Skip to content

Actions: OpenPipe/ART

Actions

Release

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
release.yml
214 workflow runs
214 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
DeepEP, torch.compile and Fix Megatron Training Bug Release #218: Pull request #646 closed by FurtherAI
1s austin/deepep_compile_and_trainability_main
austin/deepep_compile_and_trainability_main
1s
Fix flex attention compilation failures in Megatron training Release #217: Pull request #647 closed by Kovbo
1s fix-flex-attention-compilation
fix-flex-attention-compilation
1s
Add Megatron SFT internal checkpoints Release #216: Pull request #645 closed by Kovbo
1s codex/megatron-sft-internal-checkpoints
codex/megatron-sft-internal-checkpoints
1s
Share training code between backends Release #215: Pull request #626 closed by Kovbo
2s feat/shared-training-code
feat/shared-training-code
2s
[codex] Restore PR 607 KL loss removal Release #214: Pull request #639 closed by FurtherAI
1s austin/fix_pr_607_regression_from_pr_619
austin/fix_pr_607_regression_from_pr_619
1s
fix: Qwen3.5 tool-call chat-template tokenization Release #213: Pull request #634 closed by vivekkalyan
1s fix/qwen3.5-tool-call-chat-template
fix/qwen3.5-tool-call-chat-template
1s
Update RL training skill Release #212: Pull request #633 closed by Kovbo
1s fix/update-train-skills
fix/update-train-skills
1s
feat: add Novita AI as LLM provider Release #211: Pull request #630 closed by Alex-wuhu
Action required Alex-wuhu:novita-integration
Alex-wuhu:novita-integration
Action required
Megatron LoRA correctness: align distributed semantics with Megatron and validate TP/EP/ETP/DP against an oracle Release #210: Pull request #619 closed by FurtherAI
1s austin/megatron_lora_correctness_oracle_tests
austin/megatron_lora_correctness_oracle_tests
1s
feat: Add merged vLLM rollout weights Release #209: Pull request #631 closed by vivekkalyan
1s feat/merged-inference
feat/merged-inference
1s
feat: install-skills creates CLAUDE.md for AI discoverability Release #208: Pull request #574 closed by Kovbo
1s Kovbo:feat/agentic-exp
Kovbo:feat/agentic-exp
1s
feat: Upgrade to transformers v5 Release #207: Pull request #629 closed by vivekkalyan
1s feat/transformers-v5
feat/transformers-v5
1s
fix: Respect PipelineTrainer max_batch_size Release #206: Pull request #622 closed by vivekkalyan
1s fix/pipeline-max-batch
fix/pipeline-max-batch
1s
feat: Support PipelineTrainer with dedicated LocalBackend Release #205: Pull request #621 closed by vivekkalyan
1s feat/pipeline-localbackend
feat/pipeline-localbackend
1s
fix: auto-log Tinker inference costs from ART managed openai_client Release #204: Pull request #620 closed by vivekkalyan
1s fix/auto-inference-costs
fix/auto-inference-costs
1s
Fix W&B run reuse across ART models Release #203: Pull request #618 closed by Kovbo
1s fix/wandb-multi-run
fix/wandb-multi-run
1s
Release v0.5.17 Release #202: Pull request #616 closed by vivekkalyan
40s release/v0.5.17
release/v0.5.17
40s
feat: Add W&B run config API Release #201: Pull request #615 closed by vivekkalyan
1s feat/wandb-config
feat/wandb-config
1s
feat: add tenant-scoped Tinker model aliases Release #200: Pull request #614 closed by bradhilton
1s feat/tinker-model-alias-api
feat/tinker-model-alias-api
1s
feat: Update Tinker renderers Release #199: Pull request #613 closed by vivekkalyan
3s feat/add-qwen35-tinker-main-v2
feat/add-qwen35-tinker-main-v2
3s
fix: Add pyarrow to Tinker extra Release #198: Pull request #611 closed by vivekkalyan
2s fix/pyarrow-dep
fix/pyarrow-dep
2s
feat: Improved metrics in ART Release #197: Pull request #609 closed by vivekkalyan
1s feat/improved-metrics
feat/improved-metrics
1s
build: Upgrade vLLM to 0.17.0 Release #196: Pull request #610 closed by vivekkalyan
1s build/upgrade-vllm-0.17.0
build/upgrade-vllm-0.17.0
1s
feat: upgrade vLLM to 0.16.0 Release #195: Pull request #584 closed by vivekkalyan
1s feat/vllm-0.16.0
feat/vllm-0.16.0
1s
Remove beta KL divergence from training loss Release #194: Pull request #607 closed by arcticfly
1s fix/remove-beta-kl-divergence
fix/remove-beta-kl-divergence
1s