Alex995647's picture
Refresh docs for H3 Realism People
83c1009 verified
Raw History Blame Contribute Delete
11.7 kB
═══════════════════════════════════════════════════════════════════════════════
H3 REALISM PEOPLE
═══════════════════════════════════════════════════════════════════════════════
───────────────────────────────────────────────────────────────────────────────
IDENTITY
───────────────────────────────────────────────────────────────────────────────
Base model MiniMax H3
Generation mode Text-to-Video (T2VA), Image-to-Video (I2VA, first frame), Reference-to-Video (Ref2VA)
Checkpoint both families β€” see the author's builds
Category Photorealism
Author fal
Source https://huggingface.co/fal/MiniMax-H3-Realism-People-LoRA
Provenance HuggingFace Β· fal/MiniMax-H3-Realism-People-LoRA Β· h3-realism-people-t2v-i2v-r2v.safetensors
───────────────────────────────────────────────────────────────────────────────
WEIGHTS
───────────────────────────────────────────────────────────────────────────────
File h3-realism-people-t2v-i2v-r2v.safetensors
Version main
Size 125.2 MB
Format SafeTensor / unknown precision
SHA256 acc529601d2da117fb81179e76c56e488a3beab1171659d305f04fa3655b787e
Published 2026-08-12
───────────────────────────────────────────────────────────────────────────────
TRIGGER WORDS
───────────────────────────────────────────────────────────────────────────────
β€Ί "r34l1sm"
───────────────────────────────────────────────────────────────────────────────
RECOMMENDED SETTINGS
───────────────────────────────────────────────────────────────────────────────
The author did not publish specific settings.
Start at strength 0.7 and adjust; see REFERENCE.txt for the MiniMax H3
community baseline.
───────────────────────────────────────────────────────────────────────────────
ADOPTION & POPULARITY
───────────────────────────────────────────────────────────────────────────────
Downloads 53,430
Rating 100% positive (346/346)
Comments 0
Published 23 days ago
Download rate 2,312.1/day since release
Adoption rank #3 of 113 in this collection (Top tier)
Momentum rank #3 of 113 by download rate (Hot Β· fresh release)
Adoption tier is the percentile of total downloads within this collection;
momentum is the percentile of downloads-per-day since release. Momentum is
ranked rather than measured against a fixed rate, because this base model
is itself new and every LoRA looks fast on an absolute scale. Both numbers
are CivitAI's, read on the scrape date at the foot of this file.
───────────────────────────────────────────────────────────────────────────────
CHAINING / STACKING
───────────────────────────────────────────────────────────────────────────────
Stacks with others not stated
The author gave no chaining guidance. The rules below are the
general MiniMax H3 ones, not anything specific to this LoRA.
General MiniMax H3 rules that apply here (see REFERENCE.txt):
β€’ Load the LoRA with LoraLoaderModelOnly between UNETLoader and your
sampler; stacking means chaining several of those nodes in series.
β€’ Stacks normally on top of a single acceleration/turbo LoRA; keep only
one distill LoRA active in the chain.
β€’ The author ships builds for both checkpoint families β€” pair each build
with its own checkpoint and never cross-pair them.
───────────────────────────────────────────────────────────────────────────────
PROS
───────────────────────────────────────────────────────────────────────────────
+ Widely adopted β€” 53,430 downloads. [source: CivitAI stats]
+ Positively received β€” 346 thumbs-up with no down-votes. [source:
CivitAI stats]
+ Explicit trigger word(s) β€” r34l1sm β€” so the effect can be turned on and
off from the prompt. [source: model metadata]
+ Author documents use across several modes (T2V, I2V, REF2V). [source:
description]
+ Compact at 125 MB β€” cheap to stack with other LoRAs. [source: file
size]
───────────────────────────────────────────────────────────────────────────────
CONS & CAVEATS
───────────────────────────────────────────────────────────────────────────────
(none recorded)
───────────────────────────────────────────────────────────────────────────────
COMPATIBILITY WARNING
───────────────────────────────────────────────────────────────────────────────
MiniMax H3 ships two transformers that share tensor key names: the fl2va
checkpoint (driving T2VA, I2VA, L2VA and FL2VA) and the separate ref2va
checkpoint (driving Ref2VA). Loading a LoRA against the wrong family can
silently degrade output instead of raising an error, so confirm the family
above matches the checkpoint you have loaded. Likewise, LoRAs built for
full checkpoints often throw adaln_proj shape errors on pruned ones β€” match
the build to your checkpoint.
───────────────────────────────────────────────────────────────────────────────
AUTHOR'S DESCRIPTION
───────────────────────────────────────────────────────────────────────────────
A rank-32 LoRA adapter for MiniMax H3 specialised in realistic people: skin
that keeps its texture instead of smoothing out, coherent eyes and micro-
expressions, film-set lighting and a subtle handheld camera quality, while
keeping H3's native synchronized audio. Trained for 1500 steps at the high-
resolution bucket on 176 hand-curated live-action clips, and validated with
19 same-prompt same-seed adapter-on/off pairs. Because it only touches the
shared attention projections (standard H3 fused-QKV layout,
diffusion_model.blocks.N.attn.qkv_proj), it loads in ComfyUI as-is and
works across text-, image- and reference-to-video. Successor to
fal/MiniMax-H3-Realism-LoRA.
───────────────────────────────────────────────────────────────────────────────
EXAMPLE MEDIA
───────────────────────────────────────────────────────────────────────────────
1 preview(s) in ./example_images/ β€” community results for
this LoRA. Videos keep a matching _poster.jpg still frame.
No prompts are recorded below: CivitAI's public API flags these posts as
having generation metadata but does not return it, so the prompts cannot be
scraped. Open the model page to read them.
01_video.mp4 (video, )
still: 01_poster.jpg
───────────────────────────────────────────────────────────────────────────────
HOW TO USE (COMFYUI)
───────────────────────────────────────────────────────────────────────────────
1. Copy the .safetensors file into ComfyUI/models/loras/
2. Load the matching H3 checkpoint with UNETLoader.
3. Insert LoraLoaderModelOnly between the loader and your sampler.
4. Set strength as listed above, then add any trigger word to the prompt.
5. Keep CFG at 1.0 with BasicGuider unless you know you want real CFG.
───────────────────────────────────────────────────────────────────────────────
Scraped by lorakit on 2026-09-04 23:23 UTC.
Stats and description are the author's; pros/cons are derived from the
evidence tagged beside each line.
───────────────────────────────────────────────────────────────────────────────