Skip to content

Pull requests: PrimeIntellect-ai/prime-rl

New pull request New
150 Open 2,668 Closed
150 Open 2,668 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

chore: bump verifiers and research-environments to latest main
#3172 opened Aug 1, 2026 by mikasenghaas Member • Draft
exp: add Laguna XS ScaleSWE training config
#3171 opened Aug 1, 2026 by rasdani Contributor • Draft
feat(orchestrator): episode- and agent-level rollout metrics
#3165 opened Jul 30, 2026 by mikasenghaas Member • Draft
fix: checkpoint and resume the orchestrator train data position
#3164 opened Jul 30, 2026 by mikasenghaas Member Loading…
6
feat(orchestrator): per-agent metric slices, and keep reward metrics to the policy
#3163 opened Jul 30, 2026 by hallerite Member • Draft
feat!: remove env-server sidecaring, launcher spawns env servers at deterministic addresses
#3162 opened Jul 30, 2026 by mikasenghaas Member • Draft
14
fix: token_id-formatted logprob tokens in the qwen3-vl fake engine
#3161 opened Jul 29, 2026 by mikasenghaas Member • Draft
perf(docker): cut ghcr image build from 74m to 16m
#3160 opened Jul 29, 2026 by eexwhyzee Collaborator Loading…
Add custom vendored kernels
#3154 opened Jul 29, 2026 by MarioSieg Member • Draft
2
feat: add NVFP4 grouped expert recipe
#3143 opened Jul 25, 2026 by S1ro1 Collaborator • 2/2 • Draft
feat(moe): fuse gate and up expert weights at runtime
#3141 opened Jul 25, 2026 by S1ro1 Collaborator • 1/2 Loading…
5
fix(orchestrator): fail fast when an env server dies
#3139 opened Jul 24, 2026 by eligotts Contributor Loading…
4
Support v1 inline raw multimodal images
#3128 opened Jul 23, 2026 by eligotts Contributor Loading…
2
fix(ci): allow branch/commit refs in ghcr image build
#3126 opened Jul 23, 2026 by eexwhyzee Collaborator Loading…
5
fix: eval at checkpoint step on resume instead of next interval boundary
#3123 opened Jul 23, 2026 by samsja Member Loading…
3
feat: add all torchtitan EP backends (deepep_v2, hybridep, minimal_async_ep)
#3122 opened Jul 23, 2026 by samsja Member • Draft
1
feat(kernels): add SM100 NVFP4 grouped GEMM
#3120 opened Jul 23, 2026 by S1ro1 Collaborator • Draft
feat(weight-transfer): support router-discovered NIXL autoscaling
#3119 opened Jul 23, 2026 by S1ro1 Collaborator • Draft
feat(wandb): log full RLConfig to W&B instead of per-component subconfig
#3110 opened Jul 22, 2026 by samsja Member • Draft
2
feat: chunked per-layer optimizer step with stream-overlapped CPU offload
#3109 opened Jul 22, 2026 by samsja Member Loading…
17
fix(orchestrator): gate group-relative algorithms to a single trainable agent
#3108 opened Jul 22, 2026 by hallerite Member Loading…
3
fix(orchestrator): drain after final train batch
#3107 opened Jul 22, 2026 by jooooonas Loading…
2
fix: replace wandb_gql import with execute_graphql for wandb>=0.27.1 compat
#3086 opened Jul 19, 2026 by TM23-sanji Loading…
5
fix(sft): tolerate trace-shape tool calls; opt-in skip for unrenderable samples
#3081 opened Jul 18, 2026 by mikasenghaas Member Loading…
feat: DSA conversion (dense-MLA -> DeepSeek Sparse Attention), with a Kimi K2 example
#3066 opened Jul 17, 2026 by faresobeid Contributor • Draft
ProTip! Exclude everything labeled bug with -label:bug.