Upload benchmarks/bench_10_multi_teacher.json with huggingface_hub
Browse files
benchmarks/bench_10_multi_teacher.json
ADDED
|
@@ -0,0 +1,1161 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"benchmark": "multi_teacher",
|
| 3 |
+
"timestamp": "2026-03-19T12:10:52.183671+00:00",
|
| 4 |
+
"device": "cuda",
|
| 5 |
+
"n_gpus": 4,
|
| 6 |
+
"gpu_names": [
|
| 7 |
+
"NVIDIA L4",
|
| 8 |
+
"NVIDIA L4",
|
| 9 |
+
"NVIDIA L4",
|
| 10 |
+
"NVIDIA L4"
|
| 11 |
+
],
|
| 12 |
+
"gpu_placement": {
|
| 13 |
+
"2_teachers": {
|
| 14 |
+
"teachers": [
|
| 15 |
+
"smolvla-base",
|
| 16 |
+
"rdt2-fm"
|
| 17 |
+
],
|
| 18 |
+
"placement": {
|
| 19 |
+
"rdt2-fm": "cuda:0",
|
| 20 |
+
"smolvla-base": "cuda:0"
|
| 21 |
+
},
|
| 22 |
+
"total_vram_gb": 3.5,
|
| 23 |
+
"gpu_memory_mb": [
|
| 24 |
+
22563.0,
|
| 25 |
+
22563.0,
|
| 26 |
+
22563.0,
|
| 27 |
+
22563.0
|
| 28 |
+
],
|
| 29 |
+
"all_on_gpu": true
|
| 30 |
+
},
|
| 31 |
+
"3_teachers": {
|
| 32 |
+
"teachers": [
|
| 33 |
+
"smolvla-base",
|
| 34 |
+
"rdt2-fm",
|
| 35 |
+
"openvla-7b"
|
| 36 |
+
],
|
| 37 |
+
"placement": {
|
| 38 |
+
"openvla-7b": "cuda:0",
|
| 39 |
+
"rdt2-fm": "cuda:0",
|
| 40 |
+
"smolvla-base": "cuda:0"
|
| 41 |
+
},
|
| 42 |
+
"total_vram_gb": 18.7,
|
| 43 |
+
"gpu_memory_mb": [
|
| 44 |
+
22563.0,
|
| 45 |
+
22563.0,
|
| 46 |
+
22563.0,
|
| 47 |
+
22563.0
|
| 48 |
+
],
|
| 49 |
+
"all_on_gpu": true
|
| 50 |
+
},
|
| 51 |
+
"4_teachers": {
|
| 52 |
+
"teachers": [
|
| 53 |
+
"smolvla-base",
|
| 54 |
+
"rdt2-fm",
|
| 55 |
+
"openvla-7b",
|
| 56 |
+
"bitvla-base"
|
| 57 |
+
],
|
| 58 |
+
"placement": {
|
| 59 |
+
"openvla-7b": "cuda:0",
|
| 60 |
+
"rdt2-fm": "cuda:0",
|
| 61 |
+
"smolvla-base": "cuda:0",
|
| 62 |
+
"bitvla-base": "cuda:0"
|
| 63 |
+
},
|
| 64 |
+
"total_vram_gb": 19.5,
|
| 65 |
+
"gpu_memory_mb": [
|
| 66 |
+
22563.0,
|
| 67 |
+
22563.0,
|
| 68 |
+
22563.0,
|
| 69 |
+
22563.0
|
| 70 |
+
],
|
| 71 |
+
"all_on_gpu": true
|
| 72 |
+
},
|
| 73 |
+
"5_teachers": {
|
| 74 |
+
"teachers": [
|
| 75 |
+
"smolvla-base",
|
| 76 |
+
"rdt2-fm",
|
| 77 |
+
"openvla-7b",
|
| 78 |
+
"bitvla-base",
|
| 79 |
+
"pi0-base"
|
| 80 |
+
],
|
| 81 |
+
"placement": {
|
| 82 |
+
"openvla-7b": "cuda:0",
|
| 83 |
+
"pi0-base": "cuda:0",
|
| 84 |
+
"rdt2-fm": "cuda:0",
|
| 85 |
+
"smolvla-base": "cuda:0",
|
| 86 |
+
"bitvla-base": "cuda:1"
|
| 87 |
+
},
|
| 88 |
+
"total_vram_gb": 22.7,
|
| 89 |
+
"gpu_memory_mb": [
|
| 90 |
+
22563.0,
|
| 91 |
+
22563.0,
|
| 92 |
+
22563.0,
|
| 93 |
+
22563.0
|
| 94 |
+
],
|
| 95 |
+
"all_on_gpu": true
|
| 96 |
+
}
|
| 97 |
+
},
|
| 98 |
+
"multi_teacher_training": {
|
| 99 |
+
"1_teacher": {
|
| 100 |
+
"n_teachers": 1,
|
| 101 |
+
"n_steps": 50,
|
| 102 |
+
"loss_curve": [
|
| 103 |
+
0.1811,
|
| 104 |
+
0.173,
|
| 105 |
+
0.1179,
|
| 106 |
+
0.1651,
|
| 107 |
+
0.08,
|
| 108 |
+
0.2392,
|
| 109 |
+
0.2909,
|
| 110 |
+
0.3574,
|
| 111 |
+
0.1247,
|
| 112 |
+
0.0965,
|
| 113 |
+
0.3327,
|
| 114 |
+
0.2088,
|
| 115 |
+
0.2335,
|
| 116 |
+
0.1851,
|
| 117 |
+
0.1211,
|
| 118 |
+
0.1902,
|
| 119 |
+
0.255,
|
| 120 |
+
0.1509,
|
| 121 |
+
0.15,
|
| 122 |
+
0.0945,
|
| 123 |
+
0.2734,
|
| 124 |
+
0.2772,
|
| 125 |
+
0.2513,
|
| 126 |
+
0.2046,
|
| 127 |
+
0.1291,
|
| 128 |
+
0.0581,
|
| 129 |
+
0.0795,
|
| 130 |
+
0.2929,
|
| 131 |
+
0.0353,
|
| 132 |
+
0.1832,
|
| 133 |
+
0.1968,
|
| 134 |
+
0.2949,
|
| 135 |
+
0.2511,
|
| 136 |
+
0.2216,
|
| 137 |
+
0.1817,
|
| 138 |
+
0.2631,
|
| 139 |
+
0.1783,
|
| 140 |
+
0.1479,
|
| 141 |
+
0.2881,
|
| 142 |
+
0.3657,
|
| 143 |
+
0.1903,
|
| 144 |
+
0.3923,
|
| 145 |
+
0.0875,
|
| 146 |
+
0.1497,
|
| 147 |
+
0.3716,
|
| 148 |
+
0.155,
|
| 149 |
+
0.2252,
|
| 150 |
+
0.1258,
|
| 151 |
+
0.1074,
|
| 152 |
+
0.1239
|
| 153 |
+
],
|
| 154 |
+
"loss_start": 0.1811,
|
| 155 |
+
"loss_end": 0.1239,
|
| 156 |
+
"loss_reduction_pct": 31.6,
|
| 157 |
+
"router_entropy_curve": [
|
| 158 |
+
-0.0,
|
| 159 |
+
-0.0,
|
| 160 |
+
-0.0,
|
| 161 |
+
-0.0,
|
| 162 |
+
-0.0,
|
| 163 |
+
-0.0,
|
| 164 |
+
-0.0,
|
| 165 |
+
-0.0,
|
| 166 |
+
-0.0,
|
| 167 |
+
-0.0,
|
| 168 |
+
-0.0,
|
| 169 |
+
-0.0,
|
| 170 |
+
-0.0,
|
| 171 |
+
-0.0,
|
| 172 |
+
-0.0,
|
| 173 |
+
-0.0,
|
| 174 |
+
-0.0,
|
| 175 |
+
-0.0,
|
| 176 |
+
-0.0,
|
| 177 |
+
-0.0,
|
| 178 |
+
-0.0,
|
| 179 |
+
-0.0,
|
| 180 |
+
-0.0,
|
| 181 |
+
-0.0,
|
| 182 |
+
-0.0,
|
| 183 |
+
-0.0,
|
| 184 |
+
-0.0,
|
| 185 |
+
-0.0,
|
| 186 |
+
-0.0,
|
| 187 |
+
-0.0,
|
| 188 |
+
-0.0,
|
| 189 |
+
-0.0,
|
| 190 |
+
-0.0,
|
| 191 |
+
-0.0,
|
| 192 |
+
-0.0,
|
| 193 |
+
-0.0,
|
| 194 |
+
-0.0,
|
| 195 |
+
-0.0,
|
| 196 |
+
-0.0,
|
| 197 |
+
-0.0,
|
| 198 |
+
-0.0,
|
| 199 |
+
-0.0,
|
| 200 |
+
-0.0,
|
| 201 |
+
-0.0,
|
| 202 |
+
-0.0,
|
| 203 |
+
-0.0,
|
| 204 |
+
-0.0,
|
| 205 |
+
-0.0,
|
| 206 |
+
-0.0,
|
| 207 |
+
-0.0
|
| 208 |
+
],
|
| 209 |
+
"router_entropy_final": -0.0,
|
| 210 |
+
"step_time_mean_ms": 1385.9,
|
| 211 |
+
"steps_per_sec": 0.72,
|
| 212 |
+
"peak_gpu_gb": 5.68
|
| 213 |
+
},
|
| 214 |
+
"2_teacher": {
|
| 215 |
+
"n_teachers": 2,
|
| 216 |
+
"n_steps": 50,
|
| 217 |
+
"loss_curve": [
|
| 218 |
+
0.4444,
|
| 219 |
+
0.1762,
|
| 220 |
+
0.2341,
|
| 221 |
+
0.1347,
|
| 222 |
+
0.137,
|
| 223 |
+
0.2796,
|
| 224 |
+
0.3747,
|
| 225 |
+
0.2338,
|
| 226 |
+
0.3114,
|
| 227 |
+
0.1703,
|
| 228 |
+
0.2068,
|
| 229 |
+
0.3026,
|
| 230 |
+
0.094,
|
| 231 |
+
0.0867,
|
| 232 |
+
0.2434,
|
| 233 |
+
0.298,
|
| 234 |
+
0.3349,
|
| 235 |
+
0.2322,
|
| 236 |
+
0.1063,
|
| 237 |
+
0.309,
|
| 238 |
+
0.1868,
|
| 239 |
+
0.2246,
|
| 240 |
+
0.1281,
|
| 241 |
+
0.2459,
|
| 242 |
+
0.0905,
|
| 243 |
+
0.0509,
|
| 244 |
+
0.272,
|
| 245 |
+
0.1021,
|
| 246 |
+
0.1744,
|
| 247 |
+
0.1578,
|
| 248 |
+
0.0891,
|
| 249 |
+
0.1626,
|
| 250 |
+
0.082,
|
| 251 |
+
0.1939,
|
| 252 |
+
0.2086,
|
| 253 |
+
0.1851,
|
| 254 |
+
0.0636,
|
| 255 |
+
0.2347,
|
| 256 |
+
0.1209,
|
| 257 |
+
0.0883,
|
| 258 |
+
0.0913,
|
| 259 |
+
0.308,
|
| 260 |
+
0.2321,
|
| 261 |
+
0.1441,
|
| 262 |
+
0.1929,
|
| 263 |
+
0.1138,
|
| 264 |
+
0.0574,
|
| 265 |
+
0.1018,
|
| 266 |
+
0.1673,
|
| 267 |
+
0.2271
|
| 268 |
+
],
|
| 269 |
+
"loss_start": 0.4444,
|
| 270 |
+
"loss_end": 0.2271,
|
| 271 |
+
"loss_reduction_pct": 48.9,
|
| 272 |
+
"router_entropy_curve": [
|
| 273 |
+
0.6898,
|
| 274 |
+
0.033,
|
| 275 |
+
0.0259,
|
| 276 |
+
0.0199,
|
| 277 |
+
0.0143,
|
| 278 |
+
0.0109,
|
| 279 |
+
0.0082,
|
| 280 |
+
0.0059,
|
| 281 |
+
0.0044,
|
| 282 |
+
0.0035,
|
| 283 |
+
0.0027,
|
| 284 |
+
0.0022,
|
| 285 |
+
0.0018,
|
| 286 |
+
0.0015,
|
| 287 |
+
0.0013,
|
| 288 |
+
0.0011,
|
| 289 |
+
0.001,
|
| 290 |
+
0.0008,
|
| 291 |
+
0.0008,
|
| 292 |
+
0.0007,
|
| 293 |
+
0.0006,
|
| 294 |
+
0.0006,
|
| 295 |
+
0.0005,
|
| 296 |
+
0.0005,
|
| 297 |
+
0.0005,
|
| 298 |
+
0.0004,
|
| 299 |
+
0.0004,
|
| 300 |
+
0.0004,
|
| 301 |
+
0.0003,
|
| 302 |
+
0.0003,
|
| 303 |
+
0.0003,
|
| 304 |
+
0.0003,
|
| 305 |
+
0.0003,
|
| 306 |
+
0.0003,
|
| 307 |
+
0.0003,
|
| 308 |
+
0.0003,
|
| 309 |
+
0.0002,
|
| 310 |
+
0.0002,
|
| 311 |
+
0.0002,
|
| 312 |
+
0.0002,
|
| 313 |
+
0.0002,
|
| 314 |
+
0.0002,
|
| 315 |
+
0.0002,
|
| 316 |
+
0.0002,
|
| 317 |
+
0.0002,
|
| 318 |
+
0.0002,
|
| 319 |
+
0.0002,
|
| 320 |
+
0.0002,
|
| 321 |
+
0.0002,
|
| 322 |
+
0.0002
|
| 323 |
+
],
|
| 324 |
+
"router_entropy_final": 0.0002,
|
| 325 |
+
"step_time_mean_ms": 877.4,
|
| 326 |
+
"steps_per_sec": 1.14,
|
| 327 |
+
"peak_gpu_gb": 6.87
|
| 328 |
+
},
|
| 329 |
+
"3_teacher": {
|
| 330 |
+
"n_teachers": 3,
|
| 331 |
+
"n_steps": 50,
|
| 332 |
+
"loss_curve": [
|
| 333 |
+
0.2586,
|
| 334 |
+
0.2341,
|
| 335 |
+
0.1751,
|
| 336 |
+
0.3115,
|
| 337 |
+
0.0737,
|
| 338 |
+
0.1848,
|
| 339 |
+
0.1948,
|
| 340 |
+
0.2198,
|
| 341 |
+
0.2203,
|
| 342 |
+
0.1545,
|
| 343 |
+
0.1288,
|
| 344 |
+
0.086,
|
| 345 |
+
0.1256,
|
| 346 |
+
0.1571,
|
| 347 |
+
0.4571,
|
| 348 |
+
0.1713,
|
| 349 |
+
0.0936,
|
| 350 |
+
0.2297,
|
| 351 |
+
0.4955,
|
| 352 |
+
0.179,
|
| 353 |
+
0.228,
|
| 354 |
+
0.1758,
|
| 355 |
+
0.119,
|
| 356 |
+
0.1552,
|
| 357 |
+
0.152,
|
| 358 |
+
0.3176,
|
| 359 |
+
0.2286,
|
| 360 |
+
0.1916,
|
| 361 |
+
0.1012,
|
| 362 |
+
0.0847,
|
| 363 |
+
0.0541,
|
| 364 |
+
0.1735,
|
| 365 |
+
0.1299,
|
| 366 |
+
0.1959,
|
| 367 |
+
0.2083,
|
| 368 |
+
0.2462,
|
| 369 |
+
0.2322,
|
| 370 |
+
0.3797,
|
| 371 |
+
0.2763,
|
| 372 |
+
0.1415,
|
| 373 |
+
0.2821,
|
| 374 |
+
0.2788,
|
| 375 |
+
0.262,
|
| 376 |
+
0.3141,
|
| 377 |
+
0.2344,
|
| 378 |
+
0.1762,
|
| 379 |
+
0.0794,
|
| 380 |
+
0.1791,
|
| 381 |
+
0.1063,
|
| 382 |
+
0.2771
|
| 383 |
+
],
|
| 384 |
+
"loss_start": 0.2586,
|
| 385 |
+
"loss_end": 0.2771,
|
| 386 |
+
"loss_reduction_pct": -7.1,
|
| 387 |
+
"router_entropy_curve": [
|
| 388 |
+
1.0851,
|
| 389 |
+
0.2696,
|
| 390 |
+
0.0359,
|
| 391 |
+
0.0161,
|
| 392 |
+
0.0114,
|
| 393 |
+
0.0093,
|
| 394 |
+
0.0077,
|
| 395 |
+
0.0066,
|
| 396 |
+
0.0054,
|
| 397 |
+
0.0043,
|
| 398 |
+
0.0035,
|
| 399 |
+
0.0028,
|
| 400 |
+
0.0022,
|
| 401 |
+
0.0018,
|
| 402 |
+
0.0015,
|
| 403 |
+
0.0012,
|
| 404 |
+
0.001,
|
| 405 |
+
0.0009,
|
| 406 |
+
0.0007,
|
| 407 |
+
0.0006,
|
| 408 |
+
0.0006,
|
| 409 |
+
0.0005,
|
| 410 |
+
0.0005,
|
| 411 |
+
0.0004,
|
| 412 |
+
0.0004,
|
| 413 |
+
0.0003,
|
| 414 |
+
0.0003,
|
| 415 |
+
0.0003,
|
| 416 |
+
0.0003,
|
| 417 |
+
0.0003,
|
| 418 |
+
0.0002,
|
| 419 |
+
0.0002,
|
| 420 |
+
0.0002,
|
| 421 |
+
0.0002,
|
| 422 |
+
0.0002,
|
| 423 |
+
0.0002,
|
| 424 |
+
0.0002,
|
| 425 |
+
0.0002,
|
| 426 |
+
0.0002,
|
| 427 |
+
0.0002,
|
| 428 |
+
0.0002,
|
| 429 |
+
0.0002,
|
| 430 |
+
0.0002,
|
| 431 |
+
0.0001,
|
| 432 |
+
0.0001,
|
| 433 |
+
0.0001,
|
| 434 |
+
0.0001,
|
| 435 |
+
0.0001,
|
| 436 |
+
0.0001,
|
| 437 |
+
0.0001
|
| 438 |
+
],
|
| 439 |
+
"router_entropy_final": 0.0001,
|
| 440 |
+
"step_time_mean_ms": 813.3,
|
| 441 |
+
"steps_per_sec": 1.23,
|
| 442 |
+
"peak_gpu_gb": 6.86
|
| 443 |
+
}
|
| 444 |
+
},
|
| 445 |
+
"universal_distillation": {
|
| 446 |
+
"balanced": {
|
| 447 |
+
"config": {
|
| 448 |
+
"name": "balanced",
|
| 449 |
+
"alpha_task": 0.3,
|
| 450 |
+
"alpha_diversity": 0.05,
|
| 451 |
+
"alpha_consistency": 0.1
|
| 452 |
+
},
|
| 453 |
+
"n_teachers": 3,
|
| 454 |
+
"n_steps": 40,
|
| 455 |
+
"loss_curves": {
|
| 456 |
+
"total": [
|
| 457 |
+
0.8202,
|
| 458 |
+
0.2073,
|
| 459 |
+
0.1744,
|
| 460 |
+
0.1607,
|
| 461 |
+
0.7928,
|
| 462 |
+
1.1326,
|
| 463 |
+
0.2527,
|
| 464 |
+
0.1391,
|
| 465 |
+
1.0104,
|
| 466 |
+
0.6887,
|
| 467 |
+
0.647,
|
| 468 |
+
0.4462,
|
| 469 |
+
0.1474,
|
| 470 |
+
0.3148,
|
| 471 |
+
0.1989,
|
| 472 |
+
0.2598,
|
| 473 |
+
0.1787,
|
| 474 |
+
0.1352,
|
| 475 |
+
0.9852,
|
| 476 |
+
0.3683,
|
| 477 |
+
0.1881,
|
| 478 |
+
0.9172,
|
| 479 |
+
0.26,
|
| 480 |
+
0.1638,
|
| 481 |
+
0.1679,
|
| 482 |
+
0.2288,
|
| 483 |
+
0.3319,
|
| 484 |
+
0.285,
|
| 485 |
+
0.7975,
|
| 486 |
+
0.1772,
|
| 487 |
+
0.1633,
|
| 488 |
+
0.206,
|
| 489 |
+
0.1612,
|
| 490 |
+
0.1813,
|
| 491 |
+
0.7064,
|
| 492 |
+
1.0078,
|
| 493 |
+
0.1998,
|
| 494 |
+
0.4962,
|
| 495 |
+
0.2323,
|
| 496 |
+
0.1961
|
| 497 |
+
],
|
| 498 |
+
"kd": [
|
| 499 |
+
0.2179,
|
| 500 |
+
0.3024,
|
| 501 |
+
0.1595,
|
| 502 |
+
0.2015,
|
| 503 |
+
0.2628,
|
| 504 |
+
0.173,
|
| 505 |
+
0.1893,
|
| 506 |
+
0.1603,
|
| 507 |
+
0.1758,
|
| 508 |
+
0.355,
|
| 509 |
+
0.2048,
|
| 510 |
+
0.28,
|
| 511 |
+
0.1252,
|
| 512 |
+
0.3447,
|
| 513 |
+
0.2781,
|
| 514 |
+
0.1793,
|
| 515 |
+
0.2006,
|
| 516 |
+
0.1195,
|
| 517 |
+
0.1742,
|
| 518 |
+
0.1541,
|
| 519 |
+
0.1475,
|
| 520 |
+
0.1286,
|
| 521 |
+
0.106,
|
| 522 |
+
0.2127,
|
| 523 |
+
0.1964,
|
| 524 |
+
0.2001,
|
| 525 |
+
0.1626,
|
| 526 |
+
0.1501,
|
| 527 |
+
0.0717,
|
| 528 |
+
0.1333,
|
| 529 |
+
0.2157,
|
| 530 |
+
0.1164,
|
| 531 |
+
0.1196,
|
| 532 |
+
0.2087,
|
| 533 |
+
0.2284,
|
| 534 |
+
0.0941,
|
| 535 |
+
0.203,
|
| 536 |
+
0.1532,
|
| 537 |
+
0.1405,
|
| 538 |
+
0.1537
|
| 539 |
+
],
|
| 540 |
+
"task": [
|
| 541 |
+
0.0,
|
| 542 |
+
0.0,
|
| 543 |
+
0.0,
|
| 544 |
+
0.0,
|
| 545 |
+
0.0,
|
| 546 |
+
0.0,
|
| 547 |
+
0.0,
|
| 548 |
+
0.0,
|
| 549 |
+
0.0,
|
| 550 |
+
0.0,
|
| 551 |
+
0.0,
|
| 552 |
+
0.0,
|
| 553 |
+
0.0,
|
| 554 |
+
0.0,
|
| 555 |
+
0.0,
|
| 556 |
+
0.0,
|
| 557 |
+
0.0,
|
| 558 |
+
0.0,
|
| 559 |
+
0.0,
|
| 560 |
+
0.0,
|
| 561 |
+
0.0,
|
| 562 |
+
0.0,
|
| 563 |
+
0.0,
|
| 564 |
+
0.0,
|
| 565 |
+
0.0,
|
| 566 |
+
0.0,
|
| 567 |
+
0.0,
|
| 568 |
+
0.0,
|
| 569 |
+
0.0,
|
| 570 |
+
0.0,
|
| 571 |
+
0.0,
|
| 572 |
+
0.0,
|
| 573 |
+
0.0,
|
| 574 |
+
0.0,
|
| 575 |
+
0.0,
|
| 576 |
+
0.0,
|
| 577 |
+
0.0,
|
| 578 |
+
0.0,
|
| 579 |
+
0.0,
|
| 580 |
+
0.0
|
| 581 |
+
],
|
| 582 |
+
"diversity": [
|
| 583 |
+
0.0139,
|
| 584 |
+
0.0181,
|
| 585 |
+
0.0776,
|
| 586 |
+
0.0341,
|
| 587 |
+
0.0231,
|
| 588 |
+
0.0645,
|
| 589 |
+
0.0352,
|
| 590 |
+
0.0634,
|
| 591 |
+
0.035,
|
| 592 |
+
0.0922,
|
| 593 |
+
0.0768,
|
| 594 |
+
0.1247,
|
| 595 |
+
0.2491,
|
| 596 |
+
0.14,
|
| 597 |
+
0.1164,
|
| 598 |
+
0.1522,
|
| 599 |
+
0.219,
|
| 600 |
+
0.2465,
|
| 601 |
+
0.249,
|
| 602 |
+
0.2203,
|
| 603 |
+
0.2989,
|
| 604 |
+
0.2741,
|
| 605 |
+
0.3376,
|
| 606 |
+
0.2201,
|
| 607 |
+
0.3444,
|
| 608 |
+
0.292,
|
| 609 |
+
0.3873,
|
| 610 |
+
0.4255,
|
| 611 |
+
0.393,
|
| 612 |
+
0.3387,
|
| 613 |
+
0.4708,
|
| 614 |
+
0.3341,
|
| 615 |
+
0.5267,
|
| 616 |
+
0.5546,
|
| 617 |
+
0.4303,
|
| 618 |
+
0.4871,
|
| 619 |
+
0.4483,
|
| 620 |
+
0.5465,
|
| 621 |
+
0.518,
|
| 622 |
+
0.4212
|
| 623 |
+
],
|
| 624 |
+
"consistency": [
|
| 625 |
+
6.9963,
|
| 626 |
+
0.401,
|
| 627 |
+
0.8276,
|
| 628 |
+
0.4812,
|
| 629 |
+
6.4711,
|
| 630 |
+
10.3427,
|
| 631 |
+
1.4685,
|
| 632 |
+
0.4777,
|
| 633 |
+
9.12,
|
| 634 |
+
4.8882,
|
| 635 |
+
5.3051,
|
| 636 |
+
2.8592,
|
| 637 |
+
0.6606,
|
| 638 |
+
1.182,
|
| 639 |
+
0.4008,
|
| 640 |
+
1.5355,
|
| 641 |
+
0.5742,
|
| 642 |
+
0.5717,
|
| 643 |
+
8.7692,
|
| 644 |
+
2.7252,
|
| 645 |
+
0.9205,
|
| 646 |
+
8.3271,
|
| 647 |
+
1.8482,
|
| 648 |
+
0.3581,
|
| 649 |
+
0.4269,
|
| 650 |
+
1.0412,
|
| 651 |
+
2.2313,
|
| 652 |
+
1.8119,
|
| 653 |
+
7.3844,
|
| 654 |
+
0.8692,
|
| 655 |
+
0.2118,
|
| 656 |
+
1.2531,
|
| 657 |
+
0.6906,
|
| 658 |
+
0.3872,
|
| 659 |
+
5.5925,
|
| 660 |
+
9.3167,
|
| 661 |
+
0.6574,
|
| 662 |
+
3.8459,
|
| 663 |
+
1.2916,
|
| 664 |
+
0.9047
|
| 665 |
+
]
|
| 666 |
+
},
|
| 667 |
+
"loss_start": 0.8202,
|
| 668 |
+
"loss_end": 0.1961,
|
| 669 |
+
"loss_reduction_pct": 76.1,
|
| 670 |
+
"diversity_start": 0.0139,
|
| 671 |
+
"diversity_end": 0.4212,
|
| 672 |
+
"consistency_start": 6.9963,
|
| 673 |
+
"consistency_end": 0.9047,
|
| 674 |
+
"final_router_weights": [
|
| 675 |
+
0.6852,
|
| 676 |
+
0.2771,
|
| 677 |
+
0.0378
|
| 678 |
+
],
|
| 679 |
+
"peak_gpu_gb": 5.68
|
| 680 |
+
},
|
| 681 |
+
"kd_heavy": {
|
| 682 |
+
"config": {
|
| 683 |
+
"name": "kd_heavy",
|
| 684 |
+
"alpha_task": 0.1,
|
| 685 |
+
"alpha_diversity": 0.05,
|
| 686 |
+
"alpha_consistency": 0.05
|
| 687 |
+
},
|
| 688 |
+
"n_teachers": 3,
|
| 689 |
+
"n_steps": 40,
|
| 690 |
+
"loss_curves": {
|
| 691 |
+
"total": [
|
| 692 |
+
0.2962,
|
| 693 |
+
0.203,
|
| 694 |
+
0.2157,
|
| 695 |
+
0.3773,
|
| 696 |
+
0.2535,
|
| 697 |
+
0.2469,
|
| 698 |
+
0.3556,
|
| 699 |
+
0.1727,
|
| 700 |
+
0.2806,
|
| 701 |
+
0.1914,
|
| 702 |
+
0.2855,
|
| 703 |
+
0.1582,
|
| 704 |
+
0.3133,
|
| 705 |
+
0.1934,
|
| 706 |
+
0.3772,
|
| 707 |
+
0.2284,
|
| 708 |
+
0.1819,
|
| 709 |
+
0.1891,
|
| 710 |
+
0.3112,
|
| 711 |
+
0.2335,
|
| 712 |
+
0.1611,
|
| 713 |
+
0.1674,
|
| 714 |
+
0.221,
|
| 715 |
+
0.1682,
|
| 716 |
+
0.1457,
|
| 717 |
+
1.096,
|
| 718 |
+
0.2961,
|
| 719 |
+
0.1963,
|
| 720 |
+
0.2676,
|
| 721 |
+
0.2331,
|
| 722 |
+
0.2775,
|
| 723 |
+
0.147,
|
| 724 |
+
0.2782,
|
| 725 |
+
0.3111,
|
| 726 |
+
0.1405,
|
| 727 |
+
0.5473,
|
| 728 |
+
0.1188,
|
| 729 |
+
0.3153,
|
| 730 |
+
0.2511,
|
| 731 |
+
0.269
|
| 732 |
+
],
|
| 733 |
+
"kd": [
|
| 734 |
+
0.2503,
|
| 735 |
+
0.222,
|
| 736 |
+
0.2271,
|
| 737 |
+
0.2296,
|
| 738 |
+
0.2532,
|
| 739 |
+
0.229,
|
| 740 |
+
0.1399,
|
| 741 |
+
0.1868,
|
| 742 |
+
0.2751,
|
| 743 |
+
0.1493,
|
| 744 |
+
0.1777,
|
| 745 |
+
0.1576,
|
| 746 |
+
0.2593,
|
| 747 |
+
0.2025,
|
| 748 |
+
0.1483,
|
| 749 |
+
0.1816,
|
| 750 |
+
0.1433,
|
| 751 |
+
0.2145,
|
| 752 |
+
0.1792,
|
| 753 |
+
0.1472,
|
| 754 |
+
0.1165,
|
| 755 |
+
0.1657,
|
| 756 |
+
0.1607,
|
| 757 |
+
0.1714,
|
| 758 |
+
0.1048,
|
| 759 |
+
0.1592,
|
| 760 |
+
0.2288,
|
| 761 |
+
0.1744,
|
| 762 |
+
0.2124,
|
| 763 |
+
0.1997,
|
| 764 |
+
0.1947,
|
| 765 |
+
0.1225,
|
| 766 |
+
0.1971,
|
| 767 |
+
0.2119,
|
| 768 |
+
0.1391,
|
| 769 |
+
0.1351,
|
| 770 |
+
0.0765,
|
| 771 |
+
0.3373,
|
| 772 |
+
0.0688,
|
| 773 |
+
0.2348
|
| 774 |
+
],
|
| 775 |
+
"task": [
|
| 776 |
+
0.0,
|
| 777 |
+
0.0,
|
| 778 |
+
0.0,
|
| 779 |
+
0.0,
|
| 780 |
+
0.0,
|
| 781 |
+
0.0,
|
| 782 |
+
0.0,
|
| 783 |
+
0.0,
|
| 784 |
+
0.0,
|
| 785 |
+
0.0,
|
| 786 |
+
0.0,
|
| 787 |
+
0.0,
|
| 788 |
+
0.0,
|
| 789 |
+
0.0,
|
| 790 |
+
0.0,
|
| 791 |
+
0.0,
|
| 792 |
+
0.0,
|
| 793 |
+
0.0,
|
| 794 |
+
0.0,
|
| 795 |
+
0.0,
|
| 796 |
+
0.0,
|
| 797 |
+
0.0,
|
| 798 |
+
0.0,
|
| 799 |
+
0.0,
|
| 800 |
+
0.0,
|
| 801 |
+
0.0,
|
| 802 |
+
0.0,
|
| 803 |
+
0.0,
|
| 804 |
+
0.0,
|
| 805 |
+
0.0,
|
| 806 |
+
0.0,
|
| 807 |
+
0.0,
|
| 808 |
+
0.0,
|
| 809 |
+
0.0,
|
| 810 |
+
0.0,
|
| 811 |
+
0.0,
|
| 812 |
+
0.0,
|
| 813 |
+
0.0,
|
| 814 |
+
0.0,
|
| 815 |
+
0.0
|
| 816 |
+
],
|
| 817 |
+
"diversity": [
|
| 818 |
+
0.0046,
|
| 819 |
+
0.0034,
|
| 820 |
+
0.0032,
|
| 821 |
+
0.0282,
|
| 822 |
+
0.0231,
|
| 823 |
+
0.1028,
|
| 824 |
+
0.0295,
|
| 825 |
+
0.0375,
|
| 826 |
+
0.0828,
|
| 827 |
+
0.047,
|
| 828 |
+
0.0818,
|
| 829 |
+
0.082,
|
| 830 |
+
0.0881,
|
| 831 |
+
0.0623,
|
| 832 |
+
0.1426,
|
| 833 |
+
0.3155,
|
| 834 |
+
0.1484,
|
| 835 |
+
0.072,
|
| 836 |
+
0.153,
|
| 837 |
+
0.0923,
|
| 838 |
+
0.1533,
|
| 839 |
+
0.1999,
|
| 840 |
+
0.2933,
|
| 841 |
+
0.1957,
|
| 842 |
+
0.2222,
|
| 843 |
+
0.2705,
|
| 844 |
+
0.2611,
|
| 845 |
+
0.442,
|
| 846 |
+
0.3867,
|
| 847 |
+
0.312,
|
| 848 |
+
0.3789,
|
| 849 |
+
0.341,
|
| 850 |
+
0.3485,
|
| 851 |
+
0.4008,
|
| 852 |
+
0.4957,
|
| 853 |
+
0.4744,
|
| 854 |
+
0.4763,
|
| 855 |
+
0.513,
|
| 856 |
+
0.5425,
|
| 857 |
+
0.4803
|
| 858 |
+
],
|
| 859 |
+
"consistency": [
|
| 860 |
+
1.9136,
|
| 861 |
+
0.5042,
|
| 862 |
+
0.677,
|
| 863 |
+
3.8427,
|
| 864 |
+
0.9952,
|
| 865 |
+
1.1709,
|
| 866 |
+
4.8428,
|
| 867 |
+
0.4271,
|
| 868 |
+
1.1276,
|
| 869 |
+
1.3927,
|
| 870 |
+
2.7848,
|
| 871 |
+
0.5604,
|
| 872 |
+
2.0293,
|
| 873 |
+
0.5647,
|
| 874 |
+
5.0286,
|
| 875 |
+
1.3479,
|
| 876 |
+
1.1967,
|
| 877 |
+
0.2778,
|
| 878 |
+
3.2046,
|
| 879 |
+
2.2228,
|
| 880 |
+
1.2041,
|
| 881 |
+
0.4974,
|
| 882 |
+
1.5555,
|
| 883 |
+
0.4264,
|
| 884 |
+
1.0152,
|
| 885 |
+
19.1024,
|
| 886 |
+
2.0002,
|
| 887 |
+
0.6938,
|
| 888 |
+
1.5658,
|
| 889 |
+
1.1545,
|
| 890 |
+
2.057,
|
| 891 |
+
0.639,
|
| 892 |
+
2.0621,
|
| 893 |
+
2.4296,
|
| 894 |
+
0.0892,
|
| 895 |
+
8.3113,
|
| 896 |
+
0.6766,
|
| 897 |
+
0.3963,
|
| 898 |
+
3.3787,
|
| 899 |
+
1.1435
|
| 900 |
+
]
|
| 901 |
+
},
|
| 902 |
+
"loss_start": 0.2962,
|
| 903 |
+
"loss_end": 0.269,
|
| 904 |
+
"loss_reduction_pct": 9.2,
|
| 905 |
+
"diversity_start": 0.0046,
|
| 906 |
+
"diversity_end": 0.4803,
|
| 907 |
+
"consistency_start": 1.9136,
|
| 908 |
+
"consistency_end": 1.1435,
|
| 909 |
+
"final_router_weights": [
|
| 910 |
+
0.7951,
|
| 911 |
+
0.1155,
|
| 912 |
+
0.0894
|
| 913 |
+
],
|
| 914 |
+
"peak_gpu_gb": 6.88
|
| 915 |
+
},
|
| 916 |
+
"diverse": {
|
| 917 |
+
"config": {
|
| 918 |
+
"name": "diverse",
|
| 919 |
+
"alpha_task": 0.2,
|
| 920 |
+
"alpha_diversity": 0.15,
|
| 921 |
+
"alpha_consistency": 0.1
|
| 922 |
+
},
|
| 923 |
+
"n_teachers": 3,
|
| 924 |
+
"n_steps": 40,
|
| 925 |
+
"loss_curves": {
|
| 926 |
+
"total": [
|
| 927 |
+
0.3073,
|
| 928 |
+
0.3902,
|
| 929 |
+
0.1624,
|
| 930 |
+
0.3824,
|
| 931 |
+
0.3597,
|
| 932 |
+
0.1845,
|
| 933 |
+
0.1476,
|
| 934 |
+
0.2143,
|
| 935 |
+
0.31,
|
| 936 |
+
0.5431,
|
| 937 |
+
0.3286,
|
| 938 |
+
0.4385,
|
| 939 |
+
0.2854,
|
| 940 |
+
0.3086,
|
| 941 |
+
0.1937,
|
| 942 |
+
0.3604,
|
| 943 |
+
0.23,
|
| 944 |
+
0.1601,
|
| 945 |
+
0.1948,
|
| 946 |
+
0.109,
|
| 947 |
+
0.2419,
|
| 948 |
+
0.2447,
|
| 949 |
+
0.2458,
|
| 950 |
+
0.2329,
|
| 951 |
+
0.2035,
|
| 952 |
+
0.7461,
|
| 953 |
+
0.1736,
|
| 954 |
+
0.4718,
|
| 955 |
+
0.2367,
|
| 956 |
+
0.4103,
|
| 957 |
+
0.1619,
|
| 958 |
+
0.2883,
|
| 959 |
+
0.3277,
|
| 960 |
+
0.4079,
|
| 961 |
+
0.3442,
|
| 962 |
+
0.2008,
|
| 963 |
+
0.3297,
|
| 964 |
+
0.2013,
|
| 965 |
+
0.445,
|
| 966 |
+
2.0527
|
| 967 |
+
],
|
| 968 |
+
"kd": [
|
| 969 |
+
0.4063,
|
| 970 |
+
0.3409,
|
| 971 |
+
0.242,
|
| 972 |
+
0.3398,
|
| 973 |
+
0.3296,
|
| 974 |
+
0.2204,
|
| 975 |
+
0.1434,
|
| 976 |
+
0.2839,
|
| 977 |
+
0.2216,
|
| 978 |
+
0.2332,
|
| 979 |
+
0.12,
|
| 980 |
+
0.3295,
|
| 981 |
+
0.2716,
|
| 982 |
+
0.3083,
|
| 983 |
+
0.2472,
|
| 984 |
+
0.2547,
|
| 985 |
+
0.2139,
|
| 986 |
+
0.1458,
|
| 987 |
+
0.24,
|
| 988 |
+
0.109,
|
| 989 |
+
0.2901,
|
| 990 |
+
0.2125,
|
| 991 |
+
0.3603,
|
| 992 |
+
0.1676,
|
| 993 |
+
0.1387,
|
| 994 |
+
0.099,
|
| 995 |
+
0.1803,
|
| 996 |
+
0.2299,
|
| 997 |
+
0.2543,
|
| 998 |
+
0.1848,
|
| 999 |
+
0.1258,
|
| 1000 |
+
0.2009,
|
| 1001 |
+
0.249,
|
| 1002 |
+
0.3668,
|
| 1003 |
+
0.2625,
|
| 1004 |
+
0.2157,
|
| 1005 |
+
0.2359,
|
| 1006 |
+
0.2398,
|
| 1007 |
+
0.2816,
|
| 1008 |
+
0.1966
|
| 1009 |
+
],
|
| 1010 |
+
"task": [
|
| 1011 |
+
0.0,
|
| 1012 |
+
0.0,
|
| 1013 |
+
0.0,
|
| 1014 |
+
0.0,
|
| 1015 |
+
0.0,
|
| 1016 |
+
0.0,
|
| 1017 |
+
0.0,
|
| 1018 |
+
0.0,
|
| 1019 |
+
0.0,
|
| 1020 |
+
0.0,
|
| 1021 |
+
0.0,
|
| 1022 |
+
0.0,
|
| 1023 |
+
0.0,
|
| 1024 |
+
0.0,
|
| 1025 |
+
0.0,
|
| 1026 |
+
0.0,
|
| 1027 |
+
0.0,
|
| 1028 |
+
0.0,
|
| 1029 |
+
0.0,
|
| 1030 |
+
0.0,
|
| 1031 |
+
0.0,
|
| 1032 |
+
0.0,
|
| 1033 |
+
0.0,
|
| 1034 |
+
0.0,
|
| 1035 |
+
0.0,
|
| 1036 |
+
0.0,
|
| 1037 |
+
0.0,
|
| 1038 |
+
0.0,
|
| 1039 |
+
0.0,
|
| 1040 |
+
0.0,
|
| 1041 |
+
0.0,
|
| 1042 |
+
0.0,
|
| 1043 |
+
0.0,
|
| 1044 |
+
0.0,
|
| 1045 |
+
0.0,
|
| 1046 |
+
0.0,
|
| 1047 |
+
0.0,
|
| 1048 |
+
0.0,
|
| 1049 |
+
0.0,
|
| 1050 |
+
0.0
|
| 1051 |
+
],
|
| 1052 |
+
"diversity": [
|
| 1053 |
+
0.0017,
|
| 1054 |
+
0.0086,
|
| 1055 |
+
0.0089,
|
| 1056 |
+
0.0328,
|
| 1057 |
+
0.0505,
|
| 1058 |
+
0.0569,
|
| 1059 |
+
0.0662,
|
| 1060 |
+
0.0123,
|
| 1061 |
+
0.0314,
|
| 1062 |
+
0.0858,
|
| 1063 |
+
0.0396,
|
| 1064 |
+
0.0534,
|
| 1065 |
+
0.0789,
|
| 1066 |
+
0.1071,
|
| 1067 |
+
0.1216,
|
| 1068 |
+
0.0297,
|
| 1069 |
+
0.072,
|
| 1070 |
+
0.163,
|
| 1071 |
+
0.1427,
|
| 1072 |
+
0.1169,
|
| 1073 |
+
0.1055,
|
| 1074 |
+
0.1664,
|
| 1075 |
+
0.1086,
|
| 1076 |
+
0.1434,
|
| 1077 |
+
0.2265,
|
| 1078 |
+
0.1857,
|
| 1079 |
+
0.0884,
|
| 1080 |
+
0.1191,
|
| 1081 |
+
0.0493,
|
| 1082 |
+
0.0645,
|
| 1083 |
+
0.2075,
|
| 1084 |
+
0.1427,
|
| 1085 |
+
0.0942,
|
| 1086 |
+
0.061,
|
| 1087 |
+
0.1197,
|
| 1088 |
+
0.1151,
|
| 1089 |
+
0.142,
|
| 1090 |
+
0.1033,
|
| 1091 |
+
0.0755,
|
| 1092 |
+
0.1249
|
| 1093 |
+
],
|
| 1094 |
+
"consistency": [
|
| 1095 |
+
0.8359,
|
| 1096 |
+
2.0139,
|
| 1097 |
+
0.2794,
|
| 1098 |
+
1.9054,
|
| 1099 |
+
1.7085,
|
| 1100 |
+
0.5477,
|
| 1101 |
+
0.5877,
|
| 1102 |
+
0.5637,
|
| 1103 |
+
1.8343,
|
| 1104 |
+
4.0196,
|
| 1105 |
+
2.5665,
|
| 1106 |
+
2.4922,
|
| 1107 |
+
1.2415,
|
| 1108 |
+
1.2295,
|
| 1109 |
+
0.3951,
|
| 1110 |
+
2.1588,
|
| 1111 |
+
1.015,
|
| 1112 |
+
0.5543,
|
| 1113 |
+
0.4135,
|
| 1114 |
+
0.3159,
|
| 1115 |
+
0.6654,
|
| 1116 |
+
1.0287,
|
| 1117 |
+
0.3127,
|
| 1118 |
+
1.192,
|
| 1119 |
+
0.9326,
|
| 1120 |
+
6.6385,
|
| 1121 |
+
0.6114,
|
| 1122 |
+
3.2752,
|
| 1123 |
+
0.8949,
|
| 1124 |
+
2.9897,
|
| 1125 |
+
0.6156,
|
| 1126 |
+
1.5635,
|
| 1127 |
+
1.7659,
|
| 1128 |
+
1.9699,
|
| 1129 |
+
1.8183,
|
| 1130 |
+
0.6497,
|
| 1131 |
+
1.7863,
|
| 1132 |
+
0.5384,
|
| 1133 |
+
2.7877,
|
| 1134 |
+
19.259
|
| 1135 |
+
]
|
| 1136 |
+
},
|
| 1137 |
+
"loss_start": 0.3073,
|
| 1138 |
+
"loss_end": 2.0527,
|
| 1139 |
+
"loss_reduction_pct": -568.0,
|
| 1140 |
+
"diversity_start": 0.0017,
|
| 1141 |
+
"diversity_end": 0.1249,
|
| 1142 |
+
"consistency_start": 0.8359,
|
| 1143 |
+
"consistency_end": 19.259,
|
| 1144 |
+
"final_router_weights": [
|
| 1145 |
+
0.6264,
|
| 1146 |
+
0.2669,
|
| 1147 |
+
0.1067
|
| 1148 |
+
],
|
| 1149 |
+
"peak_gpu_gb": 6.87
|
| 1150 |
+
}
|
| 1151 |
+
},
|
| 1152 |
+
"teacher_per_gpu": {
|
| 1153 |
+
"n_teachers": 3,
|
| 1154 |
+
"n_gpus_used": 4,
|
| 1155 |
+
"sequential_mean_ms": 441.21,
|
| 1156 |
+
"sequential_p50_ms": 442.35,
|
| 1157 |
+
"parallel_mean_ms": 439.04,
|
| 1158 |
+
"parallel_p50_ms": 439.03,
|
| 1159 |
+
"speedup": 1.0
|
| 1160 |
+
}
|
| 1161 |
+
}
|