Training in progress, step 1500, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 847599616
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f6fe29fb072a551b80f556bf6f5e92ff2fb758c7cec9dff1e7b56d77aefdeed6
|
| 3 |
size 847599616
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 350603
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e1a9b7e2b18fd4689820a9812adc3deb3d2448ef0f2839825807a9049cf28400
|
| 3 |
size 350603
|
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:f0baeedffba25b389e46cf6c0e0d8ff516db209714f8118ad9c8ffe8798c13fb
|
| 3 |
size 14645
|
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 sha256:6e5541df11272ae1a1e9106b759745ecaab8d8edbaeecba4b2c0fe5ad109221d
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 150,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -1760,6 +1760,444 @@
|
|
| 1760 |
"eval_samples_per_second": 9.553,
|
| 1761 |
"eval_steps_per_second": 1.624,
|
| 1762 |
"step": 1200
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1763 |
}
|
| 1764 |
],
|
| 1765 |
"logging_steps": 5,
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.3,
|
| 6 |
"eval_steps": 150,
|
| 7 |
+
"global_step": 1500,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 1760 |
"eval_samples_per_second": 9.553,
|
| 1761 |
"eval_steps_per_second": 1.624,
|
| 1762 |
"step": 1200
|
| 1763 |
+
},
|
| 1764 |
+
{
|
| 1765 |
+
"epoch": 0.241,
|
| 1766 |
+
"grad_norm": 0.06398201733827591,
|
| 1767 |
+
"learning_rate": 3.6149182975598944e-06,
|
| 1768 |
+
"loss": 6.538330841064453,
|
| 1769 |
+
"step": 1205
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 0.242,
|
| 1773 |
+
"grad_norm": 0.10430450737476349,
|
| 1774 |
+
"learning_rate": 3.6110077916911736e-06,
|
| 1775 |
+
"loss": 6.606370544433593,
|
| 1776 |
+
"step": 1210
|
| 1777 |
+
},
|
| 1778 |
+
{
|
| 1779 |
+
"epoch": 0.243,
|
| 1780 |
+
"grad_norm": 0.06399216502904892,
|
| 1781 |
+
"learning_rate": 3.6070796680993503e-06,
|
| 1782 |
+
"loss": 6.5816490173339846,
|
| 1783 |
+
"step": 1215
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"epoch": 0.244,
|
| 1787 |
+
"grad_norm": 0.08809840679168701,
|
| 1788 |
+
"learning_rate": 3.6031339697417533e-06,
|
| 1789 |
+
"loss": 6.556430053710938,
|
| 1790 |
+
"step": 1220
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 0.245,
|
| 1794 |
+
"grad_norm": 0.0840051919221878,
|
| 1795 |
+
"learning_rate": 3.599170739767907e-06,
|
| 1796 |
+
"loss": 6.665988159179688,
|
| 1797 |
+
"step": 1225
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"epoch": 0.246,
|
| 1801 |
+
"grad_norm": 0.08308332413434982,
|
| 1802 |
+
"learning_rate": 3.5951900215190614e-06,
|
| 1803 |
+
"loss": 6.544137573242187,
|
| 1804 |
+
"step": 1230
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 0.247,
|
| 1808 |
+
"grad_norm": 0.0652918890118599,
|
| 1809 |
+
"learning_rate": 3.5911918585277125e-06,
|
| 1810 |
+
"loss": 6.496437835693359,
|
| 1811 |
+
"step": 1235
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"epoch": 0.248,
|
| 1815 |
+
"grad_norm": 0.07509986311197281,
|
| 1816 |
+
"learning_rate": 3.58717629451713e-06,
|
| 1817 |
+
"loss": 6.490621185302734,
|
| 1818 |
+
"step": 1240
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 0.249,
|
| 1822 |
+
"grad_norm": 0.09300875663757324,
|
| 1823 |
+
"learning_rate": 3.583143373400879e-06,
|
| 1824 |
+
"loss": 6.5218017578125,
|
| 1825 |
+
"step": 1245
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 0.25,
|
| 1829 |
+
"grad_norm": 0.09340658783912659,
|
| 1830 |
+
"learning_rate": 3.5790931392823376e-06,
|
| 1831 |
+
"loss": 6.703663635253906,
|
| 1832 |
+
"step": 1250
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 0.251,
|
| 1836 |
+
"grad_norm": 0.09822723269462585,
|
| 1837 |
+
"learning_rate": 3.5750256364542196e-06,
|
| 1838 |
+
"loss": 6.545758819580078,
|
| 1839 |
+
"step": 1255
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 0.252,
|
| 1843 |
+
"grad_norm": 0.07032929360866547,
|
| 1844 |
+
"learning_rate": 3.5709409093980814e-06,
|
| 1845 |
+
"loss": 6.586306762695313,
|
| 1846 |
+
"step": 1260
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"epoch": 0.253,
|
| 1850 |
+
"grad_norm": 0.100587859749794,
|
| 1851 |
+
"learning_rate": 3.5668390027838457e-06,
|
| 1852 |
+
"loss": 6.500570678710938,
|
| 1853 |
+
"step": 1265
|
| 1854 |
+
},
|
| 1855 |
+
{
|
| 1856 |
+
"epoch": 0.254,
|
| 1857 |
+
"grad_norm": 0.08588574081659317,
|
| 1858 |
+
"learning_rate": 3.5627199614693055e-06,
|
| 1859 |
+
"loss": 6.516835784912109,
|
| 1860 |
+
"step": 1270
|
| 1861 |
+
},
|
| 1862 |
+
{
|
| 1863 |
+
"epoch": 0.255,
|
| 1864 |
+
"grad_norm": 0.05896945297718048,
|
| 1865 |
+
"learning_rate": 3.5585838304996356e-06,
|
| 1866 |
+
"loss": 6.462747192382812,
|
| 1867 |
+
"step": 1275
|
| 1868 |
+
},
|
| 1869 |
+
{
|
| 1870 |
+
"epoch": 0.256,
|
| 1871 |
+
"grad_norm": 0.07021471858024597,
|
| 1872 |
+
"learning_rate": 3.5544306551069037e-06,
|
| 1873 |
+
"loss": 6.572966003417969,
|
| 1874 |
+
"step": 1280
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 0.257,
|
| 1878 |
+
"grad_norm": 0.09236051887273788,
|
| 1879 |
+
"learning_rate": 3.5502604807095687e-06,
|
| 1880 |
+
"loss": 6.615792083740234,
|
| 1881 |
+
"step": 1285
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"epoch": 0.258,
|
| 1885 |
+
"grad_norm": 0.07293561100959778,
|
| 1886 |
+
"learning_rate": 3.5460733529119908e-06,
|
| 1887 |
+
"loss": 6.55091552734375,
|
| 1888 |
+
"step": 1290
|
| 1889 |
+
},
|
| 1890 |
+
{
|
| 1891 |
+
"epoch": 0.259,
|
| 1892 |
+
"grad_norm": 0.0651378184556961,
|
| 1893 |
+
"learning_rate": 3.541869317503928e-06,
|
| 1894 |
+
"loss": 6.545086669921875,
|
| 1895 |
+
"step": 1295
|
| 1896 |
+
},
|
| 1897 |
+
{
|
| 1898 |
+
"epoch": 0.26,
|
| 1899 |
+
"grad_norm": 0.0653856098651886,
|
| 1900 |
+
"learning_rate": 3.537648420460038e-06,
|
| 1901 |
+
"loss": 6.5654754638671875,
|
| 1902 |
+
"step": 1300
|
| 1903 |
+
},
|
| 1904 |
+
{
|
| 1905 |
+
"epoch": 0.261,
|
| 1906 |
+
"grad_norm": 0.08197472244501114,
|
| 1907 |
+
"learning_rate": 3.5334107079393755e-06,
|
| 1908 |
+
"loss": 6.614053344726562,
|
| 1909 |
+
"step": 1305
|
| 1910 |
+
},
|
| 1911 |
+
{
|
| 1912 |
+
"epoch": 0.262,
|
| 1913 |
+
"grad_norm": 0.07018602639436722,
|
| 1914 |
+
"learning_rate": 3.5291562262848845e-06,
|
| 1915 |
+
"loss": 6.489773559570312,
|
| 1916 |
+
"step": 1310
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 0.263,
|
| 1920 |
+
"grad_norm": 0.07549969106912613,
|
| 1921 |
+
"learning_rate": 3.5248850220228957e-06,
|
| 1922 |
+
"loss": 6.4777587890625,
|
| 1923 |
+
"step": 1315
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 0.264,
|
| 1927 |
+
"grad_norm": 0.07509016990661621,
|
| 1928 |
+
"learning_rate": 3.5205971418626145e-06,
|
| 1929 |
+
"loss": 6.554988861083984,
|
| 1930 |
+
"step": 1320
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 0.265,
|
| 1934 |
+
"grad_norm": 0.07180987298488617,
|
| 1935 |
+
"learning_rate": 3.51629263269561e-06,
|
| 1936 |
+
"loss": 6.668986511230469,
|
| 1937 |
+
"step": 1325
|
| 1938 |
+
},
|
| 1939 |
+
{
|
| 1940 |
+
"epoch": 0.266,
|
| 1941 |
+
"grad_norm": 0.07208196818828583,
|
| 1942 |
+
"learning_rate": 3.511971541595307e-06,
|
| 1943 |
+
"loss": 6.480735778808594,
|
| 1944 |
+
"step": 1330
|
| 1945 |
+
},
|
| 1946 |
+
{
|
| 1947 |
+
"epoch": 0.267,
|
| 1948 |
+
"grad_norm": 0.08938033133745193,
|
| 1949 |
+
"learning_rate": 3.5076339158164635e-06,
|
| 1950 |
+
"loss": 6.487246704101563,
|
| 1951 |
+
"step": 1335
|
| 1952 |
+
},
|
| 1953 |
+
{
|
| 1954 |
+
"epoch": 0.268,
|
| 1955 |
+
"grad_norm": 0.0639733299612999,
|
| 1956 |
+
"learning_rate": 3.503279802794662e-06,
|
| 1957 |
+
"loss": 6.5667259216308596,
|
| 1958 |
+
"step": 1340
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 0.269,
|
| 1962 |
+
"grad_norm": 0.11628873646259308,
|
| 1963 |
+
"learning_rate": 3.4989092501457832e-06,
|
| 1964 |
+
"loss": 6.565748596191407,
|
| 1965 |
+
"step": 1345
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 0.27,
|
| 1969 |
+
"grad_norm": 0.0958283543586731,
|
| 1970 |
+
"learning_rate": 3.4945223056654914e-06,
|
| 1971 |
+
"loss": 6.564359283447265,
|
| 1972 |
+
"step": 1350
|
| 1973 |
+
},
|
| 1974 |
+
{
|
| 1975 |
+
"epoch": 0.27,
|
| 1976 |
+
"eval_accuracy": 0.1386959706959707,
|
| 1977 |
+
"eval_loss": 6.456244945526123,
|
| 1978 |
+
"eval_runtime": 10.3332,
|
| 1979 |
+
"eval_samples_per_second": 9.678,
|
| 1980 |
+
"eval_steps_per_second": 1.645,
|
| 1981 |
+
"step": 1350
|
| 1982 |
+
},
|
| 1983 |
+
{
|
| 1984 |
+
"epoch": 0.271,
|
| 1985 |
+
"grad_norm": 0.08361370116472244,
|
| 1986 |
+
"learning_rate": 3.4901190173287086e-06,
|
| 1987 |
+
"loss": 6.599095153808594,
|
| 1988 |
+
"step": 1355
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 0.272,
|
| 1992 |
+
"grad_norm": 0.07728071510791779,
|
| 1993 |
+
"learning_rate": 3.48569943328909e-06,
|
| 1994 |
+
"loss": 6.567552185058593,
|
| 1995 |
+
"step": 1360
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 0.273,
|
| 1999 |
+
"grad_norm": 0.23476524651050568,
|
| 2000 |
+
"learning_rate": 3.4812636018784994e-06,
|
| 2001 |
+
"loss": 6.519585418701172,
|
| 2002 |
+
"step": 1365
|
| 2003 |
+
},
|
| 2004 |
+
{
|
| 2005 |
+
"epoch": 0.274,
|
| 2006 |
+
"grad_norm": 0.08214252442121506,
|
| 2007 |
+
"learning_rate": 3.476811571606478e-06,
|
| 2008 |
+
"loss": 6.542357635498047,
|
| 2009 |
+
"step": 1370
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"epoch": 0.275,
|
| 2013 |
+
"grad_norm": 0.07677047699689865,
|
| 2014 |
+
"learning_rate": 3.4723433911597163e-06,
|
| 2015 |
+
"loss": 6.5419761657714846,
|
| 2016 |
+
"step": 1375
|
| 2017 |
+
},
|
| 2018 |
+
{
|
| 2019 |
+
"epoch": 0.276,
|
| 2020 |
+
"grad_norm": 0.11742841452360153,
|
| 2021 |
+
"learning_rate": 3.4678591094015187e-06,
|
| 2022 |
+
"loss": 6.5219581604003904,
|
| 2023 |
+
"step": 1380
|
| 2024 |
+
},
|
| 2025 |
+
{
|
| 2026 |
+
"epoch": 0.277,
|
| 2027 |
+
"grad_norm": 0.09686063230037689,
|
| 2028 |
+
"learning_rate": 3.4633587753712717e-06,
|
| 2029 |
+
"loss": 6.592768859863281,
|
| 2030 |
+
"step": 1385
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 0.278,
|
| 2034 |
+
"grad_norm": 0.09387538582086563,
|
| 2035 |
+
"learning_rate": 3.458842438283909e-06,
|
| 2036 |
+
"loss": 6.60013427734375,
|
| 2037 |
+
"step": 1390
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 0.279,
|
| 2041 |
+
"grad_norm": 0.0723787397146225,
|
| 2042 |
+
"learning_rate": 3.4543101475293682e-06,
|
| 2043 |
+
"loss": 6.4840553283691404,
|
| 2044 |
+
"step": 1395
|
| 2045 |
+
},
|
| 2046 |
+
{
|
| 2047 |
+
"epoch": 0.28,
|
| 2048 |
+
"grad_norm": 0.06756219267845154,
|
| 2049 |
+
"learning_rate": 3.449761952672055e-06,
|
| 2050 |
+
"loss": 6.542083740234375,
|
| 2051 |
+
"step": 1400
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 0.281,
|
| 2055 |
+
"grad_norm": 0.0757589340209961,
|
| 2056 |
+
"learning_rate": 3.4451979034502994e-06,
|
| 2057 |
+
"loss": 6.548152923583984,
|
| 2058 |
+
"step": 1405
|
| 2059 |
+
},
|
| 2060 |
+
{
|
| 2061 |
+
"epoch": 0.282,
|
| 2062 |
+
"grad_norm": 0.08698936551809311,
|
| 2063 |
+
"learning_rate": 3.440618049775814e-06,
|
| 2064 |
+
"loss": 6.562844085693359,
|
| 2065 |
+
"step": 1410
|
| 2066 |
+
},
|
| 2067 |
+
{
|
| 2068 |
+
"epoch": 0.283,
|
| 2069 |
+
"grad_norm": 0.08987753838300705,
|
| 2070 |
+
"learning_rate": 3.436022441733143e-06,
|
| 2071 |
+
"loss": 6.502987670898437,
|
| 2072 |
+
"step": 1415
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 0.284,
|
| 2076 |
+
"grad_norm": 0.07419227808713913,
|
| 2077 |
+
"learning_rate": 3.43141112957912e-06,
|
| 2078 |
+
"loss": 6.47925033569336,
|
| 2079 |
+
"step": 1420
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 0.285,
|
| 2083 |
+
"grad_norm": 0.0743662565946579,
|
| 2084 |
+
"learning_rate": 3.4267841637423166e-06,
|
| 2085 |
+
"loss": 6.590135955810547,
|
| 2086 |
+
"step": 1425
|
| 2087 |
+
},
|
| 2088 |
+
{
|
| 2089 |
+
"epoch": 0.286,
|
| 2090 |
+
"grad_norm": 0.09506815671920776,
|
| 2091 |
+
"learning_rate": 3.422141594822489e-06,
|
| 2092 |
+
"loss": 6.4904052734375,
|
| 2093 |
+
"step": 1430
|
| 2094 |
+
},
|
| 2095 |
+
{
|
| 2096 |
+
"epoch": 0.287,
|
| 2097 |
+
"grad_norm": 0.06352763622999191,
|
| 2098 |
+
"learning_rate": 3.4174834735900283e-06,
|
| 2099 |
+
"loss": 6.492788696289063,
|
| 2100 |
+
"step": 1435
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"epoch": 0.288,
|
| 2104 |
+
"grad_norm": 0.1327190101146698,
|
| 2105 |
+
"learning_rate": 3.4128098509854014e-06,
|
| 2106 |
+
"loss": 6.528990936279297,
|
| 2107 |
+
"step": 1440
|
| 2108 |
+
},
|
| 2109 |
+
{
|
| 2110 |
+
"epoch": 0.289,
|
| 2111 |
+
"grad_norm": 0.1287909299135208,
|
| 2112 |
+
"learning_rate": 3.4081207781185963e-06,
|
| 2113 |
+
"loss": 6.511381530761719,
|
| 2114 |
+
"step": 1445
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"epoch": 0.29,
|
| 2118 |
+
"grad_norm": 0.05822750926017761,
|
| 2119 |
+
"learning_rate": 3.4034163062685623e-06,
|
| 2120 |
+
"loss": 6.534752655029297,
|
| 2121 |
+
"step": 1450
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 0.291,
|
| 2125 |
+
"grad_norm": 0.08409831672906876,
|
| 2126 |
+
"learning_rate": 3.3986964868826498e-06,
|
| 2127 |
+
"loss": 6.494256591796875,
|
| 2128 |
+
"step": 1455
|
| 2129 |
+
},
|
| 2130 |
+
{
|
| 2131 |
+
"epoch": 0.292,
|
| 2132 |
+
"grad_norm": 0.11647513508796692,
|
| 2133 |
+
"learning_rate": 3.393961371576048e-06,
|
| 2134 |
+
"loss": 6.531385040283203,
|
| 2135 |
+
"step": 1460
|
| 2136 |
+
},
|
| 2137 |
+
{
|
| 2138 |
+
"epoch": 0.293,
|
| 2139 |
+
"grad_norm": 0.06273698806762695,
|
| 2140 |
+
"learning_rate": 3.3892110121312196e-06,
|
| 2141 |
+
"loss": 6.5295967102050785,
|
| 2142 |
+
"step": 1465
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 0.294,
|
| 2146 |
+
"grad_norm": 0.06390611082315445,
|
| 2147 |
+
"learning_rate": 3.3844454604973327e-06,
|
| 2148 |
+
"loss": 6.487610626220703,
|
| 2149 |
+
"step": 1470
|
| 2150 |
+
},
|
| 2151 |
+
{
|
| 2152 |
+
"epoch": 0.295,
|
| 2153 |
+
"grad_norm": 0.0920199602842331,
|
| 2154 |
+
"learning_rate": 3.3796647687896982e-06,
|
| 2155 |
+
"loss": 6.561590576171875,
|
| 2156 |
+
"step": 1475
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 0.296,
|
| 2160 |
+
"grad_norm": 0.07950405776500702,
|
| 2161 |
+
"learning_rate": 3.3748689892891945e-06,
|
| 2162 |
+
"loss": 6.498722839355469,
|
| 2163 |
+
"step": 1480
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 0.297,
|
| 2167 |
+
"grad_norm": 0.0956474170088768,
|
| 2168 |
+
"learning_rate": 3.3700581744416966e-06,
|
| 2169 |
+
"loss": 6.526725006103516,
|
| 2170 |
+
"step": 1485
|
| 2171 |
+
},
|
| 2172 |
+
{
|
| 2173 |
+
"epoch": 0.298,
|
| 2174 |
+
"grad_norm": 0.07354336977005005,
|
| 2175 |
+
"learning_rate": 3.3652323768575066e-06,
|
| 2176 |
+
"loss": 6.478034210205078,
|
| 2177 |
+
"step": 1490
|
| 2178 |
+
},
|
| 2179 |
+
{
|
| 2180 |
+
"epoch": 0.299,
|
| 2181 |
+
"grad_norm": 0.07010781019926071,
|
| 2182 |
+
"learning_rate": 3.360391649310772e-06,
|
| 2183 |
+
"loss": 6.511841583251953,
|
| 2184 |
+
"step": 1495
|
| 2185 |
+
},
|
| 2186 |
+
{
|
| 2187 |
+
"epoch": 0.3,
|
| 2188 |
+
"grad_norm": 0.07634287327528,
|
| 2189 |
+
"learning_rate": 3.355536044738915e-06,
|
| 2190 |
+
"loss": 6.506979370117188,
|
| 2191 |
+
"step": 1500
|
| 2192 |
+
},
|
| 2193 |
+
{
|
| 2194 |
+
"epoch": 0.3,
|
| 2195 |
+
"eval_accuracy": 0.138991452991453,
|
| 2196 |
+
"eval_loss": 6.424478530883789,
|
| 2197 |
+
"eval_runtime": 10.3935,
|
| 2198 |
+
"eval_samples_per_second": 9.621,
|
| 2199 |
+
"eval_steps_per_second": 1.636,
|
| 2200 |
+
"step": 1500
|
| 2201 |
}
|
| 2202 |
],
|
| 2203 |
"logging_steps": 5,
|