Brevitas

Breaking Changes
* Remove ONNX QOp export (https://github.com/Xilinx/brevitas/pull/917)
* QuantTensor cannot have empty metadata fields (e.g., scale, bitwidth, etc.) (https://github.com/Xilinx/brevitas/pull/819)
* Bias quantization now requires the specification of bit-width (https://github.com/Xilinx/brevitas/pull/839)
* QuantLayers do not expose quant_metadata directly. This is delegated to the proxies (https://github.com/Xilinx/brevitas/pull/883)
* QuantDropout has been removed (https://github.com/Xilinx/brevitas/pull/861)
* QuantMaxPool has been removed (https://github.com/Xilinx/brevitas/pull/858)



Highlights
* Support for OCP/FNUZ FP8 quantization
* Compatibility with QAT/PTQ, including all current PTQ algorithms implemented (GPTQ, LearnedRound, GPFQ, etc.)
* Possibility to fully customize the minifloat configuration (i.e., select mantissa/exponent bit-width, exponent bias, etc.)
* Support for ONNX QDQ export

* Support for OCP MX Quantization
* Compatibility with QAT/PTQ, including all current PTQ algorithms implemented (GPTQ, LearnedRound, GPFQ, etc.)
* Possibility to fully customize the minifloat configuration (i.e., select mantissa/exponent bit-width, exponent bias, group size, etc.)

* New QuantTensor supports:
* FloatQuantTensor: supports OCP FP formats and general minifloat quantization
* GroupwiseQuantTensor: supports for OCP MX formats and general groupwise int/minifloat quantization

* Support for Channel splitting

* Support for HQO optimization for zero point

* Support for HQO optimization for scale (prototype)

* Improved SDXL entrypoint under brevitas_examples

* Improved LLM entrypoint under brevitas_examples
* Compatibility with accelerate

* Prototype support for `torch.compile`:
* Check PR https://github.com/Xilinx/brevitas/pull/1006 for an example on how to use it

What's Changed
For a more comprehensive list of changes and fix, check the list below:

* Enhance: Importing quantized models after bias correction by costigt-dev in https://github.com/Xilinx/brevitas/pull/868
* Fix QCDQDecoupledWeightQuantProxyHandlerMixin return args by costigt-dev in https://github.com/Xilinx/brevitas/pull/870
* Fix - Speech to text: Create an empty json file by costigt-dev in https://github.com/Xilinx/brevitas/pull/871
* Feat (scaling/standalone): flag to retrieve full state dict by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/874
* Notebooks: makes notebooks deterministic and prints output of asserts by fabianandresgrob in https://github.com/Xilinx/brevitas/pull/847
* Fix (proxy): revert value tracer change by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/888
* Fix (proxy): fix for attributes retrieval by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/880
* Feat (notebook): add example for dynamic quantization to ONNX export by fabianandresgrob in https://github.com/Xilinx/brevitas/pull/877
* Fix (gpxq): handling empty tensors with GPxQ and adding unit tests by i-colbert in https://github.com/Xilinx/brevitas/pull/892
* Fix (ptq): expose uint_sym_act flag and fix issue with minifloat sign by fabianandresgrob in https://github.com/Xilinx/brevitas/pull/898
* Feat (minifloat): add support for user specified minifloat format by fabianandresgrob in https://github.com/Xilinx/brevitas/pull/821
* Feat: Add QuantConv3d and QuantConv3dTranspose by costigt-dev in https://github.com/Xilinx/brevitas/pull/805
* Add tutorial examples of per-channel quantization by OscarSavolainenDR in https://github.com/Xilinx/brevitas/pull/867
* Fix (tests): revert pytest pin by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/903
* Remove: Remove original_cat workaround by costigt-dev in https://github.com/Xilinx/brevitas/pull/902
* Infra: Update issue template by nickfraser in https://github.com/Xilinx/brevitas/pull/893
* Pull Request Template by capnramses in https://github.com/Xilinx/brevitas/pull/885
* Fix (core): add return in state_dict by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/910
* Fix (quant_tensor): fix typing and remove unused checks by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/913
* Fix (nn): removed unused caching in adaptive avgpool2d by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/911
* Fix (quant_tensor): remove unused checks by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/918
* Setup: pin ONNX to 1.15 due to ORT incompatibility by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/924
* Feat (examples): add support for Stable Diffusion XL by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/909
* Assert all ptq-common bit widths are positive integers by OscarSavolainenDR in https://github.com/Xilinx/brevitas/pull/931
* Enhance: Quant Tensor Test by costigt-dev in https://github.com/Xilinx/brevitas/pull/894
* Fix (examples/stable_diffusion): README formatting and clarification by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/932
* Fix (examples/ptq): fix for bitwidth check by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/934
* Feat: functionalize QuantTensor by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/878
* Feat (minifloat): cleanup minifloat impl by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/922
* Fix tests in dev by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/939
* Feat (proxy): scale computation delegated to bias proxy by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/938
* Fix (gpxq): adding input quant to process input by i-colbert in https://github.com/Xilinx/brevitas/pull/943
* Fix (quant): propagate device and dtype in subinjector by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/942
* Fix (gpxq): correct variable name by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/944
* Fix (quant_tensor): fix AvgPool functional implementation by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/945
* Feat (quant_tensor): support for dim() and ndim by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/947
* Fix (graph/standardize): correct check for Mean to AvgPool by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/948
* Feat (graph/standardize): default keepdim value by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/950
* Fix bullet formatting in getting started guide by timkpaine in https://github.com/Xilinx/brevitas/pull/952
* Fix (quant/float): correct scaling_impl and float_scaling_impl by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/953
* Fix/remove-numel - Remove numel is zero check from context manager exit method by costigt-dev in https://github.com/Xilinx/brevitas/pull/920
* Feat (examples/ptq): support for dynamic act quant by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/935
* Feat (quant_tensor): support for FloatQuantTensor by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/919
* Fix (examples/llm): Add all rewriters to the list by nickfraser in https://github.com/Xilinx/brevitas/pull/956
* Fix (core/quant/float): use eps to avoid log(0) by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/957
* Fix (test/actions): Excluded `torch==1.9.1`, `platform=macos-latest` tests by nickfraser in https://github.com/Xilinx/brevitas/pull/960
* Adding FP8 weight export by costigt-dev in https://github.com/Xilinx/brevitas/pull/907
* Fix (llm): fix device issue for eval when not using default device by fabianandresgrob in https://github.com/Xilinx/brevitas/pull/949
* Fix (GPFQ): using random projection for speed up/less memory usage by fabianandresgrob in https://github.com/Xilinx/brevitas/pull/964
* Fix (calibrate/minifloat): fix for act calibration by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/966
* Fix (quant/float): restore fix for log(0) by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/968
* Setup: pin numpy version by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/974
* Feat (minifloat): support for FNUZ variants by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/973
* Fix (core/float): add default for float_scaling_impl by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/972
* Feat (graph/equalize): upcast during equalization computation by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/970
* Generative improv by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/965
* Fix (requirements/setuptools): Set maximum requirement for `setuptools` by nickfraser in https://github.com/Xilinx/brevitas/pull/963
* Fix: Typo fix on SDXL command line args by nickfraser in https://github.com/Xilinx/brevitas/pull/976
* Fix (graph/bias_correction): Fix when layer parameters are offloaded to `accelerate` by nickfraser in https://github.com/Xilinx/brevitas/pull/962
* Fix (ptq/bias_correction): remove unnecessary forward pass by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/980
* Fix (export/qonnx): Fixed symbolic kwargs order. by nickfraser in https://github.com/Xilinx/brevitas/pull/988
* Various SDXL quantization fixes by nickfraser in https://github.com/Xilinx/brevitas/pull/977
* Fix (brevitas_examples/sdxl): Various fixes by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/991
* Feat (proxy/parameter_quant): cache quant weights by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/990
* Docs: Added 0.10.3 release note to README. by nickfraser in https://github.com/Xilinx/brevitas/pull/993
* Added some preliminary unit tests to the CNNs 'quantize_model' by OscarSavolainenDR in https://github.com/Xilinx/brevitas/pull/927
* Feat (tests): extended minifloat unit tests by alexredd99 in https://github.com/Xilinx/brevitas/pull/979
* Fix (proxy/runtime_quant): correct handling of mixed type quantization by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/985
* docs (readme): Fixed GH actions badges by nickfraser in https://github.com/Xilinx/brevitas/pull/996
* Feat: Update LLM entry-point by nickfraser in https://github.com/Xilinx/brevitas/pull/987
* Feat: Support for Groupwise (MX) quantization by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/971
* Feat(graph): better exclusion mechanism by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/1003
* Fix (mx): input view during quantization by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/1005
* Feat (mx): adding padding and transposed support by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/1007
* fix (nn/avg_pool): Fix for trunc quant not being applied by nickfraser in https://github.com/Xilinx/brevitas/pull/1014
* Fix (nn/conv): Fixed conversion of convolutions when `padding_mode='same'` by nickfraser in https://github.com/Xilinx/brevitas/pull/1017
* Fix (proxy): clean-up by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/1011
* Feat (mx): automatic group_dim in layerwise quant by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/1012
* fix (nn/conv): Fix regression introduced in 1017 by nickfraser in https://github.com/Xilinx/brevitas/pull/1019
* Feat (mx): gptq compatibility and quant tests by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/1013
* Feat (mx): PTQ MX + Float support by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/1010
* Fix (graph/quant): Bugfix in blacklist matching in `find_module` by nickfraser in https://github.com/Xilinx/brevitas/pull/1021
* Test (graph/quantize) Added extra prefix test to `layerwise_quantize` by nickfraser in https://github.com/Xilinx/brevitas/pull/1022
* Test (example/llm): Refactor and add basic tests for the LLM entry-point by nickfraser in https://github.com/Xilinx/brevitas/pull/1002
* Feat (examples/sdxl): Updates to SDXL entry-point by nickfraser in https://github.com/Xilinx/brevitas/pull/1020
* Feat (gptq): optimizing CPU to GPU memory transfer by i-colbert in https://github.com/Xilinx/brevitas/pull/1009
* notebooks: rerun notebooks. by nickfraser in https://github.com/Xilinx/brevitas/pull/1026
* HQO for scale/zero point by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/937
* Test calibration reference by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/1031
* Fix (sdxl): avoid suppressing checkpoint errors by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/1034
* Setup: update checkout version to v3 by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/1039
* Fix (gpxq): correct index for groupwise GPxQ by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/1040
* Fix (llm): small fixes to LLM by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/1035
* Feat (activation_calibration): speed-up by skipping quantization by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/1029
* Fix (proxy): fix for float quant properties is_ocp and is_fnuz by alexredd99 in https://github.com/Xilinx/brevitas/pull/1028
* Decoupled PerChannel/PerTensor quantization by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/1025
* Fix (examples/llm): Fix infinite loop in LLM entrypoint with WikiText2 by pablomlago in https://github.com/Xilinx/brevitas/pull/1044
* Fix po2 for float quant by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/1033
* Feat (gpfq): adding memory-efficient formulation by i-colbert in https://github.com/Xilinx/brevitas/pull/1043
* Fix weights mse by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/1047

New Contributors
* OscarSavolainenDR made their first contribution in https://github.com/Xilinx/brevitas/pull/861
* costigt-dev made their first contribution in https://github.com/Xilinx/brevitas/pull/868
* timkpaine made their first contribution in https://github.com/Xilinx/brevitas/pull/952
* alexredd99 made their first contribution in https://github.com/Xilinx/brevitas/pull/979
* pablomlago made their first contribution in https://github.com/Xilinx/brevitas/pull/1044

**Full Changelog**: https://github.com/Xilinx/brevitas/compare/v0.10.2...v0.11.0

What's Changed
* Backport: Fix (export/qonnx): Fixed symbolic kwargs order. (988) by nickfraser in https://github.com/Xilinx/brevitas/pull/992
* `numpy` version, `onnx` version and maximum `setuptools` version set

**Full Changelog**: https://github.com/Xilinx/brevitas/compare/v0.10.2...v0.10.3

What's Changed
* Fix (QuantLayer): make bias for QuantLayer optional by fabianandresgrob in https://github.com/Xilinx/brevitas/pull/846
* Fix (examples/llm): set `group_size` only for groupwise quantization by nickfraser in https://github.com/Xilinx/brevitas/pull/853
* Fix (gpfq): updating input processing and L1-norm constraints for GPFA2Q by i-colbert in https://github.com/Xilinx/brevitas/pull/852
* ImageNet PTQ example fix by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/863
* feat (gen/quantize): Added device flag to `quantize_model` by nickfraser in https://github.com/Xilinx/brevitas/pull/860
* Docs: update README for 0.10.2 release by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/865


**Full Changelog**: https://github.com/Xilinx/brevitas/compare/v0.10.1...v0.10.2

Highlights
* A2Q+ support [paper](https://arxiv.org/abs/2401.10432)
* A2Q+ examples with CIFAR10 and Super Resolution
* Support for concatenation equalization for weights and activations
* Support for GPFQ + A2Q L1 Norm bound
* Possibility to explicitly export Q node for weights in QCDQ export
* Support for float16 and bfloat16 for QCDQ export
* Support for Dynamic Activation Quantization for ONNX QDQ export
* Support for channel-splitting [paper](https://arxiv.org/pdf/1901.09504.pdf)
* (Beta) Better compatibility with Huggingface accelerate and optimum
* (Beta) Improved support and testing for minifloat quantization

What's Changed
* Fix (examples/generative): set weight_bit_width in weight_quant by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/783
* Feat (graph/equalize): improvements for llm equalization by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/784
* [graph] Fix typo in class name by nickfraser in https://github.com/Xilinx/brevitas/pull/765
* Fix (graph/equalize): refactor for act equalization by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/787
* [quant_tensor] Updates `__truediv__` behaviour to match "standard fixed point rules" by nickfraser in https://github.com/Xilinx/brevitas/pull/769
* Feat (export): (b)float16 support for qcdq export by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/776
* Feat (ptq): Adding A2Q Upper Bound clipping to GPFQ by fabianandresgrob in https://github.com/Xilinx/brevitas/pull/734
* Extended equalization by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/778
* Better Bfloat16 support by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/777
* Fix (stats): add return statement in state_dict by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/792
* Fix (equalize): improved cat eq checks by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/793
* Fix (export): add CastMixin by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/794
* Dynamic Act Quant support by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/796
* Fix (examples/quantizers): correct dynamic zero point handling by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/806
* Feat (a2q+): improving accumulator-aware weight quantization by i-colbert in https://github.com/Xilinx/brevitas/pull/797
* Feat (a2q+): adding new super resolution models to brevitas_examples by i-colbert in https://github.com/Xilinx/brevitas/pull/811
* Feat (Channel-Splitting): sets up first skeleton for channel-splitting by fabianandresgrob in https://github.com/Xilinx/brevitas/pull/772
* Feat: support for optimum by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/826
* Fix (tests): adding tests for FloatQuant by fabianandresgrob in https://github.com/Xilinx/brevitas/pull/815
* Fix (export): correct q node export by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/829
* Fix (examples/llm): correct groupwise export by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/832
* Fix (examples/super_res): updating README by i-colbert in https://github.com/Xilinx/brevitas/pull/828
* Fix (examples/export): improved export by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/838
* Fix (graph/equalize): cleanup and device management by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/840
* Feat (examples/a2q): adding CIFAR10 example by i-colbert in https://github.com/Xilinx/brevitas/pull/813
* Fix (export): check for Per Group quantization by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/848

**Full Changelog**: https://github.com/Xilinx/brevitas/compare/v0.10.0...v0.10.1

a2q_cifar10_r1
This release contains training code and pre-trained weights to demonstrate accumulator-aware quantization (A2Q) on an image classification task. Code is also provided to demonstrate Euclidean projection-based weight initialization (EP-init) as proposed in our paper ["A2Q+: Improving Accumulator-Aware Weight Quantization"](https://arxiv.org/abs/2401.10432).

Find the associated docs at [https://github.com/Xilinx/brevitas/tree/a2q_cifar10_r1/src/brevitas_examples/imagenet_classification/a2q](https://github.com/Xilinx/brevitas/tree/a2q_cifar10_r1/src/brevitas_examples/imagenet_classification/a2q).

super_res_r2
A2Q+ Super Resolution Experiments with Brevitas

This release contains training code and pre-trained weights to demonstrate accumulator-aware quantization (A2Q+) as proposed in our paper "[A2Q+: Improving Accumulator-Aware Weight Quantization](https://arxiv.org/abs/2401.10432)" on a super resolution task.

Find the associated docs at [https://github.com/Xilinx/brevitas/tree/super_res_r2/src/brevitas_examples/super_resolution](https://github.com/Xilinx/brevitas/tree/super_res_r2/src/brevitas_examples/super_resolution).

* Fix (docs): README.md for pre-commit by volcacius in https://github.com/Xilinx/brevitas/pull/781

New Contributors
* fabianandresgrob made their first contribution in https://github.com/Xilinx/brevitas/pull/717
* saadulkh made their first contribution in https://github.com/Xilinx/brevitas/pull/741

**Full Changelog**: https://github.com/Xilinx/brevitas/compare/v0.9.1...v0.10.0

super_res_r1
Integer-Quantized Super Resolution Experiments with Brevitas

This release contains scripts demonstrating how to train integer-quantized super resolution models using Brevitas.
Code is also provided to demonstrate accumulator-aware quantization (A2Q) as proposed in our ICCV 2023 paper "[A2Q: Accumulator-Aware Quantization with Guaranteed Overflow Avoidance](https://arxiv.org/abs/2308.13504)".

Find the associated docs at https://github.com/Xilinx/brevitas/tree/super_res_r1/src/brevitas_examples/super_resolution .

What's Changed
* Setup: add requirements-dev with pre-commit by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/581
* CI update by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/570
* Fix (brevitas_examples/bnn_pynq): missing 4b resnet18 link and hash fn by volcacius in https://github.com/Xilinx/brevitas/pull/583
* Docs: update READMEs by Giuseppe5 in https://github.com/Xilinx/brevitas/pull/584


**Full Changelog**: https://github.com/Xilinx/brevitas/compare/v0.9.0...v0.9.1