Commit
•
ede467d
1
Parent(s):
507c313
Upload folder using huggingface_hub (#3)
Browse files- Upload folder using huggingface_hub (d2bcdf81b4c7be220954c2f72b33d6d95238beca)
- adapter_model.bin +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +603 -3
adapter_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 39409357
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e7b701342ea311c1e25f0f6953505f048df3f56027dc5c764b95df4491bfde1c
|
| 3 |
size 39409357
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 78844421
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:da79792a601e6737b980ff24e453353405dce9a807a0964433ac96d5a97b84aa
|
| 3 |
size 78844421
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14575
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7508d4b8dd267de5cc58e972da25236687927651336a28f292c92f7f23951475
|
| 3 |
size 14575
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 627
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:04fdaf305b17644f9d215d28be45b77e6d0f8e4b5adc1c7045c50a61bd8f3c5b
|
| 3 |
size 627
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 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,
|
|
@@ -1807,13 +1807,613 @@
|
|
| 1807 |
"learning_rate": 0.00011620111731843578,
|
| 1808 |
"loss": 1.6315,
|
| 1809 |
"step": 300
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1810 |
}
|
| 1811 |
],
|
| 1812 |
"logging_steps": 1,
|
| 1813 |
"max_steps": 716,
|
| 1814 |
"num_train_epochs": 4,
|
| 1815 |
"save_steps": 100,
|
| 1816 |
-
"total_flos":
|
| 1817 |
"trial_name": null,
|
| 1818 |
"trial_params": null
|
| 1819 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.2346368715083798,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 400,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1807 |
"learning_rate": 0.00011620111731843578,
|
| 1808 |
"loss": 1.6315,
|
| 1809 |
"step": 300
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 1.68,
|
| 1813 |
+
"learning_rate": 0.00011592178770949722,
|
| 1814 |
+
"loss": 1.6802,
|
| 1815 |
+
"step": 301
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 1.69,
|
| 1819 |
+
"learning_rate": 0.00011564245810055867,
|
| 1820 |
+
"loss": 1.7174,
|
| 1821 |
+
"step": 302
|
| 1822 |
+
},
|
| 1823 |
+
{
|
| 1824 |
+
"epoch": 1.69,
|
| 1825 |
+
"learning_rate": 0.00011536312849162012,
|
| 1826 |
+
"loss": 1.5212,
|
| 1827 |
+
"step": 303
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 1.7,
|
| 1831 |
+
"learning_rate": 0.00011508379888268157,
|
| 1832 |
+
"loss": 1.5808,
|
| 1833 |
+
"step": 304
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"epoch": 1.7,
|
| 1837 |
+
"learning_rate": 0.00011480446927374303,
|
| 1838 |
+
"loss": 1.6152,
|
| 1839 |
+
"step": 305
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 1.71,
|
| 1843 |
+
"learning_rate": 0.00011452513966480447,
|
| 1844 |
+
"loss": 1.5435,
|
| 1845 |
+
"step": 306
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 1.72,
|
| 1849 |
+
"learning_rate": 0.00011424581005586592,
|
| 1850 |
+
"loss": 1.6603,
|
| 1851 |
+
"step": 307
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 1.72,
|
| 1855 |
+
"learning_rate": 0.00011396648044692737,
|
| 1856 |
+
"loss": 1.685,
|
| 1857 |
+
"step": 308
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 1.73,
|
| 1861 |
+
"learning_rate": 0.00011368715083798884,
|
| 1862 |
+
"loss": 1.6002,
|
| 1863 |
+
"step": 309
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 1.73,
|
| 1867 |
+
"learning_rate": 0.00011340782122905029,
|
| 1868 |
+
"loss": 1.6046,
|
| 1869 |
+
"step": 310
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 1.74,
|
| 1873 |
+
"learning_rate": 0.00011312849162011174,
|
| 1874 |
+
"loss": 1.5969,
|
| 1875 |
+
"step": 311
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 1.74,
|
| 1879 |
+
"learning_rate": 0.0001128491620111732,
|
| 1880 |
+
"loss": 1.5845,
|
| 1881 |
+
"step": 312
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"epoch": 1.75,
|
| 1885 |
+
"learning_rate": 0.00011256983240223464,
|
| 1886 |
+
"loss": 1.8183,
|
| 1887 |
+
"step": 313
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"epoch": 1.75,
|
| 1891 |
+
"learning_rate": 0.00011229050279329609,
|
| 1892 |
+
"loss": 1.6953,
|
| 1893 |
+
"step": 314
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 1.76,
|
| 1897 |
+
"learning_rate": 0.00011201117318435754,
|
| 1898 |
+
"loss": 1.7787,
|
| 1899 |
+
"step": 315
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 1.77,
|
| 1903 |
+
"learning_rate": 0.000111731843575419,
|
| 1904 |
+
"loss": 1.6422,
|
| 1905 |
+
"step": 316
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 1.77,
|
| 1909 |
+
"learning_rate": 0.00011145251396648045,
|
| 1910 |
+
"loss": 1.7034,
|
| 1911 |
+
"step": 317
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 1.78,
|
| 1915 |
+
"learning_rate": 0.00011117318435754192,
|
| 1916 |
+
"loss": 1.7301,
|
| 1917 |
+
"step": 318
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 1.78,
|
| 1921 |
+
"learning_rate": 0.00011089385474860337,
|
| 1922 |
+
"loss": 1.7084,
|
| 1923 |
+
"step": 319
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 1.79,
|
| 1927 |
+
"learning_rate": 0.00011061452513966482,
|
| 1928 |
+
"loss": 1.772,
|
| 1929 |
+
"step": 320
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 1.79,
|
| 1933 |
+
"learning_rate": 0.00011033519553072626,
|
| 1934 |
+
"loss": 1.5733,
|
| 1935 |
+
"step": 321
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 1.8,
|
| 1939 |
+
"learning_rate": 0.00011005586592178771,
|
| 1940 |
+
"loss": 1.6423,
|
| 1941 |
+
"step": 322
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 1.8,
|
| 1945 |
+
"learning_rate": 0.00010977653631284917,
|
| 1946 |
+
"loss": 1.5809,
|
| 1947 |
+
"step": 323
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 1.81,
|
| 1951 |
+
"learning_rate": 0.00010949720670391062,
|
| 1952 |
+
"loss": 1.6781,
|
| 1953 |
+
"step": 324
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 1.82,
|
| 1957 |
+
"learning_rate": 0.00010921787709497207,
|
| 1958 |
+
"loss": 1.6788,
|
| 1959 |
+
"step": 325
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 1.82,
|
| 1963 |
+
"learning_rate": 0.00010893854748603351,
|
| 1964 |
+
"loss": 1.6346,
|
| 1965 |
+
"step": 326
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 1.83,
|
| 1969 |
+
"learning_rate": 0.00010865921787709499,
|
| 1970 |
+
"loss": 1.6634,
|
| 1971 |
+
"step": 327
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 1.83,
|
| 1975 |
+
"learning_rate": 0.00010837988826815643,
|
| 1976 |
+
"loss": 1.7561,
|
| 1977 |
+
"step": 328
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 1.84,
|
| 1981 |
+
"learning_rate": 0.00010810055865921788,
|
| 1982 |
+
"loss": 1.66,
|
| 1983 |
+
"step": 329
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 1.84,
|
| 1987 |
+
"learning_rate": 0.00010782122905027934,
|
| 1988 |
+
"loss": 1.7298,
|
| 1989 |
+
"step": 330
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 1.85,
|
| 1993 |
+
"learning_rate": 0.00010754189944134079,
|
| 1994 |
+
"loss": 1.6893,
|
| 1995 |
+
"step": 331
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 1.85,
|
| 1999 |
+
"learning_rate": 0.00010726256983240224,
|
| 2000 |
+
"loss": 1.7631,
|
| 2001 |
+
"step": 332
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 1.86,
|
| 2005 |
+
"learning_rate": 0.00010698324022346368,
|
| 2006 |
+
"loss": 1.6633,
|
| 2007 |
+
"step": 333
|
| 2008 |
+
},
|
| 2009 |
+
{
|
| 2010 |
+
"epoch": 1.87,
|
| 2011 |
+
"learning_rate": 0.00010670391061452513,
|
| 2012 |
+
"loss": 1.5388,
|
| 2013 |
+
"step": 334
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 1.87,
|
| 2017 |
+
"learning_rate": 0.00010642458100558659,
|
| 2018 |
+
"loss": 1.6718,
|
| 2019 |
+
"step": 335
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 1.88,
|
| 2023 |
+
"learning_rate": 0.00010614525139664805,
|
| 2024 |
+
"loss": 1.5536,
|
| 2025 |
+
"step": 336
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 1.88,
|
| 2029 |
+
"learning_rate": 0.00010586592178770951,
|
| 2030 |
+
"loss": 1.6483,
|
| 2031 |
+
"step": 337
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 1.89,
|
| 2035 |
+
"learning_rate": 0.00010558659217877096,
|
| 2036 |
+
"loss": 1.5774,
|
| 2037 |
+
"step": 338
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 1.89,
|
| 2041 |
+
"learning_rate": 0.00010530726256983241,
|
| 2042 |
+
"loss": 1.6366,
|
| 2043 |
+
"step": 339
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 1.9,
|
| 2047 |
+
"learning_rate": 0.00010502793296089387,
|
| 2048 |
+
"loss": 1.5567,
|
| 2049 |
+
"step": 340
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 1.91,
|
| 2053 |
+
"learning_rate": 0.0001047486033519553,
|
| 2054 |
+
"loss": 1.5323,
|
| 2055 |
+
"step": 341
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 1.91,
|
| 2059 |
+
"learning_rate": 0.00010446927374301676,
|
| 2060 |
+
"loss": 1.4608,
|
| 2061 |
+
"step": 342
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 1.92,
|
| 2065 |
+
"learning_rate": 0.00010418994413407821,
|
| 2066 |
+
"loss": 1.5933,
|
| 2067 |
+
"step": 343
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 1.92,
|
| 2071 |
+
"learning_rate": 0.00010391061452513966,
|
| 2072 |
+
"loss": 1.6625,
|
| 2073 |
+
"step": 344
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 1.93,
|
| 2077 |
+
"learning_rate": 0.00010363128491620113,
|
| 2078 |
+
"loss": 1.7236,
|
| 2079 |
+
"step": 345
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 1.93,
|
| 2083 |
+
"learning_rate": 0.00010335195530726258,
|
| 2084 |
+
"loss": 1.759,
|
| 2085 |
+
"step": 346
|
| 2086 |
+
},
|
| 2087 |
+
{
|
| 2088 |
+
"epoch": 1.94,
|
| 2089 |
+
"learning_rate": 0.00010307262569832404,
|
| 2090 |
+
"loss": 1.7248,
|
| 2091 |
+
"step": 347
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 1.94,
|
| 2095 |
+
"learning_rate": 0.00010279329608938548,
|
| 2096 |
+
"loss": 1.5144,
|
| 2097 |
+
"step": 348
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 1.95,
|
| 2101 |
+
"learning_rate": 0.00010251396648044693,
|
| 2102 |
+
"loss": 1.6905,
|
| 2103 |
+
"step": 349
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 1.96,
|
| 2107 |
+
"learning_rate": 0.00010223463687150838,
|
| 2108 |
+
"loss": 1.6119,
|
| 2109 |
+
"step": 350
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 1.96,
|
| 2113 |
+
"learning_rate": 0.00010195530726256983,
|
| 2114 |
+
"loss": 1.5464,
|
| 2115 |
+
"step": 351
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 1.97,
|
| 2119 |
+
"learning_rate": 0.00010167597765363129,
|
| 2120 |
+
"loss": 1.6901,
|
| 2121 |
+
"step": 352
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 1.97,
|
| 2125 |
+
"learning_rate": 0.00010139664804469273,
|
| 2126 |
+
"loss": 1.3511,
|
| 2127 |
+
"step": 353
|
| 2128 |
+
},
|
| 2129 |
+
{
|
| 2130 |
+
"epoch": 1.98,
|
| 2131 |
+
"learning_rate": 0.0001011173184357542,
|
| 2132 |
+
"loss": 1.5434,
|
| 2133 |
+
"step": 354
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 1.98,
|
| 2137 |
+
"learning_rate": 0.00010083798882681566,
|
| 2138 |
+
"loss": 1.5891,
|
| 2139 |
+
"step": 355
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 1.99,
|
| 2143 |
+
"learning_rate": 0.0001005586592178771,
|
| 2144 |
+
"loss": 1.6658,
|
| 2145 |
+
"step": 356
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 1.99,
|
| 2149 |
+
"learning_rate": 0.00010027932960893855,
|
| 2150 |
+
"loss": 1.5657,
|
| 2151 |
+
"step": 357
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 2.0,
|
| 2155 |
+
"learning_rate": 0.0001,
|
| 2156 |
+
"loss": 1.7005,
|
| 2157 |
+
"step": 358
|
| 2158 |
+
},
|
| 2159 |
+
{
|
| 2160 |
+
"epoch": 2.01,
|
| 2161 |
+
"learning_rate": 9.972067039106146e-05,
|
| 2162 |
+
"loss": 1.4202,
|
| 2163 |
+
"step": 359
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 2.01,
|
| 2167 |
+
"learning_rate": 9.944134078212291e-05,
|
| 2168 |
+
"loss": 1.5262,
|
| 2169 |
+
"step": 360
|
| 2170 |
+
},
|
| 2171 |
+
{
|
| 2172 |
+
"epoch": 2.02,
|
| 2173 |
+
"learning_rate": 9.916201117318436e-05,
|
| 2174 |
+
"loss": 1.6323,
|
| 2175 |
+
"step": 361
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 2.02,
|
| 2179 |
+
"learning_rate": 9.888268156424582e-05,
|
| 2180 |
+
"loss": 1.5521,
|
| 2181 |
+
"step": 362
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 2.03,
|
| 2185 |
+
"learning_rate": 9.860335195530727e-05,
|
| 2186 |
+
"loss": 1.5762,
|
| 2187 |
+
"step": 363
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 2.03,
|
| 2191 |
+
"learning_rate": 9.832402234636872e-05,
|
| 2192 |
+
"loss": 1.613,
|
| 2193 |
+
"step": 364
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 2.04,
|
| 2197 |
+
"learning_rate": 9.804469273743018e-05,
|
| 2198 |
+
"loss": 1.4231,
|
| 2199 |
+
"step": 365
|
| 2200 |
+
},
|
| 2201 |
+
{
|
| 2202 |
+
"epoch": 2.04,
|
| 2203 |
+
"learning_rate": 9.776536312849163e-05,
|
| 2204 |
+
"loss": 1.5706,
|
| 2205 |
+
"step": 366
|
| 2206 |
+
},
|
| 2207 |
+
{
|
| 2208 |
+
"epoch": 2.05,
|
| 2209 |
+
"learning_rate": 9.748603351955308e-05,
|
| 2210 |
+
"loss": 1.5245,
|
| 2211 |
+
"step": 367
|
| 2212 |
+
},
|
| 2213 |
+
{
|
| 2214 |
+
"epoch": 2.06,
|
| 2215 |
+
"learning_rate": 9.720670391061453e-05,
|
| 2216 |
+
"loss": 1.4771,
|
| 2217 |
+
"step": 368
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 2.06,
|
| 2221 |
+
"learning_rate": 9.692737430167597e-05,
|
| 2222 |
+
"loss": 1.596,
|
| 2223 |
+
"step": 369
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"epoch": 2.07,
|
| 2227 |
+
"learning_rate": 9.664804469273744e-05,
|
| 2228 |
+
"loss": 1.537,
|
| 2229 |
+
"step": 370
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 2.07,
|
| 2233 |
+
"learning_rate": 9.636871508379889e-05,
|
| 2234 |
+
"loss": 1.4276,
|
| 2235 |
+
"step": 371
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 2.08,
|
| 2239 |
+
"learning_rate": 9.608938547486033e-05,
|
| 2240 |
+
"loss": 1.4746,
|
| 2241 |
+
"step": 372
|
| 2242 |
+
},
|
| 2243 |
+
{
|
| 2244 |
+
"epoch": 2.08,
|
| 2245 |
+
"learning_rate": 9.581005586592178e-05,
|
| 2246 |
+
"loss": 1.4374,
|
| 2247 |
+
"step": 373
|
| 2248 |
+
},
|
| 2249 |
+
{
|
| 2250 |
+
"epoch": 2.09,
|
| 2251 |
+
"learning_rate": 9.553072625698325e-05,
|
| 2252 |
+
"loss": 1.4704,
|
| 2253 |
+
"step": 374
|
| 2254 |
+
},
|
| 2255 |
+
{
|
| 2256 |
+
"epoch": 2.09,
|
| 2257 |
+
"learning_rate": 9.52513966480447e-05,
|
| 2258 |
+
"loss": 1.5997,
|
| 2259 |
+
"step": 375
|
| 2260 |
+
},
|
| 2261 |
+
{
|
| 2262 |
+
"epoch": 2.1,
|
| 2263 |
+
"learning_rate": 9.497206703910614e-05,
|
| 2264 |
+
"loss": 1.5034,
|
| 2265 |
+
"step": 376
|
| 2266 |
+
},
|
| 2267 |
+
{
|
| 2268 |
+
"epoch": 2.11,
|
| 2269 |
+
"learning_rate": 9.46927374301676e-05,
|
| 2270 |
+
"loss": 1.6392,
|
| 2271 |
+
"step": 377
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 2.11,
|
| 2275 |
+
"learning_rate": 9.441340782122905e-05,
|
| 2276 |
+
"loss": 1.5611,
|
| 2277 |
+
"step": 378
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 2.12,
|
| 2281 |
+
"learning_rate": 9.413407821229052e-05,
|
| 2282 |
+
"loss": 1.3384,
|
| 2283 |
+
"step": 379
|
| 2284 |
+
},
|
| 2285 |
+
{
|
| 2286 |
+
"epoch": 2.12,
|
| 2287 |
+
"learning_rate": 9.385474860335196e-05,
|
| 2288 |
+
"loss": 1.5732,
|
| 2289 |
+
"step": 380
|
| 2290 |
+
},
|
| 2291 |
+
{
|
| 2292 |
+
"epoch": 2.13,
|
| 2293 |
+
"learning_rate": 9.357541899441341e-05,
|
| 2294 |
+
"loss": 1.3874,
|
| 2295 |
+
"step": 381
|
| 2296 |
+
},
|
| 2297 |
+
{
|
| 2298 |
+
"epoch": 2.13,
|
| 2299 |
+
"learning_rate": 9.329608938547486e-05,
|
| 2300 |
+
"loss": 1.4139,
|
| 2301 |
+
"step": 382
|
| 2302 |
+
},
|
| 2303 |
+
{
|
| 2304 |
+
"epoch": 2.14,
|
| 2305 |
+
"learning_rate": 9.301675977653633e-05,
|
| 2306 |
+
"loss": 1.5195,
|
| 2307 |
+
"step": 383
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"epoch": 2.15,
|
| 2311 |
+
"learning_rate": 9.273743016759777e-05,
|
| 2312 |
+
"loss": 1.4371,
|
| 2313 |
+
"step": 384
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 2.15,
|
| 2317 |
+
"learning_rate": 9.245810055865922e-05,
|
| 2318 |
+
"loss": 1.4411,
|
| 2319 |
+
"step": 385
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 2.16,
|
| 2323 |
+
"learning_rate": 9.217877094972067e-05,
|
| 2324 |
+
"loss": 1.5214,
|
| 2325 |
+
"step": 386
|
| 2326 |
+
},
|
| 2327 |
+
{
|
| 2328 |
+
"epoch": 2.16,
|
| 2329 |
+
"learning_rate": 9.189944134078213e-05,
|
| 2330 |
+
"loss": 1.4971,
|
| 2331 |
+
"step": 387
|
| 2332 |
+
},
|
| 2333 |
+
{
|
| 2334 |
+
"epoch": 2.17,
|
| 2335 |
+
"learning_rate": 9.162011173184358e-05,
|
| 2336 |
+
"loss": 1.3453,
|
| 2337 |
+
"step": 388
|
| 2338 |
+
},
|
| 2339 |
+
{
|
| 2340 |
+
"epoch": 2.17,
|
| 2341 |
+
"learning_rate": 9.134078212290503e-05,
|
| 2342 |
+
"loss": 1.394,
|
| 2343 |
+
"step": 389
|
| 2344 |
+
},
|
| 2345 |
+
{
|
| 2346 |
+
"epoch": 2.18,
|
| 2347 |
+
"learning_rate": 9.106145251396648e-05,
|
| 2348 |
+
"loss": 1.5058,
|
| 2349 |
+
"step": 390
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 2.18,
|
| 2353 |
+
"learning_rate": 9.078212290502794e-05,
|
| 2354 |
+
"loss": 1.4855,
|
| 2355 |
+
"step": 391
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 2.19,
|
| 2359 |
+
"learning_rate": 9.050279329608939e-05,
|
| 2360 |
+
"loss": 1.4647,
|
| 2361 |
+
"step": 392
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 2.2,
|
| 2365 |
+
"learning_rate": 9.022346368715084e-05,
|
| 2366 |
+
"loss": 1.3435,
|
| 2367 |
+
"step": 393
|
| 2368 |
+
},
|
| 2369 |
+
{
|
| 2370 |
+
"epoch": 2.2,
|
| 2371 |
+
"learning_rate": 8.99441340782123e-05,
|
| 2372 |
+
"loss": 1.5815,
|
| 2373 |
+
"step": 394
|
| 2374 |
+
},
|
| 2375 |
+
{
|
| 2376 |
+
"epoch": 2.21,
|
| 2377 |
+
"learning_rate": 8.966480446927375e-05,
|
| 2378 |
+
"loss": 1.4742,
|
| 2379 |
+
"step": 395
|
| 2380 |
+
},
|
| 2381 |
+
{
|
| 2382 |
+
"epoch": 2.21,
|
| 2383 |
+
"learning_rate": 8.938547486033519e-05,
|
| 2384 |
+
"loss": 1.5389,
|
| 2385 |
+
"step": 396
|
| 2386 |
+
},
|
| 2387 |
+
{
|
| 2388 |
+
"epoch": 2.22,
|
| 2389 |
+
"learning_rate": 8.910614525139666e-05,
|
| 2390 |
+
"loss": 1.384,
|
| 2391 |
+
"step": 397
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 2.22,
|
| 2395 |
+
"learning_rate": 8.882681564245811e-05,
|
| 2396 |
+
"loss": 1.3967,
|
| 2397 |
+
"step": 398
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 2.23,
|
| 2401 |
+
"learning_rate": 8.854748603351956e-05,
|
| 2402 |
+
"loss": 1.442,
|
| 2403 |
+
"step": 399
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 2.23,
|
| 2407 |
+
"learning_rate": 8.8268156424581e-05,
|
| 2408 |
+
"loss": 1.396,
|
| 2409 |
+
"step": 400
|
| 2410 |
}
|
| 2411 |
],
|
| 2412 |
"logging_steps": 1,
|
| 2413 |
"max_steps": 716,
|
| 2414 |
"num_train_epochs": 4,
|
| 2415 |
"save_steps": 100,
|
| 2416 |
+
"total_flos": 2.056923590572032e+17,
|
| 2417 |
"trial_name": null,
|
| 2418 |
"trial_params": null
|
| 2419 |
}
|