{"record":{"id":"0be9398cf58e39e4","repo":"Comfy-Org/ComfyUI","slug":"sigma-min-and-sigma-max-must-not-be-0","errorCode":null,"errorMessage":"sigma_min and sigma_max must not be 0","messagePattern":"sigma_min and sigma_max must not be 0","errorType":"validation","errorClass":"ValueError","httpStatus":null,"severity":"error","filePath":"comfy/k_diffusion/sampling.py","lineNumber":624,"sourceCode":"                x = x_high + su * s_noise * noise_sampler(self.sigma(s), self.sigma(t))\n                s = t\n                info['n_accept'] += 1\n            else:\n                info['n_reject'] += 1\n            info['nfe'] += order\n            info['steps'] += 1\n\n            if self.info_callback is not None:\n                self.info_callback({'x': x, 'i': info['steps'] - 1, 't': s, 't_up': s, 'denoised': denoised, 'error': error, 'h': pid.h, **info})\n\n        return x, info\n\n\n@torch.no_grad()\ndef sample_dpm_fast(model, x, sigma_min, sigma_max, n, extra_args=None, callback=None, disable=None, eta=0., s_noise=1., noise_sampler=None):\n    \"\"\"DPM-Solver-Fast (fixed step size). See https://arxiv.org/abs/2206.00927.\"\"\"\n    if sigma_min <= 0 or sigma_max <= 0:\n        raise ValueError('sigma_min and sigma_max must not be 0')\n    with tqdm(total=n, disable=disable) as pbar:\n        dpm_solver = DPMSolver(model, extra_args, eps_callback=pbar.update)\n        if callback is not None:\n            dpm_solver.info_callback = lambda info: callback({'sigma': dpm_solver.sigma(info['t']), 'sigma_hat': dpm_solver.sigma(info['t_up']), **info})\n        return dpm_solver.dpm_solver_fast(x, dpm_solver.t(torch.tensor(sigma_max)), dpm_solver.t(torch.tensor(sigma_min)), n, eta, s_noise, noise_sampler)\n\n\n@torch.no_grad()\ndef sample_dpm_adaptive(model, x, sigma_min, sigma_max, extra_args=None, callback=None, disable=None, order=3, rtol=0.05, atol=0.0078, h_init=0.05, pcoeff=0., icoeff=1., dcoeff=0., accept_safety=0.81, eta=0., s_noise=1., noise_sampler=None, return_info=False):\n    \"\"\"DPM-Solver-12 and 23 (adaptive step size). See https://arxiv.org/abs/2206.00927.\"\"\"\n    if sigma_min <= 0 or sigma_max <= 0:\n        raise ValueError('sigma_min and sigma_max must not be 0')\n    with tqdm(disable=disable) as pbar:\n        dpm_solver = DPMSolver(model, extra_args, eps_callback=pbar.update)\n        if callback is not None:\n            dpm_solver.info_callback = lambda info: callback({'sigma': dpm_solver.sigma(info['t']), 'sigma_hat': dpm_solver.sigma(info['t_up']), **info})\n        x, info = dpm_solver.dpm_solver_adaptive(x, dpm_solver.t(torch.tensor(sigma_max)), dpm_solver.t(torch.tensor(sigma_min)), order, rtol, atol, h_init, pcoeff, icoeff, dcoeff, accept_safety, eta, s_noise, noise_sampler)\n    if return_info:","sourceCodeStart":606,"sourceCodeEnd":642,"githubUrl":"https://github.com/Comfy-Org/ComfyUI/blob/1c6d8d45b3693bfbb32385b410d813a7fd6be216/comfy/k_diffusion/sampling.py#L606-L642","documentation":"Guard at the top of sample_dpm_fast: the DPM-Solver time mapping t(sigma) = log(sigma) is undefined at sigma=0 (log of zero), and zero-sigma endpoints would also collapse the schedule. Both sigma_min and sigma_max must be strictly positive; a non-positive value raises before the solver is built.","triggerScenarios":"Calling sample_dpm_fast(model, x, sigma_min=0.0, ...) or with a negative bound. Common when code forwards a scheduler's last sigma (which is often exactly 0.0 for terminal denoise) directly as sigma_min.","commonSituations":"Custom samplers piping schedule endpoints computed with final sigmas zeroed (ComfyUI's sigmas convention ends in 0); UI widgets defaulting to 0; converting from samplers that take a full sigmas tensor and silently drop the zero.","solutions":["Clamp sigma_min to a small positive epsilon, e.g. max(sigma_min, 1e-5) or the model's sigma_min (often ~0.03)","Strip the terminal 0.0 from computed sigma schedules before passing bounds","Check widget defaults in custom nodes are not 0"],"exampleFix":"# before\nsample_dpm_fast(model, x, sigma_min=float(sigmas[-1]), sigma_max=float(sigmas[0]), n=steps)  # sigmas[-1] == 0.0\n# after\nsample_dpm_fast(model, x, sigma_min=max(float(sigmas[-2]), 1e-5), sigma_max=float(sigmas[0]), n=steps)","handlingStrategy":"validation","validationCode":"sigma_min = max(float(sigma_min), 1e-5)\nsigma_max = max(float(sigma_max), sigma_min)\nassert sigma_min > 0 and sigma_max > 0","typeGuard":null,"tryCatchPattern":null,"preventionTips":["Never forward a schedule's terminal 0.0 sigma as sigma_min","Default sigma widgets to the model's own range, not 0"],"tags":["sampling","dpm-solver","sigma","validation","schedule"],"backgroundTag":null,"analyzedSha":"1c6d8d45b3693bfbb32385b410d813a7fd6be216","analyzedAt":"2026-08-14T19:37:18.893Z","schemaVersion":2},"datasetVersion":"2026-08-15T22:17:37.221Z"}