Lincoln Stein
12cae33dcd
fix inpaint model detection ( #3843 )
...
Co-authored-by: Lincoln Stein <lstein@gmail.com>
2023-07-20 12:57:14 +12:00
Lincoln Stein
8439e30798
Merge branch 'main' into release/invokeai-3-0-beta
2023-07-19 12:09:32 -04:00
Lincoln Stein
0b6ef7eb7d
make the convert VAE available to model manager for use in UI
2023-07-19 09:05:24 -04:00
Sergey Borisov
2e7fc055c4
Support both pre and post 4.31.0 transformers
2023-07-19 06:15:17 +03:00
Lincoln Stein
a690cca5b5
make convert work with both 4.30.2 and 4.31.0
2023-07-18 22:18:13 -04:00
Sergey Borisov
0aa7193d3b
Load text_model.embeddings.position_ids outsude state_dict
2023-07-19 04:18:43 +03:00
blessedcoolant
186e98da5e
Merge branch 'main' into fix/mem_cleanup
2023-07-19 10:10:32 +12:00
Eugene Brodsky
dea9a5da7a
Avoid crash if unable to modify the model config file ( #3824 )
...
* fix whitespace; remove invisible characters
* log error and proceed if unable to modify the model config
2023-07-18 16:33:19 -04:00
Sergey Borisov
bda0000acd
Cleanup vram after models offloading, tweak to cleanup local variable references on ram offload
2023-07-18 23:21:18 +03:00
StAlKeR7779
889b77d3d6
Merge branch 'main' into save_vram
2023-07-18 16:55:48 +03:00
Sergey Borisov
bc11296a5e
Disable lazy offloading on disabled vram cache, move resulted tensors to cpu(to not stack vram tensors in cache), fix - text encoder not freed(detach)
2023-07-18 16:20:25 +03:00
Lincoln Stein
1353bf98b3
add specific exception for model probe failures
2023-07-17 23:08:39 -04:00
blessedcoolant
13da881953
Merge branch 'main' into sdxl-support
2023-07-18 13:34:07 +12:00
Lincoln Stein
0ea8d3c30c
prevent crash on rename operation on models in models directory
2023-07-17 07:50:06 -04:00
Lincoln Stein
84a13ff8e1
Merge branch 'mm-ui' of github.com:blessedcoolant/InvokeAI into mm-ui
2023-07-17 07:29:35 -04:00
Lincoln Stein
3fba262c94
expose paths as absolute to web api
2023-07-17 07:29:26 -04:00
Lincoln Stein
107ca6bf47
expose model paths as absolute to web models API
2023-07-17 07:26:05 -04:00
blessedcoolant
cbfd1d1b27
Merge branch 'main' into feat/standalone_diffusers_ti
2023-07-17 22:01:52 +12:00
blessedcoolant
aebd595607
Merge branch 'main' into mm-ui
2023-07-17 13:49:25 +12:00
Kent Keirsey
675a92401c
Merge branch 'main' into lstein/default-model-install
2023-07-16 19:32:59 -04:00
Sergey Borisov
b61c83e836
Allow bin extension to detect diffusers-ti provided as file
2023-07-17 00:32:17 +03:00
Lincoln Stein
2bc3e36bc0
add missing exception name
2023-07-16 16:14:28 -04:00
Lincoln Stein
6fbb5ce780
add renaming capabilities to model update API route
2023-07-16 14:17:05 -04:00
Lincoln Stein
ccbfa5d862
resolve conflicts
2023-07-15 19:47:50 -04:00
Lincoln Stein
6b0a158ffa
Merge branch 'main' into lstein/default-model-install
2023-07-15 18:23:34 -04:00
Lincoln Stein
e01706f5f5
add fp16 support to controlnet models
2023-07-15 10:37:11 -04:00
Lincoln Stein
32e7e52d69
Merge branch 'main' into lstein/default-model-install
2023-07-15 08:30:22 -04:00
Lincoln Stein
2faa7cee37
add rename_model route
2023-07-14 23:03:18 -04:00
Lincoln Stein
a45f7ce355
add --list-models command
2023-07-14 19:52:47 -04:00
Lincoln Stein
e71ce83e9c
Merge branch 'main' into lstein/model-manager-route-enhancements
2023-07-14 13:52:55 -04:00
Lincoln Stein
8600aad12b
multiple enhancements to model manager REACT API
...
1. add a /sync route for synchronizing the in-memory model lists to
models.yaml, the models directory, and the autoimport directories.
2. add optional destination_directories to convert_model and merge_model
operations.
3. add /ckpt_confs route for retrieving known legacy checkpoint configuration
files.
4. add /search route for finding all models in a directory located in the server
filesystem
2023-07-14 13:45:16 -04:00
Lincoln Stein
ad076b1174
add model directory search route
2023-07-14 11:14:33 -04:00
blessedcoolant
16e93c6455
Merge branch 'main' into mm-ui
2023-07-14 15:46:53 +12:00
psychedelicious
eb0d55263b
fix(mm): make model config attribute names consistent
...
Our model fields use `model_name`, but the API response uses `name`. Some places use `model_type` but the API response used `type`.
Changed the API response to provide `model_name` and `model_type`, which simplifies how we manage models on the client substantially.
2023-07-13 15:40:05 +10:00
blessedcoolant
71e34ac256
Merge branch 'main' into mm-ui
2023-07-13 12:48:43 +12:00
Sergey Borisov
67c8cf4bc2
Controlnet model detection
2023-07-12 08:50:19 -04:00
Sergey Borisov
a328986b43
Less naive model detection
2023-07-12 08:50:19 -04:00
blessedcoolant
5a6ad99d4e
feat: Restore Delete Model Functionality
2023-07-12 16:39:07 +12:00
Lincoln Stein
25591788c1
fix conflicts
2023-07-11 15:55:10 -04:00
Lincoln Stein
dab03fb646
rename gpu_mem_reserved to max_vram_cache_size
...
To be consistent with max_cache_size, the amount of memory to hold in
VRAM for model caching is now controlled by the max_vram_cache_size
configuration parameter.
2023-07-11 15:25:39 -04:00
Lincoln Stein
d32f9f7cb0
reverse logic of gpu_mem_reserved
...
- gpu_mem_reserved now indicates the amount of VRAM that will be reserved
for model caching (similar to max_cache_size).
2023-07-11 15:16:40 -04:00
Lincoln Stein
6b93c1451f
do not crash when probing an unknown model type
2023-07-11 10:56:47 -04:00
Lincoln Stein
bf2b5b5cd4
improvements to sdxl support in model manager
...
- Move SDXL-related models to models/sdxl.py
- Create separate base type BaseModelType.StableDiffusionXLRefiner for the refiner
models.
2023-07-09 20:42:03 -04:00
Lincoln Stein
5759a390f9
introduce gpu_mem_reserved configuration parameter
2023-07-09 18:35:04 -04:00
Lincoln Stein
130249a2dd
add model loading support for SDXL
2023-07-09 15:47:06 -04:00
Lincoln Stein
8d7dba937d
fix undefined variable
2023-07-09 14:37:45 -04:00
Lincoln Stein
d6cb0e54b3
don't unload models from GPU until the space is needed
2023-07-09 14:26:30 -04:00
Lincoln Stein
5f7435955e
if models.yaml doesn't exist, rebuild it
2023-07-08 15:13:51 -04:00
Lincoln Stein
69ef1e1e56
speculative change to upgrade script
2023-07-08 11:45:26 -04:00
Eugene Brodsky
97b2ec58e2
Merge branch 'main' into release/invokeai-3-0-alpha
2023-07-07 14:18:12 -04:00
