de096d0ce7
Weight initialization and More activation func
...
add weight init
add weight init option in create_hypernetwork
fstringify hypernet info
save weight initialization info for further debugging
fill bias with zero for He/Xavier
initialize LayerNorm with Normal
fix loading weight_init
2022-10-26 09:17:01 +03:00
c702d4d0df
Fix off-by-one
2022-10-26 13:43:04 +08:00
2f4c91894d
Remove activation from final layer of HNs
2022-10-26 12:10:30 +08:00
18f86e41f6
Removed two unused imports
2022-10-24 17:21:18 +02:00
e9a410b535
check length for variance
2022-10-24 09:07:39 +03:00
0d2e1dac40
convert deque -> list
...
I don't feel this being efficient
2022-10-24 09:07:39 +03:00
348f89c8d4
statistics for pbar
2022-10-24 09:07:39 +03:00
40b56c9289
cleanup some code
2022-10-24 09:07:39 +03:00
b297cc3324
Hypernetworks - fix KeyError in statistics caching
...
Statistics logging has changed to {filename : list[losses]}, so it has to use loss_info[key].pop()
2022-10-24 09:07:39 +03:00
1fbfc052eb
Update hypernetwork.py
2022-10-23 08:34:33 +03:00
48dbf99e84
Allow tracking real-time loss
...
Someone had 6000 images in their dataset, and it was shown as 0, which was confusing.
This will allow tracking real time dataset-average loss for registered objects.
2022-10-22 22:24:19 +03:00
24694e5983
Update hypernetwork.py
2022-10-22 20:25:32 +03:00
6a4fa73a38
small fix
2022-10-22 13:44:39 +00:00
97749b7c7d
Merge branch 'AUTOMATIC1111:master' into master
2022-10-22 22:00:59 +09:00
7912acef72
small fix
2022-10-22 13:00:44 +00:00
fccba4729d
add an option to avoid dying relu
2022-10-22 12:02:41 +00:00
7fd90128eb
added a guard for hypernet training that will stop early if weights are getting no gradients
2022-10-22 14:48:43 +03:00
dcb45dfecf
Merge branch 'master' of upstream
2022-10-22 11:14:46 +00:00
0e8ca8e7af
add dropout
2022-10-22 11:07:00 +00:00
272fa527bb
Remove unused variable.
2022-10-21 16:52:24 +03:00
19818f023c
Match hypernet name with filename in all cases.
2022-10-21 16:52:24 +03:00
51e3dc9cca
Sanitize hypernet name input.
2022-10-21 16:52:24 +03:00
03a1e288c4
turns out LayerNorm also has weight and bias and needs to be pre-multiplied and trained for hypernets
2022-10-21 10:13:24 +03:00
0c5522ea21
Merge branch 'master' into training-help-text
2022-10-21 09:57:55 +03:00
4ff274e1e3
Revise comments.
2022-10-21 09:55:00 +03:00
5245c7a493
Issue #2921-Give PNG info to Hypernet previews.
2022-10-21 09:55:00 +03:00
c23f666dba
a more strict check for activation type and a more reasonable check for type of layer in hypernets
2022-10-21 09:47:43 +03:00
7543cf5e3b
Fixed some typos in the code
2022-10-20 22:43:08 +02:00
8f59129847
Some changes to the tensorboard code and hypernetwork support
2022-10-20 22:37:16 +02:00
f89829ec3a
Revert "fix bugs and optimizations"
...
This reverts commit 108be15500
2022-10-21 01:37:11 +09:00
108be15500
fix bugs and optimizations
2022-10-21 01:00:41 +09:00
a71e021236
only linear
2022-10-20 23:48:52 +09:00
d8acd34f66
generalized some functions and option for ignoring first layer
2022-10-20 23:43:03 +09:00
6b38c2c19c
Merge branch 'AUTOMATIC1111:master' into master
2022-10-20 18:51:12 +09:00
930b4c64f7
allow float sizes for hypernet's layer_structure
2022-10-20 08:18:02 +03:00
6f98e89486
update
2022-10-20 00:10:45 +00:00
166be3919b
allow overwrite old hn
2022-10-20 00:09:40 +01:00
d6ea584137
change html output
2022-10-20 00:07:57 +01:00
2ce52d32e4
fix for #3086 failing to load any previous hypernet
2022-10-19 16:31:12 +00:00
c6e9fed500
fix for #3086 failing to load any previous hypernet
2022-10-19 19:21:16 +03:00
3770b8d2fa
enable to write layer structure of hn himself
2022-10-19 15:28:42 +00:00
42fbda83bb
layer options moves into create hnet ui
2022-10-19 14:30:33 +00:00
7f8670c4ef
Merge branch 'master' into master
2022-10-19 15:18:45 +09:00
da72becb13
Use training width/height when training hypernetworks.
2022-10-19 09:13:28 +03:00
e40ba281f1
update
2022-10-19 01:03:58 +09:00
a5611ea502
update
2022-10-19 01:00:01 +09:00
6021f7a75f
add options to custom hypernetwork layer structure
2022-10-19 00:51:36 +09:00
703e6d9e4e
check NaN for hypernetwork tuning
2022-10-15 17:15:26 +03:00
c7a86f7fe9
add option to use batch size for training
2022-10-15 09:24:59 +03:00
03d62538ae
remove duplicate code for log loss, add step, make it read from options rather than gradio input
2022-10-14 22:43:55 +03:00
326fe7d44b
Merge remote-tracking branch 'Melanpan/master'
2022-10-14 22:14:50 +03:00
c344ba3b32
add option to read generation params for learning previews from txt2img
2022-10-14 20:31:49 +03:00
354ef0da3b
add hypernetwork multipliers
2022-10-13 20:12:37 +03:00
8636b50aea
Add learn_rate to csv and removed a left-over debug statement
2022-10-13 12:37:58 +02:00
1cfc2a1898
Save a csv containing the loss while training
2022-10-12 23:36:29 +02:00
c3c8eef9fd
train: change filename processing to be more simple and configurable
...
train: make it possible to make text files with prompts
train: rework scheduler so that there's less repeating code in textual inversion and hypernets
train: move epochs setting to options
2022-10-12 20:49:47 +03:00
ee015a1af6
change textual inversion tab to train
...
remake train interface to use tabs
2022-10-12 11:05:57 +03:00
2d006ce16c
xy_grid: Find hypernetwork by closest name
2022-10-12 10:40:10 +03:00
6be32b31d1
reports that training with medvram is possible.
2022-10-11 23:07:09 +03:00
d6fcc6b87b
apply lr schedule to hypernets
2022-10-11 22:03:05 +03:00
6a9ea5b41c
prevent extra modules from being saved/loaded with hypernet
2022-10-11 19:22:30 +03:00
d4ea5f4d86
add an option to unload models during hypernetwork training to save VRAM
2022-10-11 19:03:08 +03:00
6d09b8d1df
produce error when training with medvram/lowvram enabled
2022-10-11 18:33:57 +03:00
d682444ecc
add option to select hypernetwork modules when creating
2022-10-11 18:04:47 +03:00
b0583be088
more renames
2022-10-11 15:54:34 +03:00
873efeed49
rename hypernetwork dir to hypernetworks to prevent clash with an old filename that people who use zip instead of git clone will have
2022-10-11 15:51:30 +03:00
AngelBottomless