Upload sigma_lambda_proto_v3/history.json with huggingface_hub
Browse files
sigma_lambda_proto_v3/history.json
CHANGED
|
@@ -3,96 +3,112 @@
|
|
| 3 |
1,
|
| 4 |
2,
|
| 5 |
3,
|
| 6 |
-
4
|
|
|
|
| 7 |
],
|
| 8 |
"gen_loss": [
|
| 9 |
18.370433433850607,
|
| 10 |
12.677154767152034,
|
| 11 |
12.081629503279022,
|
| 12 |
-
11.843723264174027
|
|
|
|
| 13 |
],
|
| 14 |
"ret_loss": [
|
| 15 |
0.0,
|
| 16 |
0.0,
|
| 17 |
0.0,
|
|
|
|
| 18 |
0.0
|
| 19 |
],
|
| 20 |
"ent_loss": [
|
| 21 |
0.6078802872065342,
|
| 22 |
0.6268522234035261,
|
| 23 |
0.6330166747714534,
|
| 24 |
-
0.6162352153026697
|
|
|
|
| 25 |
],
|
| 26 |
"ppl": [
|
| 27 |
40729.60576844635,
|
| 28 |
2535.223731428122,
|
| 29 |
337.67548432577263,
|
| 30 |
-
123.54939191255174
|
|
|
|
| 31 |
],
|
| 32 |
"tf_acc": [
|
| 33 |
9.249682110189831e-06,
|
| 34 |
0.21421828488041522,
|
| 35 |
0.39405658955515016,
|
| 36 |
-
0.4409562865464462
|
|
|
|
| 37 |
],
|
| 38 |
"fr_acc": [
|
| 39 |
0.0,
|
| 40 |
0.0,
|
| 41 |
0.11453107816089203,
|
| 42 |
-
0.09257571660973458
|
|
|
|
| 43 |
],
|
| 44 |
"def_rate": [
|
| 45 |
0.0,
|
| 46 |
0.0,
|
| 47 |
0.0,
|
| 48 |
-
0.45
|
|
|
|
| 49 |
],
|
| 50 |
"mrr": [
|
| 51 |
0.004940100014209747,
|
| 52 |
0.005048312712460756,
|
| 53 |
0.004967095330357552,
|
| 54 |
-
0.004980043042451143
|
|
|
|
| 55 |
],
|
| 56 |
"r1": [
|
| 57 |
0.0006249999860301614,
|
| 58 |
0.0006249999860301614,
|
| 59 |
0.0006249999860301614,
|
|
|
|
| 60 |
0.0006249999860301614
|
| 61 |
],
|
| 62 |
"ret_ok_pct": [
|
| 63 |
0.0,
|
| 64 |
0.0,
|
| 65 |
0.0,
|
|
|
|
| 66 |
0.0
|
| 67 |
],
|
| 68 |
"grad_norm": [
|
| 69 |
2201.0254727094825,
|
| 70 |
1418.5299411384265,
|
| 71 |
1329.5837538555174,
|
| 72 |
-
1217.5837738567411
|
|
|
|
| 73 |
],
|
| 74 |
"step_var": [
|
| 75 |
0.22542811930179596,
|
| 76 |
0.22660225629806519,
|
| 77 |
0.2280091941356659,
|
| 78 |
-
0.22952058911323547
|
|
|
|
| 79 |
],
|
| 80 |
"pos_slope": [
|
| 81 |
0.49990350008010864,
|
| 82 |
0.4996356666088104,
|
| 83 |
0.49926072359085083,
|
| 84 |
-
0.4988322854042053
|
|
|
|
| 85 |
],
|
| 86 |
"beta_mean": [
|
| 87 |
0.09998118132352829,
|
| 88 |
0.09992741793394089,
|
| 89 |
0.09985271841287613,
|
| 90 |
-
0.09976637363433838
|
|
|
|
| 91 |
],
|
| 92 |
"w_ret": [
|
| 93 |
0.0,
|
| 94 |
0.0,
|
| 95 |
0.0,
|
| 96 |
-
0.0
|
|
|
|
| 97 |
]
|
| 98 |
}
|
|
|
|
| 3 |
1,
|
| 4 |
2,
|
| 5 |
3,
|
| 6 |
+
4,
|
| 7 |
+
5
|
| 8 |
],
|
| 9 |
"gen_loss": [
|
| 10 |
18.370433433850607,
|
| 11 |
12.677154767152034,
|
| 12 |
12.081629503279022,
|
| 13 |
+
11.843723264174027,
|
| 14 |
+
11.67701819376512
|
| 15 |
],
|
| 16 |
"ret_loss": [
|
| 17 |
0.0,
|
| 18 |
0.0,
|
| 19 |
0.0,
|
| 20 |
+
0.0,
|
| 21 |
0.0
|
| 22 |
],
|
| 23 |
"ent_loss": [
|
| 24 |
0.6078802872065342,
|
| 25 |
0.6268522234035261,
|
| 26 |
0.6330166747714534,
|
| 27 |
+
0.6162352153026697,
|
| 28 |
+
0.6095350569667238
|
| 29 |
],
|
| 30 |
"ppl": [
|
| 31 |
40729.60576844635,
|
| 32 |
2535.223731428122,
|
| 33 |
337.67548432577263,
|
| 34 |
+
123.54939191255174,
|
| 35 |
+
81.4022446898097
|
| 36 |
],
|
| 37 |
"tf_acc": [
|
| 38 |
9.249682110189831e-06,
|
| 39 |
0.21421828488041522,
|
| 40 |
0.39405658955515016,
|
| 41 |
+
0.4409562865464462,
|
| 42 |
+
0.4514476024551921
|
| 43 |
],
|
| 44 |
"fr_acc": [
|
| 45 |
0.0,
|
| 46 |
0.0,
|
| 47 |
0.11453107816089203,
|
| 48 |
+
0.09257571660973458,
|
| 49 |
+
0.1671656922337565
|
| 50 |
],
|
| 51 |
"def_rate": [
|
| 52 |
0.0,
|
| 53 |
0.0,
|
| 54 |
0.0,
|
| 55 |
+
0.45,
|
| 56 |
+
1.0
|
| 57 |
],
|
| 58 |
"mrr": [
|
| 59 |
0.004940100014209747,
|
| 60 |
0.005048312712460756,
|
| 61 |
0.004967095330357552,
|
| 62 |
+
0.004980043042451143,
|
| 63 |
+
0.004973841365426779
|
| 64 |
],
|
| 65 |
"r1": [
|
| 66 |
0.0006249999860301614,
|
| 67 |
0.0006249999860301614,
|
| 68 |
0.0006249999860301614,
|
| 69 |
+
0.0006249999860301614,
|
| 70 |
0.0006249999860301614
|
| 71 |
],
|
| 72 |
"ret_ok_pct": [
|
| 73 |
0.0,
|
| 74 |
0.0,
|
| 75 |
0.0,
|
| 76 |
+
0.0,
|
| 77 |
0.0
|
| 78 |
],
|
| 79 |
"grad_norm": [
|
| 80 |
2201.0254727094825,
|
| 81 |
1418.5299411384265,
|
| 82 |
1329.5837538555174,
|
| 83 |
+
1217.5837738567411,
|
| 84 |
+
1152.3023670419057
|
| 85 |
],
|
| 86 |
"step_var": [
|
| 87 |
0.22542811930179596,
|
| 88 |
0.22660225629806519,
|
| 89 |
0.2280091941356659,
|
| 90 |
+
0.22952058911323547,
|
| 91 |
+
0.23098886013031006
|
| 92 |
],
|
| 93 |
"pos_slope": [
|
| 94 |
0.49990350008010864,
|
| 95 |
0.4996356666088104,
|
| 96 |
0.49926072359085083,
|
| 97 |
+
0.4988322854042053,
|
| 98 |
+
0.4983888268470764
|
| 99 |
],
|
| 100 |
"beta_mean": [
|
| 101 |
0.09998118132352829,
|
| 102 |
0.09992741793394089,
|
| 103 |
0.09985271841287613,
|
| 104 |
+
0.09976637363433838,
|
| 105 |
+
0.09967698156833649
|
| 106 |
],
|
| 107 |
"w_ret": [
|
| 108 |
0.0,
|
| 109 |
0.0,
|
| 110 |
0.0,
|
| 111 |
+
0.0,
|
| 112 |
+
0.03
|
| 113 |
]
|
| 114 |
}
|