mirror of
https://github.com/storytold/storyteller-ml.git
synced 2026-10-20 08:04:35 +00:00
Brandon Thomas
79 lines
3.4 KiB
Python
79 lines
3.4 KiB
Python
# Copyright (c) 2021, NVIDIA CORPORATION. All rights reserved.
|
|
#
|
|
# 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.
|
|
|
|
import torch
|
|
from torchmetrics import Metric
|
|
|
|
__all__ = ['SequencePerplexity']
|
|
|
|
|
|
class SequencePerplexity(Metric):
|
|
"""
|
|
This class computes mean perplexity across the batches of sequences.
|
|
|
|
You have to provide ``log_probs`` (float tensor of shape [batch_size x seq_length x vocab_size]) and
|
|
``labels`` (int tensor of shape [batch_size x seq_length] with values from the range [0, vocab_size-1])
|
|
to the :meth:`update` method. If some of the sequences are shorter than seq_length, you can also provide
|
|
an optional argument ``mask`` (bool tensor of shape [batch_size x seq_length]) which masks out tokens
|
|
not participating in perplexity computation.
|
|
|
|
See :doc:`PyTorch Lightning Metrics<pytorch-lightning:metrics>` for the metric usage instructions.
|
|
|
|
Args:
|
|
compute_on_step:
|
|
Forward only calls ``update()`` and returns ``None`` if this is set to ``False``. default: ``True``
|
|
dist_sync_on_step:
|
|
Synchronize metric state across processes at each ``forward()`` before returning the value at the step.
|
|
process_group:
|
|
Specify the process group on which synchronization is called. default: ``None`` (which selects the entire
|
|
world)
|
|
dist_sync_fn:
|
|
Callback that performs the allgather operation on the metric state. When ``None``, DDP will be used
|
|
to perform the allgather.
|
|
"""
|
|
|
|
def __init__(self, compute_on_step=True, dist_sync_on_step=False, process_group=None, dist_sync_fn=None):
|
|
super().__init__(
|
|
compute_on_step=compute_on_step,
|
|
dist_sync_on_step=dist_sync_on_step,
|
|
process_group=process_group,
|
|
dist_sync_fn=dist_sync_fn,
|
|
)
|
|
|
|
# Total sum of exponentiated average negative log likelihoods
|
|
self.add_state('perplexities_sum', default=torch.tensor(0.0, dtype=torch.float64), dist_reduce_fx='sum')
|
|
# Total number of sequences in all batches
|
|
self.add_state('num_sequences', default=torch.tensor(0, dtype=torch.int64), dist_reduce_fx='sum')
|
|
|
|
def update(self, log_probs: torch.Tensor, labels: torch.Tensor, mask=None):
|
|
|
|
if mask is None:
|
|
mask = torch.ones_like(labels)
|
|
if mask.dtype is not log_probs.dtype:
|
|
mask = mask.to(log_probs.dtype)
|
|
|
|
target_log_probs = log_probs.gather(2, labels.unsqueeze(2)).squeeze(2)
|
|
avg_neg_ll = -(target_log_probs * mask).sum(dim=-1) / mask.sum(dim=-1)
|
|
ppl = avg_neg_ll.exp()
|
|
self.num_sequences += ppl.numel()
|
|
self.perplexities_sum += ppl.sum()
|
|
|
|
def compute(self):
|
|
"""
|
|
Returns perplexity across all workers and resets to 0 :attr:`perplexities_sum` and :attr:`num_sequences`.
|
|
"""
|
|
if self.num_sequences.eq(0):
|
|
return None
|
|
return self.perplexities_sum / self.num_sequences
|