Comfy-Org/ComfyUI · error · ValueError

schedule '{schedule}' unknown.

Error message

schedule '{schedule}' unknown.

What it means

Raised by make_beta_schedule when the schedule name doesn't match any implemented beta schedule: 'linear', 'cosine', 'squaredcos_cap_v2', 'sqrt_linear', 'sqrt' (plus their quad variants). The function dispatches purely on the lowercase schedule string from the model/sampler config; unknown strings hit the trailing ValueError.

Source

Thrown at comfy/ldm/modules/diffusionmodules/util.py:117

        alphas = timesteps / (1 + cosine_s) * np.pi / 2
        alphas = torch.cos(alphas).pow(2)
        alphas = alphas / alphas[0]
        betas = 1 - alphas[1:] / alphas[:-1]
        betas = torch.clamp(betas, min=0, max=0.999)

    elif schedule == "squaredcos_cap_v2":  # used for karlo prior
        # return early
        return betas_for_alpha_bar(
            n_timestep,
            lambda t: math.cos((t + 0.008) / 1.008 * math.pi / 2) ** 2,
        )

    elif schedule == "sqrt_linear":
        betas = torch.linspace(linear_start, linear_end, n_timestep, dtype=torch.float64)
    elif schedule == "sqrt":
        betas = torch.linspace(linear_start, linear_end, n_timestep, dtype=torch.float64) ** 0.5
    else:
        raise ValueError(f"schedule '{schedule}' unknown.")
    return betas


def make_ddim_timesteps(ddim_discr_method, num_ddim_timesteps, num_ddpm_timesteps, verbose=True):
    if ddim_discr_method == 'uniform':
        c = num_ddpm_timesteps // num_ddim_timesteps
        ddim_timesteps = np.asarray(list(range(0, num_ddpm_timesteps, c)))
    elif ddim_discr_method == 'quad':
        ddim_timesteps = ((np.linspace(0, np.sqrt(num_ddpm_timesteps * .8), num_ddim_timesteps)) ** 2).astype(int)
    else:
        raise NotImplementedError(f'There is no ddim discretization method called "{ddim_discr_method}"')

    # assert ddim_timesteps.shape[0] == num_ddim_timesteps
    # add one to get the final alpha values right (the ones from first scale to data during sampling)
    steps_out = ddim_timesteps + 1
    if verbose:
        logging.info(f'Selected timesteps for ddim sampler: {steps_out}')
    return steps_out

View on GitHub (pinned to 1c6d8d45b3)

Solutions

1. Use one of the supported names: 'linear', 'quad', 'cosine', 'sqrt_linear', 'sqrt', 'squaredcos_cap_v2' (and their variants as implemented)
2. Read the schedule string from the checkpoint config and map it to the closest supported schedule rather than inventing one
3. Normalize the string (lowercase, strip) before dispatch

Example fix

# before
betas = make_beta_schedule('cosine-beta', 1000)
# after
betas = make_beta_schedule('cosine', 1000)

SUPPORTED = {'linear', 'quad', 'cosine', 'cosine_linear', 'squaredcos_cap_v2', 'sqrt_linear', 'sqrt'}
if schedule not in SUPPORTED:
    raise ValueError(f"unsupported schedule {schedule!r}; supported: {sorted(SUPPORTED)}")
betas = make_beta_schedule(schedule, n_timestep)

def is_supported_schedule(name: str) -> bool:
    return name in {'linear', 'quad', 'cosine', 'cosine_linear', 'squaredcos_cap_v2', 'sqrt_linear', 'sqrt'}

When it happens

Trigger: Calling make_beta_schedule('linear-beta') or passing a config's schedule field verbatim when it contains a name this implementation doesn't have (e.g. 'sigmoid' or a typo).

Common situations: Porting diffusion configs between codebases (compvis/diffusers/custom) with differing schedule naming; typos in hand-written sampler configs; renamed schedules across versions.

Related errors

• error invalid scheduler {scheduler_name}
• Unknown rope_img: {rope_img}
• Unsupported spatial_scale {scale}. Choose from {list(mapping
• Invalid {cls.__name__} string: '{value}'. Valid values are:
• Invalid normalization type: {normtype}

AI-assisted analysis of Comfy-Org/ComfyUI@1c6d8d45b3 (2026-08-14). Data as JSON: /api/errors/dca7ef91463ab761. Report an issue: GitHub.