4
I Use This!
Inactive

Commits : Listings

Analyzed 1 day ago. based on code collected 2 days ago.
Mar 26, 2023 — Mar 26, 2024
Commit Message Contributor Files Modified Lines Added Lines Removed Code Location Date
Initial commit More... over 5 years ago
update (C) date More... almost 6 years ago
Turbulence lab - Initial commit More... almost 6 years ago
improved sRNN More... about 7 years ago
improved sRNN More... about 7 years ago
minor tweaks More... about 7 years ago
minor perf improvements More... over 7 years ago
ce loss fixes, sRnn bptt fix More... over 7 years ago
using SoftMax TransofmOp for more compact code More... over 7 years ago
minor fixes More... over 7 years ago
fixed sampling code and loss update More... over 7 years ago
minor fixes More... over 7 years ago
bugfixes to bptt in srnn, minor fixes More... over 7 years ago
fixed wrong adagrad update, more cli params for nn runner More... over 7 years ago
added apprunner plugin, fixed bug in sRNN, adjusted test More... over 7 years ago
fixed sample generation More... over 7 years ago
refactored char/word/stacked RNNs in 2 classes More... over 7 years ago
added sRNN More... over 7 years ago
rnn loosing memory of previous hidden state - fixed More... over 7 years ago
char rnn refactoring, word rnn, added cv tests with different hyperparams More... over 7 years ago
minor fixes, java raw char-rnn model inspired by karpathy More... over 7 years ago
fixed bug on mini batch More... about 8 years ago
skipgram using configurable mini batch More... about 8 years ago
added nesterov momentum to skipgram, other minor tweaks More... about 8 years ago
fixed momentum impl, more appropriate softmax usage More... about 8 years ago
biases initialized to 0.001, per output word softmax, cached expanded representation in hot encoded samples, 0.5 decay on learning rate More... about 8 years ago
initialize weights to uniform distribution, adjusted test texts, added option to use momentum, minor fixes More... about 8 years ago
refactored MLN and improved SGN, commons-math 3.6 More... about 8 years ago
avoiding some matrix transpose calls More... about 8 years ago
refactored SFFNN to MLN, added ReLU function and (compact) skip-gram More... about 8 years ago