Skip to content

Actions: NetKit-Labs/netkit

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
208 workflow runs
208 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
Show only the two latest suite PNGs on the README. CI #176: Commit 38a6321 pushed by jameslavrenz
8m 46s main
main
8m 46s
Drop Kalman roadmap and put MCU ahead of MPU/CPU in About. CI #175: Commit a98c0c6 pushed by jameslavrenz
10m 0s main
main
10m 0s
Point benchmark README at tracked LinkedIn suite PNGs. CI #174: Commit 98fbe46 pushed by jameslavrenz
8m 17s main
main
8m 17s
Refresh About/README for completed benches and track suite PNGs. CI #173: Commit 6c063dc pushed by jameslavrenz
9m 17s main
main
9m 17s
Document MCU no-heap policy and close C FeatureTap getter parity. CI #172: Commit 247b06e pushed by jameslavrenz
9m 52s main
main
9m 52s
Ban MCU heap forever and reclaim flash vs TFLM on int8 A/B. CI #171: Commit 8901d53 pushed by jameslavrenz
9m 4s main
main
9m 4s
Add NUCLEO DS-CNN int8 boards and save MCU CMSIS A/B UART logs. CI #170: Commit bb7ec3f pushed by jameslavrenz
10m 39s main
main
10m 39s
Save Pi Zero 2 W float32/int8 A/B results and defer open accuracy gaps. CI #169: Commit f24c5c0 pushed by jameslavrenz
10m 33s main
main
10m 33s
Add Pi Zero 2 W int8 MPU A/B suite alongside float32. CI #168: Commit 7cba4ec pushed by jameslavrenz
10m 18s main
main
10m 18s
Add Raspberry Pi Zero 2 W float32 MPU A/B suite tooling. CI #167: Commit 905fb47 pushed by jameslavrenz
11m 2s main
main
11m 2s
Add mid-run partial checkpoints for YOLOX mini train. CI #166: Commit 9265b65 pushed by jameslavrenz
10m 12s main
main
10m 12s
Document EMA v3 YOLOX fine-tune recipe toward ~0.30 AP@0.5. CI #165: Commit 080fa4b pushed by jameslavrenz
8m 11s main
main
8m 11s
Pack YOLOX EMA v2 (~0.224 AP@0.5) and skip pretrained HF downloads. CI #164: Commit 1c3db0d pushed by jameslavrenz
9m 55s main
main
9m 55s
Allow YOLOX hold-out eval without downloading timm pretrained weights. CI #163: Commit e81b5d1 pushed by jameslavrenz
9m 22s main
main
9m 22s
Add YOLOX float32 host A/B and full XNNPACK detect subgraph. CI #162: Commit 3c2e6c7 pushed by jameslavrenz
10m 12s main
main
10m 12s
Fix YOLOX pack layout and enable large CNN host infer. CI #161: Commit 20614f1 pushed by jameslavrenz
10m 6s main
main
10m 6s
Pack YOLOX trained.nk from EMA fine-tune that cleared AP@0.5. CI #160: Commit 75be6da pushed by jameslavrenz
9m 27s main
main
9m 27s
Support 416/h128 YOLOX training with shape-safe warm-starts. CI #159: Commit fc0abc2 pushed by jameslavrenz
10m 40s main
main
10m 40s
Add Mosaic and multi-scale aug to YOLOX mini-train. CI #158: Commit 2a1740b pushed by jameslavrenz
9m 15s main
main
9m 15s
Scale YOLOX mini-train docs to 50k and gate packing on COCO AP. CI #157: Commit ded4e73 pushed by jameslavrenz
7m 42s main
main
7m 42s
Add SimOTA assign and COCO-style AP@0.5 to YOLOX mini-train. CI #156: Commit c26a870 pushed by jameslavrenz
9m 23s main
main
9m 23s
Fix int8 .nk round-trips that broke test-full CI. CI #155: Commit 546e3ab pushed by jameslavrenz
9m 22s main
main
9m 22s
Test full Test full #4: Manually run by jameslavrenz
11m 47s main
main
11m 47s
Test full Test full #3: Manually run by jameslavrenz
13m 33s main
main
13m 33s
CI CI #154: Manually run by jameslavrenz
8m 12s main
main
8m 12s