sgl-project/sglang · error · ValueError

{self.config.timestep_spacing} is not supported. Please make

Error message

{self.config.timestep_spacing} is not supported. Please make sure to choose one of 'linspace', 'leading' or 'trailing'.

What it means

Raised by set_timesteps when config.timestep_spacing is not 'linspace', 'leading', or 'trailing'. The scheduler branches on this value to compute the timestep grid; an unrecognized spacing cannot produce timesteps.

Source

Thrown at python/sglang/multimodal_gen/runtime/models/schedulers/scheduling_unipc_multistep.py:414

                (np.arange(0, num_inference_steps + 1) * step_ratio)
                .round()[::-1][:-1]
                .copy()
                .astype(np.int64)
            )
            timesteps += self.config.steps_offset
        elif self.config.timestep_spacing == "trailing":
            step_ratio = self.config.num_train_timesteps / num_inference_steps
            # creates integer timesteps by multiplying by ratio
            # casting to int to avoid issues when num_inference_step is power of 3
            timesteps = (
                np.arange(self.config.num_train_timesteps, 0, -step_ratio)
                .round()
                .copy()
                .astype(np.int64)
            )
            timesteps -= 1
        else:
            raise ValueError(
                f"{self.config.timestep_spacing} is not supported. Please make sure to choose one of 'linspace', 'leading' or 'trailing'."
            )

        sigmas = np.array(((1 - self.alphas_cumprod) / self.alphas_cumprod) ** 0.5)
        if self.config.use_karras_sigmas:
            log_sigmas = np.log(sigmas)
            sigmas = np.flip(sigmas).copy()
            sigmas = self._convert_to_karras(
                in_sigmas=sigmas, num_inference_steps=num_inference_steps
            )
            if self.config.use_flow_sigmas:
                # Karras builds sigmas in EDM space; flow-matching models expect
                # sigmas in [0, 1]. Map EDM -> flow with sigma / (sigma + 1) and
                # derive timesteps from the flow sigmas (matches diffusers >=0.38).
                sigmas = sigmas / (sigmas + 1)
                timesteps = (sigmas * self.config.num_train_timesteps).copy()
            else:
                timesteps = np.array(

View on GitHub (pinned to 0132848349)

Solutions

1. Set timestep_spacing to 'linspace', 'leading', or 'trailing'
2. For models trained with zero terminal SNR use 'trailing' along with rescale_betas_zero_snr=True
3. Validate the scheduler_config.json before loading

Example fix

// before
sched = UniPCMultistepScheduler.from_config(cfg, timestep_spacing="steps")
// after
sched = UniPCMultistepScheduler.from_config(cfg, timestep_spacing="trailing", rescale_betas_zero_snr=True)

assert cfg.get("timestep_spacing", "linspace") in {"linspace", "leading", "trailing"}

When it happens

Trigger: Calling scheduler.set_timesteps(num_inference_steps) after constructing the scheduler with timestep_spacing set to e.g. 'hop' or a typo like 'trailng'.

Common situations: Configs ported from schedulers that support additional spacings (e.g. DDIM-style or LCM 'trailing' variants), or manual config edits. Zero terminal SNR workflows usually require 'trailing'.

Understand the failure class

Background: "Invalid value" and "allowed values are" config errors: what your library rejected and how to fix it — this error's family across 41 libraries.

Related errors

• {beta_schedule} is not implemented for {self.__class__}
• {solver_type} is not implemented for {self.__class__}
• `final_sigmas_type` must be one of 'zero', or 'sigma_min', b
• missing `sample` as a required keyword argument
• prediction_type given as {self.config.prediction_type} must

AI-assisted analysis of sgl-project/sglang@0132848349 (2026-08-28). Data as JSON: /api/errors/be32a5cab9599045. Report an issue: GitHub.