Commit message (Expand) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | Added prompt dropout | Volpeon | 2023-06-22 | 1 | -1/+1 |
* | Remove training guidance_scale | Volpeon | 2023-06-22 | 1 | -7/+0 |
* | Update | Volpeon | 2023-06-21 | 1 | -8/+0 |
* | Update | Volpeon | 2023-06-21 | 1 | -220/+269 |
* | Fix LoRA training with DAdan | Volpeon | 2023-05-16 | 1 | -31/+27 |
* | Avoid model recompilation due to varying prompt lengths | Volpeon | 2023-05-16 | 1 | -1/+2 |
* | Patch xformers to cast dtypes | Volpeon | 2023-05-16 | 1 | -1/+12 |
* | LoRA: Apply to out layers as well | Volpeon | 2023-05-16 | 1 | -2/+4 |
* | Support LoRA training for token embeddings | Volpeon | 2023-05-16 | 1 | -8/+24 |
* | Fix for latest PEFT | Volpeon | 2023-05-12 | 1 | -11/+3 |
* | Update | Volpeon | 2023-05-11 | 1 | -1/+1 |
* | Update | Volpeon | 2023-05-06 | 1 | -0/+17 |
* | Update | Volpeon | 2023-05-05 | 1 | -0/+8 |
* | torch.compile won't work yet, keep code prepared | Volpeon | 2023-04-29 | 1 | -2/+17 |
* | Optional xformers | Volpeon | 2023-04-29 | 1 | -3/+10 |
* | Fixed model compilation | Volpeon | 2023-04-29 | 1 | -0/+6 |
* | Support torch.compile | Volpeon | 2023-04-28 | 1 | -0/+10 |
* | Fix | Volpeon | 2023-04-27 | 1 | -8/+19 |
* | Update | Volpeon | 2023-04-27 | 1 | -2/+11 |
* | Fix cycle loop | Volpeon | 2023-04-23 | 1 | -3/+5 |
* | Update | Volpeon | 2023-04-21 | 1 | -2/+4 |
* | Fix PTI | Volpeon | 2023-04-20 | 1 | -2/+4 |
* | Update | Volpeon | 2023-04-20 | 1 | -29/+116 |
* | Fix | Volpeon | 2023-04-17 | 1 | -3/+2 |
* | Improved automation caps | Volpeon | 2023-04-16 | 1 | -20/+33 |
* | Added option to use constant LR on cycles > 1 | Volpeon | 2023-04-16 | 1 | -2/+11 |
* | Fix | Volpeon | 2023-04-16 | 1 | -1/+10 |
* | Update | Volpeon | 2023-04-16 | 1 | -1/+5 |
* | TI via LoRA | Volpeon | 2023-04-15 | 1 | -3/+4 |
* | Added cycle LR decay | Volpeon | 2023-04-13 | 1 | -5/+7 |
* | Update | Volpeon | 2023-04-13 | 1 | -27/+46 |
* | Update | Volpeon | 2023-04-11 | 1 | -7/+16 |
* | Randomize dataset across cycles | Volpeon | 2023-04-10 | 1 | -1/+3 |
* | Update | Volpeon | 2023-04-10 | 1 | -22/+44 |
* | Update | Volpeon | 2023-04-09 | 1 | -7/+14 |
* | Update | Volpeon | 2023-04-09 | 1 | -6/+1 |
* | Update | Volpeon | 2023-04-09 | 1 | -128/+12 |
* | Fix | Volpeon | 2023-04-09 | 1 | -4/+6 |
* | Made Lora script interactive | Volpeon | 2023-04-09 | 1 | -40/+71 |
* | Update | Volpeon | 2023-04-08 | 1 | -8/+32 |
* | Update | Volpeon | 2023-04-08 | 1 | -10/+22 |
* | Update | Volpeon | 2023-04-08 | 1 | -34/+41 |
* | Fix | Volpeon | 2023-04-07 | 1 | -35/+42 |
* | Fixed Lora PTI | Volpeon | 2023-04-07 | 1 | -17/+20 |
* | Fix | Volpeon | 2023-04-07 | 1 | -5/+8 |
* | Run PTI only if placeholder tokens arg isn't empty | Volpeon | 2023-04-07 | 1 | -54/+55 |
* | Fix | Volpeon | 2023-04-07 | 1 | -1/+1 |
* | Update | Volpeon | 2023-04-07 | 1 | -7/+22 |
* | Update | Volpeon | 2023-04-07 | 1 | -42/+261 |
* | Update | Volpeon | 2023-04-06 | 1 | -19/+32 |