Upload ppo_monitor_history.json with huggingface_hub
Browse files- ppo_monitor_history.json +462 -0
ppo_monitor_history.json
ADDED
|
@@ -0,0 +1,462 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"epoch": 1,
|
| 4 |
+
"avg_ce": 1.298728307088216,
|
| 5 |
+
"avg_ppo": 0.5474145638008375,
|
| 6 |
+
"avg_reward": 0.9100000063578287,
|
| 7 |
+
"details": [
|
| 8 |
+
{
|
| 9 |
+
"ce": 1.109570860862732,
|
| 10 |
+
"ppo_score": 0.288488413242618,
|
| 11 |
+
"reward": 0.7599999904632568
|
| 12 |
+
},
|
| 13 |
+
{
|
| 14 |
+
"ce": 1.4682210683822632,
|
| 15 |
+
"ppo_score": 0.7341105341911316,
|
| 16 |
+
"reward": 1.0
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"ce": 1.3183929920196533,
|
| 20 |
+
"ppo_score": 0.6196447439687631,
|
| 21 |
+
"reward": 0.9700000286102295
|
| 22 |
+
}
|
| 23 |
+
]
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 2,
|
| 27 |
+
"avg_ce": 1.2838817834854126,
|
| 28 |
+
"avg_ppo": 0.5413803499665638,
|
| 29 |
+
"avg_reward": 0.9100000063578287,
|
| 30 |
+
"details": [
|
| 31 |
+
{
|
| 32 |
+
"ce": 1.094568133354187,
|
| 33 |
+
"ppo_score": 0.28458770423347346,
|
| 34 |
+
"reward": 0.7599999904632568
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"ce": 1.4575672149658203,
|
| 38 |
+
"ppo_score": 0.7287836074829102,
|
| 39 |
+
"reward": 1.0
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"ce": 1.2995100021362305,
|
| 43 |
+
"ppo_score": 0.6107697381833077,
|
| 44 |
+
"reward": 0.9700000286102295
|
| 45 |
+
}
|
| 46 |
+
]
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 3,
|
| 50 |
+
"avg_ce": 1.2241702477137248,
|
| 51 |
+
"avg_ppo": 0.5171198306996511,
|
| 52 |
+
"avg_reward": 0.9100000063578287,
|
| 53 |
+
"details": [
|
| 54 |
+
{
|
| 55 |
+
"ce": 1.0314350128173828,
|
| 56 |
+
"ppo_score": 0.2681730934959887,
|
| 57 |
+
"reward": 0.7599999904632568
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"ce": 1.3960256576538086,
|
| 61 |
+
"ppo_score": 0.6980128288269043,
|
| 62 |
+
"reward": 1.0
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"ce": 1.245050072669983,
|
| 66 |
+
"ppo_score": 0.5851735697760603,
|
| 67 |
+
"reward": 0.9700000286102295
|
| 68 |
+
}
|
| 69 |
+
]
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 4,
|
| 73 |
+
"avg_ce": 1.154691755771637,
|
| 74 |
+
"avg_ppo": 0.487802223386718,
|
| 75 |
+
"avg_reward": 0.9100000063578287,
|
| 76 |
+
"details": [
|
| 77 |
+
{
|
| 78 |
+
"ce": 0.9742447733879089,
|
| 79 |
+
"ppo_score": 0.25330363178973414,
|
| 80 |
+
"reward": 0.7599999904632568
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"ce": 1.3294204473495483,
|
| 84 |
+
"ppo_score": 0.664710283279419,
|
| 85 |
+
"reward": 1.0
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"ce": 1.1604100465774536,
|
| 89 |
+
"ppo_score": 0.5453927550910009,
|
| 90 |
+
"reward": 0.9700000286102295
|
| 91 |
+
}
|
| 92 |
+
]
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 5,
|
| 96 |
+
"avg_ce": 1.0816803177197774,
|
| 97 |
+
"avg_ppo": 0.4576717686419964,
|
| 98 |
+
"avg_reward": 0.9100000063578287,
|
| 99 |
+
"details": [
|
| 100 |
+
{
|
| 101 |
+
"ce": 0.9083136916160583,
|
| 102 |
+
"ppo_score": 0.23616155115782078,
|
| 103 |
+
"reward": 0.7599999904632568
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"ce": 1.2864009141921997,
|
| 107 |
+
"ppo_score": 0.6432003974914551,
|
| 108 |
+
"reward": 1.0
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"ce": 1.0503263473510742,
|
| 112 |
+
"ppo_score": 0.4936533572767132,
|
| 113 |
+
"reward": 0.9700000286102295
|
| 114 |
+
}
|
| 115 |
+
]
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 6,
|
| 119 |
+
"avg_ce": 1.0213011503219604,
|
| 120 |
+
"avg_ppo": 0.43329118010116474,
|
| 121 |
+
"avg_reward": 0.9100000063578287,
|
| 122 |
+
"details": [
|
| 123 |
+
{
|
| 124 |
+
"ce": 0.8462240099906921,
|
| 125 |
+
"ppo_score": 0.220018250024566,
|
| 126 |
+
"reward": 0.7599999904632568
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"ce": 1.2515289783477783,
|
| 130 |
+
"ppo_score": 0.6257644891738892,
|
| 131 |
+
"reward": 1.0
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"ce": 0.9661504626274109,
|
| 135 |
+
"ppo_score": 0.4540908011050391,
|
| 136 |
+
"reward": 0.9700000286102295
|
| 137 |
+
}
|
| 138 |
+
]
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 7,
|
| 142 |
+
"avg_ce": 0.9735608696937561,
|
| 143 |
+
"avg_ppo": 0.41395096229899764,
|
| 144 |
+
"avg_reward": 0.9100000063578287,
|
| 145 |
+
"details": [
|
| 146 |
+
{
|
| 147 |
+
"ce": 0.7980301976203918,
|
| 148 |
+
"ppo_score": 0.20748785926789992,
|
| 149 |
+
"reward": 0.7599999904632568
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"ce": 1.2239456176757812,
|
| 153 |
+
"ppo_score": 0.6119728088378906,
|
| 154 |
+
"reward": 1.0
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"ce": 0.8987067937850952,
|
| 158 |
+
"ppo_score": 0.42239221879120237,
|
| 159 |
+
"reward": 0.9700000286102295
|
| 160 |
+
}
|
| 161 |
+
]
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 8,
|
| 165 |
+
"avg_ce": 0.9334022601445516,
|
| 166 |
+
"avg_ppo": 0.3975095751488548,
|
| 167 |
+
"avg_reward": 0.9100000063578287,
|
| 168 |
+
"details": [
|
| 169 |
+
{
|
| 170 |
+
"ce": 0.7595431208610535,
|
| 171 |
+
"ppo_score": 0.1974812196775133,
|
| 172 |
+
"reward": 0.7599999904632568
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"ce": 1.1978540420532227,
|
| 176 |
+
"ppo_score": 0.5989269614219666,
|
| 177 |
+
"reward": 1.0
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"ce": 0.8428096175193787,
|
| 181 |
+
"ppo_score": 0.39612054434708455,
|
| 182 |
+
"reward": 0.9700000286102295
|
| 183 |
+
}
|
| 184 |
+
]
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 9,
|
| 188 |
+
"avg_ce": 0.9001490275065104,
|
| 189 |
+
"avg_ppo": 0.38373239634663037,
|
| 190 |
+
"avg_reward": 0.9100000063578287,
|
| 191 |
+
"details": [
|
| 192 |
+
{
|
| 193 |
+
"ce": 0.7288264632225037,
|
| 194 |
+
"ppo_score": 0.18949487348722016,
|
| 195 |
+
"reward": 0.7599999904632568
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"ce": 1.1680219173431396,
|
| 199 |
+
"ppo_score": 0.5840109586715698,
|
| 200 |
+
"reward": 1.0
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"ce": 0.8035987019538879,
|
| 204 |
+
"ppo_score": 0.3776913568811011,
|
| 205 |
+
"reward": 0.9700000286102295
|
| 206 |
+
}
|
| 207 |
+
]
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 10,
|
| 211 |
+
"avg_ce": 0.877958079179128,
|
| 212 |
+
"avg_ppo": 0.37494345543318275,
|
| 213 |
+
"avg_reward": 0.9100000063578287,
|
| 214 |
+
"details": [
|
| 215 |
+
{
|
| 216 |
+
"ce": 0.7012356519699097,
|
| 217 |
+
"ppo_score": 0.1823212628246722,
|
| 218 |
+
"reward": 0.7599999904632568
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"ce": 1.1389628648757935,
|
| 222 |
+
"ppo_score": 0.5694814920425415,
|
| 223 |
+
"reward": 1.0
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"ce": 0.7936757206916809,
|
| 227 |
+
"ppo_score": 0.37302761143233454,
|
| 228 |
+
"reward": 0.9700000286102295
|
| 229 |
+
}
|
| 230 |
+
]
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 11,
|
| 234 |
+
"avg_ce": 0.8554977973302206,
|
| 235 |
+
"avg_ppo": 0.36565266708344285,
|
| 236 |
+
"avg_reward": 0.9100000063578287,
|
| 237 |
+
"details": [
|
| 238 |
+
{
|
| 239 |
+
"ce": 0.6801788210868835,
|
| 240 |
+
"ppo_score": 0.176846486995899,
|
| 241 |
+
"reward": 0.7599999904632568
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"ce": 1.118119478225708,
|
| 245 |
+
"ppo_score": 0.5590597987174988,
|
| 246 |
+
"reward": 1.0
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"ce": 0.7681950926780701,
|
| 250 |
+
"ppo_score": 0.36105171553693083,
|
| 251 |
+
"reward": 0.9700000286102295
|
| 252 |
+
}
|
| 253 |
+
]
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 12,
|
| 257 |
+
"avg_ce": 0.8307044704755148,
|
| 258 |
+
"avg_ppo": 0.35559931417884627,
|
| 259 |
+
"avg_reward": 0.9100000063578287,
|
| 260 |
+
"details": [
|
| 261 |
+
{
|
| 262 |
+
"ce": 0.6544622182846069,
|
| 263 |
+
"ppo_score": 0.17016017051255972,
|
| 264 |
+
"reward": 0.7599999904632568
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"ce": 1.0980572700500488,
|
| 268 |
+
"ppo_score": 0.5490286350250244,
|
| 269 |
+
"reward": 1.0
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"ce": 0.7395939230918884,
|
| 273 |
+
"ppo_score": 0.3476091369989547,
|
| 274 |
+
"reward": 0.9700000286102295
|
| 275 |
+
}
|
| 276 |
+
]
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 13,
|
| 280 |
+
"avg_ce": 0.815015951792399,
|
| 281 |
+
"avg_ppo": 0.3490992292643682,
|
| 282 |
+
"avg_reward": 0.9100000063578287,
|
| 283 |
+
"details": [
|
| 284 |
+
{
|
| 285 |
+
"ce": 0.6398036479949951,
|
| 286 |
+
"ppo_score": 0.16634894237705566,
|
| 287 |
+
"reward": 0.7599999904632568
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"ce": 1.0827982425689697,
|
| 291 |
+
"ppo_score": 0.5413991212844849,
|
| 292 |
+
"reward": 1.0
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"ce": 0.7224459648132324,
|
| 296 |
+
"ppo_score": 0.3395496241315641,
|
| 297 |
+
"reward": 0.9700000286102295
|
| 298 |
+
}
|
| 299 |
+
]
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 14,
|
| 303 |
+
"avg_ce": 0.8029795090357462,
|
| 304 |
+
"avg_ppo": 0.34417347507580587,
|
| 305 |
+
"avg_reward": 0.9100000063578287,
|
| 306 |
+
"details": [
|
| 307 |
+
{
|
| 308 |
+
"ce": 0.6281819939613342,
|
| 309 |
+
"ppo_score": 0.16332731243913656,
|
| 310 |
+
"reward": 0.7599999904632568
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"ce": 1.0745869874954224,
|
| 314 |
+
"ppo_score": 0.5372934341430664,
|
| 315 |
+
"reward": 1.0
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"ce": 0.7061695456504822,
|
| 319 |
+
"ppo_score": 0.33189967864521464,
|
| 320 |
+
"reward": 0.9700000286102295
|
| 321 |
+
}
|
| 322 |
+
]
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 15,
|
| 326 |
+
"avg_ce": 0.7931084831555685,
|
| 327 |
+
"avg_ppo": 0.34020966878169173,
|
| 328 |
+
"avg_reward": 0.9100000063578287,
|
| 329 |
+
"details": [
|
| 330 |
+
{
|
| 331 |
+
"ce": 0.6173056960105896,
|
| 332 |
+
"ppo_score": 0.16049945957846035,
|
| 333 |
+
"reward": 0.7599999904632568
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"ce": 1.0660080909729004,
|
| 337 |
+
"ppo_score": 0.5330040454864502,
|
| 338 |
+
"reward": 1.0
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"ce": 0.6960116624832153,
|
| 342 |
+
"ppo_score": 0.3271255012801646,
|
| 343 |
+
"reward": 0.9700000286102295
|
| 344 |
+
}
|
| 345 |
+
]
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 16,
|
| 349 |
+
"avg_ce": 0.7849790453910828,
|
| 350 |
+
"avg_ppo": 0.3368807465024162,
|
| 351 |
+
"avg_reward": 0.9100000063578287,
|
| 352 |
+
"details": [
|
| 353 |
+
{
|
| 354 |
+
"ce": 0.6095706820487976,
|
| 355 |
+
"ppo_score": 0.15848837151936834,
|
| 356 |
+
"reward": 0.7599999904632568
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"ce": 1.0610578060150146,
|
| 360 |
+
"ppo_score": 0.5305287837982178,
|
| 361 |
+
"reward": 1.0
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"ce": 0.684308648109436,
|
| 365 |
+
"ppo_score": 0.3216250841896624,
|
| 366 |
+
"reward": 0.9700000286102295
|
| 367 |
+
}
|
| 368 |
+
]
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 17,
|
| 372 |
+
"avg_ce": 0.7803110082944235,
|
| 373 |
+
"avg_ppo": 0.33504825729727844,
|
| 374 |
+
"avg_reward": 0.9100000063578287,
|
| 375 |
+
"details": [
|
| 376 |
+
{
|
| 377 |
+
"ce": 0.6039620637893677,
|
| 378 |
+
"ppo_score": 0.15703011532819744,
|
| 379 |
+
"reward": 0.7599999904632568
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"ce": 1.0579419136047363,
|
| 383 |
+
"ppo_score": 0.5289709568023682,
|
| 384 |
+
"reward": 1.0
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"ce": 0.6790290474891663,
|
| 388 |
+
"ppo_score": 0.3191436997612698,
|
| 389 |
+
"reward": 0.9700000286102295
|
| 390 |
+
}
|
| 391 |
+
]
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 18,
|
| 395 |
+
"avg_ce": 0.7780356407165527,
|
| 396 |
+
"avg_ppo": 0.33396928939788023,
|
| 397 |
+
"avg_reward": 0.9100000063578287,
|
| 398 |
+
"details": [
|
| 399 |
+
{
|
| 400 |
+
"ce": 0.603886067867279,
|
| 401 |
+
"ppo_score": 0.15701035638917915,
|
| 402 |
+
"reward": 0.7599999904632568
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"ce": 1.0564583539962769,
|
| 406 |
+
"ppo_score": 0.5282291173934937,
|
| 407 |
+
"reward": 1.0
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"ce": 0.6737625002861023,
|
| 411 |
+
"ppo_score": 0.31666839441096784,
|
| 412 |
+
"reward": 0.9700000286102295
|
| 413 |
+
}
|
| 414 |
+
]
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 19,
|
| 418 |
+
"avg_ce": 0.7770752906799316,
|
| 419 |
+
"avg_ppo": 0.33356191350819603,
|
| 420 |
+
"avg_reward": 0.9100000063578287,
|
| 421 |
+
"details": [
|
| 422 |
+
{
|
| 423 |
+
"ce": 0.6029837131500244,
|
| 424 |
+
"ppo_score": 0.15677575966850554,
|
| 425 |
+
"reward": 0.7599999904632568
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"ce": 1.054537296295166,
|
| 429 |
+
"ppo_score": 0.527268648147583,
|
| 430 |
+
"reward": 1.0
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"ce": 0.6737048625946045,
|
| 434 |
+
"ppo_score": 0.3166413327084996,
|
| 435 |
+
"reward": 0.9700000286102295
|
| 436 |
+
}
|
| 437 |
+
]
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 20,
|
| 441 |
+
"avg_ce": 0.7763257821400961,
|
| 442 |
+
"avg_ppo": 0.33327399006772396,
|
| 443 |
+
"avg_reward": 0.9100000063578287,
|
| 444 |
+
"details": [
|
| 445 |
+
{
|
| 446 |
+
"ce": 0.6022559404373169,
|
| 447 |
+
"ppo_score": 0.15658653877014217,
|
| 448 |
+
"reward": 0.7599999904632568
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"ce": 1.055877447128296,
|
| 452 |
+
"ppo_score": 0.527938723564148,
|
| 453 |
+
"reward": 1.0
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"ce": 0.6708439588546753,
|
| 457 |
+
"ppo_score": 0.31529670786888175,
|
| 458 |
+
"reward": 0.9700000286102295
|
| 459 |
+
}
|
| 460 |
+
]
|
| 461 |
+
}
|
| 462 |
+
]
|