Module dfdx::feature_flags
source · [−]Expand description
Information about the available feature flags.
Default features:
- “std”
- “numpy”
“std”
Enabled by default
Enables usage of the standard library. Otherwise no_std_compat is used.
Example:
dfdx = { version = "...", default-features = false }
Note that allocations are necessary, so the no_std_compat dependency looks like:
no-std-compat = { version = "0.4.1", features = [ "alloc", "compat_hash" ] }
“intel-mkl”
Enables using the Intel MKL
libraries (assuming you installed it already) for matrix multiplication.
Linking is currently tested & verified on the following platforms:
- Windows
- Linux
- macOS
Example:
dfdx = { version = "...", features = ["intel-mkl"] }
Installing Intel MKL libraries
It’s pretty easy!
You will need to install Intel MKL on your own from this page.
build.rs
will fail helpfully if you don’t have the correct path/environment variables.
“numpy”
Enabled by default
Enables saving and loading arrays to .npy files, and saving and loading nn to .npz files.
Example:
dfdx = { version = "...", features = ["numpy"] }
“nightly”
Enables using all features that currently require the nightly rust compiler.
Example:
dfdx = { version = "...", features = ["nightly"] }