Upload reports/mbg_l1-17m_20260901-173235_report.json with huggingface_hub
Browse files
reports/mbg_l1-17m_20260901-173235_report.json
ADDED
|
@@ -0,0 +1,46 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"params": 16748928,
|
| 3 |
+
"optimizer": "gum",
|
| 4 |
+
"tokenizer": "bpe",
|
| 5 |
+
"steps": 150,
|
| 6 |
+
"wall_s": 171.14644646644592,
|
| 7 |
+
"history": [
|
| 8 |
+
{
|
| 9 |
+
"step": 1,
|
| 10 |
+
"train_loss": 6.9811859130859375,
|
| 11 |
+
"val_loss": 6.946277678012848
|
| 12 |
+
},
|
| 13 |
+
{
|
| 14 |
+
"step": 50,
|
| 15 |
+
"train_loss": 6.513549327850342,
|
| 16 |
+
"val_loss": 6.5723422010739645
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"step": 100,
|
| 20 |
+
"train_loss": 6.081512928009033,
|
| 21 |
+
"val_loss": 6.110627841949463
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"step": 150,
|
| 25 |
+
"train_loss": 5.763103008270264,
|
| 26 |
+
"val_loss": 5.738713745276133
|
| 27 |
+
}
|
| 28 |
+
],
|
| 29 |
+
"routing": {
|
| 30 |
+
"CAUSAL": {
|
| 31 |
+
"matched": 5.851107915242513,
|
| 32 |
+
"mismatched": 5.970818837483724,
|
| 33 |
+
"delta": 0.11971092224121094
|
| 34 |
+
},
|
| 35 |
+
"SPATIAL": {
|
| 36 |
+
"matched": 5.613209327061971,
|
| 37 |
+
"mismatched": 5.669022401173909,
|
| 38 |
+
"delta": 0.05581307411193848
|
| 39 |
+
},
|
| 40 |
+
"TEMPORAL": {
|
| 41 |
+
"matched": 5.666887839635213,
|
| 42 |
+
"mismatched": 5.7906317710876465,
|
| 43 |
+
"delta": 0.12374393145243327
|
| 44 |
+
}
|
| 45 |
+
}
|
| 46 |
+
}
|