finish HardSReLUE #110

S1/hli28146_#110/HardSReLUE_cuda.py

@@ -0,0 +1,107 @@
+import torch
+import torch.nn as nn
+from torch.utils.cpp_extension import load_inline
+
+cpp_source = """
+#include <torch/extension.h>
+
+torch::Tensor hardsrelue_cuda_forward(const torch::Tensor& input, float alpha);
+"""
+
+cuda_source = """
+#include <torch/extension.h>
+#include <cuda_runtime.h>
+#include <math.h>
+
+#define BLOCK_SIZE 256
+
+struct __align__(16) Float4 {
+    float x, y, z, w;
+};
+
+// HardSReLUE Logic
+__device__ __forceinline__ float compute_hardsrelue(float x, float alpha) {
+    if (x < 0.0f) {
+        return alpha * (expf(x) - 1.0f);
+    } else {
+        float hardsig = fminf(fmaxf(0.5f * x + 0.5f, 0.0f), 1.0f);
+        return alpha * x * hardsig + x;
+    }
+}
+
+__global__ void hardsrelue_kernel(
+    float* __restrict__ output,
+    const float* __restrict__ input,
+    const int n,
+    const float alpha) 
+{
+    const int idx = blockIdx.x * blockDim.x + threadIdx.x;
+    const int vec_n = n / 4;
+    
+    int i = idx;
+    const int stride = blockDim.x * gridDim.x;
+
+    for (; i < vec_n; i += stride) {
+        Float4 in_vec = reinterpret_cast<const Float4*>(input)[i];
+        Float4 out_vec;
+
+        out_vec.x = compute_hardsrelue(in_vec.x, alpha);
+        out_vec.y = compute_hardsrelue(in_vec.y, alpha);
+        out_vec.z = compute_hardsrelue(in_vec.z, alpha);
+        out_vec.w = compute_hardsrelue(in_vec.w, alpha);
+
+        reinterpret_cast<Float4*>(output)[i] = out_vec;
+    }
+
+    int start_scalar = vec_n * 4;
+    int global_tid = blockIdx.x * blockDim.x + threadIdx.x;
+    int total_threads = gridDim.x * gridDim.x;
+    
+    int current_idx = start_scalar + global_tid;
+    while (current_idx < n) {
+        output[current_idx] = compute_hardsrelue(input[current_idx], alpha);
+        current_idx += total_threads;
+    }
+}
+
+torch::Tensor hardsrelue_cuda_forward(const torch::Tensor& input, float alpha) {
+    TORCH_CHECK(input.is_cuda(), "Input must be a CUDA tensor");
+    TORCH_CHECK(input.is_contiguous(), "Input must be contiguous");
+    
+    const int n = input.numel();
+    auto output = torch::empty_like(input);
+
+    const int vec_n = n / 4;
+    const int grid_size = (vec_n + BLOCK_SIZE - 1) / BLOCK_SIZE;
+    
+    int final_grid = (grid_size < 1) ? 1 : grid_size;
+    if (final_grid > 65535) final_grid = 65535;
+
+    hardsrelue_kernel<<<final_grid, BLOCK_SIZE>>>(
+        output.data_ptr<float>(),
+        input.data_ptr<float>(),
+        n,
+        alpha
+    );
+
+    return output;
+}
+"""
+
+hardsrelue_op_module = load_inline(
+    name='hardsrelue_op',
+    cpp_sources=cpp_source,
+    cuda_sources=cuda_source,
+    functions=['hardsrelue_cuda_forward'],
+    verbose=False,
+    extra_cuda_cflags=['-O3']
+)
+
+class ModelNew(nn.Module):
+    def __init__(self, alpha=0.2):
+        super(ModelNew, self).__init__()
+        self.alpha = alpha
+        self.op = hardsrelue_op_module
+
+    def forward(self, input_tensor: torch.Tensor) -> torch.Tensor:
+        return self.op.hardsrelue_cuda_forward(input_tensor.contiguous(), self.alpha)

S1/hli28146_#110/HardSReLUE_torch.py

@@ -0,0 +1,47 @@
+import torch
+import torch.nn as nn
+
+BATCH_SIZE = 4096
+HIDDEN_DIM = 4096
+SHAPE = (BATCH_SIZE, HIDDEN_DIM)
+
+# HardSReLUE 超参数 a ,论文中设为 0.2
+ALPHA_VALUE = 0.2
+
+class HardSReLUE(nn.Module):
+    '''
+    A novel nonlinear hybrid HardSReLUE activation function
+    https://link.springer.com/content/pdf/10.1007/s11042-022-14313-w.pdf
+
+    Formula:
+      f(x) = a * (exp(x) - 1)                      if x < 0
+      f(x) = a*x*max(0, min(1, 0.5x+0.5)) + x      if x >= 0
+    '''
+    def __init__(self, alpha=0.2):
+        super(HardSReLUE, self).__init__()
+        self.alpha = alpha
+
+    def forward(self, x: torch.Tensor) -> torch.Tensor:
+        # Negative part
+        neg_part = self.alpha * (torch.exp(x) - 1.0)
+        
+        # Positive part
+        hardsig_part = torch.clamp(0.5 * x + 0.5, 0.0, 1.0)
+        pos_part = self.alpha * x * hardsig_part + x
+        
+        return torch.where(x < 0, neg_part, pos_part)
+
+class Model(nn.Module):
+    def __init__(self, alpha=0.2):
+        super(Model, self).__init__()
+        self.act = HardSReLUE(alpha=alpha)
+    
+    def forward(self, x):
+        return self.act(x)
+
+def get_inputs():
+    input_tensor = torch.randn(SHAPE, dtype=torch.float32) * 5.0
+    return [input_tensor.contiguous()]
+
+def get_init_inputs():
+    return [ALPHA_VALUE]

S1/hli28146_#110/prompt.txt

@@ -0,0 +1,74 @@
+Write a custom CUDA kernel to optimize `HardSReLUE`.
+
+Formula:
+  f(x) = a * (exp(x) - 1)                      if x < 0
+  f(x) = a*x*max(0, min(1, 0.5x+0.5)) + x    if x >= 0
+
+Problem Analysis:
+1. Computationally Intensive & Memory Bound: The operation is element-wise but has multiple branches and involves exp, min, max, mul, add.
+2. Operator Chaining: A PyTorch implementation using `torch.where` would create numerous intermediate tensors.
+
+Optimization Strategy: Fused Element-wise Kernel with Vectorization
+
+1. One-Thread-per-Element: Map each element to a CUDA thread.
+
+2. Vectorized Loads (float4): Use `float4` to process 128 bits per memory transaction.
+
+3. Fused Branching Logic:
+   - For each element `x`, check `if (x < 0)`.
+   - If true: `a * (__expf(x) - 1.0f)`.
+   - If false: `hardsig_part = fminf(fmaxf(0.5f * x + 0.5f, 0.0f), 1.0f)`, then `a*x*hardsig_part + x`.
+   - All computations are fused in registers.
+
+4. One-Pass: Fuse all steps into a single read-compute-write kernel.
+  
+Here's an example to show you the syntax of inline embedding custom CUDA operators in torch: The example given architecture is:   
+  
+```python
+import torch
+import torch.nn as nn
+
+BATCH_SIZE = 4096
+HIDDEN_DIM = 4096
+SHAPE = (BATCH_SIZE, HIDDEN_DIM)
+
+# HardSReLUE 超参数 a ,论文中设为 0.2
+ALPHA_VALUE = 0.2
+
+class HardSReLUE(nn.Module):
+    '''
+    A novel nonlinear hybrid HardSReLUE activation function
+    https://link.springer.com/content/pdf/10.1007/s11042-022-14313-w.pdf
+
+    Formula:
+      f(x) = a * (exp(x) - 1)                      if x < 0
+      f(x) = a*x*max(0, min(1, 0.5x+0.5)) + x      if x >= 0
+    '''
+    def __init__(self, alpha=0.2):
+        super(HardSReLUE, self).__init__()
+        self.alpha = alpha
+
+    def forward(self, x: torch.Tensor) -> torch.Tensor:
+        # Negative part
+        neg_part = self.alpha * (torch.exp(x) - 1.0)
+        
+        # Positive part
+        hardsig_part = torch.clamp(0.5 * x + 0.5, 0.0, 1.0)
+        pos_part = self.alpha * x * hardsig_part + x
+        
+        return torch.where(x < 0, neg_part, pos_part)
+
+class Model(nn.Module):
+    def __init__(self, alpha=0.2):
+        super(Model, self).__init__()
+        self.act = HardSReLUE(alpha=alpha)
+    
+    def forward(self, x):
+        return self.act(x)
+
+def get_inputs():
+    input_tensor = torch.randn(SHAPE, dtype=torch.float32) * 5.0
+    return [input_tensor.contiguous()]
+
+def get_init_inputs():
+    return [ALPHA_VALUE]

S1/hli28146_#110/run_code.py

@@ -0,0 +1,74 @@
+###########################################################
+# 性能和精度验证程序
+###########################################################
+import torch
+import torch.nn as nn
+import time
+from HardSReLUE_torch import Model,get_inputs,get_init_inputs
+from HardSReLUE_cuda import ModelNew
+
+def run_benchmark():
+    # 检查 CUDA 是否可用
+    if not torch.cuda.is_available():
+        print("CUDA 不可用，请确保您有可用的 NVIDIA GPU 并已正确安装 PyTorch CUDA 版本。")
+        return
+    else:
+        device = torch.device("cuda")
+
+    # 初始化模型
+    init_inputs = get_init_inputs()
+    init_inputs = [
+        x.cuda(device=device) if isinstance(x, torch.Tensor) else x for x in init_inputs
+    ]
+    inputs = get_inputs()
+    inputs = [
+        x.cuda(device=device) if isinstance(x, torch.Tensor) else x for x in inputs
+    ]
+
+    torch_model = Model(*init_inputs).cuda()
+    cuda_model = ModelNew(*init_inputs).cuda()
+
+    torch_model.eval()
+    cuda_model.eval()
+
+    print("-------------------- 精度对齐验证 --------------------")
+    with torch.no_grad():
+        output_torch = torch_model( *inputs)
+        output_cuda = cuda_model(*inputs)
+
+    precision_flag = torch.allclose(output_torch, output_cuda,rtol=1e-03)
+    if precision_flag:
+        print("✅ 精度对齐：两个模型的输出结果非常接近。")
+    else:
+        print("❌ 精度不一致！")
+        
+    print("\n-------------------- 性能加速比测试 --------------------")
+    num_iterations = 100 
+    
+    # PyTorch 模型计时
+    torch.cuda.synchronize()
+    start_time = time.time()
+    for _ in range(num_iterations):
+        _ = torch_model(*inputs)
+    torch.cuda.synchronize()
+    torch_time = (time.time() - start_time) / num_iterations
+    
+    # 自定义 CUDA 内核计时
+    torch.cuda.synchronize()
+    start_time = time.time()
+    for _ in range(num_iterations):
+        _ = cuda_model(*inputs)
+    torch.cuda.synchronize()
+    cuda_time = (time.time() - start_time) / num_iterations
+    
+    print(f"PyTorch torch.relu 平均执行时间: {torch_time:.6f} 秒")
+    print(f"自定义 CUDA 内核 平均执行时间: {cuda_time:.6f} 秒")
+    speedup = 0
+    if cuda_time > 0:
+        speedup = torch_time / cuda_time
+        print(f"加速比 (Speedup): {speedup:.2f}x")
+    else:
+        print("CUDA 内核执行时间为0，无法计算加速比。")
+    return precision_flag,speedup
+if __name__ == "__main__":
+    precision_flag,speedup = run_benchmark()