Add --image-precision option and VAE tiling support for diffusers

- Add --image-precision with choices: bf16, f32, f16, f8
- bf16 recommended for modern GPUs (RTX 30/40 series) to avoid NaN issues
- Enable VAE tiling for diffusers when --vae-tiling is specified

coderai

@@ -3741,18 +3741,29 @@ async def create_image_generation(request: ImageGenerationRequest, http_request:
         if pipeline is None:
             print(f"Loading Stable Diffusion model: {model_to_use}")
             
+            # Determine precision from CLI argument
+            precision = getattr(global_args, 'image_precision', 'f32') or 'f32'
+            precision_map = {
+                'bf16': torch.bfloat16,
+                'f32': torch.float32,
+                'f16': torch.float16,
+                'f8': torch.float8_e4m3fn,
+            }
+            torch_dtype = precision_map.get(precision, torch.float32)
+            print(f"Using precision: {precision} ({torch_dtype})")
+            
             # Try to load as Stable Diffusion XL first
             try:
                 pipeline = StableDiffusionXLPipeline.from_pretrained(
                     model_to_use,
-                    torch_dtype=torch.float32,
+                    torch_dtype=torch_dtype,
                     use_safetensors=True,
                 )
             except Exception:
                 # Try generic diffusion pipeline
                 pipeline = DiffusionPipeline.from_pretrained(
                     model_to_use,
-                    torch_dtype=torch.float32,
+                    torch_dtype=torch_dtype,
                     use_safetensors=True,
                 )
             
@@ -3766,6 +3777,11 @@ async def create_image_generation(request: ImageGenerationRequest, http_request:
             if torch.cuda.is_available():
                 pipeline.enable_attention_slicing()
             
+            # Enable VAE tiling if requested (for lower VRAM usage)
+            if getattr(global_args, 'vae_tiling', False):
+                print("Enabling VAE tiling for lower VRAM usage...")
+                pipeline.enable_vae_tiling()
+            
             multi_model_manager.add_model(model_key, pipeline)
         
         # Get timestamp BEFORE calling diffusers (to avoid scope issues)
@@ -5087,6 +5103,13 @@ def parse_args():
         default=1.0,
         help="CFG scale for image generation (default: 1.0 for Z-Image Turbo).",
     )
+    parser.add_argument(
+        "--image-precision",
+        type=str,
+        default="f32",
+        choices=["bf16", "f32", "f16", "f8"],
+        help="Model precision for image generation (default: f32). bf16 recommended for modern GPUs.",
+    )
     parser.add_argument(
         "--image-seed",
         type=int,