fp16_bwd.cpp

/*
 * Copyright (c) 2023, NVIDIA CORPORATION. All rights reserved.
 *
 * Permission is hereby granted, free of charge, to any person obtaining a
 * copy of this software and associated documentation files (the "Software"),
 * to deal in the Software without restriction, including without limitation
 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
 * and/or sell copies of the Software, and to permit persons to whom the
 * Software is furnished to do so, subject to the following conditions:
 *
 * The above copyright notice and this permission notice shall be included in
 * all copies or substantial portions of the Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
 * DEALINGS IN THE SOFTWARE.
 */

#include <catch2/catch_test_macros.hpp>
#include "../utils/helpers.h"

#include <cuda_runtime_api.h>

#include <cudnn_frontend.h>
namespace fe = cudnn_frontend;

/*
Run this example by using command:
bin/samples "Toy sdpa backward"

This example shows how to construct a sdpa backward graph->
*/

// Tensors in backward pass
#define Q_UID 1
#define K_UID 2
#define V_UID 3
#define O_UID 4
#define STATS_UID 5
#define BIAS_UID 6
#define DBIAS_UID 7
#define SEQ_LEN_Q_UID 8
#define SEQ_LEN_KV_UID 9

#define DO_UID 101
#define DQ_UID 102
#define DK_UID 103
#define DV_UID 104

// Function to create the SDPA (Scaled Dot-Product Attention) backward graph
std::shared_ptr<fe::graph::Graph>
create_sdpa_backward_graph(int64_t const b,
                           int64_t const h_q,
                           int64_t const h_k,
                           int64_t const h_v,
                           int64_t const s_q,
                           int64_t const s_kv,
                           int64_t const d_qk,
                           int64_t const d_v,
                           float const attn_scale                     = 1.0f,
                           [[maybe_unused]] bool const generate_stats = true,
                           bool const causal_mask                     = false,
                           bool const alibi_mask                      = false,
                           bool const padding_mask                    = false,
                           bool has_attn_bias                         = false,
                           bool is_deterministic                      = false) {
    // Create a graph and set common global properties
    auto graph = std::make_shared<fe::graph::Graph>();
    graph->set_io_data_type(fe::DataType_t::BFLOAT16)
        .set_intermediate_data_type(fe::DataType_t::FLOAT)
        .set_compute_data_type(fe::DataType_t::FLOAT);

    // Define input tensors Q, K, V
    auto Q = graph->tensor(fe::graph::Tensor_attributes()
                               .set_name("Q")
                               .set_uid(Q_UID)
                               .set_dim({b, h_q, s_q, d_qk})
                               .set_stride({h_q * s_q * d_qk, s_q * d_qk, d_qk, 1}));

    auto K = graph->tensor(fe::graph::Tensor_attributes()
                               .set_name("K")
                               .set_uid(K_UID)
                               .set_dim({b, h_k, s_kv, d_qk})
                               .set_stride({h_k * s_kv * d_qk, s_kv * d_qk, d_qk, 1}));

    auto V = graph->tensor(fe::graph::Tensor_attributes()
                               .set_name("V")
                               .set_uid(V_UID)
                               .set_dim({b, h_v, s_kv, d_v})
                               .set_stride({h_v * s_kv * d_v, s_kv * d_v, d_v, 1}));

    // Define output tensor O
    auto O = graph->tensor(fe::graph::Tensor_attributes()
                               .set_name("O")
                               .set_uid(O_UID)
                               .set_dim({b, h_q, s_q, d_v})
                               .set_stride({h_q * s_q * d_v, s_q * d_v, d_v, 1}));

    // Define gradient tensor dO
    auto dO = graph->tensor(fe::graph::Tensor_attributes()
                                .set_name("dO")
                                .set_uid(DO_UID)
                                .set_dim({b, h_q, s_q, d_v})
                                .set_stride({h_q * s_q * d_v, s_q * d_v, d_v, 1}));

    // Define stats tensor
    auto Stats = graph->tensor(fe::graph::Tensor_attributes()
                                   .set_name("Stats")
                                   .set_uid(STATS_UID)
                                   .set_dim({b, h_q, s_q, 1})
                                   .set_stride({h_q * s_q, s_q, 1, 1})
                                   .set_data_type(fe::DataType_t::FLOAT));

    // Set SDPA backward options
    auto sdpa_options = fe::graph::SDPA_backward_attributes()
                            .set_name("flash_attention_backward")
                            .set_alibi_mask(alibi_mask)
                            .set_attn_scale(attn_scale)
                            .set_deterministic_algorithm(is_deterministic);

    if (causal_mask) {
        sdpa_options.set_diagonal_alignment(cudnn_frontend::DiagonalAlignment_t::TOP_LEFT)
            .set_diagonal_band_right_bound(0);
    }

    // If attention bias is provided, set it
    if (has_attn_bias) {
        auto bias = graph->tensor(fe::graph::Tensor_attributes()
                                      .set_name("bias")
                                      .set_uid(BIAS_UID)
                                      .set_dim({b, 1, s_q, s_kv})
                                      .set_stride({s_q * s_kv, s_q * s_kv, s_kv, 1}));
        sdpa_options.set_bias(bias);

        auto dbias = graph->tensor(fe::graph::Tensor_attributes()
                                       .set_name("dbias")
                                       .set_uid(DBIAS_UID)
                                       .set_dim({1, h_q, s_q, s_kv})
                                       .set_stride({s_q * s_kv * h_q, s_q * s_kv, s_kv, 1}));
        sdpa_options.set_dbias(dbias);
    }

    // If padding mask is enabled, set sequence lengths
    if (padding_mask) {
        auto seq_q  = graph->tensor(fe::graph::Tensor_attributes()
                                       .set_name("seq_q")
                                       .set_uid(SEQ_LEN_Q_UID)
                                       .set_dim({b, 1, 1, 1})
                                       .set_stride({1, 1, 1, 1})
                                       .set_data_type(fe::DataType_t::INT32));
        auto seq_kv = graph->tensor(fe::graph::Tensor_attributes()
                                        .set_name("seq_kv")
                                        .set_uid(SEQ_LEN_KV_UID)
                                        .set_dim({b, 1, 1, 1})
                                        .set_stride({1, 1, 1, 1})
                                        .set_data_type(fe::DataType_t::INT32));
        sdpa_options.set_padding_mask(padding_mask).set_seq_len_q(seq_q).set_seq_len_kv(seq_kv);
    }

    // Compute SDPA backward and get gradients dQ, dK, dV
    auto [dQ, dK, dV] = graph->sdpa_backward(Q, K, V, O, dO, Stats, sdpa_options);

    // Set output tensors dQ, dK, dV
    dQ->set_output(true)
        .set_uid(DQ_UID)
        .set_dim({b, h_q, s_q, d_qk})
        .set_stride({h_q * s_q * d_qk, s_q * d_qk, d_qk, 1});
    dK->set_output(true)
        .set_uid(DK_UID)
        .set_dim({b, h_k, s_kv, d_qk})
        .set_stride({h_k * s_kv * d_qk, s_kv * d_qk, d_qk, 1});
    dV->set_output(true)
        .set_uid(DV_UID)
        .set_dim({b, h_v, s_kv, d_v})
        .set_stride({h_v * s_kv * d_v, s_kv * d_v, d_v, 1});

    return graph;
}

// Test case for the SDPA backward graph
TEST_CASE("Toy sdpa backward", "[graph][sdpa][flash][backward]") {
    int64_t b             = 3;     // batch size
    int64_t h_q           = 4;     // head dim
    int64_t h_k           = 4;     // head dim
    int64_t h_v           = 4;     // head dim
    int64_t s_q           = 1024;  // q tensor is padded to this seq length
    int64_t s_kv          = 1024;  // k and v tensor is padded to this seq length
    int64_t d_qk          = 128;   // hidden dim
    int64_t d_v           = 128;   // hidden dim
    bool generate_stats   = true;
    float attn_scale      = 0.123f;
    bool causal_mask      = true;
    bool padding_mask     = (cudnnGetVersion() >= 8903);
    bool alibi_mask       = (cudnnGetVersion() >= 8904);
    bool has_attn_bias    = (cudnnGetVersion() >= 90500);
    bool is_deterministic = true;

    if (is_deterministic) {
        // switching off because NOT SUPPORTED in deterministic algorithm
        alibi_mask    = false;
        has_attn_bias = false;
    }

    if (cudnnGetVersion() < 8903) {
        SKIP("Test requires cudnn 8.9.3 or above");
        return;
    }

    // Create a unique_ptr for the cuDNN handle
    auto handle_ptr = create_cudnn_handle();
    auto handle     = *handle_ptr;

    // Create the SDPA backward graph
    auto graph = create_sdpa_backward_graph(b,
                                            h_q,
                                            h_k,
                                            h_v,
                                            s_q,
                                            s_kv,
                                            d_qk,
                                            d_v,
                                            attn_scale,
                                            generate_stats,
                                            causal_mask,
                                            alibi_mask,
                                            padding_mask,
                                            has_attn_bias,
                                            is_deterministic);

    REQUIRE(graph->build(handle, {fe::HeurMode_t::A}).is_good());

    //// Build variant pack
    // inputs
    Surface<half> q_tensor(b * h_q * s_q * d_qk);
    Surface<half> k_tensor(b * h_k * d_qk * s_kv);
    Surface<half> v_tensor(b * h_v * d_v * s_kv);
    Surface<half> o_tensor(b * h_q * s_q * d_v);
    Surface<half> dO_tensor(b * h_q * s_q * d_v);
    Surface<float> stats_tensor(b * h_q * s_q * 1);
    // outputs
    Surface<half> dQ_tensor(b * h_q * s_q * d_qk);
    Surface<half> dK_tensor(b * h_k * s_kv * d_qk);
    Surface<half> dV_tensor(b * h_v * s_kv * d_v);

    Surface<half> bias_tensor(1 * h_q * s_q * s_kv);
    Surface<half> dbias_tensor(1 * h_q * s_q * s_kv);

    // Create variant pack with input and output tensors
    std::unordered_map<fe::graph::Tensor_attributes::uid_t, void*> variant_pack = {// inputs
                                                                                   {Q_UID, q_tensor.devPtr},
                                                                                   {K_UID, k_tensor.devPtr},
                                                                                   {V_UID, v_tensor.devPtr},
                                                                                   {O_UID, o_tensor.devPtr},
                                                                                   {DO_UID, dO_tensor.devPtr},
                                                                                   {STATS_UID, stats_tensor.devPtr},
                                                                                   // outputs
                                                                                   {DQ_UID, dQ_tensor.devPtr},
                                                                                   {DK_UID, dK_tensor.devPtr},
                                                                                   {DV_UID, dV_tensor.devPtr}};

    // If attention bias is provided, add it to the variant pack
    if (has_attn_bias) {
        variant_pack[BIAS_UID]  = bias_tensor.devPtr;
        variant_pack[DBIAS_UID] = dbias_tensor.devPtr;
    }

    // If padding mask is enabled, add sequence lengths to the variant pack
    Surface<int32_t> devActualSeqlenQ(b);
    Surface<int32_t> devActualSeqlenKV(b);
    if (padding_mask) {
        std::vector<int32_t> hostActualSeqlenQ(b, 20);
        std::vector<int32_t> hostActualSeqlenKV(b, 20);

        CUDA_CHECK(cudaMemcpy(devActualSeqlenQ.devPtr,
                              hostActualSeqlenQ.data(),
                              sizeof(hostActualSeqlenQ[0]) * b,
                              cudaMemcpyHostToDevice));
        CUDA_CHECK(cudaMemcpy(devActualSeqlenKV.devPtr,
                              hostActualSeqlenKV.data(),
                              sizeof(hostActualSeqlenKV[0]) * b,
                              cudaMemcpyHostToDevice));
        CUDA_CHECK(cudaDeviceSynchronize());

        variant_pack[SEQ_LEN_Q_UID]  = devActualSeqlenQ.devPtr;
        variant_pack[SEQ_LEN_KV_UID] = devActualSeqlenKV.devPtr;
    }

    // Allocate workspace
    int64_t workspace_size = 0;
    REQUIRE(graph->get_workspace_size(workspace_size).is_good());
    Surface<int8_t> workspace(workspace_size);

    REQUIRE(graph->execute(handle, variant_pack, workspace.devPtr).is_good());

    CUDA_CHECK(cudaDeviceSynchronize());
}