NeMo_Canary / docs /source /core /exp_manager.rst
Respair's picture
Upload folder using huggingface_hub
b386992 verified
.. _exp-manager-label:
Experiment Manager
==================
The NeMo Framework Experiment Manager leverages PyTorch Lightning for model checkpointing, TensorBoard Logging, Weights and Biases, DLLogger and MLFlow logging. The
Experiment Manager is included by default in all NeMo example scripts.
To use the Experiment Manager, call :class:`~nemo.utils.exp_manager.exp_manager` and pass in the PyTorch Lightning ``Trainer``.
.. code-block:: python
exp_dir = exp_manager(trainer, cfg.get("exp_manager", None))
The Experiment Manager is configurable using YAML with Hydra.
.. code-block:: bash
exp_manager:
exp_dir: /path/to/my/experiments
name: my_experiment_name
create_tensorboard_logger: True
create_checkpoint_callback: True
Optionally, launch TensorBoard to view the training results in ``exp_dir``, which by default is set to ``./nemo_experiments``.
.. code-block:: bash
tensorboard --bind_all --logdir nemo_experiments
..
If ``create_checkpoint_callback`` is set to ``True``, then NeMo automatically creates checkpoints during training
using PyTorch Lightning's `ModelCheckpoint <https://lightning.ai/docs/pytorch/stable/api/lightning.pytorch.callbacks.ModelCheckpoint.html>`_.
We can configure the ``ModelCheckpoint`` via YAML or CLI:
.. code-block:: yaml
exp_manager:
...
# configure the PyTorch Lightning ModelCheckpoint using checkpoint_call_back_params
# any ModelCheckpoint argument can be set here
# save the best checkpoints based on this metric
checkpoint_callback_params.monitor=val_loss
# choose how many total checkpoints to save
checkpoint_callback_params.save_top_k=5
Resume Training
---------------
To auto-resume training, configure the ``exp_manager``. This feature is important for long training runs that might be interrupted or
shut down before the procedure has completed. To auto-resume training, set the following parameters via YAML or CLI:
.. code-block:: yaml
exp_manager:
...
# resume training if checkpoints already exist
resume_if_exists: True
# to start training with no existing checkpoints
resume_ignore_no_checkpoint: True
# by default experiments will be versioned by datetime
# we can set our own version with
exp_manager.version: my_experiment_version
Experiment Loggers
------------------
Alongside Tensorboard, NeMo also supports Weights and Biases, MLFlow, DLLogger, ClearML and NeptuneLogger. To use these loggers, set the following
via YAML or :class:`~nemo.utils.exp_manager.ExpManagerConfig`.
Weights and Biases (WandB)
~~~~~~~~~~~~~~~~~~~~~~~~~~
.. _exp_manager_weights_biases-label:
.. code-block:: yaml
exp_manager:
...
create_checkpoint_callback: True
create_wandb_logger: True
wandb_logger_kwargs:
name: ${name}
project: ${project}
entity: ${entity}
<Add any other arguments supported by WandB logger here>
MLFlow
~~~~~~
.. _exp_manager_mlflow-label:
.. code-block:: yaml
exp_manager:
...
create_checkpoint_callback: True
create_mlflow_logger: True
mlflow_logger_kwargs:
experiment_name: ${name}
tags:
<Any key:value pairs>
save_dir: './mlruns'
prefix: ''
artifact_location: None
# provide run_id if resuming a previously started run
run_id: Optional[str] = None
DLLogger
~~~~~~~~
.. _exp_manager_dllogger-label:
.. code-block:: yaml
exp_manager:
...
create_checkpoint_callback: True
create_dllogger_logger: True
dllogger_logger_kwargs:
verbose: False
stdout: False
json_file: "./dllogger.json"
ClearML
~~~~~~~
.. _exp_manager_clearml-label:
.. code-block:: yaml
exp_manager:
...
create_checkpoint_callback: True
create_clearml_logger: True
clearml_logger_kwargs:
project: None # name of the project
task: None # optional name of task
connect_pytorch: False
model_name: None # optional name of model
tags: None # Should be a list of str
log_model: False # log model to clearml server
log_cfg: False # log config to clearml server
log_metrics: False # log metrics to clearml server
Neptune
~~~~~~~
.. _exp_manager_neptune-label:
.. code-block:: yaml
exp_manager:
...
create_checkpoint_callback: True
create_neptune_logger: false
neptune_logger_kwargs:
project: ${project}
name: ${name}
prefix: train
log_model_checkpoints: false # set to True if checkpoints need to be pushed to Neptune
tags: null # can specify as an array of strings in yaml array format
description: null
<Add any other arguments supported by Neptune logger here>
Exponential Moving Average
--------------------------
.. _exp_manager_ema-label:
NeMo supports using exponential moving average (EMA) for model parameters. This can be useful for improving model generalization
and stability. To use EMA, set the following parameters via YAML or :class:`~nemo.utils.exp_manager.ExpManagerConfig`.
.. code-block:: yaml
exp_manager:
...
# use exponential moving average for model parameters
ema:
enabled: True # False by default
decay: 0.999 # decay rate
cpu_offload: False # If EMA parameters should be offloaded to CPU to save GPU memory
every_n_steps: 1 # How often to update EMA weights
validate_original_weights: False # Whether to use original weights for validation calculation or EMA weights
.. Support for Preemption
----------------------
.. _exp_manager_preemption_support-label:
NeMo adds support for a callback upon preemption while running the models on clusters. The callback takes care of saving the current state of training via the ``.ckpt``
file followed by a graceful exit from the run. The checkpoint saved upon preemption has the ``*last.ckpt`` suffix and replaces the previously saved last checkpoints.
This feature is useful to increase utilization on clusters.
The ``PreemptionCallback`` is enabled by default. To disable it, add ``create_preemption_callback: False`` under exp_manager in the config YAML file.
Stragglers Detection
----------------------
.. _exp_manager_straggler_det_support-label:
.. note::
Stragglers Detection feature is included in the optional NeMo resiliency package.
Distributed training can be affected by stragglers, which are workers that slow down the overall training process.
NeMo provides a straggler detection feature that can identify slower GPUs.
This feature is implemented in the ``StragglerDetectionCallback``, which is disabled by default.
The callback computes normalized GPU performance scores, which are scalar values ranging from 0.0 (worst) to 1.0 (best).
A performance score can be interpreted as the ratio of current performance to reference performance.
There are two types of performance scores provided by the callback:
* Relative GPU performance score: The best-performing GPU in the workload is used as a reference.
* Individual GPU performance score: The best historical performance of the GPU is used as a reference.
Examples:
* If the relative performance score is 0.5, it means that a GPU is twice slower than the fastest GPU.
* If the individual performance score is 0.5, it means that a GPU is twice slower than its best observed performance.
If a GPU performance score drops below the specified threshold, it is identified as a straggler.
To enable straggler detection, add ``create_straggler_detection_callback: True`` under exp_manager in the config YAML file.
You might also want to adjust the callback parameters:
.. code-block:: yaml
exp_manager:
...
create_straggler_detection_callback: True
straggler_detection_callback_params:
report_time_interval: 300 # Interval [seconds] of the straggler check
calc_relative_gpu_perf: True # Calculate relative GPU performance
calc_individual_gpu_perf: True # Calculate individual GPU performance
num_gpu_perf_scores_to_log: 5 # Log 5 best and 5 worst GPU performance scores, even if no stragglers are detected
gpu_relative_perf_threshold: 0.7 # Threshold for relative GPU performance scores
gpu_individual_perf_threshold: 0.7 # Threshold for individual GPU performance scores
stop_if_detected: True # Terminate the workload if stragglers are detected
Straggler detection may require inter-rank synchronization and should be performed at regular intervals, such as every few minutes.
.. Fault Tolerance
---------------
.. _exp_manager_fault_tolerance_support-label:
.. note::
Fault Tolerance feature is included in the optional NeMo resiliency package.
When training Deep Neural Network (DNN models), faults may occur, hindering the progress of the entire training process.
This is particularly common in distributed, multi-node training scenarios, with many nodes and GPUs involved.
NeMo incorporates a fault tolerance mechanism to detect training halts.
In response, it can terminate a hung workload and, if requested, restart it from the last checkpoint.
Fault tolerance ("FT") relies on a special launcher (``ft_launcher``), which is a modified ``torchrun``.
The FT launcher runs background processes called rank monitors. **You need to use ft_launcher to start
your workload if you are using FT**. I.e., `NeMo-Framework-Launcher <https://github.com/NVIDIA/NeMo-Framework-Launcher>`_
can be used to generate SLURM batch scripts with FT support.
Each training process (rank) sends `heartbeats` to its monitor during training and validation steps.
If a rank monitor stops receiving `heartbeats`, a training failure is detected.
Fault detection is implemented in the ``FaultToleranceCallback`` and is disabled by default.
To enable it, add a ``create_fault_tolerance_callback: True`` option under ``exp_manager`` in the
config YAML file. Additionally, you can customize FT parameters by adding ``fault_tolerance`` section:
.. code-block:: yaml
exp_manager:
...
create_fault_tolerance_callback: True
fault_tolerance:
initial_rank_heartbeat_timeout: 600 # wait for 10 minutes for the initial heartbeat
rank_heartbeat_timeout: 300 # wait for 5 minutes for subsequent heartbeats
calculate_timeouts: True # estimate more accurate timeouts based on observed intervals
Timeouts for fault detection need to be adjusted for a given workload:
* ``initial_rank_heartbeat_timeout`` should be long enough to allow for workload initialization.
* ``rank_heartbeat_timeout`` should be at least as long as the longest possible interval between steps.
**Importantly, `heartbeats` are not sent during checkpoint loading and saving**, so time for
checkpointing related operations should be taken into account.
If ``calculate_timeouts: True``, timeouts will be automatically estimated based on observed intervals.
Estimated timeouts take precedence over timeouts defined in the config file. **Timeouts are estimated
at the end of a training run when checkpoint loading and saving were observed.** Hence, in a multi-part
training started from scratch, estimated timeouts won't be available during the initial two runs.
Estimated timeouts are stored in a separate JSON file.
``max_subsequent_job_failures`` allows for the automatic continuation of training on a SLURM cluster.
This feature requires SLURM job to be scheduled with ``NeMo-Framework-Launcher``. If ``max_subsequent_job_failures``
value is `>0` continuation job is prescheduled. It will continue the work until ``max_subsequent_job_failures``
subsequent jobs failed (SLURM job exit code is `!= 0`) or the training is completed successfully
("end of training" marker file is produced by the ``FaultToleranceCallback``, i.e. due to iters or time limit reached).
All FT configuration items summary:
* ``workload_check_interval`` (float, default=5.0) Periodic workload check interval [seconds] in the workload monitor.
* ``initial_rank_heartbeat_timeout`` (Optional[float], default=60.0 * 60.0) Timeout [seconds] for the first heartbeat from a rank.
* ``rank_heartbeat_timeout`` (Optional[float], default=45.0 * 60.0) Timeout [seconds] for subsequent heartbeats from a rank.
* ``calculate_timeouts`` (bool, default=True) Try to calculate ``rank_heartbeat_timeout`` and ``initial_rank_heartbeat_timeout``
based on the observed heartbeat intervals.
* ``safety_factor``: (float, default=5.0) When calculating the timeouts, multiply the maximum observed heartbeat interval
by this factor to obtain the timeout estimate. Can be made smaller for stable environments and larger for unstable ones.
* ``rank_termination_signal`` (signal.Signals, default=signal.SIGKILL) Signal used to terminate the rank when failure is detected.
* ``log_level`` (str, default='INFO') Log level for the FT client and server(rank monitor).
* ``max_rank_restarts`` (int, default=0) Used by FT launcher. Max number of restarts for a rank.
If ``>0`` ranks will be restarted on existing nodes in case of a failure.
* ``max_subsequent_job_failures`` (int, default=0) Used by FT launcher. How many subsequent job failures are allowed until stopping autoresuming.
``0`` means do not auto-resume.
* ``additional_ft_launcher_args`` (str, default='') Additional FT launcher params (for advanced use).
.. _nemo_multirun-label:
Hydra Multi-Run with NeMo
-------------------------
When training neural networks, it is common to perform a hyperparameter search to improve the model’s performance on validation data.
However, manually preparing a grid of experiments and managing all checkpoints and their metrics can be tedious.
To simplify these tasks, NeMo integrates with `Hydra Multi-Run support <https://hydra.cc/docs/tutorials/basic/running_your_app/multi-run/>`_,
providing a unified way to run a set of experiments directly from the configuration.
There are certain limitations to this framework, which we list below:
* All experiments are assumed to be run on a single GPU, and multi GPU for single run (model parallel models are not supported as of now).
* NeMo Multi-Run currently supports only grid search over a set of hyperparameters. Support for advanced hyperparameter search strategies will be added in the future.
* **NeMo Multi-Run requires one or more GPUs** to function and will not work without GPU devices.
Config Setup
~~~~~~~~~~~~
In order to enable NeMo Multi-Run, we first update our YAML configs with some information to let Hydra know we expect to run multiple experiments from this one config -
.. code-block:: yaml
# Required for Hydra launch of hyperparameter search via multirun
defaults:
- override hydra/launcher: nemo_launcher
# Hydra arguments necessary for hyperparameter optimization
hydra:
# Helper arguments to ensure all hyper parameter runs are from the directory that launches the script.
sweep:
dir: "."
subdir: "."
# Define all the hyper parameters here
sweeper:
params:
# Place all the parameters you wish to search over here (corresponding to the rest of the config)
# NOTE: Make sure that there are no spaces between the commas that separate the config params !
model.optim.lr: 0.001,0.0001
model.encoder.dim: 32,64,96,128
model.decoder.dropout: 0.0,0.1,0.2
# Arguments to the process launcher
launcher:
num_gpus: -1 # Number of gpus to use. Each run works on a single GPU.
jobs_per_gpu: 1 # If each GPU has large memory, you can run multiple jobs on the same GPU for faster results (until OOM).
Next, we will setup the config for ``Experiment Manager``. When we perform hyper parameter search, each run may take some time to complete.
We want to therefore avoid the case where a run ends (say due to OOM or timeout on the machine) and we need to redo all experiments.
We therefore setup the experiment manager config such that every experiment has a unique "key", whose value corresponds to a single
resumable experiment.
Let us see how to setup such a unique "key" via the experiment name. Simply attach all the hyper parameter arguments to the experiment
name as shown below -
.. code-block:: yaml
exp_manager:
exp_dir: null # Can be set by the user.
# Add a unique name for all hyper parameter arguments to allow continued training.
# NOTE: It is necessary to add all hyperparameter arguments to the name !
# This ensures successful restoration of model runs in case HP search crashes.
name: ${name}-lr-${model.optim.lr}-adim-${model.adapter.dim}-sd-${model.adapter.adapter_strategy.stochastic_depth}
...
checkpoint_callback_params:
...
save_top_k: 1 # Dont save too many .ckpt files during HP search
always_save_nemo: True # saves the checkpoints as nemo files for fast checking of results later
...
# We highly recommend use of any experiment tracking took to gather all the experiments in one location
create_wandb_logger: True
wandb_logger_kwargs:
project: "<Add some project name here>"
# HP Search may crash due to various reasons, best to attempt continuation in order to
# resume from where the last failure case occurred.
resume_if_exists: true
resume_ignore_no_checkpoint: true
Run a NeMo Multi-Run Configuration
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Once the config has been updated, we can now run it just like any normal Hydra script, with one special flag (``-m``).
.. code-block:: bash
python script.py --config-path=ABC --config-name=XYZ -m \
trainer.max_steps=5000 \ # Any additional arg after -m will be passed to all the runs generated from the config !
...
Tips and Tricks
---------------
This section provides recommendations for using the Experiment Manager.
Preserving disk space for a large number of experiments
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Some models may have a large number of parameters, making it very expensive to save numerous checkpoints on physical storage drives.
For example, if you use the Adam optimizer, each PyTorch Lightning ".ckpt" file will be three times the size of just the model
parameters. This can become exorbitant if you have multiple runs.
In the above configuration, we explicitly set ``save_top_k: 1`` and ``always_save_nemo: True``. This limits the number of ".ckpt"
files to just one and also saves a NeMo file, which contains only the model parameters without the optimizer state.
This NeMo file can be restored immediately for further work.
We can further save storage space by using NeMo's utility functions to automatically delete either ".ckpt" or NeMo files
after a training run has finished. This is sufficient if you are collecting results in an experiment tracking tool and can
simply rerun the best configuration after the search is completed.
.. code-block:: python
# Import `clean_exp_ckpt` along with exp_manager
from nemo.utils.exp_manager import clean_exp_ckpt, exp_manager
@hydra_runner(...)
def main(cfg):
...
# Keep track of the experiment directory
exp_log_dir = exp_manager(trainer, cfg.get("exp_manager", None))
... add any training code here as needed ...
# Add following line to end of the training script
# Remove PTL ckpt file, and potentially also remove .nemo file to conserve storage space.
clean_exp_ckpt(exp_log_dir, remove_ckpt=True, remove_nemo=False)
Debugging Multi-Run Scripts
~~~~~~~~~~~~~~~~~~~~~~~~~~~
When running Hydra scripts, you may encounter configuration issues that crash the program. In NeMo Multi-Run, a crash in
any single run will not crash the entire program. Instead, we will note the error and proceed to the next job. Once all
jobs are completed, we will raise the errors in the order they occurred, crashing the program with the first error’s stack trace.
To debug NeMo Multi-Run, we recommend commenting out the entire hyperparameter configuration set inside ``sweep.params``.
Instead, run a single experiment with the configuration, which will immediately raise the error.
Experiment name cannot be parsed by Hydra
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Sometimes our hyperparameters include PyTorch Lightning ``trainer`` arguments, such as the number of steps, number of epochs,
and whether to use gradient accumulation. When we attempt to add these as keys to the experiment manager's ``name``,
Hydra may complain that ``trainer.xyz`` cannot be resolved.
A simple solution is to finalize the Hydra config before you call ``exp_manager()`` as follows:
.. code-block:: python
@hydra_runner(...)
def main(cfg):
# Make any changes as necessary to the config
cfg.xyz.abc = uvw
# Finalize the config
cfg = OmegaConf.resolve(cfg)
# Carry on as normal by calling trainer and exp_manager
trainer = pl.Trainer(**cfg.trainer)
exp_log_dir = exp_manager(trainer, cfg.get("exp_manager", None))
...
ExpManagerConfig
----------------
.. autoclass:: nemo.utils.exp_manager.ExpManagerConfig
:show-inheritance:
:members:
:member-order: bysource
:noindex: