Upload task output 1
Browse files- loss.txt +1 -1
- model.safetensors +1 -1
- trainer_state.json +417 -33
- training_args.bin +1 -1
loss.txt
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
|
|
|
|
| 1 |
+
292,0.00010653198114596307
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2200119864
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:db049db56e23cb51c422a8f629d2c1a3077b5b3c714d1116de01ccfd8591f159
|
| 3 |
size 2200119864
|
trainer_state.json
CHANGED
|
@@ -2,54 +2,438 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
-
"epoch": 0.
|
| 14 |
-
"
|
| 15 |
-
"
|
| 16 |
-
"
|
| 17 |
-
"
|
| 18 |
-
"step": 4
|
| 19 |
},
|
| 20 |
{
|
| 21 |
-
"epoch":
|
| 22 |
-
"grad_norm":
|
| 23 |
-
"learning_rate": 1.
|
| 24 |
-
"loss": 0.
|
| 25 |
-
"step":
|
| 26 |
},
|
| 27 |
{
|
| 28 |
-
"epoch":
|
| 29 |
-
"
|
| 30 |
-
"
|
| 31 |
-
"
|
| 32 |
-
"
|
| 33 |
-
"step": 8
|
| 34 |
},
|
| 35 |
{
|
| 36 |
-
"epoch":
|
| 37 |
-
"grad_norm":
|
| 38 |
-
"learning_rate":
|
| 39 |
-
"loss": 0.
|
| 40 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 41 |
},
|
| 42 |
{
|
| 43 |
-
"epoch":
|
| 44 |
-
"eval_loss":
|
| 45 |
-
"eval_runtime": 0.
|
| 46 |
-
"eval_samples_per_second":
|
| 47 |
-
"eval_steps_per_second":
|
| 48 |
-
"step":
|
| 49 |
}
|
| 50 |
],
|
| 51 |
"logging_steps": 5,
|
| 52 |
-
"max_steps":
|
| 53 |
"num_input_tokens_seen": 0,
|
| 54 |
"num_train_epochs": 3,
|
| 55 |
"save_steps": 500,
|
|
@@ -65,8 +449,8 @@
|
|
| 65 |
"attributes": {}
|
| 66 |
}
|
| 67 |
},
|
| 68 |
-
"total_flos": 1.
|
| 69 |
-
"train_batch_size":
|
| 70 |
"trial_name": null,
|
| 71 |
"trial_params": null
|
| 72 |
}
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.9863945578231292,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 292,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
+
"epoch": 0.034013605442176874,
|
| 14 |
+
"grad_norm": 34.75,
|
| 15 |
+
"learning_rate": 6.430830146224649e-06,
|
| 16 |
+
"loss": 0.6785,
|
| 17 |
+
"step": 5
|
|
|
|
| 18 |
},
|
| 19 |
{
|
| 20 |
+
"epoch": 0.06802721088435375,
|
| 21 |
+
"grad_norm": 110.0,
|
| 22 |
+
"learning_rate": 1.4469367829005459e-05,
|
| 23 |
+
"loss": 0.15,
|
| 24 |
+
"step": 10
|
| 25 |
},
|
| 26 |
{
|
| 27 |
+
"epoch": 0.10204081632653061,
|
| 28 |
+
"grad_norm": 0.00640869140625,
|
| 29 |
+
"learning_rate": 2.2507905511786274e-05,
|
| 30 |
+
"loss": 0.0071,
|
| 31 |
+
"step": 15
|
|
|
|
| 32 |
},
|
| 33 |
{
|
| 34 |
+
"epoch": 0.1360544217687075,
|
| 35 |
+
"grad_norm": 0.0135498046875,
|
| 36 |
+
"learning_rate": 3.054644319456708e-05,
|
| 37 |
+
"loss": 0.0692,
|
| 38 |
+
"step": 20
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.17006802721088435,
|
| 42 |
+
"grad_norm": 0.15234375,
|
| 43 |
+
"learning_rate": 3.858498087734789e-05,
|
| 44 |
+
"loss": 0.0616,
|
| 45 |
+
"step": 25
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.20408163265306123,
|
| 49 |
+
"grad_norm": 0.003021240234375,
|
| 50 |
+
"learning_rate": 4.6623518560128706e-05,
|
| 51 |
+
"loss": 0.0001,
|
| 52 |
+
"step": 30
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.23809523809523808,
|
| 56 |
+
"grad_norm": 28.875,
|
| 57 |
+
"learning_rate": 5.466205624290951e-05,
|
| 58 |
+
"loss": 0.0322,
|
| 59 |
+
"step": 35
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.272108843537415,
|
| 63 |
+
"grad_norm": 17.75,
|
| 64 |
+
"learning_rate": 5.6259657079961125e-05,
|
| 65 |
+
"loss": 0.0862,
|
| 66 |
+
"step": 40
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.30612244897959184,
|
| 70 |
+
"grad_norm": 0.283203125,
|
| 71 |
+
"learning_rate": 5.621861520545433e-05,
|
| 72 |
+
"loss": 0.0157,
|
| 73 |
+
"step": 45
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.3401360544217687,
|
| 77 |
+
"grad_norm": 0.0096435546875,
|
| 78 |
+
"learning_rate": 5.614606786714469e-05,
|
| 79 |
+
"loss": 0.0441,
|
| 80 |
+
"step": 50
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.3741496598639456,
|
| 84 |
+
"grad_norm": 0.271484375,
|
| 85 |
+
"learning_rate": 5.604212364632182e-05,
|
| 86 |
+
"loss": 0.0507,
|
| 87 |
+
"step": 55
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.40816326530612246,
|
| 91 |
+
"grad_norm": 0.02734375,
|
| 92 |
+
"learning_rate": 5.590693811585455e-05,
|
| 93 |
+
"loss": 0.0012,
|
| 94 |
+
"step": 60
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.4421768707482993,
|
| 98 |
+
"grad_norm": 0.189453125,
|
| 99 |
+
"learning_rate": 5.5740713607345624e-05,
|
| 100 |
+
"loss": 0.0152,
|
| 101 |
+
"step": 65
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.47619047619047616,
|
| 105 |
+
"grad_norm": 0.05078125,
|
| 106 |
+
"learning_rate": 5.5543698908302936e-05,
|
| 107 |
+
"loss": 0.0634,
|
| 108 |
+
"step": 70
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.5102040816326531,
|
| 112 |
+
"grad_norm": 42.5,
|
| 113 |
+
"learning_rate": 5.531618888978025e-05,
|
| 114 |
+
"loss": 0.0458,
|
| 115 |
+
"step": 75
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.54421768707483,
|
| 119 |
+
"grad_norm": 17.25,
|
| 120 |
+
"learning_rate": 5.505852406504504e-05,
|
| 121 |
+
"loss": 0.0562,
|
| 122 |
+
"step": 80
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.5782312925170068,
|
| 126 |
+
"grad_norm": 10.3125,
|
| 127 |
+
"learning_rate": 5.47710900799337e-05,
|
| 128 |
+
"loss": 0.1128,
|
| 129 |
+
"step": 85
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.6122448979591837,
|
| 133 |
+
"grad_norm": 17.125,
|
| 134 |
+
"learning_rate": 5.445431713565707e-05,
|
| 135 |
+
"loss": 0.0146,
|
| 136 |
+
"step": 90
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.6462585034013606,
|
| 140 |
+
"grad_norm": 42.0,
|
| 141 |
+
"learning_rate": 5.4108679344920166e-05,
|
| 142 |
+
"loss": 0.1167,
|
| 143 |
+
"step": 95
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.6802721088435374,
|
| 147 |
+
"grad_norm": 0.0869140625,
|
| 148 |
+
"learning_rate": 5.373469402231971e-05,
|
| 149 |
+
"loss": 0.0209,
|
| 150 |
+
"step": 100
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.7142857142857143,
|
| 154 |
+
"grad_norm": 0.8359375,
|
| 155 |
+
"learning_rate": 5.333292091008159e-05,
|
| 156 |
+
"loss": 0.0019,
|
| 157 |
+
"step": 105
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.7482993197278912,
|
| 161 |
+
"grad_norm": 0.07373046875,
|
| 162 |
+
"learning_rate": 5.290396134029709e-05,
|
| 163 |
+
"loss": 0.0157,
|
| 164 |
+
"step": 110
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.782312925170068,
|
| 168 |
+
"grad_norm": 0.2578125,
|
| 169 |
+
"learning_rate": 5.244845733491173e-05,
|
| 170 |
+
"loss": 0.0051,
|
| 171 |
+
"step": 115
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.8163265306122449,
|
| 175 |
+
"grad_norm": 67.5,
|
| 176 |
+
"learning_rate": 5.196709064481376e-05,
|
| 177 |
+
"loss": 0.0181,
|
| 178 |
+
"step": 120
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.8503401360544217,
|
| 182 |
+
"grad_norm": 0.0478515625,
|
| 183 |
+
"learning_rate": 5.146058172946055e-05,
|
| 184 |
+
"loss": 0.1658,
|
| 185 |
+
"step": 125
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.8843537414965986,
|
| 189 |
+
"grad_norm": 3.46875,
|
| 190 |
+
"learning_rate": 5.092968867857002e-05,
|
| 191 |
+
"loss": 0.0155,
|
| 192 |
+
"step": 130
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.9183673469387755,
|
| 196 |
+
"grad_norm": 0.040283203125,
|
| 197 |
+
"learning_rate": 5.037520607749101e-05,
|
| 198 |
+
"loss": 0.0025,
|
| 199 |
+
"step": 135
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.9523809523809523,
|
| 203 |
+
"grad_norm": 0.80859375,
|
| 204 |
+
"learning_rate": 4.979796381795077e-05,
|
| 205 |
+
"loss": 0.0006,
|
| 206 |
+
"step": 140
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.9863945578231292,
|
| 210 |
+
"grad_norm": 0.0189208984375,
|
| 211 |
+
"learning_rate": 4.919882585595965e-05,
|
| 212 |
+
"loss": 0.0738,
|
| 213 |
+
"step": 145
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.9931972789115646,
|
| 217 |
+
"eval_loss": 0.05052933469414711,
|
| 218 |
+
"eval_runtime": 0.8602,
|
| 219 |
+
"eval_samples_per_second": 30.225,
|
| 220 |
+
"eval_steps_per_second": 30.225,
|
| 221 |
+
"step": 146
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 1.0204081632653061,
|
| 225 |
+
"grad_norm": 25.0,
|
| 226 |
+
"learning_rate": 4.8578688918731776e-05,
|
| 227 |
+
"loss": 0.0782,
|
| 228 |
+
"step": 150
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 1.054421768707483,
|
| 232 |
+
"grad_norm": 0.08349609375,
|
| 233 |
+
"learning_rate": 4.7938481162557315e-05,
|
| 234 |
+
"loss": 0.0015,
|
| 235 |
+
"step": 155
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 1.08843537414966,
|
| 239 |
+
"grad_norm": 0.0025177001953125,
|
| 240 |
+
"learning_rate": 4.727916078363493e-05,
|
| 241 |
+
"loss": 0.0082,
|
| 242 |
+
"step": 160
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.1224489795918366,
|
| 246 |
+
"grad_norm": 0.03515625,
|
| 247 |
+
"learning_rate": 4.6601714583943683e-05,
|
| 248 |
+
"loss": 0.0,
|
| 249 |
+
"step": 165
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.1564625850340136,
|
| 253 |
+
"grad_norm": 0.0101318359375,
|
| 254 |
+
"learning_rate": 4.590715649430069e-05,
|
| 255 |
+
"loss": 0.0031,
|
| 256 |
+
"step": 170
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 1.1904761904761905,
|
| 260 |
+
"grad_norm": 0.01348876953125,
|
| 261 |
+
"learning_rate": 4.519652605681526e-05,
|
| 262 |
+
"loss": 0.0001,
|
| 263 |
+
"step": 175
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 1.2244897959183674,
|
| 267 |
+
"grad_norm": 0.0238037109375,
|
| 268 |
+
"learning_rate": 4.447088686901071e-05,
|
| 269 |
+
"loss": 0.0119,
|
| 270 |
+
"step": 180
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 1.2585034013605443,
|
| 274 |
+
"grad_norm": 2.515625,
|
| 275 |
+
"learning_rate": 4.373132499194244e-05,
|
| 276 |
+
"loss": 0.0016,
|
| 277 |
+
"step": 185
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 1.2925170068027212,
|
| 281 |
+
"grad_norm": 0.059814453125,
|
| 282 |
+
"learning_rate": 4.2978947324695116e-05,
|
| 283 |
+
"loss": 0.0283,
|
| 284 |
+
"step": 190
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.3265306122448979,
|
| 288 |
+
"grad_norm": 9.1875,
|
| 289 |
+
"learning_rate": 4.221487994769145e-05,
|
| 290 |
+
"loss": 0.0045,
|
| 291 |
+
"step": 195
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 1.3605442176870748,
|
| 295 |
+
"grad_norm": 0.015625,
|
| 296 |
+
"learning_rate": 4.144026643729263e-05,
|
| 297 |
+
"loss": 0.005,
|
| 298 |
+
"step": 200
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 1.3945578231292517,
|
| 302 |
+
"grad_norm": 0.0034637451171875,
|
| 303 |
+
"learning_rate": 4.06562661542123e-05,
|
| 304 |
+
"loss": 0.0001,
|
| 305 |
+
"step": 205
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.4285714285714286,
|
| 309 |
+
"grad_norm": 0.016845703125,
|
| 310 |
+
"learning_rate": 3.986405250830651e-05,
|
| 311 |
+
"loss": 0.0369,
|
| 312 |
+
"step": 210
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 1.4625850340136055,
|
| 316 |
+
"grad_norm": 0.13671875,
|
| 317 |
+
"learning_rate": 3.9064811202336157e-05,
|
| 318 |
+
"loss": 0.0003,
|
| 319 |
+
"step": 215
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 1.4965986394557822,
|
| 323 |
+
"grad_norm": 5.03125,
|
| 324 |
+
"learning_rate": 3.8259738457330774e-05,
|
| 325 |
+
"loss": 0.0025,
|
| 326 |
+
"step": 220
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.5306122448979593,
|
| 330 |
+
"grad_norm": 0.796875,
|
| 331 |
+
"learning_rate": 3.7450039222209636e-05,
|
| 332 |
+
"loss": 0.0006,
|
| 333 |
+
"step": 225
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.564625850340136,
|
| 337 |
+
"grad_norm": 0.025146484375,
|
| 338 |
+
"learning_rate": 3.663692537033991e-05,
|
| 339 |
+
"loss": 0.0018,
|
| 340 |
+
"step": 230
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 1.598639455782313,
|
| 344 |
+
"grad_norm": 0.057861328125,
|
| 345 |
+
"learning_rate": 3.5821613885731e-05,
|
| 346 |
+
"loss": 0.0072,
|
| 347 |
+
"step": 235
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.6326530612244898,
|
| 351 |
+
"grad_norm": 0.02783203125,
|
| 352 |
+
"learning_rate": 3.500532504157975e-05,
|
| 353 |
+
"loss": 0.0003,
|
| 354 |
+
"step": 240
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.6666666666666665,
|
| 358 |
+
"grad_norm": 0.037841796875,
|
| 359 |
+
"learning_rate": 3.418928057389288e-05,
|
| 360 |
+
"loss": 0.0002,
|
| 361 |
+
"step": 245
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.7006802721088436,
|
| 365 |
+
"grad_norm": 0.01104736328125,
|
| 366 |
+
"learning_rate": 3.337470185291987e-05,
|
| 367 |
+
"loss": 0.0002,
|
| 368 |
+
"step": 250
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.7346938775510203,
|
| 372 |
+
"grad_norm": 0.044677734375,
|
| 373 |
+
"learning_rate": 3.25628080551334e-05,
|
| 374 |
+
"loss": 0.0156,
|
| 375 |
+
"step": 255
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.7687074829931972,
|
| 379 |
+
"grad_norm": 0.0037841796875,
|
| 380 |
+
"learning_rate": 3.175481433849315e-05,
|
| 381 |
+
"loss": 0.0002,
|
| 382 |
+
"step": 260
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 1.8027210884353742,
|
| 386 |
+
"grad_norm": 0.376953125,
|
| 387 |
+
"learning_rate": 3.0951930023724105e-05,
|
| 388 |
+
"loss": 0.001,
|
| 389 |
+
"step": 265
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.836734693877551,
|
| 393 |
+
"grad_norm": 0.00799560546875,
|
| 394 |
+
"learning_rate": 3.015535678433143e-05,
|
| 395 |
+
"loss": 0.0,
|
| 396 |
+
"step": 270
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 1.870748299319728,
|
| 400 |
+
"grad_norm": 0.00244140625,
|
| 401 |
+
"learning_rate": 2.936628684806084e-05,
|
| 402 |
+
"loss": 0.0001,
|
| 403 |
+
"step": 275
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 1.9047619047619047,
|
| 407 |
+
"grad_norm": 0.00170135498046875,
|
| 408 |
+
"learning_rate": 2.858590121249654e-05,
|
| 409 |
+
"loss": 0.0,
|
| 410 |
+
"step": 280
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.9387755102040818,
|
| 414 |
+
"grad_norm": 0.00762939453125,
|
| 415 |
+
"learning_rate": 2.7815367877467164e-05,
|
| 416 |
+
"loss": 0.0974,
|
| 417 |
+
"step": 285
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.9727891156462585,
|
| 421 |
+
"grad_norm": 0.002166748046875,
|
| 422 |
+
"learning_rate": 2.7055840096905433e-05,
|
| 423 |
+
"loss": 0.0278,
|
| 424 |
+
"step": 290
|
| 425 |
},
|
| 426 |
{
|
| 427 |
+
"epoch": 1.9863945578231292,
|
| 428 |
+
"eval_loss": 0.00010653198114596307,
|
| 429 |
+
"eval_runtime": 0.7953,
|
| 430 |
+
"eval_samples_per_second": 32.69,
|
| 431 |
+
"eval_steps_per_second": 32.69,
|
| 432 |
+
"step": 292
|
| 433 |
}
|
| 434 |
],
|
| 435 |
"logging_steps": 5,
|
| 436 |
+
"max_steps": 441,
|
| 437 |
"num_input_tokens_seen": 0,
|
| 438 |
"num_train_epochs": 3,
|
| 439 |
"save_steps": 500,
|
|
|
|
| 449 |
"attributes": {}
|
| 450 |
}
|
| 451 |
},
|
| 452 |
+
"total_flos": 1.1110365064593408e+16,
|
| 453 |
+
"train_batch_size": 3,
|
| 454 |
"trial_name": null,
|
| 455 |
"trial_params": null
|
| 456 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5560
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:aede922fcb038c2b2f294f933050cf01eb46c9c0de6a69807b9087c912c03c36
|
| 3 |
size 5560
|