell me a random fun fact about the Roman Empire."], |
|
["Show me a code snippet of a website's sticky header in CSS and JavaScript."], |
|
], |
|
cache_examples=False, |
|
) |
|
|
|
|
|
if __name__ == "__main__": |
|
demo.launch() |
