Workflow runs · AlignmentResearch/Megatron-LM · GitHub

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows

34 workflow runs

34 workflow runs

Merge pull request #3 from AlignmentResearch/tf-at/torch-gmm-full-rec… fork-base #5: Commit 7f47384 pushed by taufeeque9

15s farai/main

farai/main

15s

perf(moe): add frozen grouped GEMM for adapter training fork-base #4: Pull request #3 synchronize by taufeeque9

15s tf-at/torch-gmm-full-recompute

tf-at/torch-gmm-full-recompute

15s

Graph Update: pip in /docker/lts, /examples/post_training/modelopt, /megatron/core, /megatron/core/distributed/fsdp/src #1544878844 Dependency Graph #2: by dependabot Bot

1m 2s farai/main

farai/main

1m 2s

Graph Update: uv in /. #1544878843 Dependency Graph #1: by dependabot Bot

50s farai/main

farai/main

50s

chore(fork): prune NVIDIA-only CI and record the upstream base Community Bot #3: Issue comment #2 (comment) created by taufeeque9

8s

8s

chore(fork): prune NVIDIA-only CI and record the upstream base Request NVSkills CI #8: Issue comment #2 (comment) created by taufeeque9

1s

1s

chore(fork): prune NVIDIA-only CI and record the upstream base Claude Fix PR #3: Issue comment #2 (comment) created by taufeeque9

1s

1s

chore(fork): prune NVIDIA-only CI and record the upstream base Community Request Assignee #3: Issue comment #2 (comment) created by taufeeque9

1s

1s

chore(fork): prune NVIDIA-only CI and record the upstream base Claude Code Review #3: Issue comment #2 (comment) created by taufeeque9

1s

1s

chore(fork): prune NVIDIA-only CI and record the upstream base Claude Copy PR to Main #3: Issue comment #2 (comment) created by taufeeque9

10s

10s

Merge pull request #4 from AlignmentResearch/tf-c/fork-base-scaffoldi… fork-base #3: Commit e92d99e pushed by taufeeque9

12s farai/main

farai/main

12s

chore(fork): prune NVIDIA-only CI and record upstream base Claude Complexity Label #2: Pull request #4 ready_for_review by taufeeque9

21s

21s

chore(fork): prune NVIDIA-only CI and record upstream base fork-base #2: Pull request #4 opened by taufeeque9

16s tf-c/fork-base-scaffolding-reviewed

tf-c/fork-base-scaffolding-reviewed

16s

Auto Swap Labels Auto Swap Labels #1: completed by SamAdamDay

2s

2s

perf(moe): add frozen grouped GEMM for adapter training Review Trigger #1: Pull request #3 submitted by SamAdamDay

1s

1s

perf(moe): add frozen grouped GEMM for adapter training Claude Complexity Label #1: Pull request #3 ready_for_review by taufeeque9

35s

35s

fix(moe): reject TE op fuser with torch backend Request NVSkills CI #7: Commit 8447c0d pushed by taufeeque9

1s tf-at/torch-gmm-full-recompute

tf-at/torch-gmm-full-recompute

1s

test(moe): simplify no-token setup Request NVSkills CI #6: Commit dcee153 pushed by taufeeque9

1s tf-at/torch-gmm-full-recompute

tf-at/torch-gmm-full-recompute

1s

test(recompute): cover frozen-input adapter gradients Request NVSkills CI #5: Commit 96ecb4b pushed by taufeeque9

1s tf-at/torch-gmm-full-recompute

tf-at/torch-gmm-full-recompute

1s

build: record Megatron-LM fork base Request NVSkills CI #4: Commit 97f7d41 pushed by taufeeque9

2s tf-at/torch-gmm-full-recompute

tf-at/torch-gmm-full-recompute

2s

feat(moe+ssm): expert-LoRA GEMM stack and hybrid-recompute guard on the d12f6c8c base Community Bot #2: Issue comment #1 (comment) created by taufeeque9

1s

1s

feat(moe+ssm): expert-LoRA GEMM stack and hybrid-recompute guard on the d12f6c8c base Request NVSkills CI #3: Issue comment #1 (comment) created by taufeeque9

1s

1s

feat(moe+ssm): expert-LoRA GEMM stack and hybrid-recompute guard on the d12f6c8c base Claude Fix PR #2: Issue comment #1 (comment) created by taufeeque9

2s

2s

feat(moe+ssm): expert-LoRA GEMM stack and hybrid-recompute guard on the d12f6c8c base Community Request Assignee #2: Issue comment #1 (comment) created by taufeeque9

7s

7s

feat(moe+ssm): expert-LoRA GEMM stack and hybrid-recompute guard on the d12f6c8c base Claude Code Review #2: Issue comment #1 (comment) created by taufeeque9

1s

1s