diff --git a/nodes.py b/nodes.py
index 21d5416..3e91dc1 100644
--- a/nodes.py
+++ b/nodes.py
@@ -383,7 +383,7 @@ class InitFluxTraining:
             "loss_type": (["l1", "l2", "huber", "smooth_l1"], {"default": "l2", "tooltip": "loss type"}),
             "timestep_sampling": (["sigmoid", "uniform", "sigma"], {"tooltip": "method to sample timestep"}),
             "sigmoid_scale": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 10.0, "step": 0.1, "tooltip": "Scale factor for sigmoid timestep sampling (only used when timestep-sampling is sigmoid"}),
-            "model_prediction_type": (["raw", "additive", "sigma_scaled"], {"tooltip": "How to interpret and process the model prediction: raw (use as is), additive (add to noisy input), sigma_scaled (apply sigma scaling)."}),
+            "model_prediction_type": (["raw", "additive", "sigma_scaled", "shift"], {"tooltip": "How to interpret and process the model prediction: raw (use as is), additive (add to noisy input), sigma_scaled (apply sigma scaling), shift of sigmoid (recommend value of 3.1582 for discrete_flow_shift)."}),
             "cpu_offload_checkpointing": ("BOOLEAN", {"default": True, "tooltip": "offload the gradient checkpointing to CPU. This reduces VRAM usage for about 2GB"}),
             "optimizer_fusing": (['fused_backward_pass', 'blockwise_fused_optimizers'], {"tooltip": "reduces memory use"}),
             "single_blocks_to_swap": ("INT", {"default": 0, "min": 0, "max": 100, "step": 1, "tooltip": "number of single blocks to swap. The default is 0. This option must be combined with blockwise_fused_optimizers"}),