GitLab

This reverts commit f38bd872.

V0.5.0 dtk24.04.1 awq

See merge request dcutoolkit/deeplearing/vllm!6

# Conflicts:
#	vllm/model_executor/layers/linear.py
#	vllm/model_executor/models/baichuan.py

# Conflicts:
#	csrc/attention/attention_kernels.cu
#	csrc/attention/attention_utils.cuh
#	csrc/layernorm_kernels.cu
#	vllm/model_executor/layers/linear.py
#	vllm/model_executor/models/baichuan.py
#	vllm/model_executor/models/llama.py