0.15.4

Changelog

8149595 chore: bump version: 0.15.4rc1 -> 0.15.4
664452a docs: Release notes for 0.15.4. (#2370)
a2f531c chore: bump version: 0.15.4rc0 -> 0.15.4rc1
6a70ea3 fix: fix task pagination filters not taking effect [DET-5442] (#2367)
cd6683e chore: bump version: 0.15.4.dev0 -> 0.15.4rc0
2926056 chore: bump version: 0.15.3.dev0 -> 0.15.4.dev0
0c7cbb2 chore: fix bumpversion config not properly bumping setup.py files (#2366)
c409fb9 Revert "chore: bump version: 0.15.3.dev0 -> 0.15.4.dev0"
6a2472e Revert "chore: fix missing version bump to 0.15.4.dev0"
cbc62bd chore: fix missing version bump to 0.15.4.dev0
61a4663 fix: profiles timings graph data conversion filling empty data [DET-5433] (#2360)
b520be8 chore: lock api state for backward compatibility check
08f857e chore: bump version: 0.15.3.dev0 -> 0.15.4.dev0
1d85f24 chore: add unit tests for webui util functions [DET-5323] (#2347)
984ad9d feat: add workload status to trial infobox [DET-4289] (#2349)
2afc2e1 fix: ci/cd model-hub tests config (#2358)
f3f828b chore: fixes eslint error (#2361)
2f363ac chore: reorder migrations (#2362)
4a66da3 refactor: delete some old commands APIs (#2321)
9b5bc16 fix: ci/cd e2e tests timeout (#2353)
e37029f test: always calling read() before calling wait() (#2356)
d8c8921 feat: store the original user submitted experiment config in db (#2332)
f677fa7 feat: support transformers library in model-hub [DET-4823, 4719, 4721, 4720] (#2068)
fc27d77 fix: improvements to automatic pod spec configurator (#2306)
f6f13dc fix: hide expected network errors when nodes are terminated [DET-4822] (#2351)
63c77f2 chore: Add output printing to debug flaky test (#2350)
0f3be84 chore: drop prior_batches_processed and num_batches (#2345) [DET-5403, DET-5405]
1161b66 fix: fix test test cluster setup cmd. (#2341)
9de56d6 chore: migrate to use total_batches more in HP search viz. (#2344)
cceb764 fix: react build should depend on its public dir (#2339)
0ddac86 chore: edits to expconf before enabling it (#2342)
cd2980b0 refactor: provide support for specifying selector for element id for element list (#2255)
dc229e5 feat: tolerate missing GPU stats when running under MIG [DET-5387] (#2327)
4160745 chore: disable webui experiment archive test (#2340)
8a0ced9 feat: add internal searcher APIs [DET-5214] (#2301)
cbafb09 feat: replace "show archived" toggle with dropdown [DET-3925] (#2333)
5952d37 fix: improve uPlot chart zooming experience [DET-5395] (#2338)
97ed53c chore: add searcher type to output of experiment APIs (#2328)
434578b docs: Release notes for 0.15.3 (#2334)
e77a16a chore: fix docstring (#2337)
5d60865 chore: add viewport meta to improve WebUI mobile experience [DET-5396] (#2335)
a0242e7 fix: system metric chart fix to support milliseconds [DET-5348] (#2311)
dca96c2 chore: sort nulls last in experiment trial API (#2329) [DET-5300]
cfb46ab chore: go mod fixes (#2325)
916b75c chore: only select a single host port per container rendezvous port (#2331)
aafdcf0 chore: update package json [DET-5335] (#2314)
20263dc chore: use filelocks to guard data download (#2244)
4f12a6f chore: loosen ruamel.yaml version (#2313)
7e6f51a revert: "revert: "fix: gracefully handle Docker binding published ports to ipv4 and ipv6 for host (#2259) [DET-5295]" (#2326)" (#2330)
1df87b2 docs: add a missing word in react readme (#2324)
953f528 Revert "fix: gracefully handle Docker binding published ports to ipv4 and ipv6 for host (#2259) [DET-5295]" (#2326)
2fa12d8 chore: update Docker images, AMIs and harness for yogadl update (#2319)
0d4cb14 ci: move CUDA 11 testing to more available GPUs (#2316)
98904a1 chore: bump version: 0.15.2.dev0 -> 0.15.3.dev0
a6bfd9a chore: log incorrect rendezvous addresses (#2312)
58f3b1b fix: expconf required fields (#2309)
ce99043 docs: remove stray text from task config reference (#2305)
6bf2e0c chore: move LearningCurveChart to use uPlot shared component [DET-5331] (#2302)
bbb5007 chore: update environment images to 0.12.0. (#2304)
cf8de24 feat: harness collects profiler metrics [DET-5061] (#2198)
b8f5ef6 chore: add internal preemption API [DET-5216] (#2260)
e75f386 docs: Release notes for 0.15.2 (#2303)
0b13fc5 refactor: code split libraries [DET-5342] (#2291)
6209a57 remove endtime as required from json-schema (#2298)
57fd334 chore: apply react strict mode and upgrade to React 17 [DET-5325] (#2279)
b07a086 chore: improve master logging (#2295)
d136dff chore: minor expconf issues (#2297)
1834bbb fix: scary warning with det shell open (#2299)
0126e21 chore: add support for building with Golang race detector (#2296)
c4592c4 docs: add release notes for preemption in k8s (#2294)
f5a4f8f chore: Helm notes should recognize preemption scheduler (#2293)
64826d2 fix: correct logic for hasData in uPlotChart (#2292)
8514007 chore: fix common fields on union types (#2270)

Docker images

• docker pull determinedai/determined-master:0.15.4
• docker pull determinedai/determined-master:81495950
• docker pull determinedai/determined-master:8149595071d4efdd765b8965f9f7dee24900158f
• docker pull determinedai/determined-dev:determined-master-81495950
• docker pull determinedai/determined-dev:determined-master-8149595071d4efdd765b8965f9f7dee24900158f
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:0.15.4
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:81495950
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:8149595071d4efdd765b8965f9f7dee24900158f

0.15.3

Changelog

b42d42b chore: bump version: 0.15.3rc3 -> 0.15.3
380982a chore: bump version: 0.15.3rc2 -> 0.15.3rc3
b479d89 docs: Release notes for 0.15.3 (#2334)
07740ba chore: bump version: 0.15.3rc1 -> 0.15.3rc2
8ced4c3 chore: go mod fixes (#2325)
63e8e11 chore: only select a single host port per container rendezvous port (#2331)
6b33c76 chore: loosen ruamel.yaml version (#2313)
9bed90d chore: bump version: 0.15.3rc0 -> 0.15.3rc1
f188b8d chore: update Docker images, AMIs and harness for yogadl update (#2319)
1428672 chore: bump version: 0.15.3.dev0 -> 0.15.3rc0
b6d1bef chore: update environment images to 0.12.0. (#2304)
3807e7d chore: bump version: 0.15.2 -> 0.15.3.dev0

Docker images

• docker pull determinedai/determined-master:0.15.3
• docker pull determinedai/determined-master:b42d42bd
• docker pull determinedai/determined-master:b42d42bdb1e66daadb0dc1a2dc8454b072bab774
• docker pull determinedai/determined-dev:determined-master-b42d42bd
• docker pull determinedai/determined-dev:determined-master-b42d42bdb1e66daadb0dc1a2dc8454b072bab774
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:0.15.3
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:b42d42bd
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:b42d42bdb1e66daadb0dc1a2dc8454b072bab774

0.15.2

Changelog

89f3ee0 chore: bump version: 0.15.2rc2 -> 0.15.2
565e187 docs: Release notes for 0.15.2 (#2303)
b0e43c1 chore: bump version: 0.15.2rc1 -> 0.15.2rc2
4527077 fix: scary warning with det shell open (#2299)
6022114 docs: add release notes for preemption in k8s (#2294)
56e87c2 fix: correct logic for hasData in uPlotChart (#2292)
c675141 chore: bump version: 0.15.2rc0 -> 0.15.2rc1
fdf5abc chore: bump version: 0.15.2.dev0 -> 0.15.2rc0
76fe5c8 fix: wait for uPlot to be ready to setData or setSize [DET-5343] (#2283)
c367fc7 feat: promote custom reducers from experimental [DET-5322] [DET-5321] (#2284)
f3636f3 docs: add docs for preemption in kubernetes (#2289)
88ca463 feat: allow activation of priority scheduler in k8s (#2288)
ae13bb6 fix: lr_scheduler step when using gradient_aggregation [DET-5289] (#2271)
c1fa923 feat: add support for preemption in Kubernetes [DET-5135] (#2282)
54ff4ae fix: only warn for non-numeric np.dtypes [DET-5288] (#2287)
1f3b5e0 chore: remove svg and ttf fonts (#2286)
1454a15 chore: drop unused compose component (#2278)
1dc19d7 fix: squelch "response already committed" master log message (#2281)
b3c33a5 docs: add missing line to docker run (#2285)
7200fdf feat: expose user id as part of the user object [DET-4856] (#2265)
bea7875 fix: add support for dynamic section content via css [DET-5299] (#2277)
e1f14ba fix: improve rendering for uPlot chart with empty data [DET-5330] (#2274)
3b79dc5 chore: upgrade to labstack/echo v4.2.2 (#2266)
7239b10 chore: add support of y-axis zooming for uPlot [DET-5266] (#2268)
4b62700 expconf: fix some minor bugs in reflect code (#2267)
4bf88ba chore: fix typo in help for "experiment download". (#2269)
6c68692 fix: gracefully handle Docker binding published ports to ipv4 and ipv6 for host (#2259) [DET-5295]
6f5b86d ci: enable taiko get elements logging to help debug the disconnect from this and actual elements (#2264)
24ca0b5 fix: ignore hp-importance as a requirement for displaying hp-viz (#2258)
438a07a fix: allow agents to be set to empty [DET-5296] (#2261)
56e6f0f chore: migrate webui to use /api/v1/auth/login [DET-5287] (#2254)
9712291 chore: replace metric chart with uplot [DET-4303] (#2234)
468a70f chore: clarify API for expconf objects (#2256)
524d3a3 chore: add option to login through the new api w/ pre-hashed pwd [DET-5270] (#2253)
d08a449 chore: remove validation operations [DET-5213] (#2189)
cd64901 chore: add compression middleware to echo (#2249)
f18cb7d docs: Release notes for 0.15.1 (#2245)
ea0b372 chore: bump version: 0.15.1.dev0 -> 0.15.2.dev0
3c2186e Fix: remove quotes for Terraform 0.13 (#2231)
d23ebd0 docs: missing service-linked role [DET-5253] (#2221)
3cf12f4 feat: support configurable port and container name for Fluent Bit [DET-5272, DET-5273] (#2251)
6827390 chore: set the image used in ptl amp test through set_tf2_image (#2194)
259cff6 chore: trigger hp importance work on exp completion (#2248)
fe1540d chore: no pointers to maps or slices in expconf (#2238)
bd509a4 fix: TFKerasTrial check for tf2 behavior on 2.2.0 [DET-5277] (#2246)
afdb749 feat: per-resource-pool configs [DET-5173] (#2214)
30a6af3 chore: reset error field on hp importance success (#2242)
6c47b6c style: transpose hp heatmap to better align the plot axes (#2232)
55cfe8f ci: fix windows test with lmdb 1.2 (#2241)
eb61693 fix: update label picker when labels change [DET-5254] (#2239)
6ab2268 chore: submit partial hp importance work to pool (#2240)
c2f96cb chore: allow dev lint errors (#2218)
8de19c0 chore: fix panic from dependency creation race (#2233)
782d095 fix: select snapshot version with snapshot (#2235) [DET-5264]
cac640b chore: up circle ci timeout for e2e tests (#2237)
0b2c2c7 fix: actually support add/drop capabilities in structs (#2236)
1c926e0 chore: fix panics in hp importance actor (#2230) [DET-5263]
e21dc31 feat: support healthchecks for det deploy aws with TLS enabled. (#2207)
88b8689 docs: Release notes for 0.15.0. (#2225)
cd39429 chore: bump version: 0.15.0.dev0 -> 0.15.1.dev0
2fbe7d5 chore: bump version: 0.14.7.dev0 -> 0.15.0.dev0
9ef41f4 fix: gcp quota checks, A100 and docs. (#2220)
875ceb2 fix: fetch agents only when authenticated [DET-5259] (#2228)
7393b4f fix: force checkpoint GC to use the master's default environment (#2229)
524c895 fix: allow scatter plot to re-render when data changes initially (#2224)
a68ef2e ci: move test_hp_importance_api to distributed tests (#2212)
77b2c47 ci: stop using coscheduler for CI testing (#2216)
19e6b3e fix: correct crash when changing filter on multi-selected rows [DET-5258] (#2226)
67de93a chore: final touches to ExperimentConfig V0 (#2142)
923c8ac docs: Update custom environment example [DET-5196] (#2171)
7be726d chore: validate grid list enum value from local storage (#2223)
11e117a fix: TFKerasTrial on tf2 with tf.compat.v1.disable_v2_behavior. (#2211)

Docker images

• docker pull determinedai/determined-master:0.15.2
• docker pull determinedai/determined-master:89f3ee04
• docker pull determinedai/determined-master:89f3ee044b25619afd32e5faf62490c81c956837
• docker pull determinedai/determined-dev:determined-master-89f3ee04
• docker pull determinedai/determined-dev:determined-master-89f3ee044b25619afd32e5faf62490c81c956837
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:0.15.2
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:89f3ee04
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:89f3ee044b25619afd32e5faf62490c81c956837

0.15.1

Changelog

0e00289 chore: bump version: 0.15.1rc0 -> 0.15.1
a18a1eb Fix: remove quotes for Terraform 0.13 (#2231)
be490b5 chore: bump version: 0.15.1.dev0 -> 0.15.1rc0
5bc5826 chore: fix panic from dependency creation race (#2233)
2f73fbc chore: fix panics in hp importance actor (#2230) [DET-5263]
5dbee57 fix: select snapshot version with snapshot (#2235) [DET-5264]
922ff05 fix: TFKerasTrial on tf2 with tf.compat.v1.disable_v2_behavior. (#2211)
f8fd98d chore: bump version: 0.15.0 -> 0.15.1.dev0

Docker images

• docker pull determinedai/determined-master:0.15.1
• docker pull determinedai/determined-master:0e002898
• docker pull determinedai/determined-master:0e002898037e6a58ec764e42d5f4a611c35a718b
• docker pull determinedai/determined-dev:determined-master-0e002898
• docker pull determinedai/determined-dev:determined-master-0e002898037e6a58ec764e42d5f4a611c35a718b
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:0.15.1
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:0e002898
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:0e002898037e6a58ec764e42d5f4a611c35a718b

0.15.0

Changelog

3a04e69 chore: bump version: 0.15.0rc1 -> 0.15.0
3fc0fa6 docs: Release notes for 0.15.0. (#2225)
b9d60b0 chore: bump version: 0.15.0rc0 -> 0.15.0rc1
e7576a7 fix: force checkpoint GC to use the master's default environment (#2229)
1fa876b fix: allow scatter plot to re-render when data changes initially (#2224)
6c1d013 chore: validate grid list enum value from local storage (#2223)
cdaff4f chore: bump version: 0.15.0.dev0 -> 0.15.0rc0
9328c90 chore: bump version: 0.14.7.dev0 -> 0.15.0.dev0
99147a3 chore: lock api state for backward compatibility check
418d49a chore: bump to 2 agents on latest master [DET-5241] (#2219)
35ae080 test: add more lr scheduler tests for lightning [DET-5223] (#2184)
93ae755 chore: stop using cloudpickle to write PyTorch checkpoints [DET-5175] (#2204)
631dce2 docs: Various fixes (#2209)
0397c1c feat: add git and ide content to detignore by default [DET-2832] (#2210)
52e5755 chore: pull EE CLI features and docs into OSS [DET-3912] (#2195)
40b414c feat: move executables to the main package and update docs. (#2187)
26404b8 chore: Update to Ubuntu 20.04 for agent, master and bastion images [DET-5238] (#2208)
154e1c6 docs: clarify k8s default pod spec behavior (#2197)
f091d5f chore: avoid rerendering experiment list if api response remains the same (#2203)
364c0bc chore: show trial metrics on webui [DET-5060] (#2167)
c648f15 chore: update estimators test fixture to not reference adaptive searcher (#2205)
225ccde chore: remove sha [DET-5225] (#2181)
469fb25 refactor: consolidate global contexts (#2186)
66000fc fix: disable det deploy wait for aws cluster on circleci. (#2192)
7de7441 chore: extend timeout on HP importance test (#2193)
3ccd0ca chore: make the first glasbey color our brand color (#2190)
5e68b75 feat: add key tracker (#2188)
376baa3 feat: health check master after cluster creation [DET-5183] (#2164)
5a35fc1 chore: enable hyperparameter importance computation by default (#2159)
6aa7a04 refactor: improve experiment terminal state [DET-5202] (#2179)
bf43f63 chore: remove stoksc from codeowners (#2185)
92056b5 fix: add efs, fsx, and govcloud templates to bumpversion [DET-5200] (#2172)
4cd0612 fix: mmdetection docker image to work with torch 1.7 (#2183)
d979092 feat: local clusters to store checkpoint data in home [DET-5154] (#2170)
19ea607 fix: bug in random search leading to incorrect total trials (#2182)
8127681 chore: idempotent searcher progress API [DET-5211] (#2180)
b60ac4c feat: zoomed modal charts [DET-5111] (#2174)
fa9c773 chore: make default goal should be build. (#2177)
d241fa9 docs: various fixes (#2163)
66f75d4 fix: allow telemetry to be disabled under Helm (#2178)
07ce81a fix: default tooltip prefix to be an empty string (#2165)
bb7f1ea chore: tweak step_lr param in e2e tests (#2160)
d96aced chore: add an example for checkpoint callbacks [DET-5186] (#2173)
ad3e0a4 refactor: update context api to reduce unnecessary re-renders [DET-5185] (#2168)
255fa74 chore: store the daily/monthly filter setting in local storage for cluster historical usage page [DET-5194] (#2161)
9e3c69d chore: wire up profiling configurations [DET-5064] (#2122)
1b164fa chore: bump version: 0.14.6.dev0 -> 0.14.7.dev0
8a458c1 chore: add tab navigation to trial details page [DET-5070] (#2162)
14e9911 feat: det deploy check for sufficient gpu quotas on aws, gcp. (#2136)
f0faf47 chore: webui for resource allocation data [DET-5046] (#2062)
c636b45 chore: update codeowners (#2145)
4b7ef37 fix: add terraform files into default detignore [DET-5155] (#2146)
3537c21 fix: get cli wheel back into trail runner. (#2156)
e9892dd fix: fix an issue in wrapping lr_scheduler for lightningadapter (#2154)
8adc237 fix: roll back det and det-deploy executable move. (#2153)
5c8a17f fix: avoid loop of effect in hp-viz when experiment is not supported [DET-5189] (#2151)
95a8638 fix: e2e test for pytorch lightning examples (#2152)
d6bccf6 fix: pytorch lightning example (#2150)
323f272 fix: avoid showing no-data message for a split second in hp-viz [DET-5099] (#2144)
957ffd6 chore: add license information to Pytorch Lightning examples (#2147)

Docker images

• docker pull determinedai/determined-master:0.15.0
• docker pull determinedai/determined-master:3a04e697
• docker pull determinedai/determined-master:3a04e697706f25e6068b2bfe0f4ff3d9c8332ec9
• docker pull determinedai/determined-dev:determined-master-3a04e697
• docker pull determinedai/determined-dev:determined-master-3a04e697706f25e6068b2bfe0f4ff3d9c8332ec9
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:0.15.0
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:3a04e697
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:3a04e697706f25e6068b2bfe0f4ff3d9c8332ec9

0.14.6

Changelog

e472436 chore: bump version: 0.14.6rc4 -> 0.14.6
29d9988 docs: Release notes for 0.14.6. (#2158)
f7a4043 chore: bump version: 0.14.6rc3 -> 0.14.6rc4
2e8c364 fix: get cli wheel back into trail runner. (#2156)
f4b8e08 chore: bump version: 0.14.6rc2 -> 0.14.6rc3
0eaa03a fix: fix an issue in wrapping lr_scheduler for lightningadapter (#2154)
a451c2c fix: roll back det and det-deploy executable move. (#2153)
74b1b79 chore: bump version: 0.14.6rc1 -> 0.14.6rc2
8024fc5 fix: avoid loop of effect in hp-viz when experiment is not supported [DET-5189] (#2151)
05645af fix: e2e test for pytorch lightning examples (#2152)
78c7fe9 fix: pytorch lightning example (#2150)
9d5dff7 chore: add license information to Pytorch Lightning examples (#2147)
fb11f6b chore: bump version: 0.14.6rc0 -> 0.14.6rc1
6e22699 chore: bump version: 0.14.6.dev0 -> 0.14.6rc0
1722f28 feat: precision prop and amp support for lightning adapter [DET-5116] (#2127)
567e237 feat: add allocation aggregation by agent label and resource pool (#2141)
cd39ce7 chore: upgrade taiko [DET-5157] (#2134)
17f77ca perf: support max_concurrent_trials for random and grid search (#2137)
ae505e4 ci: adding coscheduler to static k8s test clusters, and test (#2139)
f5723da chore: move unets_tf_keras back to previous images (#2138)
ec8f02b chore: change output format of JSON aggregated resource data (#2129)
db16db4 style: trial log section filters [DET-5176] (#2133)
23b0945 fix: tweak aggregated resource allocation history endpoint (#2123)
b9d6b0c chore: downgrade dev pytorch package versions to 1.7.1 (#2135)
74be8a0 chore: Moving back to Python 3.7 and PyTorch 1.7.1 (#2132)
0b027d2 fix: package install order in requirements.txt (#2131)
169ad89 chore: ingest multiple batches to trial profiler metrics endpoint [DET-5178] (#2117)
eb45d53 fix: update scatter plot to support non-numeric values [DET-5110] (#2126)
8e13071 feat: rank hparams with hp importance [DET-5105] (#2086)
f172ed5 docs: improvements to spot instance and resource pool docs (#2113)
dc623d0 refactor: include det-deploy into det cli. [DET-5153] (#2124)
88044b0 chore: add webui tests lint step to CI (#2115)
59f48f3 feat: add pytorch checkpoint on load/save hooks [DET-5109] (#2118)
6f58a93 ci: put GPUs in a separate GKE node pool from master (#2120)
7e5c9c1 chore: remove cluster v1 page [DET-5163] (#2114)
6cb9445 chore: add server address to cli trial log download cmd [DET-5161] (#2116)
02598ce refactor: cleanup react hook dependencies [DET-5158, DET-5159, DET-5160] (#2112)
b0b03b2 style: update hp viz nav (#2093)
4b84772 refactor: combine common, cli, deploy into one python package. [DET-4756] (#2108)
b531bda build: local build improvements [DET-5118] (#2060)
202c485 feat: add trial profiler metrics APIs [DET-5065, DET-5059] (#2051)
fe87adc chore: add new ExperimentConfig objects (#2066)
3d2f54f chore: add frequency parameter to wrap_lr_scheduler [DET-5148] (#2087)
18c3994 chore: add pytorch-lightning to docs requirements (#2111)
60ab3d3 chore: Revert "Testing gang-scheduling [DET-5134]" (#2110)
52a7fb3 chore: update to new images including TensorFlow, PyTorch, Python and CUDA upgrades (#2074)
2c5beaa Testing gang-scheduling [DET-5134] (#2100)
08d9562 feat: expose resource allocation endpoints in CLI [DET-5045] (#2107)
c346301 feat: colorize output info of det-deploy [DET-4749] (#2102)
c42069d feat: add aggregated resource allocation endpoint and job [DET-5044] (#2085)
9116b36 docs: Release notes for 0.14.5. (#2098)
6470983 docs: Release notes for 0.14.4. (#2089)
726ae46 chore: bump version: 0.14.5.dev0 -> 0.14.6.dev0
83997e8 chore: bump version: 0.14.4.dev0 -> 0.14.5.dev0
4c59e05 fix: go mod tidy for releases (#2106)
4db2311 fix: master gen target (#2104)
8431039 fix: release helm chart (#2105)
78b9d94 docs: further improve k8s coscheduling docs (#2099)
b443bec fix: broken doc links [DET-5100] (#2101)
e9754de docs: add pytorch lightning adapter docs [DET-4800] (#2076)
620b5ac chore: add readmes for pl examples [DET-5149] (#2096)
2b28eedb docs: improve docs on k8s coscheduling plugin. (#2097)
cfab2e5 feat: add batch margins [DET-5073] (#2057)
1550417 feat: provide user hint on aws, gcp auth in det-deploy [DET-4846] (#2092)
4e4a874 fix: helm files naming consistency (#2095)
aefd9e3 feat: delete Tensorboards with delete API [DET-5143] (#2083)
ff192db fix: helm chart errors in absence of defaultScheduler value (#2094)
743f4c2 feat: priority-based gang-scheduling for k8s (#2091)
4daae95 chore: improve cors proxy [DET-5115] (#2047)
f03c581 docs: stopping-based variant of adaptive search (#2090)
f379582 feat: add DELETE /api/v1/experiments/:id API [DET-4022] (#2056)

Docker images

• docker pull determinedai/determined-master:0.14.6
• docker pull determinedai/determined-master:e4724367
• docker pull determinedai/determined-master:e47243675c660cbd571c78333e7ceece5f1db447
• docker pull determinedai/determined-dev:determined-master-e4724367
• docker pull determinedai/determined-dev:determined-master-e47243675c660cbd571c78333e7ceece5f1db447
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:0.14.6
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:e4724367
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:e47243675c660cbd571c78333e7ceece5f1db447

0.14.5

Changelog

f16dc9f chore: bump version: 0.14.5rc1 -> 0.14.5
60c225e fix: go mod tidy for releases (#2106)
5335fe9 Revert "chore: bump version: 0.14.5rc1 -> 0.14.5"
b84046c chore: bump version: 0.14.5rc1 -> 0.14.5
448f602 Revert "chore: bump version: 0.14.5rc1 -> 0.14.5"
46813ed fix: master gen target (#2104)
9f86bb8 fix: release helm chart (#2105)
ff5303f chore: bump version: 0.14.5rc1 -> 0.14.5
a6f0bfd chore: bump version: 0.14.5rc0 -> 0.14.5rc1
2e2cceb Revert "chore: bump version: 0.14.5rc0 -> 0.14.5"
2d16658 docs: further improve k8s coscheduling docs (#2099)
c4508da fix: broken doc links [DET-5100] (#2101)
541e5ac chore: bump version: 0.14.5rc0 -> 0.14.5
493f037 docs: Release notes for 0.14.5. (#2098)
05012f2 chore: bump version: 0.14.5.dev0 -> 0.14.5rc0
0d861f0 feat: add batch margins [DET-5073] (#2057)
b57ca59 chore: bump version: 0.14.4 -> 0.14.5.dev0
c26615c chore: bump version: 0.14.4rc3 -> 0.14.4
515162c docs: Release notes for 0.14.4. (#2089)

Docker images

• docker pull determinedai/determined-master:0.14.5
• docker pull determinedai/determined-master:f16dc9f1
• docker pull determinedai/determined-master:f16dc9f1191a6e9b1b5c992ac39c6761ed176e20
• docker pull determinedai/determined-dev:determined-master-f16dc9f1
• docker pull determinedai/determined-dev:determined-master-f16dc9f1191a6e9b1b5c992ac39c6761ed176e20
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:0.14.5
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:f16dc9f1
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:f16dc9f1191a6e9b1b5c992ac39c6761ed176e20

0.14.3

Changelog

d28d851 chore: bump version: 0.14.3rc3 -> 0.14.3
a9a778e chore: bump version: 0.14.3rc2 -> 0.14.3rc3
710319b fix: support trial filtering based on categorical hps [DET-5107] (#2045)
c94c9f5 fix: trial log viewer scrolling issues [DET-5096] (#2044)
3010fd5 fix: nas dtrain bug (#2039)
51bf576 chore: tweak resource allocation endpoint (#2040)
9e409b2 docs: Release notes for 0.14.3. (#2046)
ec29f4b chore: bump version: 0.14.3rc1 -> 0.14.3rc2
01680a4 fix: trial log viewer disappearing when re-clicking direction button [DET-5097] (#2043)
37f8d81 fix: decode categorical config hp properly [DET-5098] (#2042)
af5c6b8 fix: correct polling issues [DET-5095] (#2036)
37b0b75 docs: fix version of CUDA 11 image (#2037)
6404731 chore: remove petname dependency (#2035)
9531146 chore: bump version: 0.14.3rc0 -> 0.14.3rc1
98ddbe2 chore: bump version: 0.14.3.dev0 -> 0.14.3rc0
26fc1a9 chore: lock api state for backward compatibility check
b5b7a28 feat: add endpoint for raw resource allocation information [DET-5043] (#2026)
229a60c feat: scatter plot and heat map [DET-4453, DET-4459, DET-5085] (#2007)
e7cd561 chore: support new tags and gov images in bumpenvs (#2033)
1f6e3a5 fix: bug in deformable detr experiment config (#2031)
231ed9e chore: warn on failures to connect to persitent storage (#2030)
0e48bec fix: update experiment trials endpoint when changing filters or sorting (#2029)
a3d168b feat: HP importance implementation [DET-4465] (#1965)
2992d5b fix: prevent stray logs during checkpoint loading (#2028)
90e4d4f feat: Support PyTorch-native Automatic Mixed Precision [DET-4753] (#1914)
1423f1e fix: fix shim for stateless searchers (grid, single, etc..) (#2004)
744c4ad fix: detect parcoords filter removal properly (#2022)
867dd84 feat: deformable detr with coco (#1817)
4806f1b refactor: DETR example to use custom reducer and support finetuning (#1816)
623ab49 fix: fix forking failure on k8s [DET-5087] (#2021)
060a2b4 chore: exclude useless GPU type in resource pool API [DET-5079] (#2016)
b42dc3a docs: default agent docker image (#1977)
05d8fcd chore: update contributing.md with helm (#2019)
717b727 docs: adding page for custom k8s master [DET-5022] (#2013)
afbb982 fix: navigation disappearing when going back from wait page [DET-5020] (#2018)
9c85ce1 chore: webui uniform path generation [DET-4734] (#2010)
6cb259e chore: make experiment list page url sharable [DET-4745] (#1999)
4310747 refactor: remove app contexts [DET-2878, DET-4820] (#1996)
dd648f3 fix: remove react extra get-deps dependency target (#2015)
216f5ca feat: example using hp constraints for NAS (#2014)
66c5559 docs: HP Search Constraints Documentation [DET-4392, DET-4993] (#1998)
5542e11 build: build swagger api bindings [DET-5056] (#2011)
33252c1 chore: refactor cifar10_tf_keras example (#2006)
f5ca28b fix: fix deprecation warning on using ABCs from 'collections' (#1997)
fdc7c96 fix: fix query for GET experiments [DET-5039] (#2012)
0228826 fix: use UTC for all logs (#1985)
e00ab75 fix: correct batches column order on trial detail page [DET-5023] (#2003)
4a386cb docs: add link for enabling Kubernetes GPU support (#2008)
b397cfd fix: fix v0 experiment snapshot shim (#2002)
2824704 chore: update cluster address for react preview [DET-5008] (#1989)
e085b91 feat: use anchor tags for navigation in tables rows [DET-4746] (#1990)
81d7800 ci: bump gke version (#2000)
96203a2 docs: fix missing parens in a few checkpoint API code snippets (#2001)
9763255 docs: add documentation regarding limit behavior for pagination (#1995)
ba47fe0 fix: call Sequence.on_epoch_end after validation (#1991)
19365f6 build: add instructions to point webui to cors disabled clusters (#1988)
c0e3a66 chore: refactor trial log viewer to improve rendering performance [DET-4866] (#1974)
b4f3be3 feat: enable parcoords in webui [DET-5013] (#1994)
3c18d5d chore: tune parcoords [DET-4991] (#1973)
e4f010d feat: add v0 experiment config objects in python (#1966)
1c0cbdc chore: support task tokens in harness for authentication [DET-4897] (#1894)
b98ad55 chore: remove searcher emitted checkpoints [DET-4996] (#1972)
2ea6d16 chore: fix typo in comments (#1986)
7af2c1e feat: HP constraints harness exception handling [DET-4867] (#1875)
acc9989 chore: bump version: 0.14.2.dev0 -> 0.14.3.dev0
9043ebd docs: Release notes for 0.14.2. (#1983)
a93fbfc chore: remove parcoords temporarily (#1982)
dabbea4 fix: fix notebook state stuck in pending [DET-4988] (#1981)
e6fd27a chore: check if experiment exists in /api/v1/experiments/:id/trials (#1978)
666fed0 fix: k8s resource pool API response should not have nil field [DET-4989] (#1979)
6e75213 fix: fix quoted string bugs for non-simple AWS deployments [DET-5001] (#1980)

Docker images

• docker pull determinedai/determined-master:0.14.3
• docker pull determinedai/determined-master:d28d851d
• docker pull determinedai/determined-master:d28d851d4f6409660103a4ba29c39e2fcf71c499
• docker pull determinedai/determined-dev:determined-master-d28d851d
• docker pull determinedai/determined-dev:determined-master-d28d851d4f6409660103a4ba29c39e2fcf71c499
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:0.14.3
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:d28d851d
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:d28d851d4f6409660103a4ba29c39e2fcf71c499

0.14.2

Changelog

44e76a6 chore: bump version: 0.14.2rc3 -> 0.14.2
6045694 docs: Release notes for 0.14.2. (#1983)
8d5cb5b chore: bump version: 0.14.2rc2 -> 0.14.2rc3
fa45890 fix: fix notebook state stuck in pending [DET-4988] (#1981)
90256ac chore: remove parcoords temporarily (#1982)
0bfe2c9 chore: bump version: 0.14.2rc1 -> 0.14.2rc2
bfb6805 fix: fix quoted string bugs for non-simple AWS deployments [DET-5001] (#1980)
68e4513 fix: k8s resource pool API response should not have nil field [DET-4989] (#1979)
e50cdcb chore: check if experiment exists in /api/v1/experiments/:id/trials (#1978)
f3ca29c chore: bump version: 0.14.2rc0 -> 0.14.2rc1
98491e9 chore: bump version: 0.14.2.dev0 -> 0.14.2rc0
3df38be fix: helm chart template dashes (#1976)
9592b64 ci: deploy NGC images as part of release [DET-4910] (#1941)
1008645 chore: fix squad example README links (#1975)
94b214f chore: lock api state for backward compatibility check
5e3cb06 docs: Release notes for 0.14.1.
ac5cdc0 chore: bump version: 0.14.1.dev0 -> 0.14.2.dev0
2a34e79 chore: bump version: 0.14.0.dev0 -> 0.14.1.dev0
8e33f2a fix: add task state to task endpoints [DET-4987] (#1958)
b424102 chore: update documentation for new images [DET-4998] (#1971)
6c4cd8c fix: restore trials with operations intact to avoid unintented side-effects (#1970) [DET-4994]
f103ab8 chore: stop hardcoding python3.6 [DET-4929] [DET-4539] (#1967)
f02715d chore: commit generated python schemas (#1964)
2ce2ff0 docs: add DGX instructions for determined-deploy (#1955)
28f3c7a ci: lint schemas during ci (#1963)
5c7e279 chore: tweak json-schema tooling, schemas, and test cases (#1957)
b2d15aa chore: refactor trial logs CLI code and remove old api endpoints (#1951)
aef46c7 chore: fix code quality issues (#1940)
8224fa7 fix: add missing comment on exported function (#1962)
285c055 chore: commit make -C schemas fmt (#1961)
2cff862 chore: check for closed connections before sending [DET-4829] (#1949)
b62a0b1 chore: remove lunch-and-learn notice, tweak docs (#1960)
321b1aa chore: migrate webui launch notebook api [DET-4821] (#1943)
318bbf8 ci: fix retention of CloudWatch logs for deploy clusters (#1954)
d149986 ci: set a CI-specific prefix for and retain CloudWatch log groups (#1953)
e951363 chore: remove max agents count in resource pool card's header (#1950)
139b149 feat: allow retaining CloudWatch log groups of det-deploy clusters (#1948)
78d4b94 build: separate out docs pre-publish step (#1945) [DET-4982]
493606f chore: fix error on empty metric in metric-snapshot endpoint [DET-4923] (#1947)
84d979b chore: rename v1's to v0's (#1946)
07be510 refactor: replace step ID with total batches in the DB (#1890)
e1fbb11 ci: don't upload cloudwatch to s3 (#1938)
0c79b35 chore: tooling for experiment config structs in python and golang (#1874)
eedbbe8 refactor: reduce code duplication across task types (#1749)
3184552 fix: make trial log timestamp filters backwards compatible (#1944)
c607134 chore: bump version: 0.13.14.dev0 -> 0.13.14rc0
89acacb docs: Release notes for 0.14.0.
648c7ae fix: add backwards compatability for logs before 0.13.8 (#1942)
2302f94 feat: parallel coord plot [DET-4450, DET-4928, DET-4978] (#1878)
fea363e fix: clean up device allocation if the resources are released (#1930)
0a36789 ci: limit concurrent deploys from master (#1905)
870d80c chore: single batch evaluation for local test mode [DET-2931] (#1934)
71b5139 test: fix resourcepoolcard story not rendering [DET-4964] (#1916)
44cfe76 chore: remove unused step iotypes (#1933)
da17313 fix: show zero value metrics [DET-4686] (#1932)
283a188 chore: migrate webui tasks api endpoint [DET-4018] (#1900)
43180e3 refactor: array prototype [DET-4976] (#1935)
7a6ce63 chore: revert default images and framework versions (#1936)
ceba9ea docs: edit model debug doc (#1925)
94265ad chore: improve resource pool details presentation [DET-4968] (#1926)
cd12430 fix: remove clickable style from trial info table [DET-4967] (#1923)
3e4b74e docs: clarification instructions for TF 1.15 eager mode (#1929)
e92d612 test: webui e2e test improvements [DET-4912] (#1893)
c796178 fix: correct the comparison function when numbers are fractions [DET-4969] (#1924)
80924f9 fix: add default query limit and add missing sort by state [DET-4919] (#1921)
e830758 refactor: paginate experiment trials [DET-4900, DET-4921, DET-4922] (#1892)
6a0c76d fix: correct cancel confirm button label to confirm [DET-4966] (#1922)
f8d061b style: change notebook create icon [DET-4965] (#1920)
6c123e3 docs: fix incorrect reference in docs (#1919)
be35c4c fix: typos in model debugging doc (#1918)
76a3b94 fix: fix a utilization calculation error in hgi resource bar for cpu slots [DET-4913] (#1911)
9e0be99 docs: clean up resource pool docs and add release notes (#1917)
3e5343c feat: support resource pools in det-deploy local agent-up [DET-4938] (#1906)
436d774 refactor: update active experiments [DET-4915] (#1910)
c036414 chore: remove docs for deprecated API (#1908)
f55a4dd fix: buffer the trial log in the correct order [DET-4931] (#1912)
167b114 fix: correct best and latest metric sort by params for the GET experiment trials API (#1915) [DET-4920]
48418e6 fix: add non-scalar metric expectation to protobufs [DET-4893] [DET-4911] (#1876)
7f7aa2e feat: support more fields to sortBy in /api/v1/experiments/trials [DET-4219, DET-4920] (#1899)
4f4bca4 chore: Bump images and versions to Tensorflow 2.4.1 (#1913)
b619197 chore: let CLI verify the master using combined system/custom certs (#1859) [DET-4666]
a00492c docs: add model debug doc (#1895)
6078b0f chore: reword resource pool ui presentation [DET-4925] (#1898)

Docker images

• docker pull determinedai/determined-master:0.14.2
• docker pull determinedai/determined-master:44e76a69
• docker pull determinedai/determined-master:44e76a69ed0aba35f1a2e03aff7928dbc90c94da
• docker pull determinedai/determined-dev:determined-master-44e76a69
• docker pull determinedai/determined-dev:determined-master-44e76a69ed0aba35f1a2e03aff7928dbc90c94da
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:0.14.2
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:44e76a69
• docker pull nvcr.io/isv-ngc-partner/determined/determined-master:44e76a69ed0aba35f1a2e03aff7928dbc90c94da

0.14.1

Changelog

875429b chore: bump version: 0.14.1rc2 -> 0.14.1
cbf814f chore: bump version: 0.14.1rc1 -> 0.14.1rc2
0c229de fix: make trial log timestamp filters backwards compatible (#1944)
461288d chore: bump version: 0.14.1rc0 -> 0.14.1rc1
a695609 chore: bump version: 0.14.1.dev0 -> 0.14.1rc0
e9d51cd chore: bump version: 0.14.0 -> 0.14.1.dev0
6a90217 docs: Release notes for 0.14.1.
3e00128 fix: add backwards compatability for logs before 0.13.8 (#1942)
db67b27 docs: More changes to release notes for 0.14.0. (#1927)

Docker images

• docker pull determinedai/determined-master:0.14.1
• docker pull determinedai/determined-master:875429b1
• docker pull determinedai/determined-master:875429b1b96bedcdd0a15bbb5f40a1957e00ee6e
• docker pull determinedai/determined-dev:determined-master-875429b1
• docker pull determinedai/determined-dev:determined-master-875429b1b96bedcdd0a15bbb5f40a1957e00ee6e