kongo_junandClaude Fable 5
795c3f14eb
- blockswap.py restored byte-identical to the Wan-verified implementation
(a403b51): class BlockSwap, .blocks only - zero regression risk for Wan
- blockswap_ltx.py: class BlockSwapLTX at the fp8-verified engine (b5dc522:
transformer_blocks detection, QuantizedTensor attr-level swap, base.to()
unpin guard); distinct ON_LOAD callback key
- the 155ff95 hardening (mmap materialization / swap-set partial unload /
comfy-pin release) is reverted: it did not fix the open bf16-43GB issue
and regressed the previously passing fp8 smoke (its partial-unload
override also fired in model_unload context and crashed there)
- bundled workflows migrated to BlockSwapLTX; README: two-node table
Verified: fp8 no-LoRA 360x360x49f via BlockSwapLTX 41.6s (parity with
pre-split 40.2s); both nodes registered in /object_info.
Known issue (todo-013, BlockSwapLTX): 43GB bf16 multi-model WF still
poisons the CUDA context after pass-1 - unchanged, see README.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
8 lines
377 B
Python
8 lines
377 B
Python
from .blockswap import NODE_CLASS_MAPPINGS as _WAN_NODES, NODE_DISPLAY_NAME_MAPPINGS as _WAN_NAMES
|
|
from .blockswap_ltx import NODE_CLASS_MAPPINGS as _LTX_NODES, NODE_DISPLAY_NAME_MAPPINGS as _LTX_NAMES
|
|
|
|
NODE_CLASS_MAPPINGS = {**_WAN_NODES, **_LTX_NODES}
|
|
NODE_DISPLAY_NAME_MAPPINGS = {**_WAN_NAMES, **_LTX_NAMES}
|
|
|
|
__all__ = ["NODE_CLASS_MAPPINGS", "NODE_DISPLAY_NAME_MAPPINGS"]
|