finish AdversarialLoss #76

S1/uucoco_#76/AdversarialLoss_cuda.py

@@ -0,0 +1,84 @@
+import torch
+import torch.nn as nn
+from torch.utils.cpp_extension import load_inline
+
+
+class ModelNew(nn.Module):
+    def __init__(self, reduction='mean'):
+        super().__init__()
+        self.reduction = reduction
+        self._compile_cuda_kernel()
+
+    def _compile_cuda_kernel(self):
+        cpp_source = """
+        torch::Tensor adversarial_loss_cuda(torch::Tensor input, torch::Tensor target);
+        """
+
+        cuda_source = """
+        #include <torch/extension.h>
+        #include <cuda_runtime.h>
+        #include <math.h>
+
+        __global__ void adversarial_loss_kernel(
+            const float* __restrict__ input,
+            const float* __restrict__ target,
+            float* __restrict__ output,
+            const int n_elements)
+        {
+            const int tid = blockIdx.x * blockDim.x + threadIdx.x;
+            const int stride = blockDim.x * gridDim.x;
+
+            for (int i = tid; i < n_elements; i += stride) {
+                float x = input[i];
+                float y = target[i];
+
+                // Binary Cross Entropy with Logits Stability Formula:
+                // max(x, 0) - x * y + log(1 + exp(-abs(x)))
+
+                float max_val = fmaxf(x, 0.0f);
+                float abs_val = fabsf(x);
+                float log_term = log1pf(expf(-abs_val));
+
+                output[i] = max_val - x * y + log_term;
+            }
+        }
+
+        torch::Tensor adversarial_loss_cuda(torch::Tensor input, torch::Tensor target) {
+            auto input_c = input.contiguous();
+            auto target_c = target.contiguous();
+            const int n_elements = input_c.numel();
+
+            auto output = torch::empty_like(input_c);
+
+            const int threads = 256;
+            const int blocks = min((n_elements + threads - 1) / threads, 65535);
+
+            adversarial_loss_kernel<<<blocks, threads>>>(
+                input_c.data_ptr<float>(),
+                target_c.data_ptr<float>(),
+                output.data_ptr<float>(),
+                n_elements
+            );
+
+            return output;
+        }
+        """
+
+        self.op = load_inline(
+            name="adversarial_loss_op",
+            cpp_sources=cpp_source,
+            cuda_sources=cuda_source,
+            functions=["adversarial_loss_cuda"],
+            extra_cuda_cflags=["-O3"],
+            verbose=False
+        )
+
+    def forward(self, input, target):
+        loss_elementwise = self.op.adversarial_loss_cuda(input, target)
+
+        if self.reduction == 'mean':
+            return loss_elementwise.mean()
+        elif self.reduction == 'sum':
+            return loss_elementwise.sum()
+        else:
+            return loss_elementwise

S1/uucoco_#76/AdversarialLoss_torch.py

@@ -0,0 +1,26 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+
+class Model(nn.Module):
+    def __init__(self, reduction='mean'):
+        super().__init__()
+        self.reduction = reduction
+
+    def forward(self, input: torch.Tensor, target: torch.Tensor) -> torch.Tensor:
+        return F.binary_cross_entropy_with_logits(input, target, reduction=self.reduction)
+
+
+batch_size = 128
+feature_dim = 1
+
+
+def get_inputs():
+    pred = torch.randn(batch_size, feature_dim, dtype=torch.float32)
+    target = torch.randint(0, 2, (batch_size, feature_dim)).float()
+    return [pred, target]
+
+
+def get_init_inputs():
+    return ['mean']

S1/uucoco_#76/prompt.txt

@@ -0,0 +1,45 @@
+You have complete freedom to choose the set of operators you want to replace. You may make the decision to replace some operators with custom CUDA kernels and leave others unchanged. You may replace multiple operators with custom implementations, consider operator fusion opportunities (combining multiple operators into a single kernel, for example, combining chunk+gelu+elementwise_mul), or algorithmic changes (such as optimized memory access patterns). You are only limited by your imagination.
+PyTorch C++/CUDA Extension: Inline compilation using torch.utils.cpp_extension.load_inline.
+
+Custom CUDA Kernel: Implements a binary cross-entropy with logits loss directly on GPU.
+
+Kernel Launch Configuration: Uses 256 threads per block and dynamically calculates block count (capped at 65535).
+
+Numerically Stable Formula: Applies max(x,0) - x*y + log(1 + exp(-|x|)) for element-wise loss.
+
+Reduction Options: Supports 'none', 'mean', and 'sum' reductions after kernel computation.
+
+Memory Contiguity: Ensures input and target tensors are contiguous before kernel execution.
+
+
+
+
+
+
+Here's an example to show you the syntax of inline embedding custom CUDA operators in torch: The example given architecture is:
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+
+class Model(nn.Module):
+    def __init__(self, reduction='mean'):
+        super().__init__()
+        self.reduction = reduction
+
+    def forward(self, input: torch.Tensor, target: torch.Tensor) -> torch.Tensor:
+        return F.binary_cross_entropy_with_logits(input, target, reduction=self.reduction)
+
+
+batch_size = 128
+feature_dim = 1
+
+
+def get_inputs():
+    pred = torch.randn(batch_size, feature_dim, dtype=torch.float32)
+    target = torch.randint(0, 2, (batch_size, feature_dim)).float()
+    return [pred, target]
+
+
+def get_init_inputs():
+    return ['mean']

S1/uucoco_#76/run_code.py

@@ -0,0 +1,77 @@
+###########################################################
+# 性能和精度验证程序
+###########################################################
+import torch
+import torch.nn as nn
+import time
+from AdversarialLoss_torch import Model, get_inputs, get_init_inputs
+from AdversarialLoss_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()