Changelog
Source:NEWS.md
keras3 (development version)
User facing changes with upstream Keras v3.3.2:
new function:
op_ctc_decode()
new function:
op_eigh()
new function:
op_select()
new function:
op_vectorize()
new function:
op_image_rgb_to_grayscale()
new function:
loss_tversky()
new args:
layer_resizing(pad_to_aspect_ratio, fill_mode, fill_value)
new arg:
layer_embedding(weights)
for providing an initial weights matrixnew args:
op_nan_to_num(nan, posinf, neginf)
new args:
op_image_resize(crop_to_aspect_ratio, pad_to_aspect_ratio, fill_mode, fill_value)
new args:
op_argmax(keepdims)
andop_argmin(keepdims)
new arg:
clear_session(free_memory)
for clearing without invoking the garbage collector.metric_kl_divergence()
andloss_kl_divergence()
clip inputs (y_true
andy_pred
) to the[0, 1]
range.new
Layer()
attributes:metrics
,dtype_policy
Added initial support for float8 training
layer_conv_*d()
layers now support LoRaop_digitize()
now supports sparse tensors.Models and layers now return owned metrics recursively.
Add pickling support for Keras models. (e.g., via
reticulate::py_save_object()
) Note that pickling is not recommended, prefer using Keras saving APIs.
keras3 0.2.0
CRAN release: 2024-04-18
New functions:
quantize_weights()
: quantize model or layer weights in-place. Currently, onlyDense
,EinsumDense
, andEmbedding
layers are supported (which is enough to cover the majority of transformers today)config_set_backend()
: change the backend after Keras has initialized.-
New Ops
-
New family of linear algebra ops
audio_dataset_from_directory()
,image_dataset_from_directory()
andtext_dataset_from_directory()
gain averbose
argument (defaultTRUE
)image_dataset_from_directory()
gainspad_to_aspect_ratio
argument (defaultFALSE
)to_categorical()
,op_one_hot()
, andfit()
can now accept R factors, offset them to be 0-based (reported in#1055
).op_convert_to_numpy()
now returns unconverted NumPy arrays.op_array()
andop_convert_to_tensor()
no longer error when casting R doubles to integer types.export_savedmodel()
now works with a Jax backend.Metric()$add_variable()
method gains arg:aggregration
.Layer()$add_weight()
method gains args:autocast
,regularizer
,aggregation
.op_bincount()
,op_multi_hot()
,op_one_hot()
, andlayer_category_encoding()
now support sparse tensors.op_custom_gradient()
now supports the PyTorch backendlayer_lstm()
andlayer_gru()
gain arguse_cudnn
, default'auto'
.Fixed an issue where
application_preprocess_inputs()
would error if supplied an R array as input.Doc improvements.
keras3 0.1.0
CRAN release: 2024-02-17
- The package has been rebuilt for Keras 3.0. Refer to
for an overview and https://keras.posit.co for the current up-to-date documentation.