Arraymancer
Arraymancer copied to clipboard
A fast, ergonomic and portable tensor library in Nim with a deep learning focus for CPU, GPU and embedded devices via OpenMP, Cuda and OpenCL backends
This is necessary to build NN models by block The current TrainableLayer concept only check the first level fields for `Variable`: https://github.com/mratsim/Arraymancer/blob/90263efabdf706d7a74c1d847b7449495d67c124/src/nn_dsl/dsl_types.nim#L61-L79 Furthermore `TrainableLayers` should be ironed out and moved...
Closes #27, the JS target is not suitable as we rely a lot on shared memory. Requires a WebAssembly BLAS.
See #329, there is probably some false sharing in the GRU code (or in the Embedding `flatten_idx` proc but less likely).
Reshaping and flattening before feeding to a linear layer gets long in the tooth.
See https://scikit-learn.org/stable/modules/generated/sklearn.feature_extraction.text.TfidfVectorizer.html A very nice introduction to TF-IDF is available here: https://cran.r-project.org/web/packages/tidytext/vignettes/tf_idf.html Efficient vectorization probably requires sparse matrices support (#28)
We have an embedding layer (#312), we have GRU with sequence support (#283). We miss a dataset and an NLP example. The IMDB dataset is probably the one to have...
There is no tokenizer currently which makes parsing text and using word embeddings very hard. I.e. currently it's either a split on white spaces or regexp.
IMDB dataset takes a while to download and nothing is printed in the command-line. A progress bar would be nice. Documentation: https://nim-lang.org/docs/httpclient.html#progress-reporting > Progress reporting > ================== > > You...
Taking the definition from example 5: ```Nim network ctx, TheGreatSequencer: layers: # Note input_shape will only require the number of features in the future # Input shape = [seq_len, Batch,...
## Shuffle Deterministic shuffles are needed in general for both deep learning and machine learning. In many cases input data might be ordered, for example IMDB is all positive reviews,...