-
Notifications
You must be signed in to change notification settings - Fork 182
Pull requests: radixark/miles
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
[DO NOT MERGE] CI test run
run-ci-megatron
#1066
opened May 4, 2026 by
guapisolo
Collaborator
Loading…
[TITO] model-support: add DeepSeek V4 TITO support
#1065
opened May 4, 2026 by
zyzshishui
Contributor
Loading…
[TITO] feat (4/N): Use prompt ids from miles session server
#1064
opened May 4, 2026 by
zyzshishui
Contributor
Loading…
[feat] Add per-phase profiling for rollout, train, and weight updates
#1060
opened May 1, 2026 by
Xinyu-Kang
Loading…
[feat] Init true on policy with qwen_moe
#1059
opened May 1, 2026 by
maocheng23
Contributor
•
Draft
5 tasks
support kimi 2.5/6 full param and lora
run-ci-lora
run-ci-megatron
#1057
opened Apr 30, 2026 by
nanjiangwill
Loading…
[feat] Init true on policy with qwen_dense
#1052
opened Apr 28, 2026 by
maocheng23
Contributor
Loading…
5 tasks
build(setup): support non-x86 architectures in wheel build
#1042
opened Apr 23, 2026 by
garrett361
Loading…
Offload session-server tokenization to a threadpool
#1041
opened Apr 23, 2026 by
Shi-Dong
Contributor
Loading…
3 tasks
[swe-agent-v2] Randomize agent_name per request
#1040
opened Apr 23, 2026 by
Shi-Dong
Contributor
Loading…
2 tasks
Expose agent-server trial timeout as a configurable launcher flag
#1036
opened Apr 22, 2026 by
Shi-Dong
Contributor
Loading…
2 tasks
[alignment-miles] PR11A: Miles Ulysses CP true-on-policy data and loss contract
#1032
opened Apr 22, 2026 by
maocheng23
Contributor
•
Draft
4 tasks
[alignment-miles] PR11: Miles true-on-policy args, logprob contract, and metric
#1031
opened Apr 22, 2026 by
maocheng23
Contributor
•
Draft
6 tasks
feat: rollout-data replay and all-gather CP for megatron runner
#1030
opened Apr 21, 2026 by
yueming-yuan
Collaborator
Loading…
4 tasks
Add GLM-4.7 Full agentic async training launch script
#1029
opened Apr 21, 2026 by
Shi-Dong
Contributor
Loading…
1 of 2 tasks
tito: support agent-layer-inserted assistant messages in append segmentation
#1028
opened Apr 21, 2026 by
DavidBellamy
Loading…
wandb: raise init_timeout, add retry, fix shared-mode init for cross-region clusters
#1027
opened Apr 21, 2026 by
DavidBellamy
Loading…
[rollout] add --max-concurrent-agent-tasks to cap agentic in-flight requests
#1026
opened Apr 21, 2026 by
Shi-Dong
Contributor
Loading…
3 tasks
Security: Remote code execution risk from
trust_remote_code=True in model bridge loading
#1025
opened Apr 21, 2026 by
tomaioo
Loading…
[feat] add performance monitoring metrics for training and rollout
#1020
opened Apr 20, 2026 by
maocheng23
Contributor
•
Draft
4 tasks
Previous Next
ProTip!
Updated in the last three days: updated:>2026-05-01.