Spaces:
Runtime error
Runtime error
File size: 7,285 Bytes
77ff0b3 db8c43f 77ff0b3 fdf1396 77ff0b3 9d3530f 77ff0b3 137041a 25a1a26 8ed0878 25a1a26 2348315 77ff0b3 e0ec530 137041a e0ec530 77ff0b3 e0ec530 ed54557 b860d5b 9d3530f 95b9be1 77ff0b3 db8c43f 77ff0b3 137041a 77ff0b3 137041a 77ff0b3 db8c43f e0ec530 f5c6e14 77ff0b3 137041a 77ff0b3 137041a 77ff0b3 b80d9fe 77ff0b3 85f3d2d 77ff0b3 85f3d2d 77ff0b3 38cd733 77ff0b3 7edc905 95aaf9f 77ff0b3 dc9e813 77ff0b3 2736189 77ff0b3 2736189 77ff0b3 2736189 77ff0b3 baa7023 77ff0b3 1fe527f |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 |
# -*- coding:utf-8 -*-
import os
import logging
import sys
import gradio as gr
import torch
import gc
from app_modules.utils import *
from app_modules.presets import *
from app_modules.overwrites import *
logging.basicConfig(
level=logging.DEBUG,
format="%(asctime)s [%(levelname)s] [%(filename)s:%(lineno)d] %(message)s",
)
base_model = "decapoda-research/llama-7b-hf"
adapter_model = "project-baize/baize-lora-7B"
tokenizer,model,device = load_tokenizer_and_model(base_model,adapter_model)
total_count = 0
def predict(text,
chatbot,
history,
top_p,
temperature,
max_length_tokens,
max_context_length_tokens,):
if text=="":
yield chatbot,history,"Empty context."
return
try:
model
except:
yield [[text,"No Model Found"]],[],"No Model Found"
return
inputs = generate_prompt_with_history(text,history,tokenizer,max_length=max_context_length_tokens)
if inputs is None:
yield chatbot,history,"Input too long."
return
else:
prompt,inputs=inputs
begin_length = len(prompt)
input_ids = inputs["input_ids"][:,-max_context_length_tokens:].to(device)
torch.cuda.empty_cache()
global total_count
total_count += 1
print(total_count)
if total_count % 50 == 0 :
os.system("nvidia-smi")
with torch.no_grad():
for x in greedy_search(input_ids,model,tokenizer,stop_words=["[|Human|]", "[|AI|]"],max_length=max_length_tokens,temperature=temperature,top_p=top_p):
if is_stop_word_or_prefix(x,["[|Human|]", "[|AI|]"]) is False:
if "[|Human|]" in x:
x = x[:x.index("[|Human|]")].strip()
if "[|AI|]" in x:
x = x[:x.index("[|AI|]")].strip()
x = x.strip()
a, b= [[y[0],convert_to_markdown(y[1])] for y in history]+[[text, convert_to_markdown(x)]],history + [[text,x]]
yield a, b, "Generating..."
if shared_state.interrupted:
shared_state.recover()
try:
yield a, b, "Stop: Success"
return
except:
pass
del input_ids
gc.collect()
torch.cuda.empty_cache()
#print(text)
#print(x)
#print("="*80)
try:
yield a,b,"Generate: Success"
except:
pass
def retry(
text,
chatbot,
history,
top_p,
temperature,
max_length_tokens,
max_context_length_tokens,
):
logging.info("Retry...")
if len(history) == 0:
yield chatbot, history, f"Empty context"
return
chatbot.pop()
inputs = history.pop()[0]
for x in predict(inputs,chatbot,history,top_p,temperature,max_length_tokens,max_context_length_tokens):
yield x
gr.Chatbot.postprocess = postprocess
with open("assets/custom.css", "r", encoding="utf-8") as f:
customCSS = f.read()
with gr.Blocks(css=customCSS, theme=small_and_beautiful_theme) as demo:
history = gr.State([])
user_question = gr.State("")
with gr.Row():
gr.HTML(title)
status_display = gr.Markdown("Success", elem_id="status_display")
gr.Markdown(description_top)
with gr.Row(scale=1).style(equal_height=True):
with gr.Column(scale=5):
with gr.Row(scale=1):
chatbot = gr.Chatbot(elem_id="chuanhu_chatbot").style(height="100%")
with gr.Row(scale=1):
with gr.Column(scale=12):
user_input = gr.Textbox(
show_label=False, placeholder="Enter text"
).style(container=False)
with gr.Column(min_width=70, scale=1):
submitBtn = gr.Button("Send")
with gr.Row(scale=1):
emptyBtn = gr.Button(
"🧹 New Conversation",
)
retryBtn = gr.Button("🔄 Regenerate")
delLastBtn = gr.Button("🗑️ Remove Last Turn")
with gr.Column():
with gr.Column(min_width=50, scale=1):
with gr.Tab(label="Parameter Setting"):
gr.Markdown("# Parameters")
top_p = gr.Slider(
minimum=-0,
maximum=1.0,
value=0.95,
step=0.05,
interactive=True,
label="Top-p",
)
temperature = gr.Slider(
minimum=0.1,
maximum=2.0,
value=1,
step=0.1,
interactive=True,
label="Temperature",
)
max_length_tokens = gr.Slider(
minimum=0,
maximum=512,
value=256,
step=8,
interactive=True,
label="Max Generation Tokens",
)
max_context_length_tokens = gr.Slider(
minimum=0,
maximum=2048,
value=2048,
step=128,
interactive=True,
label="Max History Tokens",
)
gr.Markdown(description)
predict_args = dict(
fn=predict,
inputs=[
user_question,
chatbot,
history,
top_p,
temperature,
max_length_tokens,
max_context_length_tokens,
],
outputs=[chatbot, history, status_display],
show_progress=True,
)
retry_args = dict(
fn=retry,
inputs=[
user_input,
chatbot,
history,
top_p,
temperature,
max_length_tokens,
max_context_length_tokens,
],
outputs=[chatbot, history, status_display],
show_progress=True,
)
reset_args = dict(
fn=reset_textbox, inputs=[], outputs=[user_input, status_display]
)
# Chatbot
transfer_input_args = dict(
fn=transfer_input, inputs=[user_input], outputs=[user_question, user_input, submitBtn], show_progress=True
)
predict_event1 = user_input.submit(**transfer_input_args).then(**predict_args)
predict_event2 = submitBtn.click(**transfer_input_args).then(**predict_args)
emptyBtn.click(
reset_state,
outputs=[chatbot, history, status_display],
show_progress=True,
)
emptyBtn.click(**reset_args)
predict_event3 = retryBtn.click(**retry_args)
delLastBtn.click(
delete_last_conversation,
[chatbot, history],
[chatbot, history, status_display],
show_progress=True,
)
#cancelBtn.click(
# cancel_outputing, [], [status_display],
# cancels=[
# predict_event1,predict_event2,predict_event3
# ]
#)
demo.title = "Baize"
demo.queue(concurrency_count=1).launch() |