9e73151d fix: don't panic when an unknown service is encountered (1118)
a686a376 docs: various fixes for tf.keras docs. (1119)
08338f63 fix: drop null values on trial metrics (1120)
44fdf150 docs: add docs for behavior of keras's on_epoch_end callback (1105)
4be2c4af chore: publish trial details page [DET-3843] (1103)
d1114f32 fix: add explicit searcher sync before close trial [DET-3889] (1101)
b3419939 refactor: cleanup trial detail [DET-3683] (1107)
ae271c42 chore: add a shell-entrypoint.sh (1102)
c1633641 fix: address scrolling issue when scrolling up after hitting the back to top button (1108)
6314b624 style: fix table cell styles when there is only one cell (1115)
b9cb988a feat: make sharding configurable in wrap_dataset (1041)
7e129f42 feat: kill idle tensorboards [DET-3808] (1104)
79fdaeab fix: handle failed stack deletion (1109)
f2a50fff fix: actually kill trials (1111)
7a30cdbf feat: add logging to agent to track usage [DET-3878] (1110)
2364a065 fix: let shell proxying work with master TLS [DET-3882] (1100)
bde16413 feat: improve shell cli for optional args (1098)
2318ea0e fix: fix cluster resource pie chart color mapping [DET-3861] (1094)
cc9a7c1f feat: trial table [DET-3019] (1078)
c0b3fe52 chore: add kill trial endpoint [DET-3739] (1071)
48c62308 feat: allow the harness to connect to the master over TLS [DET-3775] (1076)
3fd156f1 terminating state now unaffected if transitioning to itself (1084)
0ab58b59 fix: handle when config is undefined for trial length (1096)
b3c7b8f7 fix: support shell with host networking mode (1099)
69515e7f fix: run at least one batch per searcher op (1095)
979d3528 fix: proxy shell through master via HTTP CONNECT (1067)
1e12e12c chore: update archive/unarchive endpoint to the new api (1093)
065c1b07 chore: migrate continue trial form and remove steps [DET-3780 DET-3854] (1050)
5093a2c5 feat: add total_batches_processed to /trials/{id}/details endpoint (1089)
0b3b7e8b fix: record progress in partial units (1091)
1b6c884f chore: add proxy urls to tensorboards and notebooks [DET-3736] (1070)
ef223bdd fix: support tensorboard in us-east-1 (1092) [DET-3842]
9db99369 fix: hide table pagination when unnecessary (1088)
f978a326 feat: repr methods for python classes [DET-3750] (1068)
76f2299f fix: update `num_batches` to `total_batches_processed` (1087)
7b9976d0 chore: bump version: 0.12.13.dev0 -> 0.13.0.dev0 (1033)
d0b384ab fix: handle failed checkpoints correctly in trial_workload_sequencer [DET-3853] (1083)
b30a8add feat: support user specified pod template specs [DET-3716, DET-3731] (1005)
e6ee088c feat: add trial detail chart [DET-3013] (1063)
26730b11 chore: unpin overly specific versions (1085)
68604fb7 docs: minor fixes for data access tutorial (1082)
3aac0669 fix: fix typo in error message (1081)
5351570e chore: delete useless test (1080)
59827519 fix: update logout check to check via endpoint call (1062)
701095e8 chore: extract training validation metrics (1073)
933f7f78 feat: hide experiment max slots if it's unset (1075)
45109ab9 fix: preserve rng state for all cloud access (1079)
3485d82e feat: add ability to perform initial validation [DET-3584] (1061)
f939b1a9 test: only send slack messages for master branch CI (1077)
1721ea87 docs: add two new FAQs. (1072)
ea939dbe feat: add gaea nas [DET-3553] (824)
39726d68 fix: handle failed validations correctly [DET-3838] (1069)
a5f325fe feat: enable experiment detail [DET-3788] (1065)
5ec42fe0 test: make nonroot tests stricter (1060)
038e241b fix: trigger the correct click handler (1064)
740c61e5 chore: infobox style update and refactor [DET-3796] (1058)
6751572b fix: route experiment list links to react (1059)
f6be408f feat: get checkpoints for trial endpoint [DET-3735] (1052)
f25d1e14 feat: add archive and unarchive endpoints [DET-3582] (1037)
1fd05a60 build: adjust resource deletion order [DET-3793] (1056)
379f04ca fix: update regex to handle newlines in the log messages (1053)
64393f70 feat: get checkpoints for experiment endpoint (1046)
8713952a docs: apply our own best practices in tutorials (1054)
e1520c62 docs: tweak index page text (1055)
2b0302ee feat: make agent's TLS verification level configurable [DET-3774] (1049)
b47cf66e chore: refactor API error handler to return the error. (1031)
59339f86 chore: minor fixes for PyTorch GAN example. (1047)
57f27dfd test: handle concurrent cloudwatch CI uploads (1048)
87dd08eb chore: make metric_maker test quieter (1045)
9080ff96 fix: don't ignore SHA operations from promotion of exited trials [DET-3546] (1022)
90d3b803 feat: add tensorboard sources to task list [DET-3760] (1014)
5d9b4f5d fix: fix string formatting in k8 rp (1042)
45034bda feat: enable list pages [DET-3705] (992)
1f92f9fa chore: change search_metric to val_acc for tf_keras_native_dtrain 4bb2ab6 [DET-3794] (1044)
ab17de66 feat: get checkpoint endpoint [DET-3734] (1040)
ca18eb5f chore: add nil check for validation metrics (1043)
cfe24db0 docs: various fixes. (1038)
c91614f3 refactor: convert io.null and io.undefined to ioNullOrUndefined [DET-3747] (1036)
2ea28f2c chore: migrate trial Infobox [DET-3014 DET-3016] (981)
9f1afa94 fix: dismiss trial log download warning modal upon confirmation (1025)
bdee966b fix: filter out disabled slots for reporting (1030)
eae0747d fix: show archive / unarchive options under task actions properly (1020)
fd6410ac docs: release notes for 0.12.13 (1027) (1032)
0b58bba8 chore: fix tensorpack nightly tests (1029)
81338d1c chore: silence protobuf warnings. (1028)
398cf667 feat: add experiment cancel and kill endpoints [DET-3581] (1026)
7767524e feat: add master logs endpoint [DET-3680] (1007)
920025c8 feat: launch notebook [DET-3741] (991)
e5c975f7 feat: remove steps from the UX (968)
2b88fa71 style: refine tables (993)
049a3c17 style: polish experiment detail [DET-3517, DET-3621] (974)
b68d83d6 docs: tweak intro to model registry tutorial (1023)
6f3c7658 fix: imagenet_gaea merge and rebase conflicts [DET-3769] (1021)
4f2d35a9 chore: refactor experiment creation modal and add trial continue workflow (971)
6978480a test: disable imagenet_nas_arch_pytorch for release [DET-3770] (1018)
c71cfbcd docs: promote PyTorch flexible primitives migration guide [DET-3771] (1019)
Docker images
- `docker pull determinedai/determined-master:0.13.0`
- `docker pull determinedai/determined-master:5ea6d3b9`
- `docker pull determinedai/determined-master:5ea6d3b9fa3d982df30f84582ed0c4dc3d36ddb0`
- `docker pull determinedai/determined-dev:determined-master-5ea6d3b9`
- `docker pull determinedai/determined-dev:determined-master-5ea6d3b9fa3d982df30f84582ed0c4dc3d36ddb0`