Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | patch bug (SeverianVoid's comment on 5245c7a) | timntorres | 2022-10-26 | 1 | -1/+1 |
| | |||||
* | remove duplicate keys and lowercase | AngelBottomless | 2022-10-26 | 1 | -1/+1 |
| | |||||
* | Weight initialization and More activation func | AngelBottomless | 2022-10-26 | 1 | -9/+38 |
| | | | | | | | | | | | | | | | | 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 | ||||
* | check length for variance | AngelBottomless | 2022-10-24 | 1 | -2/+10 |
| | |||||
* | convert deque -> list | AngelBottomless | 2022-10-24 | 1 | -1/+1 |
| | | | I don't feel this being efficient | ||||
* | statistics for pbar | AngelBottomless | 2022-10-24 | 1 | -2/+10 |
| | |||||
* | cleanup some code | AngelBottomless | 2022-10-24 | 1 | -11/+3 |
| | |||||
* | Hypernetworks - fix KeyError in statistics caching | AngelBottomless | 2022-10-24 | 1 | -2/+2 |
| | | | Statistics logging has changed to {filename : list[losses]}, so it has to use loss_info[key].pop() | ||||
* | Update hypernetwork.py | DepFA | 2022-10-23 | 1 | -4/+7 |
| | |||||
* | Allow tracking real-time loss | AngelBottomless | 2022-10-22 | 1 | -1/+1 |
| | | | | 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. | ||||
* | Update hypernetwork.py | AngelBottomless | 2022-10-22 | 1 | -11/+44 |
| | |||||
* | small fix | discus0434 | 2022-10-22 | 1 | -4/+3 |
| | |||||
* | Merge branch 'AUTOMATIC1111:master' into master | discus0434 | 2022-10-22 | 1 | -0/+11 |
|\ | |||||
| * | added a guard for hypernet training that will stop early if weights are ↵ | AUTOMATIC | 2022-10-22 | 1 | -0/+11 |
| | | | | | | | | getting no gradients | ||||
* | | small fix | discus0434 | 2022-10-22 | 1 | -7/+5 |
| | | |||||
* | | add an option to avoid dying relu | discus0434 | 2022-10-22 | 1 | -6/+6 |
| | | |||||
* | | Merge branch 'master' of upstream | discus0434 | 2022-10-22 | 1 | -8/+18 |
|\| | |||||
| * | Remove unused variable. | timntorres | 2022-10-21 | 1 | -1/+0 |
| | | |||||
| * | Match hypernet name with filename in all cases. | timntorres | 2022-10-21 | 1 | -1/+7 |
| | | |||||
| * | turns out LayerNorm also has weight and bias and needs to be pre-multiplied ↵ | AUTOMATIC | 2022-10-21 | 1 | -2/+2 |
| | | | | | | | | and trained for hypernets | ||||
| * | Merge branch 'master' into training-help-text | AUTOMATIC1111 | 2022-10-21 | 1 | -40/+56 |
| |\ | |||||
| | * | Revise comments. | timntorres | 2022-10-21 | 1 | -1/+1 |
| | | | |||||
| | * | Issue #2921-Give PNG info to Hypernet previews. | timntorres | 2022-10-21 | 1 | -2/+7 |
| | | | |||||
| | * | a more strict check for activation type and a more reasonable check for type ↵ | AUTOMATIC | 2022-10-21 | 1 | -3/+9 |
| | | | | | | | | | | | | of layer in hypernets | ||||
| * | | change html output | DepFA | 2022-10-19 | 1 | -1/+1 |
| | | | |||||
* | | | add dropout | discus0434 | 2022-10-22 | 1 | -26/+42 |
| | | | |||||
* | | | Revert "fix bugs and optimizations" | aria1th | 2022-10-20 | 1 | -59/+46 |
| | | | | | | | | | | | | This reverts commit 108be15500aac590b4e00420635d7b61fccfa530. | ||||
* | | | fix bugs and optimizations | AngelBottomless | 2022-10-20 | 1 | -46/+59 |
| | | | |||||
* | | | only linear | AngelBottomless | 2022-10-20 | 1 | -5/+5 |
| | | | |||||
* | | | generalized some functions and option for ignoring first layer | AngelBottomless | 2022-10-20 | 1 | -8/+15 |
| |/ |/| | |||||
* | | update | discus0434 | 2022-10-20 | 1 | -10/+19 |
| | | |||||
* | | fix for #3086 failing to load any previous hypernet | discus0434 | 2022-10-19 | 1 | -32/+28 |
|/ | |||||
* | layer options moves into create hnet ui | discus0434 | 2022-10-19 | 1 | -32/+32 |
| | |||||
* | Merge branch 'master' into master | discus0434 | 2022-10-19 | 1 | -3/+2 |
|\ | |||||
| * | Use training width/height when training hypernetworks. | Silent | 2022-10-19 | 1 | -2/+2 |
| | | |||||
* | | update | discus0434 | 2022-10-18 | 1 | -2/+2 |
| | | |||||
* | | update | discus0434 | 2022-10-18 | 1 | -6/+8 |
| | | |||||
* | | add options to custom hypernetwork layer structure | discus0434 | 2022-10-18 | 1 | -21/+67 |
|/ | |||||
* | check NaN for hypernetwork tuning | AngelBottomless | 2022-10-15 | 1 | -4/+6 |
| | |||||
* | add option to use batch size for training | AUTOMATIC | 2022-10-15 | 1 | -9/+24 |
| | |||||
* | remove duplicate code for log loss, add step, make it read from options ↵ | AUTOMATIC | 2022-10-14 | 1 | -14/+6 |
| | | | | rather than gradio input | ||||
* | Merge remote-tracking branch 'Melanpan/master' | AUTOMATIC | 2022-10-14 | 1 | -0/+15 |
|\ | |||||
| * | Add learn_rate to csv and removed a left-over debug statement | Melan | 2022-10-13 | 1 | -3/+3 |
| | | |||||
| * | Save a csv containing the loss while training | Melan | 2022-10-12 | 1 | -1/+16 |
| | | |||||
* | | add option to read generation params for learning previews from txt2img | AUTOMATIC | 2022-10-14 | 1 | -5/+16 |
| | | |||||
* | | add hypernetwork multipliers | AUTOMATIC | 2022-10-13 | 1 | -1/+7 |
|/ | |||||
* | train: change filename processing to be more simple and configurable | AUTOMATIC | 2022-10-12 | 1 | -24/+16 |
| | | | | | | 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 | ||||
* | change textual inversion tab to train | AUTOMATIC | 2022-10-12 | 1 | -1/+1 |
| | | | | remake train interface to use tabs | ||||
* | xy_grid: Find hypernetwork by closest name | Milly | 2022-10-12 | 1 | -0/+11 |
| | |||||
* | apply lr schedule to hypernets | AUTOMATIC | 2022-10-11 | 1 | -4/+15 |
| |