Merge pull request #1494 from rstudio/cran/1.3.0

CRAN Release v1.3.0

Author: t-kalinowski

DESCRIPTION

@@ -1,7 +1,7 @@
 Package: keras3
 Type: Package
 Title: R Interface to 'Keras'
-Version: 1.2.0.9000
+Version: 1.3.0
 Authors@R: c(
   person("Tomasz", "Kalinowski", role = c("aut", "cph", "cre"),
          email = "tomasz@posit.co"),

NEWS.md

@@ -1,4 +1,4 @@
-# keras3 (development version)
+# keras3 1.3.0
 
 - Keras now uses `reticulate::py_require()` to resolve Python dependencies.
   Calling `install_keras()` is no longer required (but is still supported).
@@ -13,6 +13,9 @@
 
 - `%*%` now dispatches to `op_matmul()` for tensorflow tensors, which
   has relaxed shape constraints compared to `tf$matmul()`.
+  
+- Fixed an issue where calling a `Metric` and `Loss` object 
+  with unnamed arguments would error.
 
 ## Added compatibility with Keras v3.8.0. User-facing changes:
 

R/install.R

@@ -128,23 +128,29 @@ is_linux <- function() {
 
 #' Configure a Keras backend
 #'
-#' @param backend string, can be `"tensorflow"`, `"jax"`, `"numpy"`, or `"torch"`.
-#' @param gpu bool, whether to use the GPU.
+#' @param backend string, can be `"tensorflow"`, `"jax"`, `"numpy"`, or
+#'   `"torch"`.
+#' @param gpu bool, whether to use the GPU. If `NA` (default), it will attempt
+#'   to detect GPU availability on Linux. On M-series Macs, it defaults to
+#'   `FALSE` for TensorFlow and `TRUE` for JAX. On Windows, it defaults to
+#'   `FALSE`.
 #'
 #' @details
-#' These functions allow configuring which backend keras will use.
-#' Note that only one backend can be configured at a time.
+#'
+#' These functions allow configuring which backend keras will use. Note that
+#' only one backend can be configured at a time.
 #'
 #' The function should be called after `library(keras3)` and before calling
 #' other functions within the package (see below for an example).
 #'
-#' There is experimental support for changing the backend after keras has initialized.
-#' using `config_set_backend()`.
+#' There is experimental support for changing the backend after keras has
+#' initialized. using `config_set_backend()`.
 #' ```r
 #' library(keras3)
 #' use_backend("tensorflow")
 #' ```
-#' @returns Called primarily for side effects. Returns the provided `backend`, invisibly.
+#' @returns Called primarily for side effects. Returns the provided `backend`,
+#'   invisibly.
 #' @export
 use_backend <- function(backend, gpu = NA) {
 
@@ -245,9 +251,11 @@ use_backend <- function(backend, gpu = NA) {
       if (gpu) {
         py_require(c("tensorflow-cpu", "torch", "torchvision", "torchaudio"))
       } else {
-        Sys.setenv("UV_INDEX" = "https://download.pytorch.org/whl/cpu")
+        Sys.setenv("UV_INDEX" = trimws(paste(sep = " ",
+          "https://download.pytorch.org/whl/cpu",
+          Sys.getenv("UV_INDEX")
+        )))
         py_require(c("tensorflow-cpu", "torch", "torchvision", "torchaudio"))
-                   # additional_args = c("--index", "https://download.pytorch.org/whl/cpu"))
       }
     },
 
@@ -268,12 +276,14 @@ use_backend <- function(backend, gpu = NA) {
 
     Windows_torch = {
       if (is.na(gpu))
-        gpu <- has_gpu()
+        gpu <- FALSE
 
       if (gpu) {
-        Sys.setenv("UV_INDEX" = "https://download.pytorch.org/whl/cu126")
+        Sys.setenv("UV_INDEX" = trimws(paste(sep = " ",
+            "https://download.pytorch.org/whl/cu126",
+            Sys.getenv("UV_INDEX")
+        )))
         py_require(c("torch", "torchvision", "torchaudio"))
-                   # additional_args = c("--index", "https://download.pytorch.org/whl/cu126"))
       } else {
         py_require(c("torch", "torchvision", "torchaudio"))
       }
@@ -289,6 +299,7 @@ use_backend <- function(backend, gpu = NA) {
 
 
 
+
 get_os <- function() {
   if (is_windows()) "Windows" else if (is_mac_arm64()) "macOS" else "Linux"
 }

R/losses.R

@@ -1962,7 +1962,8 @@ py_to_r_wrapper.keras.src.losses.loss.Loss <- function(x) {
   as.function.default(c(formals(x), quote({
     args <- capture_args(list(y_true = as_py_array,
                               y_pred = as_py_array,
-                              sample_weight = as_py_array))
+                              sample_weight = as_py_array),
+                         enforce_all_dots_named = FALSE)
     do.call(x, args)
   })))
 }

R/metrics.R

@@ -2978,6 +2978,13 @@ function (y_true, y_pred, from_logits = FALSE, ignore_class = NULL,
 #' ```
 #'
 #' ```{r}
+#' # calling a metric directly is equivalent to calling
+#' # m$update_state(); m$result()
+#' m <- metric_mean()
+#' m(c(1, 3, 5, 7))
+#' ```
+#'
+#' ```{r}
 #' m$reset_state()
 #' m$update_state(c(1, 3, 5, 7), sample_weight = c(1, 1, 0, 0))
 #' m$result()

R/package.R

@@ -72,6 +72,7 @@ keras <- NULL
   # the tensorflow R package calls `py_require()` to ensure GPU is usable on Linux
   # use_backend() includes py_require(action = "remove") calls to undo
   # what tensorflow:::.onLoad() did. Keep them in sync!
+  # backend <- Sys.getenv("KERAS_BACKEND", "jax")
   backend <- Sys.getenv("KERAS_BACKEND", "tensorflow")
   gpu <- NA
   if (endsWith(backend, "-cpu")) {

cran-comments.md

@@ -2,7 +2,8 @@
 
 0 errors | 0 warnings | 1 note
 
-* installed size is 12.4Mb
+* checking installed package size ... NOTE
+  installed size is 14.3Mb
   sub-directories of 1Mb or more:
     doc    3.3Mb
-    help   8.4Mb
+    help  10.1Mb