ltx23_transformer_block

LTX-2 transformer block

Description

Dual-stream (video + audio) block: modulated self-attention per modality, text cross-attention per modality (with LTX-2.3 query and key/value modulation), bidirectional audio-video cross-attention with global+per-block modulation, and modulated feed-forward.

Usage

ltx23_transformer_block(dim, num_attention_heads, attention_head_dim,
                        cross_attention_dim, audio_dim,
                        audio_num_attention_heads, audio_attention_head_dim,
                        audio_cross_attention_dim, video_gated_attn = TRUE,
                        video_cross_attn_adaln = TRUE, audio_gated_attn = TRUE,
                        audio_cross_attn_adaln = TRUE, eps = 1e-06,
                        elementwise_affine = FALSE, rope_type = "split",
                        perturbed_attn = TRUE)

Arguments

  • cross_attention_dim: Integer. Text embedding dim for video.
  • audio_cross_attention_dim: Integer. Text embedding dim for audio.
  • eps: Numeric. Norm epsilon.
  • elementwise_affine: Logical. Block norms carry weights (FALSE for LTX).
  • rope_type: “split” or “interleaved”.
  • perturbed_attn: Logical. Enable the STG perturbation arguments.
  • dim,audio_dim: Integers. Video/audio stream dimensions.
  • num_attention_heads,attention_head_dim: Video attention shape.
  • audio_num_attention_heads,audio_attention_head_dim: Audio attention shape.
  • video_gated_attn,audio_gated_attn: Logicals. Per-head output gates.
  • video_cross_attn_adaln,audio_cross_attn_adaln: Logicals. LTX-2.3 text cross-attention modulation (9 mod params instead of 6).