# Copyright (c) 2026 LightSeek Foundation # # Permission is hereby granted, free of charge, to any person obtaining a copy # of this software and associated documentation files (the "Software"), to deal # in the Software without restriction, including without limitation the rights # to use, copy, modify, merge, publish, distribute, sublicense, and/or sell # copies of the Software, and to permit persons to whom the Software is # furnished to do so, subject to the following conditions: # # The above copyright notice and this permission notice shall be included in # all copies or substantial portions of the Software. # # THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR # IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, # FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE # AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER # LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, # OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE # SOFTWARE. from tokenspeed_kernel.profiling import bootstrap_profiling_from_env bootstrap_profiling_from_env() from tokenspeed_kernel.ops.attention import ( GdnCheckpointLayout, GdnChunkPrefillResult, attn_merge_state, dsa_decode, dsa_decode_topk, dsa_plan, dsa_prefill, dsa_prefill_topk, gdn_chunk_prefill, mha_decode_with_kvcache, mha_extend_with_kvcache, mha_plan, mha_prefill, mla_decode_with_kvcache, mla_prefill, ) from tokenspeed_kernel.ops.gemm import mm from tokenspeed_kernel.ops.moe import moe_apply, moe_plan, moe_process_weights from tokenspeed_kernel.ops.quantization import ( quantize_fp8, quantize_fp8_with_scale, quantize_mxfp4, quantize_mxfp8, quantize_nvfp4, ) from tokenspeed_kernel.ops.sampling import argmax from tokenspeed_kernel.ops.transform import hadamard_transform from tokenspeed_kernel.selection import NoKernelFoundError __all__ = [ # exceptions "NoKernelFoundError", # gemm "mm", # attention "mha_plan", "mha_prefill", "mha_extend_with_kvcache", "mha_decode_with_kvcache", "mla_prefill", "mla_decode_with_kvcache", "dsa_prefill", "dsa_decode", "dsa_prefill_topk", "dsa_decode_topk", "dsa_plan", "attn_merge_state", "gdn_chunk_prefill", "GdnCheckpointLayout", "GdnChunkPrefillResult", # moe "moe_apply", "moe_plan", "moe_process_weights", # quantization "quantize_fp8", "quantize_fp8_with_scale", "quantize_mxfp8", "quantize_nvfp4", "quantize_mxfp4", # sampling "argmax", # transform "hadamard_transform", ]