Public/koboldcpp
1
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2026-09-19 17:25:07 +02:00
Code Issues Packages Projects Releases Wiki Activity

mtmd checkpoint 2

Browse Source
This commit is contained in:
Concedo
2026-06-09 17:55:17 +08:00
parent 794a271cfa
commit 90a14cecf8
4 changed files with 137 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files
gpttype_adapter.cpp
+121 -7
View File
@@ -49,6 +49,8 @@
#include "mpt_v3.cpp"
#include "tools/mtmd/mtmd.h"
#include "tools/mtmd/mtmd-helper.h"
#include "vendor/stb/stb_image.h"
#include "otherarch/sdcpp/thirdparty/stb_image_resize.h"
#include "common/common.h"
#include "ggml-rpc.h"
@@ -119,8 +121,10 @@ static llama_context * guidance_ctx = nullptr; //for classifier free guidance, w
static mtmd_context * mtmd_ctx = nullptr; //for multimodal media
static std::vector<media_object> media_objects;
static std::vector<int> last_media_mem; //for storing dummy tokens that will be consumed by mtmd
static int last_media_pos_count = 0;
static std::string media_composite_image_signature = ""; //for identifying when the media changes, we need to invalidate the cache
static int current_media_identifier = MEDIA_TOKEN_IDENTIFIER_A;
static int vision_max_res = 2048;
static bool use_mrope = false;
static kcpp_params * kcpp_data = nullptr;
@@ -2003,7 +2007,7 @@ static bool kcpp_eval_media(llama_context * ctx_llama, const media_chunk & media
*n_past,
0,
n_batch,
true,
false,
&new_n_past);
if (result != 0) {
fprintf(stderr, "\n%s : failed to eval mtmd media chunk, status %d\n", __func__, result);
@@ -2323,6 +2327,7 @@ ModelLoadResult gpttype_load_model(const load_model_inputs inputs, FileFormat in
}
kcpp_data->vision_min_tokens = inputs.visionmintokens;
kcpp_data->vision_max_tokens = inputs.visionmaxtokens;
vision_max_res = inputs.visionmaxres;
if(isGguf && kcpp_pipeline_parallelism)
{
//double the logical batch, while keeping the physical batch the same, pipeline parallel set GGML_SCHED_MAX_COPIES to 2
@@ -4388,6 +4393,92 @@ int GetThreadsToUse(bool blasmode)
return kcpp_data->n_threads;
}
static mtmd_bitmap * kcpp_mtmd_bitmap_init_image_from_buf(const unsigned char * buf, size_t len, int maxdims)
{
int nx = 0;
int ny = 0;
int nc = 0;
uint8_t * data = stbi_load_from_memory(buf, (int)len, &nx, &ny, &nc, 3);
if(data == nullptr)
{
printf("\nError: MTMD image failed to decode bytes.");
return nullptr;
}
if(maxdims > 0 && (nx > maxdims || ny > maxdims))
{
const float aspect_ratio = static_cast<float>(nx) / ny;
int new_width = nx;
int new_height = ny;
if(aspect_ratio > 1.0f)
{
new_width = maxdims;
new_height = std::max(1, static_cast<int>(maxdims / aspect_ratio));
}
else
{
new_height = maxdims;
new_width = std::max(1, static_cast<int>(maxdims * aspect_ratio));
}
printf("\nImage requires resizing: original size %d x %d scaling to max %d px", nx, ny, maxdims);
uint8_t * resized_image = (uint8_t *)malloc((size_t)new_width * new_height * 3);
if(resized_image != nullptr && stbir_resize_uint8(data, nx, ny, 0, resized_image, new_width, new_height, 0, 3))
{
stbi_image_free(data);
data = resized_image;
nx = new_width;
ny = new_height;
printf("\nResized to clamped to %d x %d", nx, ny);
}
else
{
printf("\nWarning: MTMD image resize failed, using original image.");
free(resized_image);
}
}
const float maxaspect = 4.0f;
const float aspect_ratio = static_cast<float>(nx) / ny;
int out_width = nx;
int out_height = ny;
bool need_letterbox = false;
if(aspect_ratio > maxaspect)
{
out_height = std::max(1, static_cast<int>(nx / maxaspect));
need_letterbox = true;
}
else if(aspect_ratio < 1.0f / maxaspect)
{
out_width = std::max(1, static_cast<int>(ny / maxaspect));
need_letterbox = true;
}
mtmd_bitmap * bitmap = nullptr;
if(need_letterbox)
{
printf("\nImage requires letterboxing: %d x %d changed to %d x %d", nx, ny, out_width, out_height);
std::vector<uint8_t> letterboxed((size_t)out_width * out_height * 3, 0);
int offset_x = (out_width - nx) / 2;
int offset_y = (out_height - ny) / 2;
for(int y = 0; y < ny; ++y)
{
memcpy(
letterboxed.data() + ((y + offset_y) * out_width + offset_x) * 3,
data + y * nx * 3,
(size_t)nx * 3);
}
bitmap = mtmd_bitmap_init(out_width, out_height, letterboxed.data());
}
else
{
bitmap = mtmd_bitmap_init(nx, ny, data);
}
stbi_image_free(data);
return bitmap;
}
//this function prepares the mtmd chunks for media. it's only needed when media changes
static void PrepareMediaEmbds(const int nctx, const std::vector<int> & media_intro, const std::vector<int> & media_outro)
{
@@ -4396,12 +4487,15 @@ static void PrepareMediaEmbds(const int nctx, const std::vector<int> & media_int
int introsize = media_intro.size();
int outrosize = media_outro.size();
last_media_mem.clear();
last_media_pos_count = 0;
for(int i=0;i<media_objects.size();++i)
{
std::string media_obj = media_objects[i].b64data;
const std::vector<uint8_t> media_data_buffer = kcpp_base64_decode(media_obj);
mtmd::bitmap bitmap(mtmd_helper_bitmap_init_from_buf(mtmd_ctx, media_data_buffer.data(), media_data_buffer.size()));
mtmd::bitmap bitmap(media_objects[i].is_audio
? mtmd_helper_bitmap_init_from_buf(mtmd_ctx, media_data_buffer.data(), media_data_buffer.size())
: kcpp_mtmd_bitmap_init_image_from_buf(media_data_buffer.data(), media_data_buffer.size(), vision_max_res));
if(!bitmap.ptr)
{
printf("\nError: MTMD media %d failed to load!",i);
@@ -4423,6 +4517,8 @@ static void PrepareMediaEmbds(const int nctx, const std::vector<int> & media_int
}
int mediatokensneeded = 0;
int mediaposneeded = 0;
const int boundarytokensneeded = media_objects[i].chunk_start_seq.size() + media_objects[i].chunk_end_seq.size();
for(size_t j=0;j<chunks.size();++j)
{
const mtmd_input_chunk * mtmdchunk = chunks[j];
@@ -4430,29 +4526,39 @@ static void PrepareMediaEmbds(const int nctx, const std::vector<int> & media_int
chunk.is_audio = media_objects[i].is_audio;
chunk.mtmd_chunk = mtmd_input_chunk_copy(mtmdchunk);
chunk.clp_image_tokens = mtmd_input_chunk_get_n_tokens(mtmdchunk);
chunk.clp_image_positions = mtmd_input_chunk_get_n_pos(mtmdchunk);
mediatokensneeded += chunk.clp_image_tokens;
mediaposneeded += chunk.clp_image_positions;
media_objects[i].mediachunks.push_back(chunk);
}
if(debugmode==1 && !is_quiet)
{
printf("\nMTMD Media %i used Tokens: %d",i,mediatokensneeded);
printf("\nMTMD Media %i used Tokens: %d, Positions: %d, Boundary Tokens: %d",i,mediatokensneeded,mediaposneeded,boundarytokensneeded);
}
if(mediatokensneeded>0 && mediatokensneeded < nctx)
int mediactxneeded = std::max(mediatokensneeded, mediaposneeded) + boundarytokensneeded;
if(i==0)
{
int tokcnt = mediatokensneeded;
mediactxneeded += introsize + outrosize;
}
if(mediatokensneeded>0 && mediactxneeded < nctx)
{
int tokcnt = mediatokensneeded + boundarytokensneeded;
int poscnt = mediaposneeded + boundarytokensneeded;
if(i==0)
{
tokcnt += introsize + outrosize;
poscnt += introsize + outrosize;
}
for(int n=0;n<tokcnt;++n)
{
last_media_mem.push_back(current_media_identifier);
}
last_media_pos_count += poscnt;
}
else
{
media_composite_image_signature = ""; //force invalidate
printf("\nWarning: Media excluded - Context size too low or not enough mtmd tokens! (needed %d)\nMedia will be IGNORED! You probably want to relaunch with a larger context size!\n",mediatokensneeded);
printf("\nWarning: Media excluded - Context size too low or not enough mtmd tokens! (needed %d tokens, %d positions, %d boundary tokens)\nMedia will be IGNORED! You probably want to relaunch with a larger context size!\n",mediatokensneeded,mediaposneeded,boundarytokensneeded);
}
}
}
@@ -4644,6 +4750,10 @@ generation_outputs gpttype_generate(const generation_inputs inputs)
std::vector<int> media_intro; //added before media list
std::vector<int> media_outro; //added before media list
std::string intro = "\nAttached Media:\n";
if(mtmd_ctx && kcpp_mtmd_is_gemma4uv(mtmd_ctx)) //ugly fix for gemma4uv vision coherency
{
intro = "\n<|channel><channel|>" + intro;
}
TokenizeString(intro, media_intro, file_format, true);
//clear previous run media memory, just-in-time free
@@ -4678,6 +4788,7 @@ generation_outputs gpttype_generate(const generation_inputs inputs)
media_object lv;
lv.b64data = item;
lv.is_audio = false;
TokenizeString("\n\n", lv.chunk_end_seq, file_format, false);
media_objects.push_back(lv);
new_media_composite += item;
}
@@ -4690,6 +4801,7 @@ generation_outputs gpttype_generate(const generation_inputs inputs)
media_object lv;
lv.b64data = item;
lv.is_audio = true;
TokenizeString("\n\n", lv.chunk_end_seq, file_format, false);
media_objects.push_back(lv);
new_media_composite += item;
}
@@ -4886,6 +4998,7 @@ generation_outputs gpttype_generate(const generation_inputs inputs)
if(media_composite_image_signature=="")
{
last_media_mem.clear();
last_media_pos_count = 0;
}
if(media_data_changed)
{
@@ -4916,7 +5029,8 @@ generation_outputs gpttype_generate(const generation_inputs inputs)
if(last_media_mem.size()>0) //stick the media placeholders before the added mem
{
if(last_media_mem.size() + kcpp_data->n_predict + 4 > nctx)
int media_context_size = std::max((int)last_media_mem.size(), last_media_pos_count);
if(media_context_size + kcpp_data->n_predict + 4 > nctx)
{
printf("\nWarning: Too many multimodal tokens, max context exceeded! They will be ignored!\n");
}