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

[Misc] Merge bitsandbytes_stacked_params_mapping and packed_modules_mapping (#11924)

Browse Source 
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
This commit is contained in:
Jee Jee Li
2025-01-15 07:49:49 +08:00
committed by GitHub
parent 87054a57ab
commit a3a3ee4e6f
24 changed files with 49 additions and 200 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
vllm/model_executor/models/qwen2.py
View File

@@ -418,16 +418,6 @@ class Qwen2ForCausalLM(nn.Module, SupportsLoRA, SupportsPP):
embedding_modules = {}
embedding_padding_modules = []
# BitandBytes specific attributes
bitsandbytes_stacked_params_mapping = {
# shard_name, weight_name, index
"q_proj": ("qkv_proj", 0),
"k_proj": ("qkv_proj", 1),
"v_proj": ("qkv_proj", 2),
"gate_proj": ("gate_up_proj", 0),
"up_proj": ("gate_up_proj", 1),
}
def __init__(self, *, vllm_config: VllmConfig, prefix: str = ""):
super().__init__()
config = vllm_config.model_config.hf_config