CalamitousFelicitousness 5875cdd29e feat(lora): remap anima 1.0 lora block indices onto depth-expanded checkpoints ...

Anima 2.9B interleaves twelve new blocks among the 28 of Anima 1.0, so a block
index trained against 1.0 names a different block on the expanded model. Every
such key still resolves, since blocks 0 to 27 exist either way, so the
mismatch was silent. The Anima loader now shifts base-depth indices onto the
blocks that carry those weights, keyed by (base depth, expanded depth) and
applied only when the transformer is expanded and the LoRA stays inside the
base depth. Transformer keys move; llm_adapter and text encoder keys keep
their own numbering.

2026-09-05 02:24:04 +01:00

..

anima

feat(lora): remap anima 1.0 lora block indices onto depth-expanded checkpoints

2026-09-05 02:24:04 +01:00

boogu

add boogu

2026-07-04 15:26:36 +02:00

bria

update ruff and lint

2026-07-24 18:20:41 +02:00

chroma

feat(lora): run the full adapter family chain on every native arch

2026-07-14 08:06:45 +01:00

chrono

refactor(pipelines): move per-arch transformer specs into packages

2026-05-31 00:24:33 +01:00

cogview

refactor(pipelines): move per-arch transformer specs into packages

2026-05-31 00:24:33 +01:00

cosmos

update transformers/diffusers and enhanced tests

2026-06-12 15:15:16 +02:00

ernie

feat(lora): run the full adapter family chain on every native arch

2026-07-14 08:06:45 +01:00

f_lite

minor fixes from audit

2026-07-01 11:08:04 +02:00

flex2

PEP 484 pipelines directory

2026-03-25 16:56:41 -07:00

flux

feat(lora): support the lycoris_ prefix in every native adapter arch

2026-07-14 08:06:45 +01:00

flux2_klein

refactor(native-loader): replace required_markers with reactive fallback

2026-06-04 21:20:10 +01:00

glm

refactor(pipelines): move per-arch transformer specs into packages

2026-05-31 00:24:33 +01:00

hidream

full codespell coverage

2026-06-04 12:36:10 +02:00

hunyuandit

refactor(pipelines): move per-arch transformer specs into packages

2026-05-31 00:24:33 +01:00

hyimage

refactor(pipelines): move per-arch transformer specs into packages

2026-05-31 00:24:33 +01:00

ideogram

fix ideogram4

2026-08-24 14:57:58 +02:00

joy

refactor(pipelines): move per-arch transformer specs into packages

2026-05-31 00:24:33 +01:00

kandinsky3

refactor(pipelines): split kandinsky into kandinsky3 and kandinsky5

2026-05-31 00:24:33 +01:00

kandinsky5

refactor(pipelines): split kandinsky into kandinsky3 and kandinsky5

2026-05-31 00:24:33 +01:00

krea2

feat(krea2): publish the packed token layout

2026-08-28 12:40:23 +01:00

lens

feat(pipelines): wire native_spec for Qwen-Image and custom-class arches

2026-05-31 00:24:33 +01:00

longcat

refactor(pipelines): move per-arch transformer specs into packages

2026-05-31 00:24:33 +01:00

lumina_dimmo

fix(lumina): pass to_compute_mask, not use_cache, to LLaDABlock.attention

2026-08-30 18:20:30 -07:00

mageflow

fix(mageflow): decorate the vae entry points for offload hooks

2026-08-14 23:59:35 +01:00

meissonic

minor fixes from audit

2026-07-01 11:08:04 +02:00

minimax

minimax native lora handler

2026-09-01 17:38:49 +02:00

nucleus

refactor(pipelines): move per-arch transformer specs into packages

2026-05-31 00:24:33 +01:00

omnigen2

strict ruff without exclude

2026-05-11 08:36:49 +02:00

ovis

refactor(pipelines): move per-arch transformer specs into packages

2026-05-31 00:24:33 +01:00

pixart

refactor(pipelines): move per-arch transformer specs into packages

2026-05-31 00:24:33 +01:00

prx

refactor(pipelines): move per-arch transformer specs into packages

2026-05-31 00:24:33 +01:00

qwen

update spandrel integation

2026-08-21 09:35:15 +02:00

sefi

batch processing run register after script

2026-08-23 11:34:34 +02:00

segmoe

refactor for consistent folders usage for all models

2026-06-08 08:24:21 +02:00

step1x

feat(pipelines): wire native_spec for Qwen-Image and custom-class arches

2026-05-31 00:24:33 +01:00

ultraflux

full codespell coverage

2026-06-04 12:36:10 +02:00

vibe

full codespell coverage

2026-06-04 12:36:10 +02:00

wan

PEP 484 pipelines directory

2026-03-25 16:56:41 -07:00

xomni

refactor for consistent folders usage for all models

2026-06-08 08:24:21 +02:00

z_image

fix(lora): apply full-diff norm targets and map z-image's renamed modules

2026-07-14 08:06:45 +01:00

zetachroma

ai-agents

2026-04-12 18:54:22 +02:00

generic_map.py

attempt to load incomplete model

2026-06-29 13:39:46 +02:00

generic_shared.py

minimax image and reference

2026-08-13 20:15:32 +02:00

generic_test.py

automated pipeline registrations and tests

2026-06-04 10:18:09 +02:00

generic_text_encoder.py

video loader use generic methods and auth

2026-08-16 09:02:40 +02:00

generic_transformer.py

update nunchaku-lite

2026-08-18 18:00:34 +02:00

generic_util.py

automated pipeline registrations and tests

2026-06-04 10:18:09 +02:00

generic_vae.py

fix: mutable default arguments + wrong cfg var gating eva_clip text checkpoint download

2026-06-12 05:08:03 -04:00

generic.py

attempt to load incomplete model

2026-06-29 13:39:46 +02:00

model_anima.py

feat(model): add anima 2.9b as a base reference model

2026-09-05 01:45:49 +01:00

model_auraflow.py

enhance automated testing

2026-06-03 09:27:34 +02:00

model_boogu.py

add sefi-image model

2026-07-31 15:20:45 +02:00

model_bria.py

automated pipeline registrations and tests

2026-06-04 10:18:09 +02:00

model_chroma.py

enhance automated testing

2026-06-03 09:27:34 +02:00

model_chrono.py

enhance automated testing

2026-06-03 09:27:34 +02:00

model_cogview.py

enhance automated testing

2026-06-03 09:27:34 +02:00

model_cosmos.py

update transformers/diffusers and enhanced tests

2026-06-12 15:15:16 +02:00

model_ernie.py

update nunchaku-lite

2026-08-18 18:00:34 +02:00

model_flex.py

automated pipeline registrations and tests

2026-06-04 10:18:09 +02:00

model_flite.py

automated pipeline registrations and tests

2026-06-04 10:18:09 +02:00

model_flux2_klein.py

add experimental openai interface

2026-07-02 14:36:44 +02:00

model_flux2.py

enhance automated testing

2026-06-03 09:27:34 +02:00

model_flux.py

update nunchaku-lite

2026-08-18 18:00:34 +02:00

model_glm.py

enhance automated testing

2026-06-03 09:27:34 +02:00

model_google.py

Nano Banana lite add

2026-07-01 20:51:06 +00:00

model_hidream.py

pipeline init reordering

2026-08-16 12:45:50 +02:00

model_hunyuandit.py

enhance automated testing

2026-06-03 09:27:34 +02:00

model_hyimage.py

pipeline init reordering

2026-08-16 12:45:50 +02:00

model_ideogram4.py

feat(model): secondary unet override slot and ideogram4 native loading

2026-07-11 06:40:07 +01:00

model_instaflow.py

minor fixes from audit

2026-07-01 11:08:04 +02:00

model_joy.py

add joy-image-edit-plus

2026-07-13 11:26:33 +02:00

model_kandinsky.py

enhance automated testing

2026-06-03 09:27:34 +02:00

model_kolors.py

enhance automated testing

2026-06-03 09:27:34 +02:00

model_krea2.py

fallback unpack-latents, log client auth, update login page

2026-08-19 18:03:39 +02:00

model_lens.py

pipeline init reordering

2026-08-16 12:45:50 +02:00

model_longcat.py

enhance automated testing

2026-06-03 09:27:34 +02:00

model_lumina.py

refactor for consistent folders usage for all models

2026-06-08 08:24:21 +02:00

model_mageflow.py

pipeline init reordering

2026-08-16 12:45:50 +02:00

model_meissonic.py

refactor for consistent folders usage for all models

2026-06-08 08:24:21 +02:00

model_minimax.py

minimax unpack latents

2026-09-01 13:17:03 +02:00

model_nextstep.py

enhance automated testing

2026-06-03 09:27:34 +02:00

model_nucleus.py

enhance automated testing

2026-06-03 09:27:34 +02:00

model_omnigen.py

pipeline init reordering

2026-08-16 12:45:50 +02:00

model_ovis.py

enhance automated testing

2026-06-03 09:27:34 +02:00

model_pixart.py

fixes based on full skills run

2026-06-04 11:10:26 +02:00

model_prx_pixel.py

fix hf telemetry and add prxpixel

2026-07-01 12:50:45 +02:00

model_prx.py

add sefi-image model

2026-07-31 15:20:45 +02:00

model_qwen.py

update nunchaku-lite

2026-08-18 18:00:34 +02:00

model_sana.py

refactor for consistent folders usage for all models

2026-06-08 08:24:21 +02:00

model_sd3.py

enhance automated testing

2026-06-03 09:27:34 +02:00

model_sdxs.py

lint fix

2026-06-04 11:37:44 +02:00

model_sefi.py

add sefi-image model

2026-07-31 15:20:45 +02:00

model_segmoe.py

minor fixes from audit

2026-07-01 11:08:04 +02:00

model_stablecascade.py

refactor for consistent folders usage for all models

2026-06-08 08:24:21 +02:00

model_step1x_edit.py

automated pipeline registrations and tests

2026-06-04 10:18:09 +02:00

model_ultraflux.py

refactor for consistent folders usage for all models

2026-06-08 08:24:21 +02:00

model_vibe.py

automated pipeline registrations and tests

2026-06-04 10:18:09 +02:00

model_wanai.py

feat(model): secondary unet override slot and ideogram4 native loading

2026-07-11 06:40:07 +01:00

model_xomni.py

refactor for consistent folders usage for all models

2026-06-08 08:24:21 +02:00

model_z_image.py

update nunchaku-lite

2026-08-18 18:00:34 +02:00

model_zetachroma.py

automated pipeline registrations and tests

2026-06-04 10:18:09 +02:00

native_transformer.py

feat(model): add anima 2.9b as a base reference model

2026-09-05 01:45:49 +01:00