As discussed with TVM PMC, we would like to give a summary of the project per month, so that people can get better sense of what is going on in the community.
Feedback and suggestion are welcomed so that we can further improve the report. Also correct me if I made any mistake or missed anything. Thanks!
Community
In Feb 2019, we successfully released TVM v0.5 (release notes) and made the roadmap for v0.6.
The community welcomes new Reviewer Zhao Wu (@FrozenGene), Committer Jared Roesch (@jroesch) and PMC member Lianmin Zheng (@merrymercy)
TVM community has voted through Apache Incubation proposal (#2543). Markus Weimer posted a proposal to general@incubator.apache.org seeking for ideas and suggestions. The official voting is ongoing on general@ right now.
Features and Improvements
The community keeps making progress on supporting a variety of deep learning frameworks, providing better programming interfaces by improving Hybrid Script and introducing Rust frontend. We (partially) delivered major features including Quantization, Automatic Differentiation and Algebraic Data Types support.
On the forum we also had inspiring discussion about the latest ML/DL compiler techniques & designing.
More improvements along with details are listed below.
Operator Support
- A special operator
annotation.stop_fusionto prevent it being fused with previous expressions (#2624). -
batch_matmulsupported (#2561). -
reverse_reshapesupported (#2503). - Faster-RCNN proposal operator for CUDA (#2420).
- Vision operator for YOLO
yolo_reorg(#1941). -
sliceoperator for MXNet (#2662). -
arangesupported (#2621). - Vision operator
roi_align(#2618). -
whereoperator for MXNet (#2647).
User Interface and Frontend
- Introduced
HybridModule(#2477) so that normal TVM schedule can be compiled to hybrid target, run and dumped to Hybrid Script. - Most frameworks have been supported in Relay, including ONNX, Keras, Tensorflow, Caffe2, CoreML, NNVMv1, MXNet (#2246). Siju is working on DarkNet.
- Relay now supports saving and loading parameter dictionaries. (#2620)
- Rust frontend (#2292).
- We are now supporting Tensorflow saved model for NNVM (#2493). Relay support is ongoing (#2586).
- Add
max_num_threadsto Hybrid Script, which allows users to get max number of threads for GPU targets (#2672).
Runtime and Hardware Support
- RFC for bringing TVM to Bare-Metal devices (#2563)
- Make external library extend TVM’s NDArray more easily (#2613).
Performance Improvement
- AlterOpLayout pass is now enabled for x86 on Relay (#2585). It is essential to get decent performance for CNN-based model on Intel CPUs.
Documents and Tutorials
- Tutorials for deep learning frameworks support in Relay.
- Tutorial for running AutoTVM with Relay (#2594).
- Document for Algebraic Data Types (#2575).
High-level Optimizations
- An optimization pass to eliminate expressions which have the same functionality and same inputs (#2639).
- Provided a pass to calculate the number of multiply-accumulate operations in a network (#2609).
- Algebraic Data Types (ADT) support (#2442, #2575). ADT provides an expressive, efficient, and safe way to realize recursive computation (e.g., RNN). Refer to https://docs.tvm.ai/langref/relay_adt.html for more information.
- As a part of Relay’s automatic differentiation system, we are adding primal gradients for Relay operators. Please refer to #2562 for tracking the progress. (Help wanted). For automatic differentiation in Relay, also see Automatic differentiation for tensor expressions, Higher order reverse mode automatic differentiation that work with control flow, Tensor Expression level automatic differentiation
- Low-bit quantization supported (#2116). The workflow includes annotation, calibration and transformation. Please also see Data-Aware Calibration for calibration algorithm improvement (Help wanted).
Tensor Expression
- RFC for formalizing Integer Arithmetic Analysis (#2588). It is aiming to perform better context-dependent analysis, bound analysis, centralized arithmetic logic and arithmetic simplification.
Contribution and Commits
Thanks Wei @wweic for providing the tools.
People Who Reviewed Pull Requests
- tqchen Runtime, Relay, Tensor Expression, Document, Frontend, AutoTVM
- were Runtime, Hybrid Script, TOPI
- junrushao1994 Runtime, Relay, Hybrid Script
- ZihengJiang Relay, Rust, Quantization, Tensor Expression
- sgrechanik-h Tensor Expression
- derisavi Tensor Expression
- wweic Relay
- nhynes Rust, Quantization
- ehsanmok Rust
- mjs-arm Pylint
- MarisaKirisame Relay
- srkreddy1238 Relay, Tensorflow Frontend, Golang
- vinx13 Hybrid Script, Relay, TOPI, Quatization
- kazum Relay
- FrozenGene TFLite Frontend, AutoTVM, Quantization
- jroesch Relay, Rust
- zhiics Relay, Runtime, Tensorflow Frontend
- imorinaga Relay(heterogenous annotation)
- merrymercy AutoTVM, Tensor Expression, Quantization
- yzhliu AutoTVM, Relay, Tensor Expression, Frontend
- eqy AutoTVM, Quantization, Relay
- icemelon9 AutoTVM, Tensor Expression, Runtime, Relay, Frontend
- reminisce Runtime
- kevinthesun Tensor Expression, TOPI
- Anthony-Mai Tensor Expression
- slyubomirsky Relay
- joshpoll Relay Document
- eric-haibin-lin Bugfix
- ZhennanQin Bugfix
- grwlf Runtime
- xqdan Tensor Expression, Hybrid Script
- Laurawly Relay, Hybrid Script, TOPI
- masahi Relay, TOPI, Document, CodeGen
- siju-samuel Relay
- PariksheetPinjari909 Operator
- liangfu Quantization
- lixiaoquan Quantization
- ajtulloch Quantization
- zhreshold Operator
People Who Committed
- tqchen CI, Runtime, Tensor Expression, Relay Text Printer
- ruslo Documents
- mjs-arm Pylint, CI
- vinx13 Relay, TOPI, Operator
- lixiaoquan Relay
- wweic Relay Document
- kazum Tutorial and Document
- derisavi Tensor Expression (IntSet)
- antinucleon AutoTVM
- slyubomirsky Relay (ADT, AutoDiff)
- MarisaKirisame Relay
- junrushao1994 Runtime, Build
- were Hybrid Script
- yidawang Relay (MAC Calculation)
- ariwaranosai Operator
- ZihengJiang Relay, Quantization
- jroesch Relay
- hlu1 Runtime
- headupinclouds TOPI
- zhiics Relay, Tensor Expression
- weberlo Relay (param save/load)
- icemelon9 Tensor Expression, Operator
- yzhliu Tensor Expression
- eqy Relay, Quantization, AutoTVM tutorial
- geexie Bug-fix
- abergeron Conda package
- larroy NodeEntry Implement Improvement
- ptrendx Bug-fix
- Anthony-Mai Namespace Fix
- jdavies-huawei Tensor Expression
- srkreddy1238 Tensorflow Frontend, Golang
- yongwww Tensorflow Frontend, Relay Document
- alexeyr Code Improvement
- nhynes Rust
- makihiro Caffe2 Frontend
- kevinthesun AutoTVM
- Huyuwei Tutorial
- ehsanmok Rust
- siju-samuel Operator
- sgrechanik-h Tensor Expression
- haojin2 Operator
- SiNZeRo Document
- denis0x0D CodeGen
- Laurawly Bug-fix
- apivovarov Typo-fix
- take-cheeze Typo-fix
Bug Fixes
Click me to expand
- [RELAY] Fix get_int_tuple. (#2691)
- [ARITH] Select support for integer set analysis. (#2687)
- [Relay] Fix error in ANF (too agressively inline atomic expression and create free variable). (#2665)
- [Hybrid Script] Fix name conflict and attached scope problem. (#2649)
- [Relay] Fix ANF for reference and pattern matching. (#2637)
- [Relay] Fix fusion bug when call symbol that is not an operator. (#2630)
- Fix missing header file. (#2629)
- [Relay]Fix the bug in heterogeneous annotation which mistakenly steps into the fused op. (#2622)
- [AutoTVM] Fix incorrect localhost usage in RPC mode. (#2619)
- [NNVM] Fix incorrectly getting layout attribute as a tuple. (#2610)
- [Relay] Fix mutating IF expression. (#2601)
- [Tutorial] Fix downloaded file path. (#2590)
- [Storage] Fix int32 overflow bug when input is big. (#2580)
- [NNVM] Fix non-identity problem for FInplaceIdentity. (#2572)
- [Golang] Fix compilation error. (#2558)
- [Tensor Expression] Fix missing reduction init predicates. (#2495)
- [Relay] Fix missing argument for NCHWc in Relay. (#2627)
- [TOPI] Fix Nms_ir data race. (#2600)
List of Commits
(Click me) Commits in Feb 2019 with Reviewers and Authors Listed
- [BUILD] Simplify after bind device type #2670 (author: tqchen, reviewers: )
- [RUNTIME] Enable NDArray type extension #2598 (author: tqchen, reviewers: [‘tqchen’, ‘were’, ‘junrushao1994’])
- [EXPR] Expression-template based pattern matching. #2589 (author: tqchen, reviewers: [‘tqchen’, ‘ZihengJiang’, ‘sgrechanik-h’, ‘derisavi’])
- [IR] Update HalideIR #2582 (author: tqchen, reviewers: )
- Misc refactor on graph runtime, layout node #2557 (author: tqchen, reviewers: )
- [VERSION] Move version script to the project root #2556 (author: tqchen, reviewers: )
- [RELAY] TextPrinter: Use Map Format #2553 (author: tqchen, reviewers:[‘wweic’])
- [CI] Update rust format version #2550 (author: tqchen, reviewers: [‘tqchen’, ‘nhynes’, ‘ehsanmok’])
- [CI] Enable ANTLR in CPU env #2548 (author: tqchen, reviewers: )
- Few docs fixes #2703 (author: ruslo, reviewers: [‘tqchen’])
- Pin pylint version 2.2.2 #2698 (author: mjs-arm, reviewers: [‘tqchen’, ‘mjs-arm’])
- [RELAY][PASS] Enable switching CanonicalizeOps in pass_enabled #2696 (author: vinx13, reviewers: [‘tqchen’])
- [RELAY] Fix get_int_tuple for shape like ‘(1001,)’ #2691 (author: lixiaoquan, reviewers: [‘tqchen’])
- [Relay][Doc] Separate arguments types formatting with comma #2690 (author: wweic, reviewers: [‘tqchen’, ‘MarisaKirisame’, ‘wweic’])
- [DOC] MXNet frontend tutorial #2688 (author: kazum, reviewers: [‘tqchen’])
- [TVM][ARITH] Handle Select in IntSetEvaluator #2687 (author: derisavi, reviewers: [‘tqchen’])
- Stop pylint complaining about unnecessary return statement. #2684 (author: mjs-arm, reviewers: [‘tqchen’])
- Support mean in NNVM to Relay converter. #2680 (author: lixiaoquan, reviewers: [‘tqchen’])
- [Relay] Ensure nested higher-order functions are treated correctly #2676 (author: slyubomirsky, reviewers: [‘tqchen’])
- [Relay] fix name in alpha_equal #2674 (author: MarisaKirisame, reviewers: [‘tqchen’])
- Minor fix of warnings raised by -Wreturn-std-move and -Wself-assign-overloaded #2669 (author: junrushao1994, reviewers: [‘tqchen’, ‘junrushao1994’])
- [Relay] fix error in ANF (too agressively inline atomic expression and create free variable). #2665 (author: MarisaKirisame, reviewers: [‘tqchen’, ‘MarisaKirisame’])
- [RUST] ci_cpu docker image build failure. #2657 (author: mjs-arm, reviewers: [‘tqchen’, ‘nhynes’])
- Explicitly disable pylint warning subprocess-popen-preexec-fn #2656 (author: mjs-arm, reviewers: [‘tqchen’, ‘mjs-arm’])
- Stop pylint complaining about useless import alias. #2655 (author: mjs-arm, reviewers: [‘tqchen’])
- Adopt pylint suggestion to use dict.get. #2654 (author: mjs-arm, reviewers: [‘tqchen’, ‘mjs-arm’])
- [Hybrid Script] Misc fixation #2649 (author: were, reviewers: [‘tqchen’, ‘vinx13’])
- [RELAY][PASS]use attribute registration style in the mac count pass #2645 (author: yidawang, reviewers: [‘tqchen’])
- Add CONCATENATION to tflite frontend, support Inception V3 #2643 (author: ariwaranosai, reviewers: [‘tqchen’, ‘srkreddy1238’, ‘kazum’, ‘FrozenGene’, ‘ariwaranosai’])
- Fix pylint 2.2.2 gripes. #2642 (author: mjs-arm, reviewers: [‘tqchen’])
- [RELAY][PASS] Common subexpression elimination #2639 (author: vinx13, reviewers: [‘tqchen’, ‘vinx13’, ‘kazum’])
- [Relay] fix anf for reference and pattern matching #2637 (author: MarisaKirisame, reviewers: [‘tqchen’, ‘ZihengJiang’, ‘MarisaKirisame’])
- [RELAY/PASS] Change the return value convention of gradient #2635 (author: ZihengJiang, reviewers: [‘tqchen’, ‘ZihengJiang’, ‘MarisaKirisame’, ‘jroesch’])
- [RELAY/OP] Gradient of relay level1 ops #2633 (author: ZihengJiang, reviewers: [‘tqchen’, ‘ZihengJiang’, ‘MarisaKirisame’])
- Add follow up test case for #2601 #2632 (author: jroesch, reviewers: [‘tqchen’])
- Fix fusion bug when call symbol that is not an operator. #2630 (author: jroesch, reviewers: [‘tqchen’])
- [Bugfix] Missing #include #2629 (author: hlu1, reviewers: [‘tqchen’])
- [TOPI] remove batch_norm_inference #2626 (author: headupinclouds, reviewers: [‘tqchen’])
- [RELAY] Stop_fusion annotation #2624 (author: vinx13, reviewers: [‘tqchen’, ‘vinx13’])
- [Relay]fix heterogenous annotation bug #2622 (author: zhiics, reviewers: [‘tqchen’, ‘kazum’, ‘zhiics’, ‘imorinaga’])
- [Relay] Port param dict save/load from NNVM #2620 (author: weberlo, reviewers: [‘tqchen’, ‘weberlo’])
- [autotvm] fix #2617 #2619 (author: antinucleon, reviewers: [‘tqchen’, ‘merrymercy’, ‘yzhliu’, ‘eqy’, ‘icemelon9’, ‘antinucleon’])
- [TVM][LANG] Add eager simplification for operations with FloatImm #2615 (author: icemelon9, reviewers: [‘tqchen’, ‘icemelon9’])
- [RUNTIME][NDArray] Allowing External Libraries to Subclass NDArrays #2613 (author: junrushao1994, reviewers: [‘tqchen’, ‘were’, ‘junrushao1994’, ‘reminisce’])
- [Bugfix] get layout in to_relay #2610 (author: yzhliu, reviewers: [‘tqchen’, ‘icemelon9’])
- [RELAY][PASS] add a relay pass to count #macs of a model #2609 (author: yidawang, reviewers: [‘tqchen’, ‘yzhliu’, ‘eqy’, ‘zhiics’, ‘yidawang’])
- [RELAY][TOPI]
alter_op_layoutfor x86 #2602 (author: eqy, reviewers: [‘tqchen’, ‘yzhliu’]) - Fix issue mutating if expressions #2601 (author: jroesch, reviewers: [‘tqchen’])
- [DOCS] update titles to reflect tutorial content (nnvm vs. relay) #2597 (author: eqy, reviewers: [‘tqchen’, ‘kevinthesun’])
- [TEST] Remove script that references previously removed content. #2596 (author: mjs-arm, reviewers: [‘tqchen’])
- [AUTOTVM][RELAY][DOCS] relay ports of
tune_nnvm_*autotvm tutorials #2594 (author: eqy, reviewers: [‘tqchen’, ‘eqy’, ‘FrozenGene’])
A couple of fixes for GEN #2593 (author: geexie, reviewers: [‘tqchen’, ‘geexie’]) - [TUTORIAL] Fix downloaded file path #2590 (author: kazum, reviewers: [‘tqchen’])
- [TVM][Bugfix] fix storage_rewrite bug when input is big #2580 (author: zhiics, reviewers: [‘tqchen’, ‘yzhliu’, ‘wweic’, ‘zhiics’, ‘Anthony-Mai’])
- Conda packages with cuda support #2577 (author: abergeron, reviewers: [‘tqchen’, ‘abergeron’])
- Optimize move semantics of NodeEntry reducing copies of shared_ptr wh… #2576 (author: larroy, reviewers: [‘tqchen’])
- [Relay][Docs] Documentation for Algebraic Data Types #2575 (author: slyubomirsky, reviewers: [‘tqchen’, ‘slyubomirsky’, ‘wweic’, ‘zhiics’, ‘joshpoll’, ‘junrushao1994’])
- fix typo #2574 (author: MarisaKirisame, reviewers: [‘tqchen’])
- Fix the FInplaceIdentity #2572 (author: ptrendx, reviewers: [‘tqchen’, ‘yzhliu’, ‘eric-haibin-lin’, ‘ptrendx’, ‘ZhennanQin’])
- [Graph Runtime] Run_individual for benchmarking individual layers #2569 (author: hlu1, reviewers: [‘tqchen’, ‘icemelon9’, ‘zhiics’, ‘grwlf’, ‘hlu1’])
- Tighten buffer bound for TensorComputeOp by improving EvalSet on ranges #2565 (author: jdavies-huawei, reviewers: [‘tqchen’, ‘xqdan’, ‘jdavies-huawei’])
- [Relay/TOPI][Op] Add batch_matmul in relay and TOPI #2561 (author: icemelon9, reviewers: [‘tqchen’, ‘Laurawly’, ‘icemelon9’])
[Golang] bugfix #2517 #2558 (author: srkreddy1238, reviewers: [‘tqchen’, ‘srkreddy1238’]) - [COMMUNITY] @FrozenGene -> Reviewer #2544 (author: ZihengJiang, reviewers: [‘tqchen’, ‘icemelon9’])
- [RELAY] Copy subfunction arguments to output tuple field #2537 (author: vinx13, reviewers: [‘tqchen’, ‘masahi’])
print ast w/o metadata #2533 (author: zhiics, reviewers: [‘tqchen’, ‘jroesch’, ‘zhiics’]) - [DOCS][RELAY] Sync up ops with code base #2532 (author: yongwww, reviewers: [‘tqchen’, ‘yongwww’])
- Remove duplicate as Checks and CHECK value #2531 (author: alexeyr, reviewers: [‘tqchen’, ‘alexeyr’, ‘Anthony-Mai’])
addmin_repeat_msto other CUDA tutorials #2526 (author: eqy, reviewers: [‘tqchen’, ‘vinx13’]) - [Rust] Update rustc #2524 (author: nhynes, reviewers: [‘tqchen’])
- [Lang] Layout in TVM node system #2509 (author: yzhliu, reviewers: [‘tqchen’, ‘merrymercy’, ‘yzhliu’, ‘kevinthesun’])
- [Relay][Frontend] Caffe2 Support #2507 (author: makihiro, reviewers: [‘tqchen’, ‘yzhliu’, ‘kazum’, ‘makihiro’])
- [Relay][OP] Add reverse_reshape #2503 (author: icemelon9, reviewers: [‘tqchen’, ‘yzhliu’, ‘vinx13’, ‘icemelon9’, ‘jroesch’, ‘junrushao1994’])
- [Relay] GNF #2492 (author: MarisaKirisame, reviewers: [‘tqchen’, ‘MarisaKirisame’, ‘slyubomirsky’])
- Add fallback for ApplyGraphBest #2485 (author: kevinthesun, reviewers: [‘tqchen’, ‘merrymercy’, ‘kevinthesun’])
- [TEST] Remove script that references previously removed content. #2481 (author: mjs-arm, reviewers: [‘tqchen’])
- [Tutorial][Frontend] move from_keras tutorial to frontend #2479 (author: Huyuwei, reviewers: [‘tqchen’, ‘Huyuwei’, ‘eqy’])
- [Hybrid script] Backend support #2477 (author: were, reviewers: [‘tqchen’, ‘Laurawly’, ‘vinx13’, ‘were’, ‘xqdan’, ‘junrushao1994’])
- [TOPI][CUDA] Add faster-rcnn proposal op #2420 (author: vinx13, reviewers: [‘tqchen’, ‘masahi’, ‘Laurawly’, ‘vinx13’, ‘were’, ‘kevinthesun’])
- [RUST][FRONTEND] Add rust frontend v0.1 #2292 (author: ehsanmok, reviewers: [‘tqchen’, ‘ZihengJiang’, ‘nhynes’, ‘jroesch’, ‘ehsanmok’])
- REGION op removed from topi and added in darkent frontend #2275 (author: siju-samuel, reviewers: [‘tqchen’, ‘siju-samuel’, ‘PariksheetPinjari909’, ‘jroesch’])
- [RELAY][FRONTEND] Tensorflow frontend. #2216 (author: srkreddy1238, reviewers: [‘tqchen’, ‘srkreddy1238’, ‘yzhliu’, ‘kazum’, ‘jroesch’, ‘yongwww’])
- [OPT] Low-bit Quantization #2116 (author: ZihengJiang, reviewers: [‘tqchen’, ‘ZihengJiang’, ‘merrymercy’, ‘vinx13’, ‘eqy’, ‘nhynes’, ‘jroesch’, ‘liangfu’, ‘lixiaoquan’, ‘ajtulloch’, ‘FrozenGene’])
- [RELAY]Vision ops for yolo #1941 (author: siju-samuel, reviewers: [‘tqchen’, ‘siju-samuel’, ‘yzhliu’])
- [Quantize] Skip for same input-output domain scale. #2611 (author: ZihengJiang, reviewers: [‘vinx13’, ‘eqy’])
- Version 0.5 #2604 (author: ZihengJiang, reviewers: [‘ZihengJiang’, ‘yzhliu’, ‘Laurawly’, ‘icemelon9’, ‘jroesch’])
- [BUGFIX] Fix for quantize. #2573 (author: ZihengJiang, reviewers: [‘eqy’])
- [Quantization][RELAY] Add check against NCHWc ops in the quantization pass #2646 (author: eqy, reviewers: [‘ZihengJiang’])
- [TVM][BUGFIX] Fix missing reduction init predicates #2495 (author: sgrechanik-h, reviewers: [‘ZihengJiang’, ‘sgrechanik-h’])
- [Relay] Reference #2489 (author: MarisaKirisame, reviewers: [‘ZihengJiang’, ‘MarisaKirisame’, ‘slyubomirsky’, ‘zhiics’])
- [relay][frontend] TensorFlow saved model support #2586 (author: yongwww, reviewers: [‘srkreddy1238’, ‘zhiics’, ‘yongwww’])
- [Tensorflow Frontend] Add saved model support and infer shape w/o add_shapes=True #2493 (author: yongwww, reviewers: [‘tqchen’, ‘srkreddy1238’, ‘yzhliu’, ‘zhiics’, ‘yongwww’])
- [Python dep] Add missing dep pkg for relay #2568 (author: antinucleon, reviewers: [‘srkreddy1238’])
- [COMMUNITY] @merrymercy -> PMC #2578 (author: yzhliu, reviewers: [‘eqy’])
- [DOC] CoreML frontend tutorial #2667 (author: kazum, reviewers: [‘yzhliu’, ‘kazum’, ‘zhiics’])
- [Frontend][MXNet] Add MXNet frontend converter for slice operator #2662 (author: haojin2, reviewers: [‘yzhliu’, ‘haojin2’])
- Fix missing argument for NCHWc in Relay #2627 (author: eqy, reviewers: [‘yzhliu’])
- [Relay/TOPI][OP] Add arange op in Relay and TOPI #2621 (author: icemelon9, reviewers: [‘yzhliu’, ‘Laurawly’, ‘icemelon9’, ‘zhreshold’, ‘zhiics’])
- [RELAY][DOCS] Port from_mxnet tutorial to relay #2608 (author: eqy, reviewers: [‘yzhliu’, ‘eqy’, ‘zhreshold’])
- Fix typo #2595 (author: icemelon9, reviewers: [‘yzhliu’])\
- [DOC] Using External Libraries in Relay #2694 (author: SiNZeRo, reviewers: [‘masahi’, ‘SiNZeRo’])
- [CODEGEN NVPTX] Initialize llvm before lookup for the target #2683 (author: denis0x0D, reviewers: [‘masahi’, ‘denis0x0D’])
- [RELAY][OP] ROI Align #2618 (author: vinx13, reviewers: [‘masahi’, ‘Laurawly’, ‘vinx13’])
- [Bugfix] Nms_ir data_race solved #2600 (author: Laurawly, reviewers: [‘masahi’, ‘Laurawly’, ‘vinx13’])
- [Hybrid Script] Add
max_num_threads#2672 (author: were, reviewers: [‘Laurawly’, ‘vinx13’, ‘were’]) - [DOCKER][FRONTEND] Run DarkNet tests #2673 (author: kazum, reviewers: [‘kazum’])
- [COMMUNITY] @jroesch -> Committer #2661 (author: icemelon9, reviewers: [‘eqy’])
- Fix typo in Evaluate inference time cost code #2542 (author: apivovarov, reviewers: [‘eqy’])
- Fix typo #2681 (author: take-cheeze, reviewers: [‘nhynes’])
- [Relay] add more function to prelude #2660 (author: MarisaKirisame, reviewers: [‘MarisaKirisame’, ‘wweic’])
- [Relay] add more descriptive error for checked_type #2652 (author: MarisaKirisame, reviewers: [‘MarisaKirisame’])
- [Frontend][MXNet] Add MXNet frontend converter for where operator #2647 (author: haojin2, reviewers: [‘icemelon9’])
- Docker updates #2702 (author: mjs-arm, reviewers: )