Commits

Branch selector

 master

User selector

jimlinntu

All time

Commit History

Commits on May 8, 2024

• #tf-data Restores the seed_generator_ and three seeds in global_shuffle_dataset_op.cc to ensure the iterator can generate the same result after restoring

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedMay 8, 2024
  b0def2a

Commits on May 6, 2024

• Turns off .weighted_flat_map random access compatibility because it does not work with any other ops that require the parent index mapper to be stateless (like .batch for example)

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedMay 6, 2024
  c830ed7

Commits on May 3, 2024

• Accounts for ParallelMap's initial parallelism maximum buffered bytes

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedMay 3, 2024
  1d135df

Commits on Apr 26, 2024

• Reverts 92304fe

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedApr 26, 2024
  9e89e64

Commits on Apr 22, 2024

• #tf-data Adds random access support to the zip dataset

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedApr 22, 2024
  341ed26

Commits on Apr 18, 2024

• #tf-data Optimizes the parameters only after the first root GetNext is finished.

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedApr 18, 2024
  92304fe

Commits on Apr 11, 2024

• #tf-data returns bool instead of int64_t for PrefetchAutotuner::HasElementSize

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedApr 11, 2024
  3cf3e19

Commits on Apr 9, 2024

• #tf-data copies is_legacy_prefetch_autotuned_ when snapshotting

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedApr 9, 2024
  0d18d3a

Commits on Apr 2, 2024

• #tf-data makes data_service_client.cc nonblocking when getting elements from workers

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedApr 2, 2024
  6396a03

Commits on Mar 7, 2024

• #tensorflow uses logging.error instead of warning.warn in tf.InteractiveSession to redirect users to use tf.Session in production.

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedMar 7, 2024
  198ada1

Commits on Feb 29, 2024

• No public description

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedFeb 29, 2024
  e3eda23

Commits on Feb 26, 2024

• #tf-data allow access to the model proto from an iterator

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedFeb 26, 2024
  004573c

Commits on Feb 21, 2024

• #tf-data fix ParallelBatch reporting negative buffered_bytes in its Node

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedFeb 21, 2024
  1162286

Commits on Jan 26, 2024

• #tf-data refactor _Retval

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedJan 26, 2024
  aac7015

Commits on Jan 25, 2024

• #tf-data add .fingerprint() API to tf.data.Dataset

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedJan 25, 2024
  79cddf8

Commits on Jan 23, 2024

• #tf-data improve rebatch throughput by ~2.75x (when the data are movable) via avoiding copying

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedJan 23, 2024
  fe1436c

• #tf-data optimize interleave symbolic checkpointing by avoiding saving input

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedJan 23, 2024
  29362bf

Commits on Jan 22, 2024

• #tf-data improve rebatch throughput by ~2.75x (when the data are movable) via avoiding copying

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedJan 22, 2024
  84cfa9b

Commits on Jan 12, 2024

• #tf-data improve rebatch throughput by ~2.75x (when the data are movable) via avoiding copying

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedJan 12, 2024
  1b7a667

Commits on Jan 11, 2024

• #tf-data tests when batch and rebatch have same batch sizes

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedJan 11, 2024
  b70f425

Commits on Jan 9, 2024

• No public description

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedJan 9, 2024
  04857fc

Commits on Dec 12, 2023

• #tf-data fix symbolic checkpointing behavior when d.flat_map().skip() is used

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedDec 12, 2023
  b6c7b05

Commits on Dec 7, 2023

• #tf-data improve group_by_window_dataset_op reduce_func's error message

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedDec 7, 2023
  8d1c5e1

Commits on Dec 6, 2023

• No public description

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedDec 6, 2023
  720e574

Commits on Dec 4, 2023

• #tf-data remove the cap and directly use double-typed gauge cell instead of int64-typed

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedDec 4, 2023
  fffb030

Commits on Nov 30, 2023

• #tensorflow enable double-typed gauge cell

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedNov 30, 2023
  77ffe8b

• No public description

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedNov 30, 2023
  2b918e2

Commits on Nov 10, 2023

• #tf-data export the slowest stage processing time of the input pipeline as /tensorflow/data/pipeline_processing_time{id=...}

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedNov 10, 2023
  5bc1b85

Commits on Oct 27, 2023

• No public description

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedOct 27, 2023
  cef6a27

Commits on Oct 23, 2023

• #tf-data destroys element before the mutex_lock l(*mu) is destroyed.

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedOct 23, 2023
  d5d0dee

Commits on Oct 18, 2023

• #tf-data internal-only change

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedOct 18, 2023
  e18f6bc

Commits on Oct 16, 2023

• #tf-data make expired() and node->output() atomic

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedOct 16, 2023
  c7c003e

Commits on Oct 11, 2023

• #tf-data fix the ram budget calculation from share * AvailableRam() + total_buffered_bytes to share * (AvailableRam() + total_buffered_bytes)

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedOct 11, 2023
  2be7e21

Commits on Oct 6, 2023

• No public description

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedOct 6, 2023
  478ef7c

Commits on Oct 4, 2023

• #tf-data internal-only change

  

  jimlinntu

  authored and

  tensorflower-gardener

  committedOct 4, 2023
  7c9f10e

Pagination

PreviousNext