google-research/simclr issues and pull requests

#219 - Is there a way to load the savedModel as a keras model ?

Issue - State: open - Opened by capmac18 4 months ago

#218 - Cannot find the pretrained model for cifar-10 in simclrv1

Issue - State: open - Opened by RTRodoshi 11 months ago

#217 - Bump tensorflow from 1.15.4 to 2.11.1

Pull Request - State: open - Opened by dependabot[bot] over 1 year ago
Labels: dependencies

#216 - Regarding the situation where positive samples in a batch_size are treated as negative samples.

Issue - State: open - Opened by YanziWang-dot over 1 year ago - 1 comment

#215 - Reproduce the results of SimCLR_v2 Finetune_10pct checkpoints

Issue - State: open - Opened by BierOne over 1 year ago - 1 comment

#214 - Reported imagenet results (e.g., Table 6) base on the validation set?

Issue - State: closed - Opened by BierOne over 1 year ago - 2 comments

#213 - ProjectionHead setting in tf2

Issue - State: open - Opened by geanshu over 1 year ago

#212 - Linear Fine-tuning Setting

Issue - State: open - Opened by ekurtulus over 1 year ago

#211 - Finetune on custom dataset locally using TF2

Issue - State: open - Opened by vsk-phi over 1 year ago - 1 comment

#210 - A quick question about the definition of SimCLR

Issue - State: open - Opened by HelloWorldLTY over 1 year ago - 2 comments

#209 - Bump tensorflow from 1.15.4 to 2.9.3

Pull Request - State: closed - Opened by dependabot[bot] almost 2 years ago - 1 comment
Labels: dependencies

#208 - Make simclr installable via pip

Pull Request - State: open - Opened by jayathungek almost 2 years ago - 1 comment

#207 - layer freezing during pretraining

Issue - State: open - Opened by HIM003 about 2 years ago

#206 - About the training epochs

Issue - State: open - Opened by utjune about 2 years ago - 2 comments

#205 - Why use bn in projection head?

Issue - State: open - Opened by shesung about 2 years ago

#204 - remove train_summary_steps from tf2/README.md

Pull Request - State: open - Opened by billcai about 2 years ago - 1 comment

#203 - Is it possible to use gradient accumulation to counter small GPU memory?

Issue - State: open - Opened by ChrisJWest about 2 years ago - 3 comments

#202 - Dividing contrastive loss by number of replicas -- correct?

Issue - State: open - Opened by piotrmwojcik about 2 years ago

#201 - Bump tensorflow from 1.15.4 to 2.7.2

Pull Request - State: closed - Opened by dependabot[bot] over 2 years ago - 1 comment
Labels: dependencies

#200 - Bump tensorflow from 1.15.4 to 2.6.4

Pull Request - State: closed - Opened by dependabot[bot] over 2 years ago - 1 comment
Labels: dependencies

#199 - Is the trained projection head available?

Issue - State: open - Opened by lkshrsch over 2 years ago - 7 comments

#198 - Question about defining the contrastive loss on zi’s rather than hi’s

Issue - State: open - Opened by garynlfd over 2 years ago

#197 - Dev lmu

Pull Request - State: closed - Opened by itolgaozturk over 2 years ago - 1 comment

#196 - '_UserObject' object has no attribute 'summary'

Issue - State: open - Opened by aynesss over 2 years ago - 4 comments

#195 - WARNING:absl:Importing a function (__inference_sync_batch_normalization_13_layer_call_and_return_conditional_losses_20965) with ops with custom gradients. Will likely fail if a gradient is requested.

Issue - State: open - Opened by aynesss over 2 years ago

#194 - Fine-tuning the model on a custom dataset?

Issue - State: closed - Opened by ImSo3K over 2 years ago

#193 - when using the saved_model format for storing pretrained model, model loading causes many unnecessary warnings.

Issue - State: open - Opened by aynesss over 2 years ago

#192 - Load Model from Checkpoints After Pre-training for Fine-Tuning Issue

Issue - State: open - Opened by fatmacelebi over 2 years ago

#191 - Bump tensorflow from 1.15.4 to 2.5.3

Pull Request - State: closed - Opened by dependabot[bot] over 2 years ago - 1 comment
Labels: dependencies

#190 - Error when Finetuning: WARNING:absl:Importing a function (inference_call___15879) with ops with custom gradients. Will likely fail if a gradient is requested.

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

#189 - Spikes in training loss

Issue - State: open - Opened by slala2121 over 2 years ago - 1 comment

#188 - Loss, entropy, accuracy trends

Issue - State: open - Opened by slala2121 over 2 years ago - 5 comments

#187 - simclr on medical dataset gives error

Issue - State: open - Opened by fatmacelebi over 2 years ago - 7 comments

#186 - Is training on multiple GPUs possible?

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

#185 - Models trained only with self supervision

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

#184 - Dataset Split for Linear Evaluation and Fine Tuning

Issue - State: open - Opened by vamsi231297 over 2 years ago - 1 comment

#183 - Two potential Bugs

Issue - State: open - Opened by AlexMaOLS over 2 years ago

#182 - Issue in Finetune V2 simclr on Cifar10

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

#181 - tf2 saved_models with no projection_head output

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

#180 - Update README.md

Pull Request - State: open - Opened by BalajiAI almost 3 years ago

#179 - Update README.md

Pull Request - State: closed - Opened by BalajiAI almost 3 years ago

#178 - fix typo

Pull Request - State: open - Opened by wit543 almost 3 years ago

#177 - Saving model after finetuning on colab tf2

Issue - State: open - Opened by junhuplim almost 3 years ago

#176 - Why do we still need a dataset with LABELS? Isn't that Self-supervised learning supposed to work without labels?

Issue - State: closed - Opened by myagmur01 almost 3 years ago - 1 comment

#175 - A baseline supervised model is trained using random initialization.

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

#174 - Dataset of supervised baseline

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

#173 - supervised baseline was intialized with weight = None or "Imagenet"

Issue - State: open - Opened by aynesss almost 3 years ago - 2 comments

#172 - Trainable=False or Trainable = "True" in Finetune

Issue - State: open - Opened by aynesss almost 3 years ago - 1 comment

#171 - Data augmentation details

Issue - State: open - Opened by ines321 almost 3 years ago

#170 - Are data augmentation operations required at the finetuning level?

Issue - State: open - Opened by ines321 almost 3 years ago - 2 comments

#169 - Retrieve eval prediction

Pull Request - State: closed - Opened by JohnMcSpedon about 3 years ago - 1 comment

#168 - Deploy to cloud TPU

Pull Request - State: closed - Opened by JohnMcSpedon about 3 years ago - 1 comment

#167 - where are "EMA" decay and memory buffer used in this simclr work?

Issue - State: open - Opened by fortisaqua about 3 years ago - 4 comments

#166 - Could you clarify `with proper learning rate scaling across batch size` in "Intriguing Properties of Contrastive Losses"?

Issue - State: closed - Opened by nzw0301 about 3 years ago - 4 comments

#165 - large number question

Issue - State: open - Opened by yinciki about 3 years ago - 3 comments

#164 - Bump tensorflow from 1.15.4 to 2.5.1

Pull Request - State: closed - Opened by dependabot[bot] about 3 years ago - 1 comment
Labels: dependencies

#163 - Transfer learning

Issue - State: open - Opened by fjsaezm about 3 years ago - 4 comments

#162 - Error in Generalized contrastive loss

Issue - State: open - Opened by aakash-saboo about 3 years ago - 5 comments

#161 - Command about fine-tune and semi-supervised learning on tf2

Issue - State: open - Opened by lucasliunju about 3 years ago - 2 comments

#160 - Error about tf2

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

#159 - Why knowledge distillation improves the test accuracy?

Issue - State: open - Opened by OutstanderWang about 3 years ago - 1 comment

#158 - Constrastive loss input format in the code

Issue - State: open - Opened by Darthholi about 3 years ago - 1 comment

#157 - Execution in Google Colab

Issue - State: closed - Opened by fjsaezm about 3 years ago

#156 - Unknown command line flag 'variable_schema'

Issue - State: closed - Opened by liuzheyu1998 about 3 years ago - 2 comments

#155 - A typo in center_crop funtion?

Issue - State: open - Opened by lyuyunhuan about 3 years ago - 1 comment

#154 - What about simCLR performance, it's time work on mobile devices ?

Issue - State: open - Opened by RichardLiee about 3 years ago

#153 - SimCLR, augmentation generates one images instead of two

Issue - State: open - Opened by ines321 over 3 years ago - 1 comment

#152 - Is crop and resize necessary to dataset having same images size ?

Issue - State: open - Opened by ines321 over 3 years ago - 1 comment

#151 - linear eval while pretraining gradient

Issue - State: closed - Opened by slala2121 over 3 years ago - 2 comments

#150 - I got curious while reading the paper. Why do we use just one positive sample?

Issue - State: open - Opened by babbu3682 over 3 years ago - 1 comment

#149 - Multi-GPU Finetuning

Issue - State: open - Opened by edwinytgoh over 3 years ago - 7 comments

#148 - RET_CHECK failure (tensorflow/core/tpu/graph_rewrite/distributed_tpu_rewrite_pass.cc:1768) arg_shape.handle_type != DT_INVALID when using TPU VM

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

#129 - ERROR:tensorflow:Error recorded from training_loop: From /job:worker/replica:0/task:0: File system scheme '[local]' not implemented (file: 'pretrain_model/model.ckpt-0_temp_8abb8541792446a4a05c117665873ac7/part-00000-of-00001')

Issue - State: open - Opened by Riya-11 over 3 years ago - 4 comments

GitHub / google-research/simclr issues and pull requests