Anai-Guo d556544b30 fix(lumina): pass to_compute_mask, not use_cache, to LLaDABlock.attention ...

attention() takes (q, k, v, attention_bias, layer_past, to_compute_mask) and
has no use_cache parameter. Three of the four call sites still pass
use_cache=, which raises TypeError; only LLaDALlamaBlock's non-checkpointed
branch -- the path the shipped block_type=llama config takes -- is correct.

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

..

lumina_dimoo.py

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

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

notes.md

fix(lint): drop unused f-prefix and add trailing newline

2026-05-07 22:21:58 +01:00

pipelines.py

add lumina-dimmo

2026-05-06 09:11:33 +02:00