afd9ae7712
tidy(mm): remove convenience methods from high level model manager service
...
These were added as a hold-me-over for the nodes API changes, no longer needed. A followup commit will fix the nodes API to not rely on these.
2024-03-07 10:56:59 +11:00
4eefed12f0
refactor: 🚨 please the almighty linter
2024-03-07 10:44:40 +11:00
4301a3d6fd
feat: ✨ invert scroll direction for brush size
2024-03-07 10:44:40 +11:00
99c0662e3f
fix(nodes): load config before doing anything else
...
This was preventing custom nodes from loading if a custom nodes dir was specified
Closes #5862
2024-03-07 10:36:27 +11:00
cdc0d0c182
add config_path to ModelRecordChanges
2024-03-07 10:29:29 +11:00
a00369a67a
add config path as field in model update form when model is a checkpoint
2024-03-07 10:29:29 +11:00
4f096ac3ba
feat(scripts): add frontend-types
to Makefile to generate types
2024-03-07 10:16:44 +11:00
f5e3341465
feat(scripts): add support for file path & stdin to frontend typegen script
2024-03-07 10:16:44 +11:00
474852ef7e
feat(scripts): add script to generate openapi schema
2024-03-07 10:16:44 +11:00
b1d72d411e
only show default settings on main models
2024-03-07 09:07:43 +11:00
46614ee28f
lint fix
2024-03-06 15:06:27 -05:00
b019f9bb8b
make sure all metadata in viewer is rendered at correct font size - specifically fixes control adapter metadata being too big
2024-03-06 15:06:27 -05:00
b857692073
update uploads from canvas to controlnet to be intermediates so they do not appear in gallery
2024-03-06 15:06:27 -05:00
90fb7a1a59
move linear tab to be first on workflow edit mode
2024-03-06 15:06:27 -05:00
56fcf6af78
empty state for workflow with no linear fields in view mode
2024-03-06 15:06:27 -05:00
c4fe7e697b
add right-padding to prompt textareas so that text does not go behind icons
2024-03-06 15:06:27 -05:00
2fd483dfc8
use base.800 on invokeBlue.400 for all gallery selected states
2024-03-06 15:06:27 -05:00
b9a9507422
update padding in color picker
2024-03-06 15:06:27 -05:00
f2744fd7d1
fix URL for get image workflow ( #5882 )
...
Co-authored-by: Mary Hipp <maryhipp@Marys-MacBook-Air.local >
2024-03-06 12:46:16 -05:00
fe6e879d38
fix(ui): invalidate InvocationCacheStatus query cache after clearing intermediates
2024-03-06 08:14:12 -05:00
d3ab08fe10
tests: add invocation cache tests
2024-03-06 08:14:12 -05:00
b0615bdfd4
fix(nodes): correctly serialize outputs
...
In order for delete by match to work, we need the whole invocation output to be stringified.
For some reason, the serialization of the output was set to only include the `type` field. It should instead include the whole output.
I don't understand how this ever worked unless pydantic had different serialization behaviour in v1 (though it appears to have been the same).
Closes #5805
2024-03-06 08:14:12 -05:00
bab20467fb
fix(nodes): fix invocation cache clear method args
2024-03-06 08:14:12 -05:00
e24624109e
fix(nodes): fix invocation cache ABC typing
2024-03-06 08:14:12 -05:00
458e7185b8
fix: 🐛 didn't include renamed file
2024-03-06 20:06:14 +11:00
a95128f5f2
refactor: ✏️ canvas mask compositor naming
...
changes `...MaskCompositer` spelling to `...MaskCompositor`
2024-03-06 20:06:14 +11:00
46f32c5e3c
Remove references to the no longer existing invokeai.app.services.model_metadata package
2024-03-05 19:58:25 -05:00
e30cb4b52f
updates for defaultModel
( #5866 )
...
* move defaultModel logic to modelsLoaded and update to work for key instead of name/base/type string
* lint fix
---------
Co-authored-by: Mary Hipp <maryhipp@Marys-MacBook-Air.local >
2024-03-05 09:55:22 -05:00
ba1f6bf926
chore: lint
2024-03-05 23:50:19 +11:00
4a9cca6c2d
fix(ui): format model API response data
2024-03-05 23:50:19 +11:00
b0275700b3
refactor(ui): compute prompt trigger options in the component
...
We can derive the valid trigger options in the component without needing to lift the options list into global state.
2024-03-05 23:50:19 +11:00
8319aca5f9
chore(ui): typegen
2024-03-05 23:50:19 +11:00
51a604f907
pkg(ui): do not fix knip in lint:fix
script
2024-03-05 23:50:19 +11:00
7515d73628
make trigger phrases a list of options and add lora name as description to appear in dropdown
2024-03-05 23:50:19 +11:00
2c453aa531
fix type error
2024-03-05 23:50:19 +11:00
2cca6e4c76
check if lora is enabled before adding trigger phrases
2024-03-05 23:50:19 +11:00
ef171e890a
use a listener to recalculate trigger phrases when model or lora list changes
2024-03-05 23:50:19 +11:00
caafbf2f0d
only show trigger phrase settings on main and lora
2024-03-05 23:50:19 +11:00
2db5eaf907
lint fix
2024-03-05 23:50:19 +11:00
f234bf6256
cleanup
2024-03-05 23:50:19 +11:00
cfa78b4052
adapt embedding popover to work for trigger phrases also
2024-03-05 23:50:19 +11:00
ba1dd4b02b
UI in MM to create trigger phrases
2024-03-05 23:50:19 +11:00
bcf58cac59
feat(mm): add config to skip model hash
...
This is useful for when you are using a memory DB and do not want to wait for all models to be hashed on startup.
2024-03-05 23:50:19 +11:00
e866d90ab2
tidy(mm): remove unused method on probe
2024-03-05 23:50:19 +11:00
e8797787cf
fix(mm): fix incorrect calls to update_model
2024-03-05 23:50:19 +11:00
0082ecb22b
feat(mm): add path
to ModelRecordChanges
2024-03-05 23:50:19 +11:00
656839fcd1
fix(mm): fix typing on heuristic_import
2024-03-05 23:50:19 +11:00
99407c899f
feat(ui): update UI to use new model config backend
...
- Update all queries
- Remove Advanced Add
- Removed un-editable, internal-only model attributes from model edit UI (e.g. format, repo variant, model type)
- Update model tags so the list refreshes when a model installs
- Rename some queries, components, variables, types to match backend
- Fix divide-by-zero in install queue
2024-03-05 23:50:19 +11:00
48119d9010
revert(mm): restore convert route
2024-03-05 23:50:19 +11:00
7c9128b253
tidy(mm): use canonical capitalization for all model-related enums, classes
...
For example, "Lora" -> "LoRA", "Vae" -> "VAE".
2024-03-05 23:50:19 +11:00