Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / juntang-zhuang/Adabelief-Optimizer issues and pull requests

#68 - adapt to Tensorflow >= 2.11

Pull Request - State: open - Opened by bertsky about 1 month ago

#67 - loss become nan when beta1=0

Issue - State: open - Opened by yojeep 5 months ago

#66 - AttributeError: 'AdaBeliefOptimizer' object has no attribute '_set_hyper'

Issue - State: open - Opened by SamMohel 10 months ago - 4 comments

#65 - The problem of reproducing the result of ImageNet

Issue - State: open - Opened by KaltsitI over 1 year ago - 4 comments

#64 - Suppressing weight decoupling and rectification messages

Issue - State: open - Opened by gunsodo almost 2 years ago - 1 comment

#63 - Update README.md

Pull Request - State: closed - Opened by yifan16 about 2 years ago

#62 - Update README.md

Pull Request - State: closed - Opened by yifan16 about 2 years ago

#61 - Inconsistent use of epsilon

Issue - State: closed - Opened by cossio over 2 years ago - 4 comments

#60 - weight_decouple in adabelief tf

Issue - State: closed - Opened by YannPourcenoux over 2 years ago - 1 comment

#59 - Tensorflow restoration issue

Issue - State: closed - Opened by soumen-ghosh over 2 years ago - 1 comment

#58 - Some questions related to import adabelief

Issue - State: closed - Opened by HelloWorldLTY over 2 years ago - 2 comments

#57 - Your method is just equivalent to SGD with a changable global learning rate.

Issue - State: closed - Opened by Yonghongwei almost 3 years ago - 3 comments

#56 - Inconsistent computation of weight_decay and grad_residual among pytorch versions

Issue - State: open - Opened by sjscotti almost 3 years ago - 5 comments

#55 - Compatibility with warmup

Issue - State: closed - Opened by joihn almost 3 years ago - 2 comments

#54 - Question about SGD optimizer in LSTM experiments

Issue - State: closed - Opened by yunfei-teng almost 3 years ago - 1 comment

#53 - Changing init learning rate

Issue - State: closed - Opened by Kraut-Inferences about 3 years ago - 2 comments

#52 - FileNotFoundError for ImageNet

Issue - State: closed - Opened by kchak31 over 3 years ago - 1 comment

#51 - Documentation (at least for TF) and weight_decouple is not an option

Issue - State: open - Opened by grofte over 3 years ago - 2 comments

#50 - On imagenet accuracy result 70.08

Issue - State: closed - Opened by wyzjack over 3 years ago - 1 comment

#49 - Fix Typo

Pull Request - State: closed - Opened by lorenzoprincipi over 3 years ago

#48 - Why does g_t substract m_t, instead of m_{t-1} ?

Issue - State: closed - Opened by zxteloiv over 3 years ago - 1 comment

#47 - MSVAG

Issue - State: closed - Opened by densechen over 3 years ago - 1 comment

#46 - Implementation of pure keras

Pull Request - State: open - Opened by liaoxuanzhi over 3 years ago - 6 comments

#45 - Upgrade with Adas optimizer

Issue - State: closed - Opened by DaniyarM over 3 years ago - 3 comments

#44 - Create LICENSE

Pull Request - State: closed - Opened by juntang-zhuang over 3 years ago

#43 - Please add a license

Issue - State: closed - Opened by 1e100 over 3 years ago - 1 comment

#42 - fine-tune with bert models

Issue - State: closed - Opened by JaheimLee over 3 years ago - 2 comments

#40 - Imagenette baseline for AdaBelief

Issue - State: closed - Opened by tmabraham over 3 years ago - 4 comments

#39 - Tf 0.3.0

Pull Request - State: closed - Opened by cryu854 over 3 years ago - 1 comment

#38 - i use adabelief optimizer on fine-tune efficientb4 that acc is worse than Adam?

Issue - State: closed - Opened by daixiangzi over 3 years ago - 26 comments

#37 - support for tensorflow 1.10+

Issue - State: open - Opened by chenxinhua over 3 years ago - 8 comments

#36 - update_tf_0.2.1 (add contributor)

Pull Request - State: closed - Opened by cryu854 almost 4 years ago - 1 comment

#35 - Remove in-place add of eps

Pull Request - State: closed - Opened by vpj almost 4 years ago - 1 comment

#34 - Tensorflow Implementation

Issue - State: closed - Opened by ManoharSai2000 almost 4 years ago - 14 comments

#33 - Do the weight decay before using grad

Pull Request - State: closed - Opened by vpj almost 4 years ago - 13 comments

#32 - Different usage of eps between "A quick look at the algorithm" and the code

Issue - State: closed - Opened by tatsuhiko-inoue almost 4 years ago - 10 comments

#31 - Should this work with Mixed precision training (AMP)

Issue - State: closed - Opened by Mut1nyJD almost 4 years ago - 6 comments

#30 - Update ImageNet default weight decay

Pull Request - State: closed - Opened by arrufat almost 4 years ago - 4 comments

#29 - what is details about the experiments for cifar-100

Issue - State: closed - Opened by XieBinghui almost 4 years ago - 3 comments

#28 - Remove some redundancies

Pull Request - State: closed - Opened by cryu854 almost 4 years ago - 1 comment

#27 - issues on AdaBlief-tensorflow

Issue - State: closed - Opened by dusk666 almost 4 years ago - 7 comments

#26 - raw results

Issue - State: closed - Opened by skyshoumeng almost 4 years ago - 2 comments

#25 - degenerated_to_sgd hyperparameter -- background and recommendations?

Issue - State: closed - Opened by evanatyourservice almost 4 years ago - 2 comments

#24 - Epsilon is important to Adaptive Optimizer

Issue - State: closed - Opened by yuanwei2019 almost 4 years ago - 1 comment

#23 - Is extra epsilon more important than belief?

Issue - State: closed - Opened by yasutoshi almost 4 years ago - 4 comments

#22 - Matlab implementation

Issue - State: closed - Opened by pcwhy almost 4 years ago - 8 comments

#21 - recommended experiments

Issue - State: closed - Opened by dvolgyes almost 4 years ago - 1 comment

#20 - Fix problem with sparse layers in tf0.1.0

Pull Request - State: closed - Opened by cryu854 almost 4 years ago - 8 comments

#19 - 0.1.0 changes for ranger_adabelief

Issue - State: closed - Opened by bratao almost 4 years ago - 6 comments

#17 - RangerAdaBelief setstate

Issue - State: closed - Opened by soloice almost 4 years ago - 2 comments

#16 - Similarity to AdaHessian

Issue - State: closed - Opened by davda54 almost 4 years ago - 7 comments

#15 - fix deprecation warnings for pytorch 1.6+; streamline amsgrad option …

Pull Request - State: closed - Opened by pdimitrov-thoughtriver almost 4 years ago - 5 comments

#14 - Fix a typo in README.md

Pull Request - State: closed - Opened by yueyericardo almost 4 years ago - 1 comment

#13 - torch version requirement

Issue - State: closed - Opened by leonzgtee almost 4 years ago

#12 - Make it compatible with tensorflow and keras

Pull Request - State: closed - Opened by cryu854 almost 4 years ago - 23 comments

#11 - Results on ImageNet with tuning weight decay

Issue - State: closed - Opened by XuezheMax almost 4 years ago - 11 comments

#10 - Unstability in training in RNN

Issue - State: closed - Opened by bratao almost 4 years ago - 7 comments

#9 - UserWarning: This overload of add_ is deprecated

Issue - State: closed - Opened by iiSeymour almost 4 years ago - 1 comment

#8 - Performance vs AdamW

Issue - State: closed - Opened by iiSeymour almost 4 years ago - 10 comments

#7 - keyerror exp_avg_var

Issue - State: closed - Opened by mcmingchang almost 4 years ago - 5 comments

#6 - Unfair comparison on ImageNet?

Issue - State: closed - Opened by XuezheMax almost 4 years ago - 2 comments

#5 - scripts for the toy examples?

Issue - State: closed - Opened by XuezheMax almost 4 years ago - 3 comments

#4 - Debug prints in ranger-adabelief

Issue - State: closed - Opened by iiSeymour almost 4 years ago - 4 comments

#3 - Question: How similar or dissimilar is this compared to Hypergradient Descent?

Issue - State: closed - Opened by muellerzr almost 4 years ago - 2 comments

#2 - Tensorflow implementation doesn't work

Issue - State: closed - Opened by ben-arnao almost 4 years ago - 3 comments

#1 - Rename readme.txt to readme.md

Pull Request - State: closed - Opened by RahulBhalley almost 4 years ago - 1 comment