- 31 10月, 2018 1 次提交
-
-
由 Taylor Robie 提交于
* Keras-ify TPU embedding lookup * delint * pull get_variable() out of keras lambda * delint * move get_variable under variable scope
-
- 30 10月, 2018 2 次提交
-
-
由 Tayo Oguntebi 提交于
* Merges TPU-TC optimizations into HEAD. * Split a line that went over 80 from a tab. * Remove trailing whitespace.
-
由 Reed 提交于
The option is --nouse_estimator
-
- 27 10月, 2018 1 次提交
-
-
由 Reed 提交于
--ml_perf now just changes the model to make it MLPerf compliant. --output_ml_perf_compliance_logging adds the MLPerf compliance logs.
-
- 26 10月, 2018 3 次提交
-
-
由 Taylor Robie 提交于
prevent async process from writing alive file until the main process has created the cache root (#5614)
-
由 josh11b 提交于
AllReduceCrossTowerOps -> AllReduceCrossDeviceOps
-
由 Reed 提交于
The error message was: absl.flags._exceptions.IllegalFlagValueError: flag --ml_perf=None: ('Non-boolean argument to boolean flag', 'None')
-
- 25 10月, 2018 5 次提交
-
-
由 josh11b 提交于
-
由 Taylor Robie 提交于
* move version check to a function * delint * tweak pip check * delint
-
由 josh11b 提交于
To match new terminology in DistributionStrategy.
-
由 josh11b 提交于
-
由 Taylor Robie 提交于
* first pass at __getattr__ abuse logger * first pass at adding tags to NCF * minor formatting updates * fix tag name * convert metrics to python floats * getting closer... * direct mlperf logs to a file * small tweaks and add stitching * update tags * fix tag and add a sudo call * tweak format of run.sh * delint * use distribution strategies for evaluation * address PR comments * delint and fix test * adjust flag validation for xla * add prefix to distinguish log stitching * fix index bug * fix clear cache for root user * dockerize cache drop * TIL some regex magic
-
- 20 10月, 2018 1 次提交
-
-
由 Reed 提交于
-
- 19 10月, 2018 3 次提交
-
-
由 Taylor Robie 提交于
-
由 Taylor Robie 提交于
* intermediate commit finish replacing spillover with resampled padding intermediate commit * resolve merge conflict * intermediate commit * further consolidate the data pipeline * complete first pass at data pipeline refactor * remove some leftover code * fix test * remove resampling, and move train padding logic into neumf.py * small tweaks * fix weight bug * address PR comments * fix dict zip. (Reed led me astray) * delint * make data test deterministic and delint * Reed didn't lead me astray. I just can't read. * more delinting * even more delinting * use resampling for last batch padding * pad last batch with unique data * Revert "pad last batch with unique data" This reverts commit cbdf46efcd5c7907038a24105b88d38e7f1d6da2. * move padded batch to the beginning * delint * fix step check for synthetic data
-
由 Toby Boyd 提交于
Add perf args
-
- 18 10月, 2018 6 次提交
-
-
由 josh11b 提交于
Drop references to `is_single_tower`.
-
由 josh11b 提交于
Since we plan on deleting this method, it is only used in distribution_utils_test.py.
-
由 shizhiw 提交于
Refactor neumf_model.py to support users who just need top_k and ndcg tensors.
-
由 Shawn Wang 提交于
-
由 Shawn Wang 提交于
-
由 Shawn Wang 提交于
-
- 17 10月, 2018 18 次提交
-
-
由 Chris Shallue 提交于
Improvements to AstroNet and add AstroWaveNet
-
由 Chris Shallue 提交于
PiperOrigin-RevId: 217341274
-
由 Chris Shallue 提交于
PiperOrigin-RevId: 215014489
-
由 Alex Tamkin 提交于
Add AstroWaveNet model, a generative model of astronomical light curves. AstroWaveNet's hidden states can be used in a semi-supervised fashion for downstream tasks like finding exoplanets. PiperOrigin-RevId: 214073725
-
由 Alex Tamkin 提交于
PiperOrigin-RevId: 214069349
-
由 Chris Shallue 提交于
PiperOrigin-RevId: 213963765
-
由 Chris Shallue 提交于
PiperOrigin-RevId: 213353962
-
由 Chris Shallue 提交于
PiperOrigin-RevId: 212909744
-
由 Chris Shallue 提交于
Factor out evaluate(), continuous_eval(), and continuous_train_and_eval() from estimator_util.py into estimator_runner.py. PiperOrigin-RevId: 212903406
-
由 Chris Shallue 提交于
PiperOrigin-RevId: 211832751
-
由 Chris Shallue 提交于
1. reshard_arrays(xs, ys): Reshards arrays in xs to match the lengths of arrays in ys. 2. uniform_cadence_light_curve(): Combines data into a single light curve with uniform cadence numbers. PiperOrigin-RevId: 211724321
-
由 Chris Shallue 提交于
Remove the constant LONG_CADENCE_TIME_DELTA_DAYS as this is not strictly constant between light curves (although it is very close). PiperOrigin-RevId: 211153560
-
由 Chris Shallue 提交于
This applies mainly to scrambled data (NaN time values typically come with NaN flux values, which are removed anyway, but scrambing decouples NaN time values from NaN flux values). PiperOrigin-RevId: 209029696
-
由 Alex Tamkin 提交于
Update configurations with label mapping for simulated data and new batch_size and learning_rate from vizier studies. PiperOrigin-RevId: 208902383
-
由 Alex Tamkin 提交于
PiperOrigin-RevId: 208863882
-
由 Alex Tamkin 提交于
PiperOrigin-RevId: 208862798
-
由 Alex Tamkin 提交于
Ensure consistent argument order, change NaN processing to remove NaN times as well, and do so before scrambling PiperOrigin-RevId: 207804986
-
由 Alex Tamkin 提交于
Modify processing pipeline to enable generation of scrambled lightcurves. Fix bugs to enable generation of inverted lightcurves. PiperOrigin-RevId: 207595688
-