Downloads · 30 days
4.8K
22% of all-time downloads
berryber09/MiniMax-H3-ref2va-fl2va-hybrid-w4a8
MiniMax-H3-ref2va-fl2va-hybrid-w4a8 is a image-to-video model from berryber09. Use it for the image-to-video task on the model card, and read the license before you ship it in a product. It is set up for minimax-h3. The card lists the license as other.
A 4-bit (w4a8mixed) block-splice hybrid of MiniMax-H3 that combines the first/last-frame keyframe model (fl2va) with the reference-conditioning model (ref2va) in a single checkpoint — so one gen can be driven by a fir…
Downloads · 30 days
4.8K
22% of all-time downloads
All-time downloads
21.8K
Public
Repo size
12.5 GB
Likes
14
Trending 2
Click a slice to open those files.
.safetensors12.5 GB · 100%
From the Hugging Face model README
A 4-bit (w4a8_mixed) block-splice hybrid of MiniMax-H3 that combines the first/last-frame keyframe
model (fl2va) with the reference-conditioning model (ref2va) in a single checkpoint — so one gen can
be driven by a first frame and use reference images at the same time.
~12.5 GB on disk — 4-bit (w4a8_mixed); fits a 16 GB card with headroom.
fl2va — first/last-frame keyframe conditioning, higher visual/audio quality. No reference slots.ref2va — image/video/audio reference conditioning (subject/scene/style). Lower base quality.Neither alone can do "open on this frame, and have this person walk in later." The hybrid can: the first frame stays a hard temporal anchor while the reference supplies a subject's identity that the model can introduce anywhere in the clip.
b20-49)Whole-block splice of the two official-derived w4a8 checkpoints:
| Part | Source |
|---|---|
| Input/output heads + all non-block tensors | fl2va |
| Transformer blocks 0–19 | fl2va |
| Transformer blocks 20–49 | ref2va |
b20-49 leans toward reference adherence (30 of 50 blocks from ref2va) at a small cost to base
quality — a good pick when the reference is mandatory. The quality↔reference spectrum, most-quality →
most-reference: b30-49 → b25-49 → b20-49 → b15-49.
Per-layer quant metadata is merged correctly: fl2va's _quantization_metadata is kept only for the
fl2va-sourced layers; the ref2va blocks carry their own comfy_quant tensors. Verified to load in ComfyUI
(Detected mixed precision quantization).
Place in ComfyUI/models/diffusion_models/. It routes as a ref2va model, so both paths are active:
MiniMaxH3AddKeyframes) — hard-anchors that frame.MiniMaxH3ReferenceToVideo — identity/style conditioning, not frame-locked.Both can be supplied in one generation.
Inherits the MiniMax-H3 license.