Free, open-source AI video generator for Mac. Run LTX-2.5 and MiniMax H3 (Hailuo) locally on Apple Silicon: text-to-video and image-to-video with sound, lip-sync, character LoRA training, songs from lyrics (YuE2) and image editing (Qwen-Image-Edit). One-click install via Pinokio. Offline: no cloud, no API key, no NVIDIA GPU.
Python ai-filmmakingai-videoai-video-generatorapple-siliconhailuoimage-to-videolip-synclocal-ailora-trainingltx-2ltx-2-5ltx-videomacosminimax-h3mlxmusic-generationoffline-aipinokiotext-to-videovideo-generation
fix(install): demote one-stage fps patch to advisory on upstream drift
mrbizarro7e1b74366c38
diff --git a/patch_ltx_codec.py b/patch_ltx_codec.pyindex 2a27482..3a37254 100644--- a/patch_ltx_codec.py+++ b/patch_ltx_codec.py@@ -561,7 +561,24 @@ def main() -> int:# shippable hook for the panel's explicit Long Clip Boost mode: generate# fewer semantic frames at 12fps, keep LTX audio duration aligned, then# interpolate the delivered export back to 24fps.+ #+ # Treat upstream drift as ADVISORY (same convention as patches 2/3/4):+ # when dgrauet refactors `ti2vid_one_stage.py`, the anchor disappears+ # and a hard install failure here would block every new user even+ # though the codec patch (the only one that actually affects T2V/I2V+ # correctness) applied. Without this patch, generation still works at+ # the engine's native frame rate; only the panel's explicit "Long Clip+ # Boost (12→24fps)" mode is unavailable until the patch is updated.fps_outcome = apply_one_stage_fps_patch(i2v_target)+ if fps_outcome in (OUTCOME_DRIFT, OUTCOME_MISSING):+ print(+ " [one-stage frame_rate] note: anchor not found — pipeline may "+ "have been refactored upstream. T2V / I2V at native fps work fine; "+ "the panel's Long Clip Boost (12→24fps interpolation) mode is "+ "unavailable until the patch is updated.",+ file=sys.stderr,+ )+ fps_outcome = OUTCOME_ALREADYoutcomes.append(("one-stage frame_rate (12→24fps long clips)", fps_outcome))# (Keyframe OOM patch was removed — see NOTE in the patches block above.