Skip to content

Actions: InternLM/InternEvo

Actions

pr-merged

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
pr_merged.yaml
70 workflow runs
70 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

fix(linear): remove linear op assert (#438) pr-merged #238: Commit a5b3972 pushed by blankde
1d 0h 0m 1s develop
develop
1d 0h 0m 1s
fix(convert ckpt): fix option (#437) pr-merged #237: Commit c89c9db pushed by blankde
1d 0h 0m 2s develop
develop
1d 0h 0m 2s
fix(qwen3): fix qwen3 bias (#435) pr-merged #236: Commit 47d0f54 pushed by blankde
1d 0h 0m 2s develop
develop
1d 0h 0m 2s
fix(ckpt): fix qwen3 save ckpt (#434) pr-merged #235: Commit 65f02d9 pushed by blankde
1d 0h 0m 1s develop
develop
1d 0h 0m 1s
feat(model): support qwen3 (#430) pr-merged #234: Commit 550fce9 pushed by blankde
1d 0h 0m 3s develop
develop
1d 0h 0m 3s
feat(ckpt): support universal ckpt MOE extended Edition (#423) pr-merged #233: Commit 198c333 pushed by blankde
1d 0h 0m 1s develop
develop
1d 0h 0m 1s
fix(moe): fix module prefetch for moe (#421) pr-merged #232: Commit 58f723a pushed by sunpengsdu
38m 46s develop
develop
38m 46s
feat(moe) impl w1&3 fused for moe (#422) pr-merged #231: Commit 10ede60 pushed by sunpengsdu
30m 53s develop
develop
30m 53s
fix(parallel.py): fix norm and moe gate gradient reduce check (#420) pr-merged #230: Commit 30bb508 pushed by sunpengsdu
30m 2s develop
develop
30m 2s
feat(ckpt): support compatible version universal checkpoint (#418) pr-merged #229: Commit f925cd3 pushed by sunpengsdu
11m 25s develop
develop
11m 25s
fix(moe): fix lack attribute when print moe model (#407) pr-merged #228: Commit 5b75d47 pushed by zigzagcai
35m 12s develop
develop
35m 12s
feat(isp): support layer fused isp communication (#395) pr-merged #227: Commit 3bcf372 pushed by zigzagcai
31m 15s develop
develop
31m 15s
support fp32 all_reduce and reduce_scatter (#389) pr-merged #226: Commit c52dd22 pushed by sunpengsdu
30m 28s develop
develop
30m 28s
feat(2D): support selective-checkpoint-offload for 2D attention (#396) pr-merged #225: Commit caf30d8 pushed by huangting4201
30m 38s develop
develop
30m 38s
feat(cpu_offload.py): support selective layers' activation cpu offloa… pr-merged #224: Commit 20cbcfb pushed by zigzagcai
31m 32s develop
develop
31m 32s
feat(fsdp): support FSDP training (#413) pr-merged #223: Commit b0928c7 pushed by zigzagcai
30m 38s develop
develop
30m 38s
fix(moe): fix moe act late release (#387) pr-merged #222: Commit 1a48f41 pushed by sunpengsdu
3h 2m 19s develop
develop
3h 2m 19s
feat(*): Add internlm3 config (#403) pr-merged #221: Commit 24180aa pushed by gaoyang07
3m 12s develop
develop
3m 12s
test(ci): update ci (#405) pr-merged #220: Commit 02ea919 pushed by gaoyang07
30m 28s develop
develop
30m 28s
fix(hf): fix convert_inetrnevo2hf for internlm2 model (#401) pr-merged #219: Commit 7d03512 pushed by sunpengsdu
19m 37s develop
develop
19m 37s
fix(hybrid optim): fp32_grad not scaled when use offload_cpu (#399) pr-merged #218: Commit 2874881 pushed by sunpengsdu
49m 11s develop
develop
49m 11s
feat(isp): add early_reduce_scatter_release support (#380) pr-merged #217: Commit d03c6f9 pushed by huangting4201
2h 9m 58s develop
develop
2h 9m 58s
fix(CI): fix lint check in CI (#397) pr-merged #216: Commit 1ea1555 pushed by huangting4201
1h 46m 41s develop
develop
1h 46m 41s
feat(comm/attn_offload.py): support selective ckpt and cpu offload (#… pr-merged #215: Commit e3f5001 pushed by sunpengsdu
1h 26m 5s develop
develop
1h 26m 5s
feat(loss)/add different operator types for cross_entropy (#386) pr-merged #214: Commit 141e9eb pushed by sunpengsdu
2h 46m 52s develop
develop
2h 46m 52s