Training in progress, step 3000, 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 4523108832
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3f56a2fa609ef8d154fd837952b13a51a51223d702ba30b82088ee14021769ec
|
| 3 |
size 4523108832
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2912179275
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2702881bd4a85166ca761355eab24a3cb3e686083afb5dffe3f6ed30ea707793
|
| 3 |
size 2912179275
|
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 sha256:1255b82ec9f462fef6f8c2f4a8ee876c4b6d998a53c99546a4eddcc1aabaefc7
|
| 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 sha256:5ac1c46a2776d12775d23d0f587efc112188137ce2140da35bc15d301c9f620e
|
| 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 sha256:1ff3c5ccbd008d14dbd5e50d70f5d7740fe5c810277aeac0c927a840460eeea7
|
| 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":
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -1758,6 +1758,356 @@
|
|
| 1758 |
"learning_rate": 0.00015002,
|
| 1759 |
"loss": 0.22608683109283448,
|
| 1760 |
"step": 2500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1761 |
}
|
| 1762 |
],
|
| 1763 |
"logging_steps": 10,
|
|
@@ -1777,7 +2127,7 @@
|
|
| 1777 |
"attributes": {}
|
| 1778 |
}
|
| 1779 |
},
|
| 1780 |
-
"total_flos": 2.
|
| 1781 |
"train_batch_size": 1,
|
| 1782 |
"trial_name": null,
|
| 1783 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 4.0064,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 3000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 1758 |
"learning_rate": 0.00015002,
|
| 1759 |
"loss": 0.22608683109283448,
|
| 1760 |
"step": 2500
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 2.0198,
|
| 1764 |
+
"grad_norm": 0.5102221369743347,
|
| 1765 |
+
"learning_rate": 0.00014982,
|
| 1766 |
+
"loss": 0.21579885482788086,
|
| 1767 |
+
"step": 2510
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 2.0208,
|
| 1771 |
+
"grad_norm": 0.5863536596298218,
|
| 1772 |
+
"learning_rate": 0.00014962,
|
| 1773 |
+
"loss": 0.2347024440765381,
|
| 1774 |
+
"step": 2520
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 2.0218,
|
| 1778 |
+
"grad_norm": 0.5884446501731873,
|
| 1779 |
+
"learning_rate": 0.00014942000000000002,
|
| 1780 |
+
"loss": 0.2166447401046753,
|
| 1781 |
+
"step": 2530
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 2.0228,
|
| 1785 |
+
"grad_norm": 0.5152504444122314,
|
| 1786 |
+
"learning_rate": 0.00014922,
|
| 1787 |
+
"loss": 0.23897838592529297,
|
| 1788 |
+
"step": 2540
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 2.0238,
|
| 1792 |
+
"grad_norm": 0.5819721221923828,
|
| 1793 |
+
"learning_rate": 0.00014902,
|
| 1794 |
+
"loss": 0.2348470449447632,
|
| 1795 |
+
"step": 2550
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 2.0248,
|
| 1799 |
+
"grad_norm": 0.5081148147583008,
|
| 1800 |
+
"learning_rate": 0.00014882,
|
| 1801 |
+
"loss": 0.2127830982208252,
|
| 1802 |
+
"step": 2560
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 2.0258,
|
| 1806 |
+
"grad_norm": 0.5137162804603577,
|
| 1807 |
+
"learning_rate": 0.00014862,
|
| 1808 |
+
"loss": 0.20368046760559083,
|
| 1809 |
+
"step": 2570
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 2.0268,
|
| 1813 |
+
"grad_norm": 0.45256203413009644,
|
| 1814 |
+
"learning_rate": 0.00014842,
|
| 1815 |
+
"loss": 0.19960324764251708,
|
| 1816 |
+
"step": 2580
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 2.0278,
|
| 1820 |
+
"grad_norm": 0.5101103782653809,
|
| 1821 |
+
"learning_rate": 0.00014822000000000001,
|
| 1822 |
+
"loss": 0.19512444734573364,
|
| 1823 |
+
"step": 2590
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 2.0288,
|
| 1827 |
+
"grad_norm": 0.5388202667236328,
|
| 1828 |
+
"learning_rate": 0.00014802,
|
| 1829 |
+
"loss": 0.20622694492340088,
|
| 1830 |
+
"step": 2600
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 2.0298,
|
| 1834 |
+
"grad_norm": 0.5038774013519287,
|
| 1835 |
+
"learning_rate": 0.00014782,
|
| 1836 |
+
"loss": 0.19988502264022828,
|
| 1837 |
+
"step": 2610
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 2.0308,
|
| 1841 |
+
"grad_norm": 0.5558412671089172,
|
| 1842 |
+
"learning_rate": 0.00014762,
|
| 1843 |
+
"loss": 0.20110974311828614,
|
| 1844 |
+
"step": 2620
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 3.0006,
|
| 1848 |
+
"grad_norm": 0.43655768036842346,
|
| 1849 |
+
"learning_rate": 0.00014742,
|
| 1850 |
+
"loss": 0.1968843460083008,
|
| 1851 |
+
"step": 2630
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 3.0016,
|
| 1855 |
+
"grad_norm": 0.4135885536670685,
|
| 1856 |
+
"learning_rate": 0.00014722,
|
| 1857 |
+
"loss": 0.13574039936065674,
|
| 1858 |
+
"step": 2640
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 3.0026,
|
| 1862 |
+
"grad_norm": 0.364497572183609,
|
| 1863 |
+
"learning_rate": 0.00014702,
|
| 1864 |
+
"loss": 0.12638025283813475,
|
| 1865 |
+
"step": 2650
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 3.0036,
|
| 1869 |
+
"grad_norm": 0.4627242982387543,
|
| 1870 |
+
"learning_rate": 0.00014682,
|
| 1871 |
+
"loss": 0.1232415795326233,
|
| 1872 |
+
"step": 2660
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 3.0046,
|
| 1876 |
+
"grad_norm": 0.4199890196323395,
|
| 1877 |
+
"learning_rate": 0.00014662,
|
| 1878 |
+
"loss": 0.1209063172340393,
|
| 1879 |
+
"step": 2670
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 3.0056,
|
| 1883 |
+
"grad_norm": 0.3729947507381439,
|
| 1884 |
+
"learning_rate": 0.00014642,
|
| 1885 |
+
"loss": 0.11018155813217163,
|
| 1886 |
+
"step": 2680
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 3.0066,
|
| 1890 |
+
"grad_norm": 0.49802160263061523,
|
| 1891 |
+
"learning_rate": 0.00014622,
|
| 1892 |
+
"loss": 0.12115752696990967,
|
| 1893 |
+
"step": 2690
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 3.0076,
|
| 1897 |
+
"grad_norm": 0.5387603044509888,
|
| 1898 |
+
"learning_rate": 0.00014602,
|
| 1899 |
+
"loss": 0.1478494167327881,
|
| 1900 |
+
"step": 2700
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 3.0086,
|
| 1904 |
+
"grad_norm": 0.46145540475845337,
|
| 1905 |
+
"learning_rate": 0.00014582,
|
| 1906 |
+
"loss": 0.14995505809783935,
|
| 1907 |
+
"step": 2710
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 3.0096,
|
| 1911 |
+
"grad_norm": 0.4195813834667206,
|
| 1912 |
+
"learning_rate": 0.00014562,
|
| 1913 |
+
"loss": 0.16438782215118408,
|
| 1914 |
+
"step": 2720
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 3.0106,
|
| 1918 |
+
"grad_norm": 0.4632911682128906,
|
| 1919 |
+
"learning_rate": 0.00014542,
|
| 1920 |
+
"loss": 0.16314655542373657,
|
| 1921 |
+
"step": 2730
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 3.0116,
|
| 1925 |
+
"grad_norm": 0.49666211009025574,
|
| 1926 |
+
"learning_rate": 0.00014522,
|
| 1927 |
+
"loss": 0.1643773317337036,
|
| 1928 |
+
"step": 2740
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 3.0126,
|
| 1932 |
+
"grad_norm": 0.4677850306034088,
|
| 1933 |
+
"learning_rate": 0.00014502,
|
| 1934 |
+
"loss": 0.14674897193908693,
|
| 1935 |
+
"step": 2750
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 3.0136,
|
| 1939 |
+
"grad_norm": 0.48538899421691895,
|
| 1940 |
+
"learning_rate": 0.00014482,
|
| 1941 |
+
"loss": 0.151479172706604,
|
| 1942 |
+
"step": 2760
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 3.0146,
|
| 1946 |
+
"grad_norm": 0.4430471360683441,
|
| 1947 |
+
"learning_rate": 0.00014462,
|
| 1948 |
+
"loss": 0.16841667890548706,
|
| 1949 |
+
"step": 2770
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 3.0156,
|
| 1953 |
+
"grad_norm": 0.5144022107124329,
|
| 1954 |
+
"learning_rate": 0.00014442,
|
| 1955 |
+
"loss": 0.16647714376449585,
|
| 1956 |
+
"step": 2780
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 3.0166,
|
| 1960 |
+
"grad_norm": 0.45523014664649963,
|
| 1961 |
+
"learning_rate": 0.00014422,
|
| 1962 |
+
"loss": 0.14415124654769898,
|
| 1963 |
+
"step": 2790
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 3.0176,
|
| 1967 |
+
"grad_norm": 0.5218808054924011,
|
| 1968 |
+
"learning_rate": 0.00014402,
|
| 1969 |
+
"loss": 0.15597842931747435,
|
| 1970 |
+
"step": 2800
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 3.0186,
|
| 1974 |
+
"grad_norm": 0.4452967941761017,
|
| 1975 |
+
"learning_rate": 0.00014382,
|
| 1976 |
+
"loss": 0.1488332986831665,
|
| 1977 |
+
"step": 2810
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 3.0196,
|
| 1981 |
+
"grad_norm": 0.4646563231945038,
|
| 1982 |
+
"learning_rate": 0.00014362,
|
| 1983 |
+
"loss": 0.1394293189048767,
|
| 1984 |
+
"step": 2820
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 3.0206,
|
| 1988 |
+
"grad_norm": 0.5117284059524536,
|
| 1989 |
+
"learning_rate": 0.00014342,
|
| 1990 |
+
"loss": 0.15330575704574584,
|
| 1991 |
+
"step": 2830
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 3.0216,
|
| 1995 |
+
"grad_norm": 0.4975617825984955,
|
| 1996 |
+
"learning_rate": 0.00014322,
|
| 1997 |
+
"loss": 0.1486660957336426,
|
| 1998 |
+
"step": 2840
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 3.0226,
|
| 2002 |
+
"grad_norm": 0.4680009186267853,
|
| 2003 |
+
"learning_rate": 0.00014302,
|
| 2004 |
+
"loss": 0.15574642419815063,
|
| 2005 |
+
"step": 2850
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 3.0236,
|
| 2009 |
+
"grad_norm": 0.45793306827545166,
|
| 2010 |
+
"learning_rate": 0.00014282,
|
| 2011 |
+
"loss": 0.1538609027862549,
|
| 2012 |
+
"step": 2860
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 3.0246,
|
| 2016 |
+
"grad_norm": 0.43385592103004456,
|
| 2017 |
+
"learning_rate": 0.00014261999999999999,
|
| 2018 |
+
"loss": 0.14147435426712035,
|
| 2019 |
+
"step": 2870
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 3.0256,
|
| 2023 |
+
"grad_norm": 0.48055240511894226,
|
| 2024 |
+
"learning_rate": 0.00014242,
|
| 2025 |
+
"loss": 0.13285771608352662,
|
| 2026 |
+
"step": 2880
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 3.0266,
|
| 2030 |
+
"grad_norm": 0.426279217004776,
|
| 2031 |
+
"learning_rate": 0.00014222,
|
| 2032 |
+
"loss": 0.13433635234832764,
|
| 2033 |
+
"step": 2890
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 3.0276,
|
| 2037 |
+
"grad_norm": 0.4118101894855499,
|
| 2038 |
+
"learning_rate": 0.00014202,
|
| 2039 |
+
"loss": 0.13033357858657837,
|
| 2040 |
+
"step": 2900
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 3.0286,
|
| 2044 |
+
"grad_norm": 0.4413125216960907,
|
| 2045 |
+
"learning_rate": 0.00014182,
|
| 2046 |
+
"loss": 0.1351193904876709,
|
| 2047 |
+
"step": 2910
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 3.0296,
|
| 2051 |
+
"grad_norm": 0.41682061553001404,
|
| 2052 |
+
"learning_rate": 0.00014162,
|
| 2053 |
+
"loss": 0.13452460765838622,
|
| 2054 |
+
"step": 2920
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 3.0306,
|
| 2058 |
+
"grad_norm": 0.5230453610420227,
|
| 2059 |
+
"learning_rate": 0.00014141999999999998,
|
| 2060 |
+
"loss": 0.1290982484817505,
|
| 2061 |
+
"step": 2930
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 4.0004,
|
| 2065 |
+
"grad_norm": 0.3429013788700104,
|
| 2066 |
+
"learning_rate": 0.00014122,
|
| 2067 |
+
"loss": 0.14483762979507447,
|
| 2068 |
+
"step": 2940
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 4.0014,
|
| 2072 |
+
"grad_norm": 0.32104936242103577,
|
| 2073 |
+
"learning_rate": 0.00014102,
|
| 2074 |
+
"loss": 0.10209068059921264,
|
| 2075 |
+
"step": 2950
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 4.0024,
|
| 2079 |
+
"grad_norm": 0.29722949862480164,
|
| 2080 |
+
"learning_rate": 0.00014082,
|
| 2081 |
+
"loss": 0.0967288076877594,
|
| 2082 |
+
"step": 2960
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 4.0034,
|
| 2086 |
+
"grad_norm": 0.4013476073741913,
|
| 2087 |
+
"learning_rate": 0.00014062,
|
| 2088 |
+
"loss": 0.08881974220275879,
|
| 2089 |
+
"step": 2970
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 4.0044,
|
| 2093 |
+
"grad_norm": 0.37912485003471375,
|
| 2094 |
+
"learning_rate": 0.00014042,
|
| 2095 |
+
"loss": 0.09519890546798707,
|
| 2096 |
+
"step": 2980
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 4.0054,
|
| 2100 |
+
"grad_norm": 0.31893491744995117,
|
| 2101 |
+
"learning_rate": 0.00014022,
|
| 2102 |
+
"loss": 0.08268740773200989,
|
| 2103 |
+
"step": 2990
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 4.0064,
|
| 2107 |
+
"grad_norm": 0.3440733850002289,
|
| 2108 |
+
"learning_rate": 0.00014002,
|
| 2109 |
+
"loss": 0.08511788249015809,
|
| 2110 |
+
"step": 3000
|
| 2111 |
}
|
| 2112 |
],
|
| 2113 |
"logging_steps": 10,
|
|
|
|
| 2127 |
"attributes": {}
|
| 2128 |
}
|
| 2129 |
},
|
| 2130 |
+
"total_flos": 2.5397157268094976e+17,
|
| 2131 |
"train_batch_size": 1,
|
| 2132 |
"trial_name": null,
|
| 2133 |
"trial_params": null
|