> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.nvidia.com/nemo/automodel/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.nvidia.com/nemo/automodel/_mcp/server.

# nemo_automodel.components.models.glm_moe_dsa.kernels.tilelang_indexer_fwd

## Module Contents

### Functions

| Name                                                                                                                        | Description |
| --------------------------------------------------------------------------------------------------------------------------- | ----------- |
| [`clean_logits_`](#nemo_automodel-components-models-glm_moe_dsa-kernels-tilelang_indexer_fwd-clean_logits_)                 | -           |
| [`indexer_fwd_interface`](#nemo_automodel-components-models-glm_moe_dsa-kernels-tilelang_indexer_fwd-indexer_fwd_interface) | -           |
| [`tl_indexer_fwd_impl`](#nemo_automodel-components-models-glm_moe_dsa-kernels-tilelang_indexer_fwd-tl_indexer_fwd_impl)     | -           |

### API

```python
nemo_automodel.components.models.glm_moe_dsa.kernels.tilelang_indexer_fwd.clean_logits_(
    threads: int = 512,
    block_K: int = 4096
)
```

```python
nemo_automodel.components.models.glm_moe_dsa.kernels.tilelang_indexer_fwd.indexer_fwd_interface(
    q,
    kv,
    weights,
    cu_seqlen_ks,
    cu_seqlen_ke,
    clean_logits = True
)
```

```python
nemo_automodel.components.models.glm_moe_dsa.kernels.tilelang_indexer_fwd.tl_indexer_fwd_impl(
    heads,
    index_dim,
    block_N = 256,
    num_stages = 3,
    threads = 512,
    block_Q = None
)
```