Spaces:
Running
on
Zero
Running
on
Zero
File size: 53,406 Bytes
d3c19b3 8889bbb bdbbfdb 8889bbb bdbbfdb 8889bbb d3c19b3 8889bbb 43147aa 8889bbb bdbbfdb 8889bbb d3c19b3 8889bbb 43147aa 8889bbb bdbbfdb 8889bbb d3c19b3 8889bbb 43147aa 8889bbb bdbbfdb 8889bbb |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 |
try:
import spaces
def maybe_spaces_gpu(fn):
fn = spaces.GPU(fn)
return fn
except ModuleNotFoundError:
print(f'Cannot import hf `spaces` with `import spaces`.')
def maybe_spaces_gpu(fn):
return fn
import os
from gradio.themes import ThemeClass as Theme
import numpy as np
import argparse
import gradio as gr
from typing import Any, Iterator
from typing import Iterator, List, Optional, Tuple
import filelock
import glob
import json
import time
from gradio.routes import Request
from gradio.utils import SyncToAsyncIterator, async_iteration
from gradio.helpers import special_args
import anyio
from typing import AsyncGenerator, Callable, Literal, Union, cast, Generator
from gradio_client.documentation import document, set_documentation_group
from gradio.components import Button, Component
from gradio.events import Dependency, EventListenerMethod
from typing import List, Optional, Union, Dict, Tuple
from tqdm.auto import tqdm
from huggingface_hub import snapshot_download
from gradio.components.base import Component
from .base_demo import register_demo, get_demo_class, BaseDemo
from .chat_interface import (
SYSTEM_PROMPT,
MODEL_NAME,
MAX_TOKENS,
TEMPERATURE,
CHAT_EXAMPLES,
gradio_history_to_openai_conversations,
gradio_history_to_conversation_prompt,
DATETIME_FORMAT,
get_datetime_string,
chat_response_stream_multiturn_engine,
ChatInterfaceDemo,
format_conversation,
CustomizedChatInterface,
)
from gradio.events import Events
import inspect
from typing import AsyncGenerator, Callable, Literal, Union, cast
import anyio
from gradio_client import utils as client_utils
from gradio_client.documentation import document
from gradio.blocks import Blocks
from gradio.components import (
Button,
Chatbot,
Component,
Markdown,
State,
Textbox,
get_component_instance,
)
from gradio.events import Dependency, on
from gradio.helpers import create_examples as Examples # noqa: N812
from gradio.helpers import special_args
from gradio.layouts import Accordion, Group, Row
from gradio.routes import Request
from gradio.themes import ThemeClass as Theme
from gradio.utils import SyncToAsyncIterator, async_iteration
from ..globals import MODEL_ENGINE
from ..configs import (
USE_PANEL,
IMAGE_TOKEN,
IMAGE_TOKEN_INTERACTIVE,
CHATBOT_HEIGHT,
)
CSS = """
.message-fit {
min-width: 20em;
width: fit-content !important;
}
.message.svelte-1lcyrx4.svelte-1lcyrx4.svelte-1lcyrx4 {
padding-top: 1em;
padding-bottom: 1em;
}
"""
DOC_TEMPLATE = """###
{content}
###
"""
DOC_INSTRUCTION = """Answer the following query exclusively based on the information provided in the document above. \
If the information is not found, please say so instead of making up facts! Remember to answer the question in the same language as the user query!
"""
def undo_history(history):
if len(history) == 0:
return history
if history[-1][-1] is not None:
if history[-1][0] is not None:
history[-1][-1] = None
else:
history = history[:-1]
else:
history = history[:-1]
return history
def undo_history_until_last_assistant_turn(history):
history = undo_history(history)
while len(history) > 0 and history[-1][-1] is None:
history = undo_history(history)
return history, history
class MultiModalChatInterface(CustomizedChatInterface):
def __init__(
self,
fn: Callable,
*,
chatbot: Chatbot | None = None,
textbox: Textbox | None = None,
additional_inputs: str | Component | list[str | Component] | None = None,
additional_inputs_accordion_name: str | None = None,
additional_inputs_accordion: str | Accordion | None = None,
add_multimodal_fn: Callable | None = None,
render_additional_inputs_fn: Callable | None = None,
examples: list[str] | None = None,
cache_examples: bool | None = None,
title: str | None = None,
description: str | None = None,
theme: Theme | str | None = None,
css: str | None = None,
js: str | None = None,
head: str | None = None,
analytics_enabled: bool | None = None,
submit_btn: str | None | Button = "Submit",
stop_btn: str | None | Button = "Stop",
retry_btn: str | None | Button = "🔄 Retry",
undo_btn: str | None | Button = "↩️ Undo",
clear_btn: str | None | Button = "🗑️ Clear",
autofocus: bool = True,
concurrency_limit: int | None | Literal["default"] = "default",
fill_height: bool = True,
):
"""
Parameters:
fn: The function to wrap the chat interface around. Should accept two parameters: a string input message and list of two-element lists of the form [[user_message, bot_message], ...] representing the chat history, and return a string response. See the Chatbot documentation for more information on the chat history format.
chatbot: An instance of the gr.Chatbot component to use for the chat interface, if you would like to customize the chatbot properties. If not provided, a default gr.Chatbot component will be created.
textbox: An instance of the gr.Textbox component to use for the chat interface, if you would like to customize the textbox properties. If not provided, a default gr.Textbox component will be created.
additional_inputs: An instance or list of instances of gradio components (or their string shortcuts) to use as additional inputs to the chatbot. If components are not already rendered in a surrounding Blocks, then the components will be displayed under the chatbot, in an accordion.
additional_inputs_accordion_name: Deprecated. Will be removed in a future version of Gradio. Use the `additional_inputs_accordion` parameter instead.
additional_inputs_accordion: If a string is provided, this is the label of the `gr.Accordion` to use to contain additional inputs. A `gr.Accordion` object can be provided as well to configure other properties of the container holding the additional inputs. Defaults to a `gr.Accordion(label="Additional Inputs", open=False)`. This parameter is only used if `additional_inputs` is provided.
examples: Sample inputs for the function; if provided, appear below the chatbot and can be clicked to populate the chatbot input.
cache_examples: If True, caches examples in the server for fast runtime in examples. The default option in HuggingFace Spaces is True. The default option elsewhere is False.
title: a title for the interface; if provided, appears above chatbot in large font. Also used as the tab title when opened in a browser window.
description: a description for the interface; if provided, appears above the chatbot and beneath the title in regular font. Accepts Markdown and HTML content.
theme: Theme to use, loaded from gradio.themes.
css: Custom css as a string or path to a css file. This css will be included in the demo webpage.
js: Custom js or path to js file to run when demo is first loaded. This javascript will be included in the demo webpage.
head: Custom html to insert into the head of the demo webpage. This can be used to add custom meta tags, scripts, stylesheets, etc. to the page.
analytics_enabled: Whether to allow basic telemetry. If None, will use GRADIO_ANALYTICS_ENABLED environment variable if defined, or default to True.
submit_btn: Text to display on the submit button. If None, no button will be displayed. If a Button object, that button will be used.
stop_btn: Text to display on the stop button, which replaces the submit_btn when the submit_btn or retry_btn is clicked and response is streaming. Clicking on the stop_btn will halt the chatbot response. If set to None, stop button functionality does not appear in the chatbot. If a Button object, that button will be used as the stop button.
retry_btn: Text to display on the retry button. If None, no button will be displayed. If a Button object, that button will be used.
undo_btn: Text to display on the delete last button. If None, no button will be displayed. If a Button object, that button will be used.
clear_btn: Text to display on the clear button. If None, no button will be displayed. If a Button object, that button will be used.
autofocus: If True, autofocuses to the textbox when the page loads.
concurrency_limit: If set, this is the maximum number of chatbot submissions that can be running simultaneously. Can be set to None to mean no limit (any number of chatbot submissions can be running simultaneously). Set to "default" to use the default concurrency limit (defined by the `default_concurrency_limit` parameter in `.queue()`, which is 1 by default).
fill_height: If True, the chat interface will expand to the height of window.
"""
try:
super(gr.ChatInterface, self).__init__(
analytics_enabled=analytics_enabled,
mode="chat_interface",
css=css,
title=title or "Gradio",
theme=theme,
js=js,
head=head,
fill_height=fill_height,
)
except Exception as e:
# Handle old gradio versions without fill_height
super(gr.ChatInterface, self).__init__(
analytics_enabled=analytics_enabled,
mode="chat_interface",
css=css,
title=title or "Gradio",
theme=theme,
js=js,
head=head,
# fill_height=fill_height,
)
self.concurrency_limit = concurrency_limit
self.fn = fn
self.add_multimodal_fn = add_multimodal_fn
self.render_additional_inputs_fn = render_additional_inputs_fn
self.multimodal_inputs = []
self.is_async = inspect.iscoroutinefunction(
self.fn
) or inspect.isasyncgenfunction(self.fn)
self.is_generator = inspect.isgeneratorfunction(
self.fn
) or inspect.isasyncgenfunction(self.fn)
self.examples = examples
if self.space_id and cache_examples is None:
self.cache_examples = True
else:
self.cache_examples = cache_examples or False
self.buttons: list[Button | None] = []
if additional_inputs:
if not isinstance(additional_inputs, list):
additional_inputs = [additional_inputs]
self.additional_inputs = [
get_component_instance(i)
for i in additional_inputs # type: ignore
]
else:
self.additional_inputs = []
if additional_inputs_accordion_name is not None:
print(
"The `additional_inputs_accordion_name` parameter is deprecated and will be removed in a future version of Gradio. Use the `additional_inputs_accordion` parameter instead."
)
self.additional_inputs_accordion_params = {
"label": additional_inputs_accordion_name
}
if additional_inputs_accordion is None:
self.additional_inputs_accordion_params = {
"label": "Additional Inputs",
"open": False,
}
elif isinstance(additional_inputs_accordion, str):
self.additional_inputs_accordion_params = {
"label": additional_inputs_accordion
}
elif isinstance(additional_inputs_accordion, Accordion):
self.additional_inputs_accordion_params = (
additional_inputs_accordion.recover_kwargs(
additional_inputs_accordion.get_config()
)
)
else:
raise ValueError(
f"The `additional_inputs_accordion` parameter must be a string or gr.Accordion, not {type(additional_inputs_accordion)}"
)
with self:
if title:
Markdown(
f"<h1 style='text-align: center; margin-bottom: 1rem'>{self.title}</h1>"
)
if description:
Markdown(description)
if chatbot:
self.chatbot = chatbot.render()
else:
self.chatbot = Chatbot(
label="Chatbot", scale=1, height=200 if fill_height else None
)
with Row():
for btn in [retry_btn, undo_btn, clear_btn]:
if btn is not None:
if isinstance(btn, Button):
btn.render()
elif isinstance(btn, str):
btn = Button(btn, variant="secondary", size="sm")
else:
raise ValueError(
f"All the _btn parameters must be a gr.Button, string, or None, not {type(btn)}"
)
self.buttons.append(btn) # type: ignore
with Group():
with Row():
if textbox:
textbox.container = False
textbox.show_label = False
textbox_ = textbox.render()
assert isinstance(textbox_, Textbox)
self.textbox = textbox_
else:
self.textbox = Textbox(
container=False,
show_label=False,
label="Message",
placeholder="Type a message...",
scale=7,
autofocus=autofocus,
)
if submit_btn is not None:
if isinstance(submit_btn, Button):
submit_btn.render()
elif isinstance(submit_btn, str):
submit_btn = Button(
submit_btn,
variant="primary",
scale=2,
min_width=150,
)
else:
raise ValueError(
f"The submit_btn parameter must be a gr.Button, string, or None, not {type(submit_btn)}"
)
if stop_btn is not None:
if isinstance(stop_btn, Button):
stop_btn.visible = False
stop_btn.render()
elif isinstance(stop_btn, str):
stop_btn = Button(
stop_btn,
variant="stop",
visible=False,
scale=2,
min_width=150,
)
else:
raise ValueError(
f"The stop_btn parameter must be a gr.Button, string, or None, not {type(stop_btn)}"
)
self.num_tokens = Textbox(
container=False,
show_label=False,
label="num_tokens",
placeholder="0 tokens",
scale=1,
interactive=False,
# autofocus=autofocus,
min_width=10
)
self.buttons.extend([submit_btn, stop_btn]) # type: ignore
self.fake_api_btn = Button("Fake API", visible=False)
self.fake_response_textbox = Textbox(label="Response", visible=False)
(
self.retry_btn,
self.undo_btn,
self.clear_btn,
self.submit_btn,
self.stop_btn,
) = self.buttons
any_unrendered_inputs = any(
not inp.is_rendered for inp in self.additional_inputs
)
if self.add_multimodal_fn is not None:
with Row():
self.multimodal_inputs = self.add_multimodal_fn()
if self.additional_inputs and any_unrendered_inputs:
with Accordion(**self.additional_inputs_accordion_params): # type: ignore
if self.render_additional_inputs_fn is not None:
self.render_additional_inputs_fn()
else:
for input_component in self.additional_inputs:
if not input_component.is_rendered:
input_component.render()
else:
if self.additional_inputs and any_unrendered_inputs:
with Accordion(**self.additional_inputs_accordion_params): # type: ignore
if self.render_additional_inputs_fn is not None:
self.render_additional_inputs_fn()
else:
for input_component in self.additional_inputs:
if not input_component.is_rendered:
input_component.render()
if examples:
if self.is_generator:
examples_fn = self._examples_stream_fn
else:
# examples_fn = self._examples_fn
raise NotImplementedError(f'Not streaming not impl')
self.examples_handler = Examples(
examples=examples,
inputs=[self.textbox] + self.multimodal_inputs + self.additional_inputs,
outputs=self.chatbot,
fn=examples_fn,
)
# The example caching must happen after the input components have rendered
if cache_examples:
client_utils.synchronize_async(self.examples_handler.cache)
self.saved_input = State()
self.chatbot_state = (
State(self.chatbot.value) if self.chatbot.value else State([])
)
self._setup_events()
self._setup_api()
def _clear_and_save_textbox(self, message: str, *multimodal_inputs) -> tuple[str, str]:
saved_input = [message] + list(multimodal_inputs)
outputs = [''] + [None] * len(multimodal_inputs)
return outputs + [saved_input]
def _add_inputs_to_history(self, history: List[List[Union[str, None]]], *args):
message = args[0]
multimodal_inputs = args[1:1 + len(self.multimodal_inputs)] if len(args) > 1 else None
if multimodal_inputs is not None:
is_file_exists = [(x is not None and os.path.exists(x)) for x in multimodal_inputs]
if any(is_file_exists):
file_exists = [f for f, ise in zip(multimodal_inputs, is_file_exists) if ise]
if len(file_exists) > 1:
raise gr.Error(f"Cannot have more than 1 multimodal input at a time.")
fname = file_exists[0]
history.append([(fname,), None])
if message is not None and message.strip() != "":
history.append([message, None])
return history
def _display_input(
self, saved_input: List[str], history: List[List[Union[str, None]]]
) -> Tuple[List[List[Union[str, None]]], List[List[list[Union[str, None]]]]]:
# message = saved_input[0]
# multimodal_inputs = saved_input[1:] if len(saved_input) > 1 else None
# # ! If things wrong, return original history and give warning
# if multimodal_inputs is not None:
# is_file_exists = [(x is not None and os.path.exists(x)) for x in multimodal_inputs]
# if any(is_file_exists):
# file_exists = [f for f, ise in zip(multimodal_inputs, is_file_exists) if ise]
# if len(file_exists) > 1:
# raise gr.Error(f"Cannot have more than 1 multimodal input at a time.")
# fname = file_exists[0]
# history.append([(fname,), None])
# if message is not None and message.strip() != "":
# history.append([message, None])
history = self._add_inputs_to_history(history, *saved_input)
return history, history
def _delete_prev_fn(
self, history: list[list[str | None]]
) -> tuple[list[list[str | None]], str, list[list[str | None]]]:
try:
message, _ = history.pop()
except IndexError:
message = ""
saved_input = [message or ""] + [None] * len(self.multimodal_inputs)
return history, saved_input, history
def _setup_events(self) -> None:
from gradio.components import State
has_on = False
try:
from gradio.events import Dependency, EventListenerMethod, on
has_on = True
except ImportError as ie:
has_on = False
submit_fn = self._stream_fn if self.is_generator else self._submit_fn
if not self.is_generator:
raise NotImplementedError(f'should use generator')
if has_on:
# new version
submit_triggers = (
[self.textbox.submit, self.submit_btn.click]
if self.submit_btn
else [self.textbox.submit]
)
submit_event = (
on(
submit_triggers,
self._clear_and_save_textbox,
[self.textbox] + self.multimodal_inputs,
[self.textbox] + self.multimodal_inputs + [self.saved_input],
api_name=False,
queue=False,
)
.then(
self._display_input,
[self.saved_input, self.chatbot_state],
[self.chatbot, self.chatbot_state],
api_name=False,
queue=False,
)
.success(
submit_fn,
[self.chatbot_state] + self.additional_inputs,
[self.chatbot, self.chatbot_state, self.num_tokens],
api_name=False,
)
)
self._setup_stop_events(submit_triggers, submit_event)
else:
raise ValueError(f'Better install new gradio version than 3.44.0')
if self.retry_btn:
retry_event = (
self.retry_btn.click(
self._delete_prev_fn,
[self.chatbot_state],
[self.chatbot, self.saved_input, self.chatbot_state],
api_name=False,
queue=False,
)
.then(
self._display_input,
[self.saved_input, self.chatbot_state],
[self.chatbot, self.chatbot_state],
api_name=False,
queue=False,
)
.success(
submit_fn,
[self.chatbot_state] + self.additional_inputs,
[self.chatbot, self.chatbot_state, self.num_tokens],
api_name=False,
)
)
self._setup_stop_events([self.retry_btn.click], retry_event)
if self.undo_btn:
self.undo_btn.click(
# self._delete_prev_fn,
# [self.chatbot_state],
# [self.chatbot, self.saved_input, self.chatbot_state],
undo_history_until_last_assistant_turn,
[self.chatbot_state],
[self.chatbot, self.chatbot_state],
api_name=False,
queue=False,
)
# .then(
# lambda x: x,
# [self.saved_input],
# [self.textbox],
# api_name=False,
# queue=False,
# )
async def _stream_fn(
self,
# message: str,
history_with_input,
request: Request,
*args,
) -> AsyncGenerator:
history = history_with_input[:-1]
message = history_with_input[-1][0]
inputs, _, _ = special_args(
self.fn, inputs=[history_with_input, *args], request=request
)
if self.is_async:
generator = self.fn(*inputs)
else:
generator = await anyio.to_thread.run_sync(
self.fn, *inputs, limiter=self.limiter
)
generator = SyncToAsyncIterator(generator, self.limiter)
# ! In case of error, yield the previous history & undo any generation before raising error
try:
first_response_pack = await async_iteration(generator)
if isinstance(first_response_pack, (tuple, list)):
first_response, num_tokens = first_response_pack
else:
first_response, num_tokens = first_response_pack, -1
update = history + [[message, first_response]]
yield update, update, f"{num_tokens} toks"
except StopIteration:
update = history + [[message, None]]
yield update, update, "NaN toks"
except Exception as e:
yield history, history, "NaN toks"
raise e
try:
async for response_pack in generator:
if isinstance(response_pack, (tuple, list)):
response, num_tokens = response_pack
else:
response, num_tokens = response_pack, "NaN toks"
update = history + [[message, response]]
yield update, update, f"{num_tokens} toks"
except Exception as e:
yield history, history, "NaN toks"
raise e
async def _examples_stream_fn(
self,
# message: str,
*args,
) -> AsyncGenerator:
history = []
input_len = 1 + len(self.multimodal_inputs)
saved_input = args[:input_len]
message = saved_input[0]
additional_inputs = [] if len(args) <= input_len else args[input_len:]
history = self._add_inputs_to_history(history, *saved_input)
inputs, _, _ = special_args(self.fn, inputs=[history, *additional_inputs], request=None)
if self.is_async:
generator = self.fn(*inputs)
else:
generator = await anyio.to_thread.run_sync(
self.fn, *inputs, limiter=self.limiter
)
generator = SyncToAsyncIterator(generator, self.limiter)
# async for response in generator:
# yield [[message, response]]
try:
async for response_pack in generator:
if isinstance(response_pack, (tuple, list)):
response, num_tokens = response_pack
else:
response, num_tokens = response_pack, "NaN toks"
update = history + [[message, response]]
yield update, update, f"{num_tokens} toks"
except Exception as e:
yield history, history, "NaN toks"
raise e
async def _examples_fn(self, message: str, *args) -> list[list[str | None]]:
raise NotImplementedError
inputs, _, _ = special_args(self.fn, inputs=[message, [], *args], request=None)
if self.is_async:
response = await self.fn(*inputs)
else:
response = await anyio.to_thread.run_sync(
self.fn, *inputs, limiter=self.limiter
)
return [[message, response]]
def gradio_history_to_openai_conversations(message=None, history=None, system_prompt=None):
conversations = []
system_prompt = system_prompt or SYSTEM_PROMPT
if history is not None and len(history) > 0:
for i, (prompt, res) in enumerate(history):
if prompt is not None:
conversations.append({"role": "user", "content": prompt.strip()})
if res is not None:
conversations.append({"role": "assistant", "content": res.strip()})
if message is not None:
if len(message.strip()) == 0:
raise gr.Error("The message cannot be empty!")
conversations.append({"role": "user", "content": message.strip()})
if conversations[0]['role'] != 'system':
conversations = [{"role": "system", "content": system_prompt}] + conversations
return conversations
def gradio_history_to_conversation_prompt(message=None, history=None, system_prompt=None):
global MODEL_ENGINE
full_prompt = MODEL_ENGINE.apply_chat_template(
gradio_history_to_openai_conversations(
message, history=history, system_prompt=system_prompt),
add_generation_prompt=True
)
return full_prompt
def gradio_history_to_vision_conversations_paths(
history, system_prompt=None, image_token=None
):
image_token = image_token or IMAGE_TOKEN
conversations = []
image_paths = []
for i, his in enumerate(history):
prompt, response = his
last_turn = conversations[-1] if len(conversations) > 0 else None
if prompt is not None:
if isinstance(prompt, tuple):
image_path = prompt[0]
if last_turn is not None and last_turn['role'] == 'user':
last_turn['content'] += f" {image_token}"
else:
# last_turn None or last_turn['role'] == 'assistant'
conversations.append({
"role": "user",
"content": f"{image_token}"
})
image_paths.append(image_path)
else:
assert prompt is not None and isinstance(prompt, str)
if last_turn is not None and last_turn['role'] == 'user':
last_turn['content'] += f"\n{prompt}"
else:
conversations.append({
"role": "user",
"content": prompt,
})
if response is not None:
assert isinstance(response, str)
conversations.append({
"role": "assistant",
"content": response,
})
if conversations[0]['role'] != 'system':
system_prompt = system_prompt or SYSTEM_PROMPT
conversations = [{"role": "system", "content": system_prompt}] + conversations
return conversations, image_paths
def gradio_history_to_vision_conversation_prompt_paths(
history, system_prompt=None, image_token=None
):
"""
Aggregate gradio history into openai conversations
history = [
["Hello", "Response"],
[(file,), None],
]
--->
[
{"role": "user", "content": ...}
]
"""
global MODEL_ENGINE
conversations, image_paths = gradio_history_to_vision_conversations_paths(
history, system_prompt, image_token
)
# print(f'convo: {json.dumps(conversations, indent=4, ensure_ascii=False)}\n{image_paths=}')
full_prompt = MODEL_ENGINE.apply_chat_template(
conversations,
add_generation_prompt=True
)
return full_prompt, image_paths, conversations
def is_doc(file_path):
is_doc_allowed = file_path.endswith((".pdf", ".docx", ".txt"))
return is_doc_allowed
def read_doc(file_path):
from langchain_community.document_loaders import PyPDFLoader, Docx2txtLoader, TextLoader
if file_path.endswith('.pdf'):
loader = PyPDFLoader(file_path)
elif file_path.endswith('.docx'):
loader = Docx2txtLoader(file_path)
elif file_path.endswith('.txt'):
loader = TextLoader(file_path)
texts = loader.load()
text = "\n\n".join([t.page_content for t in texts])
return text
def doc_file_to_instruct_content(file_path, doc_instruction=None):
doc_instruction = doc_instruction or DOC_INSTRUCTION
content = doc_instruction.strip() + "\n" + DOC_TEMPLATE.format(content=read_doc(file_path))
return content
def gradio_history_to_doc_conversation_prompt(
history, system_prompt=None, doc_instruction=None,
):
"""
Aggregate gradio history into openai conversations
history = [
["Hello", "Response"],
[(file,), None],
]
--->
[
{"role": "user", "content": ...}
]
"""
global MODEL_ENGINE
# image_token = image_token or IMAGE_TOKEN
doc_instruction = doc_instruction or DOC_INSTRUCTION
conversations = []
image_paths = []
for i, his in enumerate(history):
prompt, response = his
last_turn = conversations[-1] if len(conversations) > 0 else None
if prompt is not None:
if isinstance(prompt, tuple):
file_path = prompt[0]
if not is_doc(file_path):
raise gr.Error(f'file not doc {file_path}')
content = doc_file_to_instruct_content(file_path, doc_instruction)
if last_turn is not None and last_turn['role'] == 'user':
last_turn['content'] += f"{content}"
else:
# last_turn None or last_turn['role'] == 'assistant'
conversations.append({
"role": "user",
"content": f"{content}"
})
else:
assert prompt is not None and isinstance(prompt, str)
if last_turn is not None and last_turn['role'] == 'user':
last_turn['content'] += f"\n{prompt}"
else:
conversations.append({
"role": "user",
"content": prompt,
})
if response is not None:
assert isinstance(response, str)
conversations.append({
"role": "assistant",
"content": response,
})
if conversations[0]['role'] != 'system':
system_prompt = system_prompt or SYSTEM_PROMPT
conversations = [{"role": "system", "content": system_prompt}] + conversations
full_prompt = MODEL_ENGINE.apply_chat_template(
conversations,
add_generation_prompt=True
)
return full_prompt, conversations
def gradio_history_to_vision_doc_conversation_prompt_paths(
history, system_prompt=None, image_token=None, doc_instruction=None,
):
"""
Aggregate gradio history into openai conversations
history = [
["Hello", "Response"],
[(file,), None],
]
--->
[
{"role": "user", "content": ...}
]
"""
global MODEL_ENGINE
image_token = image_token or IMAGE_TOKEN
doc_instruction = doc_instruction or DOC_INSTRUCTION
conversations = []
image_paths = []
for i, his in enumerate(history):
prompt, response = his
last_turn = conversations[-1] if len(conversations) > 0 else None
if prompt is not None:
if isinstance(prompt, tuple):
file_path = prompt[0]
if is_doc(file_path):
content = doc_file_to_instruct_content(file_path, doc_instruction)
if last_turn is not None and last_turn['role'] == 'user':
last_turn['content'] += f"{content}"
else:
# last_turn None or last_turn['role'] == 'assistant'
conversations.append({
"role": "user",
"content": f"{content}"
})
else:
if last_turn is not None and last_turn['role'] == 'user':
last_turn['content'] += f" {image_token}"
else:
# last_turn None or last_turn['role'] == 'assistant'
conversations.append({
"role": "user",
"content": f"{image_token}"
})
image_paths.append(file_path)
else:
assert prompt is not None and isinstance(prompt, str)
if last_turn is not None and last_turn['role'] == 'user':
last_turn['content'] += f"\n{prompt}"
else:
conversations.append({
"role": "user",
"content": prompt,
})
if response is not None:
assert isinstance(response, str)
conversations.append({
"role": "assistant",
"content": response,
})
if conversations[0]['role'] != 'system':
system_prompt = system_prompt or SYSTEM_PROMPT
conversations = [{"role": "system", "content": system_prompt}] + conversations
full_prompt = MODEL_ENGINE.apply_chat_template(
conversations,
add_generation_prompt=True
)
return full_prompt, image_paths, conversations
@maybe_spaces_gpu
def vision_chat_response_stream_multiturn_engine(
history: List[Tuple[str, str]],
temperature: float,
max_tokens: int,
system_prompt: Optional[str] = SYSTEM_PROMPT,
image_token: Optional[str] = IMAGE_TOKEN,
):
global MODEL_ENGINE
temperature = float(temperature)
# ! remove frequency_penalty
# frequency_penalty = float(frequency_penalty)
max_tokens = int(max_tokens)
# ! skip safety
if DATETIME_FORMAT in system_prompt:
# ! This sometime works sometimes dont
system_prompt = system_prompt.format(cur_datetime=get_datetime_string())
# ! history now can have multimodal
full_prompt, image_paths, conversations = gradio_history_to_vision_conversation_prompt_paths(
history=history, system_prompt=system_prompt, image_token=image_token
)
if hasattr(MODEL_ENGINE, "get_multimodal_tokens"):
num_tokens = MODEL_ENGINE.get_multimodal_tokens(full_prompt, image_paths=image_paths)
else:
num_tokens = len(MODEL_ENGINE.tokenizer.encode(full_prompt))
if num_tokens >= MODEL_ENGINE.max_position_embeddings - 128:
raise gr.Error(f"Conversation or prompt is too long ({num_tokens} toks), please clear the chatbox or try shorter input.")
# print(f'{image_paths=}')
# print(full_prompt)
outputs = None
response = None
num_tokens = -1
for j, outputs in enumerate(MODEL_ENGINE.generate_yield_string(
prompt=full_prompt,
temperature=temperature,
max_tokens=max_tokens,
image_paths=image_paths,
)):
if isinstance(outputs, tuple):
response, num_tokens = outputs
else:
response, num_tokens = outputs, -1
yield response, num_tokens
print(format_conversation(history + [[None, response]]))
if response is not None:
yield response, num_tokens
@maybe_spaces_gpu
def doc_chat_response_stream_multiturn_engine(
history: List[Tuple[str, str]],
temperature: float,
max_tokens: int,
system_prompt: Optional[str] = SYSTEM_PROMPT,
doc_instruction: Optional[str] = DOC_INSTRUCTION,
):
global MODEL_ENGINE
temperature = float(temperature)
# ! remove frequency_penalty
# frequency_penalty = float(frequency_penalty)
max_tokens = int(max_tokens)
# ! skip safety
if DATETIME_FORMAT in system_prompt:
# ! This sometime works sometimes dont
system_prompt = system_prompt.format(cur_datetime=get_datetime_string())
# ! history now can have multimodal
full_prompt, conversations = gradio_history_to_doc_conversation_prompt(
history=history, system_prompt=system_prompt, doc_instruction=doc_instruction
)
# ! length checked
num_tokens = len(MODEL_ENGINE.tokenizer.encode(full_prompt))
if num_tokens >= MODEL_ENGINE.max_position_embeddings - 128:
raise gr.Error(f"Conversation or prompt is too long ({num_tokens} toks), please clear the chatbox or try shorter input.")
# print(full_prompt)
outputs = None
response = None
num_tokens = -1
for j, outputs in enumerate(MODEL_ENGINE.generate_yield_string(
prompt=full_prompt,
temperature=temperature,
max_tokens=max_tokens,
# image_paths=image_paths,
)):
if isinstance(outputs, tuple):
response, num_tokens = outputs
else:
response, num_tokens = outputs, -1
yield response, num_tokens
print(format_conversation(history + [[None, response]]))
if response is not None:
yield response, num_tokens
@maybe_spaces_gpu
def vision_doc_chat_response_stream_multiturn_engine(
history: List[Tuple[str, str]],
temperature: float,
max_tokens: int,
system_prompt: Optional[str] = SYSTEM_PROMPT,
image_token: Optional[str] = IMAGE_TOKEN,
doc_instruction: Optional[str] = DOC_INSTRUCTION,
):
global MODEL_ENGINE
temperature = float(temperature)
# ! remove frequency_penalty
# frequency_penalty = float(frequency_penalty)
max_tokens = int(max_tokens)
# ! skip safety
if DATETIME_FORMAT in system_prompt:
# ! This sometime works sometimes dont
system_prompt = system_prompt.format(cur_datetime=get_datetime_string())
# ! history now can have multimodal
full_prompt, image_paths, conversations = gradio_history_to_vision_doc_conversation_prompt_paths(
history=history, system_prompt=system_prompt, image_token=image_token, doc_instruction=doc_instruction
)
# ! length check
if hasattr(MODEL_ENGINE, "get_multimodal_tokens"):
num_tokens = MODEL_ENGINE.get_multimodal_tokens(full_prompt, image_paths=image_paths)
else:
num_tokens = len(MODEL_ENGINE.tokenizer.encode(full_prompt))
if num_tokens >= MODEL_ENGINE.max_position_embeddings - 128:
raise gr.Error(f"Conversation or prompt is too long ({num_tokens} toks), please clear the chatbox or try shorter input.")
# print(full_prompt)
# print(f'{image_paths=}')
outputs = None
response = None
num_tokens = -1
for j, outputs in enumerate(MODEL_ENGINE.generate_yield_string(
prompt=full_prompt,
temperature=temperature,
max_tokens=max_tokens,
image_paths=image_paths,
)):
if isinstance(outputs, tuple):
response, num_tokens = outputs
else:
response, num_tokens = outputs, -1
yield response, num_tokens
print(format_conversation(history + [[None, response]]))
if response is not None:
yield response, num_tokens
@register_demo
class VisionChatInterfaceDemo(ChatInterfaceDemo):
"""
Accept vision image
"""
@property
def tab_name(self):
return "Vision Chat"
@property
def examples(self):
return [
["What's strange about this image?", "assets/dog_monalisa.jpeg",],
["Explain why the sky is blue.", None,],
]
def create_demo(
self,
title: str | None = None,
description: str | None = None,
**kwargs
) -> gr.Blocks:
system_prompt = kwargs.get("system_prompt", SYSTEM_PROMPT)
max_tokens = kwargs.get("max_tokens", MAX_TOKENS)
temperature = kwargs.get("temperature", TEMPERATURE)
model_name = kwargs.get("model_name", MODEL_NAME)
description = description or """Upload an image to ask question about it."""
def add_multimodal_fn() -> List[Component]:
image_input = gr.Image(label="Input Image", type="filepath", )
return [image_input]
additional_inputs = [
gr.Number(value=temperature, label='Temperature', min_width=20),
gr.Number(value=max_tokens, label='Max-tokens', min_width=20),
gr.Textbox(value=system_prompt, label='System prompt', lines=1),
gr.Textbox(value=IMAGE_TOKEN, label='Visual token', lines=1, interactive=IMAGE_TOKEN_INTERACTIVE, min_width=20),
]
def render_additional_inputs_fn():
with Row():
additional_inputs[0].render()
additional_inputs[1].render()
additional_inputs[3].render()
additional_inputs[2].render()
demo_chat = MultiModalChatInterface(
vision_chat_response_stream_multiturn_engine,
chatbot=gr.Chatbot(
label=model_name,
bubble_full_width=False,
latex_delimiters=[
{ "left": "$", "right": "$", "display": False},
{ "left": "$$", "right": "$$", "display": True},
],
show_copy_button=True,
layout="panel" if USE_PANEL else "bubble",
height=CHATBOT_HEIGHT,
),
# textbox=gr.Textbox(placeholder='Type message', lines=4, max_lines=128, min_width=200),
textbox=gr.Textbox(placeholder='Type message', lines=1, max_lines=128, min_width=200, scale=8),
submit_btn=gr.Button(value='Submit', variant="primary", scale=0),
# ! consider preventing the stop button
# stop_btn=None,
add_multimodal_fn=add_multimodal_fn,
title=title,
description=description,
additional_inputs=additional_inputs,
render_additional_inputs_fn=render_additional_inputs_fn,
additional_inputs_accordion=gr.Accordion("Additional Inputs", open=True),
examples=self.examples,
cache_examples=False,
css=CSS,
)
return demo_chat
def add_document_upload():
file_input = gr.File(label='Upload pdf, docx, txt', file_count='single', file_types=['pdf', 'docx', 'txt'])
# with Group():
# file_input = gr.Textbox(value=None, label='Document path', lines=1, interactive=False)
# upload_button = gr.UploadButton("Click to Upload document", file_types=['pdf', 'docx', 'txt'], file_count="single")
# upload_button.upload(lambda x: x.name, upload_button, file_input)
return file_input
@register_demo
class DocChatInterfaceDemo(ChatInterfaceDemo):
"""
Accept document (full length no RAG)
"""
@property
def tab_name(self):
return "Doc Chat"
@property
def examples(self):
return [
["Summarize the document", "assets/attention_short.pdf",],
["Explain why the sky is blue.", None,],
]
def create_demo(
self,
title: str | None = None,
description: str | None = None,
**kwargs
) -> gr.Blocks:
system_prompt = kwargs.get("system_prompt", SYSTEM_PROMPT)
max_tokens = kwargs.get("max_tokens", MAX_TOKENS)
temperature = kwargs.get("temperature", TEMPERATURE)
model_name = kwargs.get("model_name", MODEL_NAME)
# frequence_penalty = FREQUENCE_PENALTY
# presence_penalty = PRESENCE_PENALTY
description = description or """Upload a short document to ask question about it."""
def add_multimodal_fn() -> List[Component]:
file_input = add_document_upload()
# image_input = gr.Image(label="Input Image", type="filepath", )
return [file_input]
additional_inputs = [
gr.Number(value=temperature, label='Temperature', min_width=20),
gr.Number(value=max_tokens, label='Max-tokens', min_width=20),
gr.Textbox(value=system_prompt, label='System prompt', lines=1),
gr.Textbox(value=DOC_INSTRUCTION, label='Doc instruction', lines=1),
]
def render_additional_inputs_fn():
with Row():
additional_inputs[0].render()
additional_inputs[1].render()
additional_inputs[2].render()
additional_inputs[3].render()
demo_chat = MultiModalChatInterface(
doc_chat_response_stream_multiturn_engine,
chatbot=gr.Chatbot(
label=model_name,
bubble_full_width=False,
latex_delimiters=[
{ "left": "$", "right": "$", "display": False},
{ "left": "$$", "right": "$$", "display": True},
],
show_copy_button=True,
layout="panel" if USE_PANEL else "bubble",
height=CHATBOT_HEIGHT,
),
textbox=gr.Textbox(placeholder='Type message', lines=1, max_lines=128, min_width=200, scale=8),
submit_btn=gr.Button(value='Submit', variant="primary", scale=0),
# ! consider preventing the stop button
add_multimodal_fn=add_multimodal_fn,
title=title,
description=description,
additional_inputs=additional_inputs,
render_additional_inputs_fn=render_additional_inputs_fn,
additional_inputs_accordion=gr.Accordion("Additional Inputs", open=True),
examples=self.examples,
cache_examples=False,
css=CSS,
)
return demo_chat
@register_demo
class VisionDocChatInterfaceDemo(ChatInterfaceDemo):
"""
Accept either vision image or document (full length no RAG)
"""
@property
def tab_name(self):
return "Vision Doc Chat"
@property
def examples(self):
return [
["What's strange about this image?", None, "assets/dog_monalisa.jpeg",],
["Summarize the document", "assets/attention_short.pdf", None,],
["Explain why the sky is blue.", None, None],
]
def create_demo(
self,
title: str | None = None,
description: str | None = None,
**kwargs
) -> gr.Blocks:
system_prompt = kwargs.get("system_prompt", SYSTEM_PROMPT)
max_tokens = kwargs.get("max_tokens", MAX_TOKENS)
temperature = kwargs.get("temperature", TEMPERATURE)
model_name = kwargs.get("model_name", MODEL_NAME)
# frequence_penalty = FREQUENCE_PENALTY
# presence_penalty = PRESENCE_PENALTY
description = description or """Upload either an image or short document to ask question about it."""
def add_multimodal_fn() -> List[Component]:
file_input = add_document_upload()
image_input = gr.Image(label="Input Image", type="filepath", )
return [file_input, image_input]
additional_inputs = [
gr.Number(value=temperature, label='Temperature', min_width=20),
gr.Number(value=max_tokens, label='Max-tokens', min_width=20),
gr.Textbox(value=system_prompt, label='System prompt', lines=1),
gr.Textbox(value=IMAGE_TOKEN, label='Visual token', lines=1, interactive=IMAGE_TOKEN_INTERACTIVE, min_width=2),
gr.Textbox(value=DOC_INSTRUCTION, label='Doc instruction', lines=1),
]
def render_additional_inputs_fn():
with Row():
additional_inputs[0].render()
additional_inputs[1].render()
additional_inputs[3].render()
additional_inputs[2].render()
additional_inputs[4].render()
demo_chat = MultiModalChatInterface(
vision_doc_chat_response_stream_multiturn_engine,
chatbot=gr.Chatbot(
label=MODEL_NAME,
bubble_full_width=False,
latex_delimiters=[
{ "left": "$", "right": "$", "display": False},
{ "left": "$$", "right": "$$", "display": True},
],
show_copy_button=True,
layout="panel" if USE_PANEL else "bubble",
height=CHATBOT_HEIGHT,
),
textbox=gr.Textbox(placeholder='Type message', lines=1, max_lines=128, min_width=200, scale=8),
submit_btn=gr.Button(value='Submit', variant="primary", scale=0),
add_multimodal_fn=add_multimodal_fn,
title=title,
description=description,
additional_inputs=additional_inputs,
render_additional_inputs_fn=render_additional_inputs_fn,
additional_inputs_accordion=gr.Accordion("Additional Inputs", open=True),
examples=self.examples,
cache_examples=False,
css=CSS,
)
return demo_chat |