🎉 Release Version 2 - Enhanced quality and performance

- Add next-scene_lora-v2-3000.safetensors with improved training data
- Better command responsiveness and prompt following
- Fixed black bar artifacts
- Updated README with V2 announcement and usage instructions

README.md

@@ -19,6 +19,22 @@ tags:
 
 ![Demo 1](01.gif) ![Demo 2](02.gif) ![Demo 3](03.gif)
 
+---
+
+## 🎉 ✨ **UPDATE - Version 2 Now Available!** ✨ 🎉
+
+🚀 **New Model:** `next-scene_lora-v2-3000.safetensors`
+
+**What's New in V2:**
+- 🎯 **Trained on higher quality data** for significantly improved results
+- 💪 **Better command responsiveness** - the model follows your prompts more accurately
+- 🖼️ **Fixed black bar artifacts** - no more unwanted black borders on generated images
+- ⚡ **Overall enhanced performance** - smoother transitions and better cinematic flow
+
+**Recommended:** Use V2 for all new projects.
+
+---
+
 **next-scene-qwen-image-lora-2509** is a **LoRA adapter fine-tuned on Qwen-Image-Edit (build 2509)**, purpose-built to generate cinematic image sequences with natural visual progression from frame to frame.
 
 This model enables Qwen Image Edit to think like a film director—understanding camera dynamics, visual composition, and narrative continuity to create shots that flow seamlessly into one another.
@@ -51,7 +67,9 @@ Each output frame functions as the *"Next Scene"* in an evolving visual narrativ
 ### Basic Setup:
 
 1. Load **Qwen-Image-Edit 2509** as your base model
-2. Add a **LoRA Loader** node and select `next-scene-qwen-image-lora-2509`
+2. Add a **LoRA Loader** node and select:
+   - **V2 (Recommended):** `next-scene_lora-v2-3000.safetensors`
+   - **V1 (Legacy):** `next-scene_lora_v1-3000.safetensors`
 3. Set LoRA strength: **0.7 – 0.8** (recommended)
 4. Structure your prompts with **"Next Scene:"** prefix for optimal results
 

next-scene_lora-v2-3000.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8abf2306649a62282f89968515097650acaf7b040687c9b6c56e37558e6df9e1
+size 295146176