upload new workflow for my new lora

README.md

@@ -23,6 +23,7 @@ Drag-and-drop any `.json` file directly into ComfyUI to load the workflow.
 
 | Workflow                                                                                                                                                   | Description                                                                      | Type      | Updated    |
 | ---------------------------------------------------------------------------------------------------------------------------------------------------------- | -------------------------------------------------------------------------------- | --------- | ---------- |
+| [IC-LoRA CrossView Warp (v2)](https://huggingface.co/datasets/Cseti/ComfyUI-Workflows/blob/main/ltx/2.3/ic-lora-crossview-warp-v2/README.md) | V2V new-camera-angle IC-LoRA — MoGe-2 geometry warp, keyframed virtual camera path set in the node's live preview (v2, step 6k) | Video gen | 2026-08-15 |
 | [IC-LoRA CrossView Warp (v0.9)](https://huggingface.co/datasets/Cseti/ComfyUI-Workflows/blob/main/ltx/2.3/ic-lora-crossview-warp/README.md) | V2V new-camera-angle IC-LoRA — set the virtual second camera by numbers (azimuth/elevation/distance) via a depth-warp node (PoC, step 18k) | Video gen | 2026-07-24 |
 | [IC-LoRA CrossView Prompt (v0.9)](https://huggingface.co/datasets/Cseti/ComfyUI-Workflows/blob/main/ltx/2.3/ic-lora-crossview-v1-pilot/README.md) | V2V new-camera-angle IC-LoRA — re-render a clip from a prompt-described virtual second camera (pilot, step 13.7k) | Video gen | 2026-07-11 |
 | [IC-LoRA Cameraman v2](https://huggingface.co/datasets/Cseti/ComfyUI-Workflows/blob/main/ltx/2.3/ic-lora-cameraman-v2/README.md) | I2V with IC-LoRA camera motion guidance (Cameraman v2, step 14k) — dev + distilled LoRA + 2x upscaler | Video gen | 2026-06-24 |

ltx/2.3/ic-lora-crossview-warp-v2/README.md

@@ -0,0 +1,75 @@
+# LTX Video 2.3 — IC-LoRA CrossView Warp (v2) WORKFLOW
+
+**Model:** LTX Video 2.3 — 22B Dev (INT8) + distilled speed LoRA
+**Type:** Video-to-Video
+
+Video-to-Video workflow using LTX Video 2.3 with the [CrossView Warp IC-LoRA v2](https://huggingface.co/Cseti/LTX2.3-22B_IC-LoRA-CrossView-Warp_v2) — it re-renders an input clip from a new virtual camera angle, like a second camera on the same take. The angle is set as **numbers** — azimuth / elevation / distance — on the [CrossViewWarp node](https://github.com/cseti007/ComfyUI-CrossViewWarp), which builds a geometry warp of the input clip and feeds it to the IC-LoRA as a guide; a second guide carries the source clip for identity. Runs on the 22B **dev** transformer with the distilled speed LoRA and a 2x latent spatial upscaler.
+
+Compared to the v0.9 proof-of-concept entry, this version uses **MoGe-2 metric geometry** instead of Depth-Anything-V2 relative depth, and supports a **keyframed camera path** (the camera can move during the clip) authored in the node's live warp preview.
+
+---
+
+## Preview
+
+<video src="https://huggingface.co/datasets/Cseti/ComfyUI-Workflows/resolve/main/ltx/2.3/ic-lora-crossview-warp-v2/media/preview.mp4" controls width="720"></video>
+
+3x3 grid: azimuth -45 / 0 / +45 against elevation +30 / 0 / -30, with the original source footage in the centre.
+
+<!-- Replace YOUTUBE_VIDEO_ID with your video ID after uploading -->
+<!-- [![Watch on YouTube — click to open](https://img.youtube.com/vi/YOUTUBE_VIDEO_ID/maxresdefault.jpg)](https://www.youtube.com/watch?v=YOUTUBE_VIDEO_ID) -->
+<!-- *Click to open on YouTube* -->
+
+---
+
+## Requirements
+
+- **ComfyUI:** recent stable build (the native MoGe geometry-estimation nodes are required)
+- **Model:** `ltx-2.3-22b-dev_transformer_only_int8_convrot.safetensors` — [Kijai/LTX2.3_comfy](https://huggingface.co/Kijai/LTX2.3_comfy/blob/main/diffusion_models/ltx-2.3-22b-dev_transformer_only_int8_convrot.safetensors)
+- **Distilled speed LoRA:** `ltx-2.3-22b-distilled-1.1_lora-dynamic_fro09_avg_rank_111_bf16.safetensors` — [Kijai/LTX2.3_comfy](https://huggingface.co/Kijai/LTX2.3_comfy/blob/main/loras/ltx-2.3-22b-distilled-1.1_lora-dynamic_fro09_avg_rank_111_bf16.safetensors)
+- **IC-LoRA:** `LTX2.3-22B_IC-LoRA-CrossView-Warp_v2_6000.safetensors` — [Cseti/LTX2.3-22B_IC-LoRA-CrossView-Warp_v2](https://huggingface.co/Cseti/LTX2.3-22B_IC-LoRA-CrossView-Warp_v2) (place under `models/loras/LTX2/`)
+- **Video VAE:** `LTX23_video_vae_bf16.safetensors` — [Kijai/LTX2.3_comfy](https://huggingface.co/Kijai/LTX2.3_comfy/blob/main/vae/LTX23_video_vae_bf16.safetensors)
+- **Audio VAE:** `LTX23_audio_vae_bf16.safetensors` — [Kijai/LTX2.3_comfy](https://huggingface.co/Kijai/LTX2.3_comfy/blob/main/vae/LTX23_audio_vae_bf16.safetensors)
+- **Preview VAE:** `taeltx2_3.safetensors` — [Kijai/LTX2.3_comfy](https://huggingface.co/Kijai/LTX2.3_comfy/blob/main/vae/taeltx2_3.safetensors)
+- **Text encoder:** `gemma_3_12B_it_fp8_scaled.safetensors` — [Comfy-Org/ltx-2](https://huggingface.co/Comfy-Org/ltx-2/blob/main/split_files/text_encoders/gemma_3_12B_it_fp8_scaled.safetensors)
+- **Text projection:** `ltx-2.3_text_projection_bf16.safetensors` — [Kijai/LTX2.3_comfy](https://huggingface.co/Kijai/LTX2.3_comfy/blob/main/text_encoders/ltx-2.3_text_projection_bf16.safetensors)
+- **Spatial upscaler:** `ltx-2.3-spatial-upscaler-x2-1.1.safetensors` — [Lightricks/LTX-2.3](https://huggingface.co/Lightricks/LTX-2.3/blob/main/ltx-2.3-spatial-upscaler-x2-1.1.safetensors)
+- **Geometry model:** `moge_2_vitl_normal_fp16.safetensors` — [Comfy-Org/MoGe](https://huggingface.co/Comfy-Org/MoGe/blob/main/geometry_estimation/moge_2_vitl_normal_fp16.safetensors) (place under `models/geometry_estimation/`)
+- **Custom nodes:**
+  - [ComfyUI-CrossViewWarp](https://github.com/cseti007/ComfyUI-CrossViewWarp) — the CrossView Warp node (companion to the IC-LoRA), **v0.11.0 or newer**
+  - [ComfyUI-LTXVideo](https://github.com/Lightricks/ComfyUI-LTXVideo)
+  - [ComfyUI-KJNodes](https://github.com/kijai/ComfyUI-KJNodes)
+  - [ComfyUI-VideoHelperSuite](https://github.com/Kosinkadink/ComfyUI-VideoHelperSuite)
+  - [rgthree-comfy](https://github.com/rgthree/rgthree-comfy)
+
+---
+
+## Notes
+
+- **Input:** a reference clip is loaded via `Load Video`. It goes to the **CrossViewWarp node** together with the MoGe-2 geometry of the same clip; the node outputs the warp. The warp and the source clip each get their own `LTXAddVideoICLoRAGuide`.
+- **Setting the angle:** type `azimuth` / `elevation` / `distance` on the node, or use its two interactive views — the **orbit** picker (drag the marker on a 3D sphere) and the **warp preview** (drag the actual warped clip, scrub and play it). Both views write the same widgets, so the preview shows exactly what the run will feed the model.
+- **Coverage:** stay in the green zone — azimuth up to +/-45 deg, elevation +/-30 deg, distance +/-20%. Yellow (up to +/-90 deg azimuth) still works with degrading quality. Full range table is on the [LoRA card](https://huggingface.co/Cseti/LTX2.3-22B_IC-LoRA-CrossView-Warp_v2).
+- **Camera move:** this workflow ships with `use_keyframes` enabled and a 3-point path. Position the playhead in the warp preview, drag to the pose you want, press `KEY`. With a single keyframe the camera is static; with several it interpolates between them (azimuth, elevation, distance, vertical shift and the pivot are all keyframeable).
+- **IC-LoRA strength `1`**; distilled speed LoRA at `0.6`. Both IC-LoRA guides use `latent_downscale_factor = 1`.
+- For `distance > 1`: describe the newly revealed content in the text prompt — the warp cannot know what is outside the original frame.
+- **Two-pass:** an 8-step base pass (960x544, 97 frames @ 24 fps, with audio) followed by a 3-step 2x latent spatial upscale pass, output at 1920x1080.
+- From the workflow's own note on the upscale pass: *"For the upscale pass I'm using the reference video also together with the CrossView IC-LoRA but without the warp condition. It probably provides better likeness for the upscale pass. I'm not sure, never tested without it."*
+
+---
+
+## Changelog
+
+- `2026-08-15` — Initial upload (v2 IC-LoRA, MoGe-2 geometry, keyframed camera path)
+
+---
+
+## Training
+
+The IC-LoRA was trained on [RunPod cloud GPUs](https://runpod.io?ref=vpp0cion) (RTX PRO 6000 Blackwell). Full training details are on the [LoRA card](https://huggingface.co/Cseti/LTX2.3-22B_IC-LoRA-CrossView-Warp_v2).
+
+---
+
+## Support
+
+Producing and sharing this kind of open-source work requires renting cloud GPUs, which gets expensive quickly. If you find it useful and would like me to keep contributing, your support is very much appreciated:
+
+[![Ko-fi](https://img.shields.io/badge/Ko--fi-Support-FF5E5B?style=for-the-badge&logo=ko-fi&logoColor=white)](https://ko-fi.com/chetyart) [![Liberapay](https://img.shields.io/badge/Liberapay-Donate-F6C915?style=for-the-badge&logo=liberapay&logoColor=black)](https://liberapay.com/chetyart/donate)

ltx/2.3/ic-lora-crossview-warp-v2/ltx2.3-ic-lora-crossview-warp-v2.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b123239cece720a2bf87e392713766d958af941beb0abad6c96dc87ed489868
+size 159338

ltx/2.3/ic-lora-crossview-warp-v2/media/preview.mp4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:201e4eb617161278747127937848c7ce2a817769c83e36f60833ee6bcd7f1da5
+size 3785699