feat: add ltx2.3 support (#1463)

* add GemmaTokenizer

* add basic ltx2.3 support

* change vocab file encoding

* fix ci

* fix ubuntu build

* add temporal tiling support

* add ltx audio support

* update ggml submodule url

* fix generate_video

* add i2v support

* minify bundled Gemma tokenizer vocab sources

* pass video fps into temporal rope embeddings

* fix av_ca_timestep_scale_multiplier

* add LTX2Scheduler support

* update docs

* fix ci
This commit is contained in:
leejet authored and GitHub committed 2026-05-17 16:46:20 +08:00
1 parent 3b4d26f3d9
commit 67dda3f897
44 files changed
+6393 -240

No files matched your search

+3 -1
View File
@@ -92,7 +92,9 @@ struct SDContextParams {
std::string llm_vision_path;
std::string diffusion_model_path;
std::string high_noise_diffusion_model_path;
std::string embeddings_connectors_path;
std::string vae_path;
std::string audio_vae_path;
std::string taesd_path;
std::string esrgan_path;
std::string control_net_path;
@@ -187,7 +189,7 @@ struct SDGenerationParams {
int video_frames = 1;
int fps = 16;
float vace_strength = 1.f;
sd_tiling_params_t vae_tiling_params = {false, 0, 0, 0.5f, 0.0f, 0.0f};
sd_tiling_params_t vae_tiling_params = {false, false, 0, 0, 0.5f, 0.0f, 0.0f};
std::string pm_id_images_dir;
std::string pm_id_embed_path;