jax-ml/jax · error · ValueError

Number of processes must be defined.

Error message

Number of processes must be defined.

What it means

jax.distributed.initialize requires num_processes (total world size) to be explicitly set; it has no default. Omitting it raises this ValueError before any connection is attempted.

Source

Thrown at jax/_src/distributed.py:139

      local_device_ids = list(map(int, env_ids.split(",")))

    if (cluster_detection_method != 'deactivate' and
        None in (coordinator_address, num_processes, process_id, local_device_ids)):
      (coordinator_address, num_processes, process_id, local_device_ids) = (
          clusters.ClusterEnv.auto_detect_unset_distributed_params(
              coordinator_address,
              num_processes,
              process_id,
              local_device_ids,
              cluster_detection_method,
              initialization_timeout,
          )
      )

    if coordinator_address is None:
      raise ValueError('coordinator_address should be defined.')
    if num_processes is None:
      raise ValueError('Number of processes must be defined.')
    if process_id is None:
      raise ValueError('The process id of the current process must be defined.')
    if not isinstance(process_id, int):
      raise TypeError("process_id must be a nonnegative int. "
                      f"Got process_id={process_id} of type {type(process_id)}.")
    if not isinstance(num_processes, int):
      raise TypeError("num_processes must be a positive int. "
                      f"Got num_processes={num_processes} of type {type(num_processes)}.")
    if not (0 <= process_id < num_processes):
      raise ValueError("process_id and num_processes must be nonnegative, with process_id < num_processes. "
                       f"Got process_id={process_id}, num_processes={num_processes}.")

    self.coordinator_address = coordinator_address

    # The default value of [::]:port tells the coordinator to bind to all
    # available addresses on the same port as coordinator_address.
    default_coordinator_bind_address = '[::]:' + coordinator_address.rsplit(':', 1)[1]
    coordinator_bind_address = (coordinator_bind_address or

View on GitHub (pinned to 1e1c6a8fc0)

Solutions

1. Pass num_processes equal to the launcher's world size, e.g. int(os.environ['NRANKS']) or WORLD_SIZE
2. Use a cluster_detection_method that infers process counts (slurm/torchrun)
3. Double-check that every process passes the same num_processes value

Example fix

# before
jax.distributed.initialize(coordinator_address=addr)
# after
jax.distributed.initialize(coordinator_address=addr,
    num_processes=int(os.environ['WORLD_SIZE']),
    process_id=int(os.environ['RANK']))

import os
np_ = int(os.environ.get('WORLD_SIZE', os.environ.get('NRANKS', 0)))
assert np_ > 0, 'num_processes unresolved'

When it happens

Trigger: jax.distributed.initialize(coordinator_address=...) without num_processes; passing num_processes=None explicitly.

Common situations: Single-script multi-GPU launches where the user assumed JAX infers world size; missing plumbing from torchrun/mpirun env vars (NRANKS/WORLD_SIZE).

Understand the failure class

Background: "X is required", "must be set", "cannot be empty": the missing-required-config error family, from Vertex AI project/location to WeChat keys — this error's family across 18 libraries.

Related errors

• coordinator_address should be defined.
• The process id of the current process must be defined.
• num_processes must be a positive int. Got num_processes={num
• Mapped away dimension of inputs passed to vmap should be sha
• invalid truth value {val!r} for environment {varname!r}

AI-assisted analysis of jax-ml/jax@1e1c6a8fc0 (2026-08-27). Data as JSON: /api/errors/efddeefc2771f1fa. Report an issue: GitHub.