65 Commits

Author SHA1 Message Date
comfyanonymous
cd0be2a31e Fix some performance issues with weight loading and unloading.
Lower peak memory usage when changing model.

Fix case where model weights would be unloaded and reloaded.
2024-03-28 18:04:42 -04:00
comfyanonymous
cd92ad35a9 Move cleanup_models to improve performance. 2024-03-23 17:27:10 -04:00
comfyanonymous
b513a7681e Pass extra_pnginfo as None when not in input data. 2024-03-07 15:07:47 -05:00
Rick Love
7967dc360e FIX recursive_will_execute performance (simple ~300x performance increase} (#2852)
* FIX recursive_will_execute performance

* Minimize code changes

* memo must be created outside lambda
2024-02-21 20:21:24 -05:00
comfyanonymous
edacbb2656 Cleanups. 2024-01-26 21:31:13 -05:00
realazthat
82ab3b7f0f Don't use PEP 604 type hints, to stay compatible with Python<3.10. 2024-01-17 17:16:34 -05:00
comfyanonymous
28712a5446 Rename status notes to status messages.
I think message describes them better.
2024-01-12 18:17:06 -05:00
comfyanonymous
5a89871e0c Clear status notes on execution start. 2024-01-12 17:21:22 -05:00
realazthat
e57852b5e9 Add error, status to /history endpoint 2024-01-11 10:16:42 -05:00
comfyanonymous
7f99d58877 Add a /free route to unload models or free all memory.
A POST request to /free with: {"unload_models":true}
will unload models from vram.

A POST request to /free with: {"free_memory":true}
will unload models and free all cached data from the last run workflow.
2024-01-04 17:15:22 -05:00
comfyanonymous
94346bfe12 Fix VALIDATE_INPUTS getting called multiple times.
Allow VALIDATE_INPUTS to only validate specific inputs.
2023-12-29 17:36:40 -05:00
comfyanonymous
0baa3cc687 --disable-smart-memory now unloads everything like it did originally. 2023-12-23 04:25:06 -05:00
comfyanonymous
02b4cb5a7a Do a garbage collect after the interval even if nothing is running. 2023-11-30 15:22:32 -05:00
comfyanonymous
48388a8d4d Only show last 200 elements in the UI history tab. 2023-11-20 16:56:29 -05:00
comfyanonymous
dfd72217d7 Cap maximum history size at 10000. Delete oldest entry when reached. 2023-11-20 16:38:39 -05:00
comfyanonymous
18e4504de7 Pull some small changes from the other repo. 2023-10-11 20:38:48 -04:00
pythongosssss
4f7bad4bad fix crash on node with VALIDATE_INPUTS and actual inputs 2023-09-07 18:42:21 +01:00
comfyanonymous
a216b56591 Smarter memory management.
Try to keep models on the vram when possible.

Better lowvram mode for controlnets.
2023-08-17 01:06:34 -04:00
Michael Poutre
9329b1a6b2 fix(execution): Fix support for input-less nodes 2023-08-01 12:29:01 -07:00
Michael Poutre
fc8a738779 chore: Fix typo 2023-08-01 12:27:50 -07:00
comfyanonymous
c8a5d3363c Fix issue with lora in some cases when combined with model merging. 2023-07-21 21:27:27 -04:00
comfyanonymous
be226475e0 Persist node instances between executions instead of deleting them.
If the same node id with the same class exists between two executions the
same instance will be used.

This means you can now cache things in nodes for more efficiency.
2023-06-29 23:38:56 -04:00
comfyanonymous
051f79bc8e Send websocket message only when prompt is actually done executing. 2023-06-13 13:38:43 -04:00
comfyanonymous
314abc241f Add a /history/{prompt_id} endpoint. 2023-06-12 14:34:30 -04:00
comfyanonymous
224d79d812 Fix issue with cancelling prompt. 2023-05-28 00:32:26 -04:00
space-nuko
3ba39931a0 Rename exception message field 2023-05-27 21:06:07 -05:00
space-nuko
96b9364255 Exception message 2023-05-27 21:06:07 -05:00
space-nuko
49ec8a2b17 Catch typecast errors 2023-05-27 21:06:07 -05:00
space-nuko
2d885e7019 Fix 2023-05-27 21:06:07 -05:00
space-nuko
8260e16cc7 Return null for value format 2023-05-27 21:06:07 -05:00
space-nuko
cbd9aac327 Show message in the frontend if prompt execution raises an exception 2023-05-27 21:06:07 -05:00
space-nuko
f510feec01 Send back more information about exceptions that happen during execution 2023-05-27 21:06:07 -05:00
space-nuko
c742150520 Add missing input name/config 2023-05-27 21:06:07 -05:00
space-nuko
8d5245aace Improve invalid prompt error message 2023-05-27 21:06:07 -05:00
space-nuko
82e3940028 Improve error output for failed nodes 2023-05-27 21:06:07 -05:00
comfyanonymous
6169965734 Parsing error crash. 2023-05-22 20:51:30 -04:00
comfyanonymous
5ee1ea3988 Add a node_errors to the /prompt error json response.
"node_errors" contains a dict keyed by node ids. The contents are a message
and a list of dependent outputs.
2023-05-22 13:22:38 -04:00
comfyanonymous
ea32c601ea Print min and max values in validation error message. 2023-05-21 00:24:28 -04:00
comfyanonymous
5ab1c07017 Fix outputs gone from history. 2023-05-15 00:27:28 -04:00
comfyanonymous
08bea5f9a7 Print prompt execution time. 2023-05-14 01:34:25 -04:00
comfyanonymous
773d00ab0a Add the prompt id to some websocket messages. 2023-05-13 11:17:16 -04:00
BlenderNeko
dd1be4e992 Make nodes map over input lists (#579)
* allow nodes to map over lists

* make work with IS_CHANGED and VALIDATE_INPUTS

* give list outputs distinct socket shape

* add rebatch node

* add batch index logic

* add repeat latent batch

* deal with noise mask edge cases in latentfrombatch
2023-05-13 11:15:45 -04:00
comfyanonymous
b698ff5ce4 Add the prompt_id to some websocket messages. 2023-05-11 01:22:40 -04:00
comfyanonymous
f031ed3b4e Send websocket message with list of cached nodes right before execution. 2023-05-10 15:59:24 -04:00
comfyanonymous
5bd99f38c3 Send execution_error message on websocket on execution exception. 2023-05-10 15:49:49 -04:00
comfyanonymous
d8edfe1f07 Only validate each input once. 2023-05-10 00:29:31 -04:00
comfyanonymous
853b649b1d Don't print traceback when processing interrupted. 2023-05-09 23:51:52 -04:00
comfyanonymous
26380e84df If IS_CHANGED returns exception delete the output instead of crashing. 2023-04-26 02:13:56 -04:00
comfyanonymous
aa8889aa60 Don't keep cached outputs for removed nodes. 2023-04-26 02:05:57 -04:00
comfyanonymous
0c0b4c41f7 Don't delete all outputs when execution gets interrupted. 2023-04-23 22:44:38 -04:00