Context:
homelab
Cluster:
k3s
User:
marcos
Forge Rev:
v1.0 (C17)
K8s Rev:
v1.36.5+k3s1
CPU:
18%
MEM:
32%

describe(bench/mlx-main-apple-m4-cpu-mlp-train)

Backend:       mlx-main
Device:        apple-m4-cpu
Model:         mlp
Mode:          train
Runs:          9, since 2026-10-12
Latest:        200.3k samples/s on 2026-10-23
Best:          202.0k samples/s on 2026-10-22, keras 2178dd6

Speed as % of the best run, oldest on the left

100% │  ▂   ▆█▇
 87% │  █   ███
 75% │█▇███████
 62% │█████████
 50% │█████████
 37% │█████████
 25% │█████████
  0% │█████████
     └─────────
             now
TIME↓SAMPLES/SΔ PREVKERAS
2026-10-23T03:01Z200.3k-0.9%2178dd6
2026-10-22T03:01Z202.0k+1.6%2178dd6
2026-10-21T03:01Z198.8k+29.1%2178dd6
2026-10-20T18:01Z154.0k+1.8%2178dd6
2026-10-16T03:01Z151.3k-1.6%2178dd6
2026-10-15T03:01Z153.8k-15.9%2178dd6
2026-10-14T03:01Z182.9k+21.5%2178dd6
2026-10-13T03:01Z150.5k-2.2%2178dd6
2026-10-12T06:01Z153.9k2178dd6