Upload 2 files

README.md

@@ -0,0 +1,49 @@
+---
+title: Boomer FLA T2I
+emoji: 🎨
+colorFrom: indigo
+colorTo: purple
+sdk: gradio
+sdk_version: 6.15.2
+app_file: app.py
+pinned: false
+license: other
+short_description: 1024px text-to-image with Boomer FLA on ZeroGPU
+models:
+  - akrao9/Boomer-T2I
+  - google/gemma-4-E2B-it
+  - mit-han-lab/dc-ae-f32c32-sana-1.1-diffusers
+preload_from_hub:
+  - akrao9/Boomer-T2I
+---
+
+# Boomer FLA — Text to Image Demo
+
+Interactive demo for [akrao9/Boomer-T2I](https://huggingface.co/akrao9/Boomer-T2I): a 1024×1024 text-to-image flow-matching model with Flash Linear Attention.
+
+## Space setup
+
+1. Create a new **Gradio** Space on Hugging Face.
+2. Upload the contents of this `Space/` folder (or point the Space repo here).
+3. In **Settings → Hardware**, select **ZeroGPU** (requires HF PRO for hosting). This allocates a shared **NVIDIA RTX Pro 6000 Blackwell** GPU (48 GB via `@spaces.GPU(size="large")`).
+4. Add a Space secret **`HF_TOKEN`** with a token that has accepted the [Gemma Terms of Use](https://ai.google.dev/gemma/terms).
+5. Restart the Space after the first build (model + VAE + text encoder download can take several minutes).
+
+## Local test
+
+```bash
+cd Space
+pip install -r requirements.txt
+export HF_TOKEN=hf_...
+python app.py
+```
+
+The `@spaces.GPU` decorator is a no-op locally; use a CUDA machine for full inference.
+
+Pinned deps for ZeroGPU: do **not** pin `torch` or `torchvision` in `requirements.txt` — the Space builder injects `torch<=2.11.0` automatically. Other pins: diffusers 0.38.0, transformers 5.9.0, flash-linear-attention 0.5.0.
+
+## Notes
+
+- Uses `@spaces.GPU(size="large", duration=150)` — RTX Pro 6000 Blackwell 48 GB tier — for 32-step STORK-2 at 1024px.
+- Example prompts run generation on click (`run_on_click=True`).
+- VAE and Gemma 4 2B are fetched from their upstream repos on first startup.

requirements.txt

@@ -0,0 +1,9 @@
+gradio==6.15.2
+diffusers==0.38.0
+transformers==5.9.0
+accelerate==1.13.0
+safetensors==0.8.0rc0
+pillow>=11.3.0
+huggingface-hub==1.16.1
+einops>=0.8.2
+flash-linear-attention==0.5.0