CCalamitousFelicitousnessfix(lora): apply full-diff norm targets and map z-image's renamed modules
| 文件 | 最后提交记录 | 最后更新时间 |
|---|---|---|
feat(lora): support the lycoris_ prefix in every native adapter arch The lycoris_ save format is arch-independent: LyCORIS standalone wraps the loaded diffusers model and emits the wrapped module path with dots as underscores, so verbatim passthrough is correct for any arch. Only flux2 handled it; zimage, chroma, ernie and krea2 reported such files as not loaded. - add lycoris_ to KNOWN_PREFIXES_DEFAULT and PASSTHROUGH_PREFIXES_DEFAULT - drop flux2's per-arch prefix append and resolve_targets branch - add lycoris_ to ANIMA_PREFIXES (anima replaces the default tuple); network_prefix_for already routes it to the transformer namespace - cover the passthrough with a zimage loader test | 1 个月前 | |
add boogu Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
full codespell coverage Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
feat(lora): run the full adapter family chain on every native arch zimage, chroma, ernie and krea2 chained only lora/lokr/loha/oft while flux2 and anima ran all eight families, so ia3/glora/norm/full files (e.g. full-diff extractions with diff/diff_b keys) reported not loaded on the short-chain arches. The generic family loaders are arch-agnostic; wire the missing four into each chain. - add ia3/glora/norm/full wrappers and chain entries to the four arch modules, with matching suffix/marker re-exports - add a chain-level full-diff test per suite: zimage covers the legacy attention.out alias and the fused-qkv skip, chroma the proj rename, ernie the passthrough | 1 个月前 | |
refactor(pipelines): move per-arch transformer specs into packages Specs for chrono, cogview, flux2_klein, glm, hunyuandit, hyimage, joy, kandinsky, longcat, nucleus, ovis, pixart and prx lived module-level in model_<arch>.py. Move each into pipelines/<arch>/__init__.py to match the layout used by anima, bria, ernie, f_lite, lens, qwen, step1x and vibe. Each model_<arch>.py now imports its spec lazily inside the load function, so the package only gets pulled in when that arch is actually loaded (kandinsky 2.x never goes through native dispatch and stays untouched). Also drop the "without this spec..." crash paragraph from the new package docstrings plus ernie/__init__.py and bria/__init__.py. | 3 个月前 | |
refactor(pipelines): move per-arch transformer specs into packages Specs for chrono, cogview, flux2_klein, glm, hunyuandit, hyimage, joy, kandinsky, longcat, nucleus, ovis, pixart and prx lived module-level in model_<arch>.py. Move each into pipelines/<arch>/__init__.py to match the layout used by anima, bria, ernie, f_lite, lens, qwen, step1x and vibe. Each model_<arch>.py now imports its spec lazily inside the load function, so the package only gets pulled in when that arch is actually loaded (kandinsky 2.x never goes through native dispatch and stays untouched). Also drop the "without this spec..." crash paragraph from the new package docstrings plus ernie/__init__.py and bria/__init__.py. | 3 个月前 | |
update transformers/diffusers and enhanced tests Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
feat(lora): run the full adapter family chain on every native arch zimage, chroma, ernie and krea2 chained only lora/lokr/loha/oft while flux2 and anima ran all eight families, so ia3/glora/norm/full files (e.g. full-diff extractions with diff/diff_b keys) reported not loaded on the short-chain arches. The generic family loaders are arch-agnostic; wire the missing four into each chain. - add ia3/glora/norm/full wrappers and chain entries to the four arch modules, with matching suffix/marker re-exports - add a chain-level full-diff test per suite: zimage covers the legacy attention.out alias and the fused-qkv skip, chroma the proj rename, ernie the passthrough | 1 个月前 | |
minor fixes from audit Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
PEP 484 pipelines directory | 5 个月前 | |
feat(lora): support the lycoris_ prefix in every native adapter arch The lycoris_ save format is arch-independent: LyCORIS standalone wraps the loaded diffusers model and emits the wrapped module path with dots as underscores, so verbatim passthrough is correct for any arch. Only flux2 handled it; zimage, chroma, ernie and krea2 reported such files as not loaded. - add lycoris_ to KNOWN_PREFIXES_DEFAULT and PASSTHROUGH_PREFIXES_DEFAULT - drop flux2's per-arch prefix append and resolve_targets branch - add lycoris_ to ANIMA_PREFIXES (anima replaces the default tuple); network_prefix_for already routes it to the transformer namespace - cover the passthrough with a zimage loader test | 1 个月前 | |
refactor(pipelines): move per-arch transformer specs into packages Specs for chrono, cogview, flux2_klein, glm, hunyuandit, hyimage, joy, kandinsky, longcat, nucleus, ovis, pixart and prx lived module-level in model_<arch>.py. Move each into pipelines/<arch>/__init__.py to match the layout used by anima, bria, ernie, f_lite, lens, qwen, step1x and vibe. Each model_<arch>.py now imports its spec lazily inside the load function, so the package only gets pulled in when that arch is actually loaded (kandinsky 2.x never goes through native dispatch and stays untouched). Also drop the "without this spec..." crash paragraph from the new package docstrings plus ernie/__init__.py and bria/__init__.py. | 3 个月前 | |
refactor(pipelines): move per-arch transformer specs into packages Specs for chrono, cogview, flux2_klein, glm, hunyuandit, hyimage, joy, kandinsky, longcat, nucleus, ovis, pixart and prx lived module-level in model_<arch>.py. Move each into pipelines/<arch>/__init__.py to match the layout used by anima, bria, ernie, f_lite, lens, qwen, step1x and vibe. Each model_<arch>.py now imports its spec lazily inside the load function, so the package only gets pulled in when that arch is actually loaded (kandinsky 2.x never goes through native dispatch and stays untouched). Also drop the "without this spec..." crash paragraph from the new package docstrings plus ernie/__init__.py and bria/__init__.py. | 3 个月前 | |
full codespell coverage Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
refactor(pipelines): move per-arch transformer specs into packages Specs for chrono, cogview, flux2_klein, glm, hunyuandit, hyimage, joy, kandinsky, longcat, nucleus, ovis, pixart and prx lived module-level in model_<arch>.py. Move each into pipelines/<arch>/__init__.py to match the layout used by anima, bria, ernie, f_lite, lens, qwen, step1x and vibe. Each model_<arch>.py now imports its spec lazily inside the load function, so the package only gets pulled in when that arch is actually loaded (kandinsky 2.x never goes through native dispatch and stays untouched). Also drop the "without this spec..." crash paragraph from the new package docstrings plus ernie/__init__.py and bria/__init__.py. | 3 个月前 | |
refactor(pipelines): move per-arch transformer specs into packages Specs for chrono, cogview, flux2_klein, glm, hunyuandit, hyimage, joy, kandinsky, longcat, nucleus, ovis, pixart and prx lived module-level in model_<arch>.py. Move each into pipelines/<arch>/__init__.py to match the layout used by anima, bria, ernie, f_lite, lens, qwen, step1x and vibe. Each model_<arch>.py now imports its spec lazily inside the load function, so the package only gets pulled in when that arch is actually loaded (kandinsky 2.x never goes through native dispatch and stays untouched). Also drop the "without this spec..." crash paragraph from the new package docstrings plus ernie/__init__.py and bria/__init__.py. | 3 个月前 | |
fix(ideogram4): resolve activation dtype for quantized transformers transformer.dtype reports the storage dtype for fp8-quantized models (fp8 params are floating, int8 params are not), so activations were cast to fp8 and the forward crashed. Read the compute dtype from the SDNQ dequantizer instead. | 1 个月前 | |
refactor(pipelines): move per-arch transformer specs into packages Specs for chrono, cogview, flux2_klein, glm, hunyuandit, hyimage, joy, kandinsky, longcat, nucleus, ovis, pixart and prx lived module-level in model_<arch>.py. Move each into pipelines/<arch>/__init__.py to match the layout used by anima, bria, ernie, f_lite, lens, qwen, step1x and vibe. Each model_<arch>.py now imports its spec lazily inside the load function, so the package only gets pulled in when that arch is actually loaded (kandinsky 2.x never goes through native dispatch and stays untouched). Also drop the "without this spec..." crash paragraph from the new package docstrings plus ernie/__init__.py and bria/__init__.py. | 3 个月前 | |
refactor(pipelines): split kandinsky into kandinsky3 and kandinsky5 K3 and K5 share only the marketing name: different diffusers pipeline classes (Kandinsky3Pipeline vs Kandinsky5T2I/I2IPipeline), different denoiser classes (Kandinsky3UNet vs Kandinsky5Transformer3DModel), no shared scaffolding. Separate packages so each spec lives next to the arch it serves. K2.1 and K2.2 stay outside both packages since they go through KandinskyCombinedPipeline.from_pretrained directly with no spec. | 3 个月前 | |
refactor(pipelines): split kandinsky into kandinsky3 and kandinsky5 K3 and K5 share only the marketing name: different diffusers pipeline classes (Kandinsky3Pipeline vs Kandinsky5T2I/I2IPipeline), different denoiser classes (Kandinsky3UNet vs Kandinsky5Transformer3DModel), no shared scaffolding. Separate packages so each spec lives next to the arch it serves. K2.1 and K2.2 stay outside both packages since they go through KandinskyCombinedPipeline.from_pretrained directly with no spec. | 3 个月前 | |
feat(lora): run the full adapter family chain on every native arch zimage, chroma, ernie and krea2 chained only lora/lokr/loha/oft while flux2 and anima ran all eight families, so ia3/glora/norm/full files (e.g. full-diff extractions with diff/diff_b keys) reported not loaded on the short-chain arches. The generic family loaders are arch-agnostic; wire the missing four into each chain. - add ia3/glora/norm/full wrappers and chain entries to the four arch modules, with matching suffix/marker re-exports - add a chain-level full-diff test per suite: zimage covers the legacy attention.out alias and the fused-qkv skip, chroma the proj rename, ernie the passthrough | 1 个月前 | |
feat(pipelines): wire native_spec for Qwen-Image and custom-class arches Qwen-Image: diffusers registers a no-op identity lambda for QwenImageTransformer2DModel; from_single_file silently accepted any key naming and loaded with mismatches. QWEN_SPEC sets converter=None to short-circuit the no-op. Custom transformer classes (Lens, Step1X-Edit, VIBE Sana, f-lite): runtime-registered onto diffusers, never in SINGLE_FILE_LOADABLE_CLASSES. Specs land in each arch's existing pipelines/<arch>/__init__.py. NextStep skipped: loader returns None up front. | 3 个月前 | |
refactor(pipelines): move per-arch transformer specs into packages Specs for chrono, cogview, flux2_klein, glm, hunyuandit, hyimage, joy, kandinsky, longcat, nucleus, ovis, pixart and prx lived module-level in model_<arch>.py. Move each into pipelines/<arch>/__init__.py to match the layout used by anima, bria, ernie, f_lite, lens, qwen, step1x and vibe. Each model_<arch>.py now imports its spec lazily inside the load function, so the package only gets pulled in when that arch is actually loaded (kandinsky 2.x never goes through native dispatch and stays untouched). Also drop the "without this spec..." crash paragraph from the new package docstrings plus ernie/__init__.py and bria/__init__.py. | 3 个月前 | |
schedulers remove false positives Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
minor fixes from audit Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
refactor(pipelines): move per-arch transformer specs into packages Specs for chrono, cogview, flux2_klein, glm, hunyuandit, hyimage, joy, kandinsky, longcat, nucleus, ovis, pixart and prx lived module-level in model_<arch>.py. Move each into pipelines/<arch>/__init__.py to match the layout used by anima, bria, ernie, f_lite, lens, qwen, step1x and vibe. Each model_<arch>.py now imports its spec lazily inside the load function, so the package only gets pulled in when that arch is actually loaded (kandinsky 2.x never goes through native dispatch and stays untouched). Also drop the "without this spec..." crash paragraph from the new package docstrings plus ernie/__init__.py and bria/__init__.py. | 3 个月前 | |
strict ruff without exclude Co-authored-by: Copilot <copilot@github.com> Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
refactor(pipelines): move per-arch transformer specs into packages Specs for chrono, cogview, flux2_klein, glm, hunyuandit, hyimage, joy, kandinsky, longcat, nucleus, ovis, pixart and prx lived module-level in model_<arch>.py. Move each into pipelines/<arch>/__init__.py to match the layout used by anima, bria, ernie, f_lite, lens, qwen, step1x and vibe. Each model_<arch>.py now imports its spec lazily inside the load function, so the package only gets pulled in when that arch is actually loaded (kandinsky 2.x never goes through native dispatch and stays untouched). Also drop the "without this spec..." crash paragraph from the new package docstrings plus ernie/__init__.py and bria/__init__.py. | 3 个月前 | |
refactor(pipelines): move per-arch transformer specs into packages Specs for chrono, cogview, flux2_klein, glm, hunyuandit, hyimage, joy, kandinsky, longcat, nucleus, ovis, pixart and prx lived module-level in model_<arch>.py. Move each into pipelines/<arch>/__init__.py to match the layout used by anima, bria, ernie, f_lite, lens, qwen, step1x and vibe. Each model_<arch>.py now imports its spec lazily inside the load function, so the package only gets pulled in when that arch is actually loaded (kandinsky 2.x never goes through native dispatch and stays untouched). Also drop the "without this spec..." crash paragraph from the new package docstrings plus ernie/__init__.py and bria/__init__.py. | 3 个月前 | |
refactor(pipelines): move per-arch transformer specs into packages Specs for chrono, cogview, flux2_klein, glm, hunyuandit, hyimage, joy, kandinsky, longcat, nucleus, ovis, pixart and prx lived module-level in model_<arch>.py. Move each into pipelines/<arch>/__init__.py to match the layout used by anima, bria, ernie, f_lite, lens, qwen, step1x and vibe. Each model_<arch>.py now imports its spec lazily inside the load function, so the package only gets pulled in when that arch is actually loaded (kandinsky 2.x never goes through native dispatch and stays untouched). Also drop the "without this spec..." crash paragraph from the new package docstrings plus ernie/__init__.py and bria/__init__.py. | 3 个月前 | |
docs: cleanup | 3 个月前 | |
refactor for consistent folders usage for all models Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
feat(pipelines): wire native_spec for Qwen-Image and custom-class arches Qwen-Image: diffusers registers a no-op identity lambda for QwenImageTransformer2DModel; from_single_file silently accepted any key naming and loaded with mismatches. QWEN_SPEC sets converter=None to short-circuit the no-op. Custom transformer classes (Lens, Step1X-Edit, VIBE Sana, f-lite): runtime-registered onto diffusers, never in SINGLE_FILE_LOADABLE_CLASSES. Specs land in each arch's existing pipelines/<arch>/__init__.py. NextStep skipped: loader returns None up front. | 3 个月前 | |
full codespell coverage Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
full codespell coverage Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
PEP 484 pipelines directory | 5 个月前 | |
refactor for consistent folders usage for all models Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
fix(lora): apply full-diff norm targets and map z-image's renamed modules A full-weight extraction on Z-Image bound 308 modules and applied 172 of them, silently dropping the rest, and left 71 more unmapped. assign_network_names_to_compvis_modules puts every transformer module in network_layer_mapping but skips stamping network_layer_name on norms, which is the attribute the apply pass keys off. try_load_full bound those modules through the mapping and they then never applied; stamp them loader-locally, as try_load_norm already does. Z-Image also names three module groups differently from the diffusers tree: the qk-norms (q_norm/k_norm vs norm_q/norm_k), and the patch embedder and final layer, which live in ModuleDicts keyed by "{patch_size}-{f_patch_size}" and so carry a key the checkpoint has no notion of. Read that key from the live model rather than hardcoding it. The counts close exactly: 68 qk-norms plus 3 non-block targets are the 71 that went unmapped. | 1 个月前 | |
ai-agents Signed-off-by: vladmandic <mandic00@live.com> | 4 个月前 | |
attempt to load incomplete model Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
attempt to load incomplete model Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
add boogu Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
automated pipeline registrations and tests Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
fix(te): quantize single-file text-encoder overrides under SDNQ The single-file override loaded via from_pretrained(state_dict=...) without a quantization_config and then relied on do_post_load_quant, which no-ops on a bare module (it maps Model/TE onto pipeline attributes that a lone encoder does not have). A custom TE therefore loaded full-precision regardless of the SDNQ setting. Pass quant_args into from_pretrained like the repo/default path so SDNQ quantizes during load. | 2 个月前 | |
feat(model): secondary unet override slot and ideogram4 native loading One UNET override cannot serve dual-transformer arches: ideogram4 conditional/unconditional and wan combined-stage experts need separate files, and previously a single override landed on both experts. - sd_unet_secondary option with per-slot tracking, consumed-state sync, arch-change reset, and incompatible-override fallback - dropdown renders beside the primary, follows it into quicksettings, and is visible only for dual-transformer model types - ideogram4 native single-file spec with a quant-aware fused-qkv converter; such converters run before comfy_quant detection via TransformerSpec.converter_handles_quant - quicksettings render in configured order (sort keyed on the option object and always fell back to alphabetical) - post-load dtype warning skips quantized transformers | 1 个月前 | |
automated pipeline registrations and tests Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
fix: mutable default arguments + wrong cfg var gating eva_clip text checkpoint download Co-Authored-By: Claude <noreply@anthropic.com> | 2 个月前 | |
fix(anima): resolve text encoder through shared map allow_shared=False predates the Qwen-3 0.5B entries in shared_te_map, so every Anima repo downloaded its own copy of the text encoder. Remove it so all variants share the canonical Anima-1.0-Base text encoder. | 1 个月前 | |
enhance automated testing Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
add boogu Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
automated pipeline registrations and tests Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
enhance automated testing Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
enhance automated testing Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
enhance automated testing Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
update transformers/diffusers and enhanced tests Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
enhance automated testing Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
automated pipeline registrations and tests Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
automated pipeline registrations and tests Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
enhance automated testing Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
enhance automated testing Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
add experimental openai interface Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
enhance automated testing Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
Nano Banana lite add | 2 个月前 | |
automated pipeline registrations and tests Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
enhance automated testing Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
fix: dead hasattr-on-dict path, double video-vae normalization, undefined self.device, hyimage if/elif, attention-mask attr typo, wrong face logged Co-Authored-By: Claude <noreply@anthropic.com> | 2 个月前 | |
feat(model): secondary unet override slot and ideogram4 native loading One UNET override cannot serve dual-transformer arches: ideogram4 conditional/unconditional and wan combined-stage experts need separate files, and previously a single override landed on both experts. - sd_unet_secondary option with per-slot tracking, consumed-state sync, arch-change reset, and incompatible-override fallback - dropdown renders beside the primary, follows it into quicksettings, and is visible only for dual-transformer model types - ideogram4 native single-file spec with a quant-aware fused-qkv converter; such converters run before comfy_quant detection via TransformerSpec.converter_handles_quant - quicksettings render in configured order (sort keyed on the option object and always fell back to alphabetical) - post-load dtype warning skips quantized transformers | 1 个月前 | |
minor fixes from audit Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
add joy-image-edit-plus Signed-off-by: Vladimir Mandic <mandic00@live.com> | 1 个月前 | |
enhance automated testing Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
enhance automated testing Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
feat(krea2): add Krea 2 (K2) image model support Krea 2 is a 12.9B single-stream flow-matching DiT trained from scratch, using a Qwen3-VL-4B text encoder and the Qwen-Image VAE. The transformer is vendored as a diffusers ModelMixin whose module tree mirrors the checkpoint, so weights load with no key conversion; the pipeline ports the reference encode, flow-matching denoise, and VAE decode. The text encoder is shared at runtime via the existing dedup registry, so Base and Turbo reuse one Qwen3-VL-4B copy. Covers text-to-image, image-to-image, native LoRA, and the single-file UNET override. Also completes SD.Next's partial Qwen-Image VAE support (5D decode input and TAESD preview mapping) that K2 shares. | 2 个月前 | |
update lens and bria Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
enhance automated testing Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
refactor for consistent folders usage for all models Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
refactor for consistent folders usage for all models Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
enhance automated testing Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
enhance automated testing Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
refactor for consistent folders usage for all models Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
enhance automated testing Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
fixes based on full skills run Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
enhance automated testing Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
fix hf telemetry and add prxpixel Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
enhance automated testing Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
refactor for consistent folders usage for all models Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
enhance automated testing Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
lint fix Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
minor fixes from audit Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
refactor for consistent folders usage for all models Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
automated pipeline registrations and tests Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
refactor for consistent folders usage for all models Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
automated pipeline registrations and tests Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
feat(model): secondary unet override slot and ideogram4 native loading One UNET override cannot serve dual-transformer arches: ideogram4 conditional/unconditional and wan combined-stage experts need separate files, and previously a single override landed on both experts. - sd_unet_secondary option with per-slot tracking, consumed-state sync, arch-change reset, and incompatible-override fallback - dropdown renders beside the primary, follows it into quicksettings, and is visible only for dual-transformer model types - ideogram4 native single-file spec with a quant-aware fused-qkv converter; such converters run before comfy_quant detection via TransformerSpec.converter_handles_quant - quicksettings render in configured order (sort keyed on the option object and always fell back to alphabetical) - post-load dtype warning skips quantized transformers | 1 个月前 | |
refactor for consistent folders usage for all models Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
attempt to load incomplete model Signed-off-by: Vladimir Mandic <mandic00@live.com> | 2 个月前 | |
automated pipeline registrations and tests Signed-off-by: Vladimir Mandic <mandic00@live.com> | 3 个月前 | |
SDNQ add fused Triton kernels | 1 个月前 |