[CI] update typos config for CI pre-commit and fix some spells (#20919)

Signed-off-by: Peter Pan <Peter.Pan@daocloud.io>

Author: panpan0000

.pre-commit-config.yaml

@@ -21,7 +21,7 @@ repos:
   - id: ruff-format
     files: ^(.buildkite|benchmarks|examples)/.*
 - repo: https://github.com/crate-ci/typos
-  rev: v1.32.0
+  rev: v1.34.0
   hooks:
   - id: typos
 - repo: https://github.com/PyCQA/isort

csrc/cpu/sgl-kernels/common.h

@@ -58,7 +58,7 @@ namespace {
 
 #define CHECK_CONTIGUOUS(x) TORCH_CHECK(x.is_contiguous(), #x " must be contiguous")
 #define CHECK_LAST_DIM_CONTIGUOUS(x) \
-  TORCH_CHECK(x.strides()[x.strides().size() - 1] == 1, #x "must be contiguous at last dimention")
+  TORCH_CHECK(x.strides()[x.strides().size() - 1] == 1, #x "must be contiguous at last dimension")
 
 #define CHECK_INPUT(x) \
   CHECK_CPU(x);        \

csrc/cpu/sgl-kernels/gemm.h

@@ -126,7 +126,7 @@ void fused_experts_int4_w4a16_kernel_impl(
     int64_t topk,
     int64_t num_tokens_post_pad);
 
-// shared expert implememntation for int8 w8a8
+// shared expert implementation for int8 w8a8
 template <typename scalar_t>
 void shared_expert_int8_kernel_impl(
     scalar_t* __restrict__ output,

csrc/cpu/sgl-kernels/gemm_int8.cpp

@@ -41,7 +41,7 @@ struct tinygemm_kernel_nn<at::BFloat16, has_bias, BLOCK_M, BLOCK_N> {
     __m512  vd0;
     __m512  vd1[COLS];
 
-    // oops! 4x4 spills but luckly we use 4x2
+    // oops! 4x4 spills but luckily we use 4x2
     __m512 vbias[COLS];
 
     // [NOTE]: s8s8 igemm compensation in avx512-vnni

csrc/cpu/sgl-kernels/vec.h

@@ -37,7 +37,7 @@ inline Vectorized<at::BFloat16> convert_from_float_ext<at::BFloat16>(const Vecto
 #define CVT_FP16_TO_FP32(a) \
     _mm512_cvtps_ph(a, (_MM_FROUND_TO_NEAREST_INT | _MM_FROUND_NO_EXC))
 
-// this doesn't hanel NaN.
+// this doesn't handle NaN.
 inline __m512bh cvt_e4m3_bf16_intrinsic_no_nan(__m256i fp8_vec) {
   const __m512i x = _mm512_cvtepu8_epi16(fp8_vec);
 

docker/Dockerfile

@@ -63,7 +63,7 @@ ARG PYTORCH_CUDA_NIGHTLY_INDEX_BASE_URL=https://download.pytorch.org/whl/nightly
 ARG PIP_KEYRING_PROVIDER=disabled
 ARG UV_KEYRING_PROVIDER=${PIP_KEYRING_PROVIDER}
 
-# Flag enables build-in KV-connector dependency libs into docker images
+# Flag enables built-in KV-connector dependency libs into docker images
 ARG INSTALL_KV_CONNECTORS=false
 
 #################### BASE BUILD IMAGE ####################

docs/usage/v1_guide.md

@@ -106,7 +106,7 @@ to enable simultaneous generation and embedding using the same engine instance i
 
 Models using selective state-space mechanisms instead of standard transformer attention are partially supported.
 Models that use Mamba-2 layers (e.g., `Mamba2ForCausalLM`) are supported, but models that use older Mamba-1 layers
-(e.g., `MambaForCausalLM`, `JambaForCausalLM`) are not yet suported. Please note that these models currently require
+(e.g., `MambaForCausalLM`, `JambaForCausalLM`) are not yet supported. Please note that these models currently require
 enforcing eager mode and disabling prefix caching in V1.
 
 Models that combine Mamba-2 layers with standard attention layers are also supported (e.g., `BambaForCausalLM`,

pyproject.toml

@@ -174,3 +174,186 @@ respect-ignore-files = true
 
 [tool.ty.environment]
 python = "./.venv"
+
+[tool.typos.files]
+# these files may be written in non english words
+extend-exclude = ["tests/models/fixtures/*", "tests/prompts/*",
+    "benchmarks/sonnet.txt", "tests/lora/data/*", "build/*",
+    "vllm/third_party/*"]
+ignore-hidden = true
+ignore-files = true
+ignore-dot = true
+ignore-vcs = true
+ignore-global = true
+ignore-parent = true
+
+[tool.typos.default]
+binary = false
+check-filename = false
+check-file = true
+unicode = true
+ignore-hex = true
+identifier-leading-digits = false
+locale = "en"
+extend-ignore-identifiers-re = ["NVML_*", ".*Unc.*", ".*_thw",
+    ".*UE8M0.*", ".*[UE4M3|ue4m3].*", ".*eles.*",
+     ".*[Tt]h[rR].*"]
+extend-ignore-words-re = []
+extend-ignore-re = []
+
+[tool.typos.default.extend-identifiers]
+bbc5b7ede = "bbc5b7ede"
+womens_doubles = "womens_doubles"
+v_2nd = "v_2nd"
+# splitted_input = "splitted_input"
+NOOPs = "NOOPs"
+typ = "typ"
+nin_shortcut = "nin_shortcut"
+UperNetDecoder = "UperNetDecoder"
+subtile = "subtile"
+cudaDevAttrMaxSharedMemoryPerBlockOptin = "cudaDevAttrMaxSharedMemoryPerBlockOptin"
+SFOuput = "SFOuput"
+# huggingface transformers repo uses these words
+depthwise_seperable_out_channel = "depthwise_seperable_out_channel"
+DepthWiseSeperableConv1d = "DepthWiseSeperableConv1d"
+depthwise_seperable_CNN = "depthwise_seperable_CNN"
+
+[tool.typos.default.extend-words]
+iy = "iy"
+tendencias = "tendencias"
+# intel cpu features
+tme = "tme"
+dout = "dout"
+Pn = "Pn"
+arange = "arange"
+
+[tool.typos.type.py]
+extend-glob = []
+extend-ignore-identifiers-re = []
+extend-ignore-words-re = []
+extend-ignore-re = []
+
+[tool.typos.type.py.extend-identifiers]
+arange = "arange"
+NDArray = "NDArray"
+EOFError = "EOFError"
+fo = "fo"
+ba = "ba"
+
+[tool.typos.type.py.extend-words]
+
+[tool.typos.type.cpp]
+extend-glob = ["*.cu"]
+extend-ignore-identifiers-re = []
+extend-ignore-words-re = []
+extend-ignore-re = []
+
+[tool.typos.type.cpp.extend-identifiers]
+countr_one = "countr_one"
+k_ot = "k_ot"
+ot = "ot"
+
+[tool.typos.type.cpp.extend-words]
+
+[tool.typos.type.rust]
+extend-glob = []
+extend-ignore-identifiers-re = []
+extend-ignore-words-re = []
+extend-ignore-re = []
+
+[tool.typos.type.rust.extend-identifiers]
+flate2 = "flate2"
+
+[tool.typos.type.rust.extend-words]
+ser = "ser"
+
+[tool.typos.type.lock]
+extend-glob = []
+check-file = false
+extend-ignore-identifiers-re = []
+extend-ignore-words-re = []
+extend-ignore-re = []
+
+[tool.typos.type.lock.extend-identifiers]
+
+[tool.typos.type.lock.extend-words]
+
+[tool.typos.type.jl]
+extend-glob = []
+extend-ignore-identifiers-re = []
+extend-ignore-words-re = []
+extend-ignore-re = []
+
+[tool.typos.type.jl.extend-identifiers]
+
+[tool.typos.type.jl.extend-words]
+modul = "modul"
+egals = "egals"
+usig = "usig"
+egal = "egal"
+
+[tool.typos.type.go]
+extend-glob = []
+extend-ignore-identifiers-re = []
+extend-ignore-words-re = []
+extend-ignore-re = []
+
+[tool.typos.type.go.extend-identifiers]
+flate = "flate"
+
+[tool.typos.type.go.extend-words]
+
+[tool.typos.type.css]
+extend-glob = []
+extend-ignore-identifiers-re = []
+extend-ignore-words-re = []
+extend-ignore-re = []
+
+[tool.typos.type.css.extend-identifiers]
+nd = "nd"
+
+[tool.typos.type.css.extend-words]
+
+[tool.typos.type.man]
+extend-glob = []
+extend-ignore-identifiers-re = []
+extend-ignore-words-re = []
+extend-ignore-re = []
+
+[tool.typos.type.man.extend-identifiers]
+Nd = "Nd"
+
+[tool.typos.type.man.extend-words]
+
+[tool.typos.type.cert]
+extend-glob = []
+check-file = false
+extend-ignore-identifiers-re = []
+extend-ignore-words-re = []
+extend-ignore-re = []
+
+[tool.typos.type.cert.extend-identifiers]
+
+[tool.typos.type.cert.extend-words]
+
+[tool.typos.type.sh]
+extend-glob = []
+extend-ignore-identifiers-re = []
+extend-ignore-words-re = []
+extend-ignore-re = []
+
+[tool.typos.type.sh.extend-identifiers]
+ot = "ot"
+
+[tool.typos.type.sh.extend-words]
+
+[tool.typos.type.vimscript]
+extend-glob = []
+extend-ignore-identifiers-re = []
+extend-ignore-words-re = []
+extend-ignore-re = []
+
+[tool.typos.type.vimscript.extend-identifiers]
+windo = "windo"
+
+[tool.typos.type.vimscript.extend-words]

tests/kernels/moe/modular_kernel_tools/common.py

@@ -416,7 +416,7 @@ def make_hidden_states(
         # We dequant and use that as hidden_states so the tests are stable.
         # quantizing and dequantizing yield slightly different results
         # depending on the hardware. Here we, quantize and dequantize
-        # first - so further quantize and dequantize will yeild the same
+        # first - so further quantize and dequantize will yield the same
         # values.
         if config.is_per_tensor_act_quant:
             a_q, a_scales = ops.scaled_fp8_quant(

tests/kernels/moe/test_deepgemm.py

@@ -95,7 +95,7 @@ def run_single_case(m, n, k, topk, num_experts, block_size):
     topk_weights, topk_ids = torch.topk(router_logits, k=topk, dim=-1)
     topk_weights = torch.nn.functional.softmax(topk_weights, dim=-1)
 
-    # triton referrence
+    # triton reference
     out_triton = fused_experts(
         hidden_states=tokens_bf16,
         w1=w1,