01d8eaa28d
* nemotron nano v2 vlm support added * simplified code; addressed reviews * pre-downsample position embeddings during GGUF conversion for fixed input size
35 lines
1.1 KiB
C++
35 lines
1.1 KiB
C++
#include "models.h"
|
|
|
|
ggml_cgraph * clip_graph_nemotron_v2_vl::build() {
|
|
GGML_ASSERT(model.class_embedding != nullptr);
|
|
GGML_ASSERT(model.position_embeddings != nullptr);
|
|
|
|
const int n_registers = model.class_embedding->ne[1];
|
|
const int n_pos = n_patches + n_registers;
|
|
|
|
ggml_tensor * inp = build_inp();
|
|
|
|
// add position embeddings (pre-downsampled during GGUF conversion for fixed 512x512 input)
|
|
inp = ggml_add(ctx0, inp, model.position_embeddings);
|
|
cb(inp, "inp_pos", -1);
|
|
|
|
inp = ggml_concat(ctx0, model.class_embedding, inp, 1);
|
|
|
|
ggml_tensor * cur = build_vit(inp, n_pos, NORM_TYPE_NORMAL, hparams.ffn_op, nullptr, nullptr);
|
|
|
|
cur = ggml_view_2d(ctx0, cur,
|
|
n_embd, n_patches,
|
|
ggml_row_size(cur->type, n_embd),
|
|
n_registers * ggml_row_size(cur->type, n_embd));
|
|
|
|
cur = build_patch_merge_permute(cur, model.hparams.n_merge);
|
|
|
|
{
|
|
cur = build_norm(cur, model.mm_0_w, nullptr, NORM_TYPE_RMS, 1e-6, -1);
|
|
cur = build_ffn(cur, model.mm_1_w, nullptr, nullptr, nullptr, model.mm_3_w, nullptr, FFN_RELU_SQR, -1);
|
|
}
|
|
|
|
ggml_build_forward_expand(gf, cur);
|
|
|
|
return gf;
|
|
}
|