undefined

**For a minimal diffusers integration example (~150 lines):**
```bash
python scripts/ltx_kernel_injection_example.py

python scripts/transformers_injection_example.py

from kernels import get_kernel

**For a complete HuggingFace Kernels example:**
```bash
python scripts/huggingface_kernels_example.py

python benchmark_rmsnorm.py

undefined

undefined

.claude/skills/cuda-kernels/
├── scripts/
│   ├── benchmark_example.py              # End-to-end video generation benchmark
│   ├── benchmark_rmsnorm.py              # Isolated RMSNorm micro-benchmark
│   ├── ltx_kernel_injection_example.py   # Minimal diffusers integration (~150 lines)
│   ├── transformers_injection_example.py # Minimal transformers integration (~120 lines)
│   └── huggingface_kernels_example.py    # HuggingFace Kernels Hub integration
├── references/
│   ├── diffusers-integration.md          # Complete diffusers integration guide
│   ├── transformers-integration.md       # Complete transformers integration guide
│   ├── huggingface-kernels-integration.md # HuggingFace Kernels Hub (get_kernel) guide
│   ├── troubleshooting.md                # Common issues and solutions
│   ├── kernel-templates.md               # CUDA kernel templates (includes vectorized)
│   ├── h100-optimization-guide.md        # H100 (Hopper) optimization deep dive
│   ├── a100-optimization-guide.md        # A100 (Ampere) optimization deep dive
│   └── t4-optimization-guide.md          # T4 (Turing) optimization deep dive
└── SKILL.md                              # This file

examples/ltx_video/                  # Complete working example
├── kernel_src/
│   └── rmsnorm.cu                  # Vectorized RMSNorm kernel (2.67x faster)
├── torch-ext/                      # PyTorch bindings
├── generate_video.py               # Full benchmark script
├── benchmark_rmsnorm.py            # Isolated kernel benchmark
└── setup.py                        # pip install -e .

// Load 2 bfloat16 elements at once (32-bit load)
const __nv_bfloat162* vec_input = reinterpret_cast<const __nv_bfloat162*>(row_input);

#pragma unroll 4
for (int i = tid; i < vec_hidden; i += stride) {
    __nv_bfloat162 v = vec_input[i];
    float v0 = __bfloat162float(v.x);
    float v1 = __bfloat162float(v.y);
    sum_sq += v0 * v0 + v1 * v1;
}

const __half2* vec_input = reinterpret_cast<const __half2*>(row_input);
__half2 v = vec_input[i];
float v0 = __half2float(v.x);
float v1 = __half2float(v.y);

const float4* vec_input = reinterpret_cast<const float4*>(row_input);
float4 v = vec_input[i];
sum_sq += v.x * v.x + v.y * v.y + v.z * v.z + v.w * v.w;

template <typename T>
__device__ __forceinline__ T warp_reduce_sum(T val) {
    #pragma unroll
    for (int offset = 16; offset > 0; offset >>= 1) {
        val += __shfl_xor_sync(0xffffffff, val, offset);
    }
    return val;
}

constexpr int BLOCK_SIZE = 256;
int num_blocks = (total_elements + BLOCK_SIZE - 1) / BLOCK_SIZE;

// Divide by 2 for bf16/fp16 vectorized access
int threads = min(hidden_size / 2, MAX_THREADS);
threads = max(threads, WARP_SIZE);
threads = (threads + 32 - 1) / 32 * 32;  // Round to warp boundary

nix run .#build-and-copy --max-jobs 2 --cores 8 -L

uv pip install -e .

[general]
name = "ltx_kernels"
backends = ["cuda"]

[kernel.your_kernel]
backend = "cuda"
src = ["kernel_src/your_kernel.cu"]
cuda-capabilities = ["9.0"]

from kernels import get_kernel, has_kernel

# Use the kernel
x = torch.randn((4, 4), dtype=torch.float16, device="cuda")
y = torch.empty_like(x)
activation.gelu_fast(y, x)

**Key functions:**
- `get_kernel(repo_id, version=None)` - Download and load kernel from Hub
- `has_kernel(repo_id)` - Check if compatible build exists
- `get_local_kernel(path)` - Load from local directory (development)

**Popular community kernels:**
- `kernels-community/activation` - GELU, SiLU, etc.
- `kernels-community/flash-attn` - Flash Attention 2
- `kernels-community/triton-layer-norm` - LayerNorm, RMSNorm

from transformers import AutoModelForCausalLM
from ltx_kernels import rmsnorm

def patch_rmsnorm(model):
    for name, module in model.named_modules():
        if 'RMSNorm' in type(module).__name__:
            eps = getattr(module, 'variance_epsilon', None) or getattr(module, 'eps', 1e-6)
            def make_forward(mod, epsilon):
                def forward(x):
                    return rmsnorm(x, mod.weight, eps=epsilon)
                return forward
            module.forward = make_forward(module, eps)

model = AutoModelForCausalLM.from_pretrained("meta-llama/Llama-2-7b-hf", torch_dtype=torch.bfloat16)
patch_rmsnorm(model)

undefined

**Solution:** Handle both cases:
```python
has_weight = hasattr(module, 'weight') and module.weight is not None
if has_weight:
    output = rmsnorm(x, module.weight, eps=eps)
else:
    weight = torch.ones(x.shape[-1], device=x.device, dtype=x.dtype)
    output = rmsnorm(x, weight, eps=eps)

undefined

undefined

pipe = LTXPipeline.from_pretrained(...)
pipe.to("cuda")
inject_optimized_kernels(pipe)  # BEFORE offloading
pipe.enable_model_cpu_offload()  # Now safe

from diffusers import LTXPipeline
from ltx_kernels import rmsnorm

def patch_rmsnorm_modules(model):
    """Patch all RMSNorm modules to use custom kernel."""
    for name, module in model.named_modules():
        if type(module).__name__ == 'RMSNorm':
            eps = getattr(module, 'eps', 1e-6)
            has_weight = hasattr(module, 'weight') and module.weight is not None

            if has_weight:
                def make_forward(mod, epsilon):
                    def forward(x):
                        return rmsnorm(x, mod.weight, eps=epsilon)
                    return forward
                module.forward = make_forward(module, eps)
            else:
                def make_forward(epsilon):
                    def forward(x):
                        w = torch.ones(x.shape[-1], device=x.device, dtype=x.dtype)
                        return rmsnorm(x, w, eps=epsilon)
                    return forward
                module.forward = make_forward(eps)

undefined

undefined

undefined

torch._dynamo.exc.Unsupported: Attempted to call function marked as skipped

import torch

@torch.library.custom_op("ltx_kernels::rmsnorm", mutates_args={"out"})
def rmsnorm(out: torch.Tensor, input: torch.Tensor, weight: torch.Tensor, eps: float) -> None:
    ops.rmsnorm_forward(out, input.contiguous(), weight.contiguous(), eps)

@rmsnorm.register_fake
def _(out, input, weight, eps):
    pass  # No shape changes