[Bugfix] Pass quant_config to DiffusionGemma's ParallelLMHead (#48521)

Signed-off-by: Aaron Kang <aaron.h.kang@icloud.com>
Co-authored-by: Misha Goin <mgoin64@gmail.com>
This commit is contained in:
Aaron Kang
2026-09-24 13:17:02 -07:00
committed by GitHub
co-authored by Misha Goin
parent ab3de6edf2
commit 0908116dd9
@@ -250,6 +250,8 @@ class DiffusionGemmaForConditionalGeneration(
self.lm_head = ParallelLMHead(
num_embeddings=text_config.vocab_size,
embedding_dim=text_config.hidden_size,
quant_config=vllm_config.quant_config,
prefix=maybe_prefix(prefix, "lm_head"),
)
if text_config.tie_word_embeddings: