Ecosyste.ms: Issues

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

GitHub / jcjohnson/torch-rnn issues and pull requests

#235 - Bump numpy from 1.10.4 to 1.22.0

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

#234 - Bump numpy from 1.10.4 to 1.21.0

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

#233 - Training error

Issue - State: open - Opened by admeliora over 4 years ago

#231 - Sampling output file has Chinese characters in it

Issue - State: open - Opened by Samad999777 about 5 years ago

#229 - how to optimize

Issue - State: open - Opened by goog over 5 years ago

#228 - how to set learn rate?

Issue - State: closed - Opened by goog over 5 years ago

#227 - Use of the generated json during checkpoint?

Issue - State: open - Opened by binary-person over 5 years ago

#226 - Implementation of Tensorboard?

Issue - State: open - Opened by binary-person over 5 years ago

#225 - Loss increases gradually

Issue - State: open - Opened by binary-person over 5 years ago - 1 comment

#224 - ./util/utils.lua:43: attempt to index local 'f' (a nil value)

Issue - State: open - Opened by johannaflato over 6 years ago - 1 comment

#222 - Support binary format

Pull Request - State: open - Opened by ChristianRomberg over 6 years ago

#221 - sample.lua doesn't support bytes output

Issue - State: open - Opened by ChristianRomberg over 6 years ago

#220 - Cannot serialise number: must not be NaN or Infinity

Issue - State: open - Opened by thehowl over 6 years ago - 1 comment

#219 - Added a '-seed' option to sample.lua to allow for an identical rerun.…

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

#218 - Convert to classification

Issue - State: open - Opened by rookie32 almost 7 years ago

#217 - Deterministic output of sample.lua with -start_test flag

Issue - State: closed - Opened by turbotas almost 7 years ago - 6 comments

#216 - Fixed broken lua-cjson link

Pull Request - State: open - Opened by binhonglee almost 7 years ago

#215 - Updates for compatibility with changes in new HDF5 packages

Pull Request - State: open - Opened by sharpie7 almost 7 years ago

#214 - lua:56: expected align(#) on line 579

Issue - State: open - Opened by Kirthana30 almost 7 years ago

#213 - Make the model "forget" or modify vocabulary it has been trained on?

Issue - State: closed - Opened by hackily about 7 years ago - 2 comments

#212 - HDF5 No accessibility and not valid?

Issue - State: closed - Opened by weemeghann about 7 years ago - 2 comments

#211 - Error training

Issue - State: open - Opened by HarvsG about 7 years ago - 1 comment

#210 - Arch Linux multilib install error

Issue - State: closed - Opened by laudiacay about 7 years ago - 1 comment

#209 - Running in CPU mode (-GPU -1) but it's still looking for cutorch?

Issue - State: closed - Opened by lilybradic about 7 years ago - 2 comments

#208 - Nil value call in train.lua?

Issue - State: closed - Opened by anonezumi about 7 years ago - 1 comment

#207 - train.py not properly parsing lua config strings

Issue - State: closed - Opened by mattleblanc about 7 years ago

#206 - 'Unsupported HDF5 version: 1.10.1' when running train.lua

Issue - State: closed - Opened by ajakk about 7 years ago - 5 comments

#205 - Manual Manipulation of Parameters

Issue - State: open - Opened by mattebaughman1 over 7 years ago

#204 - Error in cutorch when attempting to train (Ubuntu 16.04)

Issue - State: open - Opened by thrunq over 7 years ago - 1 comment

#203 - question] how to add top-down connections in LSTM?

Issue - State: open - Opened by cseeker over 7 years ago

#202 - Complete beginner help please

Issue - State: open - Opened by ParkertronDelta over 7 years ago - 4 comments

#201 - Error running train.lua bad argument #2 to '?'

Issue - State: closed - Opened by its5Q over 7 years ago - 3 comments

#199 - passing multiple input vectors / audio frames in VanillaRNN

Issue - State: closed - Opened by saurabhvyas over 7 years ago - 1 comment

#198 - Preprocess words

Pull Request - State: open - Opened by dgcrouse over 7 years ago - 1 comment

#197 - Created new branch to merge Py2 Preprocess improvements

Issue - State: open - Opened by dgcrouse over 7 years ago

#195 - Automatically reduce batch size for small validation/test splits

Pull Request - State: closed - Opened by antihutka over 7 years ago - 1 comment

#194 - train.lua looks for cutorch even though "-gpu 0"

Issue - State: closed - Opened by wsanders over 7 years ago - 2 comments

#193 - Won't build h5py in Ubuntu 17

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

#192 - "luarocks install cltorch" no longer supported

Issue - State: open - Opened by ryuhimora over 7 years ago - 1 comment

#190 - Various installations issues [ubuntu-server 16.04.2 LTS]

Issue - State: open - Opened by DarkMorford over 7 years ago - 1 comment

#189 - Is my training running?

Issue - State: open - Opened by Quambus over 7 years ago - 9 comments

#188 - GPU Error

Issue - State: closed - Opened by Wrongful over 7 years ago - 6 comments

#187 - LSTM: Move Wx matrix multiplication out of the loop in forward

Pull Request - State: open - Opened by antihutka over 7 years ago - 1 comment

#186 - Two questions about sampling

Issue - State: open - Opened by Wrongful over 7 years ago - 4 comments

#185 - Bring back and fix TemporalAdapter

Pull Request - State: open - Opened by antihutka over 7 years ago

#184 - Old Issues

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

#183 - Code Review Standards/Pull Request Management

Issue - State: open - Opened by dgcrouse over 7 years ago - 3 comments
Labels: help wanted

#182 - Feature: Generic algorithm for word-tokens, Unicode, and many other encodings including non-text sequences.

Issue - State: closed - Opened by qwertystop over 7 years ago - 6 comments
Labels: enhancement

#181 - Add epoch and validation loss to checkpoint

Pull Request - State: open - Opened by calittle over 7 years ago - 1 comment

#177 - Fill and write each array before creating the next one, to save memory.

Pull Request - State: open - Opened by qwertystop over 7 years ago - 4 comments

#164 - print traning_loss in the checkpoint filename

Issue - State: open - Opened by zhenyulin over 7 years ago - 3 comments
Labels: enhancement

#158 - "luajit: not enough memory" during sampling

Issue - State: open - Opened by danieleghisi almost 8 years ago - 3 comments
Labels: bug

#155 - Augmented input at time T based on output at time T-1

Issue - State: open - Opened by suryabhupa almost 8 years ago - 4 comments
Labels: question

#152 - enhancement to have keyboard command to save checkpoint now

Issue - State: closed - Opened by wrapperband almost 8 years ago - 1 comment
Labels: enhancement

#150 - location of the server

Issue - State: closed - Opened by RSstevenwong about 8 years ago - 3 comments
Labels: question

#148 - Trouble sampling on Raspberry Pi 2 model B

Issue - State: closed - Opened by JarlErikC about 8 years ago - 2 comments
Labels: duplicate

#146 - Sampling making sense text

Issue - State: closed - Opened by PauloOgraddy about 8 years ago - 25 comments

#145 - expected align(#) on line 579

Issue - State: closed - Opened by himanshuUsc about 8 years ago - 3 comments
Labels: question

#143 - can this works with glibc <2.14 ?

Issue - State: closed - Opened by foice about 8 years ago - 1 comment

#141 - Using stored checkpoint in a java program

Issue - State: closed - Opened by already-taken-m17 about 8 years ago - 4 comments

#139 - TemporalConvolution question

Issue - State: open - Opened by dgcrouse about 8 years ago
Labels: question

#136 - Feature ideas: scroing, erasures, binary data

Issue - State: open - Opened by vi about 8 years ago - 3 comments
Labels: enhancement

#135 - Crashes on small dataset.

Issue - State: closed - Opened by demikandr about 8 years ago - 5 comments
Labels: bug

#132 - Add preprocess script to use words as tokens with typo and rare word reduction

Pull Request - State: open - Opened by dgcrouse over 8 years ago - 6 comments

#131 - Feature Proposal - More robust resuming?

Issue - State: closed - Opened by dgcrouse over 8 years ago - 4 comments
Labels: enhancement

#125 - Feature request: Checkpoint on time delta

Issue - State: open - Opened by ChrisCummins over 8 years ago - 2 comments
Labels: enhancement

#123 - Preprocess

Issue - State: open - Opened by PauloOgraddy over 8 years ago
Labels: enhancement

#106 - Torch-RNN poetry

Issue - State: open - Opened by ghost over 8 years ago

#105 - Unsupported HDF5 version

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

#101 - [Q] making a simple model for a sequence learning

Issue - State: open - Opened by yongduek over 8 years ago

#100 - Memory usage of torch-rnn

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

#99 - Why do we need to ``luarocks install torch`` at the beginning

Issue - State: closed - Opened by vinhqdang over 8 years ago - 2 comments

#98 - Dealing with unseen characters?

Issue - State: open - Opened by vinhqdang over 8 years ago

#97 - Variable-length sequences

Issue - State: closed - Opened by 3rd3 over 8 years ago - 1 comment

#96 - Does torch-rnn distinguish lower and upper case?

Issue - State: closed - Opened by vinhqdang over 8 years ago - 1 comment

#95 - Plotting loss values

Issue - State: closed - Opened by vinhqdang over 8 years ago - 1 comment

#94 - baseline performance of RNN and LSTM?

Issue - State: closed - Opened by okvol over 8 years ago - 1 comment

#92 - Saving/restoring network state?

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

#91 - Sampling fails at random times giving the following error

Issue - State: open - Opened by keerthiskating over 8 years ago - 4 comments

#90 - Installing torch-rnn on Windows

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

#89 - Could I generate more than 1 character at the same time?

Issue - State: closed - Opened by vinhqdang over 8 years ago - 19 comments

#86 - Error running with OpenCL

Issue - State: open - Opened by vinhqdang over 8 years ago - 2 comments

#85 - Can gradient noise be added as an option?

Issue - State: open - Opened by AlekzNet over 8 years ago - 1 comment

#83 - makes preprocess.py python3-compatible

Pull Request - State: closed - Opened by una-dinosauria over 8 years ago - 1 comment

#82 - Starting from checkpoint and not resetting iterations raises an error

Issue - State: open - Opened by Reecer6 over 8 years ago - 3 comments

#81 - Why does number of epochs exceed the ``max_epochs`` variable

Issue - State: open - Opened by vinhqdang over 8 years ago - 1 comment

#80 - Sample text in CPU-only mode from checkpoint generated with GPU?

Issue - State: closed - Opened by kadamwhite over 8 years ago - 3 comments

#79 - How could I know the name of t7 file after training

Issue - State: closed - Opened by vinhqdang over 8 years ago - 1 comment

#78 - Generated "text" always look like "start_text'

Issue - State: closed - Opened by vinhqdang over 8 years ago - 1 comment

#77 - Working with numbers instead of character

Issue - State: open - Opened by vinhqdang over 8 years ago - 1 comment

#76 - documentation on how the faster and memory-efficient results are achieved

Issue - State: closed - Opened by blackyang over 8 years ago - 3 comments