↰ Return to documentation for file (morpheus/_lib/src/stages/preprocess_nlp.cpp)
/*
* SPDX-FileCopyrightText: Copyright (c) 2021-2023, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "morpheus/stages/preprocess_nlp.hpp"
#include "mrc/node/rx_sink_base.hpp"
#include "mrc/node/rx_source_base.hpp"
#include "mrc/node/sink_properties.hpp"
#include "mrc/node/source_properties.hpp"
#include "mrc/segment/object.hpp"
#include "mrc/types.hpp"
#include "morpheus/messages/memory/inference_memory.hpp" // for InferenceMemory
#include "morpheus/messages/multi_inference.hpp"
#include "morpheus/objects/dev_mem_info.hpp"
#include "morpheus/objects/dtype.hpp"
#include "morpheus/objects/table_info.hpp" // for TableInfo
#include "morpheus/objects/tensor.hpp"
#include "morpheus/types.hpp" // for TensorIndex, TensorMap
#include "morpheus/utilities/matx_util.hpp"
#include <cudf/column/column.hpp> // for column, column::contents
#include <cudf/column/column_factories.hpp>
#include <cudf/column/column_view.hpp>
#include <cudf/filling.hpp>
#include <cudf/reshape.hpp>
#include <cudf/scalar/scalar.hpp>
#include <cudf/strings/strings_column_view.hpp> // for strings_column_view
#include <cudf/table/table_view.hpp>
#include <cudf/types.hpp>
#include <cudf/unary.hpp>
#include <mrc/segment/builder.hpp>
#include <nvtext/normalize.hpp>
#include <nvtext/subword_tokenize.hpp>
#include <pymrc/node.hpp>
#include <rmm/cuda_stream_view.hpp>
#include <rmm/device_buffer.hpp> // for device_buffer
#include <cstdint>
#include <exception>
#include <functional>
#include <map>
#include <memory>
#include <utility>
namespace morpheus {
// Component public implementations
// ************ PreprocessNLPStage ************************* //
PreprocessNLPStage::PreprocessNLPStage(std::string vocab_hash_file,
uint32_t sequence_length,
bool truncation,
bool do_lower_case,
bool add_special_token,
int stride,
std::string column) :
PythonNode(base_t::op_factory_from_sub_fn(build_operator())),
m_vocab_hash_file(std::move(vocab_hash_file)),
m_sequence_length(sequence_length),
m_truncation(truncation),
m_do_lower_case(do_lower_case),
m_add_special_token(add_special_token),
m_stride(stride),
m_column(std::move(column))
{}
PreprocessNLPStage::subscribe_fn_t PreprocessNLPStage::build_operator()
{
return [this](rxcpp::observable<sink_type_t> input, rxcpp::subscriber<source_type_t> output) {
uint32_t stride = m_stride;
// Auto calc stride to be 75% of sequence length
if (stride < 0)
{
stride = m_sequence_length / 2;
stride = stride + stride / 2;
}
return input.subscribe(rxcpp::make_observer<sink_type_t>(
[this, &output, stride](sink_type_t x) {
// Convert to string view
auto meta = x->get_meta(this->m_column);
auto col = meta.get_column(0);
auto string_col = cudf::strings_column_view{col};
// Create the hashed vocab
thread_local std::unique_ptr<nvtext::hashed_vocabulary> vocab =
nvtext::load_vocabulary_file(this->m_vocab_hash_file);
// remove leading and trailing whitespace
auto normalized_col = nvtext::normalize_spaces(string_col);
auto normalized_col_view = cudf::strings_column_view{normalized_col->view()};
// Perform the tokenizer
nvtext::tokenizer_result token_results;
if (normalized_col_view.chars_size() > 0)
{
token_results = nvtext::subword_tokenize(normalized_col_view,
*vocab,
this->m_sequence_length,
stride,
this->m_do_lower_case,
this->m_truncation,
normalized_col_view.size() * 2);
}
else
{
// workaround for a situation where the input strings contain either no characters or only
// whitespace
auto zero = cudf::numeric_scalar<uint32_t>(0, true, rmm::cuda_stream_default);
auto ids =
cudf::make_column_from_scalar(zero, this->m_sequence_length * normalized_col_view.size());
auto mask =
cudf::make_column_from_scalar(zero, this->m_sequence_length * normalized_col_view.size());
auto metadata = [&]() {
auto iota = cudf::sequence(normalized_col_view.size(), zero);
auto zeroes = cudf::make_column_from_scalar(zero, normalized_col_view.size());
return cudf::interleave_columns(cudf::table_view{
std::vector<cudf::column_view>{iota->view(), zeroes->view(), zeroes->view()}});
}();
token_results = nvtext::tokenizer_result{static_cast<uint32_t>(normalized_col_view.size()),
this->m_sequence_length,
std::move(ids),
std::move(mask),
std::move(metadata)};
}
// Build the results
auto memory = std::make_shared<InferenceMemory>(token_results.nrows_tensor);
TensorIndex length = token_results.tensor_token_ids->size() / token_results.sequence_length;
auto input_ids_released =
cudf::cast(token_results.tensor_token_ids->view(), cudf::data_type(cudf::type_id::INT32))
->release();
memory->set_tensor("input_ids",
Tensor::create(std::move(input_ids_released.data),
DType::create<int32_t>(),
{length, static_cast<TensorIndex>(token_results.sequence_length)},
{},
0));
length = token_results.tensor_attention_mask->size() / token_results.sequence_length;
auto input_mask_released =
cudf::cast(token_results.tensor_attention_mask->view(), cudf::data_type(cudf::type_id::INT32))
->release();
memory->set_tensor("input_mask",
Tensor::create(std::move(input_mask_released.data),
DType::create<int32_t>(),
{length, static_cast<TensorIndex>(token_results.sequence_length)},
{},
0));
auto tensor_index_dtype = DType::create<TensorIndex>();
length = token_results.tensor_metadata->size() / 3;
auto seq_ids_released = cudf::cast(token_results.tensor_metadata->view(),
cudf::data_type(tensor_index_dtype.cudf_type_id()))
->release();
std::shared_ptr<rmm::device_buffer> seq_ids_data = std::move(seq_ids_released.data);
if (x->mess_offset > 0)
{
// Add an offset to the seq_ids so the message IDs line up
MatxUtil::offset_seq_ids(
DevMemInfo{seq_ids_data, tensor_index_dtype.type_id(), {length, 3}, {1, 3}}, x->mess_offset);
}
memory->set_tensor("seq_ids", Tensor::create(seq_ids_data, tensor_index_dtype, {length, 3}, {}, 0));
auto next = std::make_shared<MultiInferenceMessage>(
x->meta, x->mess_offset, x->mess_count, std::move(memory), 0, memory->count);
output.on_next(std::move(next));
},
[&](std::exception_ptr error_ptr) { output.on_error(error_ptr); },
[&]() { output.on_completed(); }));
};
}
// ************ PreprocessNLPStageInterfaceProxy *********** //
std::shared_ptr<mrc::segment::Object<PreprocessNLPStage>> PreprocessNLPStageInterfaceProxy::init(
mrc::segment::Builder& builder,
const std::string& name,
std::string vocab_hash_file,
uint32_t sequence_length,
bool truncation,
bool do_lower_case,
bool add_special_token,
int stride,
std::string column)
{
auto stage = builder.construct_object<PreprocessNLPStage>(
name, vocab_hash_file, sequence_length, truncation, do_lower_case, add_special_token, stride, column);
return stage;
}
} // namespace morpheus