Spaces:
Running
Running
File size: 3,372 Bytes
5f923cd | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 | // Copyright 2025 The ODML Authors.
//
// 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 "runtime/framework/resource_management/utils/resource_manager_utils.h"
#include <algorithm>
#include <cstddef>
#include <iterator>
#include <vector>
#include "absl/status/status.h" // from @com_google_absl
#include "runtime/util/status_macros.h" // NOLINT
namespace litert::lm {
namespace {
// Returns the comparable length of the vector, excluding the negative tokens,
// which indicates the multi modal token and just a placeholder. From token id
// level, there is no way to tell if the multi modal token is matched or not, so
// we assume the multi modal token are not matched. If all tokens are not
// negative, return vec.size().
size_t ComparableLengthExcludingNegativeTokens(const std::vector<int>& vec) {
for (size_t i = 0; i < vec.size(); ++i) {
if (vec[i] < 0) {
return i;
}
}
return vec.size();
}
} // namespace
absl::Status RemoveMatchingTokens(const std::vector<int> &processed_tokens,
std::vector<int> *input_ids, int *time_step) {
RET_CHECK_NE(input_ids, nullptr) << "input_ids is null.";
RET_CHECK_NE(time_step, nullptr) << "time_step is null.";
RET_CHECK_GE(*time_step, 0) << "Time step is negative.";
RET_CHECK_GE(processed_tokens.size(), *time_step)
<< "The processed tokens size is smaller than the time step.";
// Determine the number of tokens available in processed_tokens from the
// current time_step. *time_step is int, .size() is size_t. RET_CHECKs ensure
// *time_step is valid.
const size_t processed_tokens_comparable_len =
processed_tokens.size() - static_cast<size_t>(*time_step);
// Determine how many elements to actually compare (the minimum of the two
// effective sequence lengths).
const size_t comparison_len =
std::min(processed_tokens_comparable_len,
ComparableLengthExcludingNegativeTokens(*input_ids));
// Find the first mismatch.
// The first range for mismatch is [input_ids->begin(), input_ids->begin() +
// comparison_len). The second range for mismatch is [processed_tokens.begin()
// + *time_step, processed_tokens.begin() + *time_step + comparison_len).
auto mismatch_it =
std::mismatch(input_ids->begin(), input_ids->begin() + comparison_len,
processed_tokens.begin() + *time_step);
// Calculate the number of matching tokens.
// mismatch_it.first is an iterator to the first differing element in
// input_ids, or input_ids->begin() + comparison_len if all compared elements
// match.
int matching_tokens = std::distance(input_ids->begin(), mismatch_it.first);
// Update the input_ids and time_step.
input_ids->erase(input_ids->begin(), mismatch_it.first);
*time_step += matching_tokens;
return absl::OkStatus();
};
} // namespace litert::lm
|