def get_metrics(config, metrics):
if not config.debug_run:
metrics = common_utils.get_metrics(metrics)
else:
metrics = common_utils.stack_forest(metrics)
metrics = jax.device_get(metrics)
return metrics
def maybe_eval_and_log(self, eval_summary, master, step, tick,
train_metrics, train_summary):
"""Maybe evaluate and log based on the current step value."""
if (step % self.eval_frequency == 0) or (step == self.total_steps):
del eval_summary
del train_summary
train_metrics = common_utils.get_metrics(train_metrics)
train_summary = pipeline_utils.compute_global_mean_metrics(
train_metrics)
tock = time.time()
steps_per_sec = self.eval_frequency / (tock - tick)
tick = tock
# log train summary
if master:
self.write_train_summary(step=step,
metric_dict=train_metrics,
summary=train_summary,
steps_per_sec=steps_per_sec)
# reset metric accumulation for next evaluation cycle
del train_metrics
train_metrics = []
# sync model state across replicas
self.train_state = pipeline_utils.sync_model_state_across_replicas(
self.train_state)
# evaluate and log the results
eval_summary, _ = self.eval(step, self.train_state)
return eval_summary, train_metrics, train_summary, tick
def test(optimizer, state, p_eval_step, step, test_ds, summary_writer):
"""Test the flax module in optimizer on test_ds.
Args:
optimizer: flax optimizer (contains flax module).
state: model state, e.g. batch statistics.
p_eval_step: fn; Pmapped evaluation step function.
step: int; Number of training steps passed so far.
test_ds: tf.dataset; Test dataset.
summary_writer: tensorflow summary writer.
"""
# Test Metrics
test_metrics = []
test_iter = iter(test_ds)
for _, test_batch in zip(itertools.repeat(1), test_iter):
# pylint: disable=protected-access
test_batch = common_utils.shard(
jax.tree_map(lambda x: x._numpy(), test_batch))
# pylint: enable=protected-access
metrics = p_eval_step(optimizer.target, state, test_batch)
test_metrics.append(metrics)
test_metrics = common_utils.get_metrics(test_metrics)
test_metrics_sums = jax.tree_map(jnp.sum, test_metrics)
test_denominator = test_metrics_sums.pop('denominator')
test_summary = jax.tree_map(
lambda x: x / test_denominator, # pylint: disable=cell-var-from-loop
test_metrics_sums)
logging.info('test in step: %d, loss: %.4f, acc: %.4f', step,
test_summary['loss'], test_summary['accuracy'])
if jax.host_id() == 0:
for key, val in test_summary.items():
summary_writer.scalar(f'test_{key}', val, step)
summary_writer.flush()
def combine_metrics(step_metrics):
"""Given a list of metric dicts, combine to a single summary metrics dict.
Args:
step_metrics: A dict with (metric name, metric value) items. Contains summed
metrics and the corresponding denominator (the number of next-token
prediction instances). Each metric value have at least one dimension.
Returns:
A dict with (metric name, metric value) items containing combined metrics.
"""
metrics_all = common_utils.get_metrics(step_metrics)
lr = None
if 'learning_rate' in metrics_all:
lr = metrics_all.pop('learning_rate').mean()
metrics_sums = jax.tree_map(jnp.sum, metrics_all)
denominator = metrics_sums.pop('denominator')
summary = jax.tree_map(lambda x: x / denominator, metrics_sums) # pylint: disable=cell-var-from-loop
if lr is not None:
summary['learning_rate'] = lr
# Calculate (clipped) perplexity after averaging log-perplexities:
if 'loss' in summary:
summary['perplexity'] = jnp.clip(jnp.exp(summary['loss']), a_max=1.0e4)
return summary
def predict_once(run_configuration, optimizer=None):
"""Predict the result once for each element in the dataset."""
adapter = run_configuration.adapter
checkpoint_path = run_configuration.original_checkpoint_path
optimizer = optimizer or adapter.create_optimizer(run_configuration)
dataset = run_configuration.dataset_info.dataset
# Restore checkpoint
optimizer = checkpoint_utils.restore_checkpoint(checkpoint_path, optimizer)
# Replicate optimizer.
optimizer = flax.jax_utils.replicate(optimizer)
predict_step = adapter.make_predict_step()
predict_step_parallel = jax.pmap(predict_step, axis_name='batch')
# Perform inference
dataset_iter_raw = iter(dataset)
dataset_iter = adapter.preprocess(dataset_iter_raw)
metrics_all = []
for example in itertools.islice(dataset_iter, 200):
train_inputs = adapter.get_train_inputs(example)
metrics, logits, state = predict_step_parallel(optimizer.target,
train_inputs)
adapter.handle_predict(metrics, logits, state)
metrics_all.append(metrics)
metrics_all = common_utils.get_metrics(metrics_all)
metrics = jax.tree_map(jnp.sum, metrics_all)
return metrics
def train_for_one_epoch(
dataset_source: dataset_source_lib.DatasetSource,
optimizer: flax.optim.Optimizer, state: flax.nn.Collection,
prng_key: jnp.ndarray, pmapped_train_step: _TrainStep,
pmapped_update_ema: Optional[_EMAUpdateStep],
moving_averages: Optional[efficientnet_optim.ExponentialMovingAverage],
summary_writer: tensorboard.SummaryWriter
) -> Tuple[flax.optim.Optimizer, flax.nn.Collection,
Optional[efficientnet_optim.ExponentialMovingAverage]]:
"""Trains the model for one epoch.
Args:
dataset_source: Container for the training dataset.
optimizer: The optimizer targeting the model to train.
state: Current state associated with the model (contains the batch norm MA).
prng_key: A PRNG key to use for stochasticity (e.g. for sampling an eventual
dropout mask). Is not used for shuffling the dataset.
pmapped_train_step: A pmapped version of the `train_step` function (see its
documentation for more details).
pmapped_update_ema: Function to update the parameter moving average. Can be
None if we don't use EMA.
moving_averages: Parameters moving average if used.
summary_writer: A Tensorboard SummaryWriter to use to log metrics.
Returns:
The updated optimizer (with the associated updated model), state and PRNG
key.
"""
start_time = time.time()
cnt = 0
train_metrics = []
for batch in dataset_source.get_train(use_augmentations=True):
# Generate a PRNG key that will be rolled into the batch.
step_key = jax.random.fold_in(prng_key, optimizer.state.step[0])
# Load and shard the TF batch.
batch = tensorflow_to_numpy(batch)
batch = shard_batch(batch)
# Shard the step PRNG key.
sharded_keys = common_utils.shard_prng_key(step_key)
optimizer, state, metrics, lr = pmapped_train_step(
optimizer, state, batch, sharded_keys)
cnt += 1
if moving_averages is not None:
moving_averages = pmapped_update_ema(optimizer, state, moving_averages)
train_metrics.append(metrics)
train_metrics = common_utils.get_metrics(train_metrics)
# Get training epoch summary for logging.
train_summary = jax.tree_map(lambda x: x.mean(), train_metrics)
train_summary['learning_rate'] = lr[0]
current_step = int(optimizer.state.step[0])
info = 'Whole training step done in {} ({} steps)'.format(
time.time()-start_time, cnt)
logging.info(info)
for metric_name, metric_value in train_summary.items():
summary_writer.scalar(metric_name, metric_value, current_step)
summary_writer.flush()
return optimizer, state, moving_averages
def eval_split(self, train_state, split_name, eval_env_ids=None):
"""Evaluation loop on the specified split.
Args:
train_state: TrainState; Object containing training state.
split_name: str; Name of the data split we want to evaluate the model on.
eval_env_ids: list(int); Eval environments ids.
Returns:
eval_summary, train_state
"""
data_iters = self.task.dataset.data_iters[split_name]
if eval_env_ids is None:
eval_env_ids = list(map(int, data_iters.keys()))
eval_metrics = {}
if isinstance(self.steps_per_eval, dict):
for env_id in eval_env_ids:
env_id_str = str(env_id)
env_eval_metrics = []
for _ in range(self.steps_per_eval[split_name][env_id_str]):
env_eval_batches = self.get_next_batch(
[data_iters[env_id_str]])
e_metrics = self.pmapped_eval_step(train_state,
env_eval_batches,
env_id)
env_eval_metrics.append(e_metrics)
env_eval_metrics = common_utils.get_metrics(env_eval_metrics)
eval_metrics.update(env_eval_metrics)
eval_summary = pipeline_utils.compute_global_mean_metrics(
eval_metrics)
else:
_, data_iters = list(zip(*dict(data_iters).items()))
eval_metrics = []
for _ in range(self.steps_per_eval):
env_eval_batches = self.get_next_batch(data_iters)
e_metrics = self.pmapped_eval_step(train_state,
env_eval_batches, -1)
eval_metrics.append(e_metrics)
eval_metrics = common_utils.get_metrics(eval_metrics)
eval_summary = pipeline_utils.compute_global_mean_metrics(
eval_metrics)
return eval_summary, eval_metrics
def write_train_metric(summary_writer, train_metrics, train_time, step):
summary_writer.scalar("train_time", train_time, step)
train_metrics = get_metrics(train_metrics)
for key, vals in train_metrics.items():
tag = f"train_{key}"
for i, val in enumerate(vals):
summary_writer.scalar(tag, val, step - len(vals) + i + 1)
def eval_once(run_configuration, checkpoint_path, optimizer=None):
"""Evaluates a single checkpoint on a single epoch of data."""
config = run_configuration.config
run_dir = run_configuration.run_dir
adapter = run_configuration.adapter
optimizer = optimizer or adapter.create_optimizer(run_configuration)
dataset = run_configuration.dataset_info.dataset
info = run_configuration.dataset_info.info
eval_name = config.eval_name or 'eval'
log_dir = os.path.join(run_dir, eval_name)
if jax.host_id() == 0:
summary_writer = tensorboard.SummaryWriter(log_dir)
# Restore checkpoint
optimizer = checkpoint_utils.restore_checkpoint(checkpoint_path, optimizer)
step = int(optimizer.state.step)
# Replicate optimizer.
optimizer = flax.jax_utils.replicate(optimizer)
eval_step = adapter.make_eval_step()
eval_step_parallel = jax.pmap(eval_step, axis_name='batch')
# Perform evaluation
tick = time.time()
metrics_all = []
example = None
dataset_iter_raw = iter(dataset)
dataset_iter = adapter.preprocess(dataset_iter_raw)
for unused_eval_step, example in zip(range(config.eval_steps),
dataset_iter):
train_inputs = adapter.get_train_inputs(example)
metrics, logits, state = eval_step_parallel(optimizer.target,
train_inputs)
metrics_all.append(metrics)
# Write results.
metrics_all = common_utils.get_metrics(metrics_all)
metrics_sums = jax.tree_map(jnp.sum, metrics_all)
denominator = metrics_sums.pop('denominator')
summary = jax.tree_map(lambda x: x / denominator, metrics_sums) # pylint: disable=cell-var-from-loop
summary['perplexity'] = jnp.clip(jnp.exp(summary['loss']), a_max=1.0e4)
logging.info('eval @ train step: %d, loss: %.4f', step, summary['loss'])
if jax.host_id() == 0:
tock = time.time()
steps_per_sec = len(metrics_all) / (tock - tick)
examples_per_sec = denominator / (tock - tick)
summary_writer.scalar('per-second/steps', steps_per_sec, step)
summary_writer.scalar('per-second/examples', examples_per_sec, step)
for key, val in summary.items():
summary_writer.scalar(key, val, step)
adapter.write_summaries(example, logits, summary_writer, info, step,
state)
summary_writer.flush()
def write_metric(train_metrics, eval_metrics, train_time, step):
summary_writer.scalar("train_time", train_time, step)
train_metrics = get_metrics(train_metrics)
for key, vals in train_metrics.items():
tag = f"train_{key}"
for i, val in enumerate(vals):
summary_writer.scalar(tag, val, step - len(vals) + i + 1)
for metric_name, value in eval_metrics.items():
summary_writer.scalar(f"eval_{metric_name}", value, step)
def evaluate(p_eval_step, state, eval_ds, num_eval_steps=-1):
"""Evaluate on the given dataset."""
logging.info('Starting evaluating.')
eval_metrics = []
for step, batch in enumerate(eval_ds):
batch = jax.tree_map(np.asarray, batch)
metrics = p_eval_step(batch=batch, state=state)
eval_metrics.append(metrics)
if num_eval_steps > 0 and step + 1 == num_eval_steps:
break
eval_metrics = common_utils.get_metrics(eval_metrics)
summary = train_utils.metrics_summary(eval_metrics, 'eval')
return summary
def evaluate(p_eval_step, params, eval_ds, rng):
"""Evaluate the target and return a dictionary with the metrics."""
logging.info('Gathering evaluation metrics.')
eval_metrics = []
for eval_batch in eval_ds:
eval_batch = jax.tree_map(lambda x: x._numpy(), eval_batch) # pylint: disable=protected-access
eval_batch = common_utils.shard(eval_batch)
metrics, rng = p_eval_step(rng, params, eval_batch)
eval_metrics.append(metrics)
eval_metrics = common_utils.get_metrics(eval_metrics)
eval_summary = jax.tree_map(np.mean, eval_metrics)
return eval_summary, rng
def eval_policy(policy, rng, state, model, test_ds):
"""Evaluate the target with policy and return a dictionary with the metrics."""
eval_metrics = []
for eval_batch in test_ds:
eval_batch = jax.tree_map(lambda x: x._numpy(), eval_batch) # pylint: disable=protected-access
eval_batch = common_utils.shard(eval_batch)
metrics, rng = eval_step_policy(rng, eval_batch, state, model, policy)
# Better to leave metrics on device, and off-load after finishing epoch.
eval_metrics.append(metrics)
eval_metrics = common_utils.get_metrics(eval_metrics)
eval_summary = jax.tree_map(np.mean, eval_metrics)
return eval_summary
def write_train_metric(train_metrics, train_time, step):
train_metrics = get_metrics(train_metrics)
num_steps = len(list(train_metrics.values())[0])
for key, vals in train_metrics.items():
assert len(vals) == num_steps
train_metrics_by_step = [{} for _ in range(num_steps)]
for key, vals in train_metrics.items():
for i, val in enumerate(vals):
train_metrics_by_step[i][f"train_{key}"] = val
for i in range(num_steps):
wandb.log(train_metrics_by_step[i], step=step - num_steps + i + 1)
wandb.log({"train_time": train_time}, step=step)
def evaluate(*, p_eval_step, target, eval_ds):
"""Evaluate the target an return a dictionary with the metrics."""
eval_metrics = []
for batches in eval_ds.as_numpy_iterator():
inputs, outputs, programs, _ = common_utils.shard(batches)
metrics = p_eval_step(target, inputs, outputs, programs)
eval_metrics.append(metrics)
eval_metrics = common_utils.get_metrics(eval_metrics)
eval_metrics_sums = jax.tree_map(jnp.sum, eval_metrics)
eval_denominator = eval_metrics_sums.pop('denominator')
eval_summary = jax.tree_map(
lambda x: x / eval_denominator, # pylint: disable=cell-var-from-loop
eval_metrics_sums)
return eval_summary
def evaluate(*, p_eval_step, target, eval_ds, num_eval_steps):
"""Evaluate the target an return a dictionary with the metrics."""
logging.info('Gathering evaluation metrics.')
eval_metrics = []
eval_iter = iter(eval_ds) # pytype: disable=wrong-arg-types
for _, eval_batch in zip(range(num_eval_steps), eval_iter):
eval_batch = jax.tree_map(lambda x: x._numpy(), eval_batch) # pylint: disable=protected-access
eval_batch = common_utils.shard(eval_batch)
metrics = p_eval_step(target, eval_batch)
eval_metrics.append(metrics)
eval_metrics = common_utils.get_metrics(eval_metrics)
eval_metrics_sums = jax.tree_map(jnp.sum, eval_metrics)
eval_denominator = eval_metrics_sums.pop('denominator')
eval_summary = jax.tree_map(
lambda x: x / eval_denominator, # pylint: disable=cell-var-from-loop
eval_metrics_sums)
return eval_summary
def train_for_one_epoch(
dataset_source,
optimizer, state,
prng_key, pmapped_train_step,
summary_writer
):
"""Trains the model for one epoch.
Args:
dataset_source: Container for the training dataset.
optimizer: The optimizer targeting the model to train.
state: Current state associated with the model (contains the batch norm MA).
prng_key: A PRNG key to use for stochasticity (e.g. for sampling an eventual
dropout mask). Is not used for shuffling the dataset.
pmapped_train_step: A pmapped version of the `train_step` function (see its
documentation for more details).
summary_writer: A Tensorboard SummaryWriter to use to log metrics.
Returns:
The updated optimizer (with the associated updated model), state and PRNG
key.
"""
train_metrics = []
for batch in dataset_source.get_train(use_augmentations=True):
# Generate a PRNG key that will be rolled into the batch.
step_key, prng_key = jax.random.split(prng_key)
# Load and shard the TF batch.
batch = tensorflow_to_numpy(batch)
batch = shard_batch(batch)
# Shard the step PRNG key.
sharded_keys = common_utils.shard_prng_key(step_key)
optimizer, state, metrics, lr = pmapped_train_step(
optimizer, state, batch, sharded_keys)
train_metrics.append(metrics)
train_metrics = common_utils.get_metrics(train_metrics)
# Get training epoch summary for logging.
train_summary = jax.tree_map(lambda x: x.mean(), train_metrics)
train_summary['learning_rate'] = lr[0]
current_step = int(optimizer.state.step[0])
for metric_name, metric_value in train_summary.items():
summary_writer.scalar(metric_name, metric_value, current_step)
summary_writer.flush()
return optimizer, state, prng_key
def eval_on_dataset(
model: flax.nn.Model, state: flax.nn.Collection, dataset: tf.data.Dataset,
pmapped_eval_step: _EvalStep):
"""Evaluates the model on the whole dataset.
Args:
model: The model to evaluate.
state: Current state associated with the model (contains the batch norm MA).
dataset: Dataset on which the model should be evaluated. Should already
being batched.
pmapped_eval_step: A pmapped version of the `eval_step` function (see its
documentation for more details).
Returns:
A dictionary containing the loss and error rate on the batch. These metrics
are averaged over the samples.
"""
eval_metrics = []
total_num_samples = 0
all_host_psum = jax.pmap(lambda x: jax.lax.psum(x, 'i'), 'i')
for eval_batch in dataset:
# Load and shard the TF batch.
eval_batch = load_and_shard_tf_batch(eval_batch)
# Compute metrics and sum over all observations in the batch.
metrics = pmapped_eval_step(model, state, eval_batch)
eval_metrics.append(metrics)
if 'mask' not in eval_batch:
# Number of samples seen in num_replicas * per_replica_batch_size.
total_num_samples += (
eval_batch['label'].shape[0] * eval_batch['label'].shape[1] *
jax.host_count())
else:
total_num_samples += all_host_psum(eval_batch['mask'])[0].sum()
# Metrics are all the same across all replicas (since we applied psum in the
# eval_step). The next line will fetch the metrics on one of them.
eval_metrics = common_utils.get_metrics(eval_metrics)
# Finally, we divide by the number of samples to get the mean error rate and
# cross entropy.
eval_summary = jax.tree_map(lambda x: x.sum() / total_num_samples,
eval_metrics)
return eval_summary
def test_reproduce_paper_evals(self, num_chunks):
"""Reproduce results from https://www.aclweb.org/anthology/P18-1009.pdf."""
num_samples = self.labels.shape[0]
chunk_size = num_samples // num_chunks
metrics = []
for chunk_start in range(0, num_samples, chunk_size):
chunk_end = min(chunk_start + chunk_size, num_samples)
labels = self.labels[chunk_start:chunk_end]
predictions = self.predictions[chunk_start:chunk_end]
current_metrics = ultra_fine_entity_typing_task.get_prediction_recall_metrics(
labels, predictions)
current_metrics = jax.tree_map(lambda x: jnp.expand_dims(x, 0),
current_metrics)
metrics.append(current_metrics)
metrics = common_utils.get_metrics(metrics)
metrics_sum = jax.tree_map(jnp.sum, metrics)
processed_metrics = metric_utils.process_metrics(metrics_sum)
self.assertAlmostEqual(processed_metrics['total_precision_value'],
0.481,
places=3)
self.assertAlmostEqual(processed_metrics['total_recall_value'],
0.232,
places=3)
self.assertAlmostEqual(
processed_metrics['coarse_grained_precision_value'],
0.603,
places=3)
self.assertAlmostEqual(
processed_metrics['coarse_grained_recall_value'], 0.616, places=3)
self.assertAlmostEqual(
processed_metrics['fine_grained_precision_value'], 0.404, places=3)
self.assertAlmostEqual(processed_metrics['fine_grained_recall_value'],
0.384,
places=3)
self.assertAlmostEqual(
processed_metrics['ultra_fine_grained_precision_value'],
0.428,
places=3)
self.assertAlmostEqual(
processed_metrics['ultra_fine_grained_recall_value'],
0.088,
places=3)
def eval_split(self, train_state, split_name):
"""Evaluation loop on the specified split.
Args:
train_state: TrainState; Object containing training state.
split_name: str; Name of the data split we want to evaluate the model on.
Returns:
eval_summary, train_state
"""
data_iters = self.task.dataset.data_iters[split_name]
eval_metrics = []
for _ in range(self.steps_per_eval):
env_eval_batches = self.get_next_batch(data_iters)
e_metrics = self.pmapped_eval_step(train_state, env_eval_batches)
eval_metrics.append(e_metrics)
eval_metrics = common_utils.get_metrics(eval_metrics)
eval_summary = pipeline_utils.compute_global_mean_metrics(eval_metrics)
return eval_summary, eval_metrics
def evaluate(
eval_step_fn,
train_state: ts.TrainState,
model_vars: Dict[str, Any],
eval_data: Sequence[Dict[str, Any]],
) -> Tuple[Dict[str, Any], Sequence[Tuple[Dict[str, Any], Optional[Dict[
str, Any]]]]]:
"""Evaluate current parameters and return a dictionary with metrics.
Args:
eval_step_fn: partial eval step that takes in model params and inputs only
train_state: contains model params, loss fn, grad update fn.
model_vars: model variables that are not optimized.
eval_data: sequence of evaluation data.
Returns:
Dictionary of metrics aggregated over all evaluation steps and the info
for the very first batch (batch itself and corresponding auxiliary output).
"""
logging.info('Performing evaluation.')
eval_metrics = []
eval_auxiliary = []
for batch in eval_data:
batch = jax.tree_map(jnp.asarray, batch)
metrics, auxiliary_output = eval_step_fn(
train_state,
model_vars,
batch,
)
eval_metrics.append(metrics)
batch_auxiliary = (jax.device_get(batch), jax.device_get(auxiliary_output))
eval_auxiliary.append(batch_auxiliary)
eval_metrics = common_utils.get_metrics(eval_metrics)
eval_metrics_sums = jax.tree_map(jnp.sum, eval_metrics)
eval_summary = metric_utils.process_metrics(eval_metrics_sums, prefix='eval')
return eval_summary, eval_auxiliary
def run_eval(eval_ds, num_eval_steps=-1):
eval_metrics = []
eval_iter = iter(eval_ds)
if num_eval_steps == -1:
num_iter = itertools.count()
else:
num_iter = range(num_eval_steps)
for _, eval_batch in zip(num_iter, eval_iter):
# pylint: disable=protected-access
eval_batch = common_utils.shard(
jax.tree_map(lambda x: x._numpy(), eval_batch))
# pylint: enable=protected-access
metrics = p_eval_step(optimizer.target, eval_batch)
eval_metrics.append(metrics)
eval_metrics = common_utils.get_metrics(eval_metrics)
eval_metrics_sums = jax.tree_map(jnp.sum, eval_metrics)
eval_denominator = eval_metrics_sums.pop('denominator')
eval_summary = jax.tree_map(
lambda x: x / eval_denominator, # pylint: disable=cell-var-from-loop
eval_metrics_sums)
# Calculate (clipped) perplexity after averaging log-perplexities:
eval_summary['perplexity'] = jnp.clip(jnp.exp(eval_summary['loss']),
a_max=1.0e4)
return eval_summary
def main(argv):
if len(argv) > 1:
raise app.UsageError('Too many command-line arguments.')
tf.enable_v2_behavior()
batch_size = FLAGS.batch_size
learning_rate = FLAGS.learning_rate
num_train_steps = FLAGS.num_train_steps
num_eval_steps = FLAGS.num_eval_steps
eval_freq = FLAGS.eval_frequency
max_length = FLAGS.max_length
random_seed = FLAGS.random_seed
if not FLAGS.dev:
raise app.UsageError('Please provide path to dev set.')
if not FLAGS.train:
raise app.UsageError('Please provide path to training set.')
parameter_path = os.path.join(FLAGS.model_dir, FLAGS.experiment + '.params')
if jax.host_id() == 0:
train_summary_writer = tensorboard.SummaryWriter(
os.path.join(FLAGS.model_dir, FLAGS.experiment + '_train'))
eval_summary_writer = tensorboard.SummaryWriter(
os.path.join(FLAGS.model_dir, FLAGS.experiment + '_eval'))
if batch_size % jax.device_count() > 0:
raise ValueError('Batch size must be divisible by the number of devices')
device_batch_size = batch_size // jax.device_count()
# create the training and development dataset
vocabs = input_pipeline.create_vocabs(FLAGS.train)
attributes_input = [input_pipeline.CoNLLAttributes.FORM]
attributes_target = [input_pipeline.CoNLLAttributes.XPOS]
train_ds = input_pipeline.sentence_dataset_dict(
FLAGS.train,
vocabs,
attributes_input,
attributes_target,
batch_size=batch_size,
bucket_size=max_length)
eval_ds = input_pipeline.sentence_dataset_dict(
FLAGS.dev,
vocabs,
attributes_input,
attributes_target,
batch_size=batch_size,
bucket_size=max_length,
repeat=1)
train_iter = iter(train_ds)
bs = device_batch_size * jax.device_count()
rng = random.PRNGKey(random_seed)
rng, init_rng = random.split(rng)
input_shape = (bs, max_length)
transformer_kwargs = {
'vocab_size': len(vocabs['forms']),
'output_vocab_size': len(vocabs['xpos']),
'emb_dim': 512,
'num_heads': 8,
'num_layers': 6,
'qkv_dim': 512,
'mlp_dim': 2048,
'max_len': max_length,
}
model = create_model(init_rng, tuple(input_shape), transformer_kwargs)
optimizer = create_optimizer(model, learning_rate)
del model # don't keep a copy of the initial model
learning_rate_fn = create_learning_rate_scheduler(
base_learning_rate=learning_rate)
p_train_step = jax.pmap(
functools.partial(train_step, learning_rate_fn=learning_rate_fn),
axis_name='batch')
p_eval_step = jax.pmap(eval_step, axis_name='batch')
# We init the first set of dropout PRNG keys, but update it afterwards inside
# the main pmap'd training update for performance.
dropout_rngs = random.split(rng, jax.local_device_count())
metrics_all = []
tick = time.time()
best_dev_score = 0
for step, batch in zip(range(num_train_steps), train_iter):
batch = common_utils.shard(jax.tree_map(lambda x: x._numpy(), batch)) # pylint: disable=protected-access
optimizer, metrics, dropout_rngs = p_train_step(
optimizer, batch, dropout_rng=dropout_rngs)
metrics_all.append(metrics)
if (step + 1) % eval_freq == 0:
metrics_all = common_utils.get_metrics(metrics_all)
lr = metrics_all.pop('learning_rate').mean()
metrics_sums = jax.tree_map(jnp.sum, metrics_all)
denominator = metrics_sums.pop('denominator')
summary = jax.tree_map(lambda x: x / denominator, metrics_sums) # pylint: disable=cell-var-from-loop
summary['learning_rate'] = lr
# Calculate (clipped) perplexity after averaging log-perplexities:
summary['perplexity'] = jnp.clip(jnp.exp(summary['loss']), a_max=1.0e4)
logging.info('train in step: %d, loss: %.4f', step, summary['loss'])
if jax.host_id() == 0:
tock = time.time()
steps_per_sec = eval_freq / (tock - tick)
tick = tock
train_summary_writer.scalar('steps per second', steps_per_sec, step)
for key, val in summary.items():
train_summary_writer.scalar(key, val, step)
train_summary_writer.flush()
# reset metric accumulation for next evaluation cycle.
metrics_all = []
eval_metrics = []
eval_iter = iter(eval_ds)
if num_eval_steps == -1:
num_iter = itertools.repeat(1)
else:
num_iter = range(num_eval_steps)
for _, eval_batch in zip(num_iter, eval_iter):
eval_batch = jax.tree_map(lambda x: x._numpy(), eval_batch) # pylint: disable=protected-access
# Handle final odd-sized batch by padding instead of dropping it.
cur_pred_batch_size = eval_batch['inputs'].shape[0]
if cur_pred_batch_size != batch_size:
logging.info('Uneven batch size %d.', cur_pred_batch_size)
eval_batch = jax.tree_map(
lambda x: pad_examples(x, batch_size), eval_batch)
eval_batch = common_utils.shard(eval_batch)
metrics = p_eval_step(optimizer.target, eval_batch)
eval_metrics.append(metrics)
eval_metrics = common_utils.get_metrics(eval_metrics)
eval_metrics_sums = jax.tree_map(jnp.sum, eval_metrics)
eval_denominator = eval_metrics_sums.pop('denominator')
eval_summary = jax.tree_map(
lambda x: x / eval_denominator, # pylint: disable=cell-var-from-loop
eval_metrics_sums)
# Calculate (clipped) perplexity after averaging log-perplexities:
eval_summary['perplexity'] = jnp.clip(
jnp.exp(eval_summary['loss']), a_max=1.0e4)
logging.info('eval in step: %d, loss: %.4f, accuracy: %.4f', step,
eval_summary['loss'], eval_summary['accuracy'])
if best_dev_score < eval_summary['accuracy']:
best_dev_score = eval_summary['accuracy']
# TODO: save model.
eval_summary['best_dev_score'] = best_dev_score
logging.info('best development model score %.4f', best_dev_score)
if jax.host_id() == 0:
for key, val in eval_summary.items():
eval_summary_writer.scalar(key, val, step)
eval_summary_writer.flush()
def run_train(run_configuration):
"""Runs the training workflow."""
config = run_configuration.config
run_dir = run_configuration.run_dir
adapter = run_configuration.adapter
log_dir = os.path.join(run_dir, 'train')
checkpoint_path = run_configuration.original_checkpoint_path
dataset = run_configuration.dataset_info.dataset
info = run_configuration.dataset_info.info
random_seed = 0
rng = jax.random.PRNGKey(random_seed)
rng = jax.random.fold_in(rng, jax.host_id())
rng, init_rng = jax.random.split(rng)
dropout_rngs = jax.random.split(rng, jax.local_device_count())
# Set up optimizer.
optimizer = adapter.create_optimizer(run_configuration, rng=init_rng)
# Set up train step.
train_step = adapter.make_train_step()
if jax.host_id() == 0:
summary_writer = tensorboard.SummaryWriter(log_dir)
# Set up checkpointing.
# TODO(dbieber): Set up phoenix.
checkpoint_dir = checkpoint_utils.build_checkpoint_dir(run_dir)
if checkpoint_path is None:
checkpoint_path = checkpoint_utils.latest_checkpoint(checkpoint_dir)
optimizer = checkpoint_utils.handle_restart_behavior(
checkpoint_path, optimizer, config)
start_step = int(optimizer.state.step)
num_train_steps = config.train.total_steps
# Replicate optimizer.
optimizer = flax.jax_utils.replicate(optimizer)
# Begin training loop.
dataset_iter_raw = iter(dataset)
dataset_iter = adapter.preprocess(dataset_iter_raw)
summary_freq = config.logging.summary_freq
metrics_all = []
tick = time.time()
for step, example in zip(range(start_step, num_train_steps), dataset_iter):
train_inputs = adapter.get_train_inputs(example)
optimizer, metrics, dropout_rngs, logits, state = train_step(
optimizer, train_inputs, dropout_rngs)
metrics_all.append(metrics)
# Save a Checkpoint
if ((step % config.logging.save_freq == 0 and step > 0)
or step == num_train_steps - 1):
if jax.host_id() == 0 and config.logging.save_freq:
# Save unreplicated optimizer + model state.
checkpoint_utils.save_checkpoint(
checkpoint_dir, jax_utils.unreplicate(optimizer), step)
# Periodic metric handling.
if summary_freq and step % summary_freq == 0 and step > 0:
metrics_all = common_utils.get_metrics(metrics_all)
lr = metrics_all.pop('learning_rate').mean()
metrics_sums = jax.tree_map(jnp.sum, metrics_all)
denominator = metrics_sums.pop('denominator')
summary = jax.tree_map(lambda x: x / denominator, metrics_sums) # pylint: disable=cell-var-from-loop
summary['learning_rate'] = lr
# Calculate (clipped) perplexity after averaging log-perplexities:
summary['perplexity'] = jnp.clip(jnp.exp(summary['loss']),
a_max=1.0e4)
logging.info('train step: %d, loss: %.4f', step, summary['loss'])
if jax.host_id() == 0:
tock = time.time()
steps_per_sec = summary_freq / (tock - tick)
examples_per_sec = denominator / (tock - tick)
tick = tock
summary_writer.scalar('per-second/steps', steps_per_sec, step)
summary_writer.scalar('per-second/examples', examples_per_sec,
step)
for key, val in summary.items():
summary_writer.scalar(key, val, step)
adapter.write_summaries(example, logits, summary_writer, info,
step, state)
summary_writer.flush()
# Reset metric accumulation for next evaluation cycle.
metrics_all = []
def main(argv):
if len(argv) > 1:
raise app.UsageError('Too many command-line arguments.')
tf.enable_v2_behavior()
# make sure tf does not allocate gpu memory
tf.config.experimental.set_visible_devices([], 'GPU')
if jax.host_id() == 0:
summary_writer = tensorboard.SummaryWriter(FLAGS.model_dir)
rng = random.PRNGKey(0)
image_size = 224
batch_size = FLAGS.batch_size
if batch_size % jax.device_count() > 0:
raise ValueError(
'Batch size must be divisible by the number of devices')
local_batch_size = batch_size // jax.host_count()
device_batch_size = batch_size // jax.device_count()
platform = jax.local_devices()[0].platform
if FLAGS.half_precision:
if platform == 'tpu':
model_dtype = jnp.bfloat16
input_dtype = tf.bfloat16
else:
model_dtype = jnp.float16
input_dtype = tf.float16
else:
model_dtype = jnp.float32
input_dtype = tf.float32
train_iter = create_input_iter(local_batch_size,
image_size,
input_dtype,
train=True,
cache=FLAGS.cache)
eval_iter = create_input_iter(local_batch_size,
image_size,
input_dtype,
train=False,
cache=FLAGS.cache)
num_epochs = FLAGS.num_epochs
steps_per_epoch = input_pipeline.TRAIN_IMAGES // batch_size
steps_per_eval = input_pipeline.EVAL_IMAGES // batch_size
steps_per_checkpoint = steps_per_epoch * 10
num_steps = steps_per_epoch * num_epochs
base_learning_rate = FLAGS.learning_rate * batch_size / 256.
base_learning_rate = base_learning_rate / FLAGS.loss_scaling
model, model_state = create_model(rng, device_batch_size, image_size,
model_dtype)
optimizer = optim.Momentum(beta=FLAGS.momentum,
nesterov=True).create(model)
state = TrainState(step=0, optimizer=optimizer, model_state=model_state)
del model, model_state # do not keep a copy of the initial model
state = restore_checkpoint(state)
step_offset = int(
state.step) # step_offset > 0 if restarting from checkpoint
state = jax_utils.replicate(state)
learning_rate_fn = create_learning_rate_fn(base_learning_rate,
steps_per_epoch, num_epochs)
p_train_step = jax.pmap(functools.partial(
train_step, learning_rate_fn=learning_rate_fn),
axis_name='batch')
p_eval_step = jax.pmap(eval_step, axis_name='batch')
epoch_metrics = []
t_loop_start = time.time()
for step, batch in zip(range(step_offset, num_steps), train_iter):
state, metrics = p_train_step(state, batch)
epoch_metrics.append(metrics)
if (step + 1) % steps_per_epoch == 0:
epoch = step // steps_per_epoch
epoch_metrics = common_utils.get_metrics(epoch_metrics)
summary = jax.tree_map(lambda x: x.mean(), epoch_metrics)
logging.info('train epoch: %d, loss: %.4f, accuracy: %.2f', epoch,
summary['loss'], summary['accuracy'] * 100)
steps_per_sec = steps_per_epoch / (time.time() - t_loop_start)
t_loop_start = time.time()
if jax.host_id() == 0:
for key, vals in epoch_metrics.items():
tag = 'train_%s' % key
for i, val in enumerate(vals):
summary_writer.scalar(tag, val,
step - len(vals) + i + 1)
summary_writer.scalar('steps per second', steps_per_sec, step)
epoch_metrics = []
eval_metrics = []
# sync batch statistics across replicas
state = sync_batch_stats(state)
for _ in range(steps_per_eval):
eval_batch = next(eval_iter)
metrics = p_eval_step(state, eval_batch)
eval_metrics.append(metrics)
eval_metrics = common_utils.get_metrics(eval_metrics)
summary = jax.tree_map(lambda x: x.mean(), eval_metrics)
logging.info('eval epoch: %d, loss: %.4f, accuracy: %.2f', epoch,
summary['loss'], summary['accuracy'] * 100)
if jax.host_id() == 0:
for key, val in eval_metrics.items():
tag = 'eval_%s' % key
summary_writer.scalar(tag, val.mean(), step)
summary_writer.flush()
if (step + 1) % steps_per_checkpoint == 0 or step + 1 == num_steps:
state = sync_batch_stats(state)
save_checkpoint(state)
# Wait until computations are done before exiting
jax.random.normal(jax.random.PRNGKey(0), ()).block_until_ready()
def main(argv):
if len(argv) > 1:
raise app.UsageError('Too many command-line arguments.')
if FLAGS.jax_backend_target:
jax.config.FLAGS.jax_xla_backend = "tpu_driver"
jax.config.FLAGS.jax_backend_target = FLAGS.jax_backend_target
# This seems to be necessary even when importing TF2?
tf.enable_v2_behavior()
# Number of local devices for this host.
n_devices = jax.local_device_count()
if jax.host_id() == 0:
train_summary_writer = tensorboard.SummaryWriter(
os.path.join(FLAGS.model_dir, 'train'))
eval_summary_writer = tensorboard.SummaryWriter(
os.path.join(FLAGS.model_dir, 'eval'))
if FLAGS.batch_size % n_devices:
raise ValueError(
'Batch size must be divisible by the number of devices')
vocab_path = FLAGS.vocab_path
if vocab_path is None:
vocab_path = os.path.join(FLAGS.model_dir, 'sentencepiece_model')
# Load Dataset
logging.info('Initializing dataset.')
train_ds, eval_ds, predict_ds, encoder = input_pipeline.get_wmt_datasets(
n_devices=n_devices,
dataset_name=FLAGS.dataset_name,
eval_dataset_name=FLAGS.eval_dataset_name,
shard_idx=jax.host_id(),
shard_count=jax.host_count(),
data_dir=FLAGS.data_dir,
vocab_path=vocab_path,
batch_size=FLAGS.batch_size,
max_length=FLAGS.max_target_length,
max_eval_length=FLAGS.max_eval_target_length)
train_iter = iter(train_ds)
vocab_size = int(encoder.vocab_size())
eos_token = 2 # Default Sentencepiece EOS token.
def decode_tokens(toks):
valid_toks = toks[:np.argmax(toks == eos_token) + 1].astype(np.int32)
return encoder.detokenize(valid_toks).numpy().decode('utf-8')
logging.info('Initializing model, optimizer, and step functions.')
# Build Model and Optimizer
transformer_kwargs = {
'vocab_size': vocab_size,
'output_vocab_size': vocab_size,
'emb_dim': 1024,
'num_heads': 16,
'num_layers': 6,
'qkv_dim': 1024,
'mlp_dim': 4096,
'max_len': max(FLAGS.max_target_length, FLAGS.max_eval_target_length),
'share_embeddings': FLAGS.share_embeddings,
'logits_via_embedding': FLAGS.logits_via_embedding,
}
start_step = 0
rng = random.PRNGKey(FLAGS.random_seed)
rng, init_rng = random.split(rng)
input_shape = (FLAGS.batch_size, FLAGS.max_target_length)
target_shape = (FLAGS.batch_size, FLAGS.max_target_length)
model, cache_def = create_model(init_rng, input_shape, target_shape,
transformer_kwargs)
optimizer = create_optimizer(model, FLAGS.learning_rate,
FLAGS.weight_decay)
# We access model only from optimizer below via optimizer.target.
del model
if FLAGS.restore_checkpoints:
# Restore unreplicated optimizer + model state from last checkpoint.
optimizer = checkpoints.restore_checkpoint(FLAGS.model_dir, optimizer)
# Grab last step.
start_step = int(optimizer.state.step)
# Replicate optimizer.
optimizer = jax_utils.replicate(optimizer)
learning_rate_fn = create_learning_rate_scheduler(
base_learning_rate=FLAGS.learning_rate,
warmup_steps=FLAGS.warmup_steps)
p_train_step = jax.pmap(functools.partial(
train_step,
learning_rate_fn=learning_rate_fn,
label_smoothing=FLAGS.label_smoothing,
use_bfloat16=FLAGS.use_bfloat16),
axis_name='batch')
p_eval_step = jax.pmap(functools.partial(
eval_step,
label_smoothing=FLAGS.label_smoothing,
use_bfloat16=FLAGS.use_bfloat16),
axis_name='batch')
p_pred_step = jax.pmap(
functools.partial(predict_step, use_bfloat16=FLAGS.use_bfloat16),
axis_name='batch',
static_broadcasted_argnums=(3,
4)) # eos token, max_length are constant
# We init the first set of dropout PRNG keys, but update it afterwards inside
# the main pmap'd training update for performance.
dropout_rngs = random.split(rng, n_devices)
logging.info('Starting training loop.')
metrics_all = []
t_loop_start = time.time()
for step, batch in zip(range(start_step, FLAGS.num_train_steps),
train_iter):
# Shard data to devices and do a training step.
batch = common_utils.shard(jax.tree_map(lambda x: x._numpy(), batch)) # pylint: disable=protected-access
optimizer, metrics, dropout_rngs = p_train_step(
optimizer, batch, dropout_rng=dropout_rngs)
metrics_all.append(metrics)
# Save a checkpoint on one host after every checkpoint_freq steps.
if (FLAGS.save_checkpoints and step % FLAGS.checkpoint_freq == 0
and step > 0 and jax.host_id() == 0):
checkpoints.save_checkpoint(FLAGS.model_dir,
jax_utils.unreplicate(optimizer), step)
# Periodic metric handling.
if step % FLAGS.eval_frequency != 0:
continue
logging.info('Gathering training metrics.')
# Training Metrics
metrics_all = common_utils.get_metrics(metrics_all)
lr = metrics_all.pop('learning_rate').mean()
metrics_sums = jax.tree_map(jnp.sum, metrics_all)
denominator = metrics_sums.pop('denominator')
summary = jax.tree_map(lambda x: x / denominator, metrics_sums) # pylint: disable=cell-var-from-loop
summary['learning_rate'] = lr
steps_per_eval = FLAGS.eval_frequency if step != 0 else 1
steps_per_sec = steps_per_eval / (time.time() - t_loop_start)
t_loop_start = time.time()
if jax.host_id() == 0:
train_summary_writer.scalar('steps per second', steps_per_sec,
step)
for key, val in summary.items():
train_summary_writer.scalar(key, val, step)
train_summary_writer.flush()
metrics_all = []
logging.info('train in step: %d, loss: %.4f', step, summary['loss'])
# Eval Metrics
logging.info('Gathering evaluation metrics.')
t_eval_start = time.time()
eval_metrics = []
eval_iter = iter(eval_ds)
for _, eval_batch in zip(range(FLAGS.num_eval_steps), eval_iter):
eval_batch = jax.tree_map(lambda x: x._numpy(), eval_batch) # pylint: disable=protected-access
eval_batch = common_utils.shard(eval_batch)
metrics = p_eval_step(optimizer.target, eval_batch)
eval_metrics.append(metrics)
eval_metrics = common_utils.get_metrics(eval_metrics)
eval_metrics_sums = jax.tree_map(jnp.sum, eval_metrics)
eval_denominator = eval_metrics_sums.pop('denominator')
eval_summary = jax.tree_map(
lambda x: x / eval_denominator, # pylint: disable=cell-var-from-loop
eval_metrics_sums)
if jax.host_id() == 0:
for key, val in eval_summary.items():
eval_summary_writer.scalar(key, val, step)
eval_summary_writer.flush()
logging.info('eval in step: %d, loss: %.4f', step,
eval_summary['loss'])
logging.info('eval time: %.4f s step %d',
time.time() - t_eval_start, step)
# Translation and BLEU Score.
logging.info('Translating evaluation dataset.')
t_inference_start = time.time()
predict_iter = iter(predict_ds)
sources, references, predictions = [], [], []
for _, pred_batch in enumerate(predict_iter):
pred_batch = jax.tree_map(lambda x: x._numpy(), pred_batch) # pylint: disable=protected-access
# Handle final odd-sized batch by padding instead of dropping it.
cur_pred_batch_size = pred_batch['inputs'].shape[0]
if cur_pred_batch_size % n_devices:
padded_size = int(
np.ceil(cur_pred_batch_size / n_devices) * n_devices)
pred_batch = jax.tree_map(
lambda x: pad_examples(x, padded_size), pred_batch) # pylint: disable=cell-var-from-loop
pred_batch = common_utils.shard(pred_batch)
per_device_batchsize = pred_batch['inputs'].shape[1]
cache_dtype = jnp.bfloat16 if FLAGS.use_bfloat16 else jnp.float32
cache = jax_utils.replicate(
cache_def.initialize_cache(
(per_device_batchsize, FLAGS.max_predict_length),
dtype=cache_dtype))
predicted = p_pred_step(pred_batch['inputs'], optimizer.target,
cache, eos_token, FLAGS.max_predict_length)
predicted = tohost(predicted)
inputs = tohost(pred_batch['inputs'])
targets = tohost(pred_batch['targets'])
# Iterate through non-padding examples of batch.
for i, s in enumerate(predicted[:cur_pred_batch_size]):
sources.append(decode_tokens(inputs[i]))
references.append(decode_tokens(targets[i]))
predictions.append(decode_tokens(s))
logging.info('Translation: %d predictions %d references %d sources.',
len(predictions), len(references), len(sources))
logging.info('Translation time: %.4f s step %d.',
time.time() - t_inference_start, step)
# Calculate BLEU score for translated eval corpus against reference.
bleu_matches = bleu.bleu_partial(references, predictions)
all_bleu_matches = per_host_sum_pmap(bleu_matches)
bleu_score = bleu.complete_bleu(*all_bleu_matches)
# Save translation samples for tensorboard.
exemplars = ''
for n in np.random.choice(np.arange(len(predictions)), 8):
exemplars += f'{sources[n]}\n\n{references[n]}\n\n{predictions[n]}\n\n'
if jax.host_id() == 0:
eval_summary_writer.scalar('bleu', bleu_score, step)
eval_summary_writer.text('samples', exemplars, step)
eval_summary_writer.flush()
logging.info('Translation BLEU Score %.4f', bleu_score)
def main():
# See all possible arguments in src/transformers/training_args.py
# or by passing the --help flag to this script.
# We now keep distinct sets of args, for a cleaner separation of concerns.
parser = HfArgumentParser(
(ModelArguments, DataTrainingArguments, TrainingArguments))
if len(sys.argv) == 2 and sys.argv[1].endswith(".json"):
# If we pass only one argument to the script and it's the path to a json file,
# let's parse it to get our arguments.
model_args, data_args, training_args = parser.parse_json_file(
json_file=os.path.abspath(sys.argv[1]))
else:
model_args, data_args, training_args = parser.parse_args_into_dataclasses(
)
# Sending telemetry. Tracking the example usage helps us better allocate resources to maintain them. The
# information sent is the one passed as arguments along with your Python/PyTorch versions.
send_example_telemetry("run_t5_mlm",
model_args,
data_args,
framework="flax")
if (os.path.exists(training_args.output_dir)
and os.listdir(training_args.output_dir) and training_args.do_train
and not training_args.overwrite_output_dir):
raise ValueError(
f"Output directory ({training_args.output_dir}) already exists and is not empty."
"Use --overwrite_output_dir to overcome.")
# Setup logging
logging.basicConfig(
format="%(asctime)s - %(levelname)s - %(name)s - %(message)s",
level=logging.INFO,
datefmt="[%X]",
)
# Log on each process the small summary:
logger = logging.getLogger(__name__)
# Set the verbosity to info of the Transformers logger (on main process only):
logger.info(f"Training/evaluation parameters {training_args}")
# Set seed before initializing model.
set_seed(training_args.seed)
# Handle the repository creation
if training_args.push_to_hub:
if training_args.hub_model_id is None:
repo_name = get_full_repo_name(Path(
training_args.output_dir).absolute().name,
token=training_args.hub_token)
else:
repo_name = training_args.hub_model_id
repo = Repository(training_args.output_dir, clone_from=repo_name)
# Get the datasets: you can either provide your own CSV/JSON/TXT training and evaluation files (see below)
# or just provide the name of one of the public datasets available on the hub at https://huggingface.co/datasets/
# (the dataset will be downloaded automatically from the datasets Hub).
#
# For CSV/JSON files, this script will use the column called 'text' or the first column if no column called
# 'text' is found. You can easily tweak this behavior (see below).
if data_args.dataset_name is not None:
# Downloading and loading a dataset from the hub.
datasets = load_dataset(
data_args.dataset_name,
data_args.dataset_config_name,
cache_dir=model_args.cache_dir,
use_auth_token=True if model_args.use_auth_token else None,
)
if "validation" not in datasets.keys():
datasets["validation"] = load_dataset(
data_args.dataset_name,
data_args.dataset_config_name,
split=f"train[:{data_args.validation_split_percentage}%]",
cache_dir=model_args.cache_dir,
use_auth_token=True if model_args.use_auth_token else None,
)
datasets["train"] = load_dataset(
data_args.dataset_name,
data_args.dataset_config_name,
split=f"train[{data_args.validation_split_percentage}%:]",
cache_dir=model_args.cache_dir,
use_auth_token=True if model_args.use_auth_token else None,
)
else:
data_files = {}
if data_args.train_file is not None:
data_files["train"] = data_args.train_file
if data_args.validation_file is not None:
data_files["validation"] = data_args.validation_file
extension = data_args.train_file.split(".")[-1]
if extension == "txt":
extension = "text"
datasets = load_dataset(
extension,
data_files=data_files,
cache_dir=model_args.cache_dir,
use_auth_token=True if model_args.use_auth_token else None,
)
if "validation" not in datasets.keys():
datasets["validation"] = load_dataset(
extension,
data_files=data_files,
split=f"train[:{data_args.validation_split_percentage}%]",
cache_dir=model_args.cache_dir,
use_auth_token=True if model_args.use_auth_token else None,
)
datasets["train"] = load_dataset(
extension,
data_files=data_files,
split=f"train[{data_args.validation_split_percentage}%:]",
cache_dir=model_args.cache_dir,
use_auth_token=True if model_args.use_auth_token else None,
)
# See more about loading any type of standard or custom dataset (from files, python dict, pandas DataFrame, etc) at
# https://huggingface.co/docs/datasets/loading_datasets.html.
# Load pretrained model and tokenizer
if model_args.tokenizer_name:
tokenizer = AutoTokenizer.from_pretrained(
model_args.tokenizer_name,
cache_dir=model_args.cache_dir,
use_fast=model_args.use_fast_tokenizer,
use_auth_token=True if model_args.use_auth_token else None,
)
elif model_args.model_name_or_path:
tokenizer = AutoTokenizer.from_pretrained(
model_args.model_name_or_path,
cache_dir=model_args.cache_dir,
use_fast=model_args.use_fast_tokenizer,
use_auth_token=True if model_args.use_auth_token else None,
)
else:
raise ValueError(
"You are instantiating a new tokenizer from scratch. This is not supported by this script."
"You can do it from another script, save it, and load it from here, using --tokenizer_name."
)
if model_args.config_name:
config = T5Config.from_pretrained(
model_args.config_name,
cache_dir=model_args.cache_dir,
vocab_size=len(tokenizer),
use_auth_token=True if model_args.use_auth_token else None,
)
elif model_args.model_name_or_path:
config = T5Config.from_pretrained(
model_args.model_name_or_path,
cache_dir=model_args.cache_dir,
use_auth_token=True if model_args.use_auth_token else None,
)
else:
config = CONFIG_MAPPING[model_args.model_type]()
logger.warning(
"You are instantiating a new config instance from scratch.")
# Preprocessing the datasets.
# First we tokenize all the texts.
if training_args.do_train:
column_names = datasets["train"].column_names
else:
column_names = datasets["validation"].column_names
text_column_name = "text" if "text" in column_names else column_names[0]
max_seq_length = min(data_args.max_seq_length, tokenizer.model_max_length)
# Otherwise, we tokenize every text, then concatenate them together before splitting them in smaller parts.
# Since we make sure that all sequences are of the same length, no attention_mask is needed.
def tokenize_function(examples):
return tokenizer(examples[text_column_name],
return_attention_mask=False)
tokenized_datasets = datasets.map(
tokenize_function,
batched=True,
num_proc=data_args.preprocessing_num_workers,
remove_columns=column_names,
load_from_cache_file=not data_args.overwrite_cache,
)
# T5-like span masked language modeling will fuse consecutively masked tokens to a single sentinel token.
# To ensure that the input length is `max_seq_length`, we need to increase the maximum length
# according to `mlm_probability` and `mean_noise_span_length`. We can also define the label length accordingly.
expanded_inputs_length, targets_length = compute_input_and_target_lengths(
inputs_length=max_seq_length,
noise_density=data_args.mlm_probability,
mean_noise_span_length=data_args.mean_noise_span_length,
)
# Main data processing function that will concatenate all texts from our dataset and generate chunks of expanded_inputs_length.
def group_texts(examples):
# Concatenate all texts.
concatenated_examples = {
k: list(chain(*examples[k]))
for k in examples.keys()
}
total_length = len(concatenated_examples[list(examples.keys())[0]])
# We drop the small remainder, we could add padding if the model supported it instead of this drop, you can
# customize this part to your needs.
if total_length >= expanded_inputs_length:
total_length = (total_length //
expanded_inputs_length) * expanded_inputs_length
# Split by chunks of max_len.
result = {
k: [
t[i:i + expanded_inputs_length]
for i in range(0, total_length, expanded_inputs_length)
]
for k, t in concatenated_examples.items()
}
return result
# Note that with `batched=True`, this map processes 1,000 texts together, so group_texts throws away a
# remainder for each of those groups of 1,000 texts. You can adjust that batch_size here but a higher value
# might be slower to preprocess.
#
# To speed up this part, we use multiprocessing. See the documentation of the map method for more information:
# https://huggingface.co/docs/datasets/package_reference/main_classes.html#datasets.Dataset.map
tokenized_datasets = tokenized_datasets.map(
group_texts,
batched=True,
num_proc=data_args.preprocessing_num_workers,
load_from_cache_file=not data_args.overwrite_cache,
)
# Enable tensorboard only on the master node
has_tensorboard = is_tensorboard_available()
if has_tensorboard and jax.process_index() == 0:
try:
from flax.metrics.tensorboard import SummaryWriter
summary_writer = SummaryWriter(
log_dir=Path(training_args.output_dir))
except ImportError as ie:
has_tensorboard = False
logger.warning(
f"Unable to display metrics through TensorBoard because some package are not installed: {ie}"
)
else:
logger.warning(
"Unable to display metrics through TensorBoard because the package is not installed: "
"Please run pip install tensorboard to enable.")
# Initialize our training
rng = jax.random.PRNGKey(training_args.seed)
dropout_rngs = jax.random.split(rng, jax.local_device_count())
if model_args.model_name_or_path:
model = FlaxT5ForConditionalGeneration.from_pretrained(
model_args.model_name_or_path,
config=config,
seed=training_args.seed,
dtype=getattr(jnp, model_args.dtype),
use_auth_token=True if model_args.use_auth_token else None,
)
else:
config.vocab_size = len(tokenizer)
model = FlaxT5ForConditionalGeneration(
config,
seed=training_args.seed,
dtype=getattr(jnp, model_args.dtype),
)
# Data collator
# This one will take care of randomly masking the tokens.
data_collator = FlaxDataCollatorForT5MLM(
tokenizer=tokenizer,
noise_density=data_args.mlm_probability,
mean_noise_span_length=data_args.mean_noise_span_length,
input_length=max_seq_length,
target_length=targets_length,
pad_token_id=model.config.pad_token_id,
decoder_start_token_id=model.config.decoder_start_token_id,
)
# Store some constant
num_epochs = int(training_args.num_train_epochs)
train_batch_size = int(
training_args.per_device_train_batch_size) * jax.device_count()
per_device_eval_batch_size = int(training_args.per_device_eval_batch_size)
eval_batch_size = per_device_eval_batch_size * jax.device_count()
num_train_steps = len(
tokenized_datasets["train"]) // train_batch_size * num_epochs
num_of_hosts = jax.process_count()
current_host_idx = jax.process_index()
# Create learning rate schedule
warmup_fn = optax.linear_schedule(
init_value=0.0,
end_value=training_args.learning_rate,
transition_steps=training_args.warmup_steps)
decay_fn = optax.linear_schedule(
init_value=training_args.learning_rate,
end_value=0,
transition_steps=num_train_steps - training_args.warmup_steps,
)
linear_decay_lr_schedule_fn = optax.join_schedules(
schedules=[warmup_fn, decay_fn],
boundaries=[training_args.warmup_steps])
# We use Optax's "masking" functionality to not apply weight decay
# to bias and LayerNorm scale parameters. decay_mask_fn returns a
# mask boolean with the same structure as the parameters.
# The mask is True for parameters that should be decayed.
def decay_mask_fn(params):
flat_params = traverse_util.flatten_dict(params)
# find out all LayerNorm parameters
layer_norm_candidates = ["layernorm", "layer_norm", "ln"]
layer_norm_named_params = set([
layer[-2:] for layer_norm_name in layer_norm_candidates
for layer in flat_params.keys()
if layer_norm_name in "".join(layer).lower()
])
flat_mask = {
path: (path[-1] != "bias"
and path[-2:] not in layer_norm_named_params)
for path in flat_params
}
return traverse_util.unflatten_dict(flat_mask)
# create adam optimizer
if training_args.adafactor:
# We use the default parameters here to initialize adafactor,
# For more details about the parameters please check https://github.com/deepmind/optax/blob/ed02befef9bf81cbbf236be3d2b0e032e9ed4a40/optax/_src/alias.py#L74
optimizer = optax.adafactor(
learning_rate=linear_decay_lr_schedule_fn, )
else:
optimizer = optax.adamw(
learning_rate=linear_decay_lr_schedule_fn,
b1=training_args.adam_beta1,
b2=training_args.adam_beta2,
weight_decay=training_args.weight_decay,
mask=decay_mask_fn,
)
# Setup train state
state = train_state.TrainState.create(apply_fn=model.__call__,
params=model.params,
tx=optimizer)
# Define gradient update step fn
def train_step(state, batch, dropout_rng):
dropout_rng, new_dropout_rng = jax.random.split(dropout_rng)
def loss_fn(params):
labels = batch.pop("labels")
logits = state.apply_fn(**batch,
params=params,
dropout_rng=dropout_rng,
train=True)[0]
# compute loss
loss = optax.softmax_cross_entropy(
logits, onehot(labels, logits.shape[-1])).mean()
return loss
grad_fn = jax.value_and_grad(loss_fn)
loss, grad = grad_fn(state.params)
grad = jax.lax.pmean(grad, "batch")
new_state = state.apply_gradients(grads=grad)
metrics = jax.lax.pmean(
{
"loss": loss,
"learning_rate": linear_decay_lr_schedule_fn(state.step)
},
axis_name="batch")
return new_state, metrics, new_dropout_rng
# Create parallel version of the train step
p_train_step = jax.pmap(train_step, "batch", donate_argnums=(0, ))
# Define eval fn
def eval_step(params, batch):
labels = batch.pop("labels")
logits = model(**batch, params=params, train=False)[0]
# compute loss
loss = optax.softmax_cross_entropy(logits,
onehot(labels, logits.shape[-1]))
# compute accuracy
accuracy = jnp.equal(jnp.argmax(logits, axis=-1), labels)
# summarize metrics
metrics = {"loss": loss.mean(), "accuracy": accuracy.mean()}
metrics = jax.lax.pmean(metrics, axis_name="batch")
return metrics
p_eval_step = jax.pmap(eval_step, "batch", donate_argnums=(0, ))
# Replicate the train state on each device
state = jax_utils.replicate(state)
train_time = 0
epochs = tqdm(range(num_epochs), desc="Epoch ... ", position=0)
for epoch in epochs:
# ======================== Training ================================
train_start = time.time()
train_metrics = []
# Create sampling rng
rng, input_rng = jax.random.split(rng)
# Generate an epoch by shuffling sampling indices from the train dataset
num_train_samples = len(tokenized_datasets["train"])
# Avoid using jax.numpy here in case of TPU training
train_samples_idx = np.random.permutation(np.arange(num_train_samples))
train_batch_idx = generate_batch_splits(train_samples_idx,
train_batch_size)
# Gather the indexes for creating the batch and do a training step
for step, batch_idx in enumerate(
tqdm(train_batch_idx, desc="Training...", position=1)):
samples = [
tokenized_datasets["train"][int(idx)] for idx in batch_idx
]
model_inputs = data_collator(samples)
local_host_model_inputs = {
key: np.split(model_inputs.data[key], num_of_hosts,
axis=0)[current_host_idx]
for key, value in model_inputs.data.items()
}
# Model forward
model_inputs = shard(local_host_model_inputs)
state, train_metric, dropout_rngs = p_train_step(
state, model_inputs, dropout_rngs)
train_metrics.append(train_metric)
cur_step = epoch * (num_train_samples // train_batch_size) + step
if cur_step % training_args.logging_steps == 0 and cur_step > 0:
# Save metrics
train_metric = jax_utils.unreplicate(train_metric)
train_time += time.time() - train_start
if has_tensorboard and jax.process_index() == 0:
write_train_metric(summary_writer, train_metrics,
train_time, cur_step)
epochs.write(
f"Step... ({cur_step} | Loss: {train_metric['loss'].mean()}, Learning Rate:"
f" {train_metric['learning_rate'].mean()})")
train_metrics = []
if cur_step % training_args.eval_steps == 0 and cur_step > 0:
# ======================== Evaluating ==============================
num_eval_samples = len(tokenized_datasets["validation"])
# Avoid using jax.numpy here in case of TPU training
eval_samples_idx = np.arange(num_eval_samples)
eval_batch_idx = generate_batch_splits(eval_samples_idx,
eval_batch_size,
drop_last=False)
eval_metrics = []
for i, batch_idx in enumerate(
tqdm(eval_batch_idx, desc="Evaluating ...",
position=2)):
samples = [
tokenized_datasets["validation"][int(idx)]
for idx in batch_idx
]
model_inputs = data_collator(samples)
# Model forward
metrics = pad_shard_unpad(p_eval_step, static_return=True)(
state.params,
model_inputs.data,
min_device_batch=per_device_eval_batch_size)
eval_metrics.append(metrics)
# get eval metrics
eval_metrics = get_metrics(eval_metrics)
eval_metrics = jax.tree_map(jnp.mean, eval_metrics)
# Update progress bar
epochs.write(
f"Step... ({cur_step} | Loss: {eval_metrics['loss']}, Acc: {eval_metrics['accuracy']})"
)
# Save metrics
if has_tensorboard and jax.process_index() == 0:
write_eval_metric(summary_writer, eval_metrics, cur_step)
if cur_step % training_args.save_steps == 0 and cur_step > 0:
# save checkpoint after each epoch and push checkpoint to the hub
if jax.process_index() == 0:
params = jax.device_get(
jax.tree_map(lambda x: x[0], state.params))
model.save_pretrained(training_args.output_dir,
params=params)
tokenizer.save_pretrained(training_args.output_dir)
if training_args.push_to_hub:
repo.push_to_hub(
commit_message=
f"Saving weights and logs of step {cur_step}",
blocking=False)
# Eval after training
if training_args.do_eval:
num_eval_samples = len(tokenized_datasets["validation"])
# Avoid using jax.numpy here in case of TPU training
eval_samples_idx = np.arange(num_eval_samples)
eval_batch_idx = generate_batch_splits(eval_samples_idx,
eval_batch_size,
drop_last=False)
eval_metrics = []
for i, batch_idx in enumerate(
tqdm(eval_batch_idx, desc="Evaluating ...", position=2)):
samples = [
tokenized_datasets["validation"][int(idx)] for idx in batch_idx
]
model_inputs = data_collator(samples)
# Model forward
metrics = pad_shard_unpad(p_eval_step, static_return=True)(
state.params,
model_inputs.data,
min_device_batch=per_device_eval_batch_size)
eval_metrics.append(metrics)
# get eval metrics
eval_metrics = get_metrics(eval_metrics)
eval_metrics = jax.tree_map(lambda metric: jnp.mean(metric).item(),
eval_metrics)
if jax.process_index() == 0:
eval_metrics = {
f"eval_{metric_name}": value
for metric_name, value in eval_metrics.items()
}
path = os.path.join(training_args.output_dir, "eval_results.json")
with open(path, "w") as f:
json.dump(eval_metrics, f, indent=4, sort_keys=True)
Args:
model: The model the evaluate.
state: Model state containing state for stateful flax.nn functions, such as
batch normalization.
eval_dataset: Dataset to evaluate the model over.
Returns:
Dictionary containing the average loss and accuracy of the model on the given
dataset.
"""
p_eval_step = jax.pmap(_eval_step, axis_name='batch')
batch_sizes = []
metrics = []
for batch in eval_dataset:
batch_size = len(batch[LABELKEY])
# These are required for pmap call.
batch = _shard_batch(batch)
batch_metrics = p_eval_step(model, state, batch)
batch_sizes.append(batch_size)
metrics.append(batch_metrics)
# Note: use weighted mean, since we do mean of means with potentially
# different batch sizes otherwise.
batch_sizes = jnp.array(batch_sizes)
weights = batch_sizes / jnp.sum(batch_sizes)
eval_metrics = common_utils.get_metrics(metrics)
return jax.tree_map(lambda x: (weights * x).sum(), eval_metrics)
num_eval_samples = len(tokenized_datasets["validation"])
eval_samples_idx = jnp.arange(num_eval_samples)
eval_batch_idx = generate_batch_splits(eval_samples_idx, eval_batch_size)
eval_metrics = []
for i, batch_idx in enumerate(tqdm(eval_batch_idx, desc="Evaluating ...", position=2)):
samples = [tokenized_datasets["validation"][int(idx)] for idx in batch_idx]
model_inputs = data_collator(samples, pad_to_multiple_of=16)
# Model forward
model_inputs = shard(model_inputs.data)
metrics = p_eval_step(state.params, model_inputs)
eval_metrics.append(metrics)
# normalize eval metrics
eval_metrics = get_metrics(eval_metrics)
eval_metrics = jax.tree_map(jnp.sum, eval_metrics)
eval_normalizer = eval_metrics.pop("normalizer")
eval_metrics = jax.tree_map(lambda x: x / eval_normalizer, eval_metrics)
# Update progress bar
epochs.desc = f"Step... ({cur_step} | Loss: {eval_metrics['loss']}, Acc: {eval_metrics['accuracy']})"
# Save metrics
if has_tensorboard and jax.process_index() == 0:
write_eval_metric(summary_writer, eval_metrics, cur_step)
if cur_step % training_args.save_steps == 0 and cur_step > 0:
# save checkpoint after each epoch and push checkpoint to the hub
if jax.process_index() == 0:
params = jax.device_get(jax.tree_map(lambda x: x[0], state.params))
def train_loop(config, dropout_rngs, eval_ds, eval_freq, num_eval_steps,
num_train_steps, optimizer, state, p_eval_step, p_train_step,
start_step, train_iter, summary_writer):
"""Training loop.
Args:
config: experiment config.
dropout_rngs: float array; Jax PRNG key.
eval_ds: tf.dataset; Evaluation dataset.
eval_freq: int; Evaluation frequency;
num_eval_steps: int; Number of evaluation steps.
num_train_steps: int; Number of training steps.
optimizer: flax optimizer.
state: model state, e.g. batch statistics.
p_eval_step: fn; Pmapped evaluation step function.
p_train_step: fn; Pmapped train step function.
start_step: int; global training step.
train_iter: iter(tf.dataset); Training data iterator.
summary_writer: tensorflow summary writer.
Returns:
optimizer, global training step
"""
metrics_all = []
tick = time.time()
logging.info('Starting training')
logging.info('====================')
step = 0
for step, batch in zip(range(start_step, num_train_steps), train_iter):
batch = common_utils.shard(jax.tree_map(lambda x: x._numpy(), batch)) # pylint: disable=protected-access
optimizer, state, metrics, dropout_rngs = p_train_step(
optimizer, state, batch, dropout_rng=dropout_rngs)
metrics_all.append(metrics)
# Save a Checkpoint
if ((step % config.checkpoint_freq == 0 and step > 0)
or step == num_train_steps - 1):
if jax.host_id() == 0 and config.save_checkpoints:
# Save unreplicated optimizer + model state.
checkpoints.save_checkpoint(FLAGS.model_dir,
(jax_utils.unreplicate(optimizer),
jax_utils.unreplicate(state)),
step)
# Periodic metric handling.
if step % eval_freq == 0 and step > 0:
metrics_all = common_utils.get_metrics(metrics_all)
lr = metrics_all.pop('learning_rate').mean()
metrics_sums = jax.tree_map(jnp.sum, metrics_all)
denominator = metrics_sums.pop('denominator')
summary = jax.tree_map(lambda x: x / denominator, metrics_sums) # pylint: disable=cell-var-from-loop
summary['learning_rate'] = lr
# Calculate (clipped) perplexity after averaging log-perplexities:
logging.info('train in step: %d, loss: %.4f, acc: %.4f', step,
summary['loss'], summary['accuracy'])
if jax.host_id() == 0:
tock = time.time()
steps_per_sec = eval_freq / (tock - tick)
tick = tock
summary_writer.scalar('examples_per_second',
steps_per_sec * config.batch_size, step)
for key, val in summary.items():
summary_writer.scalar(f'train_{key}', val, step)
summary_writer.flush()
# Reset metric accumulation for next evaluation cycle.
metrics_all = []
# Eval Metrics
eval_metrics = []
eval_iter = iter(eval_ds)
if num_eval_steps == -1:
num_iter = itertools.repeat(1)
else:
num_iter = range(num_eval_steps)
for _, eval_batch in zip(num_iter, eval_iter):
# pylint: disable=protected-access
eval_batch = common_utils.shard(
jax.tree_map(lambda x: x._numpy(), eval_batch))
# pylint: enable=protected-access
metrics = p_eval_step(optimizer.target, state, eval_batch)
eval_metrics.append(metrics)
eval_metrics = common_utils.get_metrics(eval_metrics)
eval_metrics_sums = jax.tree_map(jnp.sum, eval_metrics)
eval_denominator = eval_metrics_sums.pop('denominator')
eval_summary = jax.tree_map(
lambda x: x / eval_denominator, # pylint: disable=cell-var-from-loop
eval_metrics_sums)
logging.info('eval in step: %d, loss: %.4f, acc: %.4f', step,
eval_summary['loss'], eval_summary['accuracy'])
if jax.host_id() == 0:
for key, val in eval_summary.items():
summary_writer.scalar(f'val_{key}', val, step)
summary_writer.flush()
return optimizer, state, step
