Skip to content

Actions: paiml/aprender

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
fix(train): NF4 QLoRA CUDA forward computed the WRONG MODEL — NeoX ro… CI #8136: Commit fcc0fbc pushed by github-merge-queue Bot
16m 16s main
main
16m 16s
CI CI #8135: Merge group checks requested
16m 27s
16m 27s
BEAT Speed (nightly) BEAT Speed (nightly) #19: Scheduled
7m 46s main
main
7m 46s
Nightly Bench Nightly Bench #117: Scheduled
8m 2s main
main
8m 2s
qwen-story-daily qwen-story-daily #41: Scheduled
Queued main
main
Queued
Nightly Nightly #131: Scheduled
20m 10s main
main
20m 10s
Coverage Nightly Coverage Nightly #28: Scheduled
20m 55s main
main
20m 55s
CI CI #8134: Merge group checks requested
4m 44s
4m 44s
fix(train): NF4 QLoRA CUDA forward computed the WRONG MODEL — NeoX rope pairing + Q/K/V biases + full-warp softmax + causal CPU oracle CI #8133: Pull request #2252 opened by noahgift
20m 22s fix/nf4-train-loss-parity-main
fix/nf4-train-loss-parity-main
20m 22s
fix(train): NF4 QLoRA CUDA forward computed the WRONG MODEL — NeoX rope pairing + Q/K/V biases + full-warp softmax + causal CPU oracle PR Gate #1503: Pull request #2252 opened by noahgift
5s
5s
fix(train): NF4 QLoRA CUDA forward NaN loss — per-call cuBLAS stream … CI #8132: Commit 8e33d93 pushed by github-merge-queue Bot
17m 16s main
main
17m 16s
CI CI #8131: Merge group checks requested
16m 59s
16m 59s
fix(train): NF4 QLoRA NaN loss — per-call cuBLAS stream binding kills cross-stream data race PR Gate #1502: Pull request #2251 opened by noahgift
7s
7s
fix(train): NF4 QLoRA NaN loss — per-call cuBLAS stream binding kills cross-stream data race CI #8130: Pull request #2251 opened by noahgift
20m 44s fix/nf4-cublas-stream-binding
fix/nf4-cublas-stream-binding
20m 44s
fix(finetune): CUDA loss window honors --max-seq-len — kill the hardc… CI #8129: Commit c19f7bb pushed by github-merge-queue Bot
18m 2s main
main
18m 2s
CI CI #8128: Merge group checks requested
21m 44s
21m 44s
fix(train): NF4 QLoRA forward self-deadlock — wave of 4 in fused_resi… CI #8127: Commit 05045a1 pushed by github-merge-queue Bot
20m 12s main
main
20m 12s
CI CI #8126: Merge group checks requested
19m 21s
19m 21s
CI CI #8125: Merge group checks requested
6m 13s
6m 13s
CI CI #8124: Merge group checks requested
6m 14s
6m 14s
v0.56.0 — apr-format sovereign leaf + f16-RNE sweep + honest finetune GPU + Ollama /api streaming Binary Release #57: Release v0.56.0 published by noahgift
2m 54s
2m 54s
fix(finetune): CUDA loss window honors --max-seq-len — kill hardcoded 512 clamp that silently trained nothing CI #8123: Pull request #2250 synchronize by noahgift
34m 37s fix/finetune-cuda-loss-window-512-clamp
fix/finetune-cuda-loss-window-512-clamp
34m 37s
fix(train): NF4 QLoRA forward self-deadlock — wave of 4 in fused_residual_rmsnorm_forward (unblocks GPU finetune) CI #8122: Pull request #2249 synchronize by noahgift
34m 13s fix/qlora-fused-rmsnorm-self-deadlock
fix/qlora-fused-rmsnorm-self-deadlock
34m 13s
chore(release): 0.56.0 — apr-format sovereign leaf + f16-RNE sweep + … CI #8121: Commit d712de6 pushed by github-merge-queue Bot
28m 56s main
main
28m 56s
CI CI #8120: Merge group checks requested
27m 21s
27m 21s