Add LTX2.3-IC-LORA-Dual-Character

.gitattributes

@@ -58,3 +58,10 @@ ltx-2-3-enhancers/example_images/03_video.mp4 filter=lfs diff=lfs merge=lfs -tex
 ltx-2-3-enhancers/example_images/04_video.mp4 filter=lfs diff=lfs merge=lfs -text
 ltx-2-3-enhancers/example_images/05_video.mp4 filter=lfs diff=lfs merge=lfs -text
 ltx-2-3-enhancers/example_images/06_video.mp4 filter=lfs diff=lfs merge=lfs -text
+ltx2-3-ic-lora-dual-character/example_images/01_video.mp4 filter=lfs diff=lfs merge=lfs -text
+ltx2-3-ic-lora-dual-character/example_images/02_video.mp4 filter=lfs diff=lfs merge=lfs -text
+ltx2-3-ic-lora-dual-character/example_images/03_poster.jpg filter=lfs diff=lfs merge=lfs -text
+ltx2-3-ic-lora-dual-character/example_images/03_video.mp4 filter=lfs diff=lfs merge=lfs -text
+ltx2-3-ic-lora-dual-character/example_images/04_video.mp4 filter=lfs diff=lfs merge=lfs -text
+ltx2-3-ic-lora-dual-character/example_images/05_video.mp4 filter=lfs diff=lfs merge=lfs -text
+ltx2-3-ic-lora-dual-character/example_images/06_video.mp4 filter=lfs diff=lfs merge=lfs -text

ltx2-3-ic-lora-dual-character/LTX2.3-IC-LORA-Dual-Character.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6c3199e2c95eb0aad0cea4f3e8dfa33e8f6966d85bc79e3994faf5a57406103
+size 327287347

ltx2-3-ic-lora-dual-character/example_images/01_video.mp4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:10a7439d4f3ceae0f535bdc79df8c0fa918cf69cec0d8114894e0467dd0111f7
+size 3273317

ltx2-3-ic-lora-dual-character/example_images/02_video.mp4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:09d5145fe68a7402be726852ea6501c27bc2b723a29ae5d9b1d60b006b9d002c
+size 6949922

ltx2-3-ic-lora-dual-character/example_images/03_video.mp4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b25bb538b204851dd17aabe70f030b3ff9e91f62e18136695fd132f399f3addc
+size 38951960

ltx2-3-ic-lora-dual-character/example_images/04_video.mp4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bbbab13e54fa4b2419b153a40de9d59eb2e960eb56b76059ab336f1b965d8386
+size 15516660

ltx2-3-ic-lora-dual-character/example_images/05_video.mp4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:576a616105e0473261cf8bf102cc695639ea2e8b790507d78a229f184ebaf4fd
+size 11031961

ltx2-3-ic-lora-dual-character/example_images/06_video.mp4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:657cf6dc1aca71ca3ca8ce9dcf668a2e4906a114178eac075cc47671fd4b485f
+size 12083657

ltx2-3-ic-lora-dual-character/info.txt

@@ -0,0 +1,369 @@
+═══════════════════════════════════════════════════════════════════════════════
+  LTX2.3-IC-LORA-DUAL-CHARACTER
+═══════════════════════════════════════════════════════════════════════════════
+
+───────────────────────────────────────────────────────────────────────────────
+  IDENTITY
+───────────────────────────────────────────────────────────────────────────────
+  Base model        LTX-2.3
+  Generation mode   IC-LoRA / Video-to-Video (V2V), Text-to-Video (T2V), Image-to-Video (I2V)
+  Checkpoint        ltx-2.3-22b-dev (full / base)
+  Category          Character / subject
+  Author            MaqueAI
+  Source            https://civitai.com/models/2500098
+  Provenance        CivitAI · model 2500098 · version 2810376
+
+───────────────────────────────────────────────────────────────────────────────
+  WEIGHTS
+───────────────────────────────────────────────────────────────────────────────
+  File              LTX2.3-IC-LORA-Dual-Character.safetensors
+  Version           v0.1
+  Size              312.1 MB
+  Format            SafeTensor / unknown precision
+  SHA256            b6c3199e2c95eb0aad0cea4f3e8dfa33e8f6966d85bc79e3994faf5a57406103
+  Published         2026-03-28
+
+───────────────────────────────────────────────────────────────────────────────
+  TRIGGER WORDS
+───────────────────────────────────────────────────────────────────────────────
+  No trigger word. The LoRA applies as soon as it is loaded —
+  strength is the only way to dial the effect up or down.
+
+───────────────────────────────────────────────────────────────────────────────
+  RECOMMENDED SETTINGS
+───────────────────────────────────────────────────────────────────────────────
+  Taken verbatim from the author's description:
+
+  Resolution          1280x720
+
+───────────────────────────────────────────────────────────────────────────────
+  ADOPTION & POPULARITY
+───────────────────────────────────────────────────────────────────────────────
+  Downloads               5,230
+  Rating                  100% positive (204/204)
+  Comments                0
+  Published               160 days ago
+  Download rate           32.6/day since release
+
+  Adoption rank           #6 of 150 in this collection  (Top tier)
+  Momentum rank           #5 of 150 by download rate  (Hot)
+
+  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 LTX-2.3 ones, not anything specific to this LoRA.
+
+  General LTX-2.3 rules that apply here (see REFERENCE.txt):
+
+  • Stacking an acceleration LoRA with a task LoRA is officially
+    sanctioned: the ID-LoRA template chains LoraLoaderModelOnly(ltx_2.3_22b
+    _distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16, 0.5) into
+    LoraLoaderModelOnly(ltx-2.3-id-lora-talkvid-3k, 1.0) on the dev
+    checkpoint.
+  • The Python CLI exposes `--lora <path> [strength]` as a repeatable flag
+    with default strength 1.0, so multiple adapters at once is a supported
+    configuration, not a hack.
+  • Load the acceleration/distilled LoRA on the dev checkpoint only. The
+    two-stage pipelines require it for the full model but explicitly do not
+    use it for DistilledPipeline, ICLoraPipeline or DubItPipeline, which
+    already run distilled weights.
+  • Do not stack two distillation LoRAs (e.g. the official rank-384 /
+    rank-111 one plus a community DMD LoRA). The DaSiWa DMD LoRA already
+    blends 75-80% of the official distilled delta into its audio-output
+    route and is described by its author as recreating the distilled model
+    on top of stock dev, i.e. a replacement.
+  • IC-LoRAs are not chained like ordinary LoRAs: each carries a
+    reference_downscale_factor in its safetensors metadata and needs a node
+    that reads it (GetICLoRAParameters -> LTXVAddGuide in core, or
+    LTXICLoRALoaderModelOnly + LTXAddVideoICLoRAGuide in ComfyUI-LTXVideo).
+    Without one the factor silently defaults to 1.
+  • Two IC-LoRA reference guides can be driven from a single IC-LoRA: the
+    CrossView Warp author wires both a depth-warp video and the original
+    video into separate reference guides and sets latent_downscale_factor =
+    1 on both.
+  • DubItPipeline accepts exactly one Dub-It IC-LoRA (`--lora` exactly
+    once) and applies it in both stages — that path is single-adapter by
+    design.
+  • Community convention on strengths when stacking: the IC-LoRA Dual-
+    Character author recommends 0.6-1.0 for a content LoRA used alone,
+    dropped to 0.3-0.5 when combined with others.
+  • Some LoRA pairs are designed to be blended against each other — the LTX
+    2.3 Crisp Enhance / Soft Enhance pair is described by its author as
+    "can be used together and balanced at different strengths".
+  • The gemma-3-12b-it-abliterated LoRA is not a video LoRA and does not
+    belong in the DiT chain — it patches the Gemma text encoder through a
+    model+clip LoraLoader, as the official templates wire it.
+  • The full, flexible, trainable bf16 model — ltx-2.3-22b-dev.safetensors
+    is 46.15 GB bf16, 29.15 GB as ltx-2.3-22b-dev-fp8.safetensors. This is
+    what LoRAs are trained against, and what four of the six official
+    ComfyUI LTX-2.3 templates load (t2v, i2v, ia2v, id_lora). Needs either
+    real guidance (Lightricks' own single-stage workflow sets LTXVScheduler
+    to 15 steps with MultimodalGuider) or the distill
+
+───────────────────────────────────────────────────────────────────────────────
+  PROS
+───────────────────────────────────────────────────────────────────────────────
+  + Widely adopted — 5,230 downloads. [source: CivitAI stats]
+  + Positively received — 204 thumbs-up with no down-votes. [source:
+    CivitAI stats]
+  + Author documents use across several modes (V2V_IC_LORA, T2V, I2V).
+    [source: description]
+  + Compact at 312 MB — cheap to stack with other LoRAs. [source: file
+    size]
+  + Author published concrete recommended settings. [source: description]
+
+───────────────────────────────────────────────────────────────────────────────
+  CONS & CAVEATS
+───────────────────────────────────────────────────────────────────────────────
+  - No trigger word — the LoRA is always on once loaded, so strength is the
+    only control. [source: model metadata]
+
+───────────────────────────────────────────────────────────────────────────────
+  COMPATIBILITY WARNING
+──────────────────────────────────────────────────────────────────────��────────
+  Within LTX-2.3, dev vs distilled is low risk: they are the same
+  architecture and the distillation itself is published as a LoRA that the
+  model card calls "applicable to the full model", so ordinary content LoRAs
+  load on both. Expect style/strength drift rather than breakage; the
+  distilled LoRA is meant to be applied to a dev checkpoint, not stacked on
+  an already-distilled one (the Python pipelines that start from a distilled
+  checkpoint — DistilledPipeline, ICLoraPipeline, DubItPipeline — explicitly
+  do not take --distilled-lora). The genuine risk is cross-version and cross-
+  format, and it is silent rather than a hard error: the LTX-2 repo README
+  states flatly that files "are not interchangeable between the two models,
+  and a LoRA only works with the model it was trained on", and a wrong-
+  version LoRA surfaces only as a stream of "lora key not loaded:
+  diffusion_model...." console warnings while generation proceeds as if no
+  LoRA were attached. The same silent no-op hits correctly-versioned LoRAs
+  saved in a non-ComfyUI key layout: SimpleTuner issue #2349 shows keys like
+  diffusion_model.connectors.audio_connector.transformer_blocks.N.attn1.to_q
+  going unloaded, where the LTX-2 trainer's own layout is attn1/attn2 for
+  video, audio_attn1/audio_attn2 for audio and audio_to_video_attn /
+  video_to_audio_attn for cross-modal. Community finetunes (Sulphur 2, JoyAI-
+  Echo merges) load stock LTX-2.3 LoRAs but shift the result. And an IC-LoRA
+  loaded without a metadata-reading node loses its reference scale: ComfyUI's
+  GetICLoRAParameters reads reference_downscale_factor out of the LoRA's
+  safetensors metadata and silently falls back to 1 when it is absent, so a
+  ref0.5 IC-LoRA wired straight into LTXVAddGuide without it feeds the
+  reference at the wrong latent scale.
+
+───────────────────────────────────────────────────────────────────────────────
+  AUTHOR'S DESCRIPTION
+───────────────────────────────────────────────────────────────────────────────
+  ## 中文说明
+  LTX-Video (2.3) IC-LoRA: 双人分镜头对话增强模型
+  本模型是基于 Lightricks LTX-2.3 底模训练的 IC-LoRA，专为双人同框对话、角色互动及分镜头视频生成场景深度优化。
+  ---
+  ### 一、 模型核心提升
+  1. 角色参考稳定性：显著提升双人同框时的人物特征一致性，减少角色漂移。
+  2. 分镜构图稳定性：针对影视化对话构图进行了加固，支持更精准的镜头控制。
+  3. 叙事连贯性：增强了对多段描述的理解力，使分镜间的过渡衔接更自然。
+  4. 风格兼容性：完美支持古风仙侠、现代都市、3D 动漫等主流视觉风格。
+  ---
+  ### 二、 模型基本信息
+  1. 基础模型：Lightricks/LTX-2.3
+  2. 许可证：Apache-2.0
+  3. 管道标签：image-to-video, text-to-video
+  4. 模型用途：仅供学习交流使用
+  5. 开发者：麻雀 AI
+  ---
+  ### 三、 运行指南
+  1. 推荐平台：ComfyUI
+  2. 支持工作流：ComfyUI 官方 LTX 工作流、KJ-LTX 插件工作流
+  3. 生成模式：文生视频 (T2V) 与 图生视频 (I2V) 均支持
+  4. 硬件参考：RTX 5090 显卡在 720P 分辨率下，单条视频生成耗时约 2 分钟
+  ---
+  ### 四、 推荐参数配置
+  1. 分辨率：建议使用 16:9 (如 1280x720)
+  2. 时长与帧率：建议时长 ≥10 秒，帧率设定为 24 FPS
+  3. LoRA 权重设定：
+  - 独立使用建议：0.6 - 1.0
+  - 叠加其他 LoRA 使用时建议：0.3 - 0.5
+  ---
+  ### 五、 Prompt 编写规范
+  1. 编写逻辑：需包含完整的场景描述 + 角色设定 + 分镜设计 + 镜头语言，强化双人对话互动逻辑。
+  2. 触发词说明：无需特定触发词。
+  3. 参考提示词示例（古风）：
+  [场景] 古风仙侠，云端仙境，粉紫霞光与翻涌云海交织，暖金日光穿透云层洒落，仙气缥缈朦胧，氛围梦幻唯美，尽显东方仙侠的空灵意境。
+  [角色] 女子：古典发髻挽起，点缀银质花叶发饰；身着天蓝色轻纱广袖裙，裙身绣有鎏金莲花纹样，肩带与袖摆薄如蝉翼，手持琉璃净瓶，瓶中盛有清透仙水，宛如瑶
+  池临凡的仙子。男子：墨色长发束起，银质发冠点缀；身着月白色广袖仙袍，袖口与衣摆绣有银线云纹，手持白玉柄拂尘，尘丝垂落随风轻晃，仙风道骨，仿若天界执掌时
+  序的仙君。
+  分镜 1 (全景 3 秒)
+  俯拍全景。云海翻涌如浪，粉紫霞光铺满天际，暖金日光斜洒而下，两人并肩立于云端，衣袂随仙气轻扬，与云雾霞光融为一体，尽显仙境的辽阔与缥缈。
+  分镜 2 (中景 4 秒) 正面中景。镜头缓缓推近��女子微微侧身，指尖轻托琉璃净瓶，瓶中仙水泛着细碎微光，她抬眸望向身旁男子，眉眼含柔，轻声道：“仙君
+  ，今日凡间春汛已平，这净瓶仙水可滋养人间草木，唤回生机。”
+  分镜 3 (近景 3 秒) 侧拍近景。男子手持拂尘，目光落在女子手中的净瓶上，语气平和温润：“有劳仙子费心，人间得此仙泽，定能岁岁繁茂，不负春日。”
+  分镜 4 (特写 2 秒)
+  特写镜头。聚焦女子手中的琉璃净瓶，瓶身通透如冰，仙水在光下流转泛着莹光，裙身鎏金莲花纹样若隐若现，与瓶中微光相映，细节尽显东方仙意的精致。
+  分镜 5 (全景 4 秒)
+  拉远全景。镜头缓缓拉远，两位仙子的身影渐渐融入云海霞光之中，拂尘尘丝与净瓶微光在云雾中闪烁，最终化作天际一点流光，只留缥缈仙踪于天地间。
+  ---
+  ### 六、 效果说明与局限性
+  1. 优势风格：在古风、现代、3D 动漫类双人对话场景中表现最佳。
+  2. 已知限制：受限于 LTX-2.3 底模性能，极其复杂的双人肢体互动（如缠绕、打斗）可能出现形变。
+  3. 运动幅度：建议以对话和微动作为主，大动态动作的连贯性仍有提升空间。
+  ## English Description
+  This model is an IC-LoRA trained based on the Lightricks LTX-2.3 base
+  model, specifically optimized for two-person dialogue, character
+  interaction, and multi-shot cinematic video generation.
+  ---
+  ### 1. Core Enhancements
+  1. Character Consistency: Significantly improves character feature
+  stability when two people are in the same frame, reducing character drift.
+  2. Compositional Stability: Reinforced for cinematic dialogue compositions,
+  supporting more precise camera control.
+  3. Narrative Coherence: Enhanced understanding of multi-segment
+  descriptions, making transitions between shots more natural.
+  4. Style Compatibility: Perfectly supports various styles including Ancient
+  Chinese Fantasy, Modern Urban, and 3D Anime.
+  ---
+  ### 2. Basic Information
+  1. Base Model: Lightricks/LTX-2.3
+  2. License: Apache-2.0
+  3. Pipeline Tags: image-to-video, text-to-video
+  4. Usage: For educational and exchange purposes only
+  5. Developer: Maque AI (麻雀 AI)
+  ---
+  ### 3. Running Guide
+  1. Recommended Platform: ComfyUI
+  2. Supported Workflows: ComfyUI Official LTX Workflow, KJ-LTX Plugin
+  Workflow
+  3. Generation Mode: Supports both Text-to-Video (T2V) and Image-to-Video
+  (I2V)
+  4. Hardware Reference: On an RTX 5090 at 720P resolution, a single video
+  takes approximately 2 minutes to generate.
+  ---
+  ### 4. Recommended Parameters
+  1. Resolution: 16:9 (e.g., 1280x720) is recommended.
+  2. Duration & FPS: Duration ≥ 10s, Frame Rate set to 24 FPS.
+  3. LoRA Weight:
+  - Standalone use: 0.6 - 1.0
+  - When stacked with other LoRAs: 0.3 - 0.5
+  ---
+  ### 5. Prompting Guide
+  1. Writing Logic: Prompts should include a complete Scene Description +
+  Character Settings + Storyboard Design + Camera Language to reinforce
+  dialogue logic.
+  2. Trigger Words: No specific trigger word required.
+  3. Reference Example (Ancient Chinese Fantasy):
+  - [Scene] Ancient Chinese fantasy, celestial realm above clouds, pink and
+  purple sunset glow blending with rolling clouds, warm golden sunlight
+  piercing through, ethereal and dreamy atmosphere.
+  - [Characters] Female: Classical bun hairstyle, silver floral hairpins,
+  wearing a sky-blue translucent flowing gown with gold lotus patterns,
+  holding a glazed jade bottle. Male: Black hair tied up with a silver crown,
+  wearing a moon-white wide-sleeved immortal robe with silver cloud
+  embroidery, holding a white jade whisk.
+  - Shot 1 (Wide Shot, 3s): Overhead view. Two immortals stand side by side
+  above the clouds, robes fluttering in the mist, showing the vastness of the
+  celestial realm.
+  - Shot 2 (Medium Shot, 4s): Camera slowly pushes in. The female character
+  turns slightly, looking at the male with gentle eyes, whispering: "The
+  spring floods in the mortal world have subsided; this celestial water can
+  restore life to the land."
+  - Shot 3 (Side Medium-Close Shot, 3s): The male character looks at the
+  bottle in her hand, replying warmly: "Thank you for your kindness. With
+  this blessing, the mortal world will surely flourish."
+  - Shot 4 (Close-up, 2s): Focus on the glazed jade bottle. The body is
+  translucent like ice, with light reflecting off the water inside and the
+  golden lotus patterns on the dress flickering in the background.
+  - Shot 5 (Wide Shot, 4s): Camera pulls back. The two figures gradually
+  merge into the sunset glow and clouds, eventually turning into a streak of
+  light in the sky.
+  ---
+  ### 6. Limitations & Notes
+  1. Best Styles: Performs best in Ancient Fantasy, Modern, and 3D Anime
+  dialogue scenes.
+  2. Known Limitations: Due to the constraints of the LTX-2.3 base model,
+  extremely complex physical interactions (e.g., wrestling or hugging) may
+  cause deformation.
+  3. Motion Range: Recommended for dialogue and subtle movements; high-
+  intensity action consistency still has room for improvement.
+  ---
+
+───────────────────────────────────────────────────────────────────────────────
+  VERSION NOTES
+───────────────────────────────────────────────────────────────────────────────
+  test0.1
+
+───────────────────────────────────────────────────────────────────────────────
+  EXAMPLE MEDIA
+───────────────────────────────────────────────────────────────────────────────
+  6 preview(s) in ./example_images/ — 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, 1280x704)
+      still: 01_poster.jpg
+  02_video.mp4  (video, 880x560)
+      still: 02_poster.jpg
+  03_video.mp4  (video, 1280x704)
+      still: 03_poster.jpg
+  04_video.mp4  (video, 880x560)
+      still: 04_poster.jpg
+  05_video.mp4  (video, 1024x704)
+      still: 05_poster.jpg
+  06_video.mp4  (video, 1024x704)
+      still: 06_poster.jpg
+
+───────────────────────────────────────────────────────────────────────────────
+  HOW TO USE (COMFYUI)
+───────────────────────────────────────────────────────────────────────────────
+  1. Update ComfyUI. LTX-2.3 is supported in ComfyUI core
+  (comfy_extras/nodes_lt.py provides EmptyLTXVLatentVideo, LTXVAddGuide,
+  GetICLoRAParameters, LTXVConditioning, LTXVScheduler,
+  LTXVConcatAVLatent/LTXVSeparateAVLatent, LTXVReferenceAudio and more) and
+  the six official templates need no custom nodes; install the Lightricks
+  "ComfyUI-LTXVideo" custom node pack only if you want IC-LoRA loaders, low-
+  VRAM loaders, HDR decode, audio-only nodes or Q8 nodes.
+  2. Put the base checkpoint in ComfyUI/models/checkpoints/ —
+  ltx-2.3-22b-dev-fp8.safetensors (29.15 GB) or ltx-2.3-22b-distilled-
+  fp8.safetensors (29.53 GB) from Lightricks/LTX-2.3-fp8, or the bf16
+  originals (46.15 GB) from Lightricks/LTX-2.3. It is a single file that also
+  carries the video VAE, the audio VAE and the text projection.
+  3. Put gemma_3_12B_it_fp4_mixed.safetensors in
+  ComfyUI/models/text_encoders/ (from Comfy-Org/ltx-2). The Python package
+  instead wants the full google/gemma-3-12b-it-qat-q4_0-unquantized folder.
+  4. Put ltx-2.3-spatial-upscaler-x2-1.1.safetensors in
+  ComfyUI/models/latent_upscale_models/ (needed by every two-stage template),
+  and moge_2_vitl_normal_fp16.safetensors (Comfy-Org/MoGe) in
+  ComfyUI/models/geometry_estimation/ if you use the IC-LoRA depth path.
+  5. Put every LoRA .safetensors in ComfyUI/models/loras/. Subfolders are
+  fine — Lightricks' own workflows reference
+  "ltxv/ltx2/ltx-2.3-22b-distilled-lora-384-1.1.safetensors".
+  6. Load a normal content/style LoRA with LoraLoaderModelOnly, wired between
+  CheckpointLoaderSimple's MODEL output and the guider/sampler. In the stock
+  templates you chain it after the distilled-LoRA loader. Use the full
+  LoraLoader (model + clip) only for the gemma-3-12b-it-abliterated LoRA,
+  which patches the Gemma text encoder rather than the video model.
+  7. On a dev checkpoint, also load the acceleration LoRA
+  (ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors
+  at 0.5) unless you intend to run the slow guided path.
+  8. For an IC-LoRA, the LoRA loader alone is not enough: either use
+  LTXICLoRALoaderModelOnly + LTXAddVideoICLoRAGuide (ComfyUI-LTXVideo), or —
+  as the official ic_lora template does — feed the LoRA-loaded model through
+  core's GetICLoRAParameters into LTXVAddGuide, so the
+  reference_downscale_factor stored in the file's safetensors metadata is
+  read instead of defaulting to 1.
+  9. Template Library > Video > any LTX-2.3 workflow gives you correct wiring
+  to start from; workflow JSONs are at Comfy-Org/workflow_templates/templates
+  /video_ltx2_3_{t2v,i2v,ia2v,flf2v,ic_lora,id_lora}.json and
+  Lightricks/ComfyUI-LTXVideo/example_workflows/2.3/.
+
+────��──────────────────────────────────────────────────────────────────────────
+  Scraped by lorakit on 2026-09-04 23:12 UTC.
+  Stats and description are the author's; pros/cons are derived from the
+  evidence tagged beside each line.
+───────────────────────────────────────────────────────────────────────────────

ltx2-3-ic-lora-dual-character/metadata.json

@@ -0,0 +1,143 @@
+{
+  "name": "LTX2.3-IC-LORA-Dual-Character",
+  "slug": "ltx2-3-ic-lora-dual-character",
+  "base_model": "LTX-2.3",
+  "media_type": "video",
+  "model_id": 2500098,
+  "version_id": 2810376,
+  "version_name": "v0.1",
+  "source": "civitai",
+  "url": "https://civitai.com/models/2500098",
+  "repo_id": null,
+  "file_path": null,
+  "author": "MaqueAI",
+  "category": {
+    "key": "character",
+    "label": "Character / subject"
+  },
+  "modalities": [
+    "v2v_ic_lora",
+    "t2v",
+    "i2v"
+  ],
+  "primary_modality": "v2v_ic_lora",
+  "checkpoint_family": "dev",
+  "trigger_words": [],
+  "recommended_settings": {
+    "Resolution": "1280x720"
+  },
+  "chaining": {
+    "author": {
+      "stacks": "not stated",
+      "notes": "",
+      "avoid": [],
+      "requires": [],
+      "order": "",
+      "strength": "",
+      "stated": false
+    },
+    "general_rules": [
+      "Stacking an acceleration LoRA with a task LoRA is officially sanctioned: the ID-LoRA template chains LoraLoaderModelOnly(ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16, 0.5) into LoraLoaderModelOnly(ltx-2.3-id-lora-talkvid-3k, 1.0) on the dev checkpoint.",
+      "The Python CLI exposes `--lora <path> [strength]` as a repeatable flag with default strength 1.0, so multiple adapters at once is a supported configuration, not a hack.",
+      "Load the acceleration/distilled LoRA on the dev checkpoint only. The two-stage pipelines require it for the full model but explicitly do not use it for DistilledPipeline, ICLoraPipeline or DubItPipeline, which already run distilled weights.",
+      "Do not stack two distillation LoRAs (e.g. the official rank-384 / rank-111 one plus a community DMD LoRA). The DaSiWa DMD LoRA already blends 75-80% of the official distilled delta into its audio-output route and is described by its author as recreating the distilled model on top of stock dev, i.e. a replacement.",
+      "IC-LoRAs are not chained like ordinary LoRAs: each carries a reference_downscale_factor in its safetensors metadata and needs a node that reads it (GetICLoRAParameters -> LTXVAddGuide in core, or LTXICLoRALoaderModelOnly + LTXAddVideoICLoRAGuide in ComfyUI-LTXVideo). Without one the factor silently defaults to 1.",
+      "Two IC-LoRA reference guides can be driven from a single IC-LoRA: the CrossView Warp author wires both a depth-warp video and the original video into separate reference guides and sets latent_downscale_factor = 1 on both.",
+      "DubItPipeline accepts exactly one Dub-It IC-LoRA (`--lora` exactly once) and applies it in both stages — that path is single-adapter by design.",
+      "Community convention on strengths when stacking: the IC-LoRA Dual-Character author recommends 0.6-1.0 for a content LoRA used alone, dropped to 0.3-0.5 when combined with others.",
+      "Some LoRA pairs are designed to be blended against each other — the LTX 2.3 Crisp Enhance / Soft Enhance pair is described by its author as \"can be used together and balanced at different strengths\".",
+      "The gemma-3-12b-it-abliterated LoRA is not a video LoRA and does not belong in the DiT chain — it patches the Gemma text encoder through a model+clip LoraLoader, as the official templates wire it.",
+      "The full, flexible, trainable bf16 model — ltx-2.3-22b-dev.safetensors is 46.15 GB bf16, 29.15 GB as ltx-2.3-22b-dev-fp8.safetensors. This is what LoRAs are trained against, and what four of the six official ComfyUI LTX-2.3 templates load (t2v, i2v, ia2v, id_lora). Needs either real guidance (Lightricks' own single-stage workflow sets LTXVScheduler to 15 steps with MultimodalGuider) or the distill"
+    ]
+  },
+  "pros": [
+    "Widely adopted — 5,230 downloads. [source: CivitAI stats]",
+    "Positively received — 204 thumbs-up with no down-votes. [source: CivitAI stats]",
+    "Author documents use across several modes (V2V_IC_LORA, T2V, I2V). [source: description]",
+    "Compact at 312 MB — cheap to stack with other LoRAs. [source: file size]",
+    "Author published concrete recommended settings. [source: description]"
+  ],
+  "cons": [
+    "No trigger word — the LoRA is always on once loaded, so strength is the only control. [source: model metadata]"
+  ],
+  "file": {
+    "name": "LTX2.3-IC-LORA-Dual-Character.safetensors",
+    "size_bytes": 327287347,
+    "size_mb": 312.13,
+    "sha256": "b6c3199e2c95eb0aad0cea4f3e8dfa33e8f6966d85bc79e3994faf5a57406103"
+  },
+  "stats": {
+    "downloads": 5230,
+    "thumbs_up": 204,
+    "thumbs_down": 0
+  },
+  "popularity": {
+    "downloads": 5230,
+    "thumbs_up": 204,
+    "thumbs_down": 0,
+    "comments": 0,
+    "approval": "100% positive (204/204)",
+    "age_days": 160.5,
+    "downloads_per_day": 32.59,
+    "rank_by_downloads": 6,
+    "rank_by_momentum": 5,
+    "of_total": 150,
+    "adoption_tier": "Top tier",
+    "momentum": "Hot"
+  },
+  "selection_score": 7678.0,
+  "selection_reasons": [
+    "5,230 downloads",
+    "204 thumbs-up",
+    "top-ranked for its category"
+  ],
+  "example_media": [
+    {
+      "file": "01_video.mp4",
+      "type": "video",
+      "resolution": "1280x704",
+      "prompt": "",
+      "poster": "01_poster.jpg"
+    },
+    {
+      "file": "02_video.mp4",
+      "type": "video",
+      "resolution": "880x560",
+      "prompt": "",
+      "poster": "02_poster.jpg"
+    },
+    {
+      "file": "03_video.mp4",
+      "type": "video",
+      "resolution": "1280x704",
+      "prompt": "",
+      "poster": "03_poster.jpg"
+    },
+    {
+      "file": "04_video.mp4",
+      "type": "video",
+      "resolution": "880x560",
+      "prompt": "",
+      "poster": "04_poster.jpg"
+    },
+    {
+      "file": "05_video.mp4",
+      "type": "video",
+      "resolution": "1024x704",
+      "prompt": "",
+      "poster": "05_poster.jpg"
+    },
+    {
+      "file": "06_video.mp4",
+      "type": "video",
+      "resolution": "1024x704",
+      "prompt": "",
+      "poster": "06_poster.jpg"
+    }
+  ],
+  "tags": [
+    "tool"
+  ],
+  "scraped_at": "2026-09-04T23:12:57+00:00",
+  "path": "ltxv-2.3/ltx2-3-ic-lora-dual-character"
+}