Search results

LoRA trained on base model vs distilled model

Status: Open.

#175 In Lightricks/LTX-2;

· sarlea1 opened on Mar 24, 2026

Feature Request : LTX-2 trainer Captioning tool

Status: Open.

#174 In Lightricks/LTX-2;

· seanhan19911990-source opened on Mar 23, 2026

ltx-trainer: int4-quanto can be much slower than int2-quanto for LoRA training on CUDA/bf16

Status: Open.

#173 In Lightricks/LTX-2;

· Pyro (pyros-projects) opened on Mar 22, 2026

Controlling audio generation using batch processing Image to Video

Status: Open.

#171 In Lightricks/LTX-2;

· jlitz opened on Mar 20, 2026

been working on a bunch of upgrades - Seperable Causal Diffusion / Variational Flow Maps - 1 shot distillation / streaming

Status: Open.

#168 In Lightricks/LTX-2;

· John D. Pope (johndpope) opened on Mar 19, 2026

1st test: the trainer is working under Windows 11. Any additional optimation suggestion?

Status: Open.

#167 In Lightricks/LTX-2;

· zappazack opened on Mar 18, 2026

Is it possible to achieve audio + video to video?

Status: Open.

#166 In Lightricks/LTX-2;

· Li Muyi (Limuyi-col) opened on Mar 18, 2026

LTx2.3 train output noise, pls help

Status: Open.

#165 In Lightricks/LTX-2;

· GuardSkill (GuardSkill) opened on Mar 16, 2026

Feature Request: Support for smaller text encoders (Gemma 3 4B / quantized Gemma) for <32GB VRAM GPUs for those not using ComfyUI (instead CLI or python code)

Status: Open.

#164 In Lightricks/LTX-2;

· Nick Brown (BossLevel) opened on Mar 15, 2026

Feature Request: Multi-GPU inference for a single ti2vid_two_stages job

Status: Open.

#162 In Lightricks/LTX-2;

· 余嘉軒 (cvecve147) opened on Mar 13, 2026

for those who's already download LTX-2.3 full models...

Status: Open.

#161 In Lightricks/LTX-2;

· ericssonruffy opened on Mar 12, 2026

Backward compatibility of ltx-2-19b checkpoints with current code?

Status: Open.

#160 In Lightricks/LTX-2;

· antoine-scenario opened on Mar 11, 2026