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 | 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 | 0.00445 | 128 | 4 | 10,620 | 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 | 0.00592 | 256 | 1 | 51,840 | 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 | 0.00477 | 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 | 0.00342 | 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 | 0.00352 | 128 | 512 | 3,970 | 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 | 0.00285 | 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 | 0.0056 | 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 | 0.01002 | 128 | 1 | 86,400 | 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 | 0.00358 | 128 | 1 | 8,960 | 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 | 0.00733 | 128 | 1 | 59,260 | 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 | 0.00403 | 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 | 53,248 | other | 0.00432 | 256 | 52 | 216,700 | 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 | 0.00534 | 128 | 364 | 758,660 | 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 | 0.0048 | 128 | 364 | 1,500,000 | 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<2>>(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 | 0.00691 | 128 | 1 | 43,650 | 36 |
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 | 106,496 | mul | 0.0071 | 128 | 208 | 232,700 | 16 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | copy | void at::indexFuncLargeIndex<c10::BFloat16, long, unsigned int, 2, 2, (int)-2, 1, at::<unnamed>::ReduceAdd>(cuda::TensorInfo<T1, T3>, cuda::TensorInfo<const T1, T3>, cuda::TensorInfo<const T2, T3>, int, int, T3, T3, long, long, const T8 &, T1) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 524,288 | other | 0.00874 | 128 | 832 | 432,900 | 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 | 0.00739 | 128 | 1 | 59,260 | 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 | 0.00416 | 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 | 65,536 | other | 0.00442 | 256 | 64 | 265,860 | 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 | 0.00573 | 128 | 448 | 930,690 | 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 | 448 | 1,840,000 | 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<2>>(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 | 0.00691 | 128 | 1 | 43,650 | 36 |
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 | 0.0071 | 128 | 256 | 282,110 | 16 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | copy | void at::indexFuncLargeIndex<c10::BFloat16, long, unsigned int, 2, 2, (int)-2, 1, at::<unnamed>::ReduceAdd>(cuda::TensorInfo<T1, T3>, cuda::TensorInfo<const T1, T3>, cuda::TensorInfo<const T2, T3>, int, int, T3, T3, long, long, const T8 &, T1) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 524,288 | other | 0.01142 | 128 | 864 | 531,200 | 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 | 0.0073 | 128 | 1 | 59,260 | 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 | 61,440 | other | 0.00432 | 256 | 60 | 249,600 | 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 | 0.00547 | 128 | 420 | 873,340 | 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 | 0.00499 | 128 | 420 | 1,730,000 | 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<2>>(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 | 0.00678 | 128 | 1 | 43,780 | 36 |
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 | 122,880 | mul | 0.0071 | 128 | 240 | 265,600 | 16 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | copy | void at::indexFuncLargeIndex<c10::BFloat16, long, unsigned int, 2, 2, (int)-2, 1, at::<unnamed>::ReduceAdd>(cuda::TensorInfo<T1, T3>, cuda::TensorInfo<const T1, T3>, cuda::TensorInfo<const T2, T3>, int, int, T3, T3, long, long, const T8 &, T1) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 524,288 | other | 0.01114 | 128 | 864 | 498,560 | 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 | 0.00749 | 128 | 1 | 59,260 | 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 | 0.00406 | 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 | 96,256 | other | 0.00445 | 256 | 94 | 388,860 | 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 | 0.00643 | 128 | 658 | 1,360,000 | 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 | 0.00589 | 128 | 658 | 2,700,000 | 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<2>>(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 | 0.00672 | 128 | 1 | 44,030 | 36 |
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 | 262,144 | mul | 0.00723 | 128 | 376 | 404,860 | 16 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | copy | void at::indexFuncLargeIndex<c10::BFloat16, long, unsigned int, 2, 2, (int)-2, 1, at::<unnamed>::ReduceAdd>(cuda::TensorInfo<T1, T3>, cuda::TensorInfo<const T1, T3>, cuda::TensorInfo<const T2, T3>, int, int, T3, T3, long, long, const T8 &, T1) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 524,288 | other | 0.01203 | 128 | 864 | 777,220 | 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 | 0.00742 | 128 | 1 | 59,260 | 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 | 0.00406 | 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 | 26,624 | other | 0.00419 | 256 | 26 | 109,950 | 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 | 93,184 | silu | 0.00506 | 128 | 182 | 385,920 | 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 | 93,184 | mul | 0.00413 | 128 | 182 | 751,740 | 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<2>>(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 | 0.00678 | 128 | 1 | 43,520 | 36 |
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 | 53,248 | mul | 0.00701 | 128 | 104 | 126,340 | 16 |
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 | 0.00752 | 128 | 1 | 59,260 | 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 | 0.00435 | 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 | 34,816 | other | 0.00432 | 256 | 34 | 142,850 | 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 | 121,856 | silu | 0.0049 | 128 | 238 | 500,740 | 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 | 121,856 | mul | 0.00435 | 128 | 238 | 981,120 | 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<2>>(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 | 0.00678 | 128 | 1 | 43,520 | 36 |
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 | 69,632 | mul | 0.00701 | 128 | 136 | 158,980 | 16 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | copy | void at::indexFuncLargeIndex<c10::BFloat16, long, unsigned int, 2, 2, (int)-2, 1, at::<unnamed>::ReduceAdd>(cuda::TensorInfo<T1, T3>, cuda::TensorInfo<const T1, T3>, cuda::TensorInfo<const T2, T3>, int, int, T3, T3, long, long, const T8 &, T1) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 262,144 | other | 0.00858 | 128 | 544 | 285,440 | 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 | 0.00752 | 128 | 1 | 59,260 | 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 | 0.00435 | 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 | 55,296 | other | 0.00432 | 256 | 54 | 224,900 | 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 | 0.00534 | 128 | 378 | 787,330 | 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 | 0.0048 | 128 | 378 | 1,550,000 | 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<2>>(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 | 0.00685 | 128 | 1 | 43,650 | 36 |
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 | 110,592 | mul | 0.00726 | 128 | 216 | 240,900 | 16 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | copy | void at::indexFuncLargeIndex<c10::BFloat16, long, unsigned int, 2, 2, (int)-2, 1, at::<unnamed>::ReduceAdd>(cuda::TensorInfo<T1, T3>, cuda::TensorInfo<const T1, T3>, cuda::TensorInfo<const T2, T3>, int, int, T3, T3, long, long, const T8 &, T1) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 524,288 | other | 0.00874 | 128 | 864 | 449,280 | 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 | 0.00739 | 128 | 1 | 59,260 | 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 | 0.004 | 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 | 131,072 | other | 0.0047 | 256 | 128 | 528,130 | 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 | 524,288 | silu | 0.00717 | 128 | 896 | 1,850,000 | 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 | 524,288 | mul | 0.00656 | 128 | 896 | 3,680,000 | 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<2>>(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 | 0.00675 | 128 | 1 | 44,290 | 36 |
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 | 262,144 | mul | 0.00774 | 128 | 512 | 544,130 | 16 |
Mixtral-8x7B-Instruct_prefill | A100 | Mixtral-8x7B | copy | void at::indexFuncLargeIndex<c10::BFloat16, long, unsigned int, 2, 2, (int)-2, 1, at::<unnamed>::ReduceAdd>(cuda::TensorInfo<T1, T3>, cuda::TensorInfo<const T1, T3>, cuda::TensorInfo<const T2, T3>, int, int, T3, T3, long, long, const T8 &, T1) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 524,288 | other | 0.01536 | 128 | 864 | 1,060,000 | 32 |
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 | 0.00534 | 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 | 0.00858 | 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 | 0.00573 | 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 | 0.02672 | 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 | 0.00304 | 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 | 0.00326 | 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 | 0.008 | 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 | 0.00826 | 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 | 0.00733 | 128 | 512 | 542,080 | 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 | 0.0129 | 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 | 0.00787 | 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 | 0.00691 | 128 | 128 | 148,990 | 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 | 0.00982 | 512 | 432 | 269,180 | 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 | 0.0071 | 128 | 256 | 314,750 | 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 | 0.00746 | 128 | 256 | 544,130 | 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 | 0.00858 | 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 | 0.00867 | 512 | 432 | 269,060 | 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 | 0.00541 | 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 | 0.00896 | 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 | 0.0057 | 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 | 0.0265 | 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 | 0.00301 | 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 | 0.00326 | 128 | 1 | 6,910 | 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 | 0.00784 | 128 | 2,048 | 2,110,000 | 16 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.