Skip to content

Releases: mosaicml/composer

v0.12.0

23 Dec 00:13
@bandish-shah bandish-shah
v0.12.0
d2582ec
Compare
Choose a tag to compare
v0.12.0

πŸš€ Composer v0.12.0

Composer v0.12.0 is released! Install via pip:

pip install mosaicml==0.12.0

New Features

  1. πŸͺ΅ Logging and ObjectStore Enhancements

    There are multiple improvements to our logging and object store support in this release.

    • Image visualization using our CometMLLogger (#1710)

      We've added support for using our ImageVisualizer callback with CometML to log images and segmentation masks to CometML.

      from composer.trainer import Trainer

trainer = Trainer(...,
    callbacks=[ImageVisualizer()],
    loggers=[CometMLLogger()]
)

    • Added direct support for Oracle Cloud Infrastructure (OCI) as an ObjectStore (#1774) and support for Google Cloud Storage (GCS) via URI (#1833)

      To use, you can simply set your save_folder or load_path to a URI beginning with oci:// or gs://, to save and load with OCI and GCS respectively.

      from composer.trainer import Trainer

# Checkpoint saving to Google Cloud Storage.
trainer = Trainer(
    model=model,
    save_folder="gs://my-bucket/{run_name}/checkpoints",
    run_name='my-run',
    save_interval="1ep",
    save_filename="ep{epoch}.pt",
    save_num_checkpoints_to_keep=0,  # delete all checkpoints locally
    ...
)

trainer.fit()

    • Added basic support for logging with MLFlow (#1795)

      We've added basic support for using MLFlow to log experiment metrics.

      from composer.loggers import MLFlowLogger
from composer.trainer import Trainer

mlflow_logger = MLFlowLogger(experiment_name=mlflow_exp_name,
                             run_name=mlflow_run_name,
                             tracking_uri=mlflow_uri)
trainer = Trainer(..., loggers=[mlflow_logger])

    • Simplified console and progress bar logging (#1694)

      To turn off the progress bar, set progress_bar=False. To turn on logging directly to the console, set log_to_console=True. To control the frequency of logging to console, set console_log_interval (e.g. to 1ep or 1ba).

    • getfile supports URIs (#1750)

      Our get_file utility now supports URIs directly (s3://, oci://, and gs://) for downloading files.

  2. πŸƒβ€β™€οΈ Support for Mid-Epoch Resumption with the latest release of Streaming

    We've added support in Composer for the latest release of our Streaming library. This includes awesome new features like instant mid epoch resumption and deterministic shuffling, regardless of the number of nodes. See the Streaming release notes for more!

  3. 🚨 New algorithm - GyroDropout!

    Thanks to @jelite for adding a new algorithm, GyroDropout to Composer! Please see the method card for more details.

  4. πŸ€— HuggingFace + Composer improvements

    We've added a new utility to load a πŸ€— HuggingFace model and tokenizer out of a Composer checkpoint (#1754), making the pretraining -> finetuning workflow even easier in Composer. Check out the docs for more details, and our example notebook for a full tutorial (#1775)!

  5. πŸŽ“ GradMonitor -> OptimizerMonitor

    Renames our GradMonitor callback to OptimizerMonitor, and adds the ability to track optimizer specific metrics. Check out the docs for more details, and add to your code just like any other callback!

    from composer.callbacks import OptimizerMonitor
from composer.trainer import Trainer

trainer = Trainer(
    ..., 
    callbacks=[OptimizerMonitor(log_optimizer_metrics=log_optimizer_metrics)]
)

  6. 🐳 New PyTorch and CUDA versions

    We've expanded our library of Docker images with support for PyTorch 1.13 + CUDA 11.7:

    • mosaicml/pytorch:1.13.0_cu117-python3.10-ubuntu20.04
    • mosaicml/pytorch:1.13.0_cpu-python3.10-ubuntu20.04

    The mosaicml/pytorch:latest, mosaicml/pytorch:cpu_latest and mosaicml/composer:0.12.0 tags are now built from PyTorch 1.13 based images. Please see our DockerHub repository for additional details.

API changes

  1. Replace grad_accum with device_train_microbatch_size (#1749, #1776)

    We're deprecating the grad_accum Trainer argument in favor of the more intuitive device_train_microbatch_size. Instead of thinking about how to divide your specified minibatch into microbatches, simply specify the size of your microbatch. For example, let's say you want to split your minibatch of 2048 into two microbatches of 1024:

    from composer import Trainer

trainer = Trainer(
    ...,
    device_train_microbatch_size=1024,
)

    If you want Composer to tune the microbatch for you automatically, enable automatic microbatching as follows:

    from composer import Trainer

trainer = Trainer(
    ...,
    device_train_microbatch_size='auto',
)

    The grad_accum argument is still supported but will be deprecated in the next Composer release.

  2. Renamed precisions (#1761)

    We've renamed precision attributes for clarity. The following values have been removed: ['amp', 'fp16', bf16'].

    We have added the following values, prefixed with 'amp' to clarify when an Automatic Mixed Precision type is being used: ['amp_fp16', 'amp_bf16'].

    The fp32 precision value remains unchanged.

Deprecations

  1. Removed support for YAHP (#1512)
  2. Removed COCO and SSD datasets (#1717)
  3. Fully removed Streaming v1 support, please see the mosaicml/streaming project for our next-gen streaming datasets (#1787)
  4. Deprecated FusedLayerNorm algorithm (#1789)
  5. Fully removed grad_clip_norm training argument, please use the GradientClipping algorithm instead (#1768)
  6. Removed data_fit, data_epoch, and data_batch from Logger (#1826)

Bug Fixes

  • Fix FSDP checkpoint strategy (#1734)
  • Fix gradient clipping with FSDP (#1740)
  • Adds more supported FSDP config flags (sync_module_states, forward_prefecth, limit_all_gathers) (#1794)
  • Allow FULL precision with FSDP (#1796)
  • Fix eval_microbatch modification on EVAL_BEFORE_FORWARD event (#1739)
  • Fix algorithm API backwards compatibility in checkpoints (#1741)
  • Fixes a bad None check preventing setting device_id to 0 (#1767)
  • Unregister engine to make cleaning up memory easier (#1769)
  • Fix issue if metric_names is not a list (#1798)
  • Match implementation for list and tensor batch splitting (#1804)
  • Fixes infinite eval issue (#1815)

What's Changed

Read more

Contributors

cojennin, knighton, and 17 other contributors
Assets 2
Loading

v0.11.1

16 Nov 21:41
@bandish-shah bandish-shah
v0.11.1
e470861
Compare
Choose a tag to compare
v0.11.1

πŸš€ Composer v0.11.1

Composer v0.11.1 is released! Install via pip:

pip install --upgrade mosaicml==0.11.1

Bug Fixes

  • Fixes for Notebooks (#1659)
  • Documentation updates and fixes (#1685, #1696, #1702, #1709)
  • Addressed warnings and speed improvements for Torchmetrics (#1674)
  • Fixes to Gated Linear Units method (#1575, #1689)
  • Set NCCL_ASYNC_ERROR_HANDLING ENV variable in Composer launcher to enable distributed timeout (#1695)
  • Fix epoch count when eval is called before fit (#1697)
  • Constrain PyTorch package versions to avoid unintended upgrades (#1688)
  • Fix Optimizer state sharding issue with FSDP (#1732)
  • Rase ValueError with if evaluation dataloader of infinite length is specified

Full Changelog: v0.11.0...v0.11.1

Assets 2
Loading

v0.11.0

25 Oct 00:36
@bandish-shah bandish-shah
v0.11.0
f5f9571
Compare
Choose a tag to compare
v0.11.0

πŸš€ Composer v0.11.0

Composer v0.11.0 is released! Install via pip:

pip install --upgrade mosaicml==0.11.0

New Features

  1. 🧰 FSDP Beta Support

    Composer now supports PyTorch FSDP! PyTorch FSDP is a strategy for distributed training, similar to PyTorch DDP, that distributes work using data-parallelism only. On top of this, FSDP uses model, gradient, and optimizer sharding to dramatically reduce device memory requirements, and enables users to easily scale and train large models.

    Here's how easy it is to use FSDP with Composer:

    import torch.nn as nn
from composer import Trainer

class Block (nn.Module):
    ...

# Your custom model
class Model(nn.Module):
    def __init__(self, n_layers):
        super().__init__()
        self.blocks = nn.ModuleList([
            Block(...) for _ in range(n_layers)
        ]),
        self.head = nn.Linear(...)
    def forward(self, inputs):
        ...

    # FSDP Wrap Function
    def fsdp_wrap_fn(self, module):
        return isinstance(module, Block)

    # Activation Checkpointing Function
    def activation_checkpointing_fn(self, module):
        return isinstance(module, Block)

# ComposerModel wrapper, used by the Trainer
# to compute loss, metrics, etc.
class MyComposerModel(ComposerModel):

    def __init__(self, n_layers):
        super().__init__()
        self.model = Model(n_layers)
        ...

    def forward(self, batch):
        ...

    def eval_forward(self, batch, outputs=None):
        ...

    def loss(self, outputs, batch):
        ...

# Pass your ComposerModel and fsdp_config into the Trainer
composer_model = MyComposerModel(n_layers=3)
fsdp_config = {
    'sharding_strategy': 'FULL_SHARD',
    'min_params': 1e8,
    'cpu_offload': False, # Not supported yet
    'mixed_precision': 'DEFAULT',
    'backward_prefetch': 'BACKWARD_POST',
    'activation_checkpointing': False,
    'activation_cpu_offload': False,
    'verbose': True
}

trainer = Trainer(
    model=composer_model,
    fsdp_config=fsdp_config,
    ...
)

trainer.fit()

    For more information, please see our FSDP docs.

  2. 🚰 Streaming v0.1

    We've spun off Streaming datasets into it's own repository! Streaming datasets is a high-performance drop-in for TorchΒ IterableDataset, enabling users to stream training data from cloud based object stores. Streaming is shipping with built-in support for popular open source datasets (ADE20K, C4, COCO, Enwiki, ImageNet, etc.)

    To get started, install the Streaming PyPi package:

    pip install mosaicml-streaming

    You can use the streaming Dataset class with the PyTorch native DataLoader class as follows:

    import torch
from streaming import Dataset

dataloader = torch.utils.data.DataLoader(dataset=Dataset(remote='s3://...'))

    For more information, please check out the Streaming docs.

  3. βœ”πŸ‘‰ Simplified Checkpointing Interface

    With this release we’ve greatly simplified configuration of loading and saving checkpoints in Composer.

    To save checkpoints to S3, all you need to do is:

    • Specify with save_folder your full URI to your save directory destination (e.g. 's3://my-bucket/{run_name}/checkpoints')
    • Optionally, set save_filename to the pattern you want for your checkpoint file names
    from composer.trainer import Trainer

# Checkpoint saving to S3.
trainer = Trainer(
    model=model,
    save_folder="s3://my-bucket/{run_name}/checkpoints",
        run_name='my-run',
    save_interval="1ep",
    save_filename="ep{epoch}.pt",
    save_num_checkpoints_to_keep=0,  # delete all checkpoints locally
        ...
)

trainer.fit()

    Likewise, to load checkpoints from S3, all you have to do is:

    • Set load_path to the full URI to your desired checkpoint file (e.g.'s3://my-bucket/my-run/checkpoints/epoch13.pt')
    from composer.trainer import Trainer

# Checkpoint loading from S3.
new_trainer = Trainer(
    model=model,
    train_dataloader=train_dataloader,
    max_duration="10ep",
    load_path="s3://my-bucket/my-run/checkpoints/ep13.pt",
   )

    new_trainer.fit()

    For more information, please see our Checkpointing guide.

  4. 𐄳 Improved Distributed Experience

    We’ve made it easier to write your own custom distributed entry points by exposing our distributed API. You can now leverage all of our helpful distributed functions and contexts.

    For example, let's say we want to need to download a dataset in a distributed training application. To avoid race conditions where different ranks try to write the dataset to the same place, we need to ensure that only rank 0 downloads the dataset first:

    import datetime
from composer.trainer.devices import DeviceGPU
from composer.utils import dist

dist.initialize(DeviceGPU(), datetime.timedelta(seconds=30)) # Initialize distributed module

if dist.get_local_rank() == 0: # Download dataset on rank zero
    dataset = download_my_dataset()
dist.barrier() # All ranks wait until dataset is downloaded

# Create and train your model!

    For more information, please check out our Distributed API docs.

Bug Fixes

  • fix loss and eval_forward for HF models (#1597)
  • add more robust casting to int for fsdp min_params (#1608)
  • Deepspeed Docs Typo (#1605)
  • Fix mmdet typo (#1618)
  • Blurpool idempotent (#1625)
  • When model is not on meta device, initialization should occur on compute device not CPU (#1623)
  • Auto resumption (#1615)
  • Adjust speed monitor (#1645)
  • Hot fix console logging (#1643)
  • Lazy Logging + pretty print dict for hparams (#1653)
  • Fix many failing notebook tests (#1646)

What's Changed

Read more

Contributors

nqn, dblalock, and 14 other contributors
Assets 2
Loading

v0.10.1

06 Oct 00:17
@bandish-shah bandish-shah
v0.10.1
01eb228
Compare
Choose a tag to compare
v0.10.1

πŸš€ Composer v0.10.1

Composer v0.10.1 is released! Install via pip:

pip install --upgrade mosaicml==0.10.1

New Features

  1. 𐄷 Weight Standardization

    Weight Standardization reparametrizes convolutional weights such that the fan-in dimensions have zero mean and unit standard deviation. This could slightly improve performance at the expensive of 5% lower throughput. This has been used in several papers to train with smaller batch sizes, with normalization layers besides batch norm, and for transfer learning.

    Using Weight Standardization with the Composer Trainer:

    import composer
 
# Apply Weight Standardization (when training is initialized)
weight_std = composer.algorithms.WeightStandardization()

# Train with Weight Standardization
trainer = composer.trainer.Trainer(
    ...
    algorithms=[weight_std]
)
trainer.fit()

    Using Weight Standardization with the Composer functional interface:

    import composer
from torchvision.models import resnet50
 
my_model = resnet50()
 
# Apply weight standardization to model
my_model = composer.functional.weight_standardization(my_model)

    Please see the Weight Standardization Method Card for more details.

Bug Fixes

  • Fix for checkpoints not being saved automatically at the end of a run (#1552)
  • Fix Onnx export for Composer HuggingFaceModels (#1557)
  • Fix for MIoU metric producing NaN's (#1558)
  • CometML logger documentation updates and fixes (#1567, #1570, #1571)
  • WandB image visualizer fix (#1591)

What's Changed

New Contributors

Full Changelog: v0.10.0...v0.10.1

Contributors

growlix, karan6181, and 10 other contributors
Assets 2
Loading

v0.10.0

22 Sep 06:25
@bandish-shah bandish-shah
v0.10.0
1937d2e
Compare
Choose a tag to compare
v0.10.0

πŸš€ Composer v0.10.0

Composer v0.10.0 is out! This latest release adds support for CometML Experiment tracking, automatic selection of evaluation batch size, API enhancements for Evaluation/Logging/Metrics and a preview of our new streaming datasets repository!

pip install --upgrade mosaicml==0.10.0

New Features

  1. β˜„οΈ Comet Experiment Tracking (#1490)

    We've added support for the popular Comet experiment tracker! To enable, simply create the logger and pass it to the Trainer object at initialization:

    from composer import Trainer
from composer.loggers import CometMLLogger

cometml_logger = CometMLLogger()

trainer = Trainer(
    ...
    loggers=[cometml_logger],
)

    Please see our Logging and CometMLLogger docs pages for details on usage.

  2. πŸͺ„ Automatic Evaluation Batch Size Selection (#1417)

    Composer now supports eval_batch_size='auto', which will choose the right evaluation batch size to avoid CUDA OOMs! Now, in conjunction with grad_accum='auto', you can run the same code on any hardware with no changes necessary. This makes it easy to add evaluation to a training script without having to pick and choose the right batch sizes to avoid CUDA OOMs.

  3. 🎯 Evaluation API Changes (#1479)

    The Evaluation API has been updated to be consistent with the Trainer API. If the eval_dataloader was provided to the Trainer during initialization, eval can be invoked without needing to provide anything additional:

    trainer = Trainer(
    eval_dataloader=...
)
trainer.eval()

    Alternatively, the eval_dataloader can be passed directly to the eval() method:

    trainer = Trainer(
    ...
)
trainer.eval(
    eval_dataloader=...
)

    The eval_dataloader can be a pytorch dataloader, or for multiple metrics, a list of Evaluator objects.

  4. πŸͺ΅ Simplified Logging (#1416)

    We've significantly simplified our internal logging interface:

    • Removed the use of LogLevel throughout the logging, which was a mostly unused feature. Filtering logs are the responsibility of the logger.
    • For better compatibility with external logging interfaces such as CometML or Weights & Biases, loggers now support the following methods: log_metrics, log_hyperparameters, and log_artifacts. Previous calls to data_fit, data_epeoch, .. have been removed.

  5. 🎯 validate --> eval_forward (#1411 , #1419)

    Previously, ComposerModel implemented the validate(batch: Any) -> Tuple[Any, Any] method which returns an (input, target) tuple, and the Trainer handles updating the metrics. In v0.10, we return the metrics updating control to the user.

    Now, models instead implement def eval_forward(batch: Any) which returns the outputs of evaluation, and also def update_metric(batch, outputs, metric) which updates the metric.

    An example implementation for classification can be found in our ComposerClassifer base class:

        def update_metric(self, batch: Any, outputs: Any, metric: Metric) -> None:
        _, targets = batch
        metric.update(outputs, targets)

    def eval_forward(self, batch: Any, outputs: Optional[Any] = None) -> Any:
        return outputs if outputs is not None else self.forward(batch)

  6. πŸ•΅οΈβ€β™€οΈ Evaluator changes

    The Evaluator class now stores evaluation metric names instead of metric instances. For example:

    glue_mrpc_task = Evaluator(
    label='glue_mrpc',
    dataloader=mrpc_dataloader,
    metric_names=['BinaryF1Score', 'Accuracy']
)

    These metric names are matched against the metrics returned by the ComposerModel. The metric instances are now stored as deep copies in the State class as state.train_metrics or state.eval_metrics.

  7. 🚧 Streaming Datasets Repository Preview

    We're in the process of splitting out streaming datasets into it's own repository! Streaming datasets is a high-performance drop-in replacement for Torch IterableDataset objects and enables you to stream your training data from cloud based object stores. For an early preview, please checkout the Streaming repo.

  8. ❌ YAHP deprecation

    We are deprecating support for yahp, our hyperparameter configuration tool. Support for this will be removed in the following minor version release of Composer. We recommend users migrate to OmegaConf, or Hydra as tools.

Bug Fixes

What's Changed

Read more

Contributors

moinnadeem, nqn, and 16 other contributors
Assets 2
Loading

v0.9.0

16 Aug 06:11
@bandish-shah bandish-shah
v0.9.0
8ff8ec3
Compare
Choose a tag to compare
v0.9.0

πŸš€ Composer v0.9.0

Excited to share the release of Composer v0.9.0, which comes with an Inference Export API, beta support for Apple Silicon and TPU training, as well as expanded usability of NLP-related speed-up methods. This release includes 175 commits from 34 contributors, including 10 new contributors πŸ™Œ !

pip install --upgrade mosaicml==0.9.0

Alternatively, install Composer with Conda:

conda install -c mosaicml mosaicml=0.9.0

New Features

  1. πŸ“¦ Export for inference APIs

    Train with Composer and deploy anywhere! We have added a dedicated export API as well as an export training callback to allow you to export Composer-trained models for inference, supporting popular formats such as torchscript and ONNX.

    For example, here’s how to export a model in torchscript format:

    from composer.utils import export_for_inference

# Invoking export with a trained model
export_for_inference(model=model, 
                     save_format='torchscript', 
                     save_path=model_save_path)

    Here’s an example of using the training callback, which automatically exports the model at the end of training to ONNX format:

    from composer.callbacks import ExportForInferenceCallback

# Initializing Trainer with the export callback
callback = ExportForInferenceCallback(save_format='onnx', 
                                                                            save_path=model_save_path)
trainer = Trainer(model=model,
                                callbacks=callback,
                                train_dataloader=dataloader,
                                max_duration='10ep')

# Model will be exported at the end of training
trainer.fit()

    Please see our Exporting for Inference notebook for more information.

  2. πŸ“ˆ ALiBi support for BERT training

    You can now use ALiBi (Attention with Linear Biases; Press et al., 2021) when training BERT models with Composer, delivering faster training and higher accuracy by leveraging shorter sequence lengths.

    ALiBi improves the quality of BERT pre-training, especially when pre-training uses shorter sequence lengths than the downstream (fine-tuning) task. This allows models with ALiBi to reach higher downstream accuracy with less pre-training time.

    Example of using ALiBi as an algorithm with the Composer Trainer:

    # Create an instance of a BERT masked language model
model = composer.models.create_bert_mlm()

# Apply ALiBi (when training is initialized)
alibi = composer.algorithms.alibi(max_sequence_length=1024)

# Train with ALiBi
trainer = composer.trainer.Trainer(
    model=model,
    train_dataloader=train_dataloader,
    algorithms=[alibi]
)
trainer.fit()

    Example using the Composer Functional API:

    import composer.functional as cf

# Create an instance of a BERT masked language model
model = composer.models.create_bert_mlm()

# Apply ALiBi and expand the model's maximum sequence length to 1024
cf.apply_alibi(model=model, max_sequence_length=1024)

    AliBi can also now be extended to work with custom models by registering your attention and embedding layers. Please see our ALiBi method card for more information.

  3. 🧐 Entry point for GLUE tasks pre-training and fine-tuning

    You can now easily pre-train and fine-tune NLP models across all GLUE (General Language Understanding Evaluation) tasks through one simple entry point! The entry point handles model saving and loading, spawns GLUE tasks in parallel across all available GPUs, and delivers a highly efficient evaluation of model performance.

    Example of launching the entrypoint:

    # This runs pre-training followed by fine-tuning.
# --training_scheme can take either pretrain, finetune, or all depending on the task!
python run_glue_trainer.py -f glue_example.yaml --training_scheme all

    Please see our GLUE entrypoint notebook for more information.

  4. πŸ€– TPU support (in beta)

    You can now use Composer to train your models on TPUs! Support is now available in Beta, and currently only supports single-core TPU training. Try it out, explore optimizations, and share your feedback and feature requests with us so we can make it better for you and for the community.

    To use TPUs with Composer, simply specify a tpu device:

    # Set device to `tpu`
trainer = composer.trainer.Trainer(
    model=model,
    train_dataloader=train_dataloader,
    max_duration=train_epochs,
    device='tpu')

# Run fit
trainer.fit()

    Please see our Training with TPUs notebook for more information.

  5. 🍎 Apple Silicon support (beta)

    Leverage Apple Silicon chips to train your models with Composer by providing the device='mps' argument:

    trainer = Trainer(
    ...,
    device='mps'
)

    We use the latest PyTorch MPS backend to execute the training. This requires torch version β‰₯1.12, and Max OSX 12.3+.

    For more information on training with Apple M chips, see the PyTorch 1.12 blog and our API Reference for Composer specific details.

  6. 🚧 Contrib repository

    Got a new method idea, or published a paper and want those methods to be easily accessible? We’ve created the mcontrib repository, with a lightweight process to contribute new algorithms. We’re happy to work directly with you to benchmark these methods and eventually β€œpromote” them to Composer for use by end customers.

    Please checkout the README for details on how to contribute a new algorithm. For more details on how to write speed-up methods, see our notebook on custom speed-up methods.

Additional API Changes

  1. πŸ”’ Passes Module

    The order in which algorithms are run matters significantly during composition. With this release we refactored algorithm passes into their own passes module. Users can now register custom passes (for custom algorithms) with the Engine. Please see #1377 for more information.

  2. πŸ—„οΈ Default Checkpoint Extension

    The CheckpointSaver now defaults to using the *.pt extension for checkpoint fienames. Please see #1370 for more information.

  3. πŸ‘οΈ Models Refactor

    Most vision models (ResNet, MNIST, ViT, EfficientNet) have been refactored from classes to a factory function. For example ComposerResNet -> composer_resnet.

    # before
from composer.models import ComposerResNet
model = ComposerResNet(..)

from composer.models import composer_resnet  # after
model = composer_resnet(..)

    The same refactor has been done for NLP as well, e.g. BERTModel -> create_bert_mlm and create_bert_classification.

    See #1227 (vision) and #1130 (NLP) for more details.

  4. βž• Misc API Changes

    • BreakEpochException has been removed.
    • state.is_model_deepspeed has been moved to composer.utils.is_model_deepspeed.
    • Helper function monitored_barrier has been added to composer distributed.

Bug Fixes

  • Add informative error for infer batch size issues (#1401)
  • Fix ImagenetDatasetHparams bug (#1392), resolves #1111
  • Fix hparams error condition checking (#1394)
  • Fix AMP resumption with grad scaler (#1376)
  • Auto Grad Accum Cache Clearing (#1380), fixes issue reported in #1331
  • Fix default precision (#1369)
  • Fix the profiler on multi-node training (#1358), resolves #1270
  • Retry SFTP on Size Mismatch (#1300)
  • Fix scheduler edge cases (#1350), resolves #1077
  • Fix a race condition in the object store logger (#1328)
  • Fix WandB load from checkpoint (#1326)
  • Fix Notebook Progress Bars (#1313)

Commits

What's Changed

Read more

Contributors

xloem, moinnadeem, and 27 other contributors
Assets 2
Loading

v0.8.2

27 Jul 23:36
@bandish-shah bandish-shah
v0.8.2
9e14a47
Compare
Choose a tag to compare
v0.8.2

πŸš€ Composer v0.8.2

Composer v0.8.2 is released! Install via pip:

pip install --upgrade mosaicml==0.8.2

Alternatively, install Composer with Conda:

conda install -c mosaicml mosaicml=0.8.2

πŸ› Bug Fixes

  1. Fixed Notebook Progress Bars in Colab

    Fixes a bug introduced by #1264 which causes Composer running in Colab notebooks to error out with:
    UnsupportedOperation: fileno.

    Closes #1312. Fixed in PR #1314.

Changelog

v0.8.1...v0.8.2

Assets 2
Loading

v0.8.1

22 Jul 23:23
@bandish-shah bandish-shah
v0.8.1
8418a67
This commit was created on GitHub.com and signed with GitHub’s verified signature. The key has expired.
GPG key ID: 4AEE18F83AFDEB23
Expired
Learn about vigilant mode.
Compare
Choose a tag to compare
v0.8.1

πŸš€ Composer v0.8.1

Composer v0.8.1 is released! Install via pip:

pip install --upgrade mosaicml==0.8.1

Alternatively, install Composer with Conda:

conda install -c mosaicml mosaicml=0.8.1

🎁 New Features

  1. πŸ–ΌοΈ Image Visualizer

    The ImageVisualizer callback periodically logs the training and validation images when using the WandB logger. This is great for validating your dataloader pipeline, especially if extensive data augmentations are used. Also, when training on a semantic segmentation task, the callback can log the target segmentation mask and the predicted segmentation mask by setting the argument mode='segmentation'. See PR #1266 for more details. Here is an example of using the ImageVisualizer callback:

    from composer import Trainer
from composer.callbacks import ImageVisualizer

# Callback to log 8 training images after every 100 batches
image_visualizer = ImageVisualizer()

# Construct trainer
trainer = Trainer(
    ...,
    callbacks=image_visualizer
)

# Train!
trainer.fit()

    Here is an example visualization from the training set of ADE20k:

  2. πŸ“Ά TensorBoard Logging

    You can now log metrics and losses from your Composer training runs with Tensorboard! See #1250 and #1283 for more details. All you have to do is create a TensorboardLogger object and add it
    to the list of loggers in your Trainer object like so:

    from composer import Trainer
from composer.loggers import TensorboardLogger

tb_logger = TensorboardLogger(log_dir="./my_tensorboard_logs")

trainer = Trainer(
    ...
    # Add your Tensorboard Logger to the trainer here.
    loggers=[tb_logger],
)

trainer.fit()

    For more information, see this tutorial.

  3. πŸ”™ Multiple Losses

    Adds support for multiple losses. If a model returns a tuple of losses, they are summed before the loss.backward() call. See #1240 for more details.

  4. 🌎️ Stream Datasets from HTTP URIs

    You can now specify a HTTP URI for a Streaming Dataset remote. See #1258 for more detials. For example:

    from composer.datasets.streaming import StreamingDataset
from torch.utils.data import DataLoader

# Construct the Dataset
dataset = StreamingDataset(
    ...,
    remote="https://example.com/dataset/",
)

# Construct the DataLoader
train_dl = DataLoader(dataset)

# Construct the Trainer
trainer = Trainer(
    ...,
    train_dataloader=train_dl,
)

# Train!
trainer.fit()

    For more information on streaming datasets, see this tutorial.

  5. πŸ„οΈ GPU Devices default to TF32 Matmuls

    Beginning with PyTorch 1.12, the default behavior for computing FP32 matrix multiplies on NVIDIA Ampere devices was switched from TF32 to FP32. See PyTorch documentation here.

    Since Composer is designed specifically for ML training with a focus on efficiency, we choose to preserve the old default of using TF32 on Ampere devices. This leads to significantly higher throughput when training in single precision, without impact training convergence. See PR #1275 for implementation details.

  6. πŸ‘‹ Set the Device ID for GPU Devices

    Specify the device ID within a DeviceGPU to train on when instantiating a Trainer object instead of using the local ID! For example,

    from composer.trainer.devices.device_gpu import DeviceGPU

# Specify to use GPU 3 to train 
device = DeviceGPU(device_id=3)

# Construct the Trainer
trainer = Trainer(
    ...,
    device = device
)

# Train!
trainer.fit()

  7. BERT and C4 Updates

    We make some minor adjustments to our bert-base-uncased.yaml training config. In particular, we make the global train and eval batch sizes a power of 2. This maintains divisibility when using many GPUs in multi-node training. We also adjust the max_duration so that it converts cleanly to 70,000 batches.

    We also upgrade our StreamingDataset C4 conversion script (scripts/mds/c4.py) to use a multi-threaded reader. On a 64-core machine we are able to convert the 770GB train split to .mds format in ~1.5hr.

  8. πŸ“‚ Set a prefix when using a S3ObjectStore

    When using S3ObjectStore for applications like checkpointing, it can be useful to provide path prefixes, mimicking folder/subfolder directories like on a local filesystem. When prefix is provided, any objects uploaded with S3ObjectStore will be stored at f's3://{self.bucket}/{self.prefix}{object_name}'.

  9. βš–οΈ Scale the Warmup Period of Composer Schedulers

    Added a new flag scale_warmup to schedulers that will scale the warmup period when a scale schedule ratio is applied. Default is False to mirror default behavior. See #1268 for more detials.

  10. 🧊 Stochastic Depth on Residual Blocks

    Residual blocks are detected automatically and replaced with stochastic versions. See #1253 for more details.

πŸ› Bug Fixes

  1. Fixed Progress Bars

    Fixed a bug where the the Progress Bars jumped around and did not stream properly when tailing the terminal over the network. Fixed in #1264, #1287, and #1289.

  2. Fixed S3ObjectStore in Multithreaded Environments

    Fixed a bug where the boto3 crashed when creating the default session in multiple threads simultaniously (see boto/boto3#1592). Fixed in #1260.

  3. Retry on ChannelException errors in the SFTPObjectStore

    Catch ChannelException SFTP transient error and retry. Fixed in #1245.

  4. Treating S3 Permission Denied Errors as Not Found Errors

    We update our handling of botocore 403 ClientErrors to interpret them as FileNotFoundErrors. We do this because of a situation that occurs when a user has no S3 credentials configured, and tries to read from a bucket with public files. For privacy, Amazon S3 raises 403 (Permission Denied) instead of 404 (Not Found) errors. As such, PR #1249 treats 403 ClientErrors as FileNotFoundErrors.

  5. Fixed Parsing of grad_accum in the TrainerHparams

    Fixes an error where the command line override --grad_accum lead to incorrect parsing. Fixed in #1256.

  6. Fixed Example YAML Files

    Our recipe configurations (YAML) are updated to the latest version, and a test was added to enforce correctness moving forward. Fixed in #1235 and #1257.

Changelog

v0.8.0...v0.8.1

Assets 2
Loading

v0.8.0

01 Jul 04:15
@ravi-mosaicml ravi-mosaicml
v0.8.0
80d3293
Compare
Choose a tag to compare
v0.8.0

πŸš€ Composer v0.8.0

Composer v0.8.0 is released! Install via pip:

pip install --upgrade mosaicml==0.8.0

Alternatively, install Composer with Conda:

conda install -c mosaicml mosaicml=0.8.0

New Features

  1. πŸ€— HuggingFace ComposerModel

    Train your HuggingFace models with Composer! We introduced a HuggingFaceModel that converts your existing πŸ€— Transformers models into a ComposerModel.

    For example:

    import transformers
from composer.models import HuggingFaceModel

# Define the model
hf_model = transformers.AutoModelForSequenceClassification.from_pretrained('bert-base-uncased', num_labels=2)

# Convert it into a ComposerModel
model = HuggingFaceModel(hf_model)

# Construct the trainer
trainer = Trainer(
    ...,
    model,
)

# Train!
trainer.fit()

    For more information, see the example on fine-tuning a pretrained BERT with Composer.

  2. πŸ«• Fused Layer Norm

    Fused LayerNorm replaces implementations of torch.nn.LayerNorm with a apex.normalization.fused_layer_norm. The fused kernel provides increased GPU utilization.

    For example:

    from composer.trainer import Trainer
from composer.algorithms import FusedLayerNorm

# Initialize the algorithm
alg = FusedLayerNorm()

# Construct the trainer
trainer = Trainer(
    algorithms=alg,
)

# Train!
trainer.fit()

    See the method card for more information.

  3. πŸ’Ύ Ignore Checkpoint Parameters

    If you have a checkpoint and don't want to restore some elements of the chceckpoint to the state, we added a load_ignore_keys parameter. Any specified (nested) keys will be ignored. Glob syntax is supported!

    For example, to restore a checkpoint without the seed:

    from composer import Trainer

trainer = Trainer(
    ...,
    load_path="path/to/my/checkpoint.pt",
    load_ignore_keys=["state/rank_zero_seed", "rng"],
)

    See the Trainer API Reference for more information.

  4. πŸͺ£ Object Stores

    Composer v0.8.0 introduces an abstract Object Store API to support multiple object store drivers, such as boto3 (for Amazon S3) and Paramiko (for SFTP), in addition to the existing libcloud implementation.

    For example, if you are training on AWS where credentials are available in the environment, here's how to to save checkpoints to a S3 object store via Boto3.

    from composer import Trainer
from composer.loggers import ObjectStoreLogger
from composer.utils.object_store import S3ObjectStore

logger = ObjectStoreLogger(
    object_store_cls=S3ObjectStore,
    object_store_kwargs={
        # These arguments will be passed into the S3ObjectStore -- e.g.:
        # object_store = S3ObjectStore(**object_store_kwargs)
        # Refer to the S3ObjectStore class for documentation
        'bucket': 'my-bucket',
    },
)

trainer = Trainer(
    ...,
    loggers=logger,
)

# Train!
trainer.fit()

    See the Object Store API Reference for more information.

  5. πŸͺ¨ Artifact Metadata

    Composer automatically logs the epoch, batch, sample, and token counts as metadata when storing artifacts in Weights & Biases. See the API Reference for more information.

API Changes

  1. βœ‚οΈ Gradient Clipping is now an Algorithm

    To clean up the Trainer, we moved gradient clipping into an Algorithm. The grad_clip_norm argument in the Trainer is deprecated and will be removed in a future version of Composer. Instead, use the Gradient Clipping algorithm:

    For example:

    from composer.algorithms import GradientClipping
from composer.trainer import Trainer

# Configure gradient clipping
gradient_clipping = GradientClipping()

# Configure the trainer
trainer = Trainer(
    ...,
    algorithms=gradient_clipping,
)

# Train!
trainer.fit()

    See the method card for more information.

  2. πŸ•’οΈ Removed batch_num_samples and batch_num_tokens from the state.

    State properties batch_num_samples and batch_num_tokens have been removed.
    Instead, use State.timestamp for token and sample tracking.

  3. πŸ§‘β€πŸ€β€πŸ§‘ DDP Sync Strategy

    We changed the default DDP Sync Strategy to MULTI_AUTO_SYNC, as FORCED_SYNC doesn't work with all algorithms.

  4. πŸƒ Moved the run_name into the State

    The run_name has been added to the State object, so it is persisted with checkpoints. It has been removed from the Logger.

Bug Fixes

  • In the Object Store Logger, added in retries for credential validation, and validating credentials only on global rank zero. (#1144)
  • Fixed a bug in the speed monitor where it returned negative wall clock times. (#1123)
  • Fixed how block-wise Stochastic Depth could freeze the trainer. (#1087)
  • Fixed a bug in the [MLPerfCallback] where sample counts were incorrect on per-sharded datasets. (#1156)

Changelog

v0.7.1...v0.8.0

Assets 2
Loading

v0.7.1

07 Jun 00:21
@ravi-mosaicml ravi-mosaicml
v0.7.1
20aad7e
Compare
Choose a tag to compare
v0.7.1

πŸš€ Composer v0.7.1

Composer v0.7.1 is released! Install via pip:

pip install --upgrade mosaicml==0.7.1

Alternatively, install Composer with Conda:

conda install -c mosaicml mosaicml=0.7.1

Bug Fixes

  • Upgraded wandb>=0.12.17, to fix incompatibility with protobuf >= 4 (wandb/wandb#3709)

Changelog

v0.7.0...v0.7.1

Assets 2
Loading