def CreateSamplerProtoFromFlags() -> sampler_pb2.Sampler:
sampler = sampler_pb2.Sampler(
start_text=FLAGS.clgen_seed_text,
batch_size=FLAGS.clgen_sample_batch_size,
sequence_length=FLAGS.clgen_sample_sequence_length,
temperature_micros=int(FLAGS.clgen_sample_temperature * 1000000),
termination_criteria=[
sampler_pb2.SampleTerminationCriterion(
symtok=sampler_pb2.SymmetricalTokenDepth(
depth_increase_token="{", depth_decrease_token="}",
)
),
],
)
if FLAGS.clgen_max_sample_length:
sampler.termination_criteria.extend(
[
sampler_pb2.SampleTerminationCriterion(
maxlen=sampler_pb2.MaxTokenLength(
maximum_tokens_in_sample=FLAGS.clgen_max_sample_length,
)
),
]
)
return sampler
def abc_sampler_config():
"""The sampler config for a simple Sampler."""
maxlen = sampler_pb2.MaxTokenLength(maximum_tokens_in_sample=5)
sample_stop = [sampler_pb2.SampleTerminationCriterion(maxlen=maxlen)]
return sampler_pb2.Sampler(start_text='a', batch_size=5,
termination_criteria=sample_stop,
temperature_micros=1000000)
def MakeClgenInstanceConfig(
working_dir: pathlib.Path,
encoded_db: encoded.EncodedContentFiles,
num_training_epochs: int,
seed_text: str,
neurons_per_layer: int,
num_layers: int,
) -> clgen_pb2.Instance:
"""Construct a CLgen instance.
Args:
working_dir: The directory to cache CLgen working files in.
encoded_db: The directory of encoded content files.
num_training_epochs: The number of epochs to train for.
seed_text: The text to initiate sampling with.
neurons_per_layer: Number of neurons in a layer.
"""
return clgen_pb2.Instance(
working_dir=str(working_dir),
model=model_pb2.Model(
corpus=corpus_pb2.Corpus(pre_encoded_corpus_url=encoded_db.url, ),
architecture=model_pb2.NetworkArchitecture(
backend=model_pb2.NetworkArchitecture.TENSORFLOW,
neuron_type=model_pb2.NetworkArchitecture.LSTM,
neurons_per_layer=neurons_per_layer,
num_layers=num_layers,
post_layer_dropout_micros=0,
),
training=model_pb2.TrainingOptions(
num_epochs=num_training_epochs,
sequence_length=64,
batch_size=64,
shuffle_corpus_contentfiles_between_epochs=True,
adam_optimizer=model_pb2.AdamOptimizer(
initial_learning_rate_micros=2000,
learning_rate_decay_per_epoch_micros=50000,
beta_1_micros=900000,
beta_2_micros=999000,
normalized_gradient_clip_micros=5000000,
),
),
),
sampler=sampler_pb2.Sampler(
start_text=seed_text,
batch_size=64,
sequence_length=1024,
temperature_micros=1000000, # = 1.0 real value
termination_criteria=[
sampler_pb2.SampleTerminationCriterion(
symtok=sampler_pb2.SymmetricalTokenDepth(
depth_increase_token="{",
depth_decrease_token="}",
)),
sampler_pb2.SampleTerminationCriterion(
maxlen=sampler_pb2.MaxTokenLength(
maximum_tokens_in_sample=20000, )),
],
),
)
def test_MaxlenTerminationCriterion_SampleIsComplete():
"""Test SampleIsComplete() returns expected values."""
t = samplers.MaxlenTerminationCriterion(
sampler_pb2.MaxTokenLength(maximum_tokens_in_sample=3))
assert not t.SampleIsComplete([])
assert not t.SampleIsComplete(["a"])
assert not t.SampleIsComplete(["a", "b"])
assert t.SampleIsComplete(["a", "b", "c"])
assert t.SampleIsComplete(["a", "b", "c", "d"])
assert t.SampleIsComplete(["a", "b", "c", "d", "e"])
def test_MaxlenTerminationCriterion_SampleIsComplete():
"""Test SampleIsComplete() returns expected values."""
t = samplers.MaxlenTerminationCriterion(
sampler_pb2.MaxTokenLength(maximum_tokens_in_sample=3))
assert not t.SampleIsComplete([])
assert not t.SampleIsComplete(['a'])
assert not t.SampleIsComplete(['a', 'b'])
assert t.SampleIsComplete(['a', 'b', 'c'])
assert t.SampleIsComplete(['a', 'b', 'c', 'd'])
assert t.SampleIsComplete(['a', 'b', 'c', 'd', 'e'])
def test_MaxlenTerminationCriterion_invalid_maximum_tokens_in_sample():
"""Test that error is raised if maximum_tokens_in_sample is invalid."""
config = sampler_pb2.MaxTokenLength()
# Field is missing.
with test.Raises(errors.UserError) as e_info:
samplers.MaxlenTerminationCriterion(config)
assert "MaxTokenLength.maximum_tokens_in_sample must be > 0" == str(
e_info.value)
# Value is zero.
config.maximum_tokens_in_sample = 0
with test.Raises(errors.UserError) as e_info:
samplers.MaxlenTerminationCriterion(config)
assert "MaxTokenLength.maximum_tokens_in_sample must be > 0" == str(
e_info.value)
def Train(self, *args, **kwargs) -> None:
with self.Session():
test_sampler_config = sampler_pb2.Sampler()
test_sampler_config.CopyFrom(self.sampler.config)
# Make all test samples the same 512-token length.
del test_sampler_config.termination_criteria[:]
test_sampler_config.termination_criteria.extend([
sampler_pb2.SampleTerminationCriterion(
maxlen=sampler_pb2.MaxTokenLength(
maximum_tokens_in_sample=512)),
])
test_sampler = samplers.Sampler(test_sampler_config)
# We inject the `test_sampler` argument so that we can create samples
# during training.
self.model.Train(*args, test_sampler=test_sampler, **kwargs)
def main(argv: typing.List[str]):
"""Main entry point."""
if len(argv) > 1:
raise app.UsageError("Unknown arguments: '{}'.".format(" ".join(
argv[1:])))
instance = clgen.Instance(
clgen_pb2.Instance(
working_dir=FLAGS.clgen_dir,
model=model_pb2.Model(
corpus=corpus_pb2.Corpus(
local_directory=FLAGS.clgen_corpus_dir,
ascii_character_atomizer=True,
preprocessor=[
"deeplearning.clgen.preprocessors.opencl:ClangPreprocessWithShim",
"deeplearning.clgen.preprocessors.opencl:Compile",
"deeplearning.clgen.preprocessors.opencl:NormalizeIdentifiers",
"deeplearning.clgen.preprocessors.opencl:StripDoubleUnderscorePrefixes",
"deeplearning.clgen.preprocessors.common:StripDuplicateEmptyLines",
"deeplearning.clgen.preprocessors.opencl:SanitizeKernelPrototype",
"deeplearning.clgen.preprocessors.common:StripTrailingWhitespace",
"deeplearning.clgen.preprocessors.opencl:ClangFormat",
"deeplearning.clgen.preprocessors.common:MinimumLineCount3",
"deeplearning.clgen.preprocessors.opencl:Compile",
],
contentfile_separator="\n\n",
),
architecture=model_pb2.NetworkArchitecture(
backend=model_pb2.NetworkArchitecture.TENSORFLOW,
neuron_type=model_pb2.NetworkArchitecture.LSTM,
neurons_per_layer=512,
num_layers=2,
post_layer_dropout_micros=0,
),
training=model_pb2.TrainingOptions(
num_epochs=50,
sequence_length=64,
batch_size=64,
shuffle_corpus_contentfiles_between_epochs=True,
adam_optimizer=model_pb2.AdamOptimizer(
initial_learning_rate_micros=2000,
learning_rate_decay_per_epoch_micros=50000,
beta_1_micros=900000,
beta_2_micros=999000,
normalized_gradient_clip_micros=5000000,
),
),
),
sampler=sampler_pb2.Sampler(
start_text="kernel void ",
batch_size=64,
sequence_length=1024,
temperature_micros=1000000, # = 1.0 real value
termination_criteria=[
sampler_pb2.SampleTerminationCriterion(
symtok=sampler_pb2.SymmetricalTokenDepth(
depth_increase_token="{",
depth_decrease_token="}",
)),
sampler_pb2.SampleTerminationCriterion(
maxlen=sampler_pb2.MaxTokenLength(
maximum_tokens_in_sample=20000, )),
],
),
), )
db = grewe_features_db.Database(FLAGS.db)
profile_dir = pathlib.Path(FLAGS.profile_dir)
profile_dir.mkdir(parents=True, exist_ok=True)
profiler = prof.AutoCsvProfiler(profile_dir)
with instance.Session(), multiprocessing.Pool() as pool:
while True:
Sample(instance, db, profiler, pool)