Spaces:
Running
on
Zero
Running
on
Zero
layerdiffusion
commited on
Commit
•
a1c7119
1
Parent(s):
a9cb4cf
- app.py +16 -1
- chat_interface.py +16 -10
app.py
CHANGED
@@ -28,6 +28,7 @@ from diffusers.models.attention_processor import AttnProcessor2_0
|
|
28 |
from transformers import CLIPTextModel, CLIPTokenizer
|
29 |
from lib_omost.pipeline import StableDiffusionXLOmostPipeline
|
30 |
from chat_interface import ChatInterface
|
|
|
31 |
|
32 |
import lib_omost.canvas as omost_canvas
|
33 |
|
@@ -130,9 +131,23 @@ def chat_fn(message: str, history: list, seed:int, temperature: float, top_p: fl
|
|
130 |
|
131 |
streamer = TextIteratorStreamer(llm_tokenizer, timeout=10.0, skip_prompt=True, skip_special_tokens=True)
|
132 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
133 |
generate_kwargs = dict(
|
134 |
input_ids=input_ids,
|
135 |
streamer=streamer,
|
|
|
136 |
max_new_tokens=max_new_tokens,
|
137 |
do_sample=True,
|
138 |
temperature=temperature,
|
@@ -148,7 +163,7 @@ def chat_fn(message: str, history: list, seed:int, temperature: float, top_p: fl
|
|
148 |
for text in streamer:
|
149 |
outputs.append(text)
|
150 |
# print(outputs)
|
151 |
-
yield "".join(outputs)
|
152 |
|
153 |
print(f'Chat end at {time.time() - time_stamp:.2f} seconds:', message)
|
154 |
return
|
|
|
28 |
from transformers import CLIPTextModel, CLIPTokenizer
|
29 |
from lib_omost.pipeline import StableDiffusionXLOmostPipeline
|
30 |
from chat_interface import ChatInterface
|
31 |
+
from transformers.generation.stopping_criteria import StoppingCriteriaList
|
32 |
|
33 |
import lib_omost.canvas as omost_canvas
|
34 |
|
|
|
131 |
|
132 |
streamer = TextIteratorStreamer(llm_tokenizer, timeout=10.0, skip_prompt=True, skip_special_tokens=True)
|
133 |
|
134 |
+
def interactive_stopping_criteria(input_ids: torch.LongTensor, score: torch.FloatTensor, **kwargs) -> bool:
|
135 |
+
if getattr(streamer, 'user_interrupted', False):
|
136 |
+
print('User stopped generation')
|
137 |
+
return True
|
138 |
+
else:
|
139 |
+
return False
|
140 |
+
|
141 |
+
stopping_criteria = StoppingCriteriaList([interactive_stopping_criteria])
|
142 |
+
|
143 |
+
def interrupter():
|
144 |
+
streamer.user_interrupted = True
|
145 |
+
return
|
146 |
+
|
147 |
generate_kwargs = dict(
|
148 |
input_ids=input_ids,
|
149 |
streamer=streamer,
|
150 |
+
stopping_criteria=stopping_criteria,
|
151 |
max_new_tokens=max_new_tokens,
|
152 |
do_sample=True,
|
153 |
temperature=temperature,
|
|
|
163 |
for text in streamer:
|
164 |
outputs.append(text)
|
165 |
# print(outputs)
|
166 |
+
yield "".join(outputs), interrupter
|
167 |
|
168 |
print(f'Chat end at {time.time() - time_stamp:.2f} seconds:', message)
|
169 |
return
|
chat_interface.py
CHANGED
@@ -20,7 +20,7 @@ from gradio.components import (
|
|
20 |
State,
|
21 |
Textbox,
|
22 |
get_component_instance,
|
23 |
-
Dataset
|
24 |
)
|
25 |
from gradio.events import Dependency, on
|
26 |
from gradio.helpers import special_args
|
@@ -103,6 +103,8 @@ class ChatInterface(Blocks):
|
|
103 |
self.pre_fn = pre_fn
|
104 |
self.pre_fn_kwargs = pre_fn_kwargs
|
105 |
|
|
|
|
|
106 |
self.multimodal = multimodal
|
107 |
self.concurrency_limit = concurrency_limit
|
108 |
self.fn = fn
|
@@ -287,7 +289,7 @@ class ChatInterface(Blocks):
|
|
287 |
.then(
|
288 |
submit_fn,
|
289 |
[self.saved_input, self.chatbot_state] + self.additional_inputs,
|
290 |
-
[self.chatbot, self.chatbot_state],
|
291 |
show_api=False,
|
292 |
concurrency_limit=cast(
|
293 |
Union[int, Literal["default"], None], self.concurrency_limit
|
@@ -395,6 +397,11 @@ class ChatInterface(Blocks):
|
|
395 |
def _setup_stop_events(
|
396 |
self, event_triggers: list[Callable], event_to_cancel: Dependency
|
397 |
) -> None:
|
|
|
|
|
|
|
|
|
|
|
398 |
if self.stop_btn and self.is_generator:
|
399 |
if self.submit_btn:
|
400 |
for event_trigger in event_triggers:
|
@@ -434,9 +441,8 @@ class ChatInterface(Blocks):
|
|
434 |
queue=False,
|
435 |
)
|
436 |
self.stop_btn.click(
|
437 |
-
|
438 |
-
|
439 |
-
None,
|
440 |
cancels=event_to_cancel,
|
441 |
show_api=False,
|
442 |
)
|
@@ -545,7 +551,7 @@ class ChatInterface(Blocks):
|
|
545 |
)
|
546 |
generator = SyncToAsyncIterator(generator, self.limiter)
|
547 |
try:
|
548 |
-
first_response = await async_iteration(generator)
|
549 |
if self.multimodal and isinstance(message, dict):
|
550 |
for x in message["files"]:
|
551 |
history.append([(x,), None])
|
@@ -553,21 +559,21 @@ class ChatInterface(Blocks):
|
|
553 |
yield update, update
|
554 |
else:
|
555 |
update = history + [[message, first_response]]
|
556 |
-
yield update, update
|
557 |
except StopIteration:
|
558 |
if self.multimodal and isinstance(message, dict):
|
559 |
self._append_multimodal_history(message, None, history)
|
560 |
yield history, history
|
561 |
else:
|
562 |
update = history + [[message, None]]
|
563 |
-
yield update, update
|
564 |
-
async for response in generator:
|
565 |
if self.multimodal and isinstance(message, dict):
|
566 |
update = history + [[message["text"], response]]
|
567 |
yield update, update
|
568 |
else:
|
569 |
update = history + [[message, response]]
|
570 |
-
yield update, update
|
571 |
|
572 |
async def _api_submit_fn(
|
573 |
self, message: str, history: list[list[str | None]], request: Request, *args
|
|
|
20 |
State,
|
21 |
Textbox,
|
22 |
get_component_instance,
|
23 |
+
Dataset,
|
24 |
)
|
25 |
from gradio.events import Dependency, on
|
26 |
from gradio.helpers import special_args
|
|
|
103 |
self.pre_fn = pre_fn
|
104 |
self.pre_fn_kwargs = pre_fn_kwargs
|
105 |
|
106 |
+
self.interrupter = State(None)
|
107 |
+
|
108 |
self.multimodal = multimodal
|
109 |
self.concurrency_limit = concurrency_limit
|
110 |
self.fn = fn
|
|
|
289 |
.then(
|
290 |
submit_fn,
|
291 |
[self.saved_input, self.chatbot_state] + self.additional_inputs,
|
292 |
+
[self.chatbot, self.chatbot_state, self.interrupter],
|
293 |
show_api=False,
|
294 |
concurrency_limit=cast(
|
295 |
Union[int, Literal["default"], None], self.concurrency_limit
|
|
|
397 |
def _setup_stop_events(
|
398 |
self, event_triggers: list[Callable], event_to_cancel: Dependency
|
399 |
) -> None:
|
400 |
+
def perform_interrupt(ipc):
|
401 |
+
if ipc is not None:
|
402 |
+
ipc()
|
403 |
+
return
|
404 |
+
|
405 |
if self.stop_btn and self.is_generator:
|
406 |
if self.submit_btn:
|
407 |
for event_trigger in event_triggers:
|
|
|
441 |
queue=False,
|
442 |
)
|
443 |
self.stop_btn.click(
|
444 |
+
fn=perform_interrupt,
|
445 |
+
inputs=[self.interrupter],
|
|
|
446 |
cancels=event_to_cancel,
|
447 |
show_api=False,
|
448 |
)
|
|
|
551 |
)
|
552 |
generator = SyncToAsyncIterator(generator, self.limiter)
|
553 |
try:
|
554 |
+
first_response, first_interrupter = await async_iteration(generator)
|
555 |
if self.multimodal and isinstance(message, dict):
|
556 |
for x in message["files"]:
|
557 |
history.append([(x,), None])
|
|
|
559 |
yield update, update
|
560 |
else:
|
561 |
update = history + [[message, first_response]]
|
562 |
+
yield update, update, first_interrupter
|
563 |
except StopIteration:
|
564 |
if self.multimodal and isinstance(message, dict):
|
565 |
self._append_multimodal_history(message, None, history)
|
566 |
yield history, history
|
567 |
else:
|
568 |
update = history + [[message, None]]
|
569 |
+
yield update, update, first_interrupter
|
570 |
+
async for response, interrupter in generator:
|
571 |
if self.multimodal and isinstance(message, dict):
|
572 |
update = history + [[message["text"], response]]
|
573 |
yield update, update
|
574 |
else:
|
575 |
update = history + [[message, response]]
|
576 |
+
yield update, update, interrupter
|
577 |
|
578 |
async def _api_submit_fn(
|
579 |
self, message: str, history: list[list[str | None]], request: Request, *args
|