vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

llama : add support for GLM-Edge and GLM-Edge-V series models (#10573)

Browse source 
* add glm edge chat model

* use config partial_rotary_factor as rope ratio

* support for glm edge model

* vision model support

* remove debug info

* fix format

* llava.cpp trailing whitespace

* remove unused AutoTokenizer

* Update src/llama.cpp for not contain <|end|> or </s>

Co-authored-by: Xuan Son Nguyen <thichthat@gmail.com>

* add edge template

* fix chat template

* fix confict

* fix confict

* fix ci err

* fix format err

* fix template err

* 9b hf chat support

* format

* format clip.cpp

* fix format

* Apply suggestions from code review

* Apply suggestions from code review

* Update examples/llava/clip.cpp

* fix format

* minor : style

---------

Co-authored-by: liyuhang <yuhang.li@zhipuai.cn>
Co-authored-by: piDack <pcdack@hotmail.co>
Co-authored-by: Xuan Son Nguyen <thichthat@gmail.com>
Co-authored-by: liyuhang <yuhang.li@aminer.cn>
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
This commit is contained in:
piDack 2025-02-02 15:48:46 +08:00 • committed by GitHub
parent 53debe6f3c
commit 0cec062a63
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
15 changed files with 568 additions and 67 deletions
Download patch file Download diff file Expand all files Collapse all files

17
examples/llava/llava.cpp
View file

@ -311,6 +311,20 @@ static bool encode_image_with_clip(clip_ctx * ctx_clip, int n_threads, const cli
img_res_v.size = 0;
img_res_v.data = nullptr;
}
else if (clip_is_glm(ctx_clip)){
struct clip_image_size * load_image_size = clip_image_size_init();
load_image_size->width = img_res_v.data[0].nx;
load_image_size->height = img_res_v.data[0].ny;
clip_add_load_image_size(ctx_clip, load_image_size);
bool encoded = clip_image_encode(ctx_clip, n_threads, &img_res_v.data[0], image_embd);
int pos = int(load_image_size->width/clip_patch_size(ctx_clip)/2);
*n_img_pos = (pos * pos + 2);
if (!encoded){
LOG_ERR("Unable to encode image \n");
return false;
}
}
else if (strcmp(mm_patch_merge_type, "spatial_unpad") != 0) {
// flat / default llava-1.5 type embedding
*n_img_pos = clip_n_patches(ctx_clip);
@ -395,6 +409,9 @@ bool llava_image_embed_make_with_clip_img(clip_ctx * ctx_clip, int n_threads, co
if (clip_is_minicpmv(ctx_clip)) {
num_max_patches = 10;
}
if (clip_is_glm(ctx_clip)) {
num_max_patches = 1;
}
float * image_embd;
if (clip_is_qwen2vl(ctx_clip)) {
// qwen2vl don't split image into chunks, so `num_max_patches` is not needed.