Kohya S
326c425a5b
feat: enhance LoRA weight handling in model loading and add text encoder loading function
2026-02-12 22:24:48 +09:00
Kohya S
cc1f57bc70
fix: review with Copilot
2026-02-12 08:05:41 +09:00
Kohya S.
57aa70ea9e
Update library/anima_train_utils.py
...
Remove unused import.
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com >
2026-02-12 08:02:05 +09:00
Kohya S.
07df0e54a0
Update library/safetensors_utils.py
...
Fix duplicated new_key for concat_hook.
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com >
2026-02-11 22:36:19 +09:00
Kohya S.
82f963650b
Update library/strategy_anima.py
...
Use torch.all instead of all.
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com >
2026-02-11 22:28:00 +09:00
Kohya S
59267d19f3
feat: optimize RMSNorm forward method and remove unused torch_attention_op
2026-02-11 22:06:47 +09:00
Kohya S
4b2283491e
feat: add VAE chunking and caching options to reduce memory usage
2026-02-11 21:32:00 +09:00
Kohya S
a7cd38dcaf
feat: split CFG processing in do_sample function to reduce memory usage
2026-02-11 18:00:10 +09:00
Kohya S
9349c91c89
fix: remove unused code
2026-02-11 17:23:58 +09:00
Kohya S
b67cc5a457
feat: update Anima SAI model spec metadata handling
2026-02-11 14:56:23 +09:00
Kohya S
dbb40ae4c0
fix: wait all block trasfer before siwtching offloader mode
2026-02-10 21:46:58 +09:00
Kohya S
8d2d286a13
feat: refactor Anima training script
2026-02-10 21:27:22 +09:00
Kohya S
6d08c93b23
feat: enhance block swap functionality for inference and training in Anima model
2026-02-10 21:26:50 +09:00
Kohya S
02a75944b3
fix: keep latents 4D except DiT call
2026-02-10 21:26:20 +09:00
Kohya S
58db77a488
fix: update to work fp8_scaled option
2026-02-09 23:31:33 +09:00
Kohya S
0f413974b7
feat: support LoRA format without net. prefix
2026-02-09 23:21:04 +09:00
kohya-ss
8f5b298906
feat: simplify Anima training script and update empty caption handling
2026-02-09 12:46:53 +09:00
kohya-ss
06dcb30016
fix: update argument names in anima_train_utils to align with other archtectures
2026-02-09 12:46:04 +09:00
kohya-ss
bf9806648c
feat: support 5d tensor in get_noisy_model_input_and_timesteps
2026-02-09 12:43:37 +09:00
kohya-ss
2774e7757b
feat: add dtype property and all-zero mask handling in cross-attention in LLMAdapterTransformerBlock
2026-02-09 12:43:11 +09:00
kohya-ss
4f6511bf28
feat: kept caption dropout rate in cache and handle in training script
2026-02-08 15:35:53 +09:00
kohya-ss
d992037984
format: format
2026-02-08 12:22:54 +09:00
kohya-ss
5c4643516c
feat: loading with dynamic fp8 optimization and LoRA support
2026-02-08 12:17:02 +09:00
kohya-ss
44b8d79577
feat: use unified attention module, add wrapper for state dict compatibility
2026-02-08 12:16:00 +09:00
kohya-ss
10445ff660
feat: simplify encode_tokens
2026-02-08 12:06:12 +09:00
kohya-ss
7b0ed3269a
feat: add Qwen-Image VAE
2026-02-08 11:13:09 +09:00
duongve13112002
e21a7736f8
Support Anima model ( #2260 )
...
* Support Anima model
* Update document and fix bug
* Fix latent normlization
* Fix typo
* Fix cache embedding
* fix typo in tests/test_anima_cache.py
* Remove redundant argument apply_t5_attn_mask
* Improving caching with argument caption_dropout_rate
* Fix W&B logging bugs
* Fix discrete_flow_shift default value
2026-02-08 10:18:55 +09:00
Kohya S.
c6bc632ec6
fix: metadata dataset degradation and make it work ( #2186 )
...
* fix: support dataset with metadata
* feat: support another tagger model
* fix: improve handling of image size and caption/tag processing in FineTuningDataset
* fix: enhance metadata loading to support JSONL format in FineTuningDataset
* feat: enhance image loading and processing in ImageLoadingPrepDataset with batch support and output options
* fix: improve image path handling and memory management in dataset classes
* Update finetune/tag_images_by_wd14_tagger.py
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com >
* fix: add return type annotation for process_tag_replacement function and ensure tags are returned
* feat: add artist category threshold for tagging
* doc: add comment for clarification
---------
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com >
2026-01-18 15:17:07 +09:00
urlesistiana
f7fc7ddda2
fix #2201 : lumina 2 timesteps handling
2025-10-13 16:08:28 +08:00
Kohya S
5462a6bb24
Merge branch 'dev' into sd3
2025-09-29 21:02:02 +09:00
Kohya S
63711390a0
Merge branch 'main' into dev
2025-09-29 20:56:07 +09:00
Kohya S
60bfa97b19
fix: disable_mmap_safetensors not defined in SDXL TI training
2025-09-29 20:52:48 +09:00
Kohya S.
e7b89826c5
Update library/custom_offloading_utils.py
...
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com >
2025-09-21 13:29:58 +09:00
Kohya S
806d535ef1
fix: block-wise scaling is overwritten by per-tensor scaling
2025-09-21 13:10:41 +09:00
Kohya S
3876343fad
fix: remove print statement for guidance rescale in AdaptiveProjectedGuidance
2025-09-21 13:09:38 +09:00
Kohya S
040d976597
feat: add guidance rescale options for Adaptive Projected Guidance in inference
2025-09-21 13:03:14 +09:00
Kohya S
9621d9d637
feat: add Adaptive Projected Guidance parameters and noise rescaling
2025-09-21 12:34:40 +09:00
Kohya S
f41e9e2b58
feat: add vae_chunk_size argument for memory-efficient VAE decoding and processing
2025-09-21 11:09:37 +09:00
Kohya S
b090d15f7d
feat: add multi backend attention and related update for HI2.1 models and scripts
2025-09-20 19:45:33 +09:00
Kohya S
f834b2e0d4
fix: --fp8_vl to work
2025-09-18 23:46:18 +09:00
Kohya S
f6b4bdc83f
feat: block-wise fp8 quantization
2025-09-18 21:20:54 +09:00
Kohya S
f5b004009e
fix: correct tensor indexing in HunyuanVAE2D class for blending and encoding functions
2025-09-17 21:54:25 +09:00
Kohya S
4e2a80a6ca
refactor: update imports to use safetensors_utils for memory-efficient operations
2025-09-13 21:07:11 +09:00
Kohya S
d831c88832
fix: sample generation doesn't work with block swap
2025-09-13 21:06:04 +09:00
Kohya S
bae7fa74eb
Merge branch 'sd3' into feat-hunyuan-image-2.1-inference
2025-09-13 20:13:58 +09:00
Kohya S.
e1c666e97f
Update library/safetensors_utils.py
...
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com >
2025-09-13 20:03:55 +09:00
Kohya S
8783f8aed3
feat: faster safetensors load and split safetensor utils
2025-09-13 19:51:38 +09:00
Kohya S
209c02dbb6
feat: HunyuanImage LoRA training
2025-09-12 21:40:42 +09:00
Kohya S
a0f0afbb46
fix: revert constructor signature update
2025-09-11 22:27:00 +09:00
Kohya S
7f983c558d
feat: block swap for inference and initial impl for HunyuanImage LoRA (not working)
2025-09-11 22:15:22 +09:00
