🐛[BUG]: unrecognized input to lagged ensembles

[Yair Cohen]

Version

source - main

On which installation method(s) does this occur?

Pip

Describe the issue

following the instructions in lagged ensembles main:

torchrun --nproc_per_node 2 --nnodes 1 -m earth2mip.lagged_ensembles --model sfno_73ch --inits 10 --leads 5 --lags 4

produces the following error:

usage: Run a lagged ensemble scoring

    Can be run against either a fcn model (--model), a forecast directory as
    output by earth2mip.time_collection (--forecast_dir), persistence forecast
    (--persistence), or deterministic IFS (--ifs).

    Saves data as csv files (1 per rank).

    Examples:

        torchrun --nproc_per_node 2 --nnodes 1 -m earth2mip.lagged_ensembles --model sfno_73ch --inits 10 --leads 5 --lags 4

    
__main__.py: error: unrecognized arguments: --inits 10
usage: Run a lagged ensemble scoring

    Can be run against either a fcn model (--model), a forecast directory as
    output by earth2mip.time_collection (--forecast_dir), persistence forecast
    (--persistence), or deterministic IFS (--ifs).

    Saves data as csv files (1 per rank).

    Examples:

        torchrun --nproc_per_node 2 --nnodes 1 -m earth2mip.lagged_ensembles --model sfno_73ch --inits 10 --leads 5 --lags 4

    
__main__.py: error: unrecognized arguments: --inits 10
[2023-12-06 14:26:24,496] torch.distributed.elastic.multiprocessing.api: [ERROR] failed (exitcode: 2) local_rank: 0 (pid: 922229) of binary: /usr/bin/python
Traceback (most recent call last):
  File "/usr/local/bin/torchrun", line 8, in <module>
    sys.exit(main())
  File "/usr/local/lib/python3.10/dist-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
    return f(*args, **kwargs)
  File "/usr/local/lib/python3.10/dist-packages/torch/distributed/run.py", line 806, in main
    run(args)
  File "/usr/local/lib/python3.10/dist-packages/torch/distributed/run.py", line 797, in run
    elastic_launch(
  File "/usr/local/lib/python3.10/dist-packages/torch/distributed/launcher/api.py", line 134, in __call__
    return launch_agent(self._config, self._entrypoint, list(args))
  File "/usr/local/lib/python3.10/dist-packages/torch/distributed/launcher/api.py", line 264, in launch_agent
    raise ChildFailedError(
torch.distributed.elastic.multiprocessing.errors.ChildFailedError: 

it seem that inits is no longer an argument in parse_args - I think @nbren12 might have decided to make it a fix number rather than an input by users choice.

Environment details

running on Selene interactive session with `gitlab-master.nvidia.com/earth-2/fcn-mip:latest`

[Noah D. Brenowitz]

Try running –help. Thes args are more flexible now: https://github.com/NVIDIA/earth2mip/blob/a17fd31ae15b83a052c57c88eb30a153d2995415/earth2mip/_cli_utils.py#L25 From: Yair Cohen ***@***.***> Date: Wednesday, December 6, 2023 at 2:29 PM To: NVIDIA/earth2mip ***@***.***> Cc: Noah Brenowitz ***@***.***>, Mention ***@***.***> Subject: [NVIDIA/earth2mip] 🐛[BUG]: unrecognized input to lagged ensembles (Issue #143) Version source - main On which installation method(s) does this occur? Pip Describe the issue following the instructions in lagged ensembles main: torchrun --nproc_per_node 2 --nnodes 1 -m earth2mip.lagged_ensembles --model sfno_73ch --inits 10 --leads 5 --lags 4 produces the following error: usage: Run a lagged ensemble scoring Can be run against either a fcn model (--model), a forecast directory as output by earth2mip.time_collection (--forecast_dir), persistence forecast (--persistence), or deterministic IFS (--ifs). Saves data as csv files (1 per rank). Examples: torchrun --nproc_per_node 2 --nnodes 1 -m earth2mip.lagged_ensembles --model sfno_73ch --inits 10 --leads 5 --lags 4

…

__main__.py: error: unrecognized arguments: --inits 10 usage: Run a lagged ensemble scoring Can be run against either a fcn model (--model), a forecast directory as output by earth2mip.time_collection (--forecast_dir), persistence forecast (--persistence), or deterministic IFS (--ifs). Saves data as csv files (1 per rank). Examples: torchrun --nproc_per_node 2 --nnodes 1 -m earth2mip.lagged_ensembles --model sfno_73ch --inits 10 --leads 5 --lags 4

__main__.py: error: unrecognized arguments: --inits 10 [2023-12-06 14:26:24,496] torch.distributed.elastic.multiprocessing.api: [ERROR] failed (exitcode: 2) local_rank: 0 (pid: 922229) of binary: /usr/bin/python Traceback (most recent call last): File "/usr/local/bin/torchrun", line 8, in <module> sys.exit(main()) File "/usr/local/lib/python3.10/dist-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper return f(*args, **kwargs) File "/usr/local/lib/python3.10/dist-packages/torch/distributed/run.py", line 806, in main run(args) File "/usr/local/lib/python3.10/dist-packages/torch/distributed/run.py", line 797, in run elastic_launch( File "/usr/local/lib/python3.10/dist-packages/torch/distributed/launcher/api.py", line 134, in __call__ return launch_agent(self._config, self._entrypoint, list(args)) File "/usr/local/lib/python3.10/dist-packages/torch/distributed/launcher/api.py", line 264, in launch_agent raise ChildFailedError( torch.distributed.elastic.multiprocessing.errors.ChildFailedError: it seem that inits is no longer an argument in parse_args - I think @nbren12<https://github.com/nbren12> might have decided to make it a fix number rather than an input by users choice. Environment details running on Selene interactive session with `gitlab-master.nvidia.com/earth-2/fcn-mip:latest` — Reply to this email directly, view it on GitHub<#143>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AAKSREVMIGXJWWN3IISXY7TYIDWVNAVCNFSM6AAAAABAKDNRLGVHI2DSMVQWIX3LMV43ASLTON2WKOZSGAZDSNBZHA3TOMA>. You are receiving this because you were mentioned.Message ID: ***@***.***>

[Noah D. Brenowitz]

Closing since the --inits flag is replaced by --start-time and --end-time.