Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +709 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,709 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 462,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.004329004329004329,
|
| 13 |
+
"grad_norm": 20.877743796637407,
|
| 14 |
+
"learning_rate": 4.2553191489361704e-07,
|
| 15 |
+
"loss": 1.7871,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.021645021645021644,
|
| 20 |
+
"grad_norm": 18.21795285874322,
|
| 21 |
+
"learning_rate": 2.1276595744680853e-06,
|
| 22 |
+
"loss": 1.7389,
|
| 23 |
+
"step": 5
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.04329004329004329,
|
| 27 |
+
"grad_norm": 5.701059989509366,
|
| 28 |
+
"learning_rate": 4.255319148936171e-06,
|
| 29 |
+
"loss": 1.5421,
|
| 30 |
+
"step": 10
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.06493506493506493,
|
| 34 |
+
"grad_norm": 2.789667646224892,
|
| 35 |
+
"learning_rate": 6.382978723404256e-06,
|
| 36 |
+
"loss": 1.3975,
|
| 37 |
+
"step": 15
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.08658008658008658,
|
| 41 |
+
"grad_norm": 1.613528044644943,
|
| 42 |
+
"learning_rate": 8.510638297872341e-06,
|
| 43 |
+
"loss": 1.3146,
|
| 44 |
+
"step": 20
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.10822510822510822,
|
| 48 |
+
"grad_norm": 1.222589469421322,
|
| 49 |
+
"learning_rate": 1.0638297872340426e-05,
|
| 50 |
+
"loss": 1.2876,
|
| 51 |
+
"step": 25
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.12987012987012986,
|
| 55 |
+
"grad_norm": 1.3215528280582176,
|
| 56 |
+
"learning_rate": 1.2765957446808513e-05,
|
| 57 |
+
"loss": 1.2531,
|
| 58 |
+
"step": 30
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.15151515151515152,
|
| 62 |
+
"grad_norm": 1.3282799697001735,
|
| 63 |
+
"learning_rate": 1.4893617021276596e-05,
|
| 64 |
+
"loss": 1.2096,
|
| 65 |
+
"step": 35
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.17316017316017315,
|
| 69 |
+
"grad_norm": 1.1149401355405273,
|
| 70 |
+
"learning_rate": 1.7021276595744682e-05,
|
| 71 |
+
"loss": 1.2256,
|
| 72 |
+
"step": 40
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.19480519480519481,
|
| 76 |
+
"grad_norm": 1.4141147361835624,
|
| 77 |
+
"learning_rate": 1.914893617021277e-05,
|
| 78 |
+
"loss": 1.1834,
|
| 79 |
+
"step": 45
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.21645021645021645,
|
| 83 |
+
"grad_norm": 1.1822192980158253,
|
| 84 |
+
"learning_rate": 1.9997421320309795e-05,
|
| 85 |
+
"loss": 1.2118,
|
| 86 |
+
"step": 50
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.23809523809523808,
|
| 90 |
+
"grad_norm": 1.1446997824008667,
|
| 91 |
+
"learning_rate": 1.9981667538108588e-05,
|
| 92 |
+
"loss": 1.201,
|
| 93 |
+
"step": 55
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.2597402597402597,
|
| 97 |
+
"grad_norm": 1.1200026579741698,
|
| 98 |
+
"learning_rate": 1.9951615113309076e-05,
|
| 99 |
+
"loss": 1.1952,
|
| 100 |
+
"step": 60
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.2813852813852814,
|
| 104 |
+
"grad_norm": 1.187192621847611,
|
| 105 |
+
"learning_rate": 1.9907307095720304e-05,
|
| 106 |
+
"loss": 1.2281,
|
| 107 |
+
"step": 65
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.30303030303030304,
|
| 111 |
+
"grad_norm": 1.1254345317613856,
|
| 112 |
+
"learning_rate": 1.984880695615066e-05,
|
| 113 |
+
"loss": 1.1929,
|
| 114 |
+
"step": 70
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.3246753246753247,
|
| 118 |
+
"grad_norm": 2.053142496804073,
|
| 119 |
+
"learning_rate": 1.9776198495486566e-05,
|
| 120 |
+
"loss": 1.2203,
|
| 121 |
+
"step": 75
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.3463203463203463,
|
| 125 |
+
"grad_norm": 1.2029392175636573,
|
| 126 |
+
"learning_rate": 1.9689585724648515e-05,
|
| 127 |
+
"loss": 1.2179,
|
| 128 |
+
"step": 80
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.36796536796536794,
|
| 132 |
+
"grad_norm": 1.1250217257278072,
|
| 133 |
+
"learning_rate": 1.958909271559642e-05,
|
| 134 |
+
"loss": 1.1996,
|
| 135 |
+
"step": 85
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.38961038961038963,
|
| 139 |
+
"grad_norm": 1.1037575465342937,
|
| 140 |
+
"learning_rate": 1.947486342359773e-05,
|
| 141 |
+
"loss": 1.1843,
|
| 142 |
+
"step": 90
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.41125541125541126,
|
| 146 |
+
"grad_norm": 1.6143939161369725,
|
| 147 |
+
"learning_rate": 1.9347061481012894e-05,
|
| 148 |
+
"loss": 1.1646,
|
| 149 |
+
"step": 95
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.4329004329004329,
|
| 153 |
+
"grad_norm": 1.2124944707128797,
|
| 154 |
+
"learning_rate": 1.9205869962893607e-05,
|
| 155 |
+
"loss": 1.2082,
|
| 156 |
+
"step": 100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.45454545454545453,
|
| 160 |
+
"grad_norm": 1.366998499730903,
|
| 161 |
+
"learning_rate": 1.9051491124729512e-05,
|
| 162 |
+
"loss": 1.1927,
|
| 163 |
+
"step": 105
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.47619047619047616,
|
| 167 |
+
"grad_norm": 1.120158585500849,
|
| 168 |
+
"learning_rate": 1.8884146112719208e-05,
|
| 169 |
+
"loss": 1.1637,
|
| 170 |
+
"step": 110
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.49783549783549785,
|
| 174 |
+
"grad_norm": 1.169190437583552,
|
| 175 |
+
"learning_rate": 1.8704074646980416e-05,
|
| 176 |
+
"loss": 1.1739,
|
| 177 |
+
"step": 115
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.5194805194805194,
|
| 181 |
+
"grad_norm": 1.5328426889406774,
|
| 182 |
+
"learning_rate": 1.8511534678153243e-05,
|
| 183 |
+
"loss": 1.1488,
|
| 184 |
+
"step": 120
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.5411255411255411,
|
| 188 |
+
"grad_norm": 1.0366421774380163,
|
| 189 |
+
"learning_rate": 1.830680201788836e-05,
|
| 190 |
+
"loss": 1.1756,
|
| 191 |
+
"step": 125
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.5627705627705628,
|
| 195 |
+
"grad_norm": 1.1075356705624422,
|
| 196 |
+
"learning_rate": 1.8090169943749477e-05,
|
| 197 |
+
"loss": 1.1905,
|
| 198 |
+
"step": 130
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.5844155844155844,
|
| 202 |
+
"grad_norm": 1.0847986023260712,
|
| 203 |
+
"learning_rate": 1.7861948779096047e-05,
|
| 204 |
+
"loss": 1.1665,
|
| 205 |
+
"step": 135
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.6060606060606061,
|
| 209 |
+
"grad_norm": 1.0102263118741437,
|
| 210 |
+
"learning_rate": 1.762246544854807e-05,
|
| 211 |
+
"loss": 1.137,
|
| 212 |
+
"step": 140
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.6277056277056277,
|
| 216 |
+
"grad_norm": 1.036983432238017,
|
| 217 |
+
"learning_rate": 1.7372063009669685e-05,
|
| 218 |
+
"loss": 1.1433,
|
| 219 |
+
"step": 145
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.6493506493506493,
|
| 223 |
+
"grad_norm": 1.1129097876771734,
|
| 224 |
+
"learning_rate": 1.7111100161542545e-05,
|
| 225 |
+
"loss": 1.1756,
|
| 226 |
+
"step": 150
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.670995670995671,
|
| 230 |
+
"grad_norm": 1.048990729813769,
|
| 231 |
+
"learning_rate": 1.683995073093283e-05,
|
| 232 |
+
"loss": 1.1225,
|
| 233 |
+
"step": 155
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.6926406926406926,
|
| 237 |
+
"grad_norm": 1.1501277496328752,
|
| 238 |
+
"learning_rate": 1.655900313678799e-05,
|
| 239 |
+
"loss": 1.1335,
|
| 240 |
+
"step": 160
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.7142857142857143,
|
| 244 |
+
"grad_norm": 1.1230467049226944,
|
| 245 |
+
"learning_rate": 1.626865983383037e-05,
|
| 246 |
+
"loss": 1.1359,
|
| 247 |
+
"step": 165
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.7359307359307359,
|
| 251 |
+
"grad_norm": 1.0339043537639934,
|
| 252 |
+
"learning_rate": 1.596933673604467e-05,
|
| 253 |
+
"loss": 1.1525,
|
| 254 |
+
"step": 170
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.7575757575757576,
|
| 258 |
+
"grad_norm": 1.3703136108639966,
|
| 259 |
+
"learning_rate": 1.5661462620885198e-05,
|
| 260 |
+
"loss": 1.1084,
|
| 261 |
+
"step": 175
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.7792207792207793,
|
| 265 |
+
"grad_norm": 1.1235783405767077,
|
| 266 |
+
"learning_rate": 1.534547851505627e-05,
|
| 267 |
+
"loss": 1.1454,
|
| 268 |
+
"step": 180
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.8008658008658008,
|
| 272 |
+
"grad_norm": 1.0145691819619813,
|
| 273 |
+
"learning_rate": 1.5021837062745714e-05,
|
| 274 |
+
"loss": 1.1314,
|
| 275 |
+
"step": 185
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.8225108225108225,
|
| 279 |
+
"grad_norm": 1.0366379821859082,
|
| 280 |
+
"learning_rate": 1.469100187721644e-05,
|
| 281 |
+
"loss": 1.1347,
|
| 282 |
+
"step": 190
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.8441558441558441,
|
| 286 |
+
"grad_norm": 0.9797343756954148,
|
| 287 |
+
"learning_rate": 1.435344687668489e-05,
|
| 288 |
+
"loss": 1.1267,
|
| 289 |
+
"step": 195
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.8658008658008658,
|
| 293 |
+
"grad_norm": 1.0125169520271824,
|
| 294 |
+
"learning_rate": 1.400965560543778e-05,
|
| 295 |
+
"loss": 1.1432,
|
| 296 |
+
"step": 200
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.8874458874458875,
|
| 300 |
+
"grad_norm": 1.0076883560029861,
|
| 301 |
+
"learning_rate": 1.3660120541159536e-05,
|
| 302 |
+
"loss": 1.1453,
|
| 303 |
+
"step": 205
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.9090909090909091,
|
| 307 |
+
"grad_norm": 0.9594741075750862,
|
| 308 |
+
"learning_rate": 1.3305342389462792e-05,
|
| 309 |
+
"loss": 1.1235,
|
| 310 |
+
"step": 210
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.9307359307359307,
|
| 314 |
+
"grad_norm": 1.0087650659111476,
|
| 315 |
+
"learning_rate": 1.294582936663239e-05,
|
| 316 |
+
"loss": 1.1333,
|
| 317 |
+
"step": 215
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.9523809523809523,
|
| 321 |
+
"grad_norm": 1.021522349018145,
|
| 322 |
+
"learning_rate": 1.2582096471610467e-05,
|
| 323 |
+
"loss": 1.1104,
|
| 324 |
+
"step": 220
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.974025974025974,
|
| 328 |
+
"grad_norm": 1.0148162711090376,
|
| 329 |
+
"learning_rate": 1.221466474826543e-05,
|
| 330 |
+
"loss": 1.1077,
|
| 331 |
+
"step": 225
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.9956709956709957,
|
| 335 |
+
"grad_norm": 0.9665861752821213,
|
| 336 |
+
"learning_rate": 1.1844060539001618e-05,
|
| 337 |
+
"loss": 1.1016,
|
| 338 |
+
"step": 230
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.0,
|
| 342 |
+
"eval_loss": 1.1138663291931152,
|
| 343 |
+
"eval_runtime": 4.3692,
|
| 344 |
+
"eval_samples_per_second": 34.789,
|
| 345 |
+
"eval_steps_per_second": 0.687,
|
| 346 |
+
"step": 231
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 1.0173160173160174,
|
| 350 |
+
"grad_norm": 2.0527264159154432,
|
| 351 |
+
"learning_rate": 1.1470814730778905e-05,
|
| 352 |
+
"loss": 0.9102,
|
| 353 |
+
"step": 235
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 1.0389610389610389,
|
| 357 |
+
"grad_norm": 1.455965475523599,
|
| 358 |
+
"learning_rate": 1.109546199462221e-05,
|
| 359 |
+
"loss": 0.8067,
|
| 360 |
+
"step": 240
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 1.0606060606060606,
|
| 364 |
+
"grad_norm": 1.1731794708057288,
|
| 365 |
+
"learning_rate": 1.0718540019710433e-05,
|
| 366 |
+
"loss": 0.824,
|
| 367 |
+
"step": 245
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 1.0822510822510822,
|
| 371 |
+
"grad_norm": 1.1851242426639188,
|
| 372 |
+
"learning_rate": 1.0340588743141879e-05,
|
| 373 |
+
"loss": 0.8115,
|
| 374 |
+
"step": 250
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 1.103896103896104,
|
| 378 |
+
"grad_norm": 1.078251194842658,
|
| 379 |
+
"learning_rate": 9.962149576479546e-06,
|
| 380 |
+
"loss": 0.8114,
|
| 381 |
+
"step": 255
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.1255411255411256,
|
| 385 |
+
"grad_norm": 1.0244399166247982,
|
| 386 |
+
"learning_rate": 9.583764630184311e-06,
|
| 387 |
+
"loss": 0.8198,
|
| 388 |
+
"step": 260
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 1.1471861471861473,
|
| 392 |
+
"grad_norm": 1.0118125591636071,
|
| 393 |
+
"learning_rate": 9.205975937046879e-06,
|
| 394 |
+
"loss": 0.8043,
|
| 395 |
+
"step": 265
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.1688311688311688,
|
| 399 |
+
"grad_norm": 1.0315115837254647,
|
| 400 |
+
"learning_rate": 8.82932467573106e-06,
|
| 401 |
+
"loss": 0.7932,
|
| 402 |
+
"step": 270
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 1.1904761904761905,
|
| 406 |
+
"grad_norm": 1.0047518561511999,
|
| 407 |
+
"learning_rate": 8.45435039554054e-06,
|
| 408 |
+
"loss": 0.7865,
|
| 409 |
+
"step": 275
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.2121212121212122,
|
| 413 |
+
"grad_norm": 1.0324907986018956,
|
| 414 |
+
"learning_rate": 8.081590243519682e-06,
|
| 415 |
+
"loss": 0.7959,
|
| 416 |
+
"step": 280
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.2337662337662338,
|
| 420 |
+
"grad_norm": 0.9684538095728706,
|
| 421 |
+
"learning_rate": 7.71157819499559e-06,
|
| 422 |
+
"loss": 0.7692,
|
| 423 |
+
"step": 285
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.2554112554112553,
|
| 427 |
+
"grad_norm": 1.026533916056025,
|
| 428 |
+
"learning_rate": 7.344844288663571e-06,
|
| 429 |
+
"loss": 0.8126,
|
| 430 |
+
"step": 290
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.277056277056277,
|
| 434 |
+
"grad_norm": 0.9834883699233659,
|
| 435 |
+
"learning_rate": 6.981913867311818e-06,
|
| 436 |
+
"loss": 0.7872,
|
| 437 |
+
"step": 295
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.2987012987012987,
|
| 441 |
+
"grad_norm": 1.1658729678515243,
|
| 442 |
+
"learning_rate": 6.623306825272937e-06,
|
| 443 |
+
"loss": 0.8035,
|
| 444 |
+
"step": 300
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.3203463203463204,
|
| 448 |
+
"grad_norm": 0.9850249154792221,
|
| 449 |
+
"learning_rate": 6.2695368636803075e-06,
|
| 450 |
+
"loss": 0.8043,
|
| 451 |
+
"step": 305
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.341991341991342,
|
| 455 |
+
"grad_norm": 0.9651304936317318,
|
| 456 |
+
"learning_rate": 5.92111075459616e-06,
|
| 457 |
+
"loss": 0.782,
|
| 458 |
+
"step": 310
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.3636363636363638,
|
| 462 |
+
"grad_norm": 0.9621470633936673,
|
| 463 |
+
"learning_rate": 5.5785276150654925e-06,
|
| 464 |
+
"loss": 0.7969,
|
| 465 |
+
"step": 315
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.3852813852813852,
|
| 469 |
+
"grad_norm": 0.9875120746768518,
|
| 470 |
+
"learning_rate": 5.242278192135682e-06,
|
| 471 |
+
"loss": 0.8255,
|
| 472 |
+
"step": 320
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.406926406926407,
|
| 476 |
+
"grad_norm": 0.9703641243841216,
|
| 477 |
+
"learning_rate": 4.912844159866113e-06,
|
| 478 |
+
"loss": 0.8081,
|
| 479 |
+
"step": 325
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.4285714285714286,
|
| 483 |
+
"grad_norm": 1.0333712850123837,
|
| 484 |
+
"learning_rate": 4.59069742933468e-06,
|
| 485 |
+
"loss": 0.8274,
|
| 486 |
+
"step": 330
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.4502164502164503,
|
| 490 |
+
"grad_norm": 0.9681660489299534,
|
| 491 |
+
"learning_rate": 4.276299472629735e-06,
|
| 492 |
+
"loss": 0.7975,
|
| 493 |
+
"step": 335
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.4718614718614718,
|
| 497 |
+
"grad_norm": 1.0012088021253578,
|
| 498 |
+
"learning_rate": 3.970100661795766e-06,
|
| 499 |
+
"loss": 0.7774,
|
| 500 |
+
"step": 340
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.4935064935064934,
|
| 504 |
+
"grad_norm": 0.9522339979957177,
|
| 505 |
+
"learning_rate": 3.6725396236797937e-06,
|
| 506 |
+
"loss": 0.8325,
|
| 507 |
+
"step": 345
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.5151515151515151,
|
| 511 |
+
"grad_norm": 0.9539695380134816,
|
| 512 |
+
"learning_rate": 3.384042611602605e-06,
|
| 513 |
+
"loss": 0.792,
|
| 514 |
+
"step": 350
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.5367965367965368,
|
| 518 |
+
"grad_norm": 0.9876834897890561,
|
| 519 |
+
"learning_rate": 3.105022894755003e-06,
|
| 520 |
+
"loss": 0.8234,
|
| 521 |
+
"step": 355
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.5584415584415585,
|
| 525 |
+
"grad_norm": 0.9521924923182135,
|
| 526 |
+
"learning_rate": 2.835880166193683e-06,
|
| 527 |
+
"loss": 0.7782,
|
| 528 |
+
"step": 360
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.5800865800865802,
|
| 532 |
+
"grad_norm": 0.9556390207534557,
|
| 533 |
+
"learning_rate": 2.576999970284735e-06,
|
| 534 |
+
"loss": 0.8192,
|
| 535 |
+
"step": 365
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.601731601731602,
|
| 539 |
+
"grad_norm": 0.9187916659434155,
|
| 540 |
+
"learning_rate": 2.328753150415094e-06,
|
| 541 |
+
"loss": 0.8044,
|
| 542 |
+
"step": 370
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.6233766233766234,
|
| 546 |
+
"grad_norm": 0.9663741256129553,
|
| 547 |
+
"learning_rate": 2.091495317762955e-06,
|
| 548 |
+
"loss": 0.7728,
|
| 549 |
+
"step": 375
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.645021645021645,
|
| 553 |
+
"grad_norm": 0.9519160746498837,
|
| 554 |
+
"learning_rate": 1.8655663418881587e-06,
|
| 555 |
+
"loss": 0.776,
|
| 556 |
+
"step": 380
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.6666666666666665,
|
| 560 |
+
"grad_norm": 1.0062143070641176,
|
| 561 |
+
"learning_rate": 1.6512898638723496e-06,
|
| 562 |
+
"loss": 0.7845,
|
| 563 |
+
"step": 385
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.6883116883116882,
|
| 567 |
+
"grad_norm": 0.9807077916152634,
|
| 568 |
+
"learning_rate": 1.4489728327062325e-06,
|
| 569 |
+
"loss": 0.7894,
|
| 570 |
+
"step": 390
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.70995670995671,
|
| 574 |
+
"grad_norm": 0.9521910568116111,
|
| 575 |
+
"learning_rate": 1.258905065588103e-06,
|
| 576 |
+
"loss": 0.7711,
|
| 577 |
+
"step": 395
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.7316017316017316,
|
| 581 |
+
"grad_norm": 0.9157886273812276,
|
| 582 |
+
"learning_rate": 1.0813588327634961e-06,
|
| 583 |
+
"loss": 0.7872,
|
| 584 |
+
"step": 400
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.7532467532467533,
|
| 588 |
+
"grad_norm": 0.9518282733524287,
|
| 589 |
+
"learning_rate": 9.165884675006931e-07,
|
| 590 |
+
"loss": 0.7874,
|
| 591 |
+
"step": 405
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.774891774891775,
|
| 595 |
+
"grad_norm": 0.9282109794972182,
|
| 596 |
+
"learning_rate": 7.648300017607535e-07,
|
| 597 |
+
"loss": 0.7737,
|
| 598 |
+
"step": 410
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.7965367965367967,
|
| 602 |
+
"grad_norm": 1.0188698575765125,
|
| 603 |
+
"learning_rate": 6.263008280840133e-07,
|
| 604 |
+
"loss": 0.774,
|
| 605 |
+
"step": 415
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.8181818181818183,
|
| 609 |
+
"grad_norm": 0.953904828798195,
|
| 610 |
+
"learning_rate": 5.011993881773569e-07,
|
| 611 |
+
"loss": 0.7737,
|
| 612 |
+
"step": 420
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.8398268398268398,
|
| 616 |
+
"grad_norm": 0.9612309743009354,
|
| 617 |
+
"learning_rate": 3.897048886483934e-07,
|
| 618 |
+
"loss": 0.7734,
|
| 619 |
+
"step": 425
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.8614718614718615,
|
| 623 |
+
"grad_norm": 0.9540258820581294,
|
| 624 |
+
"learning_rate": 2.9197704429370976e-07,
|
| 625 |
+
"loss": 0.7728,
|
| 626 |
+
"step": 430
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.883116883116883,
|
| 630 |
+
"grad_norm": 0.9391259666010853,
|
| 631 |
+
"learning_rate": 2.0815584930895972e-07,
|
| 632 |
+
"loss": 0.7846,
|
| 633 |
+
"step": 435
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.9047619047619047,
|
| 637 |
+
"grad_norm": 0.923039210320182,
|
| 638 |
+
"learning_rate": 1.3836137674854254e-07,
|
| 639 |
+
"loss": 0.7948,
|
| 640 |
+
"step": 440
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.9264069264069263,
|
| 644 |
+
"grad_norm": 0.9661085243975593,
|
| 645 |
+
"learning_rate": 8.269360652209291e-08,
|
| 646 |
+
"loss": 0.775,
|
| 647 |
+
"step": 445
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.948051948051948,
|
| 651 |
+
"grad_norm": 0.8801221387422419,
|
| 652 |
+
"learning_rate": 4.1232282174229475e-08,
|
| 653 |
+
"loss": 0.805,
|
| 654 |
+
"step": 450
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.9696969696969697,
|
| 658 |
+
"grad_norm": 0.9540642633518202,
|
| 659 |
+
"learning_rate": 1.403679665270108e-08,
|
| 660 |
+
"loss": 0.7993,
|
| 661 |
+
"step": 455
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.9913419913419914,
|
| 665 |
+
"grad_norm": 0.938537521673335,
|
| 666 |
+
"learning_rate": 1.1461072285490205e-09,
|
| 667 |
+
"loss": 0.8082,
|
| 668 |
+
"step": 460
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 2.0,
|
| 672 |
+
"eval_loss": 1.1257315874099731,
|
| 673 |
+
"eval_runtime": 5.4011,
|
| 674 |
+
"eval_samples_per_second": 28.142,
|
| 675 |
+
"eval_steps_per_second": 0.555,
|
| 676 |
+
"step": 462
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 2.0,
|
| 680 |
+
"step": 462,
|
| 681 |
+
"total_flos": 96733400924160.0,
|
| 682 |
+
"train_loss": 0.9972741247771623,
|
| 683 |
+
"train_runtime": 3285.0208,
|
| 684 |
+
"train_samples_per_second": 9.0,
|
| 685 |
+
"train_steps_per_second": 0.141
|
| 686 |
+
}
|
| 687 |
+
],
|
| 688 |
+
"logging_steps": 5,
|
| 689 |
+
"max_steps": 462,
|
| 690 |
+
"num_input_tokens_seen": 0,
|
| 691 |
+
"num_train_epochs": 2,
|
| 692 |
+
"save_steps": 500,
|
| 693 |
+
"stateful_callbacks": {
|
| 694 |
+
"TrainerControl": {
|
| 695 |
+
"args": {
|
| 696 |
+
"should_epoch_stop": false,
|
| 697 |
+
"should_evaluate": false,
|
| 698 |
+
"should_log": false,
|
| 699 |
+
"should_save": true,
|
| 700 |
+
"should_training_stop": true
|
| 701 |
+
},
|
| 702 |
+
"attributes": {}
|
| 703 |
+
}
|
| 704 |
+
},
|
| 705 |
+
"total_flos": 96733400924160.0,
|
| 706 |
+
"train_batch_size": 8,
|
| 707 |
+
"trial_name": null,
|
| 708 |
+
"trial_params": null
|
| 709 |
+
}
|