kohya-ss/sd-scripts

Pull Requests Commits

feat: normalize key names when extracting tensors from tuned model

kohya-ss committed 244 days ago

65096900

Merge branch 'dev' into sd3

kohya-ss committed 244 days ago

5462a6bb

Merge branch 'main' into dev

kohya-ss committed 244 days ago

63711390

Merge pull request #2216 from kohya-ss/fix-sdxl-textual-inversion-training-disable-mmap

kohya-ss committed 244 days ago

Verified 206adb64

fix: disable_mmap_safetensors not defined in SDXL TI training

kohya-ss committed 244 days ago

60bfa97b

Merge pull request #2213 from kohya-ss/doc-hunyuan-image-training-text-encoder-cpu-note

kohya-ss committed 246 days ago

Verified f0c767e0

docs: enhance text encoder CPU usage instructions for HunyuanImage-2.1 training

kohya-ss committed 246 days ago

a0c26a0e

Merge pull request #2212 from kohya-ss/fix-hunyuan-image-sample-generation

kohya-ss committed 246 days ago

Verified 67d06213

feat: add new parameters for sample image inference configuration

kohya-ss committed 246 days ago

6a826d21

Merge pull request #2207 from kohya-ss/fix-flux-extract-lora-metadata-failed

kohya-ss committed 249 days ago

Verified 4c197a53

fix: update metadata construction to include model_config for flux

kohya-ss committed 249 days ago

4b79d735

Merge pull request #2198 from kohya-ss/feat-hunyuan-image-2.1-inference

kohya-ss committed 251 days ago

Verified 121853ca

doc: update README with HunyuanImage-2.1 LoRA training details and requirements

kohya-ss committed 251 days ago

58df9dff

doc: add --network_train_unet_only option for HunyuanImage-2.1 training

kohya-ss committed 251 days ago

31f7df3b

Update hunyuan_image_train_network.py

kohya-ss committed 253 days ago

Verified 753c7945

Update library/custom_offloading_utils.py

kohya-ss committed 253 days ago

Verified e7b89826

fix: block-wise scaling is overwritten by per-tensor scaling

kohya-ss committed 253 days ago

806d535e

fix: remove print statement for guidance rescale in AdaptiveProjectedGuidance

kohya-ss committed 253 days ago

3876343f

feat: add guidance rescale options for Adaptive Projected Guidance in inference

kohya-ss committed 253 days ago

040d9765

feat: add Adaptive Projected Guidance parameters and noise rescaling

kohya-ss committed 253 days ago

9621d9d6

doc: add --vae_chunk_size option for training and inference

kohya-ss committed 253 days ago

e7b8e9a7

feat: add vae_chunk_size argument for memory-efficient VAE decoding and processing

kohya-ss committed 253 days ago

f41e9e2b

feat: add --text_encoder_cpu option to reduce VRAM usage by running text encoders on CPU for training

kohya-ss committed 253 days ago

8f20c379

feat: add multi backend attention and related update for HI2.1 models and scripts

kohya-ss committed 253 days ago

b090d15f

fix: --fp8_vl to work

kohya-ss committed 255 days ago

f834b2e0

feat: block-wise fp8 quantization

kohya-ss committed 255 days ago

f6b4bdc8

fix: fp8 casting not working

kohya-ss committed 255 days ago

2ce506e1

fix: correct tensor indexing in HunyuanVAE2D class for blending and encoding functions

kohya-ss committed 256 days ago

f5b00400

feat: add conversion script for LoRA models to ComfyUI format with reverse option

kohya-ss committed 257 days ago

cbe2a9da

docs: update HunyuanImage-2.1 training guide with model download instructions and VRAM optimization settings (by Claude)

kohya-ss committed 257 days ago

f318ddae

Older