Upload folder using huggingface_hub
Browse files- pytorch_model.bin +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +713 -49
- training_args.bin +1 -1
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1200772613
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ba16c41d31871267dfe01b92e672f8b95982397770b24e22928cf1e04c386b2a
|
| 3 |
size 1200772613
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14575
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:111e3239bd042ea8b1ade1dda895306214751247cef7da6e07148eeb29217a46
|
| 3 |
size 14575
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 627
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d041190d35d52dd601ecb437b71c4265d56df4c38de1bca76e19afcb93a9aef7
|
| 3 |
size 627
|
trainer_state.json
CHANGED
|
@@ -1,147 +1,811 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.05,
|
| 13 |
-
"learning_rate": 4.
|
| 14 |
-
"loss": 19.
|
| 15 |
"step": 500
|
| 16 |
},
|
| 17 |
{
|
| 18 |
"epoch": 0.1,
|
| 19 |
-
"learning_rate": 4.
|
| 20 |
-
"loss": 2.
|
| 21 |
"step": 1000
|
| 22 |
},
|
| 23 |
{
|
| 24 |
"epoch": 0.14,
|
| 25 |
-
"learning_rate": 4.
|
| 26 |
-
"loss":
|
| 27 |
"step": 1500
|
| 28 |
},
|
| 29 |
{
|
| 30 |
"epoch": 0.19,
|
| 31 |
-
"learning_rate": 4.
|
| 32 |
-
"loss": 0.
|
| 33 |
"step": 2000
|
| 34 |
},
|
| 35 |
{
|
| 36 |
"epoch": 0.24,
|
| 37 |
-
"learning_rate":
|
| 38 |
-
"loss": 0.
|
| 39 |
"step": 2500
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.29,
|
| 43 |
-
"learning_rate":
|
| 44 |
-
"loss": 0.
|
| 45 |
"step": 3000
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.34,
|
| 49 |
-
"learning_rate":
|
| 50 |
-
"loss": 0.
|
| 51 |
"step": 3500
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.39,
|
| 55 |
-
"learning_rate":
|
| 56 |
-
"loss": 0.
|
| 57 |
"step": 4000
|
| 58 |
},
|
| 59 |
{
|
| 60 |
"epoch": 0.43,
|
| 61 |
-
"learning_rate":
|
| 62 |
-
"loss": 0.
|
| 63 |
"step": 4500
|
| 64 |
},
|
| 65 |
{
|
| 66 |
"epoch": 0.48,
|
| 67 |
-
"learning_rate":
|
| 68 |
-
"loss": 0.
|
| 69 |
"step": 5000
|
| 70 |
},
|
| 71 |
{
|
| 72 |
"epoch": 0.53,
|
| 73 |
-
"learning_rate":
|
| 74 |
-
"loss": 0.
|
| 75 |
"step": 5500
|
| 76 |
},
|
| 77 |
{
|
| 78 |
"epoch": 0.58,
|
| 79 |
-
"learning_rate":
|
| 80 |
-
"loss": 0.
|
| 81 |
"step": 6000
|
| 82 |
},
|
| 83 |
{
|
| 84 |
"epoch": 0.63,
|
| 85 |
-
"learning_rate":
|
| 86 |
-
"loss": 0.
|
| 87 |
"step": 6500
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.67,
|
| 91 |
-
"learning_rate":
|
| 92 |
-
"loss": 0.
|
| 93 |
"step": 7000
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.72,
|
| 97 |
-
"learning_rate":
|
| 98 |
-
"loss": 0.
|
| 99 |
"step": 7500
|
| 100 |
},
|
| 101 |
{
|
| 102 |
"epoch": 0.77,
|
| 103 |
-
"learning_rate":
|
| 104 |
-
"loss": 0.
|
| 105 |
"step": 8000
|
| 106 |
},
|
| 107 |
{
|
| 108 |
"epoch": 0.82,
|
| 109 |
-
"learning_rate":
|
| 110 |
-
"loss": 0.
|
| 111 |
"step": 8500
|
| 112 |
},
|
| 113 |
{
|
| 114 |
"epoch": 0.87,
|
| 115 |
-
"learning_rate":
|
| 116 |
-
"loss": 0.
|
| 117 |
"step": 9000
|
| 118 |
},
|
| 119 |
{
|
| 120 |
"epoch": 0.92,
|
| 121 |
-
"learning_rate": 4.
|
| 122 |
-
"loss": 0.
|
| 123 |
"step": 9500
|
| 124 |
},
|
| 125 |
{
|
| 126 |
"epoch": 0.96,
|
| 127 |
-
"learning_rate":
|
| 128 |
-
"loss": 0.
|
| 129 |
"step": 10000
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 1.0,
|
| 133 |
-
"eval_loss": 0.
|
| 134 |
-
"eval_runtime": 169.
|
| 135 |
-
"eval_samples_per_second":
|
| 136 |
-
"eval_steps_per_second": 8.
|
| 137 |
"step": 10378
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 138 |
}
|
| 139 |
],
|
| 140 |
"logging_steps": 500,
|
| 141 |
-
"max_steps":
|
| 142 |
-
"num_train_epochs":
|
| 143 |
"save_steps": 500,
|
| 144 |
-
"total_flos":
|
| 145 |
"trial_name": null,
|
| 146 |
"trial_params": null
|
| 147 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 6.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 62268,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.05,
|
| 13 |
+
"learning_rate": 4.95985096678872e-05,
|
| 14 |
+
"loss": 19.0199,
|
| 15 |
"step": 500
|
| 16 |
},
|
| 17 |
{
|
| 18 |
"epoch": 0.1,
|
| 19 |
+
"learning_rate": 4.91970193357744e-05,
|
| 20 |
+
"loss": 2.0644,
|
| 21 |
"step": 1000
|
| 22 |
},
|
| 23 |
{
|
| 24 |
"epoch": 0.14,
|
| 25 |
+
"learning_rate": 4.879552900366159e-05,
|
| 26 |
+
"loss": 0.8825,
|
| 27 |
"step": 1500
|
| 28 |
},
|
| 29 |
{
|
| 30 |
"epoch": 0.19,
|
| 31 |
+
"learning_rate": 4.839403867154879e-05,
|
| 32 |
+
"loss": 0.6416,
|
| 33 |
"step": 2000
|
| 34 |
},
|
| 35 |
{
|
| 36 |
"epoch": 0.24,
|
| 37 |
+
"learning_rate": 4.799254833943599e-05,
|
| 38 |
+
"loss": 0.5382,
|
| 39 |
"step": 2500
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.29,
|
| 43 |
+
"learning_rate": 4.759105800732319e-05,
|
| 44 |
+
"loss": 0.4854,
|
| 45 |
"step": 3000
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.34,
|
| 49 |
+
"learning_rate": 4.7189567675210385e-05,
|
| 50 |
+
"loss": 0.4543,
|
| 51 |
"step": 3500
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.39,
|
| 55 |
+
"learning_rate": 4.678807734309758e-05,
|
| 56 |
+
"loss": 0.4062,
|
| 57 |
"step": 4000
|
| 58 |
},
|
| 59 |
{
|
| 60 |
"epoch": 0.43,
|
| 61 |
+
"learning_rate": 4.6386587010984776e-05,
|
| 62 |
+
"loss": 0.3786,
|
| 63 |
"step": 4500
|
| 64 |
},
|
| 65 |
{
|
| 66 |
"epoch": 0.48,
|
| 67 |
+
"learning_rate": 4.5985096678871974e-05,
|
| 68 |
+
"loss": 0.3586,
|
| 69 |
"step": 5000
|
| 70 |
},
|
| 71 |
{
|
| 72 |
"epoch": 0.53,
|
| 73 |
+
"learning_rate": 4.558360634675917e-05,
|
| 74 |
+
"loss": 0.3445,
|
| 75 |
"step": 5500
|
| 76 |
},
|
| 77 |
{
|
| 78 |
"epoch": 0.58,
|
| 79 |
+
"learning_rate": 4.518211601464637e-05,
|
| 80 |
+
"loss": 0.3206,
|
| 81 |
"step": 6000
|
| 82 |
},
|
| 83 |
{
|
| 84 |
"epoch": 0.63,
|
| 85 |
+
"learning_rate": 4.478062568253357e-05,
|
| 86 |
+
"loss": 0.301,
|
| 87 |
"step": 6500
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.67,
|
| 91 |
+
"learning_rate": 4.437913535042076e-05,
|
| 92 |
+
"loss": 0.2895,
|
| 93 |
"step": 7000
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.72,
|
| 97 |
+
"learning_rate": 4.397764501830796e-05,
|
| 98 |
+
"loss": 0.2801,
|
| 99 |
"step": 7500
|
| 100 |
},
|
| 101 |
{
|
| 102 |
"epoch": 0.77,
|
| 103 |
+
"learning_rate": 4.357615468619516e-05,
|
| 104 |
+
"loss": 0.2765,
|
| 105 |
"step": 8000
|
| 106 |
},
|
| 107 |
{
|
| 108 |
"epoch": 0.82,
|
| 109 |
+
"learning_rate": 4.317466435408236e-05,
|
| 110 |
+
"loss": 0.2644,
|
| 111 |
"step": 8500
|
| 112 |
},
|
| 113 |
{
|
| 114 |
"epoch": 0.87,
|
| 115 |
+
"learning_rate": 4.2773174021969556e-05,
|
| 116 |
+
"loss": 0.2631,
|
| 117 |
"step": 9000
|
| 118 |
},
|
| 119 |
{
|
| 120 |
"epoch": 0.92,
|
| 121 |
+
"learning_rate": 4.237168368985675e-05,
|
| 122 |
+
"loss": 0.2477,
|
| 123 |
"step": 9500
|
| 124 |
},
|
| 125 |
{
|
| 126 |
"epoch": 0.96,
|
| 127 |
+
"learning_rate": 4.1970193357743946e-05,
|
| 128 |
+
"loss": 0.2512,
|
| 129 |
"step": 10000
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 1.0,
|
| 133 |
+
"eval_loss": 0.19760668277740479,
|
| 134 |
+
"eval_runtime": 169.1284,
|
| 135 |
+
"eval_samples_per_second": 66.039,
|
| 136 |
+
"eval_steps_per_second": 8.26,
|
| 137 |
"step": 10378
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.01,
|
| 141 |
+
"learning_rate": 4.1568703025631145e-05,
|
| 142 |
+
"loss": 0.2432,
|
| 143 |
+
"step": 10500
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 1.06,
|
| 147 |
+
"learning_rate": 4.116721269351834e-05,
|
| 148 |
+
"loss": 0.2329,
|
| 149 |
+
"step": 11000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 1.11,
|
| 153 |
+
"learning_rate": 4.076572236140554e-05,
|
| 154 |
+
"loss": 0.2353,
|
| 155 |
+
"step": 11500
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 1.16,
|
| 159 |
+
"learning_rate": 4.0364232029292733e-05,
|
| 160 |
+
"loss": 0.2274,
|
| 161 |
+
"step": 12000
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.2,
|
| 165 |
+
"learning_rate": 3.996274169717993e-05,
|
| 166 |
+
"loss": 0.2215,
|
| 167 |
+
"step": 12500
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 1.25,
|
| 171 |
+
"learning_rate": 3.956125136506713e-05,
|
| 172 |
+
"loss": 0.2187,
|
| 173 |
+
"step": 13000
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 1.3,
|
| 177 |
+
"learning_rate": 3.915976103295433e-05,
|
| 178 |
+
"loss": 0.2151,
|
| 179 |
+
"step": 13500
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 1.35,
|
| 183 |
+
"learning_rate": 3.875827070084153e-05,
|
| 184 |
+
"loss": 0.218,
|
| 185 |
+
"step": 14000
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.4,
|
| 189 |
+
"learning_rate": 3.8356780368728726e-05,
|
| 190 |
+
"loss": 0.2106,
|
| 191 |
+
"step": 14500
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 1.45,
|
| 195 |
+
"learning_rate": 3.795529003661592e-05,
|
| 196 |
+
"loss": 0.2109,
|
| 197 |
+
"step": 15000
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 1.49,
|
| 201 |
+
"learning_rate": 3.7553799704503117e-05,
|
| 202 |
+
"loss": 0.2061,
|
| 203 |
+
"step": 15500
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 1.54,
|
| 207 |
+
"learning_rate": 3.7152309372390315e-05,
|
| 208 |
+
"loss": 0.204,
|
| 209 |
+
"step": 16000
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 1.59,
|
| 213 |
+
"learning_rate": 3.6750819040277514e-05,
|
| 214 |
+
"loss": 0.2047,
|
| 215 |
+
"step": 16500
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 1.64,
|
| 219 |
+
"learning_rate": 3.634932870816471e-05,
|
| 220 |
+
"loss": 0.2028,
|
| 221 |
+
"step": 17000
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 1.69,
|
| 225 |
+
"learning_rate": 3.5947838376051904e-05,
|
| 226 |
+
"loss": 0.1929,
|
| 227 |
+
"step": 17500
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 1.73,
|
| 231 |
+
"learning_rate": 3.55463480439391e-05,
|
| 232 |
+
"loss": 0.1984,
|
| 233 |
+
"step": 18000
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.78,
|
| 237 |
+
"learning_rate": 3.51448577118263e-05,
|
| 238 |
+
"loss": 0.1901,
|
| 239 |
+
"step": 18500
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 1.83,
|
| 243 |
+
"learning_rate": 3.47433673797135e-05,
|
| 244 |
+
"loss": 0.1974,
|
| 245 |
+
"step": 19000
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 1.88,
|
| 249 |
+
"learning_rate": 3.43418770476007e-05,
|
| 250 |
+
"loss": 0.2006,
|
| 251 |
+
"step": 19500
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 1.93,
|
| 255 |
+
"learning_rate": 3.394038671548789e-05,
|
| 256 |
+
"loss": 0.1885,
|
| 257 |
+
"step": 20000
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 1.98,
|
| 261 |
+
"learning_rate": 3.353889638337509e-05,
|
| 262 |
+
"loss": 0.1827,
|
| 263 |
+
"step": 20500
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 2.0,
|
| 267 |
+
"eval_loss": 0.1765761375427246,
|
| 268 |
+
"eval_runtime": 169.0515,
|
| 269 |
+
"eval_samples_per_second": 66.069,
|
| 270 |
+
"eval_steps_per_second": 8.264,
|
| 271 |
+
"step": 20756
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 2.02,
|
| 275 |
+
"learning_rate": 3.313740605126229e-05,
|
| 276 |
+
"loss": 0.1909,
|
| 277 |
+
"step": 21000
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 2.07,
|
| 281 |
+
"learning_rate": 3.2735915719149485e-05,
|
| 282 |
+
"loss": 0.1796,
|
| 283 |
+
"step": 21500
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 2.12,
|
| 287 |
+
"learning_rate": 3.2334425387036684e-05,
|
| 288 |
+
"loss": 0.1785,
|
| 289 |
+
"step": 22000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 2.17,
|
| 293 |
+
"learning_rate": 3.193293505492388e-05,
|
| 294 |
+
"loss": 0.1767,
|
| 295 |
+
"step": 22500
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 2.22,
|
| 299 |
+
"learning_rate": 3.1531444722811074e-05,
|
| 300 |
+
"loss": 0.1764,
|
| 301 |
+
"step": 23000
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 2.26,
|
| 305 |
+
"learning_rate": 3.112995439069827e-05,
|
| 306 |
+
"loss": 0.1777,
|
| 307 |
+
"step": 23500
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 2.31,
|
| 311 |
+
"learning_rate": 3.072846405858547e-05,
|
| 312 |
+
"loss": 0.1722,
|
| 313 |
+
"step": 24000
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 2.36,
|
| 317 |
+
"learning_rate": 3.032697372647267e-05,
|
| 318 |
+
"loss": 0.1759,
|
| 319 |
+
"step": 24500
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 2.41,
|
| 323 |
+
"learning_rate": 2.9925483394359865e-05,
|
| 324 |
+
"loss": 0.172,
|
| 325 |
+
"step": 25000
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 2.46,
|
| 329 |
+
"learning_rate": 2.9523993062247064e-05,
|
| 330 |
+
"loss": 0.172,
|
| 331 |
+
"step": 25500
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 2.51,
|
| 335 |
+
"learning_rate": 2.912250273013426e-05,
|
| 336 |
+
"loss": 0.1763,
|
| 337 |
+
"step": 26000
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 2.55,
|
| 341 |
+
"learning_rate": 2.8721012398021457e-05,
|
| 342 |
+
"loss": 0.1712,
|
| 343 |
+
"step": 26500
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 2.6,
|
| 347 |
+
"learning_rate": 2.8319522065908656e-05,
|
| 348 |
+
"loss": 0.1666,
|
| 349 |
+
"step": 27000
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 2.65,
|
| 353 |
+
"learning_rate": 2.791803173379585e-05,
|
| 354 |
+
"loss": 0.1704,
|
| 355 |
+
"step": 27500
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 2.7,
|
| 359 |
+
"learning_rate": 2.751654140168305e-05,
|
| 360 |
+
"loss": 0.1695,
|
| 361 |
+
"step": 28000
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 2.75,
|
| 365 |
+
"learning_rate": 2.7115051069570248e-05,
|
| 366 |
+
"loss": 0.1643,
|
| 367 |
+
"step": 28500
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 2.79,
|
| 371 |
+
"learning_rate": 2.6713560737457443e-05,
|
| 372 |
+
"loss": 0.1666,
|
| 373 |
+
"step": 29000
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 2.84,
|
| 377 |
+
"learning_rate": 2.6312070405344642e-05,
|
| 378 |
+
"loss": 0.1711,
|
| 379 |
+
"step": 29500
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"epoch": 2.89,
|
| 383 |
+
"learning_rate": 2.5910580073231837e-05,
|
| 384 |
+
"loss": 0.1662,
|
| 385 |
+
"step": 30000
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 2.94,
|
| 389 |
+
"learning_rate": 2.5509089741119035e-05,
|
| 390 |
+
"loss": 0.1663,
|
| 391 |
+
"step": 30500
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 2.99,
|
| 395 |
+
"learning_rate": 2.5107599409006234e-05,
|
| 396 |
+
"loss": 0.1661,
|
| 397 |
+
"step": 31000
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 3.0,
|
| 401 |
+
"eval_loss": 0.1699543297290802,
|
| 402 |
+
"eval_runtime": 168.4223,
|
| 403 |
+
"eval_samples_per_second": 66.315,
|
| 404 |
+
"eval_steps_per_second": 8.295,
|
| 405 |
+
"step": 31134
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 3.04,
|
| 409 |
+
"learning_rate": 2.470610907689343e-05,
|
| 410 |
+
"loss": 0.1641,
|
| 411 |
+
"step": 31500
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 3.08,
|
| 415 |
+
"learning_rate": 2.4304618744780628e-05,
|
| 416 |
+
"loss": 0.1577,
|
| 417 |
+
"step": 32000
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 3.13,
|
| 421 |
+
"learning_rate": 2.3903128412667826e-05,
|
| 422 |
+
"loss": 0.1588,
|
| 423 |
+
"step": 32500
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 3.18,
|
| 427 |
+
"learning_rate": 2.350163808055502e-05,
|
| 428 |
+
"loss": 0.1569,
|
| 429 |
+
"step": 33000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 3.23,
|
| 433 |
+
"learning_rate": 2.310014774844222e-05,
|
| 434 |
+
"loss": 0.1616,
|
| 435 |
+
"step": 33500
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 3.28,
|
| 439 |
+
"learning_rate": 2.2698657416329415e-05,
|
| 440 |
+
"loss": 0.1531,
|
| 441 |
+
"step": 34000
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 3.32,
|
| 445 |
+
"learning_rate": 2.2297167084216614e-05,
|
| 446 |
+
"loss": 0.1642,
|
| 447 |
+
"step": 34500
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 3.37,
|
| 451 |
+
"learning_rate": 2.1895676752103812e-05,
|
| 452 |
+
"loss": 0.1571,
|
| 453 |
+
"step": 35000
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 3.42,
|
| 457 |
+
"learning_rate": 2.1494186419991007e-05,
|
| 458 |
+
"loss": 0.1525,
|
| 459 |
+
"step": 35500
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 3.47,
|
| 463 |
+
"learning_rate": 2.1092696087878206e-05,
|
| 464 |
+
"loss": 0.1509,
|
| 465 |
+
"step": 36000
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 3.52,
|
| 469 |
+
"learning_rate": 2.06912057557654e-05,
|
| 470 |
+
"loss": 0.1537,
|
| 471 |
+
"step": 36500
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 3.57,
|
| 475 |
+
"learning_rate": 2.02897154236526e-05,
|
| 476 |
+
"loss": 0.1554,
|
| 477 |
+
"step": 37000
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 3.61,
|
| 481 |
+
"learning_rate": 1.9888225091539798e-05,
|
| 482 |
+
"loss": 0.1512,
|
| 483 |
+
"step": 37500
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 3.66,
|
| 487 |
+
"learning_rate": 1.9486734759426993e-05,
|
| 488 |
+
"loss": 0.151,
|
| 489 |
+
"step": 38000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 3.71,
|
| 493 |
+
"learning_rate": 1.9085244427314192e-05,
|
| 494 |
+
"loss": 0.1519,
|
| 495 |
+
"step": 38500
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 3.76,
|
| 499 |
+
"learning_rate": 1.868375409520139e-05,
|
| 500 |
+
"loss": 0.1534,
|
| 501 |
+
"step": 39000
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 3.81,
|
| 505 |
+
"learning_rate": 1.8282263763088585e-05,
|
| 506 |
+
"loss": 0.1551,
|
| 507 |
+
"step": 39500
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 3.85,
|
| 511 |
+
"learning_rate": 1.7880773430975784e-05,
|
| 512 |
+
"loss": 0.1506,
|
| 513 |
+
"step": 40000
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 3.9,
|
| 517 |
+
"learning_rate": 1.747928309886298e-05,
|
| 518 |
+
"loss": 0.1479,
|
| 519 |
+
"step": 40500
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 3.95,
|
| 523 |
+
"learning_rate": 1.7077792766750178e-05,
|
| 524 |
+
"loss": 0.1485,
|
| 525 |
+
"step": 41000
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 4.0,
|
| 529 |
+
"learning_rate": 1.6676302434637376e-05,
|
| 530 |
+
"loss": 0.1545,
|
| 531 |
+
"step": 41500
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 4.0,
|
| 535 |
+
"eval_loss": 0.16708502173423767,
|
| 536 |
+
"eval_runtime": 168.3042,
|
| 537 |
+
"eval_samples_per_second": 66.362,
|
| 538 |
+
"eval_steps_per_second": 8.3,
|
| 539 |
+
"step": 41512
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 4.05,
|
| 543 |
+
"learning_rate": 1.627481210252457e-05,
|
| 544 |
+
"loss": 0.1467,
|
| 545 |
+
"step": 42000
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 4.1,
|
| 549 |
+
"learning_rate": 1.587332177041177e-05,
|
| 550 |
+
"loss": 0.1478,
|
| 551 |
+
"step": 42500
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 4.14,
|
| 555 |
+
"learning_rate": 1.547183143829897e-05,
|
| 556 |
+
"loss": 0.1452,
|
| 557 |
+
"step": 43000
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 4.19,
|
| 561 |
+
"learning_rate": 1.5070341106186164e-05,
|
| 562 |
+
"loss": 0.1464,
|
| 563 |
+
"step": 43500
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 4.24,
|
| 567 |
+
"learning_rate": 1.4668850774073362e-05,
|
| 568 |
+
"loss": 0.15,
|
| 569 |
+
"step": 44000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 4.29,
|
| 573 |
+
"learning_rate": 1.4267360441960559e-05,
|
| 574 |
+
"loss": 0.1433,
|
| 575 |
+
"step": 44500
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 4.34,
|
| 579 |
+
"learning_rate": 1.3865870109847756e-05,
|
| 580 |
+
"loss": 0.1431,
|
| 581 |
+
"step": 45000
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 4.38,
|
| 585 |
+
"learning_rate": 1.3464379777734953e-05,
|
| 586 |
+
"loss": 0.1424,
|
| 587 |
+
"step": 45500
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 4.43,
|
| 591 |
+
"learning_rate": 1.3062889445622151e-05,
|
| 592 |
+
"loss": 0.14,
|
| 593 |
+
"step": 46000
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 4.48,
|
| 597 |
+
"learning_rate": 1.2661399113509348e-05,
|
| 598 |
+
"loss": 0.1443,
|
| 599 |
+
"step": 46500
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 4.53,
|
| 603 |
+
"learning_rate": 1.2259908781396545e-05,
|
| 604 |
+
"loss": 0.1461,
|
| 605 |
+
"step": 47000
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 4.58,
|
| 609 |
+
"learning_rate": 1.1858418449283742e-05,
|
| 610 |
+
"loss": 0.1431,
|
| 611 |
+
"step": 47500
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 4.63,
|
| 615 |
+
"learning_rate": 1.145692811717094e-05,
|
| 616 |
+
"loss": 0.1403,
|
| 617 |
+
"step": 48000
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 4.67,
|
| 621 |
+
"learning_rate": 1.1055437785058137e-05,
|
| 622 |
+
"loss": 0.1442,
|
| 623 |
+
"step": 48500
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 4.72,
|
| 627 |
+
"learning_rate": 1.0653947452945334e-05,
|
| 628 |
+
"loss": 0.1468,
|
| 629 |
+
"step": 49000
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 4.77,
|
| 633 |
+
"learning_rate": 1.0252457120832531e-05,
|
| 634 |
+
"loss": 0.1483,
|
| 635 |
+
"step": 49500
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 4.82,
|
| 639 |
+
"learning_rate": 9.85096678871973e-06,
|
| 640 |
+
"loss": 0.1402,
|
| 641 |
+
"step": 50000
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 4.87,
|
| 645 |
+
"learning_rate": 9.449476456606926e-06,
|
| 646 |
+
"loss": 0.1428,
|
| 647 |
+
"step": 50500
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 4.91,
|
| 651 |
+
"learning_rate": 9.047986124494123e-06,
|
| 652 |
+
"loss": 0.1447,
|
| 653 |
+
"step": 51000
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 4.96,
|
| 657 |
+
"learning_rate": 8.64649579238132e-06,
|
| 658 |
+
"loss": 0.1472,
|
| 659 |
+
"step": 51500
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 5.0,
|
| 663 |
+
"eval_loss": 0.16224625706672668,
|
| 664 |
+
"eval_runtime": 168.3955,
|
| 665 |
+
"eval_samples_per_second": 66.326,
|
| 666 |
+
"eval_steps_per_second": 8.296,
|
| 667 |
+
"step": 51890
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 5.01,
|
| 671 |
+
"learning_rate": 8.245005460268517e-06,
|
| 672 |
+
"loss": 0.1418,
|
| 673 |
+
"step": 52000
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 5.06,
|
| 677 |
+
"learning_rate": 7.843515128155714e-06,
|
| 678 |
+
"loss": 0.1358,
|
| 679 |
+
"step": 52500
|
| 680 |
+
},
|
| 681 |
+
{
|
| 682 |
+
"epoch": 5.11,
|
| 683 |
+
"learning_rate": 7.442024796042911e-06,
|
| 684 |
+
"loss": 0.1332,
|
| 685 |
+
"step": 53000
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 5.16,
|
| 689 |
+
"learning_rate": 7.040534463930108e-06,
|
| 690 |
+
"loss": 0.1417,
|
| 691 |
+
"step": 53500
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 5.2,
|
| 695 |
+
"learning_rate": 6.639044131817306e-06,
|
| 696 |
+
"loss": 0.1368,
|
| 697 |
+
"step": 54000
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 5.25,
|
| 701 |
+
"learning_rate": 6.2375537997045036e-06,
|
| 702 |
+
"loss": 0.1422,
|
| 703 |
+
"step": 54500
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 5.3,
|
| 707 |
+
"learning_rate": 5.8360634675917e-06,
|
| 708 |
+
"loss": 0.1329,
|
| 709 |
+
"step": 55000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 5.35,
|
| 713 |
+
"learning_rate": 5.434573135478898e-06,
|
| 714 |
+
"loss": 0.1347,
|
| 715 |
+
"step": 55500
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 5.4,
|
| 719 |
+
"learning_rate": 5.033082803366095e-06,
|
| 720 |
+
"loss": 0.1468,
|
| 721 |
+
"step": 56000
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 5.44,
|
| 725 |
+
"learning_rate": 4.631592471253293e-06,
|
| 726 |
+
"loss": 0.1348,
|
| 727 |
+
"step": 56500
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 5.49,
|
| 731 |
+
"learning_rate": 4.2301021391404895e-06,
|
| 732 |
+
"loss": 0.14,
|
| 733 |
+
"step": 57000
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 5.54,
|
| 737 |
+
"learning_rate": 3.828611807027687e-06,
|
| 738 |
+
"loss": 0.1414,
|
| 739 |
+
"step": 57500
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 5.59,
|
| 743 |
+
"learning_rate": 3.4271214749148845e-06,
|
| 744 |
+
"loss": 0.1415,
|
| 745 |
+
"step": 58000
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 5.64,
|
| 749 |
+
"learning_rate": 3.0256311428020817e-06,
|
| 750 |
+
"loss": 0.1346,
|
| 751 |
+
"step": 58500
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 5.69,
|
| 755 |
+
"learning_rate": 2.624140810689279e-06,
|
| 756 |
+
"loss": 0.1375,
|
| 757 |
+
"step": 59000
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 5.73,
|
| 761 |
+
"learning_rate": 2.2226504785764763e-06,
|
| 762 |
+
"loss": 0.1379,
|
| 763 |
+
"step": 59500
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"epoch": 5.78,
|
| 767 |
+
"learning_rate": 1.8211601464636733e-06,
|
| 768 |
+
"loss": 0.1365,
|
| 769 |
+
"step": 60000
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 5.83,
|
| 773 |
+
"learning_rate": 1.4196698143508704e-06,
|
| 774 |
+
"loss": 0.1462,
|
| 775 |
+
"step": 60500
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 5.88,
|
| 779 |
+
"learning_rate": 1.0181794822380677e-06,
|
| 780 |
+
"loss": 0.1356,
|
| 781 |
+
"step": 61000
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 5.93,
|
| 785 |
+
"learning_rate": 6.166891501252649e-07,
|
| 786 |
+
"loss": 0.137,
|
| 787 |
+
"step": 61500
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 5.97,
|
| 791 |
+
"learning_rate": 2.1519881801246228e-07,
|
| 792 |
+
"loss": 0.1429,
|
| 793 |
+
"step": 62000
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 6.0,
|
| 797 |
+
"eval_loss": 0.1622563898563385,
|
| 798 |
+
"eval_runtime": 168.6575,
|
| 799 |
+
"eval_samples_per_second": 66.223,
|
| 800 |
+
"eval_steps_per_second": 8.283,
|
| 801 |
+
"step": 62268
|
| 802 |
}
|
| 803 |
],
|
| 804 |
"logging_steps": 500,
|
| 805 |
+
"max_steps": 62268,
|
| 806 |
+
"num_train_epochs": 6,
|
| 807 |
"save_steps": 500,
|
| 808 |
+
"total_flos": 2.6337445102485504e+17,
|
| 809 |
"trial_name": null,
|
| 810 |
"trial_params": null
|
| 811 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4155
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2aaf4dacd231a1b34262783141b42c9b905294cc1b5a94a80eba4089f04783bd
|
| 3 |
size 4155
|