Update README.md

README.md

@@ -1,3 +1,111 @@
----
-license: creativeml-openrail-m
----
+---
+license: creativeml-openrail-m
+language:
+- en
+pipeline_tag: text-to-image
+library_name: diffusers
+---
+
+# LiteVision-v1
+
+LiteVision-v1 is a lightweight, fast Stable Diffusion 1.5-based model optimized for low-step inference using LCM scheduler.
+
+Built for speed.  
+Minimal steps.  
+Clean results.
+
+---
+
+## 🚀 Key Features
+
+- 🔹 Based on Stable Diffusion 1.5 architecture
+- 🔹 Optimized for LCM (Latent Consistency Models)
+- 🔹 High-quality output in **4–8 inference steps**
+- 🔹 Works with standard `StableDiffusionPipeline`
+- 🔹 No custom pipeline required
+- 🔹 Fully compatible with Diffusers 0.36.0
+
+---
+
+## ⚡ Recommended Settings
+
+For best results:
+
+```python
+num_inference_steps = 6
+guidance_scale = 1.5
+scheduler = LCMScheduler
+Lower guidance gives cleaner results.
+Higher guidance may introduce instability due to LCM behavior.
+
+🧠 Why LiteVision?
+Traditional SD 1.5 models require 20–30 steps for good quality.
+
+LiteVision-v1 is tuned for:
+
+Faster sampling
+
+Lower compute
+
+Minimal VRAM usage
+
+Rapid prototyping
+
+Perfect for:
+
+Real-time applications
+
+Low-power GPUs
+
+Quick generation workflows
+
+🛠 Usage
+python
+Copy code
+import torch
+from diffusers import DiffusionPipeline
+
+pipe = DiffusionPipeline.from_pretrained(
+    "HyHorX/LiteVision-v1",
+    torch_dtype=torch.float16
+).to("cuda")
+
+image = pipe(
+    "cinematic cyberpunk city, ultra detailed",
+    num_inference_steps=6,
+    guidance_scale=1.5,
+    height=512,
+    width=512
+).images[0]
+
+image.save("litevision_output.png")
+📦 Architecture
+UNet: SD 1.5 compatible
+
+VAE: AutoencoderKL
+
+Text Encoder: CLIP
+
+Scheduler: LCMScheduler (recommended)
+
+Safety Checker: StableDiffusionSafetyChecker
+
+🧪 Benchmark (LCM)
+Model	Steps	Time	Quality
+SD 1.5	20	Slow	High
+LiteVision-v1	6	Fast	High
+
+(Tested on RTX-class GPU)
+
+⚠ Notes
+Designed specifically for LCM scheduler.
+
+Not tuned for traditional DDIM/PNDM high-step sampling.
+
+Use FP16 for optimal performance.
+
+👤 Author
+HyHorX
+
+License
+Same as Stable Diffusion 1.5.