Skip to content

Actions: THUDM/slime

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
2,500+ workflow runs
2,500+ workflow runs

Filter by Workflow

⌥ + click/return to exclude

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
[Bug] slime-v0.3.0 版本在跑 qwen3.6 35B A3B 模型的时候，在第二次 rollout 会有乱码。怀疑 镜像&sglang 版本导致 Slash Command Handler #905: Issue comment #2091 (comment) edited by yuanlehome
9s
9s
[Bug] slime-v0.3.0 版本在跑 qwen3.6 35B A3B 模型的时候，在第二次 rollout 会有乱码。怀疑 镜像&sglang 版本导致 Slash Command Handler #904: Issue comment #2091 (comment) created by yuanlehome
12s
12s
Add critic load fallback for resuming training PR Test #2614: Pull request #2200 opened by coding-famer
4m 59s coding-famer:fix/critic_load
coding-famer:fix/critic_load
4m 59s
Add critic load fallback for resuming training conda CI #3911: Pull request #2200 opened by coding-famer
1s coding-famer:fix/critic_load
coding-famer:fix/critic_load
1s
Add critic load fallback for resuming training pre-commit #4639: Pull request #2200 opened by coding-famer
1m 2s coding-famer:fix/critic_load
coding-famer:fix/critic_load
1m 2s
[Bug] offload_train train worker dies on CUDA 13 with libcudart.so.12: cannot open shared object file — the torch_memory_saver LD_PRELOAD .so is chosen by filename, not by CUDA runtime Slash Command Handler #903: Issue comment #2186 (comment) created by littlemex
9s
9s
fix(npu): bracket IPv6 hosts in distributed init methods conda CI #3910: Pull request #2199 opened by EazyReal
1s EazyReal:fix/npu-ipv6-init-methods
EazyReal:fix/npu-ipv6-init-methods
1s
fix(npu): bracket IPv6 hosts in distributed init methods PR Test #2613: Pull request #2199 opened by EazyReal
3m 6s EazyReal:fix/npu-ipv6-init-methods
EazyReal:fix/npu-ipv6-init-methods
3m 6s
fix(npu): bracket IPv6 hosts in distributed init methods pre-commit #4638: Pull request #2199 opened by EazyReal
58s EazyReal:fix/npu-ipv6-init-methods
EazyReal:fix/npu-ipv6-init-methods
58s
fix(ppo): clamp log-ratios before exp() conda CI #3909: Pull request #2198 opened by EazyReal
1s EazyReal:fix/ppo-ratio-nan-clamp
EazyReal:fix/ppo-ratio-nan-clamp
1s
fix(ppo): clamp log-ratios before exp() pre-commit #4637: Pull request #2198 opened by EazyReal
1m 0s EazyReal:fix/ppo-ratio-nan-clamp
EazyReal:fix/ppo-ratio-nan-clamp
1m 0s
fix(ppo): clamp log-ratios before exp() PR Test #2612: Pull request #2198 opened by EazyReal
3m 39s EazyReal:fix/ppo-ratio-nan-clamp
EazyReal:fix/ppo-ratio-nan-clamp
3m 39s
fix: guard zero rollout temperature logprob scaling PR Test #2611: Pull request #2197 opened by EazyReal
2m 33s EazyReal:fix/rollout-temp-zero-guard
EazyReal:fix/rollout-temp-zero-guard
2m 33s
fix: guard zero rollout temperature logprob scaling conda CI #3908: Pull request #2197 opened by EazyReal
2s EazyReal:fix/rollout-temp-zero-guard
EazyReal:fix/rollout-temp-zero-guard
2s
fix: guard zero rollout temperature logprob scaling pre-commit #4636: Pull request #2197 opened by EazyReal
1m 5s EazyReal:fix/rollout-temp-zero-guard
EazyReal:fix/rollout-temp-zero-guard
1m 5s
fix(rollout): honor per-call args in generate_rollout_async (GenerateState caches first args) pre-commit #4635: Pull request #2196 synchronize by sdc17
1m 12s sdc17:fix/generate-state-stale-args
sdc17:fix/generate-state-stale-args
1m 12s
fix(rollout): honor per-call args in generate_rollout_async (GenerateState caches first args) conda CI #3907: Pull request #2196 synchronize by sdc17
1s sdc17:fix/generate-state-stale-args
sdc17:fix/generate-state-stale-args
1s
fix(rollout): honor per-call args in generate_rollout_async (GenerateState caches first args) PR Test #2610: Pull request #2196 synchronize by sdc17
2m 42s sdc17:fix/generate-state-stale-args
sdc17:fix/generate-state-stale-args
2m 42s
fix(rollout): honor per-call args in generate_rollout_async (GenerateState caches first args) PR Test #2609: Pull request #2196 opened by sdc17
3m 47s sdc17:fix/generate-state-stale-args
sdc17:fix/generate-state-stale-args
3m 47s
fix(rollout): honor per-call args in generate_rollout_async (GenerateState caches first args) pre-commit #4634: Pull request #2196 opened by sdc17
1m 0s sdc17:fix/generate-state-stale-args
sdc17:fix/generate-state-stale-args
1m 0s
fix(rollout): honor per-call args in generate_rollout_async (GenerateState caches first args) conda CI #3906: Pull request #2196 opened by sdc17
1s sdc17:fix/generate-state-stale-args
sdc17:fix/generate-state-stale-args
1s
Slash Command Handler Slash Command Handler #902: created by xwuShirley
9s
9s
fix(ray): dynamically resolve torch_memory_saver hook preload path for CUDA compatibility Slash Command Handler #901: Issue comment #2195 (comment) created by chethanuk
7s
7s
fix(ray): dynamically resolve torch_memory_saver hook preload path for CUDA compatibility conda CI #3905: Pull request #2195 synchronize by chethanuk
1s chethanuk:fix/issue-2186-cuda13-preload-resolver
chethanuk:fix/issue-2186-cuda13-preload-resolver
1s
fix(ray): dynamically resolve torch_memory_saver hook preload path for CUDA compatibility pre-commit #4633: Pull request #2195 synchronize by chethanuk
1m 4s chethanuk:fix/issue-2186-cuda13-preload-resolver
chethanuk:fix/issue-2186-cuda13-preload-resolver
1m 4s