Merge tag 'v0.3.3' into vllm-v0.3.2-dtk23.10 and add gfx

a130cf33 · zhuwenwen · a2d181be · 82091b86 · a130cf33 · a130cf33
Commit a130cf33 authored Mar 06, 2024 by zhuwenwen
20 changed files
--- a/vllm/lora/punica.py
+++ b/vllm/lora/punica.py
--- a/vllm/model_executor/__init__.py
+++ b/vllm/model_executor/__init__.py
--- a/vllm/model_executor/guided_decoding.py
+++ b/vllm/model_executor/guided_decoding.py
--- a/vllm/model_executor/guided_logits_processors.py
+++ b/vllm/model_executor/guided_logits_processors.py
--- a/vllm/model_executor/layers/activation.py
+++ b/vllm/model_executor/layers/activation.py
--- a/vllm/model_executor/layers/attention.py
+++ b/vllm/model_executor/layers/attention.py
--- a/vllm/model_executor/layers/fused_moe/__init__.py
+++ b/vllm/model_executor/layers/fused_moe/__init__.py
--- a/vllm/model_executor/layers/fused_moe/configs/E=8,N=3584,device_name=NVIDIA_A100-SXM4-80GB.json
+++ b/vllm/model_executor/layers/fused_moe/configs/E=8,N=3584,device_name=NVIDIA_A100-SXM4-80GB.json
--- a/vllm/model_executor/layers/fused_moe/configs/E=8,N=7168,device_name=NVIDIA_H100_80GB_HBM3.json
+++ b/vllm/model_executor/layers/fused_moe/configs/E=8,N=7168,device_name=NVIDIA_H100_80GB_HBM3.json
--- a/vllm/model_executor/layers/fused_moe/configs/README
+++ b/vllm/model_executor/layers/fused_moe/configs/README
--- a/vllm/model_executor/layers/fused_moe.py
+++ b/vllm/model_executor/layers/fused_moe.py
--- a/vllm/model_executor/layers/linear.py
+++ b/vllm/model_executor/layers/linear.py
--- a/vllm/model_executor/layers/quantization/__init__.py
+++ b/vllm/model_executor/layers/quantization/__init__.py
--- a/vllm/model_executor/layers/quantization/gptq.py
+++ b/vllm/model_executor/layers/quantization/gptq.py
--- a/vllm/model_executor/layers/quantization/marlin.py
+++ b/vllm/model_executor/layers/quantization/marlin.py
--- a/vllm/model_executor/layers/rotary_embedding.py
+++ b/vllm/model_executor/layers/rotary_embedding.py
--- a/vllm/model_executor/layers/sampler.py
+++ b/vllm/model_executor/layers/sampler.py
--- a/vllm/model_executor/layers/triton_kernel/prefix_prefill.py
+++ b/vllm/model_executor/layers/triton_kernel/prefix_prefill.py
--- a/vllm/model_executor/model_loader.py
+++ b/vllm/model_executor/model_loader.py
--- a/vllm/model_executor/models/__init__.py
+++ b/vllm/model_executor/models/__init__.py