Update README for flux1-schnell-fp8-blackwell

README.md

@@ -1,52 +1,128 @@
-# FLUX1-SCHNELL-FP8-BLACKWELL
+---
+library_name: tensorrt-rtx
+license: apache-2.0
+base_model: black-forest-labs/FLUX.1-schnell
+tags:
+- tensorrt-rtx
+- flux1
+- fp8
+- schnell
+- optimized
+inference: false
+---
 
-TensorRT-RTX engines for FLUX1-SCHNELL optimized for Blackwell architecture with FP8 quantization.
+# FLUX1 TensorRT-RTX: SCHNELL-Fp8 🔨 Building
 
-## Specifications
+Optimized TensorRT-RTX engines for **FLUX1** on **Fp8** architecture with **SCHNELL** quantization.
 
-- **Model**: FLUX1-SCHNELL
-- **Base Model**: black-forest-labs/FLUX.1-schnell
-- **Architecture**: Blackwell
-- **Quantization**: FP8
-- **Batch Size**: 1 (optimized)
-- **Resolution**: 1024x1024 (optimized)
-- **Estimated Size**: 6.0 GB
+## 🎯 This Repository
 
-## Performance Estimates
+**One variant, one download** - only get exactly what you need!
 
-Based on architecture and quantization:
-- **Memory Usage**: ~6.0GB VRAM
-- **Speed**: ~2.1s (H200)
+- **Model**: FLUX1
+- **Architecture**: Fp8 (Compute Capability 8.0+)  
+- **Quantization**: SCHNELL
+- **Memory**: TBD
+- **Speed**: TBD for 1024x1024 generation
 
-## Files
+## 🚀 Quick Start
 
-- `engines/`: TensorRT engine files (.plan)
-- `config.json`: Configuration metadata  
-- `README.md`: This file
+### Automatic (Recommended)
 
-## Usage
+```bash
+# ImageAI server downloads automatically
+curl -X POST "http://localhost:8001/generate" \
+  -H "Content-Type: application/json" \
+  -d '{
+    "prompt": "a beautiful landscape",
+    "model": "flux1-tensorrt_rtx:schnell",
+    "width": 1024,
+    "height": 1024
+  }'
+```
+
+### Manual Download
+
+```python
+from huggingface_hub import snapshot_download
+
+# Download this specific variant only
+engines_path = snapshot_download(
+    repo_id="imgailab/flux1-trtx-schnell-fp8-blackwell"
+)
+
+# Engines are in: engines_path/engines/*.plan
+```
+
+### Direct Integration
 
 ```python
 from imageai_server.tensorrt.nvidia_sdxl_pipeline import NVIDIASDXLPipeline
 
 pipeline = NVIDIASDXLPipeline()
 pipeline.load_engines(
-    engine_dir="./engines",
-    framework_model_dir="./framework",
-    onnx_dir="./onnx"
+    engine_dir=f"{engines_path}/engines",
+    framework_model_dir=f"{engines_path}/framework",  
+    onnx_dir=f"{engines_path}/onnx"
 )
 pipeline.activate_engines()
 
 images, time_ms = pipeline.infer(
-    prompt="a beautiful landscape",
+    prompt="a serene mountain landscape",
     height=1024,
     width=1024
 )
 ```
 
-## Requirements
+## 📊 Performance
+
+| Metric | Value |
+|--------|-------|
+| **Memory Usage** | TBD |
+| **Inference Speed** | TBD |
+| **Resolution** | 1024x1024 (optimized) |
+| **Batch Size** | 1 (optimized) |
+| **Precision** | SCHNELL |
+
+## 🔧 Requirements
+
+### Hardware
+- **GPU**: Fp8 architecture
+  - Ampere: RTX 3090, A100, etc.
+  - Ada Lovelace: RTX 4090, etc.
+  - Blackwell: H200, etc.
+- **VRAM**: TBD minimum
+- **Compute Capability**: 8.0+
 
-- **GPU**: Blackwell architecture (Compute Capability 8.0+)
-- **VRAM**: 6.0GB+
+### Software  
 - **TensorRT-RTX**: 1.0.0.21+
 - **CUDA**: 12.0+
+- **Python**: 3.8+
+
+## 📁 Repository Structure
+
+```
+flux1-trtx-schnell-fp8-blackwell/
+├── engines/           # TensorRT engine files
+│   ├── *.plan        # Optimized engines
+├── config.json       # Configuration metadata
+└── README.md         # This file
+```
+
+## 🌐 Related Repositories
+
+Other variants for FLUX1:
+- [Ampere BF16](https://huggingface.co/imgailab/flux1-trtx-bf16-ampere)\n- [Ada FP8](https://huggingface.co/imgailab/flux1-trtx-fp8-ada)\n- [Ada BF16](https://huggingface.co/imgailab/flux1-trtx-bf16-ada)\n- [Blackwell FP4](https://huggingface.co/imgailab/flux1-trtx-fp4-blackwell)\n- [Blackwell FP8](https://huggingface.co/imgailab/flux1-trtx-fp8-blackwell)\n- [Blackwell BF16](https://huggingface.co/imgailab/flux1-trtx-bf16-blackwell)\n
+
+## 📝 License
+
+Inherits license from base model: [black-forest-labs/FLUX.1-schnell](https://huggingface.co/black-forest-labs/FLUX.1-schnell)
+
+## 🔄 Updates
+
+- **2025-08-12**: Initial release
+- Optimized for single-variant downloads
+
+---
+
+*Part of the ImageAI TensorRT-RTX engine collection*