forked from shibing624/MedicalGPT
-
Notifications
You must be signed in to change notification settings - Fork 0
/
gradio_demo.py
209 lines (188 loc) · 6.92 KB
/
gradio_demo.py
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
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
# -*- coding: utf-8 -*-
"""
@author:XuMing(xuming624@qq.com)
@description:
pip install gradio
pip install mdtex2html
"""
import argparse
import os
import gradio as gr
import mdtex2html
import torch
from peft import PeftModel
from transformers import (
AutoModel,
AutoTokenizer,
AutoModelForCausalLM,
BloomForCausalLM,
BloomTokenizerFast,
LlamaTokenizer,
LlamaForCausalLM,
)
from supervised_finetuning import get_conv_template
MODEL_CLASSES = {
"bloom": (BloomForCausalLM, BloomTokenizerFast),
"chatglm": (AutoModel, AutoTokenizer),
"llama": (LlamaForCausalLM, LlamaTokenizer),
"baichuan": (AutoModelForCausalLM, AutoTokenizer),
"auto": (AutoModelForCausalLM, AutoTokenizer),
}
@torch.inference_mode()
def generate_answer(
model,
tokenizer,
prompt,
device,
max_new_tokens=256,
temperature=0.2,
top_k=40,
top_p=0.9,
do_sample=True,
num_beams=1,
repetition_penalty=1.3,
context_len=2048,
):
generation_config = dict(
max_new_tokens=max_new_tokens,
temperature=temperature,
top_k=top_k,
top_p=top_p,
do_sample=do_sample,
num_beams=num_beams,
repetition_penalty=repetition_penalty,
)
input_ids = tokenizer(prompt).input_ids
max_src_len = context_len - max_new_tokens - 8
input_ids = input_ids[-max_src_len:]
generation_output = model.generate(
input_ids=torch.as_tensor([input_ids]).to(device),
**generation_config,
)
output_ids = generation_output[0]
output = tokenizer.decode(output_ids, skip_special_tokens=False)
stop_str = tokenizer.eos_token
l_prompt = len(tokenizer.decode(input_ids, skip_special_tokens=False))
pos = output.rfind(stop_str, l_prompt)
if pos != -1:
output = output[l_prompt:pos]
else:
output = output[l_prompt:]
return output
def main():
parser = argparse.ArgumentParser()
parser.add_argument('--model_type', default=None, type=str, required=True)
parser.add_argument('--base_model', default=None, type=str, required=True)
parser.add_argument('--lora_model', default="", type=str, help="If None, perform inference on the base model")
parser.add_argument('--tokenizer_path', default=None, type=str)
parser.add_argument('--template_name', default="alpaca", type=str, help="Prompt template name")
parser.add_argument('--gpus', default="0", type=str)
parser.add_argument('--only_cpu', action='store_true', help='only use CPU for inference')
parser.add_argument('--resize_emb', action='store_true', help='Whether to resize model token embeddings')
args = parser.parse_args()
if args.only_cpu is True:
args.gpus = ""
os.environ["CUDA_VISIBLE_DEVICES"] = args.gpus
def postprocess(self, y):
if y is None:
return []
for i, (message, response) in enumerate(y):
y[i] = (
None if message is None else mdtex2html.convert((message)),
None if response is None else mdtex2html.convert(response),
)
return y
gr.Chatbot.postprocess = postprocess
load_type = torch.float16
if torch.cuda.is_available():
device = torch.device(0)
else:
device = torch.device('cpu')
if args.tokenizer_path is None:
args.tokenizer_path = args.base_model
model_class, tokenizer_class = MODEL_CLASSES[args.model_type]
tokenizer = tokenizer_class.from_pretrained(args.tokenizer_path, trust_remote_code=True)
base_model = model_class.from_pretrained(
args.base_model,
load_in_8bit=False,
torch_dtype=load_type,
low_cpu_mem_usage=True,
device_map='auto',
trust_remote_code=True,
)
if args.resize_emb:
model_vocab_size = base_model.get_input_embeddings().weight.size(0)
tokenzier_vocab_size = len(tokenizer)
print(f"Vocab of the base model: {model_vocab_size}")
print(f"Vocab of the tokenizer: {tokenzier_vocab_size}")
if model_vocab_size != tokenzier_vocab_size:
print("Resize model embeddings to fit tokenizer")
base_model.resize_token_embeddings(tokenzier_vocab_size)
if args.lora_model:
model = PeftModel.from_pretrained(base_model, args.lora_model, torch_dtype=load_type, device_map='auto')
print("loaded lora model")
else:
model = base_model
if device == torch.device('cpu'):
model.float()
model.eval()
def reset_user_input():
return gr.update(value='')
def reset_state():
return [], []
conv = get_conv_template(args.template_name)
def predict(
input,
chatbot,
history,
max_new_tokens,
temperature,
top_p
):
now_input = input
chatbot.append((input, ""))
history = history or []
conv.append_message(conv.roles[0], now_input)
conv.append_message(conv.roles[1], '')
prompt = conv.get_prompt()
output = generate_answer(
model,
tokenizer,
prompt,
device,
max_new_tokens=max_new_tokens,
temperature=temperature,
top_p=top_p,
)
conv.messages[-1][-1] = output.strip()
history.append((now_input, output))
chatbot[-1] = (now_input, output)
return chatbot, history
with gr.Blocks() as demo:
gr.HTML("""<h1 align="center">MedicalGPT</h1>""")
gr.Markdown(
"> 为了促进医疗行业大模型的开放研究,本项目开源了MedicalGPT医疗大模型")
chatbot = gr.Chatbot()
with gr.Row():
with gr.Column(scale=4):
with gr.Column(scale=12):
user_input = gr.Textbox(show_label=False, placeholder="Input...", lines=10).style(
container=False)
with gr.Column(min_width=32, scale=1):
submitBtn = gr.Button("Submit", variant="primary")
with gr.Column(scale=1):
emptyBtn = gr.Button("Clear History")
max_length = gr.Slider(
0, 4096, value=128, step=1.0, label="Maximum length", interactive=True)
top_p = gr.Slider(0, 1, value=0.8, step=0.01,
label="Top P", interactive=True)
temperature = gr.Slider(
0, 1, value=0.7, step=0.01, label="Temperature", interactive=True)
history = gr.State([])
submitBtn.click(predict, [user_input, chatbot, history, max_length, temperature, top_p], [chatbot, history],
show_progress=True)
submitBtn.click(reset_user_input, [], [user_input])
emptyBtn.click(reset_state, outputs=[chatbot, history], show_progress=True)
demo.queue().launch(share=False, inbrowser=True, server_name='0.0.0.0', server_port=8081)
if __name__ == '__main__':
main()