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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::neg_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 2)]::operator ()() const::[lambda() (instance 9)]::operator ()() const::[lambda(c10::BFloat16) (instance 1)]>(at::TensorIteratorBase &, const T1 &)::[l... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 458,752 | neg | 0.00611 | 128 | 512 | 547,580 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | copy | void at::native::<unnamed>::CatArrayBatchedCopy<at::native::<unnamed>::OpaqueType<(unsigned int)2>, unsigned int, (int)4, (int)64, (int)64>(T1 *, at::native::<unnamed>::CatArrInputTensorMetadata<T1, T2, T4, T5>, at::native::<unnamed>::TensorSizeStride<T2, (unsigned int)4>, int, T2) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | other | 0.00848 | 512 | 528 | 1,060,000 | 25 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 458,752 | mul | 0.0063 | 128 | 1,024 | 1,140,000 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::CUDAFunctor_add<c10::BFloat16>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 458,752 | residual | 0.00672 | 128 | 1,024 | 2,120,000 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 458,752 | mul | 0.0064 | 128 | 512 | 616,960 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::neg_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 2)]::operator ()() const::[lambda() (instance 9)]::operator ()() const::[lambda(c10::BFloat16) (instance 1)]>(at::TensorIteratorBase &, const T1 &)::[l... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 131,072 | neg | 0.00598 | 128 | 256 | 285,180 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | copy | void at::native::<unnamed>::CatArrayBatchedCopy<at::native::<unnamed>::OpaqueType<(unsigned int)2>, unsigned int, (int)4, (int)64, (int)64>(T1 *, at::native::<unnamed>::CatArrInputTensorMetadata<T1, T2, T4, T5>, at::native::<unnamed>::TensorSizeStride<T2, (unsigned int)4>, int, T2) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | other | 0.00742 | 512 | 528 | 531,460 | 25 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 458,752 | mul | 0.0064 | 128 | 512 | 616,700 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::CUDAFunctor_add<c10::BFloat16>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 458,752 | residual | 0.00653 | 128 | 512 | 1,080,000 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | copy | void at::native::<unnamed>::CatArrayBatchedCopy<at::native::<unnamed>::OpaqueType<(unsigned int)2>, unsigned int, (int)4, (int)64, (int)64>(T1 *, at::native::<unnamed>::CatArrInputTensorMetadata<T1, T2, T4, T5>, at::native::<unnamed>::TensorSizeStride<T2, (unsigned int)4>, int, T2) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | other | 0.00726 | 512 | 528 | 531,200 | 25 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | copy | void at::native::<unnamed>::CatArrayBatchedCopy<at::native::<unnamed>::OpaqueType<(unsigned int)2>, unsigned int, (int)4, (int)64, (int)64>(T1 *, at::native::<unnamed>::CatArrInputTensorMetadata<T1, T2, T4, T5>, at::native::<unnamed>::TensorSizeStride<T2, (unsigned int)4>, int, T2) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | other | 0.00736 | 512 | 528 | 531,460 | 25 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | cast | void at::native::unrolled_elementwise_kernel<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 7)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>, (int)4, TrivialOffsetCalculator<(int)1, unsigned int... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 458,752 | cast | 0.00637 | 128 | 896 | 950,270 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, void at::native::<unnamed>::pow_tensor_scalar_kernel_impl<float, float>(at::TensorIteratorBase &, T2)::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 229,376 | rmsnorm | 0.00435 | 128 | 448 | 1,840,000 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | reduce | void at::native::reduce_kernel<(int)512, (int)1, at::native::ReduceOp<float, at::native::MeanOps<float, float, float, float>, unsigned int, float, (int)4, (int)4>>(T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 16,384 | rmsnorm | 0.01728 | 512 | 8 | 1,900,000 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::CUDAFunctorOnSelf_add<float>, std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | residual | 0.00368 | 128 | 1 | 5,890 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::rsqrt_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 2)]::operator ()() const::[lambda() (instance 2)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | rmsnorm | 0.0039 | 128 | 1 | 7,170 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)2, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<float, float, float, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | mul | 0.00605 | 128 | 1,792 | 1,850,000 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | cast | void at::native::unrolled_elementwise_kernel<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 7)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>, (int)4, TrivialOffsetCalculator<(int)1, unsigned int... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 3,584 | cast | 0.00598 | 128 | 7 | 39,680 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::CUDAFunctorOnSelf_add<float>, std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,048 | residual | 0.00467 | 128 | 4 | 19,710 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)2, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<float, float, float, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | mul | 0.00605 | 128 | 1,792 | 1,860,000 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | cast | void at::native::vectorized_elementwise_kernel<(int)8, at::native::bfloat16_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 229,376 | cast | 0.0049 | 128 | 448 | 1,840,000 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::CUDAFunctor_add<c10::BFloat16>, std::array<char *, (unsigned long)3>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 229,376 | residual | 0.0047 | 128 | 448 | 1,840,000 | 34 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | cast | void at::native::unrolled_elementwise_kernel<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 7)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>, (int)4, TrivialOffsetCalculator<(int)1, unsigned int... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 458,752 | cast | 0.00653 | 128 | 896 | 950,270 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, void at::native::<unnamed>::pow_tensor_scalar_kernel_impl<float, float>(at::TensorIteratorBase &, T2)::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 229,376 | rmsnorm | 0.0049 | 128 | 448 | 1,840,000 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | reduce | void at::native::reduce_kernel<(int)512, (int)1, at::native::ReduceOp<float, at::native::MeanOps<float, float, float, float>, unsigned int, float, (int)4, (int)4>>(T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 16,384 | rmsnorm | 0.0175 | 512 | 8 | 1,900,000 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::CUDAFunctorOnSelf_add<float>, std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | residual | 0.00378 | 128 | 1 | 5,890 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::rsqrt_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 2)]::operator ()() const::[lambda() (instance 2)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | rmsnorm | 0.00406 | 128 | 1 | 7,170 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)2, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<float, float, float, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | mul | 0.0063 | 128 | 1,792 | 1,850,000 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | cast | void at::native::unrolled_elementwise_kernel<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 7)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>, (int)4, TrivialOffsetCalculator<(int)1, unsigned int... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 3,584 | cast | 0.00595 | 128 | 7 | 39,940 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::CUDAFunctorOnSelf_add<float>, std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,048 | residual | 0.00384 | 128 | 4 | 19,710 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)2, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<float, float, float, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | mul | 0.00598 | 128 | 1,792 | 1,860,000 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | cast | void at::native::vectorized_elementwise_kernel<(int)8, at::native::bfloat16_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 229,376 | cast | 0.0048 | 128 | 448 | 1,840,000 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::GeluCUDAKernelImpl(at::TensorIteratorBase &, at::native::GeluType)::[lambda() (instance 1)]::operator ()() const::[lambda() (instance 4)]::operator ()() const::[lambda(c10::BFloat16) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | other | 0.00637 | 128 | 1,792 | 3,680,000 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, at::native::binary_internal::MulFunctor<float>>, std::array<char *, (unsigned long)3>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | mul | 0.00707 | 128 | 1,792 | 7,350,000 | 40 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | cast | void at::native::unrolled_elementwise_kernel<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 7)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>, (int)4, TrivialOffsetCalculator<(int)1, unsigned int... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 458,752 | cast | 0.00659 | 128 | 896 | 950,020 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, void at::native::<unnamed>::pow_tensor_scalar_kernel_impl<float, float>(at::TensorIteratorBase &, T2)::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 229,376 | rmsnorm | 0.00486 | 128 | 448 | 1,840,000 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | reduce | void at::native::reduce_kernel<(int)512, (int)1, at::native::ReduceOp<float, at::native::MeanOps<float, float, float, float>, unsigned int, float, (int)4, (int)4>>(T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 16,384 | rmsnorm | 0.01706 | 512 | 8 | 1,900,000 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::CUDAFunctorOnSelf_add<float>, std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | residual | 0.00394 | 128 | 1 | 5,890 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::rsqrt_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 2)]::operator ()() const::[lambda() (instance 2)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | rmsnorm | 0.00378 | 128 | 1 | 7,170 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)2, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<float, float, float, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | mul | 0.00614 | 128 | 1,792 | 1,850,000 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | cast | void at::native::unrolled_elementwise_kernel<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 7)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>, (int)4, TrivialOffsetCalculator<(int)1, unsigned int... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 3,584 | cast | 0.00589 | 128 | 7 | 39,940 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::CUDAFunctorOnSelf_add<float>, std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,048 | residual | 0.00435 | 128 | 4 | 19,710 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)2, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<float, float, float, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | mul | 0.00602 | 128 | 1,792 | 1,860,000 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | cast | void at::native::vectorized_elementwise_kernel<(int)8, at::native::bfloat16_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 229,376 | cast | 0.00502 | 128 | 448 | 1,840,000 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::CUDAFunctor_add<c10::BFloat16>, std::array<char *, (unsigned long)3>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 229,376 | residual | 0.00477 | 128 | 448 | 1,840,000 | 34 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | cast | void at::native::unrolled_elementwise_kernel<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 7)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>, (int)4, TrivialOffsetCalculator<(int)1, unsigned int... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 458,752 | cast | 0.0065 | 128 | 896 | 950,270 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, void at::native::<unnamed>::pow_tensor_scalar_kernel_impl<float, float>(at::TensorIteratorBase &, T2)::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 229,376 | rmsnorm | 0.00493 | 128 | 448 | 1,840,000 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | reduce | void at::native::reduce_kernel<(int)512, (int)1, at::native::ReduceOp<float, at::native::MeanOps<float, float, float, float>, unsigned int, float, (int)4, (int)4>>(T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 16,384 | rmsnorm | 0.01738 | 512 | 8 | 1,900,000 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::CUDAFunctorOnSelf_add<float>, std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | residual | 0.00435 | 128 | 1 | 5,630 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::rsqrt_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 2)]::operator ()() const::[lambda() (instance 2)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | rmsnorm | 0.0039 | 128 | 1 | 7,170 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)2, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<float, float, float, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | mul | 0.00602 | 128 | 1,792 | 1,850,000 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | cast | void at::native::unrolled_elementwise_kernel<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 7)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>, (int)4, TrivialOffsetCalculator<(int)1, unsigned int... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 3,584 | cast | 0.00618 | 128 | 7 | 39,940 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::CUDAFunctorOnSelf_add<float>, std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,048 | residual | 0.00378 | 128 | 4 | 19,710 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)2, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<float, float, float, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | mul | 0.00608 | 128 | 1,792 | 1,860,000 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | cast | void at::native::vectorized_elementwise_kernel<(int)8, at::native::bfloat16_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 229,376 | cast | 0.00515 | 128 | 448 | 1,840,000 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 458,752 | mul | 0.00691 | 128 | 1,024 | 1,140,000 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::neg_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 2)]::operator ()() const::[lambda() (instance 9)]::operator ()() const::[lambda(c10::BFloat16) (instance 1)]>(at::TensorIteratorBase &, const T1 &)::[l... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 458,752 | neg | 0.00608 | 128 | 512 | 547,580 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | copy | void at::native::<unnamed>::CatArrayBatchedCopy<at::native::<unnamed>::OpaqueType<(unsigned int)2>, unsigned int, (int)4, (int)64, (int)64>(T1 *, at::native::<unnamed>::CatArrInputTensorMetadata<T1, T2, T4, T5>, at::native::<unnamed>::TensorSizeStride<T2, (unsigned int)4>, int, T2) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | other | 0.00854 | 512 | 528 | 1,060,000 | 25 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 458,752 | mul | 0.00634 | 128 | 1,024 | 1,140,000 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::CUDAFunctor_add<c10::BFloat16>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 458,752 | residual | 0.00666 | 128 | 1,024 | 2,120,000 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 458,752 | mul | 0.00637 | 128 | 512 | 616,960 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::neg_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 2)]::operator ()() const::[lambda() (instance 9)]::operator ()() const::[lambda(c10::BFloat16) (instance 1)]>(at::TensorIteratorBase &, const T1 &)::[l... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 131,072 | neg | 0.00598 | 128 | 256 | 285,440 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | copy | void at::native::<unnamed>::CatArrayBatchedCopy<at::native::<unnamed>::OpaqueType<(unsigned int)2>, unsigned int, (int)4, (int)64, (int)64>(T1 *, at::native::<unnamed>::CatArrInputTensorMetadata<T1, T2, T4, T5>, at::native::<unnamed>::TensorSizeStride<T2, (unsigned int)4>, int, T2) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | other | 0.0073 | 512 | 528 | 531,710 | 25 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 458,752 | mul | 0.00627 | 128 | 512 | 616,700 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)4, void at::native::gpu_kernel_impl_nocast<at::native::CUDAFunctor_add<c10::BFloat16>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 458,752 | residual | 0.00646 | 128 | 512 | 1,080,000 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | copy | void at::native::<unnamed>::CatArrayBatchedCopy<at::native::<unnamed>::OpaqueType<(unsigned int)2>, unsigned int, (int)4, (int)64, (int)64>(T1 *, at::native::<unnamed>::CatArrInputTensorMetadata<T1, T2, T4, T5>, at::native::<unnamed>::TensorSizeStride<T2, (unsigned int)4>, int, T2) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | other | 0.0072 | 512 | 528 | 531,200 | 25 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | copy | void at::native::<unnamed>::CatArrayBatchedCopy<at::native::<unnamed>::OpaqueType<(unsigned int)2>, unsigned int, (int)4, (int)64, (int)64>(T1 *, at::native::<unnamed>::CatArrInputTensorMetadata<T1, T2, T4, T5>, at::native::<unnamed>::TensorSizeStride<T2, (unsigned int)4>, int, T2) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | other | 0.00698 | 512 | 528 | 531,200 | 25 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | cast | void at::native::unrolled_elementwise_kernel<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 7)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>, (int)4, TrivialOffsetCalculator<(int)1, unsigned int... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 458,752 | cast | 0.00637 | 128 | 896 | 950,270 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, void at::native::<unnamed>::pow_tensor_scalar_kernel_impl<float, float>(at::TensorIteratorBase &, T2)::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 229,376 | rmsnorm | 0.00445 | 128 | 448 | 1,840,000 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | reduce | void at::native::reduce_kernel<(int)512, (int)1, at::native::ReduceOp<float, at::native::MeanOps<float, float, float, float>, unsigned int, float, (int)4, (int)4>>(T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 16,384 | rmsnorm | 0.01747 | 512 | 8 | 1,900,000 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::CUDAFunctorOnSelf_add<float>, std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | residual | 0.00368 | 128 | 1 | 5,890 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::rsqrt_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 2)]::operator ()() const::[lambda() (instance 2)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | rmsnorm | 0.00387 | 128 | 1 | 6,910 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)2, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<float, float, float, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | mul | 0.00608 | 128 | 1,792 | 1,850,000 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | cast | void at::native::unrolled_elementwise_kernel<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 7)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>, (int)4, TrivialOffsetCalculator<(int)1, unsigned int... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 3,584 | cast | 0.00624 | 128 | 7 | 39,940 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::CUDAFunctorOnSelf_add<float>, std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,048 | residual | 0.00378 | 128 | 4 | 19,710 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)2, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<float, float, float, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | mul | 0.00605 | 128 | 1,792 | 1,860,000 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | cast | void at::native::vectorized_elementwise_kernel<(int)8, at::native::bfloat16_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 229,376 | cast | 0.00426 | 128 | 448 | 1,840,000 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::CUDAFunctor_add<c10::BFloat16>, std::array<char *, (unsigned long)3>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 229,376 | residual | 0.00509 | 128 | 448 | 1,840,000 | 34 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | cast | void at::native::unrolled_elementwise_kernel<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 7)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>, (int)4, TrivialOffsetCalculator<(int)1, unsigned int... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 458,752 | cast | 0.00614 | 128 | 896 | 950,270 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, void at::native::<unnamed>::pow_tensor_scalar_kernel_impl<float, float>(at::TensorIteratorBase &, T2)::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 229,376 | rmsnorm | 0.00474 | 128 | 448 | 1,840,000 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | reduce | void at::native::reduce_kernel<(int)512, (int)1, at::native::ReduceOp<float, at::native::MeanOps<float, float, float, float>, unsigned int, float, (int)4, (int)4>>(T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 16,384 | rmsnorm | 0.01712 | 512 | 8 | 1,900,000 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::CUDAFunctorOnSelf_add<float>, std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | residual | 0.0039 | 128 | 1 | 5,630 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::rsqrt_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 2)]::operator ()() const::[lambda() (instance 2)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | rmsnorm | 0.00374 | 128 | 1 | 7,170 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)2, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<float, float, float, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | mul | 0.00621 | 128 | 1,792 | 1,850,000 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | cast | void at::native::unrolled_elementwise_kernel<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 7)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>, (int)4, TrivialOffsetCalculator<(int)1, unsigned int... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 3,584 | cast | 0.00566 | 128 | 7 | 39,940 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::CUDAFunctorOnSelf_add<float>, std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,048 | residual | 0.00483 | 128 | 4 | 19,710 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)2, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<float, float, float, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | mul | 0.00614 | 128 | 1,792 | 1,860,000 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | cast | void at::native::vectorized_elementwise_kernel<(int)8, at::native::bfloat16_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 229,376 | cast | 0.00483 | 128 | 448 | 1,840,000 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::GeluCUDAKernelImpl(at::TensorIteratorBase &, at::native::GeluType)::[lambda() (instance 1)]::operator ()() const::[lambda() (instance 4)]::operator ()() const::[lambda(c10::BFloat16) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | other | 0.00621 | 128 | 1,792 | 3,680,000 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)8, at::native::BinaryFunctor<c10::BFloat16, c10::BFloat16, c10::BFloat16, at::native::binary_internal::MulFunctor<float>>, std::array<char *, (unsigned long)3>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | mul | 0.00691 | 128 | 1,792 | 7,350,000 | 40 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | cast | void at::native::unrolled_elementwise_kernel<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 7)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>, (int)4, TrivialOffsetCalculator<(int)1, unsigned int... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 458,752 | cast | 0.00653 | 128 | 896 | 950,270 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, void at::native::<unnamed>::pow_tensor_scalar_kernel_impl<float, float>(at::TensorIteratorBase &, T2)::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 229,376 | rmsnorm | 0.00442 | 128 | 448 | 1,840,000 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | reduce | void at::native::reduce_kernel<(int)512, (int)1, at::native::ReduceOp<float, at::native::MeanOps<float, float, float, float>, unsigned int, float, (int)4, (int)4>>(T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 16,384 | rmsnorm | 0.0169 | 512 | 8 | 1,900,000 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::CUDAFunctorOnSelf_add<float>, std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | residual | 0.00442 | 128 | 1 | 5,890 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::rsqrt_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 2)]::operator ()() const::[lambda() (instance 2)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 512 | rmsnorm | 0.00384 | 128 | 1 | 7,170 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)2, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<float, float, float, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | mul | 0.00621 | 128 | 1,792 | 1,850,000 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | cast | void at::native::unrolled_elementwise_kernel<at::native::direct_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda() (instance 3)]::operator ()() const::[lambda() (instance 7)]::operator ()() const::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>, (int)4, TrivialOffsetCalculator<(int)1, unsigned int... | bf16 | false | null | null | null | 1 | 128 | null | null | null | 3,584 | cast | 0.00598 | 128 | 7 | 39,940 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::vectorized_elementwise_kernel<(int)4, at::native::CUDAFunctorOnSelf_add<float>, std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 2,048 | residual | 0.0039 | 128 | 4 | 19,460 | 32 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | elementwise | void at::native::elementwise_kernel<(int)128, (int)2, void at::native::gpu_kernel_impl_nocast<at::native::BinaryFunctor<float, float, float, at::native::binary_internal::MulFunctor<float>>>(at::TensorIteratorBase &, const T1 &)::[lambda(int) (instance 1)]>(int, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 917,504 | mul | 0.00614 | 128 | 1,792 | 1,860,000 | 18 |
gemma-2-9b-it_prefill | H100 | Gemma-9B | cast | void at::native::vectorized_elementwise_kernel<(int)8, at::native::bfloat16_copy_kernel_cuda(at::TensorIteratorBase &)::[lambda(float) (instance 1)], std::array<char *, (unsigned long)2>>(int, T2, T3) | bf16 | false | null | null | null | 1 | 128 | null | null | null | 229,376 | cast | 0.00474 | 128 | 448 | 1,840,000 | 32 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.