For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /max/get-started.md).
Python class
RaggedAttention
RaggedAttention
class max.nn.attention.RaggedAttention(*, mask_variant, num_attention_heads, num_key_value_heads, hidden_size, kv_params, devices=None, dtype=float32, linear_cls=<class 'max.nn.linear.Linear'>, stacked_qkv=False, scale=None, has_bias=False, clip_qkv=None)
Bases: Module
Layer that computes the self attention score for ragged inputs.
Initializes the attention layer.
-
Parameters:
-
- rope – The rope layer to borrow the freqs_cis value from.
- num_attention_heads (int) – The number of attention heads.
- num_key_value_heads (int) – Number of key/value heads.
- hidden_size (int) – The dimension of the hidden states.
- kv_params (KVCacheParams) – KV Cache Params, including the number of kv heads, the head dim, and data type.
- dtype (DType) – DType of the
- devices (list[DeviceRef] | None) – Device to place the weights and run the computation. If multiple are provided, the first device is used.
- linear_cls (Callable[..., Linear]) – Linear class to use for the outputs dense layer.
- stacked_qkv (bool) – Whether the weights are stacked together.
- scale (float | None) – Value used to scale the results of the attention output.
- has_bias (bool) – Whether to use an attention bias.
- clip_qkv (float | None) – If provided, the QKV weights are clamped between [-clip_qkv, clip_qkv]
- mask_variant (MHAMaskVariant)
Was this page helpful?
Thank you! We'll create more content like this.
Thank you for helping us improve!