End of training
Browse files- README.md +14 -2
- all_results.json +16 -0
- eval_results.json +10 -0
- train_results.json +9 -0
- trainer_state.json +483 -0
README.md
CHANGED
|
@@ -2,11 +2,23 @@
|
|
| 2 |
library_name: transformers
|
| 3 |
tags:
|
| 4 |
- generated_from_trainer
|
|
|
|
|
|
|
| 5 |
metrics:
|
| 6 |
- accuracy
|
| 7 |
model-index:
|
| 8 |
- name: test_seed-42_1e-3
|
| 9 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
---
|
| 11 |
|
| 12 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -14,7 +26,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 14 |
|
| 15 |
# test_seed-42_1e-3
|
| 16 |
|
| 17 |
-
This model was trained from scratch on
|
| 18 |
It achieves the following results on the evaluation set:
|
| 19 |
- Loss: 3.0025
|
| 20 |
- Accuracy: 0.4223
|
|
|
|
| 2 |
library_name: transformers
|
| 3 |
tags:
|
| 4 |
- generated_from_trainer
|
| 5 |
+
datasets:
|
| 6 |
+
- qing-yao/slightly-cleaner-babylm
|
| 7 |
metrics:
|
| 8 |
- accuracy
|
| 9 |
model-index:
|
| 10 |
- name: test_seed-42_1e-3
|
| 11 |
+
results:
|
| 12 |
+
- task:
|
| 13 |
+
name: Causal Language Modeling
|
| 14 |
+
type: text-generation
|
| 15 |
+
dataset:
|
| 16 |
+
name: qing-yao/slightly-cleaner-babylm
|
| 17 |
+
type: qing-yao/slightly-cleaner-babylm
|
| 18 |
+
metrics:
|
| 19 |
+
- name: Accuracy
|
| 20 |
+
type: accuracy
|
| 21 |
+
value: 0.4223382514736395
|
| 22 |
---
|
| 23 |
|
| 24 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 26 |
|
| 27 |
# test_seed-42_1e-3
|
| 28 |
|
| 29 |
+
This model was trained from scratch on the qing-yao/slightly-cleaner-babylm dataset.
|
| 30 |
It achieves the following results on the evaluation set:
|
| 31 |
- Loss: 3.0025
|
| 32 |
- Accuracy: 0.4223
|
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 19.98937410022623,
|
| 3 |
+
"eval_accuracy": 0.4223382514736395,
|
| 4 |
+
"eval_loss": 3.0025322437286377,
|
| 5 |
+
"eval_runtime": 102.5726,
|
| 6 |
+
"eval_samples": 45143,
|
| 7 |
+
"eval_samples_per_second": 440.108,
|
| 8 |
+
"eval_steps_per_second": 6.883,
|
| 9 |
+
"perplexity": 20.136462849313748,
|
| 10 |
+
"total_flos": 1.21900898304e+18,
|
| 11 |
+
"train_loss": 3.21239372584039,
|
| 12 |
+
"train_runtime": 46406.0037,
|
| 13 |
+
"train_samples": 466779,
|
| 14 |
+
"train_samples_per_second": 201.172,
|
| 15 |
+
"train_steps_per_second": 0.786
|
| 16 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 19.98937410022623,
|
| 3 |
+
"eval_accuracy": 0.4223382514736395,
|
| 4 |
+
"eval_loss": 3.0025322437286377,
|
| 5 |
+
"eval_runtime": 102.5726,
|
| 6 |
+
"eval_samples": 45143,
|
| 7 |
+
"eval_samples_per_second": 440.108,
|
| 8 |
+
"eval_steps_per_second": 6.883,
|
| 9 |
+
"perplexity": 20.136462849313748
|
| 10 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 19.98937410022623,
|
| 3 |
+
"total_flos": 1.21900898304e+18,
|
| 4 |
+
"train_loss": 3.21239372584039,
|
| 5 |
+
"train_runtime": 46406.0037,
|
| 6 |
+
"train_samples": 466779,
|
| 7 |
+
"train_samples_per_second": 201.172,
|
| 8 |
+
"train_steps_per_second": 0.786
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,483 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 3.0025322437286377,
|
| 3 |
+
"best_model_checkpoint": "models/test_seed-42_1e-3/checkpoint-36460",
|
| 4 |
+
"epoch": 19.98937410022623,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 36460,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.5484335367107699,
|
| 13 |
+
"grad_norm": 4.604539394378662,
|
| 14 |
+
"learning_rate": 3.125e-05,
|
| 15 |
+
"loss": 6.1409,
|
| 16 |
+
"step": 1000
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 1.0,
|
| 20 |
+
"eval_accuracy": 0.307168896400241,
|
| 21 |
+
"eval_loss": 4.236142635345459,
|
| 22 |
+
"eval_runtime": 102.5735,
|
| 23 |
+
"eval_samples_per_second": 440.104,
|
| 24 |
+
"eval_steps_per_second": 6.883,
|
| 25 |
+
"step": 1824
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 1.0965243024610956,
|
| 29 |
+
"grad_norm": 6.257113933563232,
|
| 30 |
+
"learning_rate": 6.25e-05,
|
| 31 |
+
"loss": 4.4438,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 1.6449578391718653,
|
| 36 |
+
"grad_norm": 5.381688117980957,
|
| 37 |
+
"learning_rate": 9.375e-05,
|
| 38 |
+
"loss": 4.0561,
|
| 39 |
+
"step": 3000
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 2.0,
|
| 43 |
+
"eval_accuracy": 0.34904844865531887,
|
| 44 |
+
"eval_loss": 3.71162748336792,
|
| 45 |
+
"eval_runtime": 102.7226,
|
| 46 |
+
"eval_samples_per_second": 439.465,
|
| 47 |
+
"eval_steps_per_second": 6.873,
|
| 48 |
+
"step": 3648
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 2.193048604922191,
|
| 52 |
+
"grad_norm": 5.227675437927246,
|
| 53 |
+
"learning_rate": 0.000125,
|
| 54 |
+
"loss": 3.7917,
|
| 55 |
+
"step": 4000
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 2.741482141632961,
|
| 59 |
+
"grad_norm": 4.001857757568359,
|
| 60 |
+
"learning_rate": 0.00015625,
|
| 61 |
+
"loss": 3.6138,
|
| 62 |
+
"step": 5000
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 3.0,
|
| 66 |
+
"eval_accuracy": 0.3723173375413121,
|
| 67 |
+
"eval_loss": 3.453321695327759,
|
| 68 |
+
"eval_runtime": 103.2591,
|
| 69 |
+
"eval_samples_per_second": 437.182,
|
| 70 |
+
"eval_steps_per_second": 6.837,
|
| 71 |
+
"step": 5472
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"epoch": 3.2895729073832864,
|
| 75 |
+
"grad_norm": 4.385268688201904,
|
| 76 |
+
"learning_rate": 0.0001875,
|
| 77 |
+
"loss": 3.4747,
|
| 78 |
+
"step": 6000
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 3.8380064440940562,
|
| 82 |
+
"grad_norm": 3.377948522567749,
|
| 83 |
+
"learning_rate": 0.00021875,
|
| 84 |
+
"loss": 3.3863,
|
| 85 |
+
"step": 7000
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 4.0,
|
| 89 |
+
"eval_accuracy": 0.3849880966497314,
|
| 90 |
+
"eval_loss": 3.3194580078125,
|
| 91 |
+
"eval_runtime": 103.8766,
|
| 92 |
+
"eval_samples_per_second": 434.583,
|
| 93 |
+
"eval_steps_per_second": 6.797,
|
| 94 |
+
"step": 7296
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 4.386097209844382,
|
| 98 |
+
"grad_norm": 3.0919594764709473,
|
| 99 |
+
"learning_rate": 0.00025,
|
| 100 |
+
"loss": 3.2982,
|
| 101 |
+
"step": 8000
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 4.934530746555152,
|
| 105 |
+
"grad_norm": 3.083603858947754,
|
| 106 |
+
"learning_rate": 0.00028121875,
|
| 107 |
+
"loss": 3.2557,
|
| 108 |
+
"step": 9000
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 5.0,
|
| 112 |
+
"eval_accuracy": 0.3929661428845069,
|
| 113 |
+
"eval_loss": 3.2404134273529053,
|
| 114 |
+
"eval_runtime": 103.0746,
|
| 115 |
+
"eval_samples_per_second": 437.964,
|
| 116 |
+
"eval_steps_per_second": 6.849,
|
| 117 |
+
"step": 9120
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 5.482621512305477,
|
| 121 |
+
"grad_norm": 2.775636672973633,
|
| 122 |
+
"learning_rate": 0.00031246875000000003,
|
| 123 |
+
"loss": 3.188,
|
| 124 |
+
"step": 10000
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 6.0,
|
| 128 |
+
"eval_accuracy": 0.3972814928421361,
|
| 129 |
+
"eval_loss": 3.1924350261688232,
|
| 130 |
+
"eval_runtime": 103.0708,
|
| 131 |
+
"eval_samples_per_second": 437.981,
|
| 132 |
+
"eval_steps_per_second": 6.85,
|
| 133 |
+
"step": 10944
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 6.030712278055803,
|
| 137 |
+
"grad_norm": 2.7608652114868164,
|
| 138 |
+
"learning_rate": 0.00034368749999999997,
|
| 139 |
+
"loss": 3.1706,
|
| 140 |
+
"step": 11000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 6.579145814766573,
|
| 144 |
+
"grad_norm": 2.680748462677002,
|
| 145 |
+
"learning_rate": 0.00037490625,
|
| 146 |
+
"loss": 3.1185,
|
| 147 |
+
"step": 12000
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 7.0,
|
| 151 |
+
"eval_accuracy": 0.40085627676407826,
|
| 152 |
+
"eval_loss": 3.15944504737854,
|
| 153 |
+
"eval_runtime": 102.6889,
|
| 154 |
+
"eval_samples_per_second": 439.609,
|
| 155 |
+
"eval_steps_per_second": 6.875,
|
| 156 |
+
"step": 12768
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 7.1272365805168985,
|
| 160 |
+
"grad_norm": 2.6895594596862793,
|
| 161 |
+
"learning_rate": 0.00040615625,
|
| 162 |
+
"loss": 3.1032,
|
| 163 |
+
"step": 13000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 7.675670117227669,
|
| 167 |
+
"grad_norm": 2.445477247238159,
|
| 168 |
+
"learning_rate": 0.00043740625,
|
| 169 |
+
"loss": 3.0753,
|
| 170 |
+
"step": 14000
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 8.0,
|
| 174 |
+
"eval_accuracy": 0.40288859845380237,
|
| 175 |
+
"eval_loss": 3.1403636932373047,
|
| 176 |
+
"eval_runtime": 102.725,
|
| 177 |
+
"eval_samples_per_second": 439.455,
|
| 178 |
+
"eval_steps_per_second": 6.873,
|
| 179 |
+
"step": 14592
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 8.223760882977993,
|
| 183 |
+
"grad_norm": 2.3262434005737305,
|
| 184 |
+
"learning_rate": 0.00046865625,
|
| 185 |
+
"loss": 3.0553,
|
| 186 |
+
"step": 15000
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 8.772194419688764,
|
| 190 |
+
"grad_norm": 2.2922704219818115,
|
| 191 |
+
"learning_rate": 0.00049990625,
|
| 192 |
+
"loss": 3.047,
|
| 193 |
+
"step": 16000
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 9.0,
|
| 197 |
+
"eval_accuracy": 0.40464988600495244,
|
| 198 |
+
"eval_loss": 3.1230103969573975,
|
| 199 |
+
"eval_runtime": 102.505,
|
| 200 |
+
"eval_samples_per_second": 440.398,
|
| 201 |
+
"eval_steps_per_second": 6.887,
|
| 202 |
+
"step": 16416
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 9.32028518543909,
|
| 206 |
+
"grad_norm": 2.122209310531616,
|
| 207 |
+
"learning_rate": 0.000531125,
|
| 208 |
+
"loss": 3.0169,
|
| 209 |
+
"step": 17000
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 9.86871872214986,
|
| 213 |
+
"grad_norm": 2.0698060989379883,
|
| 214 |
+
"learning_rate": 0.0005623749999999999,
|
| 215 |
+
"loss": 3.0232,
|
| 216 |
+
"step": 18000
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 10.0,
|
| 220 |
+
"eval_accuracy": 0.4059791694627921,
|
| 221 |
+
"eval_loss": 3.112029790878296,
|
| 222 |
+
"eval_runtime": 102.491,
|
| 223 |
+
"eval_samples_per_second": 440.458,
|
| 224 |
+
"eval_steps_per_second": 6.888,
|
| 225 |
+
"step": 18240
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 10.416809487900185,
|
| 229 |
+
"grad_norm": 2.0590832233428955,
|
| 230 |
+
"learning_rate": 0.000593625,
|
| 231 |
+
"loss": 2.9872,
|
| 232 |
+
"step": 19000
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 10.965243024610954,
|
| 236 |
+
"grad_norm": 1.9800970554351807,
|
| 237 |
+
"learning_rate": 0.0006248437500000001,
|
| 238 |
+
"loss": 3.008,
|
| 239 |
+
"step": 20000
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 11.0,
|
| 243 |
+
"eval_accuracy": 0.4073983632839087,
|
| 244 |
+
"eval_loss": 3.1057026386260986,
|
| 245 |
+
"eval_runtime": 102.4282,
|
| 246 |
+
"eval_samples_per_second": 440.728,
|
| 247 |
+
"eval_steps_per_second": 6.893,
|
| 248 |
+
"step": 20064
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 11.51333379036128,
|
| 252 |
+
"grad_norm": 1.9862931966781616,
|
| 253 |
+
"learning_rate": 0.00065609375,
|
| 254 |
+
"loss": 2.9609,
|
| 255 |
+
"step": 21000
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 12.0,
|
| 259 |
+
"eval_accuracy": 0.4079114170090427,
|
| 260 |
+
"eval_loss": 3.0999560356140137,
|
| 261 |
+
"eval_runtime": 102.6416,
|
| 262 |
+
"eval_samples_per_second": 439.812,
|
| 263 |
+
"eval_steps_per_second": 6.878,
|
| 264 |
+
"step": 21888
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 12.061424556111605,
|
| 268 |
+
"grad_norm": 1.9215582609176636,
|
| 269 |
+
"learning_rate": 0.0006873125,
|
| 270 |
+
"loss": 2.9846,
|
| 271 |
+
"step": 22000
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 12.609858092822376,
|
| 275 |
+
"grad_norm": 1.8263485431671143,
|
| 276 |
+
"learning_rate": 0.0007185625,
|
| 277 |
+
"loss": 2.954,
|
| 278 |
+
"step": 23000
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 13.0,
|
| 282 |
+
"eval_accuracy": 0.40866657719065297,
|
| 283 |
+
"eval_loss": 3.092158079147339,
|
| 284 |
+
"eval_runtime": 102.7763,
|
| 285 |
+
"eval_samples_per_second": 439.236,
|
| 286 |
+
"eval_steps_per_second": 6.869,
|
| 287 |
+
"step": 23712
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 13.157948858572702,
|
| 291 |
+
"grad_norm": 1.9183614253997803,
|
| 292 |
+
"learning_rate": 0.00074978125,
|
| 293 |
+
"loss": 2.9608,
|
| 294 |
+
"step": 24000
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 13.706382395283471,
|
| 298 |
+
"grad_norm": 1.7683203220367432,
|
| 299 |
+
"learning_rate": 0.0007810312499999999,
|
| 300 |
+
"loss": 2.953,
|
| 301 |
+
"step": 25000
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 14.0,
|
| 305 |
+
"eval_accuracy": 0.4089403042966295,
|
| 306 |
+
"eval_loss": 3.0896799564361572,
|
| 307 |
+
"eval_runtime": 102.684,
|
| 308 |
+
"eval_samples_per_second": 439.63,
|
| 309 |
+
"eval_steps_per_second": 6.875,
|
| 310 |
+
"step": 25536
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 14.254473161033797,
|
| 314 |
+
"grad_norm": 1.7568416595458984,
|
| 315 |
+
"learning_rate": 0.00081228125,
|
| 316 |
+
"loss": 2.9428,
|
| 317 |
+
"step": 26000
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 14.802906697744568,
|
| 321 |
+
"grad_norm": 1.6089884042739868,
|
| 322 |
+
"learning_rate": 0.00084353125,
|
| 323 |
+
"loss": 2.952,
|
| 324 |
+
"step": 27000
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 15.0,
|
| 328 |
+
"eval_accuracy": 0.40930029322940215,
|
| 329 |
+
"eval_loss": 3.087488889694214,
|
| 330 |
+
"eval_runtime": 102.8286,
|
| 331 |
+
"eval_samples_per_second": 439.012,
|
| 332 |
+
"eval_steps_per_second": 6.866,
|
| 333 |
+
"step": 27360
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 15.350997463494892,
|
| 337 |
+
"grad_norm": 1.7014449834823608,
|
| 338 |
+
"learning_rate": 0.0008747500000000001,
|
| 339 |
+
"loss": 2.9285,
|
| 340 |
+
"step": 28000
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 15.899431000205663,
|
| 344 |
+
"grad_norm": 1.6123307943344116,
|
| 345 |
+
"learning_rate": 0.000906,
|
| 346 |
+
"loss": 2.9527,
|
| 347 |
+
"step": 29000
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 16.0,
|
| 351 |
+
"eval_accuracy": 0.409807874149815,
|
| 352 |
+
"eval_loss": 3.0875635147094727,
|
| 353 |
+
"eval_runtime": 102.7641,
|
| 354 |
+
"eval_samples_per_second": 439.288,
|
| 355 |
+
"eval_steps_per_second": 6.87,
|
| 356 |
+
"step": 29184
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 16.447521765955987,
|
| 360 |
+
"grad_norm": 1.6590194702148438,
|
| 361 |
+
"learning_rate": 0.00093725,
|
| 362 |
+
"loss": 2.9172,
|
| 363 |
+
"step": 30000
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 16.995955302666758,
|
| 367 |
+
"grad_norm": 1.7436003684997559,
|
| 368 |
+
"learning_rate": 0.00096846875,
|
| 369 |
+
"loss": 2.9549,
|
| 370 |
+
"step": 31000
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 17.0,
|
| 374 |
+
"eval_accuracy": 0.40985817183130036,
|
| 375 |
+
"eval_loss": 3.0855603218078613,
|
| 376 |
+
"eval_runtime": 102.7791,
|
| 377 |
+
"eval_samples_per_second": 439.224,
|
| 378 |
+
"eval_steps_per_second": 6.869,
|
| 379 |
+
"step": 31008
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"epoch": 17.544046068417085,
|
| 383 |
+
"grad_norm": 1.626744031906128,
|
| 384 |
+
"learning_rate": 0.00099971875,
|
| 385 |
+
"loss": 2.9073,
|
| 386 |
+
"step": 32000
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 18.0,
|
| 390 |
+
"eval_accuracy": 0.4126600741087255,
|
| 391 |
+
"eval_loss": 3.0625429153442383,
|
| 392 |
+
"eval_runtime": 102.7386,
|
| 393 |
+
"eval_samples_per_second": 439.397,
|
| 394 |
+
"eval_steps_per_second": 6.872,
|
| 395 |
+
"step": 32832
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 18.09213683416741,
|
| 399 |
+
"grad_norm": 1.553916573524475,
|
| 400 |
+
"learning_rate": 0.0007780269058295965,
|
| 401 |
+
"loss": 2.9208,
|
| 402 |
+
"step": 33000
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 18.64057037087818,
|
| 406 |
+
"grad_norm": 1.514714002609253,
|
| 407 |
+
"learning_rate": 0.000554035874439462,
|
| 408 |
+
"loss": 2.8458,
|
| 409 |
+
"step": 34000
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 19.0,
|
| 413 |
+
"eval_accuracy": 0.4182678746797215,
|
| 414 |
+
"eval_loss": 3.021615743637085,
|
| 415 |
+
"eval_runtime": 102.6485,
|
| 416 |
+
"eval_samples_per_second": 439.783,
|
| 417 |
+
"eval_steps_per_second": 6.878,
|
| 418 |
+
"step": 34656
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 19.188661136628504,
|
| 422 |
+
"grad_norm": 1.5944099426269531,
|
| 423 |
+
"learning_rate": 0.00032982062780269057,
|
| 424 |
+
"loss": 2.7882,
|
| 425 |
+
"step": 35000
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 19.737094673339275,
|
| 429 |
+
"grad_norm": 1.5419167280197144,
|
| 430 |
+
"learning_rate": 0.00010560538116591928,
|
| 431 |
+
"loss": 2.7073,
|
| 432 |
+
"step": 36000
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 19.98937410022623,
|
| 436 |
+
"eval_accuracy": 0.4223382514736395,
|
| 437 |
+
"eval_loss": 3.0025322437286377,
|
| 438 |
+
"eval_runtime": 102.7159,
|
| 439 |
+
"eval_samples_per_second": 439.494,
|
| 440 |
+
"eval_steps_per_second": 6.873,
|
| 441 |
+
"step": 36460
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 19.98937410022623,
|
| 445 |
+
"step": 36460,
|
| 446 |
+
"total_flos": 1.21900898304e+18,
|
| 447 |
+
"train_loss": 3.21239372584039,
|
| 448 |
+
"train_runtime": 46406.0037,
|
| 449 |
+
"train_samples_per_second": 201.172,
|
| 450 |
+
"train_steps_per_second": 0.786
|
| 451 |
+
}
|
| 452 |
+
],
|
| 453 |
+
"logging_steps": 1000,
|
| 454 |
+
"max_steps": 36460,
|
| 455 |
+
"num_input_tokens_seen": 0,
|
| 456 |
+
"num_train_epochs": 20,
|
| 457 |
+
"save_steps": 500,
|
| 458 |
+
"stateful_callbacks": {
|
| 459 |
+
"EarlyStoppingCallback": {
|
| 460 |
+
"args": {
|
| 461 |
+
"early_stopping_patience": 3,
|
| 462 |
+
"early_stopping_threshold": 0.0
|
| 463 |
+
},
|
| 464 |
+
"attributes": {
|
| 465 |
+
"early_stopping_patience_counter": 0
|
| 466 |
+
}
|
| 467 |
+
},
|
| 468 |
+
"TrainerControl": {
|
| 469 |
+
"args": {
|
| 470 |
+
"should_epoch_stop": false,
|
| 471 |
+
"should_evaluate": false,
|
| 472 |
+
"should_log": false,
|
| 473 |
+
"should_save": true,
|
| 474 |
+
"should_training_stop": true
|
| 475 |
+
},
|
| 476 |
+
"attributes": {}
|
| 477 |
+
}
|
| 478 |
+
},
|
| 479 |
+
"total_flos": 1.21900898304e+18,
|
| 480 |
+
"train_batch_size": 32,
|
| 481 |
+
"trial_name": null,
|
| 482 |
+
"trial_params": null
|
| 483 |
+
}
|