Easily train or fine-tune SOTA computer vision models with one open source training library. The home of Yolo-NAS.
APACHE-2.0 License
Published by deci-services 7 months ago
This GitHub Release was done automatically by CircleCI
Published by deci-services 7 months ago
This GitHub Release was done automatically by CircleCI
Published by deci-services 8 months ago
pycocotools
dependency removed by @BloodAxe in https://github.com/Deci-AI/super-gradients/pull/1791
Full Changelog: https://github.com/Deci-AI/super-gradients/compare/3.6.0...3.6.1
Published by shaydeci 8 months ago
Full Changelog: https://github.com/Deci-AI/super-gradients/compare/3.6.0...3.6.1
Published by deci-services 9 months ago
Hey @channel
We're pleased to announce the release of Super-Gradients 3.6.0 🙂 .
This update includes several important changes and improvements:
Changes and Enhancements
DetectionMetrics
as an enhancement by @DimaBir.ImagePermute
processing inclusion, by @BloodAxe.predict()
support for segmentation models, by @BloodAxe.PPYoloELoss
, removing the requirement for a reg_max
parameter, by @BloodAxe.onnxsim
instead of onnx-simplifier
for consistency in naming, thanks to @BloodAxe.OhemLoss
thanks to @danielafrimi.[context.sg](http://context.sg/)_logger
is available, by @shaydeci.set_device
value to prevent unintentional launch of DDP, updated bymultigpu=None
with device=cpu
wasn't functioning as expected, thanks toDetectionMixup
that affected YoloXTrainingStageSwitchCallback
, by @BloodAxe.DetectionDataset
are contained within a trivial container, by @BloodAxe.ExtremeBatchDetectionVisualizationCallback
for multiscale collate function, by @BloodAxe.DistanceBasedDetectionMetrics
and DetectionMetrics
, by @BloodAxe.DetectionMetrics
as an enhancement by @DimaBir.ImagePermute
processing inclusion, by @BloodAxe.predict()
support for segmentation models, by @BloodAxe.PPYoloELoss
, removing the requirement for a reg_max
parameter, by @BloodAxe.onnxsim
instead of onnx-simplifier
for consistency in naming, thanks to @BloodAxe.OhemLoss
thanks to @danielafrimi.[context.sg](http://context.sg/)_logger
is available, by @shaydeci.set_device
value to prevent unintentional launch of DDP, updated bymultigpu=None
with device=cpu
wasn't functioning as expected, thanks toDetectionMixup
that affected YoloXTrainingStageSwitchCallback
, by @BloodAxe.DetectionDataset
are contained within a trivial container, by @BloodAxe.ExtremeBatchDetectionVisualizationCallback
for multiscale collate function, by @BloodAxe.DistanceBasedDetectionMetrics
and DetectionMetrics
, by @BloodAxe.Published by deci-services 11 months ago
This GitHub Release was done automatically by CircleCI
model.predict(video)
does not cause OOM anymore by @hakuryuu96 in https://github.com/Deci-AI/super-gradients/pull/1621
model.predict
can now take skip_resize
argument to run forward in the original image resolution (Good for large images & small objects) by @Louis-Dupont in https://github.com/Deci-AI/super-gradients/pull/1605
_pad_image
by @BloodAxe in https://github.com/Deci-AI/super-gradients/pull/1632
model.export()
that led to crash when FP16 export was requested and model was on CPU device by @BloodAxe in https://github.com/Deci-AI/super-gradients/pull/1643
Full Changelog: https://github.com/Deci-AI/super-gradients/compare/3.4.1...3.5.0
checkpoint_num_classes
is propagated from YAML to model by @BloodAxe in https://github.com/Deci-AI/super-gradients/pull/1533
np.bool
which is not supported in latest np versions by @Louis-Dupont in https://github.com/Deci-AI/super-gradients/pull/1558
CSPDarknet53.foward
by @Louis-Dupont in https://github.com/Deci-AI/super-gradients/pull/1564
class_names
to model.predict
for detection by @Louis-Dupont in https://github.com/Deci-AI/super-gradients/pull/1529
Published by deci-services 12 months ago
convert_recipe_to_code
script to convert YAML recipe to self-contained train script by @BloodAxe in https://github.com/Deci-AI/super-gradients/pull/1568
export_recipe
script to convert a YAML recipe to a single big YAML file by @shaydeci in https://github.com/Deci-AI/super-gradients/pull/1560
class_names
to model.predict
for detection by @Louis-Dupont in https://github.com/Deci-AI/super-gradients/pull/1529
checkpoint_num_classes
is propagated from YAML to model by @BloodAxe in https://github.com/Deci-AI/super-gradients/pull/1533
_scale_fn_ref
is missing in CyclicLR by @BloodAxe in https://github.com/Deci-AI/super-gradients/pull/1575
arange
not supporting fp16 for CPU device. by @BloodAxe in https://github.com/Deci-AI/super-gradients/pull/1574
CSPDarknet53.foward
by @Louis-Dupont in https://github.com/Deci-AI/super-gradients/pull/1564
data-gradients
version bumped up to 0.2.2np.bool
which is not supported in latest np versions by @Louis-Dupont in https://github.com/Deci-AI/super-gradients/pull/1558
Full Changelog: https://github.com/Deci-AI/super-gradients/compare/3.3.0...3.4.0
Published by deci-services almost 1 year ago
Super Gradients 3.3.1
class_names
to model.predict
for detection by @Louis-Dupont in https://github.com/Deci-AI/super-gradients/pull/1529
max_train_batches
/max_valid_batches
by @hakuryuu96 https://github.com/Deci-AI/super-gradients/pull/1554
np.bool
-> bool
which is not supported in latest np versions by @Louis-Dupont in https://github.com/Deci-AI/super-gradients/pull/1558
checkpoint_num_classes
is propagated from YAML files to models.get
by @BloodAxe in https://github.com/Deci-AI/super-gradients/pull/1533
CSPDarknet53.foward
by @Louis-Dupont in https://github.com/Deci-AI/super-gradients/pull/1564
Full Changelog: https://github.com/Deci-AI/super-gradients/compare/3.3.0...3.3.1
Published by deci-services about 1 year ago
This GitHub Release was done automatically by CircleCI
metric_to_watch
is wrong by @Louis-Dupont in https://github.com/Deci-AI/super-gradients/pull/1437
master_320
and other release branches (#1436) by @BloodAxe in https://github.com/Deci-AI/super-gradients/pull/1438
_get_installed_distributions
by @Louis-Dupont in https://github.com/Deci-AI/super-gradients/pull/1494
Full Changelog: https://github.com/Deci-AI/super-gradients/compare/3.1.3...3.3.0
Published by deci-services about 1 year ago
TLDR:
master_320
and other release branches by @BloodAxe in https://github.com/Deci-AI/super-gradients/pull/1436
Full Changelog: https://github.com/Deci-AI/super-gradients/compare/3.1.3...3.2.1
Published by deci-services about 1 year ago
This GitHub Release was done automatically by CircleCI
Full Changelog: https://github.com/Deci-AI/super-gradients/compare/3.1.3...3.2.0
Published by shaydeci over 1 year ago
New features
predict() support for pose estimation (PR: #1142) and classification (PR: #1220)
The possibility to log scalar values with explicit time units associated with it: log losses per step, use TimerCallback that measures and log batch time (PR: #1181)
Support torch compile (PR: #1172)
Class agnostic NMS (PR: #1232)
Allow to compute & log per-class mAP scores (PR: #1238)
Small features
Add batch_size option to predict (PR: #1273)
Allows to pass user-defined target of IoU thresholds without asking user to edit IoUThresholds enum in SG codebase (PR: #1223)
Add DetectionVerticalFlip for bounding box detection augmentation (PR: #1237)
Support the option for multiple ignored indices for segmentation metrics (PR: #1177)
Simpify torch schedulers integration (PR: #1230)
support single output in YoloX NMS (PR: #813)
Compute the best threshold for each class in an efficient manner without added loops (PR: #973)
Bug fix
Hotfix/alg 1470 drop boxes padding by (PR: #1107)
Fix evaluate from recipe by (PR: #1170)
fixed bug of whitespace by (PR: #1173)
Fix collections.Iterable -> typing.Iterable to fix crash in python 3.10 by @BloodAxe in #1178
Fix missing encoding (PR: #1185)
default quantization params set in qat from config (PR: #1192)
Fix mkdirs in checkpoint (PR: #1198)
Fix DEKR’s replace_head & improve repr for keypoints transforms (PR: #1227)
Getting rid of “module.” heritage (PR: #1184)
Fix LayerNorm have a bias parameter attribute but is not instance of torch primitive modules (PR: #1229)
moved exceptions from training (PR: #1260)
Added explicit casting of input in predict pipeline. (PR: #1281)
kd ema model unrwap model fix (PR: #1283)
New Contributors
@allankouidri made their first contribution in #1185
@djm93dev made their first contribution in #1199
@danielafrimi made their first contribution in #1220
@RanZilberstein made their first contribution in #1228
@itaylevy-deci made their first contribution in #973
@LukeAI made their first contribution in #1237
@jorgectf made their first contribution in #1240
@jacobmarks made their first contribution in #1278
Published by deci-services over 1 year ago
Full Changelog: https://github.com/Deci-AI/super-gradients/compare/3.1.1...3.1.2
Published by deci-services over 1 year ago
This GitHub Release was done automatically by CircleCI
Published by deci-services over 1 year ago
This GitHub Release was done automatically by CircleCI
Published by deci-services over 1 year ago
Release notes 3.0.9
Features
Object Detection (Yolox, PPyoloE) predict function on images, videos, GIF, and folders paths. The predict function works out of the box, no need to define pre and post-processing. Both pre and post-processing are taken from the training. Examples here. PR: #815, #804, #807, #829, #845
Support yolov5 format detection dataset by for YoloX, PPyoloE https://github.com/Deci-AI/super-gradients/pull/847
DetectionOutputAdapter tutorial. The DetectionOutputAdapter is a class that converts the output of a detection model into a user-appropriate format.
Predict function on images and videos - taking the pre+post processing from the training recipe.
Added student_adapter feature to KDModule, which lets the user pass a manipulated version of the input to the student model. PR: #820
Introduce min_samples param to dataloader_params, which repeats the dataset in case its size is smaller than the value of this parameter. PR: #838
Bug fixes:
Published by deci-services over 1 year ago
Bug fixes:
Fix the compatibility of ONNX conversion with torch 1.11+
Fix the docker publication
Fix Pre-Prediciton callback override bug
Fix dataloader multiprocessing (for Mac OS and Windows)
Fix strict_load not being passed in evaluate_from_recipe
Resolve many potential circular imports
Correct albumentations import log
Fix incorrect tensor type & device when entire batch has no targets
Fix documentation format in many files
Features:
New documentation for many features (EMA, AMP, QAT, Knowledge-Distillation, Dataloaders, Optimizer, Schedulers…)
Added Lion optimizer
Added registration for SG Logger (to allow custom-defined loggers to be used from YAML)
DDRNet 39 pre-trained segmentation model
Mapillary dataset
RoboFlow100 benchmark and all 100 datasets
QATTrainer for training with QAT
QAT from recipe script
PP-YoloE pre-trained models
Pose Estimation models loss functions
Added Auxiliary heads to Unet
Channel wise distillation
ONNX Simplifier on ONNX conversion
Unet and Loss function for binary segmentation
New Transforms: DetectionPadToSize, DetectionImagePermute
Allow activation factory type to take resolved type as inp
New strict_load mode: key_matching.
Published by deci-services over 1 year ago
Full Changelog: https://github.com/Deci-AI/super-gradients/compare/3.0.6...3.0.7
Published by deci-services almost 2 years ago
Full Changelog: https://github.com/Deci-AI/super-gradients/compare/3.0.5...3.0.6