xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2025-12-25 22:28:44 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Bugfix] Fix GLM rotary_dim issue and support v1 (#16912)

Browse Source 
Signed-off-by: isotr0py <2037008807@qq.com>
This commit is contained in:
Isotr0py 2025-04-21 22:26:34 +08:00 committed by GitHub
parent 7272bfae77
commit 55d6d3fdb8
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 2 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
vllm/model_executor/models/glm.py
View File

@ -3,13 +3,13 @@
from vllm.config import VllmConfig
from vllm.model_executor.models.llama import LlamaForCausalLM
from .interfaces import SupportsV0Only
from .utils import PPMissingLayer
class GlmForCausalLM(LlamaForCausalLM, SupportsV0Only):
class GlmForCausalLM(LlamaForCausalLM):
def __init__(self, *, vllm_config: VllmConfig, prefix: str = ""):
vllm_config.model_config.hf_config.partial_rotary_factor = 0.5
super().__init__(vllm_config=vllm_config, prefix=prefix)
# Hack Llama model to fit HF format GLM implementation
# Attention difference between GLM and Llama:
@ -17,7 +17,6 @@ class GlmForCausalLM(LlamaForCausalLM, SupportsV0Only):
# 2. There is no bias for o_proj in attention
for layer in self.model.layers:
if not isinstance(layer, PPMissingLayer):
layer.self_attn.rotary_emb.rotary_dim //= 2
layer.self_attn.rotary_emb.is_neox_style = False
layer.self_attn.o_proj.bias = None
layer.self_attn.o_proj.skip_bias_add = True