From e7da0ec71b9868ed185273a59f1ae7dd851a4685 Mon Sep 17 00:00:00 2001 From: yurekami Date: Fri, 26 Dec 2025 00:23:44 +0900 Subject: [PATCH] fix: correct typos in error messages MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit - "kvcahe" -> "KV cache" (line 337) - "is_fp8_cache" -> "is_fp8_kvcache" for consistency (line 255) 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 --- csrc/pybind.cpp | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/csrc/pybind.cpp b/csrc/pybind.cpp index 13541d40..f2b4819c 100644 --- a/csrc/pybind.cpp +++ b/csrc/pybind.cpp @@ -252,7 +252,7 @@ fwd_kvcache_mla( int bytes_per_token = 512 + 64*2 + (512/128)*4; CHECK_SHAPE(kcache, num_blocks, page_block_size, num_heads_k, bytes_per_token); TORCH_CHECK(num_heads_k == 1, "Currently the number of k heads must be 1 when is_fp8_kvcache is True"); - TORCH_CHECK(kcache.stride(1) == bytes_per_token, "The whole block must be contiguous when is_fp8_cache is True"); + TORCH_CHECK(kcache.stride(1) == bytes_per_token, "The whole block must be contiguous when is_fp8_kvcache is True"); } CHECK_SHAPE(seqlens_k, batch_size); CHECK_SHAPE(block_table, batch_size, max_num_blocks_per_seq); @@ -334,7 +334,7 @@ fwd_kvcache_mla( TORCH_CHECK(q_dtype == torch::kBFloat16, "Sparse FP8 MLA only supports BFloat16 on SM90"); sm90::run_flash_splitkv_mla_fp8_sparse_kernel(params, stream); } else { - TORCH_CHECK(false, "Only FP8 kvcahe is supported for sparse MLA on SM90"); + TORCH_CHECK(false, "Only FP8 KV cache is supported for sparse MLA on SM90"); } } else { if (is_fp8) {