[Mini-vLLM] Cut dead subsystems: models, entrypoints, backends, utils, tests

- Drop dead model modules, keeping files live imports require
- Remove anthropic, cohere, codegen entrypoints and batch STT modes
- Delete unused attention backends, MoE experts, and NPU/XPU util leaves
- Strip dead third_party trees and ROCm leaves with no CUDA path
- Remove mooncake EC connectors: dependency missing since branch base
- Re-audit verified by full import sweep and Qwen3.5 smoke test

Co-authored-by: Hinata <hinata@idihore.id>
Signed-off-by: MOVZX <movzx@yahoo.com>
This commit is contained in:
2026-10-19 00:20:48 +07:00
co-authored by Hinata
parent 6a95f62e2f
commit 81074e94a3
940 changed files with 11 additions and 380533 deletions
-48
View File
@@ -30,30 +30,6 @@ def vllm_version_matches_substr(substr: str) -> bool:
return substr in vllm_version
def tpu_platform_plugin() -> str | None:
logger.debug("Checking if TPU platform is available.")
# Check for Pathways TPU proxy
if envs.VLLM_TPU_USING_PATHWAYS:
logger.debug("Confirmed TPU platform is available via Pathways proxy.")
return "tpu_inference.platforms.tpu_platform.TpuPlatform"
# Check for libtpu installation
try:
# While it's technically possible to install libtpu on a
# non-TPU machine, this is a very uncommon scenario. Therefore,
# we assume that libtpu is installed only if the machine
# has TPUs.
import libtpu # noqa: F401
logger.debug("Confirmed TPU platform is available.")
return "vllm.platforms.tpu.TpuPlatform"
except Exception as e:
logger.debug("TPU platform is not available because: %s", str(e))
return None
def cuda_platform_plugin() -> str | None:
is_cuda = False
logger.debug("Checking if CUDA platform is available.")
@@ -140,28 +116,6 @@ def rocm_platform_plugin() -> str | None:
return "vllm.platforms.rocm.RocmPlatform" if is_rocm else None
def xpu_platform_plugin() -> str | None:
is_xpu = False
logger.debug("Checking if XPU platform is available.")
try:
import torch
if torch.distributed.is_xccl_available():
dist_backend = "xccl"
from vllm.platforms.xpu import XPUPlatform
XPUPlatform.dist_backend = dist_backend
logger.debug("Confirmed %s backend is available.", XPUPlatform.dist_backend)
if hasattr(torch, "xpu") and torch.xpu.is_available():
is_xpu = True
logger.debug("Confirmed XPU platform is available.")
except Exception as e:
logger.debug("XPU platform is not available because: %s", str(e))
return "vllm.platforms.xpu.XPUPlatform" if is_xpu else None
def _is_amd_zen_cpu() -> bool:
"""Detect AMD CPU with AVX-512 via /proc/cpuinfo."""
if not os.path.exists("/proc/cpuinfo"):
@@ -229,10 +183,8 @@ def cpu_platform_plugin() -> str | None:
builtin_platform_plugins = {
"tpu": tpu_platform_plugin,
"cuda": cuda_platform_plugin,
"rocm": rocm_platform_plugin,
"xpu": xpu_platform_plugin,
"cpu": cpu_platform_plugin,
}