Determined

Changelog

* d214a34df chore: bump version: 0.18.2-rc2 -> 0.18.2
* 10b172bcc docs: add release notes for 0.18.2 (4318)
* 38e0efb15 chore: bump version: 0.18.2-rc1 -> 0.18.2-rc2
* c5cb23fc9 fix: harness azure dependencies (4319)
* f623c1c1e chore: bump version: 0.18.2-rc0 -> 0.18.2-rc1
* 350de3838 chore: cleanup container terminations resent on agent exit (4309) [DET-7533]
* f02f11bc5 fix: skip allocation check on checkpoint save to fix pulling preemption (4308)
* cfc0f8108 fix: remove limit of length 4 for task id (4304)
* 0645ba968 fix: harness dependency class fore azure (4302)
* 9991528a2 chore: bump version: 0.18.2-dev0 -> 0.18.2-rc0
* 86b6c28a4 chore: lock api state for backward compatibility check
* 271ae41c7 chore: ensure container terminations aren't lost in the event of network failures [DET-7440, DET-7441] (4272)
* 2d5a11865 fix: check for errors from actor asks in createResourcePoolSummary [DET-7494] (4294)
* daee03661 fix: stale agent state crashes resource pool [DET-7493] (4295)
* 4722caa95 fix: limit waiting for logger to 30 seconds (4289)
* b8f9463c6 fix: jupyterlab modal stale values, not sending values to api
* 42615b4b1 feat: add cli and api for delete checkpoints [DET-7119], [DET-7120] (4246)
* ad9bcd809 chore: restore determined_version in Trial checkpoint metadata (4288)
* 0b7dcdbd2 chore: widen node version detection to 16.29 (4290)
* 118a850da fix: pass signals through wrapper processes (4286)
* 0089ede8c fix: enable Full Configuration in appropriate modals [DET-7495] (4280)
* 45c62f20c feat: show only selected trials in learning curve
* 922b11349 chore: reduce use of any in InteractiveTable
* 5b03672a4 chore: update docs to accurate node support (4279)
* 59f686c96 fix: make trials table sort properly (4275)
* 695b30117 feat: incrementally release resources (4278)
* c0f53fa26 perf: avoid Seq Scanning raw steps (4244)
* b36596f13 fix: backslashes in show_ssh_command for windows (4260)
* ec503e06d fix: correct error message when command list fails (4235)
* 15b0525fc feat: add master-side verification for agent mTLS (4220)
* c6e824937 fix: don't hardcode /bin/which in entrypoints (4257)
* 236c24117 refactor: migrating tables to use InteractiveTable component [DET-7382] (4229)
* 156ff504b fix: reshow drop targets for customize columns modal (4199)
* 5e9962686 docs: delete mnist_tf_layers example (4263)
* 9d8223113 Revert "add autosync action"
* 8288461f5 add autosync action
* 70e534f9e fix: det task logs unable to use trial task IDs and checkpoint GC task IDs [DET-7424] (4258)
* 12d21c2e0 fix: gcs storage upstream test failure (4253)
* b09ec407e feat: improve trial logs to have some system events [DET-5885] (4215)
* c20ad4334 ci: Lints migrations to ensure new migrations have higher timestamps than old ones [DET-7146] (4250)
* da83e6509 chore: small cleanups for slurm (4211)
* b71f179aa feat: det deploy now can use --yes to skip prompts [DET-7408] (4255)
* 7b3fee44c chore: better slurm option override support (4254)
* 7af3cfe48 feat: add google cloud storage (gcs) prefix support [DET-6883] (4238)
* b5c2b14f8 docs: add enhanced launcher user guide (4248)
* eb1d9aaac chore: add dev server support for embedded tasks view (4243)
* 9d19a8704 perf: dont repeatedly reprocess profiler data
* b6582c02b ci: add check to prevent ssh git url (4240)
* 50bdbc3a8 chore: bumpenvs (4239)
* 1065314ea chore: add node_modules to eslintignore (4237)
* c74dee646 feat: Add theme toggle to user settings [DET-7321] (4204)
* 3016fc103 chore: remove legacy code/docs for NCCL/Gloo port range config (4187)
* f1e8d3cdc ci: check ulimit before 4x4 distributed test on macs (4234)
* 7d9e999cd fix: adjust page to preserve props.children (4231)
* 35258efd7 feat: Enable sending empty string for displayname with fallback to username [DET-7031] (4140)
* 2e77ec58c fix: det shell start/open, in windows (4227)
* 7217cfd94 feat: k8s detect non-det tasks (4154)
* a88f5c450 chore: share webui base page (4218)
* 15bd75895 fix: use custom image for tensor board [DET-7242] (4123)
* de926b1eb chore: fix rendezvous timeout logic (4226)
* c29e97a45 chore: base Dockerfile TensorFlow 2.6, 2.7, 2.8 security patches [DET-7325] (4223)
* 7882381cd fix: authenticate pprof endpoints [DET-7402]
* 2b0ccb830 chore: bump version: 0.18.1-dev0 -> 0.18.2-dev0
* bcbab4d92 docs: add release notes for 0.18.1 (4216)
* 034b9570c chore: revert rename of `RestoreResourcesFailure -> ResourcesFailure`. (4210)
* a7c4c2afe feat: enable agent-side mTLS for connection to master (4212)
* c9e13b64a feat: save connection in context (4213)
* 15f65ab37 feat: pix2pix example (4125)
* db987de7d chore: delete "conditional" json-schema extension (4177)
* 9b33f825c fix: use bigint for checkpoint size in `proto_get_trials_plus` (4208)
* 2c4c847d7 docs: update release note instructions with important admonition (4207)
* 138caf8f3 fix: pool detail page tab count when loading (4200)
* c5f685f18 feat: move task logs to embedded view [DET-7169] (4179)
* 54982c9be perf: tweak proto_get_trials_plus plan (4206)
* e585ebbb1 refactor: cleanup task logging shell scripts (4113)
* 4e2913f4a chore: update entrypoint in expconf docs (4198)
* fcff1c2aa fix: agent panic on commands with unusal formatted environment variables [DET-6649] (4202)
* 4172a4690 refactor: pull in user service code changes from EE (4183)
* 1c48fa618 docs: improve OpenTelemetry docs slightly (4182)
* 7f508e462 fix: allow `internal: null` for pre-0.15.6 experiments (4197)
* 55957fe06 fix: add restarts back to get_trial_ids for sorting
* dd8d3f377 feat: add det experiment logs <EXP_ID> [DET-7145] (4190)
* 3ef36d56b chore: refactor action dropdown comp to be reused [DET-7171] (4164)
* 623e60d37 ci: bust circleci cache (4189)
* af56e0103 docs: document using AWS Load Balancer on EKS [DET-6669] (4174)
* 36e56670d feat: allow enabling Prometheus monitoring through helm [DET-6993] (4158)
* 3bb7bb1d9 style: minor theme fixes and style adjustments [DET-7349] (4161)
* 2166dfc88 docs: update screen shots for cluster UI (4188)
* b7a327819 style: address new `flake8-comprehensions`, `pyzmq==23.0.0`. (4185)
* 5e1a81cb2 feat: allow setting of checkpointStorage.prefix through helm [DET-7152] (4152)
* 526e1dcc1 feat: display trial restarts [DET-7347] (4160)
* 2fdc6d710 fix: agent can now be control-C while connecting to master [DET-6287] (4178)
* d339f7be2 chore: migrate det a list to new api and bindings (4186)
* ed257d38e refactor: rip out `UseFluentLogging`. (4184)
* 21b85907b docs: update fluent-bit version. (4181)
* 4b325a558 docs: document database SSL options (4169)
* 5d228f8a5 chore: make core-api tutorial Windows-friendly (4176)
* 187505148 fix: sync slot usage for k8s [DET-7350] (4172)
* 92a944f96 chore: add .dccache to .gitignore (4173)
* 5e7a30c43 docs: fix typo in release note (4170)
* a52210f9e feat: chart sync provider [DET-7309] (4139)
* a7fafbb9a fix: enable currently active side nav item (4167)
* 0a5d54d45 chore: fix hardcoded url in schema logic (4171)
* 7df89b293 chore: allow deleting delete failed experiments (4141) [DET-7070]
* 4ef2b6771 perf: fixup query for latest training per trial (4166) [DET-7352]
* 3d3fe1cd0 fix: include both old and new checkpoints in total checkpoint size (4165)
* 4d98cee3d fix: replace carriage returns with newlines in task output [DET-5302] (3945)
* a2f878aaf chore: only warn on invalid calls to daemonize resources for slurm (4108)
* c00ce0a4c chore: check git state in lock-api-state.sh (4163)
* ec743d754 ci: turn off github annotations. (4146)
* 0efd44d08 doc: fix a broken file reference (4131)
* b911d85ed fix: avoid potential race between AllocationReady and Running state (4159)
* cc59985dc revert: partial revert of 96e0e584 (4162)
* f07633b4e fix: port collisions for multiple shared-non distributed jobs (4120) [HAL-2894]
* 80d1bb152 feat: Add embedded experience for JupyterLab and TensorBoard [DET-7162] (4134)
* 83cc1ec8f fix: prevent experiment name in header from flowing entire vertical space of screen during resize (4157)


Docker images

- `docker pull determinedai/determined-master:0.18.2`
- `docker pull determinedai/determined-master:d214a34df`
- `docker pull determinedai/determined-master:d214a34df0c0eb2e5e38ae63d1359862fd2af8f1`
- `docker pull determinedai/determined-dev:determined-master-d214a34df`
- `docker pull determinedai/determined-dev:determined-master-d214a34df0c0eb2e5e38ae63d1359862fd2af8f1`
- `docker pull nvcr.io/isv-ngc-partner/determined/determined-master:0.18.2`
- `docker pull nvcr.io/isv-ngc-partner/determined/determined-master:d214a34df`
- `docker pull nvcr.io/isv-ngc-partner/determined/determined-master:d214a34df0c0eb2e5e38ae63d1359862fd2af8f1`

Changelog

* 9284a3aa6 chore: bump version: 0.18.1-rc7 -> 0.18.1
* eb09c3449 docs: add release notes for 0.18.1 (4216)
* bad9a336a chore: bump version: 0.18.1-rc6 -> 0.18.1-rc7
* 6cff1b0a5 fix: use bigint for checkpoint size in `proto_get_trials_plus` (4208)
* 60291e934 chore: bump version: 0.18.1-rc5 -> 0.18.1-rc6
* 8f4a7977d perf: tweak proto_get_trials_plus plan (4206)
* 414bcd2a4 chore: bump version: 0.18.1-rc4 -> 0.18.1-rc5
* 789b39c9c fix: allow `internal: null` for pre-0.15.6 experiments (4197)
* fd27bac05 fix: add restarts back to get_trial_ids for sorting
* 845f2f06e chore: bump version: 0.18.1-rc3 -> 0.18.1-rc4
* eed09e911 docs: update screen shots for cluster UI (4188)
* 88271dd14 style: minor theme fixes and style adjustments [DET-7349] (4161)
* f2a4e5ee4 feat: display trial restarts [DET-7347] (4160)
* eaf84e681 chore: bump version: 0.18.1-rc2 -> 0.18.1-rc3
* a8ddc8208 fix: sync slot usage for k8s [DET-7350] (4172)
* 1f13710b3 fix: enable currently active side nav item (4167)
* e9333d2f1 perf: fixup query for latest training per trial (4166) [DET-7352]
* 764ef2d06 fix: include both old and new checkpoints in total checkpoint size (4165)
* 4157c82a9 chore: bump version: 0.18.1-rc1 -> 0.18.1-rc2
* e2f949ab2 chore: bump version: 0.18.1-rc0 -> 0.18.1-rc1
* 26ede20fb chore: revert scheduling docs
* 8ea2a5256 fix: prevent experiment name in header from flowing entire vertical space of screen during resize (4157)
* 1e244e048 chore: bump version: 0.18.1-dev0 -> 0.18.1-rc0
* 96e0e5840 chore: lock api state for backward compatibility check
* 82f0366dc feat: allow NaN validation metrics [DET-7177] (4150)
* 0bbeec1f7 feat: upload all tb files DET-7139 (4155)
* 90b918acf fix: adjust upscaling of column widths [DET-7220] (4138)
* 4b66bf07f feat: rolling upgrades v0 [DET-6548] (4031)
* beea2451a ci: disable most checks on ci-only changes (4118)
* 5f7e74a6c fix: upstream test failures due to config being admin protected (4153)
* da1dcd76a fix: return user data when new user is created [DET-7255] (4149)
* 1eba7a20c docs: a vain attempt to pass ci test on already approved pr4110 content changes (4151)
* 7aea015d0 fix: No redirecting url when model name is changed (4127)
* 00171f82b feat: Cluster UI improvement [DET-7072, DET-7073] (4009)
* 27e04e4d8 feat: require admin privileges for cluster managment [DET-7186] (4129)
* 09a8ff698 ci: update gke version. (4147)
* fa3a95960 ci: Increase package-and-push-system-local resource class (4143)
* 0d4fe2385 chore: fix boolean urlparams for grpc (4136)
* b1829b8c8 feat: enable SLURM preemption (4114) [FOUNDENG-21]
* b4ef2735f chore: add a local docs server (4117)
* 5dea21112 refactor: theme architecture [DET-6211] (4004)
* aef66b088 build: make docs build incremental and idempotent (4116)
* ec7007cef ci: persist debs and rpms in circleci for dev, rc, and release builds (4124)
* b89b0a359 fix: user filter on dashboard [DET-7251] (4132)
* e3e50a916 chore: add job ID and experiment labels to prometheus endpoint mappings [DET-6964] (4119)
* b0d8a935e ci: make codecov information for sure now. (4130)
* b313503d0 chore: restructure shareable webui utils and types (4112)
* 78505d6e6 ci: turn off codecov bot PR comments (4122)
* a91866fdb chore: fix rank determination for horovod with mpi (4109)
* 3368a7727 fix: NCCL interface in distributed tests (4111)
* 18aadd0dd chore: bump version: 0.18.0-dev0 -> 0.18.1-dev0
* 7500d6fc6 docs: add release notes for 0.18.0 (4102)
* 1f4a64260 chore: bump version: 0.17.16-dev0 -> 0.18.0-dev0
* 59928ec4f chore: explicit naming of preemption and coscheduler resources [DET-7140] (4101)
* ac2564b87 fix: add missing task log teardown for trials (4107)
* c41e1dcc0 docs: rework quickstart for ml developers (4091)
* 2c7564d13 chore: use reported slots available for on prem deployments (4095)
* 62049aeab chore: change codecov to informational only (4105)
* 5061a8fd3 fix: mark distributed tests as parallel (4093)
* 87b8e5396 chore: enable codecov enforcement (4084)
* 15dd7e3d4 fix: bindings sessions in experiment apis. (4096)
* eb65b09ff chore: cleanup and fixes for "det deploy" (4103)
* d78a7120a perf: improve plan for proto_get_trial_plus.sql (4073)
* a980c560b build: update submodules on webui get-deps (4082)
* 78d4e8b7d chore: HAL-2879 Cleanly shutdown all sshd servers on exit (176) (4087)
* 940f8f703 chore: Refactor JupyterLabModal pattern [DET-6276] (4072)
* ec5553ab4 chore: make container proxy support more flexible, for slurm (3948)
* 3614c83a6 chore: wait for process substition log filters [DET-6712] (3930)
* 474742f9a chore: clean up useCallback dependency (4092)
* 39588ee25 feat: add det.LOG_FORMAT constant (4090)
* bcc50f97c fix: Support rendering rank of 0 (4083)
* c523523e8 fix: consistent total slot calculation for cluster overview [DET-7182] (4080)
* 9c163497c feat: add wrap_rank helper script (4086)
* ea4a94911 fix: dont show archived in column picker [DET-7187] (4085)
* a4e5f8410 feat: authenticate task proxies (4071)
* 5fab38417 fix: wrap torch.distributed launch in pid server/client (4077)
* e73f06309 fix: use displayNames in ClusterHistoricalUsage (4059)
* 3368dd86a chore: filter out NaN, +/- Infinity metric values for charts for now. (4076)
* f8b5bf50e chore: add and consolidate code coverage to codecov (4064)
* 487b04c21 docs: release note for core api (4069)
* 15a668be9 fix: add user column back to experiment list (4070)
* 880b769ec feat: break workload info from trial endpoint into a new endpoint [DET-6729] (3635)
* d703c968c fix: show notification when delete experiment fail [DET-6811] (4051)


Docker images

- `docker pull determinedai/determined-master:0.18.1`
- `docker pull determinedai/determined-master:9284a3aa6`
- `docker pull determinedai/determined-master:9284a3aa6e307c61426c93b5e09730c664725604`
- `docker pull determinedai/determined-dev:determined-master-9284a3aa6`
- `docker pull determinedai/determined-dev:determined-master-9284a3aa6e307c61426c93b5e09730c664725604`
- `docker pull nvcr.io/isv-ngc-partner/determined/determined-master:0.18.1`
- `docker pull nvcr.io/isv-ngc-partner/determined/determined-master:9284a3aa6`
- `docker pull nvcr.io/isv-ngc-partner/determined/determined-master:9284a3aa6e307c61426c93b5e09730c664725604`

Changelog

* 3c00fc281 chore: bump version: 0.18.0-rc3 -> 0.18.0
* 797ceeca4 docs: add release notes for 0.18.0 (4102)
* 505102457 chore: bump version: 0.18.0-rc2 -> 0.18.0-rc3
* bd87faa20 perf: improve plan for proto_get_trial_plus.sql (4073)
* c2a3ba920 fix: Support rendering rank of 0 (4083)
* d46995f27 fix: consistent total slot calculation for cluster overview [DET-7182] (4080)
* 1b946122f feat: add det.LOG_FORMAT constant (4090)
* aacd23be7 feat: add wrap_rank helper script (4086)
* 8fbf0a78a fix: dont show archived in column picker [DET-7187] (4085)
* a190a1f88 chore: bump version: 0.18.0-rc1 -> 0.18.0-rc2
* 88fc38b7f chore: bump version: 0.18.0-rc0 -> 0.18.0-rc1
* 4d73ceaa6 feat: authenticate task proxies (4071)
* 3e594e47b chore: filter out NaN, +/- Infinity metric values for charts for now. (4076)
* 1038840af fix: wrap torch.distributed launch in pid server/client (4077)
* 7df0f0013 docs: release note for core api (4069)
* 7721553cc fix: add user column back to experiment list (4070)
* db5f434de chore: bump version: 0.17.16-dev0 -> 0.18.0-rc0
* 02114cf37 chore: lock api state for backward compatibility check
* 7be795af1 docs: Core API reference and cookbook docs (4054)
* 5e364d1a6 chore: ancient checkpoints for very old pytorch (4068)
* 51e80d8ea perf: minimize create/destroy of uPlots [DET-6972] [DET-6972] [DET-6796] [DET-6853] [DET-6672] (3935)
* 531af28cc chore: update removed reducer methods (4067)
* 6f63f13ec chore: remove det.pytorch.reset_parameters() (4066)
* 4fb1886b8 feat: generic checkpoints and making Core API public (3859)
* db0f8ef5c fix: correct the return type for readStream (4063)
* 33a1a6c66 chore: remove PBT searcher (4058)
* a35d49a77 feat: support for torch native dtrain (3807)
* 29385bc09 chore: update k8s scheduler to run latest image (4061)
* e7f328962 chore: remove remainder of native api (4055)
* 87008acad chore: deprecate data layer (4056)
* 75a6bf59d chore: remove deprecated experimental custom reducer methods (4060)
* 948518cee chore: remove unnecessary use of username in webui (DET-6922) (4049)
* d2cc8d425 refactor: simplify apiConfig to reduce redundancy (4043)
* 707dd2680 chore: Refactor CheckpointModal to be hook based [DET-7136] (4034)
* ea535b035 fix: make "Show full config" modal larger (4053)
* 625070edb chore: move webui codecov upload to use env var instead of hardcoded token (4045)
* 0c1821eed fix: remove dead shell start code [DET-7131] (4016)
* 59e5b9b9a chore: Recommend git clone --recurse-submodules for submodules (4036)
* ad06cd431 fix: move allocation resources migration to the top.
* a6bf58b65 refactor: rewrite ndjson streamer [DET-7121] (4014)
* 6d018f09e fix: Alter Boolean arg default handling (4038)
* b278ef040 chore: store allocation resources and agent RM containers in DB. (3946)
* 7afc75f49 ci: Add build/coverage badge for webui/react [DET-6767] (4028)
* 628f07c96 move up profile-pics migration to assure it happens
* 600d77d52 fix: put profile pic migrations in correct directory
* 3569a6f80 build: set up shared-web submodule [DET-6961] (4006)
* 0686befc4 test: ensure that agent disabling doesn't count for experiment restarts [DET-5916] (4029)
* b61a83dec refactor: remove `NewAllocationID`. (3959)
* b605da88d fix: add another missing key fix (4033)
* 64e6f7d9e feat: allow position modification in k8s [DET-6967] [DET-6968] (3938)
* 24bbec93f feat: Creating a table for user profile pictures
* b831dbd34 fix: Human-readable option for empty filters in logs [DET-6781, DET-6999] (4017)
* 5d04a258c fix: Prevent archived models from appearing in the Register Checkpoint modal [DET-7132] (4015)
* c1523dab1 fix: Ensure table offset does not exceed pagination total [DET-6829] (4011)
* 84401b26c fix: change timeout in e2e_tests/tests/cluster/test_logging.py/test_trial_logs (4019)
* ad6f3d54a fix: add key for cancel operation (4023)
* 9ba0f0459 chore: remove deprecated type provider for moment-timezone (4018)
* be3125b8d docs: update copyright year (4005)
* ea4ab8854 feat: add product feedback link [DET-5811]
* 616241874 chore: fix documentation comment for /ws/data-layer
* 482ecd61c docs: fix grammar in training-run index (3988)
* 5480903ae docs: fix indent in pytorch-porting-tutorial (3989)
* 9a6d75260 chore: warn about ambiguous enum params (3997)
* 4de96b955 chore: give the latest-master deploy job a name (3900)
* 64b46627c fix: pass task time not from logCtx (3993)
* 59796b027 chore: update node version to active LTS (DET-7046) (3932)
* 07de42689 chore: show appropriate severity level on job launch failures (4002)
* c68b7a429 test: add option to disable `compare_stats`. (4008)
* 915184c77 chore: Removing CODEOWNERS entry for release notes (3978)
* 93b98c269 fix: handle zeros correctly in HpTrialTable metricSorter (4003)
* cd0e53149 chore: bump version: 0.17.15-dev0 -> 0.17.16-dev0
* 7d1493d00 docs: add release notes for 0.17.15 (3986)
* 63eb86c92 feat: add modal to explain why users cannot delete items [DET-6998] (3994)
* f6fbc05b6 fix: guard trial and allocation exit logic correctness (3983)
* 08218c128 fix: do not default to `noverify` for `bindings` sessions in CLI. (3991)
* 4c5dad8a1 chore: add expconf environment.slurm (3966)
* d2610c557 fix: handle infinite metrics in searcher snapshots [DET-7122] (3999)
* 57d1b38f9 chore: handle rank id in log entries (3995)
* f915a6952 fix: handle infinite validation metric values in more cases (3992)
* 8527e60c5 fix: Experiment columns filter is still applied after closing [DET-6837] (3982)
* c7f4610f0 fix: avoid permanent filtered state in model registry [DET-6946] (3984)
* 35ac5ff88 docs: update release note process (3990)
* d565f842e ci: bump profiling test timeout back up (3981)
* a43fcffed chore: handle errors from starting allocations [DET-5862] (3975)
* bd5484561 feat: add det support bundle [DET-5886] (3904)
* 899db1998 feat: add drag and drop functionality to experiment list column table [DET-7044] (3956)
* cf5c94fb6 chore: document usage of /ws/data-layer [DET-6685] (3971)
* 6f6e4a26a feat: Add overall allocation bar to new cluster page [DET-7074] (3955)
* ff78aa985 chore: add error type for non retry-able resource manager errors (3947)
* e9d26af42 chore: apply filtering to task logs (3963)
* 2a981a198 fix: add test warmup command e2e [DET-5803] (3965)
* a77b8c189 docs: fix tutorial link swap (3979)
* 5d65f6443 ci: remove old semantic PR app config. (3980)
* 896725bc7 fix: notebook logs filter by level (3967)
* f3660c2cb chore: remove old notebook logs endpoint (3960)
* 8a99e03c8 fix: task log level parsing (3973)
* 005d14402 fix: forward job.DeleteJob through agentRM (3968)
* 8e21bf0f8 docs: add release notes for PR 3914 (3962)
* 28ab2fa42 chore: Update list of false alarms in Docker image scanning (3933)
* 80a1dc7f6 ci: new semantic pull request check. (3958)
* 189f148f7 docs: update package versions, add ROCm, edit to style guide (3950)
* 0d180fe96 fix: update HPE logo sizes (3953)
* 393de717f chore: add message to cleanup external RM resources on delete, for slurm (3902)


Docker images

- `docker pull determinedai/determined-master:0.18.0`
- `docker pull determinedai/determined-master:3c00fc281`
- `docker pull determinedai/determined-master:3c00fc281542c272c1591c7d1c86eb53db8f230c`
- `docker pull determinedai/determined-dev:determined-master-3c00fc281`
- `docker pull determinedai/determined-dev:determined-master-3c00fc281542c272c1591c7d1c86eb53db8f230c`
- `docker pull nvcr.io/isv-ngc-partner/determined/determined-master:0.18.0`
- `docker pull nvcr.io/isv-ngc-partner/determined/determined-master:3c00fc281`
- `docker pull nvcr.io/isv-ngc-partner/determined/determined-master:3c00fc281542c272c1591c7d1c86eb53db8f230c`

Changelog

* 9b74e5444 chore: bump version: 0.17.15-rc4 -> 0.17.15
* 931983ac6 chore: bump version: 0.17.15-rc3 -> 0.17.15-rc4
* 3bd2b4afa fix: handle infinite metrics in searcher snapshots [DET-7122] (3999)
* 153221eda chore: bump version: 0.17.15-rc2 -> 0.17.15-rc3
* 1e9f5c4b6 chore: handle rank id in log entries (3995)
* d12c8921e fix: handle infinite validation metric values in more cases (3992)
* 59b180d26 docs: add release notes for 0.17.15 (3986)
* 9e3e5419f chore: bump version: 0.17.15-rc1 -> 0.17.15-rc2
* 1c97bc76a chore: apply filtering to task logs (3963)
* 8dd01c87f fix: task log level parsing (3973)
* 9c241218e docs: add release notes for PR 3914 (3962)
* 507cfe75d fix: update HPE logo sizes (3953)
* e1a3de646 chore: bump version: 0.17.15-rc0 -> 0.17.15-rc1
* dc3ef476f chore: bump version: 0.17.15-dev0 -> 0.17.15-rc0
* 6c38aad74 chore: lock api state for backward compatibility check
* f206919e8 fix: take out job summary caching [DET-6695] (3849)
* c9941ac85 chore: add missing icons to mobile navbar [DET-7009]
* 927ef9417 fix: parse different time format in compare stats script [DET-7039] (3909)
* 55fcf65f4 fix: checkpoint gc job should close (3943)
* b8f1073e1 feat: track task stats [DET-6872, DET-6926, DET-6927] (3852)
* b1a470d7c chore: add key attribute to avatar ActionCard action (3939)
* 7c7375c26 chore: give det job a default command to run (3934)
* 9028fe001 chore: mask registry auth password in harness [DET-6279] (3867)
* 7d0234b16 chore: add filtering by userIds to API endpoints (DET-7019) (3898)
* 180a79f47 chore: mask registry creds in webui [DET-7013] (3881)
* 2d4456439 replace username with userId for user API (3914)
* b6b2c7172 test: add interaction tests for spinner [DET-6665] (3826)
* d86d09ba5 chore: prevent InteractiveTable scroll from moving pagination or other controls [DET-7037] (3923)
* c9a24582f fix: crash on upgrade to InteractiveTable [DET-7036] (3922)
* 4c0ef956c hide archived experiments unless using --all (3918)
* 586bba4a2 docs: tweak docs for socket activation (3926)
* 64ae7ca2a perf: add infiniband-related libraries to environment (3832)
* bcc954a0d chore: disallow getting metadata from dummy checkpoint context (3920)
* 575cc21f7 chore: update dep requirements for react (3901)
* e359843b3 chore: demote _get_last_validation to internal (3919)
* 5c93cb68a chore: log checkpoint uuids in core, not in wlsq (3924)
* 7086df4b7 chore: add missing docstrings in core api (3911)
* ea0db53a9 feat: add slurm rendezvous (3777)
* d5e793ba1 chore: make store_path auto-create the directory (3916)
* 1f3c4b232 feat: add core.DownloadMode (3910)
* c334a5385 docs: fix install cli typo (3917)
* 069eb6f27 fix: run scheduling on agent connection/enable events, reconnectBacklog replay. (3906)
* b75afe6d9 chore: bump version: 0.17.15-dev0-dev0 -> 0.17.15-dev0
* d06dc9429 refactor: remove dependency of settings in the updateSettings call (3894)
* 1b4c8e951 chore: remove pr preview cluster address [DET-7040] (3907)
* e5723c7ea docs: Release notes for 0.17.14 (3912)
* 43b9a7c06 chore: bump version: 0.17.14 -> 0.17.15-dev0
* 5c4554437 fix: RM crashes when setting cmd priority (3908)
* 0c5236786 chore: fix deepspeed nightly tests (3897)
* 3e6267f0d chore: update StorageManager and extend CheckpointContext (3829)
* f76cc4599 docs: fix description of scheduling_unit behavior (3890)
* 1815ee3cb chore: sweeping rename of Core API components (3896)
* 0b7141b88 feat: deepspeed DCGAN example (3758)
* 015a8e07c feat: use enums instead of chief_only bool in Core API (3888)
* 129d841cb fix: use otel only if enabled (3893)
* 0ab3288d8 hide sizeChanger on RoutePagination (3892)
* ae50d3cec fix: match reported rp name for k8 across endpoints [DET-7006] (3870)
* c49561c6f docs: various fixes for master configuration and k8s docs (3889)
* c716759b1 chore: bump version: 0.17.13-dev0 -> 0.17.14-dev0
* 2fdb2efc7 docs: add release notes for 0.17.13 (3879)
* a34962262 chore: Clean up tests with fewer Optional types and asserts (3872)
* b0e0c96ea feat: make core.Searcher multiworker-safe (3871)
* 50c9f669d feat: show agent version in `/agents` and `det agent list` [DET-6847] (3873)
* 954cf9714 fix: avoid double-timestamps in logs (3876)
* b00124a41 feat: Drag to Reorder and Resize Experiment List columns [DET-6438] [DET-6809] (3765)
* be5848588 chore: use better NCCL SOCKET setting for gpt-neox (3874)
* 65e1a97d3 chore: remove internal flag from det.launch.horovod --help (3875)
* 9b1fa3bd0 feat: Search models by name and description substring [DET-6939] (3869)
* 78ba4a33a feat: add opentel to determined master [DET-6775] (3851)
* 789f16dce chore: cleanup stray changes (3868)


Docker images

- `docker pull determinedai/determined-master:0.17.15`
- `docker pull determinedai/determined-master:9b74e5444`
- `docker pull determinedai/determined-master:9b74e54448d64009ce574e1a68b52149c1d00fe7`
- `docker pull determinedai/determined-dev:determined-master-9b74e5444`
- `docker pull determinedai/determined-dev:determined-master-9b74e54448d64009ce574e1a68b52149c1d00fe7`
- `docker pull nvcr.io/isv-ngc-partner/determined/determined-master:0.17.15`
- `docker pull nvcr.io/isv-ngc-partner/determined/determined-master:9b74e5444`
- `docker pull nvcr.io/isv-ngc-partner/determined/determined-master:9b74e54448d64009ce574e1a68b52149c1d00fe7`

Changelog

* cf1cc00b1 chore: bump version: 0.17.14-rc0 -> 0.17.14
* f06304521 docs: Release notes for 0.17.14 (3912)
* 83868bc34 chore: bump version: 0.17.14-dev0 -> 0.17.14-rc0
* d64bbf762 chore: bump version: 0.17.13 -> 0.17.14-dev0
* 4f42c1b4b fix: RM crashes when setting cmd priority (3908)


Docker images

- `docker pull determinedai/determined-master:0.17.14`
- `docker pull determinedai/determined-master:cf1cc00b1`
- `docker pull determinedai/determined-master:cf1cc00b17f8e5d59effbd0e2f25d32328ade8b5`
- `docker pull determinedai/determined-dev:determined-master-cf1cc00b1`
- `docker pull determinedai/determined-dev:determined-master-cf1cc00b17f8e5d59effbd0e2f25d32328ade8b5`
- `docker pull nvcr.io/isv-ngc-partner/determined/determined-master:0.17.14`
- `docker pull nvcr.io/isv-ngc-partner/determined/determined-master:cf1cc00b1`
- `docker pull nvcr.io/isv-ngc-partner/determined/determined-master:cf1cc00b17f8e5d59effbd0e2f25d32328ade8b5`

Changelog

* 7629d330a chore: bump version: 0.17.13-rc3 -> 0.17.13
* 0de91c795 docs: add release notes for 0.17.13 (3879)
* e02363582 chore: bump version: 0.17.13-rc2 -> 0.17.13-rc3
* 8ac3a499d fix: avoid double-timestamps in logs (3876)
* 11e84f829 feat: show agent version in `/agents` and `det agent list` [DET-6847] (3873)
* e00ebb775 chore: bump version: 0.17.13-rc1 -> 0.17.13-rc2
* 79d96c7a3 chore: remove internal flag from det.launch.horovod --help (3875)
* 6f705b7b2 chore: use better NCCL SOCKET setting for gpt-neox (3874)
* 6ebd3c796 chore: bump version: 0.17.13-rc0 -> 0.17.13-rc1
* a95b892ad chore: cleanup stray changes (3868)
* 5d85dc806 chore: bump version: 0.17.13-dev0 -> 0.17.13-rc0
* 4ffc52871 chore: lock api state for backward compatibility check
* cde70d40b fix: e2e test stats for gke machine (3865)
* c6d557495 fix: deepspeed cifar moe example ds_config (3864)
* 9ff7aa86d feat: make launch layers generally useful (3853)
* 5c5c7b161 feat: allow the CLI to accept incomplete task UUIDs [DET-6706] (3836)
* bb8245dc4 Revert "chore: use userId as url param for user endpoints (3806)" (3855)
* 4f5f3221e feat: add table to store unused GPU [DET-6697] (3735)
* 7f64b0ec3 fix: gpt-neox example multinode on aws (3848)
* be5232a10 refactor: simplify `resourcemanagers.container`. (3840)
* cc763f8af set ExperimentCreate modal width in AdvancedMode (3831)
* 3176e9e7c add conditionalRender prop to Spinner (3837)
* 733873899 fix: remove duplicated code in useModal (3845)
* 3444f4aab fix: outdated deepspeed launcher rest api call (3834)
* bb4050fe6 ci: fail building docs when sphinx-build fails (3843)
* 63d86d38e fix: docs build failure for deepspeed reference (3844)
* 998103d83 fix: remove python 3.10 union-type syntax (3839)
* 6751860fa chore: convert ptrs.* to generics-based ptrs.Ptr (3792)
* c2142e410 chore: handle race between experiment persistent state, actor existence in ask, take 2 (3835)
* 470f98b39 chore: add missing FK constraints for `trials` and `allocation_sessions` (3810)
* c6697a585 refactor: consolidate allocation state. (3833)
* 607b565ae chore: add generic math max/min (3823)
* 9777130a9 chore: handle race between experiment persistent state, actor existence in ask (3821)
* ba56d2410 use RoutePagination to navigate between trials (3828)
* 31ecb4a4d fix: Set default project for experiments.project_id [DET-6958] (3825)
* 91a528369 doc: nested hps (3827)
* 02c55857e chore: clean up get_experiment.sql query (3822)
* 8467f936f fix: ci failures for deepspeed and pytorch geometric (3824)
* 39e629d63 fix: keep page from crashing when trial/hyperparameters are not loaded yet [DET-6951]
* 4175f96cc chore: bump version: 0.17.12-dev0 -> 0.17.13-dev0
* 69e396056 docs: add release notes for 0.17.12 (3815)
* 7d64e11ae chore: update pytorch data loader args (3775)
* 3f9188667 ci: fix click/black conflict. (3818)
* fc56db09c docs: Add examples and documentation for AMP [DET-5170] (3757)
* 0639e8017 chore: add trialIds to GET experiment response (3809)
* 4f93158bf feat: Accept model.id for lookups, show model.id in UI [DPS-49] (3723)
* 95ea7a29f fix: use max slots provisioning agents [DET-5725] (3788)
* e35972d83 chore: use userId as url param for user endpoints (3806)
* 1c6c84894 fix: change menus for mobile resolution (DET-6869) (3801)
* 0ff423940 fix: remember originally requested page on login [DET-6928] (3790)
* 2c263bc2c chore: release notes for job queue (3782)
* d70550495 ci: delete useless and slow noop test (3812)
* 7f94c988a fix: bugs in PIDServer (3800)
* 9592e0417 chore: add bun.DB singleton in master (again) (3798)
* f7ac95506 fix: add pytorch as requirement before deepspeed is installed (3811)
* 78b50fb26 fix: update references to branding (DET-6870) (3802)
* 4146b021e chore: upgrade to support panoptic segmentation for coco [DET-6945] (3545)
* d920fed12 chore: allow cors by default in dev cluster. (3794)
* b4ded4b28 fix: add dependency for harness tests (3804)
* c4cfaa045 chore: API requests that return username should also return userId (3796)
* c3008e33b chore: bump deepspeed version to 0.6.0 (3803)
* 7bec5b39e fix: update HPE logos (DET-6868) (3789)
* c2904b32a fix: cli win tests (3799)
* 704c17384 chore: reorder jobs.q_position migrations [DET-6934] (3797)
* 476b12d2d revert: add bun.DB singleton in master (3769)
* db4062c82 test: turn off timeout for now on logs tests (3793)
* 9f15d2468 docs: fixes for concepts index page (3756)
* 03c68f36c chore: add bun.DB singleton in master (3769)
* bb8186955 feat: add deepspeed support (3235)
* 67f0ca409 restore getExperimentModelDefinition (3791)
* 21c2d2768 chore: Update required fields, remove repeat fields from model registry [DET-6861] (3755)
* a8d541bce fix: wl_output with zero trials (3786)
* 4297aa596 docs: update job queue docs to mention k8 (3781)
* 1542afff8 chore: add ProtoConverter (3766)
* 86e029652 fix: kubernetes job queue unexpected message (3783)
* e89162da7 chore: disable showing queue position controls for k8 [DET-6917] (3784)
* 6df7fe587 docs: fix typo (3780)


Docker images

- `docker pull determinedai/determined-master:0.17.13`
- `docker pull determinedai/determined-master:7629d330a`
- `docker pull determinedai/determined-master:7629d330abbbf74b45865d524bf9d14761d263af`
- `docker pull determinedai/determined-dev:determined-master-7629d330a`
- `docker pull determinedai/determined-dev:determined-master-7629d330abbbf74b45865d524bf9d14761d263af`
- `docker pull nvcr.io/isv-ngc-partner/determined/determined-master:0.17.13`
- `docker pull nvcr.io/isv-ngc-partner/determined/determined-master:7629d330a`
- `docker pull nvcr.io/isv-ngc-partner/determined/determined-master:7629d330abbbf74b45865d524bf9d14761d263af`