Lumpiasty/llama.cpp
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

model : add Tencent Hy 4 (hy_v4) preview architecture support (#28127)

Browse Source 
* model: add Tencent Hy 4 (hy_v4) preview architecture support

Adds support for the Tencent Hy 4 model (Hugging Face architecture
HYV4ForCausalLM, GGUF arch hy_v4):

Add HF -> GGUF conversion script (conversion/hy_v4.py) and wire it into the conversion registry
Register hy_v4 GGUF constants, arch enum, and writer support
Implement the hy-v4 model graph, hparams, vocab and context changes
Register the new arch in llama-arch and models registry
Extend arch tests to cover hy_v4

Assisted by Claude Opus 5

* Update convert_hf_to_gguf_update.py

Co-authored-by: fairydreaming <166155368+fairydreaming@users.noreply.github.com>

* Update conversion/base.py

Co-authored-by: fairydreaming <166155368+fairydreaming@users.noreply.github.com>

* convert : move hy_v4 entry to the same place as in convert_hf_to_gguf_update.py

* model : apply changes related to n_ff_exp becoming per-layer in Hy4-preview

* n_layer_all

---------

Co-authored-by: fairydreaming <166155368+fairydreaming@users.noreply.github.com>
Co-authored-by: Stanisław Szymczyk <sszymczy@gmail.com>
Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@huggingface.co>
This commit is contained in:
HongHuang
2026-09-04 14:31:36 +02:00
committed by GitHub
co-authored by fairydreaming Stanisław Szymczyk Sigbjørn Skjæret
parent 5266f24da7
commit 49c0dc82b8
18 changed files with 1124 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/llama-context.cpp
+2 -1
View File
@@ -2317,7 +2317,8 @@ uint32_t llama_context::graph_max_nodes(uint32_t n_tokens) const {
(model.arch == LLM_ARCH_DFLASH && model.hparams.dsv4_hc_mult > 0) ||
model.arch == LLM_ARCH_NANBEIGE ||
model.arch == LLM_ARCH_MINIMAX_01 ||
model.arch == LLM_ARCH_MINIMAX_M3) {
model.arch == LLM_ARCH_MINIMAX_M3 ||
model.arch == LLM_ARCH_HY_V4) {
res = std::max<uint32_t>(n_tokens * 40, 32u * model.n_tensors());
} else if (model.arch == LLM_ARCH_DFLASH && model.hparams.dflash_selector_rank > 0) {
// DFlash2's convolutions and selector are shape work rather than matmuls,