Training in progress, step 300, 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 156926880
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dc7286aa5d3dbad64d85a654e02ec6abb323ce766baf6cab90b26ccb96ef5b88
|
| 3 |
size 156926880
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d556d7686cdc715557c46a6feac604685c6335f78a17d8925699393ed886ad08
|
| 3 |
+
size 79968964
|
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 sha256:2048aaf57bb187834c76a7432add62afe0cc9d7280cfffbad101876335a6abc0
|
| 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:d056f50c300296f18058e2a8a1b334aad77be2fed2c8f2f3b8eb3d2280d311ea
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "miner_id_24/checkpoint-
|
| 4 |
-
"epoch": 2.
|
| 5 |
"eval_steps": 50,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1805,6 +1805,364 @@
|
|
| 1805 |
"eval_samples_per_second": 28.314,
|
| 1806 |
"eval_steps_per_second": 7.113,
|
| 1807 |
"step": 250
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1808 |
}
|
| 1809 |
],
|
| 1810 |
"logging_steps": 1,
|
|
@@ -1833,7 +2191,7 @@
|
|
| 1833 |
"attributes": {}
|
| 1834 |
}
|
| 1835 |
},
|
| 1836 |
-
"total_flos":
|
| 1837 |
"train_batch_size": 8,
|
| 1838 |
"trial_name": null,
|
| 1839 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.3370027542114258,
|
| 3 |
+
"best_model_checkpoint": "miner_id_24/checkpoint-300",
|
| 4 |
+
"epoch": 2.4937759336099585,
|
| 5 |
"eval_steps": 50,
|
| 6 |
+
"global_step": 300,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1805 |
"eval_samples_per_second": 28.314,
|
| 1806 |
"eval_steps_per_second": 7.113,
|
| 1807 |
"step": 250
|
| 1808 |
+
},
|
| 1809 |
+
{
|
| 1810 |
+
"epoch": 2.087136929460581,
|
| 1811 |
+
"grad_norm": 4.143763065338135,
|
| 1812 |
+
"learning_rate": 2.233787796473432e-05,
|
| 1813 |
+
"loss": 0.0556,
|
| 1814 |
+
"step": 251
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 2.095435684647303,
|
| 1818 |
+
"grad_norm": 4.3261847496032715,
|
| 1819 |
+
"learning_rate": 2.1966197430775053e-05,
|
| 1820 |
+
"loss": 0.1249,
|
| 1821 |
+
"step": 252
|
| 1822 |
+
},
|
| 1823 |
+
{
|
| 1824 |
+
"epoch": 2.103734439834025,
|
| 1825 |
+
"grad_norm": 4.310375690460205,
|
| 1826 |
+
"learning_rate": 2.1596762663442218e-05,
|
| 1827 |
+
"loss": 0.0994,
|
| 1828 |
+
"step": 253
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 2.112033195020747,
|
| 1832 |
+
"grad_norm": 1.23174250125885,
|
| 1833 |
+
"learning_rate": 2.122960325787432e-05,
|
| 1834 |
+
"loss": 0.0315,
|
| 1835 |
+
"step": 254
|
| 1836 |
+
},
|
| 1837 |
+
{
|
| 1838 |
+
"epoch": 2.120331950207469,
|
| 1839 |
+
"grad_norm": 1.8426965475082397,
|
| 1840 |
+
"learning_rate": 2.086474862693244e-05,
|
| 1841 |
+
"loss": 0.0491,
|
| 1842 |
+
"step": 255
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 2.128630705394191,
|
| 1846 |
+
"grad_norm": 3.062448501586914,
|
| 1847 |
+
"learning_rate": 2.050222799884387e-05,
|
| 1848 |
+
"loss": 0.0968,
|
| 1849 |
+
"step": 256
|
| 1850 |
+
},
|
| 1851 |
+
{
|
| 1852 |
+
"epoch": 2.136929460580913,
|
| 1853 |
+
"grad_norm": 2.411597967147827,
|
| 1854 |
+
"learning_rate": 2.0142070414860704e-05,
|
| 1855 |
+
"loss": 0.0283,
|
| 1856 |
+
"step": 257
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 2.145228215767635,
|
| 1860 |
+
"grad_norm": 2.3557469844818115,
|
| 1861 |
+
"learning_rate": 1.9784304726933383e-05,
|
| 1862 |
+
"loss": 0.0897,
|
| 1863 |
+
"step": 258
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 2.1535269709543567,
|
| 1867 |
+
"grad_norm": 8.79699420928955,
|
| 1868 |
+
"learning_rate": 1.942895959539939e-05,
|
| 1869 |
+
"loss": 0.0328,
|
| 1870 |
+
"step": 259
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 2.1618257261410787,
|
| 1874 |
+
"grad_norm": 1.4276204109191895,
|
| 1875 |
+
"learning_rate": 1.9076063486687256e-05,
|
| 1876 |
+
"loss": 0.016,
|
| 1877 |
+
"step": 260
|
| 1878 |
+
},
|
| 1879 |
+
{
|
| 1880 |
+
"epoch": 2.1701244813278007,
|
| 1881 |
+
"grad_norm": 2.912771701812744,
|
| 1882 |
+
"learning_rate": 1.8725644671036126e-05,
|
| 1883 |
+
"loss": 0.1233,
|
| 1884 |
+
"step": 261
|
| 1885 |
+
},
|
| 1886 |
+
{
|
| 1887 |
+
"epoch": 2.1784232365145226,
|
| 1888 |
+
"grad_norm": 4.887458801269531,
|
| 1889 |
+
"learning_rate": 1.837773122023114e-05,
|
| 1890 |
+
"loss": 0.0717,
|
| 1891 |
+
"step": 262
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"epoch": 2.186721991701245,
|
| 1895 |
+
"grad_norm": 1.242602825164795,
|
| 1896 |
+
"learning_rate": 1.803235100535452e-05,
|
| 1897 |
+
"loss": 0.0208,
|
| 1898 |
+
"step": 263
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 2.195020746887967,
|
| 1902 |
+
"grad_norm": 0.9755971431732178,
|
| 1903 |
+
"learning_rate": 1.7689531694552863e-05,
|
| 1904 |
+
"loss": 0.0192,
|
| 1905 |
+
"step": 264
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 2.203319502074689,
|
| 1909 |
+
"grad_norm": 2.5533509254455566,
|
| 1910 |
+
"learning_rate": 1.734930075082076e-05,
|
| 1911 |
+
"loss": 0.0384,
|
| 1912 |
+
"step": 265
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"epoch": 2.211618257261411,
|
| 1916 |
+
"grad_norm": 0.9231969118118286,
|
| 1917 |
+
"learning_rate": 1.7011685429800595e-05,
|
| 1918 |
+
"loss": 0.0152,
|
| 1919 |
+
"step": 266
|
| 1920 |
+
},
|
| 1921 |
+
{
|
| 1922 |
+
"epoch": 2.219917012448133,
|
| 1923 |
+
"grad_norm": 1.7765648365020752,
|
| 1924 |
+
"learning_rate": 1.6676712777599273e-05,
|
| 1925 |
+
"loss": 0.041,
|
| 1926 |
+
"step": 267
|
| 1927 |
+
},
|
| 1928 |
+
{
|
| 1929 |
+
"epoch": 2.228215767634855,
|
| 1930 |
+
"grad_norm": 2.0227739810943604,
|
| 1931 |
+
"learning_rate": 1.6344409628621484e-05,
|
| 1932 |
+
"loss": 0.1056,
|
| 1933 |
+
"step": 268
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"epoch": 2.236514522821577,
|
| 1937 |
+
"grad_norm": 1.8746280670166016,
|
| 1938 |
+
"learning_rate": 1.6014802603420044e-05,
|
| 1939 |
+
"loss": 0.0542,
|
| 1940 |
+
"step": 269
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 2.2448132780082988,
|
| 1944 |
+
"grad_norm": 1.8291066884994507,
|
| 1945 |
+
"learning_rate": 1.5687918106563326e-05,
|
| 1946 |
+
"loss": 0.0206,
|
| 1947 |
+
"step": 270
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 2.2531120331950207,
|
| 1951 |
+
"grad_norm": 7.360498905181885,
|
| 1952 |
+
"learning_rate": 1.536378232452003e-05,
|
| 1953 |
+
"loss": 0.1436,
|
| 1954 |
+
"step": 271
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 2.2614107883817427,
|
| 1958 |
+
"grad_norm": 4.359367847442627,
|
| 1959 |
+
"learning_rate": 1.5042421223561432e-05,
|
| 1960 |
+
"loss": 0.075,
|
| 1961 |
+
"step": 272
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 2.2697095435684647,
|
| 1965 |
+
"grad_norm": 1.2709429264068604,
|
| 1966 |
+
"learning_rate": 1.4723860547681162e-05,
|
| 1967 |
+
"loss": 0.0172,
|
| 1968 |
+
"step": 273
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 2.2780082987551866,
|
| 1972 |
+
"grad_norm": 3.57598876953125,
|
| 1973 |
+
"learning_rate": 1.440812581653298e-05,
|
| 1974 |
+
"loss": 0.0619,
|
| 1975 |
+
"step": 274
|
| 1976 |
+
},
|
| 1977 |
+
{
|
| 1978 |
+
"epoch": 2.2863070539419086,
|
| 1979 |
+
"grad_norm": 1.036085605621338,
|
| 1980 |
+
"learning_rate": 1.4095242323386303e-05,
|
| 1981 |
+
"loss": 0.0604,
|
| 1982 |
+
"step": 275
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 2.2946058091286305,
|
| 1986 |
+
"grad_norm": 2.2307868003845215,
|
| 1987 |
+
"learning_rate": 1.3785235133100088e-05,
|
| 1988 |
+
"loss": 0.037,
|
| 1989 |
+
"step": 276
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 2.3029045643153525,
|
| 1993 |
+
"grad_norm": 1.6419095993041992,
|
| 1994 |
+
"learning_rate": 1.3478129080114848e-05,
|
| 1995 |
+
"loss": 0.0735,
|
| 1996 |
+
"step": 277
|
| 1997 |
+
},
|
| 1998 |
+
{
|
| 1999 |
+
"epoch": 2.3112033195020745,
|
| 2000 |
+
"grad_norm": 2.744889259338379,
|
| 2001 |
+
"learning_rate": 1.3173948766463145e-05,
|
| 2002 |
+
"loss": 0.0294,
|
| 2003 |
+
"step": 278
|
| 2004 |
+
},
|
| 2005 |
+
{
|
| 2006 |
+
"epoch": 2.3195020746887964,
|
| 2007 |
+
"grad_norm": 3.0023012161254883,
|
| 2008 |
+
"learning_rate": 1.2872718559798853e-05,
|
| 2009 |
+
"loss": 0.0446,
|
| 2010 |
+
"step": 279
|
| 2011 |
+
},
|
| 2012 |
+
{
|
| 2013 |
+
"epoch": 2.327800829875519,
|
| 2014 |
+
"grad_norm": 4.369894027709961,
|
| 2015 |
+
"learning_rate": 1.257446259144494e-05,
|
| 2016 |
+
"loss": 0.0467,
|
| 2017 |
+
"step": 280
|
| 2018 |
+
},
|
| 2019 |
+
{
|
| 2020 |
+
"epoch": 2.336099585062241,
|
| 2021 |
+
"grad_norm": 1.3997019529342651,
|
| 2022 |
+
"learning_rate": 1.2279204754460493e-05,
|
| 2023 |
+
"loss": 0.0205,
|
| 2024 |
+
"step": 281
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 2.3443983402489628,
|
| 2028 |
+
"grad_norm": 3.3227059841156006,
|
| 2029 |
+
"learning_rate": 1.1986968701726491e-05,
|
| 2030 |
+
"loss": 0.083,
|
| 2031 |
+
"step": 282
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 2.3526970954356847,
|
| 2035 |
+
"grad_norm": 2.5781474113464355,
|
| 2036 |
+
"learning_rate": 1.1697777844051105e-05,
|
| 2037 |
+
"loss": 0.0682,
|
| 2038 |
+
"step": 283
|
| 2039 |
+
},
|
| 2040 |
+
{
|
| 2041 |
+
"epoch": 2.3609958506224067,
|
| 2042 |
+
"grad_norm": 0.8888638019561768,
|
| 2043 |
+
"learning_rate": 1.1411655348294247e-05,
|
| 2044 |
+
"loss": 0.0169,
|
| 2045 |
+
"step": 284
|
| 2046 |
+
},
|
| 2047 |
+
{
|
| 2048 |
+
"epoch": 2.3692946058091287,
|
| 2049 |
+
"grad_norm": 3.076343059539795,
|
| 2050 |
+
"learning_rate": 1.1128624135511712e-05,
|
| 2051 |
+
"loss": 0.0746,
|
| 2052 |
+
"step": 285
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"epoch": 2.3775933609958506,
|
| 2056 |
+
"grad_norm": 2.557579755783081,
|
| 2057 |
+
"learning_rate": 1.0848706879118892e-05,
|
| 2058 |
+
"loss": 0.0715,
|
| 2059 |
+
"step": 286
|
| 2060 |
+
},
|
| 2061 |
+
{
|
| 2062 |
+
"epoch": 2.3858921161825726,
|
| 2063 |
+
"grad_norm": 3.306203842163086,
|
| 2064 |
+
"learning_rate": 1.0571926003074561e-05,
|
| 2065 |
+
"loss": 0.063,
|
| 2066 |
+
"step": 287
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 2.3941908713692945,
|
| 2070 |
+
"grad_norm": 1.7703243494033813,
|
| 2071 |
+
"learning_rate": 1.0298303680084448e-05,
|
| 2072 |
+
"loss": 0.0303,
|
| 2073 |
+
"step": 288
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 2.4024896265560165,
|
| 2077 |
+
"grad_norm": 3.1793651580810547,
|
| 2078 |
+
"learning_rate": 1.0027861829824952e-05,
|
| 2079 |
+
"loss": 0.0253,
|
| 2080 |
+
"step": 289
|
| 2081 |
+
},
|
| 2082 |
+
{
|
| 2083 |
+
"epoch": 2.4107883817427385,
|
| 2084 |
+
"grad_norm": 2.1019341945648193,
|
| 2085 |
+
"learning_rate": 9.760622117187235e-06,
|
| 2086 |
+
"loss": 0.0759,
|
| 2087 |
+
"step": 290
|
| 2088 |
+
},
|
| 2089 |
+
{
|
| 2090 |
+
"epoch": 2.4190871369294604,
|
| 2091 |
+
"grad_norm": 1.9458683729171753,
|
| 2092 |
+
"learning_rate": 9.496605950541676e-06,
|
| 2093 |
+
"loss": 0.0259,
|
| 2094 |
+
"step": 291
|
| 2095 |
+
},
|
| 2096 |
+
{
|
| 2097 |
+
"epoch": 2.4273858921161824,
|
| 2098 |
+
"grad_norm": 3.2663235664367676,
|
| 2099 |
+
"learning_rate": 9.235834480022787e-06,
|
| 2100 |
+
"loss": 0.071,
|
| 2101 |
+
"step": 292
|
| 2102 |
+
},
|
| 2103 |
+
{
|
| 2104 |
+
"epoch": 2.435684647302905,
|
| 2105 |
+
"grad_norm": 2.9131743907928467,
|
| 2106 |
+
"learning_rate": 8.978328595834984e-06,
|
| 2107 |
+
"loss": 0.0265,
|
| 2108 |
+
"step": 293
|
| 2109 |
+
},
|
| 2110 |
+
{
|
| 2111 |
+
"epoch": 2.4439834024896268,
|
| 2112 |
+
"grad_norm": 4.122836589813232,
|
| 2113 |
+
"learning_rate": 8.724108926578999e-06,
|
| 2114 |
+
"loss": 0.0449,
|
| 2115 |
+
"step": 294
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 2.4522821576763487,
|
| 2119 |
+
"grad_norm": 2.857977867126465,
|
| 2120 |
+
"learning_rate": 8.473195837599418e-06,
|
| 2121 |
+
"loss": 0.0248,
|
| 2122 |
+
"step": 295
|
| 2123 |
+
},
|
| 2124 |
+
{
|
| 2125 |
+
"epoch": 2.4605809128630707,
|
| 2126 |
+
"grad_norm": 4.170780181884766,
|
| 2127 |
+
"learning_rate": 8.225609429353187e-06,
|
| 2128 |
+
"loss": 0.0373,
|
| 2129 |
+
"step": 296
|
| 2130 |
+
},
|
| 2131 |
+
{
|
| 2132 |
+
"epoch": 2.4688796680497926,
|
| 2133 |
+
"grad_norm": 2.490182638168335,
|
| 2134 |
+
"learning_rate": 7.981369535799354e-06,
|
| 2135 |
+
"loss": 0.0751,
|
| 2136 |
+
"step": 297
|
| 2137 |
+
},
|
| 2138 |
+
{
|
| 2139 |
+
"epoch": 2.4771784232365146,
|
| 2140 |
+
"grad_norm": 3.713991165161133,
|
| 2141 |
+
"learning_rate": 7.740495722810271e-06,
|
| 2142 |
+
"loss": 0.0964,
|
| 2143 |
+
"step": 298
|
| 2144 |
+
},
|
| 2145 |
+
{
|
| 2146 |
+
"epoch": 2.4854771784232366,
|
| 2147 |
+
"grad_norm": 3.0123841762542725,
|
| 2148 |
+
"learning_rate": 7.503007286604069e-06,
|
| 2149 |
+
"loss": 0.0414,
|
| 2150 |
+
"step": 299
|
| 2151 |
+
},
|
| 2152 |
+
{
|
| 2153 |
+
"epoch": 2.4937759336099585,
|
| 2154 |
+
"grad_norm": 4.080419063568115,
|
| 2155 |
+
"learning_rate": 7.268923252198989e-06,
|
| 2156 |
+
"loss": 0.1049,
|
| 2157 |
+
"step": 300
|
| 2158 |
+
},
|
| 2159 |
+
{
|
| 2160 |
+
"epoch": 2.4937759336099585,
|
| 2161 |
+
"eval_loss": 0.3370027542114258,
|
| 2162 |
+
"eval_runtime": 7.1786,
|
| 2163 |
+
"eval_samples_per_second": 28.278,
|
| 2164 |
+
"eval_steps_per_second": 7.104,
|
| 2165 |
+
"step": 300
|
| 2166 |
}
|
| 2167 |
],
|
| 2168 |
"logging_steps": 1,
|
|
|
|
| 2191 |
"attributes": {}
|
| 2192 |
}
|
| 2193 |
},
|
| 2194 |
+
"total_flos": 1.192097835122688e+17,
|
| 2195 |
"train_batch_size": 8,
|
| 2196 |
"trial_name": null,
|
| 2197 |
"trial_params": null
|