Merge pull request #1308 from rstudio/release/2.8

Release/2.8

.github/workflows/R-CMD-check.yaml

@@ -54,7 +54,7 @@ jobs:
     continue-on-error: ${{ matrix.tf == 'nightly' || contains(matrix.tf, 'rc') || matrix.r == 'devel' }}
     env:
       R_REMOTES_NO_ERRORS_FROM_WARNINGS: 'true'
-      R_COMPILE_AND_INSTALL_PACKAGES: 'never'
+      # R_COMPILE_AND_INSTALL_PACKAGES: 'never' # commented out until CRAN builds TF 2.8
       GITHUB_PAT: ${{ secrets.GITHUB_TOKEN }}
 
     steps:
@@ -82,7 +82,7 @@ jobs:
         id: r-package-cache
         with:
           path: ${{ env.R_LIBS_USER }}
-          key: ${{ matrix.os }}-${{ steps.setup-r.outputs.installed-r-version }}-${{ steps.get-date.outputs.year-week }}-2
+          key: ${{ matrix.os }}-${{ steps.setup-r.outputs.installed-r-version }}-${{ steps.get-date.outputs.year-week }}-3
 
       - name: Install remotes
         if: steps.r-package-cache.outputs.cache-hit != 'true'

DESCRIPTION

@@ -1,7 +1,7 @@
 Package: keras
 Type: Package
 Title: R Interface to 'Keras'
-Version: 2.7.0.9000
+Version: 2.8.0
 Authors@R: c(
   person("Tomasz", "Kalinowski", role = c("ctb", "cph", "cre"),
          email = "[email protected]"),
@@ -51,6 +51,3 @@ Suggests:
 Roxygen: list(markdown = TRUE, r6 = FALSE)
 RoxygenNote: 7.1.2
 VignetteBuilder: knitr
-Remotes:
-    rstudio/reticulate,
-    rstudio/tensorflow

NEWS.md

@@ -1,17 +1,20 @@
-# keras (development version)
+# keras 2.8.0
 
 - Breaking change: The semantics of passing a named list to `keras_model()` have changed.
 
   Previously, `keras_model()` would `unname()` supplied `inputs` and `outputs`.
   Then, if a named list was passed to subsequent
   `fit()`/`evaluate()`/`call()`/`predict()` invocations,
-  matching of `x` and `y` was done to the model's `tensor$name`'s.
-  Now, matching is done to `names()` of `inputs` and/or `outputs`.
+  matching of `x` and `y` was done to the model's input and outpt `tensor$name`'s.
+  Now, matching is done to `names()` of `inputs` and/or `outputs` supplied to `keras_model()`.
   Call `unname()` on `inputs` and `outputs` to restore the old behavior, e.g.:
-    ```keras_model(unname(inputs), unname(outputs))```
+    ```
+    keras_model(unname(inputs), unname(outputs))
+    ```
 
   `keras_model()` can now accept a named list for multi-input and/or multi-output
-  models. (Requires Tensorflow >= 2.4). The named list is converted to a `dict` in python.
+  models. The named list is converted to a `dict` in python.
+   (Requires Tensorflow >= 2.4, Python >= 3.7).
 
   If `inputs` is a named list:
     - `call()`, `fit()`, `evaluate()`, and `predict()` methods can also
@@ -26,14 +29,19 @@
 
 - New layer `layer_depthwise_conv_1d()`.
 
-- Models gain default `format()` and `print()` S3 methods for compatibility
+- Models gain `format()` and `print()` S3 methods for compatibility
   with the latest reticulate. Both are powered by `model$summary()`.
 
+- `summary()` method for Models gains arguments `expand_nested` and `show_trainable`,
+  both default to `FALSE`.
+
 - `keras_model_custom()` is soft deprecated. Please define custom models by
   subclassing `keras$Model` directly using `%py_class%` or `R6::R6Class()`.
 
 - Fixed warning issued by `k_random_binomial()`.
+
 - Fixed error raised when `k_random_binomial()` was passed a non-floating dtype.
+
 - Added `k_random_bernouli()` as an alias for `k_random_binomial()`.
 
 - `image_load()` gains a `color_mode` argument.

R/layers-recurrent.R

@@ -239,7 +239,7 @@ layer_cudnn_gru <- function(object, units,
 #' @template roxlate-recurrent-layer
 #'
 #' @section References:
-#' - [Long short-term memory](https://www.bioinf.jku.at/publications/older/2604.pdf) (original 1997 paper)
+#' - [Long short-term memory](http://www.bioinf.jku.at/publications/older/2604.pdf) (original 1997 paper)
 #' - [Supervised sequence labeling with recurrent neural networks](https://www.cs.toronto.edu/~graves/preprint.pdf)
 #' - [A Theoretically Grounded Application of Dropout in Recurrent Neural Networks](https://arxiv.org/abs/1512.05287)
 #'
@@ -291,7 +291,7 @@ function(object,
 #' @inheritParams layer_dense
 #'
 #' @section References:
-#' - [Long short-term memory](https://www.bioinf.jku.at/publications/older/2604.pdf) (original 1997 paper)
+#' - [Long short-term memory](http://www.bioinf.jku.at/publications/older/2604.pdf) (original 1997 paper)
 #' - [Supervised sequence labeling with recurrent neural networks](https://www.cs.toronto.edu/~graves/preprint.pdf)
 #' - [A Theoretically Grounded Application of Dropout in Recurrent Neural Networks](https://arxiv.org/abs/1512.05287)
 #'

vignettes/new-guides/working_with_rnns.Rmd

@@ -49,7 +49,7 @@ timestep is to be fed to the next timestep.
 [Cho et al., 2014](https://arxiv.org/abs/1406.1078).
 
 3. `layer_lstm()`, first proposed in
-[Hochreiter & Schmidhuber, 1997](https://www.bioinf.jku.at/publications/older/2604.pdf).
+[Hochreiter & Schmidhuber, 1997](http://www.bioinf.jku.at/publications/older/2604.pdf).
 
 Here is a simple example of a sequential model that processes sequences of integers,
 embeds each integer into a 64-dimensional vector, then processes the sequence of