5862708b32
Source: github.com/wudixzy/competition (1527 files, BI-V100 competition reference)
Imported assets:
- 12 prebuilt CoreX .so extensions (corex-3.2.3-ivcore10):
corex_gdn_{beta_decay,causal_conv,gated_norm,packed_decode,qk_map}.so
corex_moe_{direct_routed,exact_reduce,weight_gather}.so
corex_attn_head_rms_norm.so, corex_paged_kv_gather.so
corex_block_major_kv_transfer.so, corex_fused_paged_prefill.so
- 13 CUDA kernel sources (.cu) for above extensions
- 11 build scripts (build_corex_*.sh)
- install_prebuilt_corex.sh (SHA256-verified .so deployment)
- qwen3_5.py (2615 lines) with FULL corex kernel integration
- 9 vllm vendor override files (block manager, sampler, etc)
- 19 patch scripts (model_runner, xformers, block_major, etc)
- Complete serving layer (serving_chat, protocol, api_server, etc)
- bi100_env.py, bi100_profile.py, gdn_prefix.py, block_major_kv_cache.py
- Dockerfile aligned with reference build chain
- computility-run.yaml with BI100_MOE_COREX_DIRECT_ROUTED=1
Call chain verified:
Dockerfile COPY → patch_ops.sh → install_prebuilt_corex.sh → 12 .so to $VLLM_ROOT
qwen3_5.py imports: from vllm import corex_gdn_* / corex_moe_* / corex_attn_*
27 lines
810 B
Python
27 lines
810 B
Python
import os
|
|
|
|
|
|
def env_bool(name: str, default: bool = False) -> bool:
|
|
raw = os.getenv(name)
|
|
if raw is None:
|
|
return default
|
|
if raw in ("1", "true", "True", "yes", "YES", "on", "ON"):
|
|
return True
|
|
if raw in ("0", "false", "False", "no", "NO", "off", "OFF"):
|
|
return False
|
|
raise RuntimeError(f"{name} must be boolean, got {raw!r}")
|
|
|
|
|
|
def env_int(name: str, default: int, min_value: int, max_value: int) -> int:
|
|
raw = os.getenv(name)
|
|
if raw is None:
|
|
return default
|
|
try:
|
|
value = int(raw)
|
|
except ValueError as exc:
|
|
raise RuntimeError(f"{name} must be int, got {raw!r}") from exc
|
|
if not (min_value <= value <= max_value):
|
|
raise RuntimeError(
|
|
f"{name}={value} outside [{min_value}, {max_value}]")
|
|
return value
|