Training in progress, step 2000, checkpoint
Browse files
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 228140600
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:875e11864c60557b1ce9d0f4a3628b1921ba20dcfcb047f1194317ca21dd647e
|
| 3 |
size 228140600
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 117931203
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:de1ef9fce3501f8a10d1279e16882931ece02414376645b57e1c3a181bf8a440
|
| 3 |
size 117931203
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b9eb46347e03fd2a32788474d53b64aa40655ea04df926d70dd4416068652168
|
| 3 |
size 14645
|
last-checkpoint/scaler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1383
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:61bb68517c2e5d425f2cd920b30f02d4e60fd1e393f4dd6c263b9f530746bef3
|
| 3 |
size 1383
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9cbbe8c194b3272da66f1fba8ab4ba395d75f317a59ad44137b928cbb13dbc0e
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": 750,
|
| 3 |
"best_metric": 0.5089643597602844,
|
| 4 |
"best_model_checkpoint": "./adapter-phase1/checkpoint-750",
|
| 5 |
-
"epoch": 2
|
| 6 |
"eval_steps": 400,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -1729,6 +1729,417 @@
|
|
| 1729 |
"eval_samples_per_second": 2.039,
|
| 1730 |
"eval_steps_per_second": 0.51,
|
| 1731 |
"step": 1600
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1732 |
}
|
| 1733 |
],
|
| 1734 |
"logging_steps": 10,
|
|
@@ -1748,7 +2159,7 @@
|
|
| 1748 |
"attributes": {}
|
| 1749 |
}
|
| 1750 |
},
|
| 1751 |
-
"total_flos":
|
| 1752 |
"train_batch_size": 1,
|
| 1753 |
"trial_name": null,
|
| 1754 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": 750,
|
| 3 |
"best_metric": 0.5089643597602844,
|
| 4 |
"best_model_checkpoint": "./adapter-phase1/checkpoint-750",
|
| 5 |
+
"epoch": 3.2,
|
| 6 |
"eval_steps": 400,
|
| 7 |
+
"global_step": 2000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 1729 |
"eval_samples_per_second": 2.039,
|
| 1730 |
"eval_steps_per_second": 0.51,
|
| 1731 |
"step": 1600
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"entropy": 0.27413347605615856,
|
| 1735 |
+
"epoch": 2.576,
|
| 1736 |
+
"grad_norm": 0.6262645125389099,
|
| 1737 |
+
"learning_rate": 4.8544e-05,
|
| 1738 |
+
"loss": 0.291,
|
| 1739 |
+
"mean_token_accuracy": 0.909802608937025,
|
| 1740 |
+
"num_tokens": 289137.0,
|
| 1741 |
+
"step": 1610
|
| 1742 |
+
},
|
| 1743 |
+
{
|
| 1744 |
+
"entropy": 0.3372902118600905,
|
| 1745 |
+
"epoch": 2.592,
|
| 1746 |
+
"grad_norm": 0.6019719243049622,
|
| 1747 |
+
"learning_rate": 4.8224000000000004e-05,
|
| 1748 |
+
"loss": 0.3089,
|
| 1749 |
+
"mean_token_accuracy": 0.9065854378044605,
|
| 1750 |
+
"num_tokens": 317789.0,
|
| 1751 |
+
"step": 1620
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"entropy": 0.37745234509930015,
|
| 1755 |
+
"epoch": 2.608,
|
| 1756 |
+
"grad_norm": 0.6852167248725891,
|
| 1757 |
+
"learning_rate": 4.790400000000001e-05,
|
| 1758 |
+
"loss": 0.3237,
|
| 1759 |
+
"mean_token_accuracy": 0.9017773322761059,
|
| 1760 |
+
"num_tokens": 340977.0,
|
| 1761 |
+
"step": 1630
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"entropy": 0.3725322958081961,
|
| 1765 |
+
"epoch": 2.624,
|
| 1766 |
+
"grad_norm": 0.7118895053863525,
|
| 1767 |
+
"learning_rate": 4.7584000000000004e-05,
|
| 1768 |
+
"loss": 0.3207,
|
| 1769 |
+
"mean_token_accuracy": 0.9077424634248018,
|
| 1770 |
+
"num_tokens": 360098.0,
|
| 1771 |
+
"step": 1640
|
| 1772 |
+
},
|
| 1773 |
+
{
|
| 1774 |
+
"entropy": 0.4033573804423213,
|
| 1775 |
+
"epoch": 2.64,
|
| 1776 |
+
"grad_norm": 1.0586738586425781,
|
| 1777 |
+
"learning_rate": 4.7264e-05,
|
| 1778 |
+
"loss": 0.3174,
|
| 1779 |
+
"mean_token_accuracy": 0.9044062152504921,
|
| 1780 |
+
"num_tokens": 373200.0,
|
| 1781 |
+
"step": 1650
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"entropy": 0.2776737127453089,
|
| 1785 |
+
"epoch": 2.656,
|
| 1786 |
+
"grad_norm": 0.6017902493476868,
|
| 1787 |
+
"learning_rate": 4.6944e-05,
|
| 1788 |
+
"loss": 0.2942,
|
| 1789 |
+
"mean_token_accuracy": 0.9093959752470255,
|
| 1790 |
+
"num_tokens": 413938.0,
|
| 1791 |
+
"step": 1660
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"entropy": 0.33967588590458037,
|
| 1795 |
+
"epoch": 2.672,
|
| 1796 |
+
"grad_norm": 0.6162438988685608,
|
| 1797 |
+
"learning_rate": 4.6624e-05,
|
| 1798 |
+
"loss": 0.3075,
|
| 1799 |
+
"mean_token_accuracy": 0.905268831551075,
|
| 1800 |
+
"num_tokens": 442794.0,
|
| 1801 |
+
"step": 1670
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"entropy": 0.37314077839255333,
|
| 1805 |
+
"epoch": 2.6879999999999997,
|
| 1806 |
+
"grad_norm": 0.6455461382865906,
|
| 1807 |
+
"learning_rate": 4.6304e-05,
|
| 1808 |
+
"loss": 0.312,
|
| 1809 |
+
"mean_token_accuracy": 0.9044175367802382,
|
| 1810 |
+
"num_tokens": 465992.0,
|
| 1811 |
+
"step": 1680
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"entropy": 0.3640971322543919,
|
| 1815 |
+
"epoch": 2.7039999999999997,
|
| 1816 |
+
"grad_norm": 0.7681553959846497,
|
| 1817 |
+
"learning_rate": 4.5984000000000006e-05,
|
| 1818 |
+
"loss": 0.3049,
|
| 1819 |
+
"mean_token_accuracy": 0.9096171893179417,
|
| 1820 |
+
"num_tokens": 484580.0,
|
| 1821 |
+
"step": 1690
|
| 1822 |
+
},
|
| 1823 |
+
{
|
| 1824 |
+
"entropy": 0.39063505809754134,
|
| 1825 |
+
"epoch": 2.7199999999999998,
|
| 1826 |
+
"grad_norm": 0.9511684775352478,
|
| 1827 |
+
"learning_rate": 4.5664e-05,
|
| 1828 |
+
"loss": 0.3225,
|
| 1829 |
+
"mean_token_accuracy": 0.9034549340605735,
|
| 1830 |
+
"num_tokens": 497612.0,
|
| 1831 |
+
"step": 1700
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"entropy": 0.2883146867156029,
|
| 1835 |
+
"epoch": 2.7359999999999998,
|
| 1836 |
+
"grad_norm": 0.6692296862602234,
|
| 1837 |
+
"learning_rate": 4.5344000000000005e-05,
|
| 1838 |
+
"loss": 0.2935,
|
| 1839 |
+
"mean_token_accuracy": 0.9078109141439199,
|
| 1840 |
+
"num_tokens": 537755.0,
|
| 1841 |
+
"step": 1710
|
| 1842 |
+
},
|
| 1843 |
+
{
|
| 1844 |
+
"entropy": 0.34244058514013886,
|
| 1845 |
+
"epoch": 2.752,
|
| 1846 |
+
"grad_norm": 0.5983220934867859,
|
| 1847 |
+
"learning_rate": 4.5024e-05,
|
| 1848 |
+
"loss": 0.3076,
|
| 1849 |
+
"mean_token_accuracy": 0.9057810723781585,
|
| 1850 |
+
"num_tokens": 566325.0,
|
| 1851 |
+
"step": 1720
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"entropy": 0.3659200777299702,
|
| 1855 |
+
"epoch": 2.768,
|
| 1856 |
+
"grad_norm": 0.7049655318260193,
|
| 1857 |
+
"learning_rate": 4.4704000000000004e-05,
|
| 1858 |
+
"loss": 0.3059,
|
| 1859 |
+
"mean_token_accuracy": 0.9072589132934809,
|
| 1860 |
+
"num_tokens": 589517.0,
|
| 1861 |
+
"step": 1730
|
| 1862 |
+
},
|
| 1863 |
+
{
|
| 1864 |
+
"entropy": 0.35552563723176717,
|
| 1865 |
+
"epoch": 2.784,
|
| 1866 |
+
"grad_norm": 0.7242270112037659,
|
| 1867 |
+
"learning_rate": 4.4384e-05,
|
| 1868 |
+
"loss": 0.3013,
|
| 1869 |
+
"mean_token_accuracy": 0.912841784581542,
|
| 1870 |
+
"num_tokens": 608224.0,
|
| 1871 |
+
"step": 1740
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"entropy": 0.4027377144433558,
|
| 1875 |
+
"epoch": 2.8,
|
| 1876 |
+
"grad_norm": 1.5430299043655396,
|
| 1877 |
+
"learning_rate": 4.4064e-05,
|
| 1878 |
+
"loss": 0.3223,
|
| 1879 |
+
"mean_token_accuracy": 0.9028574671596289,
|
| 1880 |
+
"num_tokens": 621051.0,
|
| 1881 |
+
"step": 1750
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"entropy": 0.2703737439122051,
|
| 1885 |
+
"epoch": 2.816,
|
| 1886 |
+
"grad_norm": 0.7151817083358765,
|
| 1887 |
+
"learning_rate": 4.3744e-05,
|
| 1888 |
+
"loss": 0.2894,
|
| 1889 |
+
"mean_token_accuracy": 0.9102732315659523,
|
| 1890 |
+
"num_tokens": 662133.0,
|
| 1891 |
+
"step": 1760
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"entropy": 0.32695954395458104,
|
| 1895 |
+
"epoch": 2.832,
|
| 1896 |
+
"grad_norm": 0.6097021698951721,
|
| 1897 |
+
"learning_rate": 4.3424e-05,
|
| 1898 |
+
"loss": 0.2967,
|
| 1899 |
+
"mean_token_accuracy": 0.9080837737768889,
|
| 1900 |
+
"num_tokens": 690682.0,
|
| 1901 |
+
"step": 1770
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"entropy": 0.36010922444984317,
|
| 1905 |
+
"epoch": 2.848,
|
| 1906 |
+
"grad_norm": 0.7698465585708618,
|
| 1907 |
+
"learning_rate": 4.3104e-05,
|
| 1908 |
+
"loss": 0.3064,
|
| 1909 |
+
"mean_token_accuracy": 0.9076121047139167,
|
| 1910 |
+
"num_tokens": 713519.0,
|
| 1911 |
+
"step": 1780
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"entropy": 0.369490017183125,
|
| 1915 |
+
"epoch": 2.864,
|
| 1916 |
+
"grad_norm": 0.997474730014801,
|
| 1917 |
+
"learning_rate": 4.2784e-05,
|
| 1918 |
+
"loss": 0.3153,
|
| 1919 |
+
"mean_token_accuracy": 0.9070124924182892,
|
| 1920 |
+
"num_tokens": 731712.0,
|
| 1921 |
+
"step": 1790
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"entropy": 0.41184745989739896,
|
| 1925 |
+
"epoch": 2.88,
|
| 1926 |
+
"grad_norm": 0.9906476736068726,
|
| 1927 |
+
"learning_rate": 4.2464000000000005e-05,
|
| 1928 |
+
"loss": 0.3325,
|
| 1929 |
+
"mean_token_accuracy": 0.9020481187850237,
|
| 1930 |
+
"num_tokens": 744149.0,
|
| 1931 |
+
"step": 1800
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"entropy": 0.28201086847111584,
|
| 1935 |
+
"epoch": 2.896,
|
| 1936 |
+
"grad_norm": 0.6134458184242249,
|
| 1937 |
+
"learning_rate": 4.2144e-05,
|
| 1938 |
+
"loss": 0.2988,
|
| 1939 |
+
"mean_token_accuracy": 0.9069436389952898,
|
| 1940 |
+
"num_tokens": 782193.0,
|
| 1941 |
+
"step": 1810
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"entropy": 0.33303718706592916,
|
| 1945 |
+
"epoch": 2.912,
|
| 1946 |
+
"grad_norm": 0.6062189936637878,
|
| 1947 |
+
"learning_rate": 4.1824000000000005e-05,
|
| 1948 |
+
"loss": 0.3086,
|
| 1949 |
+
"mean_token_accuracy": 0.9056244477629661,
|
| 1950 |
+
"num_tokens": 809927.0,
|
| 1951 |
+
"step": 1820
|
| 1952 |
+
},
|
| 1953 |
+
{
|
| 1954 |
+
"entropy": 0.3643056120723486,
|
| 1955 |
+
"epoch": 2.928,
|
| 1956 |
+
"grad_norm": 0.6338886618614197,
|
| 1957 |
+
"learning_rate": 4.1504e-05,
|
| 1958 |
+
"loss": 0.3035,
|
| 1959 |
+
"mean_token_accuracy": 0.911867779865861,
|
| 1960 |
+
"num_tokens": 832745.0,
|
| 1961 |
+
"step": 1830
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"entropy": 0.35973973935469983,
|
| 1965 |
+
"epoch": 2.944,
|
| 1966 |
+
"grad_norm": 0.8483228087425232,
|
| 1967 |
+
"learning_rate": 4.1184e-05,
|
| 1968 |
+
"loss": 0.3084,
|
| 1969 |
+
"mean_token_accuracy": 0.9093430683016777,
|
| 1970 |
+
"num_tokens": 851193.0,
|
| 1971 |
+
"step": 1840
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"entropy": 0.4053435407578945,
|
| 1975 |
+
"epoch": 2.96,
|
| 1976 |
+
"grad_norm": 0.9516308903694153,
|
| 1977 |
+
"learning_rate": 4.0864e-05,
|
| 1978 |
+
"loss": 0.332,
|
| 1979 |
+
"mean_token_accuracy": 0.8999160658568144,
|
| 1980 |
+
"num_tokens": 863867.0,
|
| 1981 |
+
"step": 1850
|
| 1982 |
+
},
|
| 1983 |
+
{
|
| 1984 |
+
"entropy": 0.2989065528847277,
|
| 1985 |
+
"epoch": 2.976,
|
| 1986 |
+
"grad_norm": 0.6929520964622498,
|
| 1987 |
+
"learning_rate": 4.0544000000000003e-05,
|
| 1988 |
+
"loss": 0.2943,
|
| 1989 |
+
"mean_token_accuracy": 0.9087879080325365,
|
| 1990 |
+
"num_tokens": 898118.0,
|
| 1991 |
+
"step": 1860
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"entropy": 0.3597102670930326,
|
| 1995 |
+
"epoch": 2.992,
|
| 1996 |
+
"grad_norm": 0.7972533106803894,
|
| 1997 |
+
"learning_rate": 4.0224e-05,
|
| 1998 |
+
"loss": 0.3215,
|
| 1999 |
+
"mean_token_accuracy": 0.902438759058714,
|
| 2000 |
+
"num_tokens": 918026.0,
|
| 2001 |
+
"step": 1870
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"entropy": 0.3693191984202713,
|
| 2005 |
+
"epoch": 3.008,
|
| 2006 |
+
"grad_norm": 0.4952141344547272,
|
| 2007 |
+
"learning_rate": 3.9904e-05,
|
| 2008 |
+
"loss": 0.3109,
|
| 2009 |
+
"mean_token_accuracy": 0.9047053713351488,
|
| 2010 |
+
"num_tokens": 946468.0,
|
| 2011 |
+
"step": 1880
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"entropy": 0.30884325662627815,
|
| 2015 |
+
"epoch": 3.024,
|
| 2016 |
+
"grad_norm": 0.6402750015258789,
|
| 2017 |
+
"learning_rate": 3.9584000000000006e-05,
|
| 2018 |
+
"loss": 0.287,
|
| 2019 |
+
"mean_token_accuracy": 0.9127614002674818,
|
| 2020 |
+
"num_tokens": 978498.0,
|
| 2021 |
+
"step": 1890
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"entropy": 0.3251019007526338,
|
| 2025 |
+
"epoch": 3.04,
|
| 2026 |
+
"grad_norm": 0.7701610326766968,
|
| 2027 |
+
"learning_rate": 3.9264e-05,
|
| 2028 |
+
"loss": 0.3012,
|
| 2029 |
+
"mean_token_accuracy": 0.9117080509662628,
|
| 2030 |
+
"num_tokens": 1004128.0,
|
| 2031 |
+
"step": 1900
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"entropy": 0.3512966329231858,
|
| 2035 |
+
"epoch": 3.056,
|
| 2036 |
+
"grad_norm": 0.934260368347168,
|
| 2037 |
+
"learning_rate": 3.8944000000000005e-05,
|
| 2038 |
+
"loss": 0.2996,
|
| 2039 |
+
"mean_token_accuracy": 0.9139776781201363,
|
| 2040 |
+
"num_tokens": 1025136.0,
|
| 2041 |
+
"step": 1910
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"entropy": 0.36649829614907503,
|
| 2045 |
+
"epoch": 3.072,
|
| 2046 |
+
"grad_norm": 1.147735357284546,
|
| 2047 |
+
"learning_rate": 3.8624e-05,
|
| 2048 |
+
"loss": 0.3172,
|
| 2049 |
+
"mean_token_accuracy": 0.90965236723423,
|
| 2050 |
+
"num_tokens": 1041157.0,
|
| 2051 |
+
"step": 1920
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"entropy": 0.33526935083791615,
|
| 2055 |
+
"epoch": 3.088,
|
| 2056 |
+
"grad_norm": 0.6278552412986755,
|
| 2057 |
+
"learning_rate": 3.8304e-05,
|
| 2058 |
+
"loss": 0.294,
|
| 2059 |
+
"mean_token_accuracy": 0.914416927471757,
|
| 2060 |
+
"num_tokens": 1069401.0,
|
| 2061 |
+
"step": 1930
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"entropy": 0.2916401638649404,
|
| 2065 |
+
"epoch": 3.104,
|
| 2066 |
+
"grad_norm": 0.7106419205665588,
|
| 2067 |
+
"learning_rate": 3.7984e-05,
|
| 2068 |
+
"loss": 0.2833,
|
| 2069 |
+
"mean_token_accuracy": 0.9128728475421667,
|
| 2070 |
+
"num_tokens": 1101705.0,
|
| 2071 |
+
"step": 1940
|
| 2072 |
+
},
|
| 2073 |
+
{
|
| 2074 |
+
"entropy": 0.31783650666475294,
|
| 2075 |
+
"epoch": 3.12,
|
| 2076 |
+
"grad_norm": 0.6372864246368408,
|
| 2077 |
+
"learning_rate": 3.7664e-05,
|
| 2078 |
+
"loss": 0.2808,
|
| 2079 |
+
"mean_token_accuracy": 0.9190873377025127,
|
| 2080 |
+
"num_tokens": 1127173.0,
|
| 2081 |
+
"step": 1950
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"entropy": 0.33883463945239783,
|
| 2085 |
+
"epoch": 3.136,
|
| 2086 |
+
"grad_norm": 0.7593994736671448,
|
| 2087 |
+
"learning_rate": 3.7344e-05,
|
| 2088 |
+
"loss": 0.2932,
|
| 2089 |
+
"mean_token_accuracy": 0.9133320480585099,
|
| 2090 |
+
"num_tokens": 1147878.0,
|
| 2091 |
+
"step": 1960
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"entropy": 0.36267717741429806,
|
| 2095 |
+
"epoch": 3.152,
|
| 2096 |
+
"grad_norm": 0.9578737616539001,
|
| 2097 |
+
"learning_rate": 3.7024e-05,
|
| 2098 |
+
"loss": 0.3018,
|
| 2099 |
+
"mean_token_accuracy": 0.9135202784091234,
|
| 2100 |
+
"num_tokens": 1164084.0,
|
| 2101 |
+
"step": 1970
|
| 2102 |
+
},
|
| 2103 |
+
{
|
| 2104 |
+
"entropy": 0.33903956757858394,
|
| 2105 |
+
"epoch": 3.168,
|
| 2106 |
+
"grad_norm": 0.5553727746009827,
|
| 2107 |
+
"learning_rate": 3.6704e-05,
|
| 2108 |
+
"loss": 0.2962,
|
| 2109 |
+
"mean_token_accuracy": 0.9128197953104973,
|
| 2110 |
+
"num_tokens": 1192486.0,
|
| 2111 |
+
"step": 1980
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"entropy": 0.2897605660371482,
|
| 2115 |
+
"epoch": 3.184,
|
| 2116 |
+
"grad_norm": 0.7067289352416992,
|
| 2117 |
+
"learning_rate": 3.6384e-05,
|
| 2118 |
+
"loss": 0.2867,
|
| 2119 |
+
"mean_token_accuracy": 0.9137052699923516,
|
| 2120 |
+
"num_tokens": 1224540.0,
|
| 2121 |
+
"step": 1990
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"entropy": 0.32448912151157855,
|
| 2125 |
+
"epoch": 3.2,
|
| 2126 |
+
"grad_norm": 0.7603920102119446,
|
| 2127 |
+
"learning_rate": 3.6064000000000006e-05,
|
| 2128 |
+
"loss": 0.2908,
|
| 2129 |
+
"mean_token_accuracy": 0.9150090869516134,
|
| 2130 |
+
"num_tokens": 1249827.0,
|
| 2131 |
+
"step": 2000
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 3.2,
|
| 2135 |
+
"eval_entropy": 0.4150727687478066,
|
| 2136 |
+
"eval_loss": 0.5455561280250549,
|
| 2137 |
+
"eval_mean_token_accuracy": 0.857409807562828,
|
| 2138 |
+
"eval_num_tokens": 1249827.0,
|
| 2139 |
+
"eval_runtime": 982.2461,
|
| 2140 |
+
"eval_samples_per_second": 2.036,
|
| 2141 |
+
"eval_steps_per_second": 0.509,
|
| 2142 |
+
"step": 2000
|
| 2143 |
}
|
| 2144 |
],
|
| 2145 |
"logging_steps": 10,
|
|
|
|
| 2159 |
"attributes": {}
|
| 2160 |
}
|
| 2161 |
},
|
| 2162 |
+
"total_flos": 3.452158742886605e+17,
|
| 2163 |
"train_batch_size": 1,
|
| 2164 |
"trial_name": null,
|
| 2165 |
"trial_params": null
|