Commit
•
6d1bd4a
1
Parent(s):
81fc83a
End of training
Browse files
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"eval_accuracy": 0.796813717099257,
|
| 4 |
+
"eval_loss": 1.4686633348464966,
|
| 5 |
+
"eval_runtime": 1611.9954,
|
| 6 |
+
"eval_samples": 71199,
|
| 7 |
+
"eval_samples_per_second": 44.168,
|
| 8 |
+
"eval_steps_per_second": 1.38,
|
| 9 |
+
"perplexity": 4.3434255536068305,
|
| 10 |
+
"train_loss": 1.455913220122233,
|
| 11 |
+
"train_runtime": 36506.3836,
|
| 12 |
+
"train_samples": 1353936,
|
| 13 |
+
"train_samples_per_second": 37.088,
|
| 14 |
+
"train_steps_per_second": 1.159
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"eval_accuracy": 0.796813717099257,
|
| 4 |
+
"eval_loss": 1.4686633348464966,
|
| 5 |
+
"eval_runtime": 1611.9954,
|
| 6 |
+
"eval_samples": 71199,
|
| 7 |
+
"eval_samples_per_second": 44.168,
|
| 8 |
+
"eval_steps_per_second": 1.38,
|
| 9 |
+
"perplexity": 4.3434255536068305
|
| 10 |
+
}
|
runs/Oct22_09-31-40_hkmustamd-Super-Server/events.out.tfevents.1729599649.hkmustamd-Super-Server.3603284.2
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1fe16ef686542b12150e6e220d0e391719dd4e7d9a5b6a965ce17f5b64618108
|
| 3 |
+
size 417
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"train_loss": 1.455913220122233,
|
| 4 |
+
"train_runtime": 36506.3836,
|
| 5 |
+
"train_samples": 1353936,
|
| 6 |
+
"train_samples_per_second": 37.088,
|
| 7 |
+
"train_steps_per_second": 1.159
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,529 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"global_step": 42311,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.01,
|
| 12 |
+
"learning_rate": 5.907372400756144e-06,
|
| 13 |
+
"loss": 1.9951,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.02,
|
| 18 |
+
"learning_rate": 1.1814744801512288e-05,
|
| 19 |
+
"loss": 1.4127,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.04,
|
| 24 |
+
"learning_rate": 1.7722117202268433e-05,
|
| 25 |
+
"loss": 1.3188,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.05,
|
| 30 |
+
"learning_rate": 2.3629489603024576e-05,
|
| 31 |
+
"loss": 1.2728,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.06,
|
| 36 |
+
"learning_rate": 2.953686200378072e-05,
|
| 37 |
+
"loss": 1.248,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.07,
|
| 42 |
+
"learning_rate": 3.5444234404536866e-05,
|
| 43 |
+
"loss": 1.2485,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.08,
|
| 48 |
+
"learning_rate": 4.135160680529301e-05,
|
| 49 |
+
"loss": 1.2476,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.09,
|
| 54 |
+
"learning_rate": 4.725897920604915e-05,
|
| 55 |
+
"loss": 1.259,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.11,
|
| 60 |
+
"learning_rate": 4.9648100002626124e-05,
|
| 61 |
+
"loss": 1.2954,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.12,
|
| 66 |
+
"learning_rate": 4.899157015677933e-05,
|
| 67 |
+
"loss": 1.3149,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.13,
|
| 72 |
+
"learning_rate": 4.833504031093254e-05,
|
| 73 |
+
"loss": 1.3251,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.14,
|
| 78 |
+
"learning_rate": 4.767851046508575e-05,
|
| 79 |
+
"loss": 1.342,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.15,
|
| 84 |
+
"learning_rate": 4.702198061923895e-05,
|
| 85 |
+
"loss": 1.344,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.17,
|
| 90 |
+
"learning_rate": 4.636545077339216e-05,
|
| 91 |
+
"loss": 1.335,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.18,
|
| 96 |
+
"learning_rate": 4.570892092754537e-05,
|
| 97 |
+
"loss": 1.3276,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.19,
|
| 102 |
+
"learning_rate": 4.5052391081698575e-05,
|
| 103 |
+
"loss": 1.3311,
|
| 104 |
+
"step": 8000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.2,
|
| 108 |
+
"learning_rate": 4.4395861235851784e-05,
|
| 109 |
+
"loss": 1.3312,
|
| 110 |
+
"step": 8500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.21,
|
| 114 |
+
"learning_rate": 4.373933139000499e-05,
|
| 115 |
+
"loss": 1.3298,
|
| 116 |
+
"step": 9000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.22,
|
| 120 |
+
"learning_rate": 4.30828015441582e-05,
|
| 121 |
+
"loss": 1.3236,
|
| 122 |
+
"step": 9500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.24,
|
| 126 |
+
"learning_rate": 4.242627169831141e-05,
|
| 127 |
+
"loss": 1.3251,
|
| 128 |
+
"step": 10000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.25,
|
| 132 |
+
"learning_rate": 4.176974185246462e-05,
|
| 133 |
+
"loss": 1.3347,
|
| 134 |
+
"step": 10500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.26,
|
| 138 |
+
"learning_rate": 4.1113212006617825e-05,
|
| 139 |
+
"loss": 1.3309,
|
| 140 |
+
"step": 11000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.27,
|
| 144 |
+
"learning_rate": 4.0456682160771034e-05,
|
| 145 |
+
"loss": 1.3392,
|
| 146 |
+
"step": 11500
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.28,
|
| 150 |
+
"learning_rate": 3.9800152314924235e-05,
|
| 151 |
+
"loss": 1.3407,
|
| 152 |
+
"step": 12000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.3,
|
| 156 |
+
"learning_rate": 3.914362246907744e-05,
|
| 157 |
+
"loss": 1.3494,
|
| 158 |
+
"step": 12500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.31,
|
| 162 |
+
"learning_rate": 3.848709262323065e-05,
|
| 163 |
+
"loss": 1.3534,
|
| 164 |
+
"step": 13000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.32,
|
| 168 |
+
"learning_rate": 3.783056277738386e-05,
|
| 169 |
+
"loss": 1.3561,
|
| 170 |
+
"step": 13500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.33,
|
| 174 |
+
"learning_rate": 3.717403293153707e-05,
|
| 175 |
+
"loss": 1.3614,
|
| 176 |
+
"step": 14000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.34,
|
| 180 |
+
"learning_rate": 3.651750308569028e-05,
|
| 181 |
+
"loss": 1.3675,
|
| 182 |
+
"step": 14500
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.35,
|
| 186 |
+
"learning_rate": 3.5860973239843485e-05,
|
| 187 |
+
"loss": 1.3728,
|
| 188 |
+
"step": 15000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.37,
|
| 192 |
+
"learning_rate": 3.520444339399669e-05,
|
| 193 |
+
"loss": 1.3832,
|
| 194 |
+
"step": 15500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.38,
|
| 198 |
+
"learning_rate": 3.45479135481499e-05,
|
| 199 |
+
"loss": 1.3866,
|
| 200 |
+
"step": 16000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.39,
|
| 204 |
+
"learning_rate": 3.389138370230311e-05,
|
| 205 |
+
"loss": 1.4043,
|
| 206 |
+
"step": 16500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.4,
|
| 210 |
+
"learning_rate": 3.323485385645632e-05,
|
| 211 |
+
"loss": 1.4023,
|
| 212 |
+
"step": 17000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.41,
|
| 216 |
+
"learning_rate": 3.2578324010609527e-05,
|
| 217 |
+
"loss": 1.4125,
|
| 218 |
+
"step": 17500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.43,
|
| 222 |
+
"learning_rate": 3.192179416476273e-05,
|
| 223 |
+
"loss": 1.4165,
|
| 224 |
+
"step": 18000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.44,
|
| 228 |
+
"learning_rate": 3.1265264318915936e-05,
|
| 229 |
+
"loss": 1.4195,
|
| 230 |
+
"step": 18500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.45,
|
| 234 |
+
"learning_rate": 3.0608734473069145e-05,
|
| 235 |
+
"loss": 1.4239,
|
| 236 |
+
"step": 19000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.46,
|
| 240 |
+
"learning_rate": 2.9952204627222353e-05,
|
| 241 |
+
"loss": 1.4311,
|
| 242 |
+
"step": 19500
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.47,
|
| 246 |
+
"learning_rate": 2.9295674781375565e-05,
|
| 247 |
+
"loss": 1.4391,
|
| 248 |
+
"step": 20000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.48,
|
| 252 |
+
"learning_rate": 2.8639144935528773e-05,
|
| 253 |
+
"loss": 1.4343,
|
| 254 |
+
"step": 20500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.5,
|
| 258 |
+
"learning_rate": 2.7982615089681978e-05,
|
| 259 |
+
"loss": 1.4432,
|
| 260 |
+
"step": 21000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.51,
|
| 264 |
+
"learning_rate": 2.7326085243835186e-05,
|
| 265 |
+
"loss": 1.4515,
|
| 266 |
+
"step": 21500
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.52,
|
| 270 |
+
"learning_rate": 2.6669555397988395e-05,
|
| 271 |
+
"loss": 1.4579,
|
| 272 |
+
"step": 22000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.53,
|
| 276 |
+
"learning_rate": 2.60130255521416e-05,
|
| 277 |
+
"loss": 1.4601,
|
| 278 |
+
"step": 22500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.54,
|
| 282 |
+
"learning_rate": 2.5356495706294808e-05,
|
| 283 |
+
"loss": 1.4647,
|
| 284 |
+
"step": 23000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.56,
|
| 288 |
+
"learning_rate": 2.469996586044802e-05,
|
| 289 |
+
"loss": 1.4735,
|
| 290 |
+
"step": 23500
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.57,
|
| 294 |
+
"learning_rate": 2.4043436014601224e-05,
|
| 295 |
+
"loss": 1.4769,
|
| 296 |
+
"step": 24000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.58,
|
| 300 |
+
"learning_rate": 2.3386906168754433e-05,
|
| 301 |
+
"loss": 1.4876,
|
| 302 |
+
"step": 24500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.59,
|
| 306 |
+
"learning_rate": 2.273037632290764e-05,
|
| 307 |
+
"loss": 1.49,
|
| 308 |
+
"step": 25000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.6,
|
| 312 |
+
"learning_rate": 2.2073846477060846e-05,
|
| 313 |
+
"loss": 1.4942,
|
| 314 |
+
"step": 25500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.61,
|
| 318 |
+
"learning_rate": 2.1417316631214058e-05,
|
| 319 |
+
"loss": 1.4988,
|
| 320 |
+
"step": 26000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.63,
|
| 324 |
+
"learning_rate": 2.0760786785367266e-05,
|
| 325 |
+
"loss": 1.5122,
|
| 326 |
+
"step": 26500
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.64,
|
| 330 |
+
"learning_rate": 2.010425693952047e-05,
|
| 331 |
+
"loss": 1.5134,
|
| 332 |
+
"step": 27000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.65,
|
| 336 |
+
"learning_rate": 1.944772709367368e-05,
|
| 337 |
+
"loss": 1.5191,
|
| 338 |
+
"step": 27500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.66,
|
| 342 |
+
"learning_rate": 1.8791197247826888e-05,
|
| 343 |
+
"loss": 1.5193,
|
| 344 |
+
"step": 28000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.67,
|
| 348 |
+
"learning_rate": 1.8134667401980096e-05,
|
| 349 |
+
"loss": 1.5262,
|
| 350 |
+
"step": 28500
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.69,
|
| 354 |
+
"learning_rate": 1.7478137556133304e-05,
|
| 355 |
+
"loss": 1.5383,
|
| 356 |
+
"step": 29000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.7,
|
| 360 |
+
"learning_rate": 1.682160771028651e-05,
|
| 361 |
+
"loss": 1.5349,
|
| 362 |
+
"step": 29500
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.71,
|
| 366 |
+
"learning_rate": 1.6165077864439717e-05,
|
| 367 |
+
"loss": 1.5411,
|
| 368 |
+
"step": 30000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.72,
|
| 372 |
+
"learning_rate": 1.5508548018592926e-05,
|
| 373 |
+
"loss": 1.5512,
|
| 374 |
+
"step": 30500
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.73,
|
| 378 |
+
"learning_rate": 1.4852018172746136e-05,
|
| 379 |
+
"loss": 1.5535,
|
| 380 |
+
"step": 31000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.74,
|
| 384 |
+
"learning_rate": 1.4195488326899342e-05,
|
| 385 |
+
"loss": 1.5574,
|
| 386 |
+
"step": 31500
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.76,
|
| 390 |
+
"learning_rate": 1.3538958481052549e-05,
|
| 391 |
+
"loss": 1.5595,
|
| 392 |
+
"step": 32000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.77,
|
| 396 |
+
"learning_rate": 1.2882428635205757e-05,
|
| 397 |
+
"loss": 1.5647,
|
| 398 |
+
"step": 32500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.78,
|
| 402 |
+
"learning_rate": 1.2225898789358966e-05,
|
| 403 |
+
"loss": 1.566,
|
| 404 |
+
"step": 33000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.79,
|
| 408 |
+
"learning_rate": 1.1569368943512172e-05,
|
| 409 |
+
"loss": 1.5691,
|
| 410 |
+
"step": 33500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.8,
|
| 414 |
+
"learning_rate": 1.091283909766538e-05,
|
| 415 |
+
"loss": 1.5736,
|
| 416 |
+
"step": 34000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.82,
|
| 420 |
+
"learning_rate": 1.0256309251818589e-05,
|
| 421 |
+
"loss": 1.5762,
|
| 422 |
+
"step": 34500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.83,
|
| 426 |
+
"learning_rate": 9.599779405971796e-06,
|
| 427 |
+
"loss": 1.5778,
|
| 428 |
+
"step": 35000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.84,
|
| 432 |
+
"learning_rate": 8.943249560125004e-06,
|
| 433 |
+
"loss": 1.5796,
|
| 434 |
+
"step": 35500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.85,
|
| 438 |
+
"learning_rate": 8.286719714278212e-06,
|
| 439 |
+
"loss": 1.574,
|
| 440 |
+
"step": 36000
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.86,
|
| 444 |
+
"learning_rate": 7.63018986843142e-06,
|
| 445 |
+
"loss": 1.5776,
|
| 446 |
+
"step": 36500
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.87,
|
| 450 |
+
"learning_rate": 6.973660022584627e-06,
|
| 451 |
+
"loss": 1.5834,
|
| 452 |
+
"step": 37000
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.89,
|
| 456 |
+
"learning_rate": 6.3171301767378345e-06,
|
| 457 |
+
"loss": 1.5949,
|
| 458 |
+
"step": 37500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.9,
|
| 462 |
+
"learning_rate": 5.660600330891042e-06,
|
| 463 |
+
"loss": 1.582,
|
| 464 |
+
"step": 38000
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.91,
|
| 468 |
+
"learning_rate": 5.00407048504425e-06,
|
| 469 |
+
"loss": 1.5904,
|
| 470 |
+
"step": 38500
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.92,
|
| 474 |
+
"learning_rate": 4.347540639197458e-06,
|
| 475 |
+
"loss": 1.595,
|
| 476 |
+
"step": 39000
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.93,
|
| 480 |
+
"learning_rate": 3.691010793350666e-06,
|
| 481 |
+
"loss": 1.5847,
|
| 482 |
+
"step": 39500
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.95,
|
| 486 |
+
"learning_rate": 3.0344809475038736e-06,
|
| 487 |
+
"loss": 1.5795,
|
| 488 |
+
"step": 40000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.96,
|
| 492 |
+
"learning_rate": 2.3779511016570815e-06,
|
| 493 |
+
"loss": 1.5827,
|
| 494 |
+
"step": 40500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.97,
|
| 498 |
+
"learning_rate": 1.7214212558102892e-06,
|
| 499 |
+
"loss": 1.5895,
|
| 500 |
+
"step": 41000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.98,
|
| 504 |
+
"learning_rate": 1.064891409963497e-06,
|
| 505 |
+
"loss": 1.5856,
|
| 506 |
+
"step": 41500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.99,
|
| 510 |
+
"learning_rate": 4.083615641167047e-07,
|
| 511 |
+
"loss": 1.5763,
|
| 512 |
+
"step": 42000
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 1.0,
|
| 516 |
+
"step": 42311,
|
| 517 |
+
"total_flos": 4.127720706618163e+17,
|
| 518 |
+
"train_loss": 1.455913220122233,
|
| 519 |
+
"train_runtime": 36506.3836,
|
| 520 |
+
"train_samples_per_second": 37.088,
|
| 521 |
+
"train_steps_per_second": 1.159
|
| 522 |
+
}
|
| 523 |
+
],
|
| 524 |
+
"max_steps": 42311,
|
| 525 |
+
"num_train_epochs": 1,
|
| 526 |
+
"total_flos": 4.127720706618163e+17,
|
| 527 |
+
"trial_name": null,
|
| 528 |
+
"trial_params": null
|
| 529 |
+
}
|