Search is not available for this dataset
text stringlengths 75 104k |
|---|
def self_attention_expert(x,
batch_coordinate,
mask_right=True,
split_batch=False,
attention_num_head=1,
attention_kq_size=None,
attention_v_size=None):
"""Implem... |
def local_expert_attention(x,
k,
loss_coef,
attention_num_experts,
train=True,
batch_coordinate=None,
**kwargs):
"""Attention using a mixture of experts.
... |
def expert_dot_product(q, k, v, info_q, info_k):
"""Perform dot product on a subset of the sequence.
Can add a mask to the attention to prevent sequences to attend to each other
and to prevent attention to the future.
Args:
q (tf.Tensor): Queries of shape [length_expert_q, depth_k]
k (tf.Tensor): Keys... |
def dot_product_single_head(q, k, v, gates_q, gates_k, bi):
"""Perform a dot product attention on a single sequence on a single head.
This function dispatch the q, k, v and loop over the buckets to compute the
attention dot product on each subsequences.
Args:
q (tf.Tensor): [length_q, depth_q]
k (tf.T... |
def map_fn_switch(fn, elems, use_map_fn=True, **kwargs):
"""Construct the graph with either tf.map_fn or a python for loop.
This function is mainly for for benchmarking purpose.
tf.map_fn is dynamic but is much slower than creating a static graph with
for loop. However, having a for loop make the graph much l... |
def sparse_dot_product_attention(q, k, v, bi, use_map_fn, experts_params):
"""Sparse multihead self attention.
Perform an approximation of the full multihead attention by dispatching
the tokens using their keys/values. Thus the attention matrix are only
computed each times on a subset of the tokens.
Notes:
... |
def dot_product_batched_head(q, k, v, gates_q, gates_k, mask_right=False):
"""Perform a dot product attention on a single sequence on a single head.
This function dispatch the q, k, v and loop over the buckets to compute the
attention dot product on each subsequences.
Args:
q (tf.Tensor): [batch*heads, le... |
def sparse_dot_product_attention_truncated(
q,
k,
v,
bi, # Unused
experts_params,
use_map_fn=False, # Unused
mask_right=False,
): # pylint: disable=unused-argument
"""Sparse multihead self attention.
Perform an approximation of the full multihead attention by dispatching
the tokens... |
def deconv_elems_1d(x, factor, out_depth=None):
"""Increase the length and change the dimensionality.
Expand/project each positions of dim depth of the input into
factor*tokens of dim out_depth
Args:
x (tf.Tensor): shape [batch_size, length, depth]
factor (int): Multiplicative factor of each tokens.
... |
def conv_elems_1d(x, factor, out_depth=None):
"""Decrease the length and change the dimensionality.
Merge/restore/compress factors positions of dim depth of the input into
a single position of dim out_depth.
This is basically just a strided convolution without overlap
between each strides. The original lengt... |
def local_reduction_attention(x, block_length, multihead_params):
"""Reduce the length dimension using self attention.
Args:
x (tf.Tensor): float32 of shape [batch, length, depth]
block_length (int): Block length for local attention (Compression factor)
multihead_params (dict): parameters for multihead... |
def multihead_self_attention_reduced(
x,
memory_antecedent=None,
bias=None,
factor=None,
multihead_params=None,
nonlinearity="none",
reduction_type="conv",
add_mask=True,
):
"""Reduce the length dimension by compressing with conv.
Args:
x (tf.Tensor): float32 of shape [batch, le... |
def scaled_dot_product_attention_simple(q, k, v, bias, name=None):
"""Scaled dot-product attention. One head. One spatial dimension.
Args:
q: a Tensor with shape [batch, length_q, depth_k]
k: a Tensor with shape [batch, length_kv, depth_k]
v: a Tensor with shape [batch, length_kv, depth_v]
bias: op... |
def multihead_self_attention_memory_efficient(x,
bias,
num_heads,
head_size=None,
epsilon=1e-6,
... |
def _idx_to_bits(self, i):
"""Convert an group index to its bit representation."""
bits = bin(i)[2:].zfill(self.nb_hyperplanes) # Pad the bits str with 0
return [-1.0 if b == "0" else 1.0 for b in bits] |
def get_gates(self, x):
"""Return the bucket id of the given tensor.
Args:
x (tf.Tensor): float32 of shape [length, depth]
Returns:
tf.Tensor: One-hot vector int64 of shape [heads, length, nb_buckets]
containing the id of the bucket
"""
# The balance loss don't propagate to th... |
def van_image_enc_2d(x, first_depth, reuse=False, hparams=None):
"""The image encoder for the VAN.
Similar architecture as Ruben's paper
(http://proceedings.mlr.press/v70/villegas17a/villegas17a.pdf).
Args:
x: The image to encode.
first_depth: The depth of the first layer. Depth is increased in subseq... |
def van_enc_2d(x, first_depth, reuse=False):
"""The higher level structure encoder for the VAN.
The high level structure is a vector instead of an image.
Args:
x: The higher level structure to encode.
first_depth: The depth of the first layer. Depth is increased in subsequent
layers.
reuse: To... |
def van_dec_2d(x, skip_connections, output_shape, first_depth, hparams=None):
"""The VAN decoder.
Args:
x: The analogy information to decode.
skip_connections: The encoder layers which can be used as skip connections.
output_shape: The shape of the desired output image.
first_depth: The depth of th... |
def analogy_computation_2d(f_first_enc,
f_first_frame,
f_current_enc,
first_depth):
"""Implements the deep analogy computation."""
with tf.variable_scope('analogy_computation'):
frame_enc_diff = f_first_frame - f_first_enc
fr... |
def van(first_enc,
first_frame,
current_enc,
gt_image,
reuse=False,
scope_prefix='',
hparams=None):
"""Implements a VAN.
Args:
first_enc: The first encoding.
first_frame: The first ground truth frame.
current_enc: The encoding of the frame to generate.
... |
def encoder_vgg(x, enc_final_size, reuse=False, scope_prefix='', hparams=None,
is_training=True):
"""VGG network to use as encoder without the top few layers.
Can be pretrained.
Args:
x: The image to encode. In the range 0 to 1.
enc_final_size: The desired size of the encoding.
reuse... |
def predictor(enc_flat,
action,
lstm_states,
pred_depth,
reuse=False,
scope_prefix='',
hparams=None):
"""LSTM predictor network."""
with tf.variable_scope(scope_prefix + 'predict', reuse=reuse):
enc_final_size = enc_flat.get_sh... |
def construct_model(images,
actions=None,
context_frames=2,
hparams=None,
is_training=True):
"""Constructs the tensorflow graph of the hierarchical model."""
pred_depth = 20
enc_out_all, pred_out_all, van_out_all, van_on_enc_all = [... |
def peak_signal_to_noise_ratio(true, pred):
"""Image quality metric based on maximal signal power vs. power of the noise.
Args:
true: the ground truth image.
pred: the predicted image.
Returns:
peak signal to noise ratio (PSNR)
"""
return 10.0 * tf.log(1.0 / mean_squared_error(true, pred)) / tf.l... |
def mean_squared_error(true, pred):
"""L2 distance between tensors true and pred.
Args:
true: the ground truth image.
pred: the predicted image.
Returns:
mean squared error between ground truth and predicted image.
"""
result = tf.reduce_sum(
tf.squared_difference(true, pred)) / tf.to_float... |
def l1_error(true, pred):
"""L1 distance between tensors true and pred."""
return tf.reduce_sum(tf.abs(true - pred)) / tf.to_float(tf.size(pred)) |
def calc_loss_psnr(gen_images, images, name, hparams=None, use_l1_loss=False):
"""Calculates loss and psnr for predictions over multiple timesteps."""
del hparams
with tf.name_scope(name):
loss, error, psnr_all = 0.0, 0.0, 0.0
for _, x, gx in zip(range(len(gen_images)), images, gen_images):
recon_co... |
def next_frame_sv2p():
"""SV2P model hparams."""
hparams = basic_stochastic.next_frame_basic_stochastic()
hparams.optimizer = "true_adam"
hparams.learning_rate_schedule = "constant"
hparams.learning_rate_constant = 1e-3
hparams.video_num_input_frames = 1
hparams.video_num_target_frames = 3
hparams.batch... |
def next_frame_sv2p_discrete():
"""SV2P discrete model hparams."""
hparams = next_frame_sv2p()
hparams.action_injection = "multiplicative"
hparams.small_mode = True
hparams.add_hparam("bottleneck_bits", 128)
hparams.add_hparam("bottleneck_noise", 0.02)
hparams.add_hparam("discrete_warmup_steps", 40000)
... |
def next_frame_sv2p_atari():
"""SV2P model for atari."""
hparams = next_frame_sv2p()
hparams.video_num_input_frames = 4
hparams.video_num_target_frames = 4
hparams.action_injection = "multiplicative"
hparams.num_iterations_1st_stage = 12000
hparams.num_iterations_2nd_stage = 12000
hparams.anneal_end = 4... |
def next_frame_sv2p_atari_softmax():
"""SV2P model for atari with softmax."""
hparams = next_frame_sv2p_atari()
hparams.bottom = {}
hparams.loss = {}
hparams.top = {}
hparams.internal_loss = True
return hparams |
def next_frame_sv2p_tiny():
"""Tiny SV2P model."""
hparams = next_frame_sv2p_atari_softmax()
hparams.batch_size = 2
hparams.tiny_mode = True
hparams.num_masks = 1
hparams.video_modality_loss_cutoff = 0.4
hparams.video_num_input_frames = 4
hparams.video_num_target_frames = 4
return hparams |
def next_frame_sv2p_cutoff():
"""SV2P model with additional cutoff in L2 loss for environments like pong."""
hparams = next_frame_sv2p()
hparams.video_modality_loss_cutoff = 0.4
hparams.video_num_input_frames = 4
hparams.video_num_target_frames = 1
return hparams |
def _get_mscoco(directory):
"""Download and extract MSCOCO datasets to directory unless it is there."""
for url in _MSCOCO_URLS:
filename = os.path.basename(url)
download_url = os.path.join(_MSCOCO_ROOT_URL, url)
path = generator_utils.maybe_download(directory, filename, download_url)
unzip_dir = os... |
def mscoco_generator(data_dir,
tmp_dir,
training,
how_many,
start_from=0,
eos_list=None,
vocab_filename=None):
"""Image generator for MSCOCO captioning problem with token-wise captions.
Arg... |
def flags_as_args():
"""Convert FLAGS to list of args suitable for passing on cmd line."""
if hasattr(FLAGS, "flag_values_dict"):
args_dict = FLAGS.flag_values_dict()
else:
args_dict = dict(FLAGS.__dict__["__flags"])
del args_dict["cloud_mlengine"]
# Configured later
del args_dict["t2t_usr_dir"]
a... |
def get_default_master_type(num_gpus=1):
"""Returns master_type for trainingInput."""
gpus_to_master_map = {
0: "standard",
1: "standard_p100",
4: "complex_model_m_p100",
8: "complex_model_l_gpu",
}
if num_gpus not in gpus_to_master_map:
raise ValueError("Num gpus must be in %s" %
... |
def configure_job():
"""Construct jobSpec for ML Engine job."""
# See documentation:
# https://cloud.google.com/ml-engine/reference/rest/v1/projects.jobs#traininginput
training_input = {
"pythonModule": "tensor2tensor.bin.t2t_trainer",
"args": flags_as_args(),
"region": text_encoder.native_to_... |
def launch_job(job_spec):
"""Launch job on ML Engine."""
project_id = "projects/{}".format(
text_encoder.native_to_unicode(default_project()))
credentials = GoogleCredentials.get_application_default()
cloudml = discovery.build("ml", "v1", credentials=credentials,
cache_discover... |
def _tar_and_copy(src_dir, target_dir):
"""Tar and gzip src_dir and copy to GCS target_dir."""
src_dir = src_dir.rstrip("/")
target_dir = target_dir.rstrip("/")
tmp_dir = tempfile.gettempdir().rstrip("/")
src_base = os.path.basename(src_dir)
shell_run(
"tar --exclude=.git -zcf {tmp_dir}/{src_base}.tar... |
def tar_and_copy_t2t(train_dir):
"""Tar Tensor2Tensor and cp to train_dir."""
tf.logging.info("Tarring and pushing local Tensor2Tensor package.")
output = text_encoder.native_to_unicode(shell_output(
"pip show tensor2tensor")).split("\n")
assert output[1].startswith("Version")
assert output[7].startswi... |
def tar_and_copy_usr_dir(usr_dir, train_dir):
"""Package, tar, and copy usr_dir to GCS train_dir."""
tf.logging.info("Tarring and pushing t2t_usr_dir.")
usr_dir = os.path.abspath(os.path.expanduser(usr_dir))
# Copy usr dir to a temp location
top_dir = os.path.join(tempfile.gettempdir(), "t2t_usr_container")
... |
def validate_flags():
"""Validates flags are set to acceptable values for CloudML Engine runs."""
assert not job_dir()
assert FLAGS.output_dir.startswith("gs://")
assert FLAGS.data_dir.startswith("gs://")
assert FLAGS.worker_replicas <= 1
assert FLAGS.ps_replicas <= 0
if FLAGS.hparams_range:
assert FL... |
def launch():
"""Launch t2t_trainer on Cloud ML Engine."""
validate_flags()
job_spec = configure_job()
job_name = job_spec["jobId"]
tf.logging.info("Launching job %s with ML Engine spec:\n%s", job_name,
pprint.pformat(job_spec))
assert confirm()
train_dir = FLAGS.output_dir
t2t_tar = t... |
def add_weight(cls):
"""Decorator for Layers, overriding add_weight for trainable initializers."""
@functools.wraps(cls.add_weight)
def _add_weight(self,
name=None,
shape=None,
dtype=None,
initializer=None,
regularizer=None,... |
def get_beta(self, kl_loss=0.0):
"""Get the KL multiplier, either dynamically or schedule based.
if hparams.latent_loss_multiplier_dynamic is set to true, then beta
is being adjusted to keep KL under hparams.latent_loss_multiplier_epsilon.
In order to do so, the beta is being updated at each iteration
... |
def get_kl_loss(self, means, log_vars, means_p=None, log_vars_p=None):
"""Get KL loss for all the predicted Gaussians."""
kl_loss = 0.0
if means_p is None:
means_p = tf.unstack(tf.zeros_like(means))
if log_vars_p is None:
log_vars_p = tf.unstack(tf.zeros_like(log_vars))
enumerated_inputs... |
def construct_latent_tower(self, images, time_axis):
"""Create the latent tower."""
# No latent in the first phase
first_phase = tf.less(
self.get_iteration_num(), self.hparams.num_iterations_1st_stage)
# use all frames by default but this allows more
# predicted frames at inference time
... |
def transformer_encode(encoder_function, inputs, target_space, hparams,
attention_weights=None, features=None, losses=None,
**kwargs):
"""Encode transformer inputs.
Args:
encoder_function: the encoder function
inputs: Transformer inputs [batch_size, input_lengt... |
def transformer_decode(decoder_function,
decoder_input,
encoder_output,
encoder_decoder_attention_bias,
decoder_self_attention_bias,
hparams,
attention_weights=None,
... |
def _init_transformer_cache(cache, hparams, batch_size, attention_init_length,
encoder_output, encoder_decoder_attention_bias,
scope_prefix):
"""Create the initial cache for Transformer fast decoding."""
key_channels = hparams.attention_key_channels or hparams... |
def fast_decode_tpu(encoder_output,
encoder_decoder_attention_bias,
symbols_to_logits_fn,
hparams,
decode_length,
vocab_size,
init_cache_fn=_init_transformer_cache,
beam_size=1,
... |
def fast_decode(encoder_output,
encoder_decoder_attention_bias,
symbols_to_logits_fn,
hparams,
decode_length,
vocab_size,
init_cache_fn=_init_transformer_cache,
beam_size=1,
top_beams=1,
... |
def transformer_prepare_decoder(targets, hparams, features=None):
"""Prepare one shard of the model for the decoder.
Args:
targets: a Tensor.
hparams: run hyperparameters
features: optionally pass the entire features dictionary as well. This is
needed now for "packed" datasets.
Returns:
de... |
def transformer_decoder(decoder_input,
encoder_output,
decoder_self_attention_bias,
encoder_decoder_attention_bias,
hparams,
cache=None,
decode_loop_step=None,
... |
def transformer_base_v1():
"""Set of hyperparameters."""
hparams = common_hparams.basic_params1()
hparams.norm_type = "layer"
hparams.hidden_size = 512
hparams.batch_size = 4096
hparams.max_length = 256
hparams.clip_grad_norm = 0. # i.e. no gradient clipping
hparams.optimizer_adam_epsilon = 1e-9
hpar... |
def transformer_base_v2():
"""Set of hyperparameters."""
hparams = transformer_base_v1()
hparams.layer_preprocess_sequence = "n"
hparams.layer_postprocess_sequence = "da"
hparams.layer_prepostprocess_dropout = 0.1
hparams.attention_dropout = 0.1
hparams.relu_dropout = 0.1
hparams.learning_rate_warmup_st... |
def transformer_base_vq_ada_32ex_packed():
"""Set of hyperparameters for lm1b packed following tpu params."""
hparams = transformer_base_v2()
expert_utils.update_hparams_for_vq_gating(hparams)
hparams.moe_num_experts = 32
hparams.gating_type = "vq"
# this gives us a batch size of 16 because each seq is len ... |
def transformer_base_vq1_16_nb1_packed_nda_b01_scales():
"""Set of hyperparameters."""
hparams = transformer_base_vq_ada_32ex_packed()
hparams.use_scales = int(True)
hparams.moe_num_experts = 16
hparams.moe_k = 1
hparams.beta = 0.1
hparams.layer_preprocess_sequence = "n"
hparams.layer_postprocess_sequen... |
def transformer_base_vq1_16_nb1_packed_dan_b01_scales():
"""Set of hyperparameters."""
hparams = transformer_base_vq_ada_32ex_packed()
hparams.use_scales = int(True)
hparams.moe_num_experts = 16
hparams.moe_k = 1
hparams.beta = 0.1
hparams.ema = False
return hparams |
def transformer_base_vq1_16_nb1_packed_nda_b01_scales_dialog():
"""Set of hyperparameters."""
hparams = transformer_base_vq1_16_nb1_packed_nda_b01_scales()
hparams.batch_size = 2048
hparams.max_length = 1024
hparams.filter_size = 3072
return hparams |
def transformer_ada_lmpackedbase_dialog():
"""Set of hyperparameters."""
hparams = transformer_base_vq_ada_32ex_packed()
hparams.max_length = 1024
hparams.ffn_layer = "dense_relu_dense"
hparams.batch_size = 4096
return hparams |
def transformer_base_v3():
"""Base parameters for Transformer model."""
# Update parameters here, then occasionally cut a versioned set, e.g.
# transformer_base_v2.
hparams = transformer_base_v2()
hparams.optimizer_adam_beta2 = 0.997
# New way of specifying learning rate schedule.
# Equivalent to previous... |
def transformer_big():
"""HParams for transformer big model on WMT."""
hparams = transformer_base()
hparams.hidden_size = 1024
hparams.filter_size = 4096
# Reduce batch size to 2048 from 4096 to be able to train the model on a GPU
# with 12 GB memory. For example, NVIDIA TITAN V GPU.
hparams.batch_size = ... |
def transformer_tall():
"""Hparams for transformer on LM for pretraining/finetuning/mixing."""
hparams = transformer_base()
hparams.batch_size = 2048
hparams.hidden_size = 768
hparams.filter_size = 3072
hparams.num_hidden_layers = 12
hparams.num_heads = 12
hparams.label_smoothing = 0.0
hparams.max_len... |
def transformer_tall_finetune_tied():
"""Tied means fine-tune CNN/DM summarization as LM."""
hparams = transformer_tall()
hparams.multiproblem_max_input_length = 750
hparams.multiproblem_max_target_length = 100
hparams.multiproblem_schedule_max_examples = 0
hparams.learning_rate_schedule = ("linear_warmup*c... |
def transformer_tall_finetune_uniencdec():
"""Fine-tune CNN/DM with a unidirectional encoder and decoder."""
hparams = transformer_tall()
hparams.max_input_seq_length = 750
hparams.max_target_seq_length = 100
hparams.optimizer = "true_adam"
hparams.learning_rate_schedule = ("linear_warmup*constant*cosdecay"... |
def transformer_tall_train_uniencdec():
"""Train CNN/DM with a unidirectional encoder and decoder."""
hparams = transformer_tall()
hparams.max_input_seq_length = 750
hparams.max_target_seq_length = 100
hparams.optimizer = "true_adam"
hparams.learning_rate_schedule = ("linear_warmup*constant*cosdecay")
hpa... |
def transformer_tall_finetune_textclass():
"""Hparams for transformer on LM for finetuning on text class problems."""
hparams = transformer_tall()
hparams.learning_rate_constant = 6.25e-5
hparams.learning_rate_schedule = ("linear_warmup*constant*linear_decay")
hparams.multiproblem_schedule_max_examples = 0
... |
def transformer_tall_pretrain_lm():
"""Hparams for transformer on LM pretraining (with 64k vocab)."""
hparams = transformer_tall()
hparams.learning_rate_constant = 2e-4
hparams.learning_rate_schedule = ("linear_warmup*constant*cosdecay")
hparams.optimizer = "adam_w"
hparams.optimizer_adam_beta1 = 0.9
hpar... |
def transformer_tall_pretrain_lm_tpu_adafactor():
"""Hparams for transformer on LM pretraining (with 64k vocab) on TPU."""
hparams = transformer_tall_pretrain_lm()
update_hparams_for_tpu(hparams)
hparams.max_length = 1024
# For multi-problem on TPU we need it in absolute examples.
hparams.batch_size = 8
h... |
def transformer_tall_pretrain_lm_tpu_adafactor_large():
"""Hparams for transformer on LM pretraining on TPU, large model."""
hparams = transformer_tall_pretrain_lm_tpu_adafactor()
hparams.hidden_size = 1024
hparams.num_heads = 16
hparams.filter_size = 32768 # max fitting in 16G memory is 49152, batch 2
hpa... |
def transformer_tall_pretrain_lm_tpu():
"""Hparams for transformer on LM pretraining on TPU with AdamW."""
hparams = transformer_tall_pretrain_lm_tpu_adafactor()
# Optimizer gets reset in update_hparams_for_tpu so we set it again here.
hparams.learning_rate_constant = 2e-4
hparams.learning_rate_schedule = ("l... |
def transformer_base_single_gpu():
"""HParams for transformer base model for single GPU."""
hparams = transformer_base()
hparams.batch_size = 1024
hparams.learning_rate_schedule = "constant*linear_warmup*rsqrt_decay"
hparams.learning_rate_constant = 0.1
hparams.learning_rate_warmup_steps = 16000
return hp... |
def transformer_parsing_base():
"""HParams for parsing on WSJ only."""
hparams = transformer_base()
hparams.attention_dropout = 0.2
hparams.layer_prepostprocess_dropout = 0.2
hparams.max_length = 512
hparams.learning_rate_warmup_steps = 16000
hparams.hidden_size = 1024
hparams.learning_rate = 0.05
hpa... |
def transformer_parsing_big():
"""HParams for parsing on WSJ semi-supervised."""
hparams = transformer_big()
hparams.max_length = 512
hparams.shared_source_target_embedding = False
hparams.learning_rate_warmup_steps = 4000
hparams.layer_prepostprocess_dropout = 0.1
hparams.batch_size = 2048
hparams.lear... |
def transformer_base_range(rhp):
"""Small range of hyperparameters."""
# After starting from base, set intervals for some parameters.
rhp.set_float("learning_rate", 0.3, 3.0, scale=rhp.LOG_SCALE)
rhp.set_discrete("learning_rate_warmup_steps",
[1000, 2000, 4000, 8000, 16000])
rhp.set_float("... |
def transformer_relative():
"""Use relative position embeddings instead of absolute position encodings."""
hparams = transformer_base()
hparams.pos = None
hparams.self_attention_type = "dot_product_relative"
hparams.max_relative_position = 20
return hparams |
def transformer_mlperf_tpu():
"""HParams for Transformer model on TPU for MLPerf on TPU 2x2."""
hparams = transformer_base_v3()
hparams.mlperf_mode = True
hparams.symbol_modality_num_shards = 1
hparams.max_length = 256 # ignored when using "_packed" problems
hparams.batch_size = 2048 # per-chip batch size... |
def update_hparams_for_tpu(hparams):
"""Change hparams to be compatible with TPU training."""
# Adafactor uses less memory than Adam.
# switch to Adafactor with its recommended learning rate scheme.
hparams.optimizer = "Adafactor"
hparams.learning_rate_schedule = "rsqrt_decay"
hparams.learning_rate_warmup_... |
def transformer_tpu_range(rhp):
"""Small range of hyperparameters."""
# After starting from base, set intervals for some parameters.
rhp.set_float("learning_rate", 0.3, 3.0, scale=rhp.LOG_SCALE)
rhp.set_discrete("learning_rate_warmup_steps",
[1000, 2000, 4000, 8000, 16000])
rhp.set_float("i... |
def transformer_clean():
"""No dropout, label smoothing, max_length."""
hparams = transformer_base_v2()
hparams.label_smoothing = 0.0
hparams.layer_prepostprocess_dropout = 0.0
hparams.attention_dropout = 0.0
hparams.relu_dropout = 0.0
hparams.max_length = 0
return hparams |
def transformer_lm_tpu_0():
"""HParams for training languagemodel_lm1b8k on tpu. 92M Params."""
hparams = transformer_clean_big()
update_hparams_for_tpu(hparams)
hparams.num_heads = 4 # Heads are expensive on TPUs.
hparams.batch_size = 4096
hparams.shared_embedding_and_softmax_weights = False
hparams.la... |
def transformer_librispeech_v1():
"""HParams for training ASR model on LibriSpeech V1."""
hparams = transformer_base()
hparams.num_heads = 4
hparams.filter_size = 1024
hparams.hidden_size = 256
hparams.num_encoder_layers = 5
hparams.num_decoder_layers = 3
hparams.learning_rate = 0.15
hparams.batch_si... |
def transformer_librispeech_v2():
"""HParams for training ASR model on LibriSpeech V2."""
hparams = transformer_base()
hparams.max_length = 1240000
hparams.max_input_seq_length = 1550
hparams.max_target_seq_length = 350
hparams.batch_size = 16
hparams.num_decoder_layers = 4
hparams.num_encoder_layers =... |
def transformer_librispeech_tpu_v1():
"""HParams for training ASR model on Librispeech on TPU v1."""
hparams = transformer_librispeech_v1()
update_hparams_for_tpu(hparams)
hparams.batch_size = 16
librispeech.set_librispeech_length_hparams(hparams)
return hparams |
def transformer_librispeech_tpu_v2():
"""HParams for training ASR model on Librispeech on TPU v2."""
hparams = transformer_librispeech_v2()
update_hparams_for_tpu(hparams)
hparams.batch_size = 16
librispeech.set_librispeech_length_hparams(hparams)
return hparams |
def transformer_tpu_1b():
"""Hparams for machine translation with ~1.1B parameters."""
hparams = transformer_tpu()
hparams.hidden_size = 2048
hparams.filter_size = 8192
hparams.num_hidden_layers = 8
# smaller batch size to avoid OOM
hparams.batch_size = 1024
hparams.activation_dtype = "bfloat16"
hpara... |
def transformer_wikitext103_l4k_v0():
"""HParams for training languagemodel_wikitext103_l4k."""
hparams = transformer_big()
# Adafactor uses less memory than Adam.
# switch to Adafactor with its recommended learning rate scheme.
hparams.optimizer = "Adafactor"
hparams.learning_rate_schedule = "rsqrt_decay"... |
def transformer_wikitext103_l4k_memory_v0():
"""HParams for training languagemodel_wikitext103_l4k with memory."""
hparams = transformer_wikitext103_l4k_v0()
hparams.split_targets_chunk_length = 64
hparams.split_targets_max_chunks = 64
hparams.split_targets_strided_training = True
hparams.add_hparam("memor... |
def transformer_wikitext103_l16k_memory_v0():
"""HParams for training languagemodel_wikitext103_l16k with memory."""
hparams = transformer_wikitext103_l4k_memory_v0()
hparams.max_length = 16384
hparams.split_targets_chunk_length = 64
hparams.split_targets_max_chunks = int(
hparams.max_length / hparams.... |
def transformer_cifar10_memory_v0():
"""HParams for training image_cifar10_plain_gen_flat_rev with memory."""
hparams = transformer_wikitext103_l4k_memory_v0()
hparams.num_hidden_layers = 6
hparams.max_length = 32 * 32 * 3
hparams.split_targets_chunk_length = 64 * 3
hparams.split_targets_max_chunks = int(... |
def transformer_imagenet64_memory_v0():
"""HParams for training image_imagenet64_gen_flat_rev with memory."""
hparams = transformer_cifar10_memory_v0()
hparams.max_length = 64 * 64 * 3
hparams.split_targets_chunk_length = 64 * 3
hparams.split_targets_max_chunks = int(
hparams.max_length / hparams.split... |
def maybe_reshape_4d_to_3d(x):
"""Reshape input from 4D to 3D if necessary."""
x_shape = common_layers.shape_list(x)
is_4d = False
if len(x_shape) == 4:
x = tf.reshape(x, [x_shape[0], x_shape[1]*x_shape[2], x_shape[3]])
is_4d = True
return x, x_shape, is_4d |
def local_attention_2d(x, hparams, attention_type="local_attention_2d"):
"""Local 2d, self attention layer."""
# self-attention
with tf.variable_scope("local_2d_self_att"):
y = common_attention.multihead_attention_2d(
x,
None,
hparams.attention_key_channels or hparams.hidden_size,
... |
def local_within_block_attention(x,
self_attention_bias,
hparams,
attention_type="local_within_block_mask_right",
q_padding="VALID",
kv_padding="VALID"):
... |
def local_attention_1d(x,
hparams,
attention_type="local_unmasked",
q_padding="VALID",
kv_padding="VALID"):
"""Local 1d self attention."""
# self-attention
x, x_shape, is_4d = maybe_reshape_4d_to_3d(x)
with tf.variable_s... |
def get_dilated_1d_attention_mask(
num_heads, block_size,
num_blocks, memory_size, gap_size,
name="dilated_mask"):
"""Dilated attention with a masking strategy."""
mask = np.ones((num_heads, block_size, 2*block_size), np.bool)
# now going over every row to do the right assignment of
# memory blocks... |
def dilated_attention_1d(x,
hparams,
attention_type="masked_dilated_1d",
q_padding="VALID",
kv_padding="VALID",
gap_size=2):
"""Dilated 1d self attention."""
# self-attention
x, x_shape, is... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.