def fake_dim(d):
if core.is_constant_dim(d):
return d
else:
if not isinstance(d, DimVar):
raise TypeError(
f"Encountered unexpected shape dimension {d}")
return 8
def make_tensorspec(a_jax):
a_tf_dtype = jax2tf_internal._to_tf_dtype(a_jax.dtype)
if any(not core.is_constant_dim(d) for d in a_jax.shape):
msg = ("call_tf cannot be applied to shape-polymorphic arguments. "
f"Found argument shape: {a_jax.shape}. "
"See https://github.com/google/jax/blob/main/jax/experimental/jax2tf/README.md#limitations-of-call-tf for a discussion.")
raise ValueError(msg)
return tf.TensorSpec(a_jax.shape, a_tf_dtype)
def threefry_random_bits(key: jnp.ndarray, bit_width, shape):
"""Sample uniform random bits of given width and shape using PRNG key."""
if not _is_threefry_prng_key(key):
raise TypeError("_random_bits got invalid prng key.")
if bit_width not in (8, 16, 32, 64):
raise TypeError("requires 8-, 16-, 32- or 64-bit field width.")
shape = core.as_named_shape(shape)
for name, size in shape.named_items:
real_size = lax.psum(1, name)
if real_size != size:
raise ValueError(
f"The shape of axis {name} was specified as {size}, "
f"but it really is {real_size}")
axis_index = lax.axis_index(name)
key = threefry_fold_in(key, axis_index)
size = prod(shape.positional)
# Compute ceil(bit_width * size / 32) in a way that is friendly to shape
# polymorphism
max_count, r = divmod(bit_width * size, 32)
if r > 0:
max_count += 1
if core.is_constant_dim(max_count):
nblocks, rem = divmod(max_count, jnp.iinfo(np.uint32).max)
else:
nblocks, rem = 0, max_count
if not nblocks:
bits = threefry_2x32(key, lax.iota(np.uint32, rem))
else:
keys = threefry_split(key, nblocks + 1)
subkeys, last_key = keys[:-1], keys[-1]
blocks = vmap(threefry_2x32,
in_axes=(0, None))(subkeys,
lax.iota(np.uint32,
jnp.iinfo(np.uint32).max))
last = threefry_2x32(last_key, lax.iota(np.uint32, rem))
bits = lax.concatenate([blocks.ravel(), last], 0)
dtype = UINT_DTYPES[bit_width]
if bit_width == 64:
bits = [lax.convert_element_type(x, dtype) for x in jnp.split(bits, 2)]
bits = lax.shift_left(bits[0], dtype(32)) | bits[1]
elif bit_width in [8, 16]:
# this is essentially bits.view(dtype)[:size]
bits = lax.bitwise_and(
np.uint32(np.iinfo(dtype).max),
lax.shift_right_logical(
lax.broadcast(bits, (1, )),
lax.mul(
np.uint32(bit_width),
lax.broadcasted_iota(np.uint32, (32 // bit_width, 1), 0))))
bits = lax.reshape(bits, (np.uint32(max_count * 32 // bit_width), ),
(1, 0))
bits = lax.convert_element_type(bits, dtype)[:size]
return lax.reshape(bits, shape)
def fake_dim(d):
if core.is_constant_dim(d):
return d
else:
if not isinstance(d, _DimPolynomial):
raise TypeError(f"Encountered unexpected shape dimension {d}")
# It is Ok to replace all polynomials with the same value. We may miss
# here some errors due to non-equal dimensions, but we catch them
# later.
return 8
