source stringclasses 13
values | gpu stringclasses 4
values | model stringclasses 13
values | kernel_family stringclasses 7
values | kernel_name stringclasses 328
values | dtype stringclasses 2
values | held_out bool 2
classes | M float64 1 16.4k ⌀ | N float64 8 248k ⌀ | K float64 2.05k 29.6k ⌀ | bs float64 1 1 ⌀ | seq float64 1 16.4k ⌀ | n_heads float64 8 64 ⌀ | head_dim float64 64 256 ⌀ | kv_heads float64 1 64 ⌀ | numel float64 64 531M ⌀ | op_type stringclasses 177
values | gpu_time_duration_ms float64 0 703 | launch_block_size float64 16 512 | launch_grid_size float64 1 1.04M | dram_bytes_sum float64 1.27 94.3B ⌀ | launch_registers_per_thread float64 16 255 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | splitk_reduce | void cublasLt::splitKreduce_kernel<32, 16, int, float, __nv_bfloat16, float, __nv_bfloat16, 0, __nv_bfloat16, __nv_bfloat16, __nv_bfloat16, 1, 0, 0>(cublasLt::cublasSplitKParams<T6>, const T4 *, const T10 *, T9 *, T5 *, const T6 *, const T6 *, const T11 *, const T4 *, T11 *, void *, long, T6 *, int *, T6 *, T6 *, const... | bf16 | false | 128 | 8 | 4,096 | null | null | null | null | null | null | null | 0.00582 | 512 | 8 | 139,010 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | splitk_reduce | void cublasLt::splitKreduce_kernel<32, 16, int, __nv_bfloat16, __nv_bfloat16, float, __nv_bfloat16, 0, __nv_bfloat16, __nv_bfloat16, __nv_bfloat16, 1, 0, 0>(cublasLt::cublasSplitKParams<T6>, const T4 *, const T10 *, T9 *, T5 *, const T6 *, const T6 *, const T11 *, const T4 *, T11 *, void *, long, T6 *, int *, T6 *, T6 ... | bf16 | false | 9 | 4,096 | 14,336 | null | null | null | null | null | null | null | 0.00659 | 512 | 448 | 438,530 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | splitk_reduce | void cublasLt::splitKreduce_kernel<32, 16, int, __nv_bfloat16, __nv_bfloat16, float, __nv_bfloat16, 0, __nv_bfloat16, __nv_bfloat16, __nv_bfloat16, 1, 0, 0>(cublasLt::cublasSplitKParams<T6>, const T4 *, const T10 *, T9 *, T5 *, const T6 *, const T6 *, const T11 *, const T4 *, T11 *, void *, long, T6 *, int *, T6 *, T6 ... | bf16 | false | 9 | 4,096 | 14,336 | null | null | null | null | null | null | null | 0.00662 | 512 | 448 | 438,530 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | splitk_reduce | void cublasLt::splitKreduce_kernel<32, 16, int, __nv_bfloat16, __nv_bfloat16, float, __nv_bfloat16, 0, __nv_bfloat16, __nv_bfloat16, __nv_bfloat16, 1, 0, 0>(cublasLt::cublasSplitKParams<T6>, const T4 *, const T10 *, T9 *, T5 *, const T6 *, const T6 *, const T11 *, const T4 *, T11 *, void *, long, T6 *, int *, T6 *, T6 ... | bf16 | false | 54 | 4,096 | 14,336 | null | null | null | null | null | null | null | 0.00746 | 512 | 640 | 1,700,000 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | splitk_reduce | void cublasLt::splitKreduce_kernel<32, 16, int, float, __nv_bfloat16, float, __nv_bfloat16, 0, __nv_bfloat16, __nv_bfloat16, __nv_bfloat16, 1, 0, 0>(cublasLt::cublasSplitKParams<T6>, const T4 *, const T10 *, T9 *, T5 *, const T6 *, const T6 *, const T11 *, const T4 *, T11 *, void *, long, T6 *, int *, T6 *, T6 *, const... | bf16 | false | 27 | 4,096 | 14,336 | null | null | null | null | null | null | null | 0.00592 | 512 | 384 | 1,680,000 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | splitk_reduce | void cublasLt::splitKreduce_kernel<32, 16, int, __nv_bfloat16, __nv_bfloat16, float, __nv_bfloat16, 0, __nv_bfloat16, __nv_bfloat16, __nv_bfloat16, 1, 0, 0>(cublasLt::cublasSplitKParams<T6>, const T4 *, const T10 *, T9 *, T5 *, const T6 *, const T6 *, const T11 *, const T4 *, T11 *, void *, long, T6 *, int *, T6 *, T6 ... | bf16 | false | 18 | 4,096 | 14,336 | null | null | null | null | null | null | null | 0.00675 | 512 | 448 | 868,610 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | splitk_reduce | void cublasLt::splitKreduce_kernel<32, 16, int, __nv_bfloat16, __nv_bfloat16, float, __nv_bfloat16, 0, __nv_bfloat16, __nv_bfloat16, __nv_bfloat16, 1, 0, 0>(cublasLt::cublasSplitKParams<T6>, const T4 *, const T10 *, T9 *, T5 *, const T6 *, const T6 *, const T11 *, const T4 *, T11 *, void *, long, T6 *, int *, T6 *, T6 ... | bf16 | false | 18 | 4,096 | 14,336 | null | null | null | null | null | null | null | 0.00685 | 512 | 448 | 868,610 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | splitk_reduce | void cublasLt::splitKreduce_kernel<32, 16, int, __nv_bfloat16, __nv_bfloat16, float, __nv_bfloat16, 0, __nv_bfloat16, __nv_bfloat16, __nv_bfloat16, 1, 0, 0>(cublasLt::cublasSplitKParams<T6>, const T4 *, const T10 *, T9 *, T5 *, const T6 *, const T6 *, const T11 *, const T4 *, T11 *, void *, long, T6 *, int *, T6 *, T6 ... | bf16 | false | 9 | 4,096 | 14,336 | null | null | null | null | null | null | null | 0.00646 | 512 | 448 | 438,530 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | splitk_reduce | void cublasLt::splitKreduce_kernel<32, 16, int, __nv_bfloat16, __nv_bfloat16, float, __nv_bfloat16, 0, __nv_bfloat16, __nv_bfloat16, __nv_bfloat16, 1, 0, 0>(cublasLt::cublasSplitKParams<T6>, const T4 *, const T10 *, T9 *, T5 *, const T6 *, const T6 *, const T11 *, const T4 *, T11 *, void *, long, T6 *, int *, T6 *, T6 ... | bf16 | false | 9 | 4,096 | 14,336 | null | null | null | null | null | null | null | 0.00653 | 512 | 448 | 438,660 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | splitk_reduce | void cublasLt::splitKreduce_kernel<32, 16, int, __nv_bfloat16, __nv_bfloat16, float, __nv_bfloat16, 0, __nv_bfloat16, __nv_bfloat16, __nv_bfloat16, 1, 0, 0>(cublasLt::cublasSplitKParams<T6>, const T4 *, const T10 *, T9 *, T5 *, const T6 *, const T6 *, const T11 *, const T4 *, T11 *, void *, long, T6 *, int *, T6 *, T6 ... | bf16 | false | 53 | 4,096 | 14,336 | null | null | null | null | null | null | null | 0.00746 | 512 | 640 | 1,660,000 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | splitk_reduce | void cublasLt::splitKreduce_kernel<32, 16, int, __nv_bfloat16, __nv_bfloat16, float, __nv_bfloat16, 0, __nv_bfloat16, __nv_bfloat16, __nv_bfloat16, 1, 0, 0>(cublasLt::cublasSplitKParams<T6>, const T4 *, const T10 *, T9 *, T5 *, const T6 *, const T6 *, const T11 *, const T4 *, T11 *, void *, long, T6 *, int *, T6 *, T6 ... | bf16 | false | 24 | 4,096 | 14,336 | null | null | null | null | null | null | null | 0.0073 | 512 | 448 | 1,130,000 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | splitk_reduce | void cublasLt::splitKreduce_kernel<32, 16, int, __nv_bfloat16, __nv_bfloat16, float, __nv_bfloat16, 0, __nv_bfloat16, __nv_bfloat16, __nv_bfloat16, 1, 0, 0>(cublasLt::cublasSplitKParams<T6>, const T4 *, const T10 *, T9 *, T5 *, const T6 *, const T6 *, const T11 *, const T4 *, T11 *, void *, long, T6 *, int *, T6 *, T6 ... | bf16 | false | 23 | 4,096 | 14,336 | null | null | null | null | null | null | null | 0.00691 | 512 | 448 | 1,130,000 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | splitk_reduce | void cublasLt::splitKreduce_kernel<32, 16, int, float, __nv_bfloat16, float, __nv_bfloat16, 0, __nv_bfloat16, __nv_bfloat16, __nv_bfloat16, 1, 0, 0>(cublasLt::cublasSplitKParams<T6>, const T4 *, const T10 *, T9 *, T5 *, const T6 *, const T6 *, const T11 *, const T4 *, T11 *, void *, long, T6 *, int *, T6 *, T6 *, const... | bf16 | false | 128 | 1,024 | 4,096 | null | null | null | null | null | null | null | 0.00749 | 512 | 512 | 2,610,000 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01939 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.0192 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01882 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01866 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01984 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01933 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01962 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01955 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.0192 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01946 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01933 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01952 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01917 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.0191 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01984 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.0193 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.0192 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01958 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.0193 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01971 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01914 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01936 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01901 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01936 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01923 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01882 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01917 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01907 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01936 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01968 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01923 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | flash_attn | void pytorch_flash::flash_fwd_kernel<Flash_fwd_kernel_traits<128, 128, 64, 4, 0, 0, cutlass::bfloat16_t, Flash_kernel_traits<128, 128, 64, 4, cutlass::bfloat16_t>>, 0, 1, 0, 0, 1, 1, 0, 0>(pytorch_flash::Flash_fwd_params) | bf16 | false | null | null | null | 1 | 128 | 32 | 128 | 8 | null | flash_attn | 0.01917 | 128 | 32 | 1,650,000 | 255 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | copy | void at::vectorized_gather_kernel<16, long>(char *, char *, T2 *, int, long, long, long, long, bool) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 262,144 | other | 0.00826 | 256 | 256 | 1,060,000 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | cast | void at::unrolled_elementwise_kernel<at::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 11)]::operator ()() const::[lambda(bool) (instance 1)], std::array<char *, 2>, 4, TrivialOffsetCalculator<1, unsigned int>, TrivialOffsetCalculator<1, unsigned in... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | cast | 0.00451 | 128 | 1 | 34,940 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | reduce | void at::reduce_kernel<512, 1, at::ReduceOp<bool, at::func_wrapper_t<bool, at::and_kernel_cuda(at::TensorIterator &)::[lambda() (instance 1)]::operator ()() const::[lambda() (instance 12)]::operator ()() const::[lambda(bool, bool) (instance 1)]>, unsigned int, bool, 4, 4>>(T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 128 | other | 0.00582 | 32 | 1 | 54,270 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | cast | void at::unrolled_elementwise_kernel<at::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 7)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, 2>, 4, TrivialOffsetCalculator<1, unsigned int>, TrivialOffsetCalculator<1, unsigned in... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | cast | 0.0055 | 128 | 1 | 33,410 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | copy | void at::<unnamed>::CatArrayBatchedCopy<at::<unnamed>::OpaqueType<4>, unsigned int, 3, 64, 64>(T1 *, at::<unnamed>::CatArrInputTensorMetadata<T1, T2, T4, T5>, at::<unnamed>::TensorSizeStride<T2, 4>, int, T2) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 1,048,576 | other | 0.00762 | 512 | 432 | 38,530 | 30 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::vectorized_elementwise_kernel<4, at::cos_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 2)]::operator ()() const::[lambda() (instance 2)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, 2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 8,192 | rope | 0.00608 | 128 | 16 | 105,860 | 30 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::vectorized_elementwise_kernel<4, at::AUnaryFunctor<float, float, float, binary_internal::MulFunctor<float>>, std::array<char *, 2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 8,192 | mul | 0.00368 | 128 | 16 | 70,270 | 28 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::vectorized_elementwise_kernel<4, at::sin_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 2)]::operator ()() const::[lambda() (instance 2)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, 2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 8,192 | rope | 0.00544 | 128 | 16 | 105,340 | 30 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::vectorized_elementwise_kernel<4, at::AUnaryFunctor<float, float, float, binary_internal::MulFunctor<float>>, std::array<char *, 2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 8,192 | mul | 0.00362 | 128 | 16 | 70,400 | 28 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | cast | void at::vectorized_elementwise_kernel<4, at::bfloat16_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda(float) (instance 1)], std::array<char *, 2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 8,192 | cast | 0.00355 | 128 | 16 | 70,530 | 28 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | cast | void at::vectorized_elementwise_kernel<4, at::bfloat16_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda(float) (instance 1)], std::array<char *, 2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 8,192 | cast | 0.00365 | 128 | 16 | 70,660 | 28 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | cast | void at::unrolled_elementwise_kernel<at::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 7)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, 2>, 4, TrivialOffsetCalculator<1, unsigned int>, TrivialOffsetCalculator<1, unsigned in... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 524,288 | cast | 0.01005 | 128 | 1,024 | 1,080,000 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::vectorized_elementwise_kernel<4, void at::<unnamed>::pow_tensor_scalar_kernel_impl<float, float>(at::TensorIteratorBase &, T2)::[lambda(float) (instance 1)], std::array<char *, 2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 262,144 | rmsnorm | 0.00566 | 128 | 512 | 2,100,000 | 28 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | reduce | void at::reduce_kernel<512, 1, at::ReduceOp<float, at::MeanOps<float, float, float, float>, unsigned int, float, 4, 4>>(T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 16,384 | rmsnorm | 0.02755 | 512 | 8 | 2,150,000 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::vectorized_elementwise_kernel<4, at::CUDAFunctorOnSelf_add<float>, std::array<char *, 2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | residual | 0.00419 | 128 | 1 | 5,630 | 30 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::vectorized_elementwise_kernel<4, at::rsqrt_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 2)]::operator ()() const::[lambda() (instance 2)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, 2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | rmsnorm | 0.00323 | 128 | 1 | 6,780 | 22 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::elementwise_kernel<128, 2, void at::gpu_kernel_impl_nocast<at::BinaryFunctor<float, float, float, binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 1,048,576 | mul | 0.00755 | 128 | 2,048 | 2,110,000 | 16 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | cast | void at::vectorized_elementwise_kernel<4, at::bfloat16_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda(float) (instance 1)], std::array<char *, 2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 262,144 | cast | 0.00582 | 128 | 512 | 2,100,000 | 28 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::elementwise_kernel<128, 4, void at::gpu_kernel_impl_nocast<at::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 524,288 | mul | 0.00794 | 128 | 1,024 | 1,080,000 | 16 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::elementwise_kernel<128, 4, void at::gpu_kernel_impl_nocast<at::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 524,288 | mul | 0.00822 | 128 | 1,024 | 1,100,000 | 16 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::elementwise_kernel<128, 4, void at::gpu_kernel_impl_nocast<at::neg_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 2)]::operator ()() const::[lambda() (instance 9)]::operator ()() const::[lambda(c10::BFloat16) (instance 1)]>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 262,144 | neg | 0.00746 | 128 | 512 | 543,490 | 16 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | copy | void at::<unnamed>::CatArrayBatchedCopy<at::<unnamed>::OpaqueType<2>, unsigned int, 4, 64, 64>(T1 *, at::<unnamed>::CatArrInputTensorMetadata<T1, T2, T4, T5>, at::<unnamed>::TensorSizeStride<T2, 4>, int, T2) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 1,048,576 | other | 0.01299 | 512 | 432 | 1,060,000 | 23 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::elementwise_kernel<128, 4, void at::gpu_kernel_impl_nocast<at::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 524,288 | mul | 0.00794 | 128 | 1,024 | 1,100,000 | 16 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::elementwise_kernel<128, 4, void at::gpu_kernel_impl_nocast<at::CUDAFunctor_add<c10::BFloat16>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 524,288 | residual | 0.00851 | 128 | 1,024 | 2,120,000 | 16 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::elementwise_kernel<128, 4, void at::gpu_kernel_impl_nocast<at::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 131,072 | mul | 0.00726 | 128 | 256 | 314,620 | 16 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::elementwise_kernel<128, 4, void at::gpu_kernel_impl_nocast<at::neg_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 2)]::operator ()() const::[lambda() (instance 9)]::operator ()() const::[lambda(c10::BFloat16) (instance 1)]>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 65,536 | neg | 0.00694 | 128 | 128 | 148,860 | 16 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | copy | void at::<unnamed>::CatArrayBatchedCopy<at::<unnamed>::OpaqueType<2>, unsigned int, 4, 64, 64>(T1 *, at::<unnamed>::CatArrInputTensorMetadata<T1, T2, T4, T5>, at::<unnamed>::TensorSizeStride<T2, 4>, int, T2) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 1,048,576 | other | 0.00986 | 512 | 432 | 269,060 | 23 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::elementwise_kernel<128, 4, void at::gpu_kernel_impl_nocast<at::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 131,072 | mul | 0.0072 | 128 | 256 | 314,620 | 16 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::elementwise_kernel<128, 4, void at::gpu_kernel_impl_nocast<at::CUDAFunctor_add<c10::BFloat16>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 131,072 | residual | 0.00778 | 128 | 256 | 544,000 | 16 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | copy | void at::<unnamed>::CatArrayBatchedCopy<at::<unnamed>::OpaqueType<2>, unsigned int, 4, 64, 64>(T1 *, at::<unnamed>::CatArrInputTensorMetadata<T1, T2, T4, T5>, at::<unnamed>::TensorSizeStride<T2, 4>, int, T2) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 1,048,576 | other | 0.00864 | 512 | 432 | 269,180 | 23 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | copy | void at::<unnamed>::CatArrayBatchedCopy<at::<unnamed>::OpaqueType<2>, unsigned int, 4, 64, 64>(T1 *, at::<unnamed>::CatArrInputTensorMetadata<T1, T2, T4, T5>, at::<unnamed>::TensorSizeStride<T2, 4>, int, T2) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 1,048,576 | other | 0.0087 | 512 | 432 | 269,180 | 23 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::vectorized_elementwise_kernel<4, at::CUDAFunctor_add<c10::BFloat16>, std::array<char *, 3>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 262,144 | residual | 0.00544 | 128 | 512 | 2,100,000 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | cast | void at::unrolled_elementwise_kernel<at::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 7)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, 2>, 4, TrivialOffsetCalculator<1, unsigned int>, TrivialOffsetCalculator<1, unsigned in... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 524,288 | cast | 0.00886 | 128 | 1,024 | 1,080,000 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::vectorized_elementwise_kernel<4, void at::<unnamed>::pow_tensor_scalar_kernel_impl<float, float>(at::TensorIteratorBase &, T2)::[lambda(float) (instance 1)], std::array<char *, 2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 262,144 | rmsnorm | 0.00563 | 128 | 512 | 2,100,000 | 28 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | reduce | void at::reduce_kernel<512, 1, at::ReduceOp<float, at::MeanOps<float, float, float, float>, unsigned int, float, 4, 4>>(T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 16,384 | rmsnorm | 0.02618 | 512 | 8 | 2,150,000 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::vectorized_elementwise_kernel<4, at::CUDAFunctorOnSelf_add<float>, std::array<char *, 2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | residual | 0.00307 | 128 | 1 | 5,250 | 30 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::vectorized_elementwise_kernel<4, at::rsqrt_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 2)]::operator ()() const::[lambda() (instance 2)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, 2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | rmsnorm | 0.00333 | 128 | 1 | 6,780 | 22 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::elementwise_kernel<128, 2, void at::gpu_kernel_impl_nocast<at::BinaryFunctor<float, float, float, binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 1,048,576 | mul | 0.00768 | 128 | 2,048 | 2,110,000 | 16 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | cast | void at::vectorized_elementwise_kernel<4, at::bfloat16_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda(float) (instance 1)], std::array<char *, 2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 262,144 | cast | 0.00534 | 128 | 512 | 2,100,000 | 28 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::elementwise_kernel<128, 4, void at::gpu_kernel_impl_nocast<at::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 524,288 | mul | 0.00803 | 128 | 1,024 | 1,080,000 | 16 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | cast | void at::unrolled_elementwise_kernel<at::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 7)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, 2>, 4, TrivialOffsetCalculator<1, unsigned int>, TrivialOffsetCalculator<1, unsigned in... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 1,024 | cast | 0.00768 | 128 | 2 | 34,180 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | reduce | void <unnamed>::softmax_warp_forward<float, float, float, 3, 0, 0>(T2 *, const T1 *, int, int, int, const bool *, int, bool) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,048 | other | 0.0039 | 128 | 4 | 10,370 | 21 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | reduce | void at::reduce_kernel<512, 1, at::ReduceOp<float, at::func_wrapper_t<float, at::sum_functor<float, float, float>::operator ()(at::TensorIterator &)::[lambda(float, float) (instance 1)]>, unsigned int, float, 4, 4>>(T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 1,024 | other | 0.00656 | 256 | 1 | 52,220 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::elementwise_kernel<128, 2, void at::gpu_kernel_impl_nocast<at::BinaryFunctor<float, float, float, binary_internal::DivFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | other | 0.00464 | 128 | 1 | 14,590 | 19 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | cast | void at::vectorized_elementwise_kernel<4, at::bfloat16_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda(float) (instance 1)], std::array<char *, 2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | cast | 0.0032 | 128 | 1 | 6,020 | 28 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::vectorized_elementwise_kernel<4, at::FillFunctor<c10::BFloat16>, std::array<char *, 1>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 262,144 | fill | 0.00349 | 128 | 512 | 4,100 | 16 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::vectorized_elementwise_kernel<2, at::FillFunctor<long>, std::array<char *, 1>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 1,024 | fill | 0.00301 | 128 | 2 | 4,220 | 22 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::_scatter_gather_elementwise_kernel<128, 8, void at::_cuda_scatter_fill_internal_kernel<at::OpaqueType<8>, long>::operator ()<at::TensorAssign>(at::TensorIterator &, at::OpaqueType<8>, long, long, long, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | other | 0.00605 | 128 | 1 | 40,580 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | reduce | void at::reduce_kernel<128, 4, at::ReduceOp<long, at::func_wrapper_t<long, at::sum_functor<long, long, long>::operator ()(at::TensorIterator &)::[lambda(long, long) (instance 1)]>, unsigned int, long, 4, 4>>(T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | other | 0.01037 | 128 | 1 | 86,140 | 77 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::vectorized_elementwise_kernel<4, void at::compare_scalar_kernel<long>(at::TensorIteratorBase &, at::<unnamed>::OpType, T1)::[lambda(long) (instance 1)], std::array<char *, 2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | other | 0.00416 | 128 | 1 | 9,220 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | copy | void at::index_elementwise_kernel<128, 4, void at::gpu_index_kernel<void at::index_kernel_impl<at::OpaqueType<8>>(at::TensorIteratorBase &, c10::ArrayRef<long>, c10::ArrayRef<long>)::[lambda(char *, const char *, long) (instance 1)]>(at::TensorIteratorBase &, c10::ArrayRef<long>, c10::ArrayRef<long>, const T1 &, bool):... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | other | 0.00704 | 128 | 1 | 59,140 | 36 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | cast | void at::elementwise_kernel<128, 2, void at::gpu_kernel_impl_nocast<at::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 4)]::operator ()() const::[lambda(long) (instance 1)]>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | cast | 0.00426 | 128 | 1 | 12,160 | 16 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | copy | void at::vectorized_gather_kernel<16, long>(char *, char *, T2 *, int, long, long, long, long, bool) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 69,632 | other | 0.00442 | 256 | 68 | 282,240 | 32 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::vectorized_elementwise_kernel<4, at::<unnamed>::silu_kernel(at::TensorIteratorBase &)::[lambda() (instance 1)]::operator ()() const::[lambda() (instance 6)]::operator ()() const::[lambda(c10::BFloat16) (instance 1)], std::array<char *, 2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 262,144 | silu | 0.00579 | 128 | 476 | 988,160 | 28 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | elementwise | void at::vectorized_elementwise_kernel<4, at::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, binary_internal::MulFunctor<float>>, std::array<char *, 3>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 262,144 | mul | 0.00509 | 128 | 476 | 1,960,000 | 32 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.