Skip to content

Actions: NVIDIA-NeMo/Megatron-Bridge

Actions

Community Bot

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
community-bot.yml
5,094 workflow run results
5,094 workflow run results

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

[Regression] Low QWEN3 30B B200 and H100 performance with dropless Community Bot #5094: Issue comment #1998 (comment) created by erhoo82
12s
12s
Nemotron-3-Nano Model Support Community Bot #5093: Issue comment #1914 (comment) created by liding-nv
10s
10s
[Test][patch] NVFP4 support DSv3 and QWEN3 Community Bot #5092: Issue comment #1728 (comment) created by erhoo82
11s
11s
export checkpoint error: too many values to unpack (expected 2) Community Bot #5091: Issue comment #1996 (comment) created by yaoyu-33
19s
19s
Support and switching mechanism for multiple LoRA adapters Community Bot #5090: Issue comment #1986 (comment) created by yaoyu-33
12s
12s
Is parallel (multi-process) checkpoint conversion supported? Community Bot #5089: Issue comment #1984 (comment) created by yaoyu-33
10s
10s
Context Parallel mismatch: CP set to 2, runtime slices with CP=4 causing reshape error in get_batch_on_this_cp_rank Community Bot #5088: Issue comment #1982 (comment) edited by yaoyu-33
14s
14s
Context Parallel mismatch: CP set to 2, runtime slices with CP=4 causing reshape error in get_batch_on_this_cp_rank Community Bot #5087: Issue comment #1982 (comment) created by yaoyu-33
20s
20s
Support for GPT-OSS pretraining script in Megatron-Bridge Community Bot #5086: Issue comment #1944 (comment) edited by yaoyu-33
22s
22s
Support for GPT-OSS pretraining script in Megatron-Bridge Community Bot #5085: Issue comment #1944 (comment) created by yaoyu-33
17s
17s
Incorrect Tensor Parallel sharding and semantic ordering for GDN in_proj and conv1d weights in Qwen3Next Community Bot #5084: Issue comment #2013 (comment) created by yaoyu-33
12s
12s
[M4] feat: Add M4 end2end support and qwen3 examples Community Bot #5083: Issue comment #2011 (comment) created by yaoyu-33
15s
15s
Refactor the optimizer override function so that users can swap in their own Community Bot #5082: Issue comment #2010 (comment) created by jstjohn
16s
16s
Nemotron-3-Nano Model Support Community Bot #5081: Issue comment #1914 (comment) created by liding-nv
15s
15s
[M4] feat: Add M4 end2end support and qwen3 examples Community Bot #5080: Issue comment #2011 (comment) created by yaoyu-33
14s
14s
Refactor the optimizer override function so that users can swap in their own Community Bot #5079: Issue comment #2010 (comment) created by copy-pr-bot bot
12s
12s
[Bug] Nan grad norm in MXFP8 with param AG overlap Community Bot #5078: Issue comment #1771 (comment) created by erhoo82
11s
11s
[Bug] Nan grad norm in MXFP8 with param AG overlap Community Bot #5077: Issue #1771 closed by erhoo82
12s
12s
Path-based huggingface tokenizers do not load from the checkpoint when using load_tokenizer(...) Community Bot #5076: Issue comment #1900 (comment) created by jstjohn
13s
13s
Path-based huggingface tokenizers do not load from the checkpoint when using load_tokenizer(...) Community Bot #5075: Issue comment #1900 (comment) created by jstjohn
13s
13s
[Bug] H100 + DeepSeekV3 error- DeepEP error: timeout (dispatch CPU) Community Bot #5074: Issue comment #1335 (comment) created by malay-nagda
19s
19s
Utilize B300 cluster PCT cores Community Bot #5073: Issue comment #1928 (comment) created by erhoo82
14s
14s
Proposal for more scalable seq-pack data format Community Bot #5072: Issue comment #2019 (comment) created by copy-pr-bot bot
12s
12s
Fix experimental attention variant spec Community Bot #5071: Issue comment #2018 (comment) created by copy-pr-bot bot
10s
10s
Training works fine but checkpointing OOMs out Community Bot #5070: Issue comment #1748 (comment) edited by dimapihtar
13s
13s