Spaces:
Running
Running
File size: 3,519 Bytes
f8f9857 b33166a f234576 f8f9857 2ddacc1 f8f9857 db111cc f8f9857 2ddacc1 b33166a f8f9857 b33166a f8f9857 b33166a f8f9857 b33166a f8f9857 b33166a f8f9857 b33166a f8f9857 b33166a f8f9857 b33166a f8f9857 b33166a f8f9857 b33166a f8f9857 b33166a |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 |
# ADD DISCLAIMERS
# AND LOGGING
import gradio as gr
import os
os.system("apt install libopenblas-dev")
os.system("make clean && LLAMA_OPENBLAS=1 make")
os.system('CMAKE_ARGS="-DLLAMA_BLAS=ON -DLLAMA_BLAS_VENDOR=OpenBLAS" pip install llama-cpp-python transformers')
from llama_cpp import Llama
from transformers import AutoTokenizer
model_id = "Elijahbodden/eliGPTv1.1"
# MODEL
model = Llama.from_pretrained(
repo_id=model_id,
filename="eliGPTv1.1-unsloth.Q5_K_M.gguf",
verbose=True,
n_threads = 2,
n_threads_batch = 2,
n_ctx=8192,
)
# TOKENIZER AND TEMPLATE
tokenizer = AutoTokenizer.from_pretrained(model_id)
sys_prompt = """SUMMARY - ELIJAH:
Age: 16
Interests: space flight, cybernetics, consciousness, philosophy, psychonautism, biotech, AI
Likes: thinking and learning, building stuff, interesting conversations, red hot chili peppers and techno, humanism
Traits: incredibly intelligent, funny, interesting, caffeine fiend, very ambitious, militant atheist, self-taught/homeschooled, casual
Aspirations: creating transhumanist utopia, understanding the universe more, becoming smarter and better\n"""
your_name = "elijah"
custom_template = \
f"{{% if {'sys_prompt' in locals()} %}}"\
"{{ '<|im_start|>system\n" + sys_prompt + "<|im_end|>\n' }}"\
"{% endif %}"\
"{% for message in messages %}"\
"{% if message['role'] == 'user' %}"\
"{{'<|im_start|>user\n' + message['content'] + '<|im_end|>\n'}}"\
"{% elif message['role'] == 'assistant' %}"\
"{{'<|im_start|>" + your_name + "\n' + message['content'] + eos_token + '<|im_end|>\n' }}"\
"{% else %}"\
"{{ '<|im_start|>system\n' + message['content'] + '<|im_end|>\n' }}"\
"{% endif %}"\
"{% endfor %}"\
"{{ '<|im_start|>" + your_name + "\n' }}"\
tokenizer.chat_template = custom_template
def respond(
message,
history: list[tuple[str, str]],
max_tokens,
temperature,
mirostat_tau,
mirostat_eta,
):
for val in history:
if val[0]:
messages.append({"from": "human", "content": val[0]})
if val[1]:
messages.append({"from": "gpt", "content": val[1]})
messages.append({"from": "human", "content": message})
response = ""
tokenizer.apply_chat_template(messages, tokenize=False)
for message in model.create_completion(
convo,
temperature=0.75,
stream=True,
stop=["<|im_end|>"],
mirostat_mode=1,
mirostat_tau=mirostat_tau,
mirostat_eta=mirostat_eta,
max_tokens=128
):
token = message["choices"][0]["text"]
response += token
yield response
"""
For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
"""
demo = gr.ChatInterface(
respond,
additional_inputs=[
gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
gr.Slider(
minimum=0.0,
maximum=10.0,
value=3.0,
step=0.5,
label="Mirostat tau",
),
gr.Slider(
minimum=0.0,
maximum=1.0,
value=0.1,
step=0.01,
label="Mirostat eta",
),
],
)
if __name__ == "__main__":
demo.launch() |