Upload evals for think-d12-r11.25-ctx4096
Browse files
experiments/think-d12-r11.25-ctx4096/evals/val_bpb.json
CHANGED
|
@@ -6,85 +6,85 @@
|
|
| 6 |
{
|
| 7 |
"start": 0,
|
| 8 |
"end": 256,
|
| 9 |
-
"bpb": 1.
|
| 10 |
},
|
| 11 |
{
|
| 12 |
"start": 256,
|
| 13 |
"end": 512,
|
| 14 |
-
"bpb": 1.
|
| 15 |
},
|
| 16 |
{
|
| 17 |
"start": 512,
|
| 18 |
"end": 768,
|
| 19 |
-
"bpb": 1.
|
| 20 |
},
|
| 21 |
{
|
| 22 |
"start": 768,
|
| 23 |
"end": 1024,
|
| 24 |
-
"bpb": 1.
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"start": 1024,
|
| 28 |
"end": 1280,
|
| 29 |
-
"bpb": 1.
|
| 30 |
},
|
| 31 |
{
|
| 32 |
"start": 1280,
|
| 33 |
"end": 1536,
|
| 34 |
-
"bpb": 1.
|
| 35 |
},
|
| 36 |
{
|
| 37 |
"start": 1536,
|
| 38 |
"end": 1792,
|
| 39 |
-
"bpb": 1.
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"start": 1792,
|
| 43 |
"end": 2048,
|
| 44 |
-
"bpb": 1.
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"start": 2048,
|
| 48 |
"end": 2304,
|
| 49 |
-
"bpb": 1.
|
| 50 |
},
|
| 51 |
{
|
| 52 |
"start": 2304,
|
| 53 |
"end": 2560,
|
| 54 |
-
"bpb": 1.
|
| 55 |
},
|
| 56 |
{
|
| 57 |
"start": 2560,
|
| 58 |
"end": 2816,
|
| 59 |
-
"bpb": 1.
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"start": 2816,
|
| 63 |
"end": 3072,
|
| 64 |
-
"bpb": 1.
|
| 65 |
},
|
| 66 |
{
|
| 67 |
"start": 3072,
|
| 68 |
"end": 3328,
|
| 69 |
-
"bpb": 1.
|
| 70 |
},
|
| 71 |
{
|
| 72 |
"start": 3328,
|
| 73 |
"end": 3584,
|
| 74 |
-
"bpb": 1.
|
| 75 |
},
|
| 76 |
{
|
| 77 |
"start": 3584,
|
| 78 |
"end": 3840,
|
| 79 |
-
"bpb": 1.
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"start": 3840,
|
| 83 |
"end": 4096,
|
| 84 |
-
"bpb": 1.
|
| 85 |
}
|
| 86 |
],
|
| 87 |
-
"val": 1.
|
| 88 |
},
|
| 89 |
"core_metric": null,
|
| 90 |
"core_results": null,
|
|
|
|
| 6 |
{
|
| 7 |
"start": 0,
|
| 8 |
"end": 256,
|
| 9 |
+
"bpb": 1.130142936359072
|
| 10 |
},
|
| 11 |
{
|
| 12 |
"start": 256,
|
| 13 |
"end": 512,
|
| 14 |
+
"bpb": 1.065738621390211
|
| 15 |
},
|
| 16 |
{
|
| 17 |
"start": 512,
|
| 18 |
"end": 768,
|
| 19 |
+
"bpb": 1.0495748692031812
|
| 20 |
},
|
| 21 |
{
|
| 22 |
"start": 768,
|
| 23 |
"end": 1024,
|
| 24 |
+
"bpb": 1.0430983472198936
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"start": 1024,
|
| 28 |
"end": 1280,
|
| 29 |
+
"bpb": 1.0366271994096037
|
| 30 |
},
|
| 31 |
{
|
| 32 |
"start": 1280,
|
| 33 |
"end": 1536,
|
| 34 |
+
"bpb": 1.0297924628551816
|
| 35 |
},
|
| 36 |
{
|
| 37 |
"start": 1536,
|
| 38 |
"end": 1792,
|
| 39 |
+
"bpb": 1.0281174126636314
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"start": 1792,
|
| 43 |
"end": 2048,
|
| 44 |
+
"bpb": 1.0244993749695677
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"start": 2048,
|
| 48 |
"end": 2304,
|
| 49 |
+
"bpb": 1.020344951749226
|
| 50 |
},
|
| 51 |
{
|
| 52 |
"start": 2304,
|
| 53 |
"end": 2560,
|
| 54 |
+
"bpb": 1.017900061022826
|
| 55 |
},
|
| 56 |
{
|
| 57 |
"start": 2560,
|
| 58 |
"end": 2816,
|
| 59 |
+
"bpb": 1.017925884634094
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"start": 2816,
|
| 63 |
"end": 3072,
|
| 64 |
+
"bpb": 1.0206878528813161
|
| 65 |
},
|
| 66 |
{
|
| 67 |
"start": 3072,
|
| 68 |
"end": 3328,
|
| 69 |
+
"bpb": 1.0171357038345088
|
| 70 |
},
|
| 71 |
{
|
| 72 |
"start": 3328,
|
| 73 |
"end": 3584,
|
| 74 |
+
"bpb": 1.0150316846511822
|
| 75 |
},
|
| 76 |
{
|
| 77 |
"start": 3584,
|
| 78 |
"end": 3840,
|
| 79 |
+
"bpb": 1.0124244006389753
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"start": 3840,
|
| 83 |
"end": 4096,
|
| 84 |
+
"bpb": 1.0101240407476038
|
| 85 |
}
|
| 86 |
],
|
| 87 |
+
"val": 1.0336899522729732
|
| 88 |
},
|
| 89 |
"core_metric": null,
|
| 90 |
"core_results": null,
|