Commit
•
98f3990
1
Parent(s):
8b2fd61
End of training
Browse files- README.md +5 -1
- all_results.json +13 -0
- eval_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +652 -0
README.md
CHANGED
|
@@ -3,6 +3,8 @@ license: mit
|
|
| 3 |
base_model: unicamp-dl/ptt5-small-t5-vocab
|
| 4 |
tags:
|
| 5 |
- generated_from_trainer
|
|
|
|
|
|
|
| 6 |
model-index:
|
| 7 |
- name: t5_small-qg-ctx-a
|
| 8 |
results: []
|
|
@@ -13,7 +15,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 13 |
|
| 14 |
# t5_small-qg-ctx-a
|
| 15 |
|
| 16 |
-
This model is a fine-tuned version of [unicamp-dl/ptt5-small-t5-vocab](https://huggingface.co/unicamp-dl/ptt5-small-t5-vocab) on
|
|
|
|
|
|
|
| 17 |
|
| 18 |
## Model description
|
| 19 |
|
|
|
|
| 3 |
base_model: unicamp-dl/ptt5-small-t5-vocab
|
| 4 |
tags:
|
| 5 |
- generated_from_trainer
|
| 6 |
+
datasets:
|
| 7 |
+
- tiagoblima/qg_squad_v1_pt
|
| 8 |
model-index:
|
| 9 |
- name: t5_small-qg-ctx-a
|
| 10 |
results: []
|
|
|
|
| 15 |
|
| 16 |
# t5_small-qg-ctx-a
|
| 17 |
|
| 18 |
+
This model is a fine-tuned version of [unicamp-dl/ptt5-small-t5-vocab](https://huggingface.co/unicamp-dl/ptt5-small-t5-vocab) on the tiagoblima/qg_squad_v1_pt dataset.
|
| 19 |
+
It achieves the following results on the evaluation set:
|
| 20 |
+
- Loss: 1.4559
|
| 21 |
|
| 22 |
## Model description
|
| 23 |
|
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.0,
|
| 3 |
+
"eval_loss": 1.4559422731399536,
|
| 4 |
+
"eval_runtime": 78.5069,
|
| 5 |
+
"eval_samples": 8961,
|
| 6 |
+
"eval_samples_per_second": 114.143,
|
| 7 |
+
"eval_steps_per_second": 14.279,
|
| 8 |
+
"train_loss": 1.3101274637649682,
|
| 9 |
+
"train_runtime": 7625.5065,
|
| 10 |
+
"train_samples": 52135,
|
| 11 |
+
"train_samples_per_second": 13.674,
|
| 12 |
+
"train_steps_per_second": 6.837
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.0,
|
| 3 |
+
"eval_loss": 1.4559422731399536,
|
| 4 |
+
"eval_runtime": 78.5069,
|
| 5 |
+
"eval_samples": 8961,
|
| 6 |
+
"eval_samples_per_second": 114.143,
|
| 7 |
+
"eval_steps_per_second": 14.279
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.0,
|
| 3 |
+
"train_loss": 1.3101274637649682,
|
| 4 |
+
"train_runtime": 7625.5065,
|
| 5 |
+
"train_samples": 52135,
|
| 6 |
+
"train_samples_per_second": 13.674,
|
| 7 |
+
"train_steps_per_second": 6.837
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,652 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.9999616380550493,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 52134,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.02,
|
| 13 |
+
"learning_rate": 9.904093298039668e-05,
|
| 14 |
+
"loss": 2.2133,
|
| 15 |
+
"step": 500
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.04,
|
| 19 |
+
"learning_rate": 9.808186596079335e-05,
|
| 20 |
+
"loss": 1.7686,
|
| 21 |
+
"step": 1000
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.06,
|
| 25 |
+
"learning_rate": 9.712279894119002e-05,
|
| 26 |
+
"loss": 1.6543,
|
| 27 |
+
"step": 1500
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.08,
|
| 31 |
+
"learning_rate": 9.616373192158669e-05,
|
| 32 |
+
"loss": 1.5913,
|
| 33 |
+
"step": 2000
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.1,
|
| 37 |
+
"learning_rate": 9.520466490198336e-05,
|
| 38 |
+
"loss": 1.5555,
|
| 39 |
+
"step": 2500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.12,
|
| 43 |
+
"learning_rate": 9.424559788238001e-05,
|
| 44 |
+
"loss": 1.5263,
|
| 45 |
+
"step": 3000
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.13,
|
| 49 |
+
"learning_rate": 9.32865308627767e-05,
|
| 50 |
+
"loss": 1.541,
|
| 51 |
+
"step": 3500
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.15,
|
| 55 |
+
"learning_rate": 9.232746384317337e-05,
|
| 56 |
+
"loss": 1.502,
|
| 57 |
+
"step": 4000
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.17,
|
| 61 |
+
"learning_rate": 9.136839682357004e-05,
|
| 62 |
+
"loss": 1.5094,
|
| 63 |
+
"step": 4500
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.19,
|
| 67 |
+
"learning_rate": 9.04093298039667e-05,
|
| 68 |
+
"loss": 1.4669,
|
| 69 |
+
"step": 5000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.21,
|
| 73 |
+
"learning_rate": 8.945026278436338e-05,
|
| 74 |
+
"loss": 1.4625,
|
| 75 |
+
"step": 5500
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.23,
|
| 79 |
+
"learning_rate": 8.849119576476005e-05,
|
| 80 |
+
"loss": 1.439,
|
| 81 |
+
"step": 6000
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.25,
|
| 85 |
+
"learning_rate": 8.753212874515672e-05,
|
| 86 |
+
"loss": 1.4443,
|
| 87 |
+
"step": 6500
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.27,
|
| 91 |
+
"learning_rate": 8.657306172555338e-05,
|
| 92 |
+
"loss": 1.4184,
|
| 93 |
+
"step": 7000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.29,
|
| 97 |
+
"learning_rate": 8.561399470595005e-05,
|
| 98 |
+
"loss": 1.4656,
|
| 99 |
+
"step": 7500
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.31,
|
| 103 |
+
"learning_rate": 8.465492768634673e-05,
|
| 104 |
+
"loss": 1.4078,
|
| 105 |
+
"step": 8000
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.33,
|
| 109 |
+
"learning_rate": 8.36958606667434e-05,
|
| 110 |
+
"loss": 1.4266,
|
| 111 |
+
"step": 8500
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.35,
|
| 115 |
+
"learning_rate": 8.273679364714006e-05,
|
| 116 |
+
"loss": 1.4053,
|
| 117 |
+
"step": 9000
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.36,
|
| 121 |
+
"learning_rate": 8.177772662753673e-05,
|
| 122 |
+
"loss": 1.4144,
|
| 123 |
+
"step": 9500
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.38,
|
| 127 |
+
"learning_rate": 8.081865960793342e-05,
|
| 128 |
+
"loss": 1.3931,
|
| 129 |
+
"step": 10000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.4,
|
| 133 |
+
"learning_rate": 7.985959258833007e-05,
|
| 134 |
+
"loss": 1.3731,
|
| 135 |
+
"step": 10500
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.42,
|
| 139 |
+
"learning_rate": 7.890052556872674e-05,
|
| 140 |
+
"loss": 1.3473,
|
| 141 |
+
"step": 11000
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.44,
|
| 145 |
+
"learning_rate": 7.794145854912341e-05,
|
| 146 |
+
"loss": 1.3953,
|
| 147 |
+
"step": 11500
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.46,
|
| 151 |
+
"learning_rate": 7.698239152952008e-05,
|
| 152 |
+
"loss": 1.4022,
|
| 153 |
+
"step": 12000
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.48,
|
| 157 |
+
"learning_rate": 7.602332450991675e-05,
|
| 158 |
+
"loss": 1.3876,
|
| 159 |
+
"step": 12500
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.5,
|
| 163 |
+
"learning_rate": 7.506425749031343e-05,
|
| 164 |
+
"loss": 1.355,
|
| 165 |
+
"step": 13000
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.52,
|
| 169 |
+
"learning_rate": 7.41051904707101e-05,
|
| 170 |
+
"loss": 1.3965,
|
| 171 |
+
"step": 13500
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.54,
|
| 175 |
+
"learning_rate": 7.314612345110677e-05,
|
| 176 |
+
"loss": 1.3699,
|
| 177 |
+
"step": 14000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.56,
|
| 181 |
+
"learning_rate": 7.218705643150344e-05,
|
| 182 |
+
"loss": 1.3564,
|
| 183 |
+
"step": 14500
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.58,
|
| 187 |
+
"learning_rate": 7.122798941190011e-05,
|
| 188 |
+
"loss": 1.3441,
|
| 189 |
+
"step": 15000
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.59,
|
| 193 |
+
"learning_rate": 7.026892239229678e-05,
|
| 194 |
+
"loss": 1.3664,
|
| 195 |
+
"step": 15500
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.61,
|
| 199 |
+
"learning_rate": 6.930985537269345e-05,
|
| 200 |
+
"loss": 1.3171,
|
| 201 |
+
"step": 16000
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.63,
|
| 205 |
+
"learning_rate": 6.835078835309012e-05,
|
| 206 |
+
"loss": 1.3574,
|
| 207 |
+
"step": 16500
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.65,
|
| 211 |
+
"learning_rate": 6.739172133348679e-05,
|
| 212 |
+
"loss": 1.3537,
|
| 213 |
+
"step": 17000
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.67,
|
| 217 |
+
"learning_rate": 6.643265431388346e-05,
|
| 218 |
+
"loss": 1.3226,
|
| 219 |
+
"step": 17500
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.69,
|
| 223 |
+
"learning_rate": 6.547358729428013e-05,
|
| 224 |
+
"loss": 1.3274,
|
| 225 |
+
"step": 18000
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.71,
|
| 229 |
+
"learning_rate": 6.451452027467679e-05,
|
| 230 |
+
"loss": 1.3106,
|
| 231 |
+
"step": 18500
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.73,
|
| 235 |
+
"learning_rate": 6.355545325507347e-05,
|
| 236 |
+
"loss": 1.3084,
|
| 237 |
+
"step": 19000
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.75,
|
| 241 |
+
"learning_rate": 6.259638623547014e-05,
|
| 242 |
+
"loss": 1.3071,
|
| 243 |
+
"step": 19500
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.77,
|
| 247 |
+
"learning_rate": 6.163731921586681e-05,
|
| 248 |
+
"loss": 1.3135,
|
| 249 |
+
"step": 20000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.79,
|
| 253 |
+
"learning_rate": 6.067825219626348e-05,
|
| 254 |
+
"loss": 1.2827,
|
| 255 |
+
"step": 20500
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.81,
|
| 259 |
+
"learning_rate": 5.9719185176660154e-05,
|
| 260 |
+
"loss": 1.305,
|
| 261 |
+
"step": 21000
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.82,
|
| 265 |
+
"learning_rate": 5.876011815705682e-05,
|
| 266 |
+
"loss": 1.3273,
|
| 267 |
+
"step": 21500
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.84,
|
| 271 |
+
"learning_rate": 5.780105113745349e-05,
|
| 272 |
+
"loss": 1.2855,
|
| 273 |
+
"step": 22000
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.86,
|
| 277 |
+
"learning_rate": 5.684198411785015e-05,
|
| 278 |
+
"loss": 1.313,
|
| 279 |
+
"step": 22500
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.88,
|
| 283 |
+
"learning_rate": 5.588291709824682e-05,
|
| 284 |
+
"loss": 1.3025,
|
| 285 |
+
"step": 23000
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.9,
|
| 289 |
+
"learning_rate": 5.49238500786435e-05,
|
| 290 |
+
"loss": 1.3051,
|
| 291 |
+
"step": 23500
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 0.92,
|
| 295 |
+
"learning_rate": 5.396478305904017e-05,
|
| 296 |
+
"loss": 1.354,
|
| 297 |
+
"step": 24000
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.94,
|
| 301 |
+
"learning_rate": 5.3005716039436834e-05,
|
| 302 |
+
"loss": 1.294,
|
| 303 |
+
"step": 24500
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.96,
|
| 307 |
+
"learning_rate": 5.2046649019833505e-05,
|
| 308 |
+
"loss": 1.2994,
|
| 309 |
+
"step": 25000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.98,
|
| 313 |
+
"learning_rate": 5.108758200023018e-05,
|
| 314 |
+
"loss": 1.262,
|
| 315 |
+
"step": 25500
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.0,
|
| 319 |
+
"learning_rate": 5.012851498062685e-05,
|
| 320 |
+
"loss": 1.2515,
|
| 321 |
+
"step": 26000
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 1.02,
|
| 325 |
+
"learning_rate": 4.9169447961023517e-05,
|
| 326 |
+
"loss": 1.2339,
|
| 327 |
+
"step": 26500
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 1.04,
|
| 331 |
+
"learning_rate": 4.8210380941420194e-05,
|
| 332 |
+
"loss": 1.2182,
|
| 333 |
+
"step": 27000
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.05,
|
| 337 |
+
"learning_rate": 4.725131392181686e-05,
|
| 338 |
+
"loss": 1.2314,
|
| 339 |
+
"step": 27500
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.07,
|
| 343 |
+
"learning_rate": 4.629224690221353e-05,
|
| 344 |
+
"loss": 1.245,
|
| 345 |
+
"step": 28000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.09,
|
| 349 |
+
"learning_rate": 4.53331798826102e-05,
|
| 350 |
+
"loss": 1.2328,
|
| 351 |
+
"step": 28500
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 1.11,
|
| 355 |
+
"learning_rate": 4.437411286300687e-05,
|
| 356 |
+
"loss": 1.2263,
|
| 357 |
+
"step": 29000
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 1.13,
|
| 361 |
+
"learning_rate": 4.341504584340354e-05,
|
| 362 |
+
"loss": 1.2093,
|
| 363 |
+
"step": 29500
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 1.15,
|
| 367 |
+
"learning_rate": 4.245597882380021e-05,
|
| 368 |
+
"loss": 1.2192,
|
| 369 |
+
"step": 30000
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 1.17,
|
| 373 |
+
"learning_rate": 4.149691180419688e-05,
|
| 374 |
+
"loss": 1.236,
|
| 375 |
+
"step": 30500
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.19,
|
| 379 |
+
"learning_rate": 4.053784478459355e-05,
|
| 380 |
+
"loss": 1.2455,
|
| 381 |
+
"step": 31000
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.21,
|
| 385 |
+
"learning_rate": 3.957877776499022e-05,
|
| 386 |
+
"loss": 1.2021,
|
| 387 |
+
"step": 31500
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.23,
|
| 391 |
+
"learning_rate": 3.8619710745386886e-05,
|
| 392 |
+
"loss": 1.2401,
|
| 393 |
+
"step": 32000
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 1.25,
|
| 397 |
+
"learning_rate": 3.766064372578356e-05,
|
| 398 |
+
"loss": 1.2256,
|
| 399 |
+
"step": 32500
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.27,
|
| 403 |
+
"learning_rate": 3.670157670618023e-05,
|
| 404 |
+
"loss": 1.2342,
|
| 405 |
+
"step": 33000
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 1.29,
|
| 409 |
+
"learning_rate": 3.57425096865769e-05,
|
| 410 |
+
"loss": 1.2115,
|
| 411 |
+
"step": 33500
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 1.3,
|
| 415 |
+
"learning_rate": 3.478344266697357e-05,
|
| 416 |
+
"loss": 1.2227,
|
| 417 |
+
"step": 34000
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.32,
|
| 421 |
+
"learning_rate": 3.382437564737024e-05,
|
| 422 |
+
"loss": 1.2467,
|
| 423 |
+
"step": 34500
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.34,
|
| 427 |
+
"learning_rate": 3.286530862776691e-05,
|
| 428 |
+
"loss": 1.206,
|
| 429 |
+
"step": 35000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.36,
|
| 433 |
+
"learning_rate": 3.190624160816358e-05,
|
| 434 |
+
"loss": 1.1981,
|
| 435 |
+
"step": 35500
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 1.38,
|
| 439 |
+
"learning_rate": 3.094717458856025e-05,
|
| 440 |
+
"loss": 1.224,
|
| 441 |
+
"step": 36000
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 1.4,
|
| 445 |
+
"learning_rate": 2.998810756895692e-05,
|
| 446 |
+
"loss": 1.2236,
|
| 447 |
+
"step": 36500
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 1.42,
|
| 451 |
+
"learning_rate": 2.902904054935359e-05,
|
| 452 |
+
"loss": 1.2348,
|
| 453 |
+
"step": 37000
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 1.44,
|
| 457 |
+
"learning_rate": 2.806997352975026e-05,
|
| 458 |
+
"loss": 1.2012,
|
| 459 |
+
"step": 37500
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.46,
|
| 463 |
+
"learning_rate": 2.7110906510146932e-05,
|
| 464 |
+
"loss": 1.2015,
|
| 465 |
+
"step": 38000
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.48,
|
| 469 |
+
"learning_rate": 2.61518394905436e-05,
|
| 470 |
+
"loss": 1.2002,
|
| 471 |
+
"step": 38500
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.5,
|
| 475 |
+
"learning_rate": 2.5192772470940267e-05,
|
| 476 |
+
"loss": 1.2298,
|
| 477 |
+
"step": 39000
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 1.52,
|
| 481 |
+
"learning_rate": 2.423370545133694e-05,
|
| 482 |
+
"loss": 1.2269,
|
| 483 |
+
"step": 39500
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 1.53,
|
| 487 |
+
"learning_rate": 2.327463843173361e-05,
|
| 488 |
+
"loss": 1.2181,
|
| 489 |
+
"step": 40000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 1.55,
|
| 493 |
+
"learning_rate": 2.231557141213028e-05,
|
| 494 |
+
"loss": 1.2213,
|
| 495 |
+
"step": 40500
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 1.57,
|
| 499 |
+
"learning_rate": 2.1356504392526952e-05,
|
| 500 |
+
"loss": 1.2233,
|
| 501 |
+
"step": 41000
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 1.59,
|
| 505 |
+
"learning_rate": 2.039743737292362e-05,
|
| 506 |
+
"loss": 1.2186,
|
| 507 |
+
"step": 41500
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.61,
|
| 511 |
+
"learning_rate": 1.943837035332029e-05,
|
| 512 |
+
"loss": 1.1961,
|
| 513 |
+
"step": 42000
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.63,
|
| 517 |
+
"learning_rate": 1.847930333371696e-05,
|
| 518 |
+
"loss": 1.2368,
|
| 519 |
+
"step": 42500
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 1.65,
|
| 523 |
+
"learning_rate": 1.752023631411363e-05,
|
| 524 |
+
"loss": 1.1773,
|
| 525 |
+
"step": 43000
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 1.67,
|
| 529 |
+
"learning_rate": 1.65611692945103e-05,
|
| 530 |
+
"loss": 1.2225,
|
| 531 |
+
"step": 43500
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 1.69,
|
| 535 |
+
"learning_rate": 1.5602102274906972e-05,
|
| 536 |
+
"loss": 1.2317,
|
| 537 |
+
"step": 44000
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 1.71,
|
| 541 |
+
"learning_rate": 1.4643035255303641e-05,
|
| 542 |
+
"loss": 1.1884,
|
| 543 |
+
"step": 44500
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 1.73,
|
| 547 |
+
"learning_rate": 1.3683968235700311e-05,
|
| 548 |
+
"loss": 1.219,
|
| 549 |
+
"step": 45000
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.75,
|
| 553 |
+
"learning_rate": 1.2724901216096982e-05,
|
| 554 |
+
"loss": 1.1947,
|
| 555 |
+
"step": 45500
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.76,
|
| 559 |
+
"learning_rate": 1.176583419649365e-05,
|
| 560 |
+
"loss": 1.1989,
|
| 561 |
+
"step": 46000
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 1.78,
|
| 565 |
+
"learning_rate": 1.0806767176890321e-05,
|
| 566 |
+
"loss": 1.2412,
|
| 567 |
+
"step": 46500
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 1.8,
|
| 571 |
+
"learning_rate": 9.847700157286992e-06,
|
| 572 |
+
"loss": 1.2042,
|
| 573 |
+
"step": 47000
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 1.82,
|
| 577 |
+
"learning_rate": 8.888633137683662e-06,
|
| 578 |
+
"loss": 1.21,
|
| 579 |
+
"step": 47500
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 1.84,
|
| 583 |
+
"learning_rate": 7.929566118080333e-06,
|
| 584 |
+
"loss": 1.2041,
|
| 585 |
+
"step": 48000
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.86,
|
| 589 |
+
"learning_rate": 6.970499098477001e-06,
|
| 590 |
+
"loss": 1.1775,
|
| 591 |
+
"step": 48500
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.88,
|
| 595 |
+
"learning_rate": 6.0114320788736716e-06,
|
| 596 |
+
"loss": 1.1626,
|
| 597 |
+
"step": 49000
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.9,
|
| 601 |
+
"learning_rate": 5.052365059270342e-06,
|
| 602 |
+
"loss": 1.1835,
|
| 603 |
+
"step": 49500
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 1.92,
|
| 607 |
+
"learning_rate": 4.093298039667013e-06,
|
| 608 |
+
"loss": 1.1925,
|
| 609 |
+
"step": 50000
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 1.94,
|
| 613 |
+
"learning_rate": 3.134231020063682e-06,
|
| 614 |
+
"loss": 1.1509,
|
| 615 |
+
"step": 50500
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 1.96,
|
| 619 |
+
"learning_rate": 2.1751640004603525e-06,
|
| 620 |
+
"loss": 1.2008,
|
| 621 |
+
"step": 51000
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 1.98,
|
| 625 |
+
"learning_rate": 1.2160969808570224e-06,
|
| 626 |
+
"loss": 1.1643,
|
| 627 |
+
"step": 51500
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 1.99,
|
| 631 |
+
"learning_rate": 2.5702996125369244e-07,
|
| 632 |
+
"loss": 1.2051,
|
| 633 |
+
"step": 52000
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 2.0,
|
| 637 |
+
"step": 52134,
|
| 638 |
+
"total_flos": 1.0583864216911872e+16,
|
| 639 |
+
"train_loss": 1.3101274637649682,
|
| 640 |
+
"train_runtime": 7625.5065,
|
| 641 |
+
"train_samples_per_second": 13.674,
|
| 642 |
+
"train_steps_per_second": 6.837
|
| 643 |
+
}
|
| 644 |
+
],
|
| 645 |
+
"logging_steps": 500,
|
| 646 |
+
"max_steps": 52134,
|
| 647 |
+
"num_train_epochs": 2,
|
| 648 |
+
"save_steps": 500,
|
| 649 |
+
"total_flos": 1.0583864216911872e+16,
|
| 650 |
+
"trial_name": null,
|
| 651 |
+
"trial_params": null
|
| 652 |
+
}
|