Struct GlmVit 

pub struct GlmVit {
    pub out_hidden_size: usize,
    /* private fields */
}

Fields

out_hidden_size: usize

Implementations

impl GlmVit

pub fn forward_batched( &self, images: &[(&[f32], usize, usize)], gpu: &dyn GpuBackend, stream: u64, ) -> Result<Vec<(usize, usize, usize)>>

2026-09-25: Encode images (pixels, grid h, grid w). Returns per-image (post_h, post_w, merged_p) in image order, the VisionTower::forward_batched contract.

Images are encoded in greedy chunks whose Σpatches fits p_max, and each chunk’s merged rows land after the previous chunk’s in buf_out. Refuses a grid not divisible by the merge size, an image larger than p_max, and a batch whose merged rows exceed buf_out.

impl GlmVit

pub fn new( patch_embed_w: DevicePtr, patch_embed_b: DevicePtr, blocks: Vec<GlmVitBlock>, merger: GlmVitMerger, geo: &GlmVitGeometry, gpu: &dyn GpuBackend, ) -> Result<Self>

pub fn out_row(&self, row: usize) -> DevicePtr

2026-09-25: Device pointer to row row of the packed merged output (buf_out). Panics when the scratch is not allocated.