forked from ccf-ai-infra/GPUCodeForces
46 lines
1.7 KiB
Python
46 lines
1.7 KiB
Python
import torch
|
|
import torch.nn as nn
|
|
from torch.utils.cpp_extension import load_inline
|
|
|
|
cuda_source = """
|
|
#include <torch/extension.h>
|
|
#include <cuda_runtime.h>
|
|
|
|
__global__ void irl_kernel(const float* pred_rewards, const float* expert_log_probs, const float* policy_log_probs, float* expert_out, float* policy_out, int size) {
|
|
int idx = blockIdx.x * blockDim.x + threadIdx.x;
|
|
if (idx < size) {
|
|
expert_out[idx] = -pred_rewards[idx] * expert_log_probs[idx];
|
|
policy_out[idx] = pred_rewards[idx] * policy_log_probs[idx];
|
|
}
|
|
}
|
|
|
|
torch::Tensor irl_cuda(torch::Tensor pred_rewards, torch::Tensor expert_log_probs, torch::Tensor policy_log_probs) {
|
|
auto size = pred_rewards.numel();
|
|
auto expert_out = torch::empty_like(pred_rewards);
|
|
auto policy_out = torch::empty_like(pred_rewards);
|
|
const int block_size = 256;
|
|
int num_blocks = (size + block_size - 1) / block_size;
|
|
irl_kernel<<<num_blocks, block_size>>>(pred_rewards.data_ptr<float>(), expert_log_probs.data_ptr<float>(), policy_log_probs.data_ptr<float>(), expert_out.data_ptr<float>(), policy_out.data_ptr<float>(), size);
|
|
return expert_out.mean() + policy_out.mean();
|
|
}
|
|
"""
|
|
|
|
cpp_source = """
|
|
torch::Tensor irl_cuda(torch::Tensor pred_rewards, torch::Tensor expert_log_probs, torch::Tensor policy_log_probs);
|
|
"""
|
|
|
|
irl_module = load_inline(
|
|
name="inverse_rl",
|
|
cpp_sources=cpp_source,
|
|
cuda_sources=cuda_source,
|
|
functions=["irl_cuda"],
|
|
verbose=True
|
|
)
|
|
|
|
class ModelNew(torch.nn.Module):
|
|
def __init__(self):
|
|
super(ModelNew, self).__init__()
|
|
self.irl_module = irl_module
|
|
|
|
def forward(self, pred_rewards, expert_log_probs, policy_log_probs):
|
|
return self.irl_module.irl_cuda(pred_rewards, expert_log_probs, policy_log_probs) |