0fe66c610e
val
2026-05-28 03:15:10 -04:00
80776baff7
val
2026-05-28 03:07:35 -04:00
7530a86979
Upload files to "train"
2026-05-28 02:59:17 -04:00
9dd29ac589
Workflow: pin numpy<2 AFTER requirements.txt (torch 2.2.2 NumPy 1.x ABI)
2026-05-28 02:54:44 -04:00
b92bc91bea
Workflow: pin numpy<2 (torch 2.2.2 ABI incompatible with numpy 2.x)
2026-05-28 02:52:21 -04:00
0c98d365d2
Workflow: pre-install CPU torch from pytorch.org/whl/cpu (saves ~600MB)
2026-05-28 02:47:38 -04:00
98926eb4a1
Re-render with ubuntu-latest label (routes to dawson-standard CPU runner)
2026-05-28 02:39:26 -04:00
0844b20ee4
Re-render workflow with dataset download_items + dawson-standard CPU runner
2026-05-28 02:36:43 -04:00
c5ed26372d
Workflow: pin RIAHUB_BASE_URL to logan LAN IP
2026-05-28 01:58:59 -04:00
feb908e84e
Re-trigger after repo transfer to qoherent (runner scope fix)
2026-05-28 01:53:45 -04:00
501a8c5e1c
Add adapt_dataset.py + re-trigger training canary
...
Previous run (2577) was waiting; missing scripts/adapt_dataset.py.
Adds the WAC-side adapter so the workflow's actions/checkout@v5
sparse-checkout finds the file at /scripts/adapt_dataset.py.
2026-05-28 01:48:03 -04:00
b8fd7d73d9
Trigger WavesFM Linear Probe training (canary)
...
Dataset: icc_canary_2026_05_28-v1.0.0.h5 (3 recs, 144 slices)
Model: qoherent/wavesfm-base/wavesfm-v1p0.pth @ 48787da4
Task: rml, epochs=3, batch_size=16, device=cpu
2026-05-28 01:42:59 -04:00
9f87fa9fe2
Add canary dataset: icc_canary_2026_05_28 v1.0.0 (3 recs, 144 slices)
2026-05-28 01:40:55 -04:00
e2351eaaeb
Initialize ICC demo workspace with LFS tracking + README
2026-05-28 01:35:31 -04:00
