Training in progress, step 283, 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 1140880624
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eb4636f56b84a29bc498149f8fa29ab96cd149578b787eaec38b88594a715d2c
|
| 3 |
size 1140880624
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2281891834
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f440c1bf2cce0bb3890a1db85b5dfa6f57d9db3073b20a1ea822a1fd304bdb03
|
| 3 |
size 2281891834
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:64af67540b37ea7895c7f0895b6fd3530a619682b8fb6206680dab892bb34ea0
|
| 3 |
size 14244
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4512787257fe5eeb9de23bd70dafa8e125bcb4548e42b7e1a007385ee165b1c3
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0
|
| 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,
|
|
@@ -1680,6 +1680,314 @@
|
|
| 1680 |
"learning_rate": 1.3598609942150765e-05,
|
| 1681 |
"loss": 2.6368,
|
| 1682 |
"step": 239
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1683 |
}
|
| 1684 |
],
|
| 1685 |
"logging_steps": 1,
|
|
@@ -1694,12 +2002,12 @@
|
|
| 1694 |
"should_evaluate": false,
|
| 1695 |
"should_log": false,
|
| 1696 |
"should_save": true,
|
| 1697 |
-
"should_training_stop":
|
| 1698 |
},
|
| 1699 |
"attributes": {}
|
| 1700 |
}
|
| 1701 |
},
|
| 1702 |
-
"total_flos": 1.
|
| 1703 |
"train_batch_size": 4,
|
| 1704 |
"trial_name": null,
|
| 1705 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 283,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1680 |
"learning_rate": 1.3598609942150765e-05,
|
| 1681 |
"loss": 2.6368,
|
| 1682 |
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.8480565371024735,
|
| 1686 |
+
"grad_norm": 4.957668781280518,
|
| 1687 |
+
"learning_rate": 1.3015556956751669e-05,
|
| 1688 |
+
"loss": 2.4514,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.8515901060070671,
|
| 1693 |
+
"grad_norm": 5.872159004211426,
|
| 1694 |
+
"learning_rate": 1.2443403456474017e-05,
|
| 1695 |
+
"loss": 2.6689,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.8551236749116607,
|
| 1700 |
+
"grad_norm": 5.210415363311768,
|
| 1701 |
+
"learning_rate": 1.1882318057580489e-05,
|
| 1702 |
+
"loss": 2.5006,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.8586572438162544,
|
| 1707 |
+
"grad_norm": 4.810140132904053,
|
| 1708 |
+
"learning_rate": 1.1332466114513512e-05,
|
| 1709 |
+
"loss": 2.193,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 0.8621908127208481,
|
| 1714 |
+
"grad_norm": 5.142510414123535,
|
| 1715 |
+
"learning_rate": 1.0794009671164484e-05,
|
| 1716 |
+
"loss": 2.3454,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 0.8657243816254417,
|
| 1721 |
+
"grad_norm": 5.115345478057861,
|
| 1722 |
+
"learning_rate": 1.0267107413118742e-05,
|
| 1723 |
+
"loss": 2.5682,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 0.8692579505300353,
|
| 1728 |
+
"grad_norm": 5.960536479949951,
|
| 1729 |
+
"learning_rate": 9.751914620890206e-06,
|
| 1730 |
+
"loss": 2.5434,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 0.872791519434629,
|
| 1735 |
+
"grad_norm": 4.932643890380859,
|
| 1736 |
+
"learning_rate": 9.248583124159438e-06,
|
| 1737 |
+
"loss": 2.4929,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 0.8763250883392226,
|
| 1742 |
+
"grad_norm": 5.753042221069336,
|
| 1743 |
+
"learning_rate": 8.757261257028777e-06,
|
| 1744 |
+
"loss": 2.5304,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.8798586572438163,
|
| 1749 |
+
"grad_norm": 6.017014026641846,
|
| 1750 |
+
"learning_rate": 8.278093814307637e-06,
|
| 1751 |
+
"loss": 2.5224,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 0.8833922261484098,
|
| 1756 |
+
"grad_norm": 6.0025434494018555,
|
| 1757 |
+
"learning_rate": 7.81122200884072e-06,
|
| 1758 |
+
"loss": 2.7302,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 0.8869257950530035,
|
| 1763 |
+
"grad_norm": 5.745126247406006,
|
| 1764 |
+
"learning_rate": 7.356783429892023e-06,
|
| 1765 |
+
"loss": 2.328,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 0.8904593639575972,
|
| 1770 |
+
"grad_norm": 6.124932289123535,
|
| 1771 |
+
"learning_rate": 6.9149120025965905e-06,
|
| 1772 |
+
"loss": 2.3842,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.8939929328621908,
|
| 1777 |
+
"grad_norm": 5.819494724273682,
|
| 1778 |
+
"learning_rate": 6.4857379484922375e-06,
|
| 1779 |
+
"loss": 2.5136,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.8975265017667845,
|
| 1784 |
+
"grad_norm": 4.685893535614014,
|
| 1785 |
+
"learning_rate": 6.069387747142591e-06,
|
| 1786 |
+
"loss": 2.2869,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 0.901060070671378,
|
| 1791 |
+
"grad_norm": 7.096498966217041,
|
| 1792 |
+
"learning_rate": 5.665984098862992e-06,
|
| 1793 |
+
"loss": 2.3135,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 0.9045936395759717,
|
| 1798 |
+
"grad_norm": 4.7295732498168945,
|
| 1799 |
+
"learning_rate": 5.275645888560232e-06,
|
| 1800 |
+
"loss": 2.1201,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 0.9081272084805654,
|
| 1805 |
+
"grad_norm": 4.792962551116943,
|
| 1806 |
+
"learning_rate": 4.898488150696467e-06,
|
| 1807 |
+
"loss": 2.4277,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 0.911660777385159,
|
| 1812 |
+
"grad_norm": 5.9998345375061035,
|
| 1813 |
+
"learning_rate": 4.534622035388214e-06,
|
| 1814 |
+
"loss": 2.589,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 0.9151943462897526,
|
| 1819 |
+
"grad_norm": 5.028961181640625,
|
| 1820 |
+
"learning_rate": 4.184154775649768e-06,
|
| 1821 |
+
"loss": 2.3586,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 0.9187279151943463,
|
| 1826 |
+
"grad_norm": 4.820951461791992,
|
| 1827 |
+
"learning_rate": 3.8471896557912e-06,
|
| 1828 |
+
"loss": 2.2727,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.9222614840989399,
|
| 1833 |
+
"grad_norm": 5.404629707336426,
|
| 1834 |
+
"learning_rate": 3.523825980979989e-06,
|
| 1835 |
+
"loss": 2.3514,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 0.9257950530035336,
|
| 1840 |
+
"grad_norm": 4.80362606048584,
|
| 1841 |
+
"learning_rate": 3.2141590479753236e-06,
|
| 1842 |
+
"loss": 2.2188,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.9293286219081273,
|
| 1847 |
+
"grad_norm": 5.06864070892334,
|
| 1848 |
+
"learning_rate": 2.918280117043709e-06,
|
| 1849 |
+
"loss": 2.1602,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 0.9328621908127208,
|
| 1854 |
+
"grad_norm": 5.121250629425049,
|
| 1855 |
+
"learning_rate": 2.636276385064157e-06,
|
| 1856 |
+
"loss": 2.4903,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 0.9363957597173145,
|
| 1861 |
+
"grad_norm": 5.582918167114258,
|
| 1862 |
+
"learning_rate": 2.3682309598308747e-06,
|
| 1863 |
+
"loss": 2.379,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 0.9399293286219081,
|
| 1868 |
+
"grad_norm": 5.544982433319092,
|
| 1869 |
+
"learning_rate": 2.114222835560986e-06,
|
| 1870 |
+
"loss": 2.8578,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 0.9434628975265018,
|
| 1875 |
+
"grad_norm": 4.647549629211426,
|
| 1876 |
+
"learning_rate": 1.8743268696145954e-06,
|
| 1877 |
+
"loss": 2.2086,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 0.9469964664310954,
|
| 1882 |
+
"grad_norm": 4.952742576599121,
|
| 1883 |
+
"learning_rate": 1.6486137604339813e-06,
|
| 1884 |
+
"loss": 2.3761,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 0.950530035335689,
|
| 1889 |
+
"grad_norm": 5.95998477935791,
|
| 1890 |
+
"learning_rate": 1.4371500267084338e-06,
|
| 1891 |
+
"loss": 2.5603,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 0.9540636042402827,
|
| 1896 |
+
"grad_norm": 5.027519226074219,
|
| 1897 |
+
"learning_rate": 1.2399979877708745e-06,
|
| 1898 |
+
"loss": 2.3833,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 0.9575971731448764,
|
| 1903 |
+
"grad_norm": 5.592859745025635,
|
| 1904 |
+
"learning_rate": 1.0572157452321097e-06,
|
| 1905 |
+
"loss": 2.7075,
|
| 1906 |
+
"step": 271
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 0.9611307420494699,
|
| 1910 |
+
"grad_norm": 4.911261081695557,
|
| 1911 |
+
"learning_rate": 8.888571658579703e-07,
|
| 1912 |
+
"loss": 2.2938,
|
| 1913 |
+
"step": 272
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 0.9646643109540636,
|
| 1917 |
+
"grad_norm": 5.232736587524414,
|
| 1918 |
+
"learning_rate": 7.349718656945504e-07,
|
| 1919 |
+
"loss": 2.1393,
|
| 1920 |
+
"step": 273
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 0.9681978798586572,
|
| 1924 |
+
"grad_norm": 5.2621331214904785,
|
| 1925 |
+
"learning_rate": 5.956051954461472e-07,
|
| 1926 |
+
"loss": 2.7253,
|
| 1927 |
+
"step": 274
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 0.9717314487632509,
|
| 1931 |
+
"grad_norm": 4.635681629180908,
|
| 1932 |
+
"learning_rate": 4.7079822711015296e-07,
|
| 1933 |
+
"loss": 2.2413,
|
| 1934 |
+
"step": 275
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 0.9752650176678446,
|
| 1938 |
+
"grad_norm": 5.374542236328125,
|
| 1939 |
+
"learning_rate": 3.605877418729975e-07,
|
| 1940 |
+
"loss": 2.6089,
|
| 1941 |
+
"step": 276
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 0.9787985865724381,
|
| 1945 |
+
"grad_norm": 4.751713275909424,
|
| 1946 |
+
"learning_rate": 2.6500621927054715e-07,
|
| 1947 |
+
"loss": 2.3142,
|
| 1948 |
+
"step": 277
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 0.9823321554770318,
|
| 1952 |
+
"grad_norm": 5.122497081756592,
|
| 1953 |
+
"learning_rate": 1.840818276162226e-07,
|
| 1954 |
+
"loss": 2.3476,
|
| 1955 |
+
"step": 278
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 0.9858657243816255,
|
| 1959 |
+
"grad_norm": 6.561892509460449,
|
| 1960 |
+
"learning_rate": 1.1783841569968367e-07,
|
| 1961 |
+
"loss": 2.6703,
|
| 1962 |
+
"step": 279
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 0.9893992932862191,
|
| 1966 |
+
"grad_norm": 4.851873874664307,
|
| 1967 |
+
"learning_rate": 6.629550575847354e-08,
|
| 1968 |
+
"loss": 2.3439,
|
| 1969 |
+
"step": 280
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 0.9929328621908127,
|
| 1973 |
+
"grad_norm": 5.513582229614258,
|
| 1974 |
+
"learning_rate": 2.946828772473764e-08,
|
| 1975 |
+
"loss": 2.6915,
|
| 1976 |
+
"step": 281
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 0.9964664310954063,
|
| 1980 |
+
"grad_norm": 5.608765602111816,
|
| 1981 |
+
"learning_rate": 7.36761474865455e-09,
|
| 1982 |
+
"loss": 2.5523,
|
| 1983 |
+
"step": 282
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 1.0,
|
| 1987 |
+
"grad_norm": 5.803535461425781,
|
| 1988 |
+
"learning_rate": 0.0,
|
| 1989 |
+
"loss": 2.6483,
|
| 1990 |
+
"step": 283
|
| 1991 |
}
|
| 1992 |
],
|
| 1993 |
"logging_steps": 1,
|
|
|
|
| 2002 |
"should_evaluate": false,
|
| 2003 |
"should_log": false,
|
| 2004 |
"should_save": true,
|
| 2005 |
+
"should_training_stop": true
|
| 2006 |
},
|
| 2007 |
"attributes": {}
|
| 2008 |
}
|
| 2009 |
},
|
| 2010 |
+
"total_flos": 1.7503790819180544e+16,
|
| 2011 |
"train_batch_size": 4,
|
| 2012 |
"trial_name": null,
|
| 2013 |
"trial_params": null
|