File size: 3,331 Bytes
2622840 30d64ae 2622840 dad512b 2622840 74dd3bf dad512b 2622840 dad512b 2622840 dad512b 2622840 dad512b 2622840 dad512b 2622840 |
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 |
import gradio as gr
import os
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig, GenerationConfig
model_path = os.environ.get("HF_REPO_ID")
access_token = os.environ.get("HF_TOKEN")
tokenizer = AutoTokenizer.from_pretrained(model_path, token=access_token)
bnb_config = BitsAndBytesConfig(
load_in_4bit=True,
# load_in_8bit=use_8_bit,
bnb_4bit_quant_type="nf4",
bnb_4bit_compute_dtype=getattr(torch, "bfloat16"),
bnb_4bit_use_double_quant=True,
)
model = AutoModelForCausalLM.from_pretrained(model_path, token=access_token,
quantization_config=bnb_config,
torch_dtype=torch.float16,
# attn_implementation="flash_attention_2",
device_map='auto')
if torch.cuda.is_available():
device = "cuda"
else:
device = "cpu"
def generate(
question,
context=None,
temperature=0.7,
top_p=0.7,
top_k=40,
num_beams=4,
max_new_tokens=256,):
prompt = f"### CONTEXT:\n{context}\n\n### QUESTION:\n{question}\n\n### ANSWER:"
inputs = tokenizer(prompt, return_tensors="pt")
input_ids = inputs["input_ids"].to(device)
generation_config = GenerationConfig(
temperature=temperature,
top_p=top_p,
top_k=top_k,
num_beams=num_beams,
)
# with torch.autocast("cuda"):
with torch.no_grad():
generation_output = model.generate(
input_ids=input_ids,
generation_config=generation_config,
return_dict_in_generate=True,
output_scores=True,
max_new_tokens=max_new_tokens,
)
seq = generation_output.sequences[0]
output = tokenizer.decode(seq)
return output
"""
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
"""
def respond(
message,
history: list[tuple[str, str]],
system_message,
max_tokens,
temperature,
top_p,
):
context = ""
for chat in history:
context += f"রোগী: {chat[0]}\nথেরাপিস্ট: {chat[1]}\n"
answer = generate(message, context,
temperature=temperature,
top_p=top_p,
max_new_tokens=max_tokens).split('### ANSWER:')[1]
if '</s>' in answer:
answer = answer.split('</s>')[0].strip()
return answer
"""
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=256, 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=0.1,
maximum=1.0,
value=0.7,
step=0.05,
label="Top-p (nucleus sampling)",
),
],
)
if __name__ == "__main__":
demo.launch() |