TensorRT-LLMs/cpp/tensorrt_llm/kernels/contextFusedMultiHeadAttention/cubin
2024-08-13 22:34:33 +08:00
..
fmha_cubin.h Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_bf16_64_32_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_bf16_64_64_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_bf16_128_32_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_bf16_128_64_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_bf16_256_32_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_bf16_256_64_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_bf16_384_32_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_bf16_384_64_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_bf16_512_32_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_bf16_512_64_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_q_paged_kv_160_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_q_paged_kv_160_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_q_paged_kv_160_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_q_paged_kv_192_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_q_paged_kv_192_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_q_paged_kv_192_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_q_paged_kv_256_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_q_paged_kv_256_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_q_paged_kv_256_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_q_paged_kv_256_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_q_paged_kv_256_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_q_paged_kv_256_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_qkv_160_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_qkv_160_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_qkv_160_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_qkv_160_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_qkv_192_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_qkv_192_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_qkv_192_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_qkv_192_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_qkv_256_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_qkv_256_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_qkv_256_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_qkv_256_qk_tanh_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_qkv_256_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_qkv_256_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_qkv_256_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_16_S_qkv_256_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_kv_64_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_kv_64_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_kv_64_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_kv_128_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_kv_128_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_kv_128_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_kv_128_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_kv_128_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_kv_128_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_paged_kv_40_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_paged_kv_40_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_paged_kv_40_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_paged_kv_64_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_paged_kv_64_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_paged_kv_64_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_paged_kv_80_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_paged_kv_80_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_paged_kv_80_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_paged_kv_96_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_paged_kv_96_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_paged_kv_96_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_paged_kv_104_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_paged_kv_104_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_paged_kv_104_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_paged_kv_128_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_paged_kv_128_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_paged_kv_128_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_paged_kv_128_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_paged_kv_128_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_q_paged_kv_128_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_40_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_40_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_40_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_40_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_64_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_64_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_64_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_64_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_80_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_80_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_80_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_80_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_96_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_96_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_96_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_96_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_104_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_104_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_104_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_104_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_128_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_128_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_128_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_128_qk_tanh_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_128_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_128_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_128_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_32_S_qkv_128_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_q_kv_32_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_q_kv_32_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_q_kv_32_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_q_paged_kv_16_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_q_paged_kv_16_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_q_paged_kv_16_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_q_paged_kv_32_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_q_paged_kv_32_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_q_paged_kv_32_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_q_paged_kv_160_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_q_paged_kv_160_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_q_paged_kv_192_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_q_paged_kv_192_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_q_paged_kv_256_alibi_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_q_paged_kv_256_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_q_paged_kv_256_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_q_paged_kv_256_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_qkv_16_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_qkv_16_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_qkv_16_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_qkv_16_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_qkv_32_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_qkv_32_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_qkv_32_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_qkv_32_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_qkv_160_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_qkv_160_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_qkv_192_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_qkv_192_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_qkv_256_alibi_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_qkv_256_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_qkv_256_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_64_S_qkv_256_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_kv_128_alibi_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_kv_128_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_kv_128_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_kv_128_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_kv_128_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_kv_128_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_kv_128_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_kv_128_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_kv_128_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_kv_128_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_80_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_80_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_80_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_80_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_80_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_96_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_96_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_96_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_96_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_96_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_104_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_104_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_104_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_104_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_104_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_128_alibi_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_128_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_128_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_128_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_128_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_128_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_128_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_128_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_128_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_128_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_160_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_160_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_160_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_192_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_192_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_192_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_256_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_256_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_256_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_256_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_256_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_q_paged_kv_256_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_80_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_80_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_80_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_80_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_80_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_80_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_96_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_96_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_96_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_96_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_96_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_96_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_104_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_104_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_104_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_104_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_104_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_104_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_128_alibi_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_128_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_128_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_128_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_128_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_128_qk_tanh_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_128_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_128_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_128_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_128_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_128_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_128_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_160_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_160_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_160_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_160_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_192_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_192_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_192_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_192_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_256_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_256_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_256_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_256_qk_tanh_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_256_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_256_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_256_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_128_S_qkv_256_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_256_S_q_kv_32_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_256_S_q_kv_32_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_256_S_q_kv_64_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_256_S_q_kv_64_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_256_S_q_paged_kv_32_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_256_S_q_paged_kv_32_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_256_S_q_paged_kv_40_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_256_S_q_paged_kv_40_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_256_S_q_paged_kv_64_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_256_S_q_paged_kv_64_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_256_S_qkv_32_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_256_S_qkv_32_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_256_S_qkv_40_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_256_S_qkv_40_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_256_S_qkv_64_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_64_256_S_qkv_64_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_q_kv_32_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_q_kv_32_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_q_kv_32_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_q_kv_64_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_q_kv_64_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_q_kv_64_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_q_paged_kv_16_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_q_paged_kv_16_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_q_paged_kv_16_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_q_paged_kv_32_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_q_paged_kv_32_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_q_paged_kv_32_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_q_paged_kv_40_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_q_paged_kv_40_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_q_paged_kv_40_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_q_paged_kv_64_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_q_paged_kv_64_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_q_paged_kv_64_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_qkv_16_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_qkv_16_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_qkv_16_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_qkv_16_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_qkv_32_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_qkv_32_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_qkv_32_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_qkv_32_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_qkv_40_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_qkv_40_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_qkv_40_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_qkv_40_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_qkv_64_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_qkv_64_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_qkv_64_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_bf16_128_128_S_qkv_64_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_128_S_q_paged_kv_160_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_128_S_q_paged_kv_160_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_128_S_q_paged_kv_192_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_128_S_q_paged_kv_192_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_128_S_q_paged_kv_256_alibi_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_128_S_q_paged_kv_256_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_128_S_q_paged_kv_256_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_128_S_q_paged_kv_256_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_128_S_qkv_160_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_128_S_qkv_160_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_128_S_qkv_192_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_128_S_qkv_192_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_128_S_qkv_256_alibi_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_128_S_qkv_256_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_128_S_qkv_256_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_128_S_qkv_256_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_kv_32_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_kv_32_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_kv_64_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_kv_64_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_kv_128_alibi_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_kv_128_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_kv_128_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_kv_128_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_paged_kv_32_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_paged_kv_32_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_paged_kv_40_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_paged_kv_40_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_paged_kv_64_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_paged_kv_64_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_paged_kv_80_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_paged_kv_80_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_paged_kv_96_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_paged_kv_96_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_paged_kv_104_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_paged_kv_104_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_paged_kv_128_alibi_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_paged_kv_128_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_paged_kv_128_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_q_paged_kv_128_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_qkv_32_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_qkv_32_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_qkv_40_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_qkv_40_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_qkv_64_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_qkv_64_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_qkv_80_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_qkv_80_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_qkv_96_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_qkv_96_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_qkv_104_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_qkv_104_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_qkv_128_alibi_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_qkv_128_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_qkv_128_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_64_256_S_qkv_128_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_e4m3_fp32_64_32_S_q_kv_128_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_e4m3_fp32_64_32_S_q_paged_kv_80_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_e4m3_fp32_64_32_S_q_paged_kv_96_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_e4m3_fp32_64_32_S_q_paged_kv_104_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_e4m3_fp32_64_32_S_q_paged_kv_128_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_e4m3_fp32_64_32_S_q_paged_kv_160_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_e4m3_fp32_64_32_S_q_paged_kv_192_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_e4m3_fp32_64_32_S_q_paged_kv_256_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_e4m3_fp32_64_32_S_qkv_80_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_e4m3_fp32_64_32_S_qkv_96_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_e4m3_fp32_64_32_S_qkv_104_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_e4m3_fp32_64_32_S_qkv_128_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_e4m3_fp32_64_32_S_qkv_160_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_e4m3_fp32_64_32_S_qkv_192_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_e4m3_fp32_64_32_S_qkv_256_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_e4m3_fp32_128_128_S_q_kv_32_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_e4m3_fp32_128_128_S_q_kv_64_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_e4m3_fp32_128_128_S_q_paged_kv_32_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_e4m3_fp32_128_128_S_q_paged_kv_40_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_e4m3_fp32_128_128_S_q_paged_kv_64_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_e4m3_fp32_128_128_S_qkv_32_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_e4m3_fp32_128_128_S_qkv_40_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_e4m3_fp32_128_128_S_qkv_64_sm89.cubin.cpp Update TensorRT-LLM (#2110) 2024-08-13 22:34:33 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_kv_128_qk_tanh_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_kv_128_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_paged_kv_80_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_paged_kv_96_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_paged_kv_104_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_paged_kv_128_qk_tanh_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_paged_kv_128_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_paged_kv_160_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_paged_kv_160_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_paged_kv_160_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_paged_kv_160_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_paged_kv_192_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_paged_kv_192_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_paged_kv_192_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_paged_kv_192_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_paged_kv_256_qk_tanh_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_paged_kv_256_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_paged_kv_256_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_paged_kv_256_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_paged_kv_256_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_paged_kv_256_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_paged_kv_256_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_q_paged_kv_256_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_80_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_96_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_104_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_128_qk_tanh_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_128_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_160_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_160_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_160_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_160_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_160_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_192_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_192_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_192_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_192_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_192_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_256_qk_tanh_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_256_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_256_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_256_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_256_qk_tanh_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_256_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_256_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_256_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_256_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_16_S_qkv_256_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_kv_64_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_kv_64_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_kv_64_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_kv_64_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_kv_128_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_kv_128_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_kv_128_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_kv_128_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_kv_128_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_kv_128_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_40_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_40_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_40_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_40_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_64_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_64_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_64_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_64_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_80_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_80_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_80_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_96_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_96_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_96_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_104_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_104_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_104_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_128_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_128_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_128_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_128_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_128_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_q_paged_kv_128_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_40_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_40_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_40_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_40_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_40_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_64_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_64_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_64_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_64_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_64_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_80_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_80_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_80_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_80_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_96_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_96_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_96_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_96_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_104_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_104_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_104_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_104_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_128_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_128_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_128_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_128_qk_tanh_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_128_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_128_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_128_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_32_S_qkv_128_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_q_kv_32_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_q_kv_32_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_q_kv_32_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_q_kv_32_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_q_paged_kv_16_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_q_paged_kv_16_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_q_paged_kv_16_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_q_paged_kv_32_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_q_paged_kv_32_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_q_paged_kv_32_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_q_paged_kv_32_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_q_paged_kv_160_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_q_paged_kv_160_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_q_paged_kv_192_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_q_paged_kv_192_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_q_paged_kv_256_alibi_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_q_paged_kv_256_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_q_paged_kv_256_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_q_paged_kv_256_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_qkv_16_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_qkv_16_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_qkv_16_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_qkv_16_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_qkv_32_sm70.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_qkv_32_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_qkv_32_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_qkv_32_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_qkv_32_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_qkv_160_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_qkv_160_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_qkv_192_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_qkv_192_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_qkv_256_alibi_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_qkv_256_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_qkv_256_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_64_S_qkv_256_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_kv_128_alibi_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_kv_128_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_kv_128_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_kv_128_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_kv_128_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_kv_128_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_kv_128_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_kv_128_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_kv_128_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_kv_128_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_80_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_80_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_80_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_80_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_80_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_96_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_96_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_96_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_96_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_96_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_104_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_104_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_104_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_104_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_104_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_128_alibi_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_128_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_128_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_128_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_128_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_128_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_128_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_128_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_128_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_128_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_160_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_160_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_160_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_192_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_192_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_192_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_256_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_256_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_256_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_256_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_256_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_q_paged_kv_256_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_80_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_80_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_80_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_80_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_80_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_80_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_96_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_96_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_96_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_96_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_96_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_96_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_104_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_104_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_104_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_104_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_104_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_104_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_128_alibi_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_128_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_128_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_128_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_128_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_128_qk_tanh_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_128_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_128_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_128_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_128_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_128_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_128_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_160_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_160_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_160_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_160_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_192_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_192_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_192_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_192_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_256_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_256_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_256_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_256_qk_tanh_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_256_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_256_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_256_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_128_S_qkv_256_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_256_S_q_kv_32_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_256_S_q_kv_32_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_256_S_q_kv_64_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_256_S_q_kv_64_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_256_S_q_paged_kv_32_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_256_S_q_paged_kv_32_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_256_S_q_paged_kv_40_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_256_S_q_paged_kv_40_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_256_S_q_paged_kv_64_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_256_S_q_paged_kv_64_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_256_S_qkv_32_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_256_S_qkv_32_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_256_S_qkv_40_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_256_S_qkv_40_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_256_S_qkv_64_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_64_256_S_qkv_64_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_q_kv_32_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_q_kv_32_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_q_kv_32_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_q_kv_64_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_q_kv_64_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_q_kv_64_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_q_paged_kv_16_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_q_paged_kv_16_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_q_paged_kv_16_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_q_paged_kv_32_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_q_paged_kv_32_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_q_paged_kv_32_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_q_paged_kv_40_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_q_paged_kv_40_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_q_paged_kv_40_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_q_paged_kv_64_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_q_paged_kv_64_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_q_paged_kv_64_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_qkv_16_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_qkv_16_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_qkv_16_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_qkv_16_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_qkv_32_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_qkv_32_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_qkv_32_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_qkv_32_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_qkv_40_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_qkv_40_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_qkv_40_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_qkv_40_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_qkv_64_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_qkv_64_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_qkv_64_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_128_128_S_qkv_64_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_q_paged_kv_160_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_q_paged_kv_160_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_q_paged_kv_160_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_q_paged_kv_192_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_q_paged_kv_192_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_q_paged_kv_192_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_q_paged_kv_256_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_q_paged_kv_256_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_q_paged_kv_256_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_q_paged_kv_256_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_q_paged_kv_256_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_q_paged_kv_256_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_qkv_160_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_qkv_160_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_qkv_160_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_qkv_160_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_qkv_192_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_qkv_192_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_qkv_192_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_qkv_192_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_qkv_256_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_qkv_256_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_qkv_256_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_qkv_256_qk_tanh_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_qkv_256_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_qkv_256_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_qkv_256_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_16_S_qkv_256_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_kv_64_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_kv_64_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_kv_64_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_kv_128_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_kv_128_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_kv_128_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_kv_128_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_kv_128_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_kv_128_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_paged_kv_40_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_paged_kv_40_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_paged_kv_40_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_paged_kv_64_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_paged_kv_64_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_paged_kv_64_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_paged_kv_80_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_paged_kv_80_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_paged_kv_80_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_paged_kv_96_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_paged_kv_96_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_paged_kv_96_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_paged_kv_104_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_paged_kv_104_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_paged_kv_104_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_paged_kv_128_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_paged_kv_128_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_paged_kv_128_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_paged_kv_128_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_paged_kv_128_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_q_paged_kv_128_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_40_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_40_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_40_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_40_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_64_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_64_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_64_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_64_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_80_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_80_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_80_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_80_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_96_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_96_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_96_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_96_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_104_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_104_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_104_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_104_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_128_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_128_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_128_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_128_qk_tanh_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_128_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_128_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_128_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_32_S_qkv_128_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_q_kv_32_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_q_kv_32_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_q_kv_32_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_q_paged_kv_16_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_q_paged_kv_16_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_q_paged_kv_16_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_q_paged_kv_32_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_q_paged_kv_32_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_q_paged_kv_32_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_q_paged_kv_160_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_q_paged_kv_160_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_q_paged_kv_192_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_q_paged_kv_192_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_q_paged_kv_256_alibi_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_q_paged_kv_256_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_q_paged_kv_256_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_q_paged_kv_256_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_qkv_16_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_qkv_16_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_qkv_16_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_qkv_16_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_qkv_32_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_qkv_32_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_qkv_32_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_qkv_32_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_qkv_160_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_qkv_160_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_qkv_192_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_qkv_192_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_qkv_256_alibi_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_qkv_256_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_qkv_256_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_64_S_qkv_256_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_kv_128_alibi_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_kv_128_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_kv_128_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_kv_128_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_kv_128_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_kv_128_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_kv_128_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_kv_128_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_kv_128_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_kv_128_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_80_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_80_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_80_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_80_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_80_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_96_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_96_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_96_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_96_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_96_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_104_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_104_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_104_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_104_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_104_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_128_alibi_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_128_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_128_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_128_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_128_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_128_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_128_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_128_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_128_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_128_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_160_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_160_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_160_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_192_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_192_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_192_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_256_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_256_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_256_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_256_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_256_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_q_paged_kv_256_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_80_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_80_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_80_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_80_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_80_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_80_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_96_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_96_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_96_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_96_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_96_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_96_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_104_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_104_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_104_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_104_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_104_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_104_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_128_alibi_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_128_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_128_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_128_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_128_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_128_qk_tanh_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_128_qk_tanh_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_128_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_128_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_128_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_128_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_128_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_160_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_160_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_160_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_160_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_192_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_192_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_192_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_192_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_256_qk_tanh_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_256_qk_tanh_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_256_qk_tanh_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_256_qk_tanh_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_256_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_256_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_256_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_128_S_qkv_256_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_256_S_q_kv_32_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_256_S_q_kv_32_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_256_S_q_kv_64_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_256_S_q_kv_64_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_256_S_q_paged_kv_32_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_256_S_q_paged_kv_32_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_256_S_q_paged_kv_40_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_256_S_q_paged_kv_40_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_256_S_q_paged_kv_64_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_256_S_q_paged_kv_64_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_256_S_qkv_32_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_256_S_qkv_32_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_256_S_qkv_40_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_256_S_qkv_40_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_256_S_qkv_64_alibi_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_64_256_S_qkv_64_tma_ws_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_q_kv_32_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_q_kv_32_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_q_kv_32_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_q_kv_64_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_q_kv_64_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_q_kv_64_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_q_paged_kv_16_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_q_paged_kv_16_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_q_paged_kv_16_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_q_paged_kv_32_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_q_paged_kv_32_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_q_paged_kv_32_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_q_paged_kv_40_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_q_paged_kv_40_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_q_paged_kv_40_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_q_paged_kv_64_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_q_paged_kv_64_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_q_paged_kv_64_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_qkv_16_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_qkv_16_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_qkv_16_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_qkv_16_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_qkv_32_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_qkv_32_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_qkv_32_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_qkv_32_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_qkv_40_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_qkv_40_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_qkv_40_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_qkv_40_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_qkv_64_sm80.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_qkv_64_sm86.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_qkv_64_sm89.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_flash_attention_fp16_fp32_128_128_S_qkv_64_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_fp16_64_32_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_fp16_64_64_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_fp16_128_32_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_fp16_128_64_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_fp16_256_32_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_fp16_256_64_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_fp16_384_32_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_fp16_384_64_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_fp16_512_32_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_fp16_512_64_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_fp16_fp32_64_32_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_fp16_fp32_64_64_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_fp16_fp32_128_32_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_fp16_fp32_128_64_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_fp16_fp32_256_32_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_fp16_fp32_256_64_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_fp16_fp32_384_32_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_fp16_fp32_384_64_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_fp16_fp32_512_32_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00
fmha_v2_fp16_fp32_512_64_ldgsts_sm90.cubin.cpp Update TensorRT-LLM (#2094) 2024-08-07 16:44:43 +08:00