[Bugfix] Pass drafter quant_config to ParallelLMHead in Eagle3 (#37280)
Signed-off-by: Matthias Gehre <matthias.gehre@amd.com>
This commit is contained in:
@@ -294,6 +294,7 @@ class Eagle3LlamaForCausalLM(LlamaForCausalLM):
|
||||
self.lm_head = ParallelLMHead(
|
||||
self.config.draft_vocab_size,
|
||||
self.config.hidden_size,
|
||||
quant_config=get_draft_quant_config(vllm_config),
|
||||
prefix=maybe_prefix(prefix, "lm_head"),
|
||||
)
|
||||
self.logits_processor = LogitsProcessor(
|
||||
|
||||
Reference in New Issue
Block a user