One 21 GB ComfyUI diffusion-model file that does MiniMax-H3 text/image-to-video **and** reference-to-video in **4 steps**, with a distilled turbo and a motion-smoothing style LoRA already folded into the weights. - **Base:** the pruned `fl2va` transformer with a rank-1024 SVD of the (`ref2va` - `fl2va`) weight delta fused in, so a single partition serves both first/last-frame and reference conditioning (`xmarre/MiniMax-H3-Pruned-Ref-Delta-Fused-r1024-ComfyUI`, ComfyUI conversion of `diffusers-modular/MiniMax-H3-Pruned-Ref-Delta-Fused-r1024`). - **Merged in:** lightx2v FL2VA Turbo **8-step v1.0** at 1.0 (rank-24 resize from `Kijai/MiniMax-H3_comfy`) and **Mystic v2.0** at 0.7 (Civitai model 2856467, merged for motion smoothness). - **Quantization:** ONE pass after the merge. All four heavy Linear weights in each of the 50 blocks (`qkv_proj`, `out_proj`, `fc1`, `fc2`) are INT8 with ConvRot (group 256, per-channel) in ComfyUI's native `comfy_quant` layout; everything else stays BF16/F32. No custom loader: `UNETLoader`, weight_dtype `default`. `diffusion_models/minimax_h3_fused_refdelta_r1024_turbo8_mystic07_int8_convrot.safetensors` -> `ComfyUI/models/diffusion_models/`
번역 대기 중 · 원문 표시
분류 준비 중
입력 미확인→ 출력 미확인
A rank-16 LoRA that makes the de-rope pass in ComfyUI-MAINodes behave better on fast motion. It needs that pass; it is not a general-purpose motion LoRA. Watch the bottom two panels, not the stills. Without the adapter the fighters advance and snap, advance and snap, frame to frame. With it they keep moving. Rate 1.011 means the motion comes out at the speed the shot called for; 1.416 means the pass invented about 40 percent more movement than it should have. Two things this exhibit is not. It is not the adapter on its own: the stretched clock comes from the ComfyUI-MAINodes de-rope pass, and the adapter only changes how the model behaves once that pass has stretched it. Load this file on a plain text-to-video render and you get the costs listed below with none of the benefit. And the settings shown here are 0.75 / inject 0.30, the lower-invention cell; see "Settings that work" for when to prefer strength 1.0.
번역 대기 중 · 원문 표시
입력 이미지→ 출력 영상
An NVFP4 quantisation of the MiniMax H3 pruned FL2VA diffusion model for ComfyUI. Every one of the 200 block linears (qkv, out, fc1, fc2 in all 50 blocks) is stored as fp4 e2m1 with fp8 e4m3 block scales of 16 and one fp32 per-tensor scale; embeds, token refiner, norms and heads stay bf16. On Blackwell (sm120) the GEMMs run natively on the fp4 tensor cores. **Modification notice (required by the license): this repository contains a modified version of MiniMax H3. The modification is post-training weight quantisation of the block linears to NVFP4, performed 2026-08-19. Built from `diffusion_models/minimax_h3_fl2va_pruned_bf16.safetensors` in Comfy-Org/MiniMax-H3, revision `3f57e8291d2ef846f9a074b1b76d2767db434abe`.** Runs on stock ComfyUI, no custom nodes required: the file carries per-layer `comfy_quant` metadata, the same mechanism as the NVFP4 text encoder that Comfy-Org already ships.
번역 대기 중 · 원문 표시
분류 준비 중
입력 미확인→ 출력 미확인
Rank-32 LoRAs for MiniMax-H3 **Ref2VA**. Give it a photo, get a smooth clockwise 360-degree camera orbit around the frozen subject (`ORB360_CW`), or, with the CardSpin files, turn the photo in space like a thin physical card (`ORB360_CARDSPIN`). Both CardSpin files can still orbit with the orbit prompt, but they are noticeably softer; see Which file for a clean 360?. `minimax_h3_orb360_cardspin_v2_step50.safetensors`, one reference image (`examples/whitecat.png`), the prompt in `prompts/cardspin_caption.txt`, 1024 x 768, 124 frames, 20 steps, seed 20260926.
번역 대기 중 · 원문 표시
입력 이미지→ 출력 영상
The MOSS-Image **SFT** model often draws faint, evenly spaced **vertical color stripes** across smooth areas like skies, fog, backdrops and walls. This repo has two **banding-fix LoRAs** that remove them while keeping the image close to what SFT would have made, plus **preset workflows** that pair the fix with community speed LoRAs for 10–14 step (or 4–8 step) generation. > **Update, October 2026.** Since this repo went up, nynxz's Destripe LoRA (built on this investigation) has become the gentlest fix: it reduces the stripes while keeping images almost exactly as SFT made them. This update adds **DeBand Strong** (experimental: the strongest cleanup short of Base, with more drift), an **interactive comparison of every fix** (https://matlowai.github.io/MOSS-DeBand/, also in `web/` to run locally), the **full research write-up** (RESEARCH.md) including what didn't work, and the **measurement tools** (`research/`). Quick guide: minimal change → nynxz Destripe; cleanest flat backgrounds → DeBand Strong; one-file int8 → the int8 repo below. **Just want one file that works?** MOSS-Image-SFT-DeBand-int8 has the fix (and the Fast preset) merged into ready-to-run 5.4 GB int8 models. No LoRAs or custom nodes.
번역 대기 중 · 원문 표시
입력 텍스트→ 출력 이미지
MOSS-Image **SFT** with its **vertical color stripes ("banding") fixed**, as single files you load like any other model: no LoRAs, no custom nodes. Each file is 5.4 GB int8 (the same format as Comfy-Org's `int8_convrot` files) and loads with the stock **Load Diffusion Model** node. These are the merged, ready-to-run versions of MATLOWAI/MOSS-Image-SFT-DeBand. That repo has the LoRAs, the full results and the write-up. 1. Put the file in `ComfyUI/models/diffusion_models/`. 2. You also need the MOSS-Image text encoder and VAE from Comfy-Org/MOSS-Image: - `qwen3.5_9b_int8_convrot.safetensors` → `models/text_encoders` - `flux2-vae.safetensors` → `models/vae` 3. Drag `workflows/moss_deband_int8.json` into ComfyUI. Pick your file in **Model** and the same **Version** at the top; the version sets the steps.
번역 대기 중 · 원문 표시
입력 텍스트→ 출력 이미지
