biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[LoRA][1/N]Remove LoRA extra vocab (#28382)

Browse Source 
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
This commit is contained in:
Jee Jee Li
2025-11-12 03:06:21 +08:00
committed by GitHub
parent 8c32c6e4b4
commit 9d1c474704
65 changed files with 197 additions and 754 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
vllm/model_executor/models/llama_eagle3.py
View File

@@ -15,7 +15,6 @@ from vllm.model_executor.layers.linear import QKVParallelLinear
from vllm.model_executor.layers.logits_processor import LogitsProcessor
from vllm.model_executor.layers.quantization.base_config import QuantizationConfig
from vllm.model_executor.layers.vocab_parallel_embedding import (
DEFAULT_VOCAB_PADDING_SIZE,
ParallelLMHead,
VocabParallelEmbedding,
)
@@ -252,8 +251,6 @@ class Eagle3LlamaForCausalLM(LlamaForCausalLM):
self.lm_head = ParallelLMHead(
self.config.draft_vocab_size,
self.config.hidden_size,
org_num_embeddings=self.config.draft_vocab_size,
padding_size=(DEFAULT_VOCAB_PADDING_SIZE),
prefix=maybe_prefix(prefix, "lm_head"),
)
self.logits_processor = LogitsProcessor(