Commits

Breadcrumbs

History for

onmaster

User selector

All users

Datepicker

All time

Commit History

Commits on Aug 6, 2023

• support sai model spec

  kohya-ss

  committed
  c142dad

  Copy full SHA for c142dad

Commits on Jul 30, 2023

• add adding v-pred like loss for noise pred

  kohya-ss

  committed
  0636399

  Copy full SHA for 0636399

Commits on Jul 22, 2023

• fix to work text encoder only network with bf16

  kohya-ss

  committed
  8ba02ac

  Copy full SHA for 8ba02ac

Commits on Jul 20, 2023

• Merge pull request bmaltais#630 from ddPn08/sdxl

  Show description for 73a08c0
  kohya-ss

  authored
  73a08c0

  Copy full SHA for 73a08c0

• make to work with PyTorch 1.12

  kohya-ss

  committed
  acf16c0

  Copy full SHA for acf16c0

Commits on Jul 18, 2023

• enable full bf16 trainint in train_network

  kohya-ss

  committed
  225e871

  Copy full SHA for 225e871

• Merge pull request bmaltais#645 from Ttl/prepare_order

  Show description for 7875ca8
  kohya-ss

  authored
  7875ca8

  Copy full SHA for 7875ca8

• add zero_terminal_snr option

  kohya-ss

  committed
  6d2d8df

  Copy full SHA for 6d2d8df

Commits on Jul 13, 2023

• Cast weights to correct precision before transferring them to GPU

  Ttl

  committed
  cdffd19

  Copy full SHA for cdffd19

Commits on Jul 11, 2023

• make tracker init_kwargs configurable

  ddPn08

  committed
  b841dd7

  Copy full SHA for b841dd7

Commits on Jul 9, 2023

• support multi gpu in caching text encoder outputs

  kohya-ss

  committed
  0416f26

  Copy full SHA for 0416f26

Commits on Jul 3, 2023

• add min/max_timestep

  kohya-ss

  committed
  ea18246

  Copy full SHA for ea18246

Commits on Jun 29, 2023

• fix max_token_length not works for sdxl

  kohya-ss

  committed
  d395bc0

  Copy full SHA for d395bc0

Commits on Jun 26, 2023

• fix typos

  kohya-ss

  committed
  9ebebb2

  Copy full SHA for 9ebebb2

• Add no_half_vae for SDXL training, add nan check

  kohya-ss

  committed
  2c461e4

  Copy full SHA for 2c461e4

Commits on Jun 25, 2023

• add sdxl fine-tuning and LoRA

  kohya-ss

  committed
  747af14

  Copy full SHA for 747af14

Commits on Jun 21, 2023

• fix training scripts except controlnet not working

  kohya-ss

  committed
  5114e8d

  Copy full SHA for 5114e8d

Commits on Jun 17, 2023

• Merge branch 'original-u-net' into dev

  kohya-ss

  authored
  92e5013

  Copy full SHA for 92e5013

Commits on Jun 16, 2023

• Merge branch 'main' into original-u-net

  kohya-ss

  committed
  19dfa24

  Copy full SHA for 19dfa24

Commits on Jun 15, 2023

• Support for Prodigy(Dadapt variety for Dylora) (bmaltais#585)

  Show description for e97d67a
  sdbds

  authored
  e97d67a

  Copy full SHA for e97d67a

• add arbitrary dataset feature to each script

  kohya-ss

  committed
  9806b00

  Copy full SHA for 9806b00

Commits on Jun 14, 2023

• support arbitrary dataset for train_network.py

  kohya-ss

  committed
  9aee793

  Copy full SHA for 9aee793

Commits on Jun 11, 2023

• support sdpa

  kohya-ss

  committed
  9e1683c

  Copy full SHA for 9e1683c

• remove workaround for accelerator=0.15, fix XTI

  kohya-ss

  committed
  0315611

  Copy full SHA for 0315611

Commits on Jun 6, 2023

• fix to work LyCORIS<0.1.6

  kohya-ss

  committed
  bb91a10

  Copy full SHA for bb91a10

Commits on Jun 3, 2023

• loss表示追加

  u-haru

  committed
  5907bbd

  Copy full SHA for 5907bbd

• move max_norm to lora to avoid crashing in lycoris

  kohya-ss

  committed
  5bec05e

  Copy full SHA for 5bec05e

• scale v-pred loss like noise pred

  kohya-ss

  committed
  ec2efe5

  Copy full SHA for ec2efe5

Commits on Jun 1, 2023

• update diffusers to 1.16 | train_network

  ddPn08

  committed
  c8d209d

  Copy full SHA for c8d209d

• fix crash gen script, change to network_dropout

  kohya-ss

  committed
  f8e8df5

  Copy full SHA for f8e8df5

• add scaling to max norm

  kohya-ss

  committed
  f4c9276

  Copy full SHA for f4c9276

• fix crashing when max_norm is diabled

  kohya-ss

  committed
  a5c38e5

  Copy full SHA for a5c38e5

• Dropout and Max Norm Regularization for LoRA training (bmaltais#545)

  Show description for 9c72371
  

  AI-Casanova

  and

  kohya-ss

  authored
  9c72371

  Copy full SHA for 9c72371

• Adjusted English grammar in logs to be more clear (bmaltais#554)

  Show description for 5931948
  TingTingin

  authored
  5931948

  Copy full SHA for 5931948

Commits on May 31, 2023

• show multiplier for base weights to console

  kohya-ss

  committed
  6fbd526

  Copy full SHA for 6fbd526

Pagination

Previous

Next