99a848fe feat: support configuring priority scheduler in det-deploy [DET-4508] (1682)
90548653 feat: read logs from elasticsearch [DET-4621] (1637)
7da12666 fix: let shells work through an HTTP proxy/load balancer [DET-4469] (1677)
c4b2df06 feat: expose some Fluent Bit logs in the agent (1680)
7106bef2 fix: flush searcher events more often (1673) [DET-4644]
cc351eff fix: template merging for bind mounts [DET-4630] (1678)
63e13a4e feat: webui trial logs improvements [DET-4228, DET-4480, DET-4481, DET-4482, DET-4483, DET-4594] (1650)
f4a3afd4 feat: display task priorities in CLI [DET-4515, DET-4516] (1639)
7ed3d9c9 feat: make priority scheduler configurable and add docs [DET-4641] (1667)
33a1885f fix: limit concurrent restores to avoid resource exhaustion [DET-4556] (1666)
a2a807e9 feat: allow task requests to receive a resource pool field [DET-4342] (1600)
912c3756 chore: Improvements to metrics streaming endpoints [DET-4532] (1664)
85fee6a2 docs: update k8s limitations [DET-4582] (1671)
6bc6bd69 docs: reorganize the documentation contents (1663)
155a08bc fix: allow dots in map config keys (1665)
64550ef9 chore: Work around possible bug when PyTorch opens checkpoint files [DET-4614] (1661)
3c276f09 ci: add taiko video plugin (1632)
f1dbf701 chore: migrate fork/continue experiment endpoint [DET-4023] (1659)
ab6bdfb4 test: integration test trial logs API (1456)
6e0dae15 chore: Ensuring examples always call .contiguous() before .view() [DET-4613] (1653)
d05e222b feat: priority scheduler with preemption [DET-4512] (1634)
2e5f4afb8 chore: migrate wait page to react [DET-4521] (1559)
7aecb058 test: add responsive navbar test [DET-4603] (1656)
4ce46236 test: avoid swallowing errors in taiko (1657)
466a2264 test: reduce state sharing between e2e tests [DET-4487] (1543)
764e969e build: set server address for react preview build. (1654)
d7fb07bf feat: allow default password in kubernetes [DET-4435] (1624)
80187b9b feat: support elasticsearch as a trial logging backend [DET-4179] (1542)
0929c8a7 refactor: update filters to dynamically collapse when needed [DET-4584] (1629)
8f011f36 fix: don't replay duplicates in master restart [DET-4525, DET-4599] (1655)
24428f63 feat: add telemetry for schedulers [DET-4517] (1645)
ef4e9aef chore: extend on complete hook for overflow actions (1651)
93007ed2 chore: webui migrate ActiveExperiemnt context polling to new API [DET-4068] (1560)
d94824e6 refactor: add abort api calls [DET-4585] (1630)
e7321ee8 docs: fix rest-api side menu links [DET-4598] (1640)
03228072 ci: run some E2E AWS CI tests with the master using TLS [DET-4606] (1529)
78ec837f feat: support validation_steps in configure_fit() [DET-4529] (1649)
68b7b081 feat: support max zero slot containers for resource pools [DET-4309,DET-4340] (1507)
7cbd5dad docs: fixes for EKS docs. (1636)
dd89200d fix: limit --local --test mode to 1 gpu [DET-4602] (1648)
f0eba7d4 chore: lock api state for backward compatibility check (1644)
c26de486 chore: bump version: 0.13.8.dev0 -> 0.13.9.dev0 (1641)
1ad40425 chore: fix fluent lua filter (1642)
00d38b91 fix: fix wait page url for notebook launch request [DET-4586] (1631)
fe901d56 fix: track best validation metric for darts_cnn hp search benchmark (1638)
1d58515f feat: add documentation for AWS custom tags (1621)
4dcea3e3 chore: Checking for more instances of EagerTensors [DET-4566] (1633)
a77657d0 fix: unets_tf_keras example data download (1635)
1210b771 feat: return resource_pool in agent GET APIs [DET-4567] (1616)
422b900a chore: Adding NVIDIA Tesla A100 details for GCE [DET-4583] (1628)
e3c240af style: responsive webui [DET-4417, DET-4420] (1501)
2166ad79 test: fix CircleCI timing-based splitting for E2E tests (1622)
be1ed46a chore: fix keras validation for dtrain (1626)
8384e5c3 chore: increase timeout for metrics stream tests that are flaky on CI infra [DET-4581] (1627)
fb86450f ci: add gauge taiko [DET-4576] (1619)
4855079d feat: Add custom tags to AWSClusterConfig (1561)
2e2c121c fix: clean up data handling in TFKerasTrial (1564)
74d41962 feat: support per command shmSize settings [DET-4577] (1620)
46017942 fix: tensorboard to load from experiment list via table batch (1617)
4e7e7383 ci: update task and authentication tests to be more reliable (1614)
b323c24a Fix typo (1611)
bff6f277 feat: ALBERT on SQuAD 2.0 example (1609)
d63897fd chore: include offset in trial log IDs returned to webui [DET-4561] (1608)
e1e514a5 fix: webui unarchive button loading state [DET-4017] (1594)
7509e1ae chore: webui migrate killTask to new API [DET-4019] (1589)
8d1eb2d6 docs: Release notes for 0.13.8. (1603)
4fbb6e87 fix: honor DET_MASTER_CERT_NAME with shells (1604)
43f4901d chore: update CODEOWNERS to be opt-in (1595)
eb45feed chore: add endpoint to stream trial log fields [DET-4479] (1537)
769a5d83 chore: bumpenvs (1599)
55c6ccb7 fix: allow clients to override the expected master cert address [DET-4547] (1588)
da38010b feat: change the priority scheduler to round robin scheduler [DET-4514] (1596)
6a37a5f9 chore: bumpenvs for NCCL update (1593)
ec9abc31 feat: implement API endpoint for sampling streams of metrics from the best trials [DET-4441] (1571)
d8f536ee feat: InvalidHP Searcher Ability [DET-4333] (1550)
8b6ab794 feat: propagate task priorities to resource pools [DET-4510] (1577)
e8a6784e chore: Tag metrics streaming APIs as Internal due to their less-stable or less-supported status [DET-4546] (1592)
175f2125 fix: issue data type warnings in to_device (1591)
5e9b6d38 fix: tensorboard with absolute storage_path (1590)
aedab511 chore: webui migrate getAgents to new API [DET-3844] (1576)
1bb19906 feat: add configurations for priority scheduler [DET-4507, DET-4509] (1565)
Docker images
- `docker pull determinedai/determined-master:0.13.10`
- `docker pull determinedai/determined-master:93b93697`
- `docker pull determinedai/determined-master:93b9369791d9278be68e720ab65c5328d3fed5b9`
- `docker pull determinedai/determined-dev:determined-master-93b93697`
- `docker pull determinedai/determined-dev:determined-master-93b9369791d9278be68e720ab65c5328d3fed5b9`