Skip to content

Actions: NovaSky-AI/SkyRL

Actions

SkyRL-GPU

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
gpu_skyrl.yaml
1,790 workflow runs
1,790 workflow runs

Filter by Event

⌥ + click/return to exclude

Filter by Status

⌥ + click/return to exclude

Filter by Branch

⌥ + click/return to exclude

Filter by Actor

⌥ + click/return to exclude
[feat] Add a Tinker-compatible forward(loss_fn=...) API to SkyRL's FSDP and Megatron policy workers and refactor worker outputs SkyRL-GPU #1790: Pull request #1667 synchronize by SumanthRH
4m 26s pr1-forward-loss
pr1-forward-loss
4m 26s
[feat] Add a Tinker-compatible forward(loss_fn=...) API to SkyRL's FSDP and Megatron policy workers and refactor worker outputs SkyRL-GPU #1789: Pull request #1667 synchronize by SumanthRH
4m 28s pr1-forward-loss
pr1-forward-loss
4m 28s
[CI] add E2E CI for SFT on the Tulu3 dataset with the Megatron backend SkyRL-GPU #1788: Pull request #1670 opened by SumanthRH
5m 13s add-sft-ci
add-sft-ci
5m 13s
[tinker] Abort/retry recovery for forwarded multi-tenant LoRA samples SkyRL-GPU #1787: Pull request #1669 opened by erictang000
33s erictang000:multi-tenant-async-pause-tinker-forwarding
erictang000:multi-tenant-async-pause-tinker-forwarding
33s
[feat] Add support for evaluation dataset in SFTTrainer SkyRL-GPU #1786: Pull request #1668 opened by SumanthRH
4m 34s pr2-sft-eval
pr2-sft-eval
4m 34s
[feat] Add a Tinker-compatible forward(loss_fn=...) API to SkyRL's FSDP and Megatron policy workers and refactor worker outputs SkyRL-GPU #1785: Pull request #1667 opened by SumanthRH
5m 41s pr1-forward-loss
pr1-forward-loss
5m 41s
[multi-tenant-lora] Expose per-LoRA pause state over HTTP SkyRL-GPU #1784: Pull request #1666 opened by erictang000
26s erictang000:multi-tenant-async-pause-server
erictang000:multi-tenant-async-pause-server
26s
[models] Add nemotron-nano-30b-a3b to CI SkyRL-GPU #1783: Pull request #1603 synchronize by erictang000
23s erictang000:nemotron3_nano_vllm020
erictang000:nemotron3_nano_vllm020
23s
[lora][tinker] Add pause and resume for multi-tenant lora (#1657) SkyRL-GPU #1782: Commit 77f92c2 pushed by erictang000
4m 44s main
main
4m 44s
[tinker] Forward sample requests directly to backend vLLM (non-coloca… SkyRL-GPU #1781: Commit 29f0fdf pushed by erictang000
6m 26s main
main
6m 26s
[cleanup] Remove FSDP1 support + make 'fsdp' default to fsdp2 (#1659) SkyRL-GPU #1780: Commit 2751aab pushed by erictang000
26m 7s main
main
26m 7s
[vllm] test out ray v2 executor SkyRL-GPU #1779: Pull request #1660 opened by erictang000
21s erictang000:test_ray_v2_executor
erictang000:test_ray_v2_executor
21s
[cleanup] Remove FSDP1 support + make 'fsdp' default to fsdp2 SkyRL-GPU #1778: Pull request #1659 synchronize by erictang000
30s erictang000:remove_fsdp1
erictang000:remove_fsdp1
30s
[lora][tinker] Add pause and resume for multi-tenant lora SkyRL-GPU #1777: Pull request #1657 synchronize by hao-aaron
29s hao-aaron:multi-lora-pause
hao-aaron:multi-lora-pause
29s
[lora][tinker] Add pause and resume for multi-tenant lora SkyRL-GPU #1776: Pull request #1657 synchronize by hao-aaron
29s hao-aaron:multi-lora-pause
hao-aaron:multi-lora-pause
29s
[CI] Add E2E CI for Tinker API with SkyRLTrainBackend (#1616) SkyRL-GPU #1775: Commit 4e0ba0e pushed by SumanthRH
4m 20s main
main
4m 20s
[cleanup] Remove FSDP1 support + make 'fsdp' default to fsdp2 SkyRL-GPU #1774: Pull request #1659 opened by erictang000
23s erictang000:remove_fsdp1
erictang000:remove_fsdp1
23s
[lora][tinker] Add pause and resume for multi-tenant lora SkyRL-GPU #1773: Pull request #1657 opened by hao-aaron
34s hao-aaron:multi-lora-pause
hao-aaron:multi-lora-pause
34s
[train] Fix chunked-logprob backward and skip chunking for short sequences SkyRL-GPU #1772: Pull request #1650 synchronize by SumanthRH
4m 7s fix-chunk-logprobs-small-seq
fix-chunk-logprobs-small-seq
4m 7s
[tinker] Forward sample requests directly to backend vLLM (non-colocated) SkyRL-GPU #1771: Pull request #1638 synchronize by erictang000
56s erictang000:async_sample_routing_clean
erictang000:async_sample_routing_clean
56s
[feat] Add prometheus logging to new inference, add vllm metrics to w… SkyRL-GPU #1770: Commit fb87f35 pushed by SumanthRH
8m 1s main
main
8m 1s
[feat] Add prometheus logging to new inference, add vllm metrics to wandb SkyRL-GPU #1769: Pull request #1636 synchronize by hao-aaron
24s hao-aaron:vllm-metrics
hao-aaron:vllm-metrics
24s
[feat] Add prometheus logging to new inference, add vllm metrics to wandb SkyRL-GPU #1768: Pull request #1636 synchronize by hao-aaron
29s hao-aaron:vllm-metrics
hao-aaron:vllm-metrics
29s
[feat] Add prometheus logging to new inference, add vllm metrics to wandb SkyRL-GPU #1767: Pull request #1636 synchronize by hao-aaron
22s hao-aaron:vllm-metrics
hao-aaron:vllm-metrics
22s
[models] Add nemotron-nano-30b-a3b to CI SkyRL-GPU #1766: Pull request #1603 synchronize by erictang000
29s erictang000:nemotron3_nano_vllm020
erictang000:nemotron3_nano_vllm020
29s