Upload sigma_lambda_v5/history.json with huggingface_hub
Browse files- sigma_lambda_v5/history.json +20 -8
sigma_lambda_v5/history.json
CHANGED
|
@@ -11,7 +11,8 @@
|
|
| 11 |
5,
|
| 12 |
6,
|
| 13 |
7,
|
| 14 |
-
8
|
|
|
|
| 15 |
],
|
| 16 |
"gen_loss": [
|
| 17 |
16.586591267879488,
|
|
@@ -25,7 +26,8 @@
|
|
| 25 |
74.47382958961273,
|
| 26 |
116.53298210471324,
|
| 27 |
170.46740706801785,
|
| 28 |
-
214.5222631372081
|
|
|
|
| 29 |
],
|
| 30 |
"ret_loss": [
|
| 31 |
0.0,
|
|
@@ -39,6 +41,7 @@
|
|
| 39 |
0.0,
|
| 40 |
0.0,
|
| 41 |
0.0,
|
|
|
|
| 42 |
0.0
|
| 43 |
],
|
| 44 |
"ppl": [
|
|
@@ -53,6 +56,7 @@
|
|
| 53 |
1624266.5983328498,
|
| 54 |
485165195.4097903,
|
| 55 |
485165195.4097903,
|
|
|
|
| 56 |
485165195.4097903
|
| 57 |
],
|
| 58 |
"tf_acc": [
|
|
@@ -67,7 +71,8 @@
|
|
| 67 |
0.4732017211019178,
|
| 68 |
0.46131388743580864,
|
| 69 |
0.45428323030789747,
|
| 70 |
-
0.44858981632247086
|
|
|
|
| 71 |
],
|
| 72 |
"fr_acc": [
|
| 73 |
0.12302286153389094,
|
|
@@ -81,7 +86,8 @@
|
|
| 81 |
0.24578219982631747,
|
| 82 |
0.27369709538827186,
|
| 83 |
0.12652472119383884,
|
| 84 |
-
0.14813148452854336
|
|
|
|
| 85 |
],
|
| 86 |
"mrr": [
|
| 87 |
0.006059600971639156,
|
|
@@ -95,7 +101,8 @@
|
|
| 95 |
0.006042419467121363,
|
| 96 |
0.006052091252058744,
|
| 97 |
0.006054983474314213,
|
| 98 |
-
0.006042384542524815
|
|
|
|
| 99 |
],
|
| 100 |
"r1": [
|
| 101 |
0.0007812500116415322,
|
|
@@ -109,6 +116,7 @@
|
|
| 109 |
0.0007812500116415322,
|
| 110 |
0.0007812500116415322,
|
| 111 |
0.0007812500116415322,
|
|
|
|
| 112 |
0.0007812500116415322
|
| 113 |
],
|
| 114 |
"step_var": [
|
|
@@ -123,7 +131,8 @@
|
|
| 123 |
0.22339501976966858,
|
| 124 |
0.24468941986560822,
|
| 125 |
0.26731032133102417,
|
| 126 |
-
0.28763458132743835
|
|
|
|
| 127 |
],
|
| 128 |
"pos_slope": [
|
| 129 |
0.4985616207122803,
|
|
@@ -137,7 +146,8 @@
|
|
| 137 |
0.48436233401298523,
|
| 138 |
0.4812909960746765,
|
| 139 |
0.4783017337322235,
|
| 140 |
-
0.47553473711013794
|
|
|
|
| 141 |
],
|
| 142 |
"grad_norm": [
|
| 143 |
NaN,
|
|
@@ -151,6 +161,7 @@
|
|
| 151 |
NaN,
|
| 152 |
NaN,
|
| 153 |
NaN,
|
|
|
|
| 154 |
Infinity
|
| 155 |
],
|
| 156 |
"w_ret": [
|
|
@@ -165,6 +176,7 @@
|
|
| 165 |
0.02,
|
| 166 |
0.03,
|
| 167 |
0.04,
|
| 168 |
-
0.05
|
|
|
|
| 169 |
]
|
| 170 |
}
|
|
|
|
| 11 |
5,
|
| 12 |
6,
|
| 13 |
7,
|
| 14 |
+
8,
|
| 15 |
+
9
|
| 16 |
],
|
| 17 |
"gen_loss": [
|
| 18 |
16.586591267879488,
|
|
|
|
| 26 |
74.47382958961273,
|
| 27 |
116.53298210471324,
|
| 28 |
170.46740706801785,
|
| 29 |
+
214.5222631372081,
|
| 30 |
+
274.12421057906226
|
| 31 |
],
|
| 32 |
"ret_loss": [
|
| 33 |
0.0,
|
|
|
|
| 41 |
0.0,
|
| 42 |
0.0,
|
| 43 |
0.0,
|
| 44 |
+
0.0,
|
| 45 |
0.0
|
| 46 |
],
|
| 47 |
"ppl": [
|
|
|
|
| 56 |
1624266.5983328498,
|
| 57 |
485165195.4097903,
|
| 58 |
485165195.4097903,
|
| 59 |
+
485165195.4097903,
|
| 60 |
485165195.4097903
|
| 61 |
],
|
| 62 |
"tf_acc": [
|
|
|
|
| 71 |
0.4732017211019178,
|
| 72 |
0.46131388743580864,
|
| 73 |
0.45428323030789747,
|
| 74 |
+
0.44858981632247086,
|
| 75 |
+
0.44045584150537476
|
| 76 |
],
|
| 77 |
"fr_acc": [
|
| 78 |
0.12302286153389094,
|
|
|
|
| 86 |
0.24578219982631747,
|
| 87 |
0.27369709538827186,
|
| 88 |
0.12652472119383884,
|
| 89 |
+
0.14813148452854336,
|
| 90 |
+
0.1719167780748663
|
| 91 |
],
|
| 92 |
"mrr": [
|
| 93 |
0.006059600971639156,
|
|
|
|
| 101 |
0.006042419467121363,
|
| 102 |
0.006052091252058744,
|
| 103 |
0.006054983474314213,
|
| 104 |
+
0.006042384542524815,
|
| 105 |
+
0.006041090935468674
|
| 106 |
],
|
| 107 |
"r1": [
|
| 108 |
0.0007812500116415322,
|
|
|
|
| 116 |
0.0007812500116415322,
|
| 117 |
0.0007812500116415322,
|
| 118 |
0.0007812500116415322,
|
| 119 |
+
0.0007812500116415322,
|
| 120 |
0.0007812500116415322
|
| 121 |
],
|
| 122 |
"step_var": [
|
|
|
|
| 131 |
0.22339501976966858,
|
| 132 |
0.24468941986560822,
|
| 133 |
0.26731032133102417,
|
| 134 |
+
0.28763458132743835,
|
| 135 |
+
0.3055805563926697
|
| 136 |
],
|
| 137 |
"pos_slope": [
|
| 138 |
0.4985616207122803,
|
|
|
|
| 146 |
0.48436233401298523,
|
| 147 |
0.4812909960746765,
|
| 148 |
0.4783017337322235,
|
| 149 |
+
0.47553473711013794,
|
| 150 |
+
0.47306424379348755
|
| 151 |
],
|
| 152 |
"grad_norm": [
|
| 153 |
NaN,
|
|
|
|
| 161 |
NaN,
|
| 162 |
NaN,
|
| 163 |
NaN,
|
| 164 |
+
Infinity,
|
| 165 |
Infinity
|
| 166 |
],
|
| 167 |
"w_ret": [
|
|
|
|
| 176 |
0.02,
|
| 177 |
0.03,
|
| 178 |
0.04,
|
| 179 |
+
0.05,
|
| 180 |
+
0.08
|
| 181 |
]
|
| 182 |
}
|