v1.5.0

What's Changed

• Fix NCCL_SOCKET_IFNAME typo in values.yaml under nccltest/gke by @hmhv1222 in #357
• Replace hardcoded parameters with environment variables in litgpt_container.sh by @samcmho in #359
• Update setup_and_launch_training.sh by @samcmho in #361
• Update README.md for all customers to cover all-to-all by @samcmho in #362
• remove default node pool deletion by @stevenBorisko in #351
• Pirillo/litgpt nvtx by @Chris113113 in #354
• Update lit_gpt commit to PyTorch 2.2 by @Chris113113 in #364
• Update setup_and_launch_training.sh by @samcmho in #363
• Merging Develop -> Main for sample_workloads changes by @Chris113113 in #366
• Add A3-Megagpu-8g SKU to tool. by @Chris113113 in #367
• Update NCCL link and rename a3-mega GKE in terraform module by @samcmho in #370
• Add A3-Mega support by @Chris113113 in #371
• Bump version to 1.5.0 by @Chris113113 in #373

Full Changelog: v1.4.2...v1.5.0

v1.4.2

What's Changed

• Merging main to develop by @soumyapani in #327
• staging changes for litgpt by @Chris113113 in #330
• Add litgpt readme by @Chris113113 in #332
• Small fixes to Lit-GPT demo by @gkroiz in #334
• remove profiling setup (currently not used) by @gkroiz in #335
• Adding details to explain MFU calculation by @parambole in #339
• Update rxdm image version by @Chris113113 in #341
• Fix unsupported envvar are set for SLURM cluster #343 by @parambole in #344
• Adding SLURM scripts to setup and launch lit-gpt training by @parambole in #342
• Adding a simple Multi-Node Pingpong PyTorch Workload by @parambole in #347
• Update litgpt LKG, more params for injection by @Chris113113 in #348
• Add a nccl-test sample workload by @Chris113113 in #345

New Contributors

• @parambole made their first contribution in #339

Full Changelog: v1.4.1...v1.4.2

v1.4.1

What's Changed

• Specify tcpx image versions in mig-cos by @soumyapani in #326

Full Changelog: v1.4.0...v1.4.1

v1.4.0

What's Changed

• litgpt demo by @Chris113113 in #319
• litgpt demo with tcpx lkg by @gkroiz in #320
• Add env var injection of num_nodes, datapath, batchsize, mbs by @Chris113113 in #321
• Adding support for existing resource policy. by @soumyapani in #322
• updates to fix issues with sb cluster creation by @stevenBorisko in #323

New Contributors

• @Chris113113 made their first contribution in #319
• @gkroiz made their first contribution in #320

Full Changelog: v1.3.0...v1.4.0

Release V1.3.0

What's Changed

• Borisko/update mtu by @stevenBorisko in #314
• Borisko/disable deletion protection by @stevenBorisko in #315
• Adding support for A2 MIGs in CPT by @soumyapani in #316

Full Changelog: v1.2.1...v1.3.0

Release v1.2.1

What's Changed

• Enabling continuous test integration with GCB by @soumyapani in #297
• Using fixup daemonset for TCPX support and secodary IP range in gke-beta by @soumyapani in #304
• Fixing test failure for secondary Ip range. by @soumyapani in #309
• Fixing continuous test break. by @soumyapani in #311

Full Changelog: v1.1.0...v1.2.1

Release v1.1.0

What's Changed

• Adding support for using stable fleet in GKE node pools. by @soumyapani in #292
• Adding capability to resize node pool in GKE beta. by @soumyapani in #293

Full Changelog: v1.0.1...v1.1.0

v1.0.1

What's Changed

• fix example urls by @stevenBorisko in #289

Full Changelog: v1.0.0...v1.0.1

v1.0.0

What's Changed

• Adding support for GKE BYOPP. by @soumyapani in #283
• split repo by machine type by @stevenBorisko in #284
• update gke to have installer daemonsets by @stevenBorisko in #286

Full Changelog: v0.15.0...v1.0.0

v0.15.0

What's Changed

• Reducing GKE nodepool name by removing prefix. by @soumyapani in #278
• Supporting multi-NIC network for GKE cluster by @soumyapani in #279
• Borisko/a3 docs by @stevenBorisko in #280

Full Changelog: v0.14.0...v0.15.0