Training in progress, step 319, 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 237546776
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:01a2408b511d3e72357df79b7a07ad06ad9230d6c19b84c21feaf195f605743f
|
| 3 |
size 237546776
|
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 sha256:9468afcac283b917fc2c10feb94175b8ad9038c8b7ba16949e23ec5bb2939333
|
| 3 |
+
size 223237720
|
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 sha256:37e80c0bb7385d41535be59c31232d72d36d5990a511e813fe84cb14b3359947
|
| 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 sha256:4222df64063a0da99a3d9b170b238ff21de6b814178c3cb5d1a90ab8274aa2a3
|
| 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":
|
| 5 |
"eval_steps": 80,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1711,6 +1711,559 @@
|
|
| 1711 |
"eval_samples_per_second": 43.102,
|
| 1712 |
"eval_steps_per_second": 21.551,
|
| 1713 |
"step": 240
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1714 |
}
|
| 1715 |
],
|
| 1716 |
"logging_steps": 1,
|
|
@@ -1725,12 +2278,12 @@
|
|
| 1725 |
"should_evaluate": false,
|
| 1726 |
"should_log": false,
|
| 1727 |
"should_save": true,
|
| 1728 |
-
"should_training_stop":
|
| 1729 |
},
|
| 1730 |
"attributes": {}
|
| 1731 |
}
|
| 1732 |
},
|
| 1733 |
-
"total_flos": 2.
|
| 1734 |
"train_batch_size": 2,
|
| 1735 |
"trial_name": null,
|
| 1736 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.002356637863315,
|
| 5 |
"eval_steps": 80,
|
| 6 |
+
"global_step": 319,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1711 |
"eval_samples_per_second": 43.102,
|
| 1712 |
"eval_steps_per_second": 21.551,
|
| 1713 |
"step": 240
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 0.7572663000785546,
|
| 1717 |
+
"grad_norm": 4.266700744628906,
|
| 1718 |
+
"learning_rate": 2.98305571716907e-05,
|
| 1719 |
+
"loss": 1.1874,
|
| 1720 |
+
"step": 241
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 0.7604084838963079,
|
| 1724 |
+
"grad_norm": 5.862624168395996,
|
| 1725 |
+
"learning_rate": 2.9109822211964043e-05,
|
| 1726 |
+
"loss": 1.0646,
|
| 1727 |
+
"step": 242
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 0.7635506677140613,
|
| 1731 |
+
"grad_norm": 4.90443754196167,
|
| 1732 |
+
"learning_rate": 2.8396414908880098e-05,
|
| 1733 |
+
"loss": 1.3927,
|
| 1734 |
+
"step": 243
|
| 1735 |
+
},
|
| 1736 |
+
{
|
| 1737 |
+
"epoch": 0.7666928515318147,
|
| 1738 |
+
"grad_norm": 6.397622585296631,
|
| 1739 |
+
"learning_rate": 2.769040900472488e-05,
|
| 1740 |
+
"loss": 1.2935,
|
| 1741 |
+
"step": 244
|
| 1742 |
+
},
|
| 1743 |
+
{
|
| 1744 |
+
"epoch": 0.769835035349568,
|
| 1745 |
+
"grad_norm": 5.6756591796875,
|
| 1746 |
+
"learning_rate": 2.699187747672899e-05,
|
| 1747 |
+
"loss": 1.2046,
|
| 1748 |
+
"step": 245
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"epoch": 0.7729772191673213,
|
| 1752 |
+
"grad_norm": 5.524738788604736,
|
| 1753 |
+
"learning_rate": 2.630089252952427e-05,
|
| 1754 |
+
"loss": 1.0745,
|
| 1755 |
+
"step": 246
|
| 1756 |
+
},
|
| 1757 |
+
{
|
| 1758 |
+
"epoch": 0.7761194029850746,
|
| 1759 |
+
"grad_norm": 5.8002119064331055,
|
| 1760 |
+
"learning_rate": 2.5617525587680402e-05,
|
| 1761 |
+
"loss": 1.1491,
|
| 1762 |
+
"step": 247
|
| 1763 |
+
},
|
| 1764 |
+
{
|
| 1765 |
+
"epoch": 0.779261586802828,
|
| 1766 |
+
"grad_norm": 6.8815484046936035,
|
| 1767 |
+
"learning_rate": 2.4941847288321797e-05,
|
| 1768 |
+
"loss": 1.5067,
|
| 1769 |
+
"step": 248
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 0.7824037706205813,
|
| 1773 |
+
"grad_norm": 7.29502010345459,
|
| 1774 |
+
"learning_rate": 2.427392747382623e-05,
|
| 1775 |
+
"loss": 1.5713,
|
| 1776 |
+
"step": 249
|
| 1777 |
+
},
|
| 1778 |
+
{
|
| 1779 |
+
"epoch": 0.7855459544383346,
|
| 1780 |
+
"grad_norm": 8.774285316467285,
|
| 1781 |
+
"learning_rate": 2.3613835184605525e-05,
|
| 1782 |
+
"loss": 1.701,
|
| 1783 |
+
"step": 250
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"epoch": 0.788688138256088,
|
| 1787 |
+
"grad_norm": 2.379939556121826,
|
| 1788 |
+
"learning_rate": 2.2961638651968975e-05,
|
| 1789 |
+
"loss": 1.0794,
|
| 1790 |
+
"step": 251
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 0.7918303220738413,
|
| 1794 |
+
"grad_norm": 3.104327440261841,
|
| 1795 |
+
"learning_rate": 2.231740529107057e-05,
|
| 1796 |
+
"loss": 1.4833,
|
| 1797 |
+
"step": 252
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"epoch": 0.7949725058915946,
|
| 1801 |
+
"grad_norm": 2.5866000652313232,
|
| 1802 |
+
"learning_rate": 2.1681201693940668e-05,
|
| 1803 |
+
"loss": 1.2774,
|
| 1804 |
+
"step": 253
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 0.798114689709348,
|
| 1808 |
+
"grad_norm": 2.862161874771118,
|
| 1809 |
+
"learning_rate": 2.1053093622602404e-05,
|
| 1810 |
+
"loss": 1.2606,
|
| 1811 |
+
"step": 254
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"epoch": 0.8012568735271013,
|
| 1815 |
+
"grad_norm": 3.9734158515930176,
|
| 1816 |
+
"learning_rate": 2.043314600227425e-05,
|
| 1817 |
+
"loss": 1.4184,
|
| 1818 |
+
"step": 255
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 0.8043990573448547,
|
| 1822 |
+
"grad_norm": 2.7739901542663574,
|
| 1823 |
+
"learning_rate": 1.982142291465896e-05,
|
| 1824 |
+
"loss": 1.0968,
|
| 1825 |
+
"step": 256
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 0.8075412411626081,
|
| 1829 |
+
"grad_norm": 2.9346413612365723,
|
| 1830 |
+
"learning_rate": 1.921798759131953e-05,
|
| 1831 |
+
"loss": 1.4055,
|
| 1832 |
+
"step": 257
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 0.8106834249803614,
|
| 1836 |
+
"grad_norm": 3.191915273666382,
|
| 1837 |
+
"learning_rate": 1.8622902407143394e-05,
|
| 1838 |
+
"loss": 1.2101,
|
| 1839 |
+
"step": 258
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 0.8138256087981147,
|
| 1843 |
+
"grad_norm": 3.06551456451416,
|
| 1844 |
+
"learning_rate": 1.8036228873894746e-05,
|
| 1845 |
+
"loss": 1.1478,
|
| 1846 |
+
"step": 259
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"epoch": 0.816967792615868,
|
| 1850 |
+
"grad_norm": 3.3214335441589355,
|
| 1851 |
+
"learning_rate": 1.7458027633856478e-05,
|
| 1852 |
+
"loss": 1.2364,
|
| 1853 |
+
"step": 260
|
| 1854 |
+
},
|
| 1855 |
+
{
|
| 1856 |
+
"epoch": 0.8201099764336214,
|
| 1857 |
+
"grad_norm": 3.1710729598999023,
|
| 1858 |
+
"learning_rate": 1.6888358453561648e-05,
|
| 1859 |
+
"loss": 1.0582,
|
| 1860 |
+
"step": 261
|
| 1861 |
+
},
|
| 1862 |
+
{
|
| 1863 |
+
"epoch": 0.8232521602513747,
|
| 1864 |
+
"grad_norm": 3.025092363357544,
|
| 1865 |
+
"learning_rate": 1.6327280217615792e-05,
|
| 1866 |
+
"loss": 1.2433,
|
| 1867 |
+
"step": 262
|
| 1868 |
+
},
|
| 1869 |
+
{
|
| 1870 |
+
"epoch": 0.826394344069128,
|
| 1871 |
+
"grad_norm": 3.2119741439819336,
|
| 1872 |
+
"learning_rate": 1.577485092261012e-05,
|
| 1873 |
+
"loss": 1.2365,
|
| 1874 |
+
"step": 263
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 0.8295365278868814,
|
| 1878 |
+
"grad_norm": 3.112516403198242,
|
| 1879 |
+
"learning_rate": 1.5231127671126677e-05,
|
| 1880 |
+
"loss": 1.0492,
|
| 1881 |
+
"step": 264
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"epoch": 0.8326787117046347,
|
| 1885 |
+
"grad_norm": 2.9350504875183105,
|
| 1886 |
+
"learning_rate": 1.4696166665835853e-05,
|
| 1887 |
+
"loss": 1.1789,
|
| 1888 |
+
"step": 265
|
| 1889 |
+
},
|
| 1890 |
+
{
|
| 1891 |
+
"epoch": 0.835820895522388,
|
| 1892 |
+
"grad_norm": 3.688014030456543,
|
| 1893 |
+
"learning_rate": 1.4170023203686878e-05,
|
| 1894 |
+
"loss": 1.3643,
|
| 1895 |
+
"step": 266
|
| 1896 |
+
},
|
| 1897 |
+
{
|
| 1898 |
+
"epoch": 0.8389630793401414,
|
| 1899 |
+
"grad_norm": 3.518484115600586,
|
| 1900 |
+
"learning_rate": 1.3652751670192077e-05,
|
| 1901 |
+
"loss": 1.2082,
|
| 1902 |
+
"step": 267
|
| 1903 |
+
},
|
| 1904 |
+
{
|
| 1905 |
+
"epoch": 0.8421052631578947,
|
| 1906 |
+
"grad_norm": 3.1499722003936768,
|
| 1907 |
+
"learning_rate": 1.3144405533805138e-05,
|
| 1908 |
+
"loss": 0.8437,
|
| 1909 |
+
"step": 268
|
| 1910 |
+
},
|
| 1911 |
+
{
|
| 1912 |
+
"epoch": 0.845247446975648,
|
| 1913 |
+
"grad_norm": 3.3172833919525146,
|
| 1914 |
+
"learning_rate": 1.2645037340394284e-05,
|
| 1915 |
+
"loss": 1.1557,
|
| 1916 |
+
"step": 269
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 0.8483896307934015,
|
| 1920 |
+
"grad_norm": 3.544373035430908,
|
| 1921 |
+
"learning_rate": 1.215469870781093e-05,
|
| 1922 |
+
"loss": 1.1555,
|
| 1923 |
+
"step": 270
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 0.8515318146111548,
|
| 1927 |
+
"grad_norm": 3.739354133605957,
|
| 1928 |
+
"learning_rate": 1.167344032055394e-05,
|
| 1929 |
+
"loss": 1.3633,
|
| 1930 |
+
"step": 271
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 0.8546739984289081,
|
| 1934 |
+
"grad_norm": 3.5768179893493652,
|
| 1935 |
+
"learning_rate": 1.120131192453069e-05,
|
| 1936 |
+
"loss": 1.0143,
|
| 1937 |
+
"step": 272
|
| 1938 |
+
},
|
| 1939 |
+
{
|
| 1940 |
+
"epoch": 0.8578161822466615,
|
| 1941 |
+
"grad_norm": 3.85493540763855,
|
| 1942 |
+
"learning_rate": 1.0738362321914997e-05,
|
| 1943 |
+
"loss": 1.3486,
|
| 1944 |
+
"step": 273
|
| 1945 |
+
},
|
| 1946 |
+
{
|
| 1947 |
+
"epoch": 0.8609583660644148,
|
| 1948 |
+
"grad_norm": 3.8633315563201904,
|
| 1949 |
+
"learning_rate": 1.02846393661026e-05,
|
| 1950 |
+
"loss": 1.2283,
|
| 1951 |
+
"step": 274
|
| 1952 |
+
},
|
| 1953 |
+
{
|
| 1954 |
+
"epoch": 0.8641005498821681,
|
| 1955 |
+
"grad_norm": 3.5946390628814697,
|
| 1956 |
+
"learning_rate": 9.840189956764677e-06,
|
| 1957 |
+
"loss": 1.2316,
|
| 1958 |
+
"step": 275
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 0.8672427336999214,
|
| 1962 |
+
"grad_norm": 4.018091201782227,
|
| 1963 |
+
"learning_rate": 9.405060035000135e-06,
|
| 1964 |
+
"loss": 1.0729,
|
| 1965 |
+
"step": 276
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 0.8703849175176748,
|
| 1969 |
+
"grad_norm": 3.750617265701294,
|
| 1970 |
+
"learning_rate": 8.979294578586738e-06,
|
| 1971 |
+
"loss": 0.983,
|
| 1972 |
+
"step": 277
|
| 1973 |
+
},
|
| 1974 |
+
{
|
| 1975 |
+
"epoch": 0.8735271013354281,
|
| 1976 |
+
"grad_norm": 4.0061936378479,
|
| 1977 |
+
"learning_rate": 8.562937597331899e-06,
|
| 1978 |
+
"loss": 1.4571,
|
| 1979 |
+
"step": 278
|
| 1980 |
+
},
|
| 1981 |
+
{
|
| 1982 |
+
"epoch": 0.8766692851531814,
|
| 1983 |
+
"grad_norm": 4.811827182769775,
|
| 1984 |
+
"learning_rate": 8.156032128523694e-06,
|
| 1985 |
+
"loss": 1.5216,
|
| 1986 |
+
"step": 279
|
| 1987 |
+
},
|
| 1988 |
+
{
|
| 1989 |
+
"epoch": 0.8798114689709348,
|
| 1990 |
+
"grad_norm": 3.8524036407470703,
|
| 1991 |
+
"learning_rate": 7.758620232482084e-06,
|
| 1992 |
+
"loss": 1.1154,
|
| 1993 |
+
"step": 280
|
| 1994 |
+
},
|
| 1995 |
+
{
|
| 1996 |
+
"epoch": 0.8829536527886881,
|
| 1997 |
+
"grad_norm": 4.497751712799072,
|
| 1998 |
+
"learning_rate": 7.370742988211365e-06,
|
| 1999 |
+
"loss": 1.3045,
|
| 2000 |
+
"step": 281
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 0.8860958366064414,
|
| 2004 |
+
"grad_norm": 3.6167821884155273,
|
| 2005 |
+
"learning_rate": 6.992440489154051e-06,
|
| 2006 |
+
"loss": 1.2766,
|
| 2007 |
+
"step": 282
|
| 2008 |
+
},
|
| 2009 |
+
{
|
| 2010 |
+
"epoch": 0.8892380204241949,
|
| 2011 |
+
"grad_norm": 4.365938663482666,
|
| 2012 |
+
"learning_rate": 6.623751839046455e-06,
|
| 2013 |
+
"loss": 1.1343,
|
| 2014 |
+
"step": 283
|
| 2015 |
+
},
|
| 2016 |
+
{
|
| 2017 |
+
"epoch": 0.8923802042419482,
|
| 2018 |
+
"grad_norm": 3.732172966003418,
|
| 2019 |
+
"learning_rate": 6.264715147876743e-06,
|
| 2020 |
+
"loss": 1.0509,
|
| 2021 |
+
"step": 284
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"epoch": 0.8955223880597015,
|
| 2025 |
+
"grad_norm": 5.462162494659424,
|
| 2026 |
+
"learning_rate": 5.915367527945615e-06,
|
| 2027 |
+
"loss": 1.4873,
|
| 2028 |
+
"step": 285
|
| 2029 |
+
},
|
| 2030 |
+
{
|
| 2031 |
+
"epoch": 0.8986645718774549,
|
| 2032 |
+
"grad_norm": 3.5439860820770264,
|
| 2033 |
+
"learning_rate": 5.575745090030138e-06,
|
| 2034 |
+
"loss": 0.9738,
|
| 2035 |
+
"step": 286
|
| 2036 |
+
},
|
| 2037 |
+
{
|
| 2038 |
+
"epoch": 0.9018067556952082,
|
| 2039 |
+
"grad_norm": 5.106057167053223,
|
| 2040 |
+
"learning_rate": 5.245882939651181e-06,
|
| 2041 |
+
"loss": 1.6314,
|
| 2042 |
+
"step": 287
|
| 2043 |
+
},
|
| 2044 |
+
{
|
| 2045 |
+
"epoch": 0.9049489395129615,
|
| 2046 |
+
"grad_norm": 4.478841781616211,
|
| 2047 |
+
"learning_rate": 4.92581517344457e-06,
|
| 2048 |
+
"loss": 1.3549,
|
| 2049 |
+
"step": 288
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 0.9080911233307148,
|
| 2053 |
+
"grad_norm": 4.549522876739502,
|
| 2054 |
+
"learning_rate": 4.61557487563673e-06,
|
| 2055 |
+
"loss": 1.1305,
|
| 2056 |
+
"step": 289
|
| 2057 |
+
},
|
| 2058 |
+
{
|
| 2059 |
+
"epoch": 0.9112333071484682,
|
| 2060 |
+
"grad_norm": 5.698983669281006,
|
| 2061 |
+
"learning_rate": 4.315194114624888e-06,
|
| 2062 |
+
"loss": 1.2778,
|
| 2063 |
+
"step": 290
|
| 2064 |
+
},
|
| 2065 |
+
{
|
| 2066 |
+
"epoch": 0.9143754909662215,
|
| 2067 |
+
"grad_norm": 5.114129066467285,
|
| 2068 |
+
"learning_rate": 4.0247039396622e-06,
|
| 2069 |
+
"loss": 1.364,
|
| 2070 |
+
"step": 291
|
| 2071 |
+
},
|
| 2072 |
+
{
|
| 2073 |
+
"epoch": 0.9175176747839748,
|
| 2074 |
+
"grad_norm": 5.043893337249756,
|
| 2075 |
+
"learning_rate": 3.7441343776484117e-06,
|
| 2076 |
+
"loss": 1.6245,
|
| 2077 |
+
"step": 292
|
| 2078 |
+
},
|
| 2079 |
+
{
|
| 2080 |
+
"epoch": 0.9206598586017282,
|
| 2081 |
+
"grad_norm": 6.31735897064209,
|
| 2082 |
+
"learning_rate": 3.473514430026026e-06,
|
| 2083 |
+
"loss": 1.0683,
|
| 2084 |
+
"step": 293
|
| 2085 |
+
},
|
| 2086 |
+
{
|
| 2087 |
+
"epoch": 0.9238020424194815,
|
| 2088 |
+
"grad_norm": 5.165215015411377,
|
| 2089 |
+
"learning_rate": 3.212872069782513e-06,
|
| 2090 |
+
"loss": 1.6659,
|
| 2091 |
+
"step": 294
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 0.9269442262372348,
|
| 2095 |
+
"grad_norm": 4.504268169403076,
|
| 2096 |
+
"learning_rate": 2.9622342385589254e-06,
|
| 2097 |
+
"loss": 1.322,
|
| 2098 |
+
"step": 295
|
| 2099 |
+
},
|
| 2100 |
+
{
|
| 2101 |
+
"epoch": 0.9300864100549883,
|
| 2102 |
+
"grad_norm": 5.05787467956543,
|
| 2103 |
+
"learning_rate": 2.7216268438649773e-06,
|
| 2104 |
+
"loss": 1.0772,
|
| 2105 |
+
"step": 296
|
| 2106 |
+
},
|
| 2107 |
+
{
|
| 2108 |
+
"epoch": 0.9332285938727416,
|
| 2109 |
+
"grad_norm": 6.266513347625732,
|
| 2110 |
+
"learning_rate": 2.4910747564010685e-06,
|
| 2111 |
+
"loss": 1.1698,
|
| 2112 |
+
"step": 297
|
| 2113 |
+
},
|
| 2114 |
+
{
|
| 2115 |
+
"epoch": 0.9363707776904949,
|
| 2116 |
+
"grad_norm": 5.642542362213135,
|
| 2117 |
+
"learning_rate": 2.2706018074875045e-06,
|
| 2118 |
+
"loss": 1.2399,
|
| 2119 |
+
"step": 298
|
| 2120 |
+
},
|
| 2121 |
+
{
|
| 2122 |
+
"epoch": 0.9395129615082483,
|
| 2123 |
+
"grad_norm": 6.315053462982178,
|
| 2124 |
+
"learning_rate": 2.060230786601225e-06,
|
| 2125 |
+
"loss": 1.4454,
|
| 2126 |
+
"step": 299
|
| 2127 |
+
},
|
| 2128 |
+
{
|
| 2129 |
+
"epoch": 0.9426551453260016,
|
| 2130 |
+
"grad_norm": 7.884403705596924,
|
| 2131 |
+
"learning_rate": 1.8599834390199855e-06,
|
| 2132 |
+
"loss": 1.3084,
|
| 2133 |
+
"step": 300
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 0.9457973291437549,
|
| 2137 |
+
"grad_norm": 2.967442512512207,
|
| 2138 |
+
"learning_rate": 1.6698804635747579e-06,
|
| 2139 |
+
"loss": 1.3336,
|
| 2140 |
+
"step": 301
|
| 2141 |
+
},
|
| 2142 |
+
{
|
| 2143 |
+
"epoch": 0.9489395129615082,
|
| 2144 |
+
"grad_norm": 3.1417999267578125,
|
| 2145 |
+
"learning_rate": 1.4899415105101067e-06,
|
| 2146 |
+
"loss": 1.3268,
|
| 2147 |
+
"step": 302
|
| 2148 |
+
},
|
| 2149 |
+
{
|
| 2150 |
+
"epoch": 0.9520816967792616,
|
| 2151 |
+
"grad_norm": 2.946615695953369,
|
| 2152 |
+
"learning_rate": 1.3201851794530373e-06,
|
| 2153 |
+
"loss": 1.4632,
|
| 2154 |
+
"step": 303
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"epoch": 0.9552238805970149,
|
| 2158 |
+
"grad_norm": 2.756666660308838,
|
| 2159 |
+
"learning_rate": 1.160629017490389e-06,
|
| 2160 |
+
"loss": 0.9778,
|
| 2161 |
+
"step": 304
|
| 2162 |
+
},
|
| 2163 |
+
{
|
| 2164 |
+
"epoch": 0.9583660644147682,
|
| 2165 |
+
"grad_norm": 3.684561252593994,
|
| 2166 |
+
"learning_rate": 1.0112895173551185e-06,
|
| 2167 |
+
"loss": 1.5126,
|
| 2168 |
+
"step": 305
|
| 2169 |
+
},
|
| 2170 |
+
{
|
| 2171 |
+
"epoch": 0.9615082482325216,
|
| 2172 |
+
"grad_norm": 3.060070514678955,
|
| 2173 |
+
"learning_rate": 8.721821157214316e-07,
|
| 2174 |
+
"loss": 1.1442,
|
| 2175 |
+
"step": 306
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 0.9646504320502749,
|
| 2179 |
+
"grad_norm": 3.5914061069488525,
|
| 2180 |
+
"learning_rate": 7.433211916092142e-07,
|
| 2181 |
+
"loss": 0.918,
|
| 2182 |
+
"step": 307
|
| 2183 |
+
},
|
| 2184 |
+
{
|
| 2185 |
+
"epoch": 0.9677926158680282,
|
| 2186 |
+
"grad_norm": 3.4093003273010254,
|
| 2187 |
+
"learning_rate": 6.247200648976991e-07,
|
| 2188 |
+
"loss": 1.2567,
|
| 2189 |
+
"step": 308
|
| 2190 |
+
},
|
| 2191 |
+
{
|
| 2192 |
+
"epoch": 0.9709347996857817,
|
| 2193 |
+
"grad_norm": 4.0583600997924805,
|
| 2194 |
+
"learning_rate": 5.163909949486234e-07,
|
| 2195 |
+
"loss": 1.2847,
|
| 2196 |
+
"step": 309
|
| 2197 |
+
},
|
| 2198 |
+
{
|
| 2199 |
+
"epoch": 0.974076983503535,
|
| 2200 |
+
"grad_norm": 3.779723644256592,
|
| 2201 |
+
"learning_rate": 4.1834517933907467e-07,
|
| 2202 |
+
"loss": 1.0873,
|
| 2203 |
+
"step": 310
|
| 2204 |
+
},
|
| 2205 |
+
{
|
| 2206 |
+
"epoch": 0.9772191673212883,
|
| 2207 |
+
"grad_norm": 3.453876256942749,
|
| 2208 |
+
"learning_rate": 3.3059275270396207e-07,
|
| 2209 |
+
"loss": 1.092,
|
| 2210 |
+
"step": 311
|
| 2211 |
+
},
|
| 2212 |
+
{
|
| 2213 |
+
"epoch": 0.9803613511390417,
|
| 2214 |
+
"grad_norm": 3.9864912033081055,
|
| 2215 |
+
"learning_rate": 2.5314278568850935e-07,
|
| 2216 |
+
"loss": 1.1474,
|
| 2217 |
+
"step": 312
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 0.983503534956795,
|
| 2221 |
+
"grad_norm": 3.627704381942749,
|
| 2222 |
+
"learning_rate": 1.8600328401061629e-07,
|
| 2223 |
+
"loss": 0.9844,
|
| 2224 |
+
"step": 313
|
| 2225 |
+
},
|
| 2226 |
+
{
|
| 2227 |
+
"epoch": 0.9866457187745483,
|
| 2228 |
+
"grad_norm": 3.843008041381836,
|
| 2229 |
+
"learning_rate": 1.2918118763335373e-07,
|
| 2230 |
+
"loss": 0.9021,
|
| 2231 |
+
"step": 314
|
| 2232 |
+
},
|
| 2233 |
+
{
|
| 2234 |
+
"epoch": 0.9897879025923016,
|
| 2235 |
+
"grad_norm": 5.430519104003906,
|
| 2236 |
+
"learning_rate": 8.268237004757096e-08,
|
| 2237 |
+
"loss": 1.1688,
|
| 2238 |
+
"step": 315
|
| 2239 |
+
},
|
| 2240 |
+
{
|
| 2241 |
+
"epoch": 0.992930086410055,
|
| 2242 |
+
"grad_norm": 5.529233932495117,
|
| 2243 |
+
"learning_rate": 4.651163766484779e-08,
|
| 2244 |
+
"loss": 1.309,
|
| 2245 |
+
"step": 316
|
| 2246 |
+
},
|
| 2247 |
+
{
|
| 2248 |
+
"epoch": 0.9960722702278083,
|
| 2249 |
+
"grad_norm": 6.51206636428833,
|
| 2250 |
+
"learning_rate": 2.0672729320581065e-08,
|
| 2251 |
+
"loss": 2.0086,
|
| 2252 |
+
"step": 317
|
| 2253 |
+
},
|
| 2254 |
+
{
|
| 2255 |
+
"epoch": 0.9992144540455616,
|
| 2256 |
+
"grad_norm": 5.897537708282471,
|
| 2257 |
+
"learning_rate": 5.1683158875937e-09,
|
| 2258 |
+
"loss": 0.881,
|
| 2259 |
+
"step": 318
|
| 2260 |
+
},
|
| 2261 |
+
{
|
| 2262 |
+
"epoch": 1.002356637863315,
|
| 2263 |
+
"grad_norm": 25.450618743896484,
|
| 2264 |
+
"learning_rate": 0.0,
|
| 2265 |
+
"loss": 3.048,
|
| 2266 |
+
"step": 319
|
| 2267 |
}
|
| 2268 |
],
|
| 2269 |
"logging_steps": 1,
|
|
|
|
| 2278 |
"should_evaluate": false,
|
| 2279 |
"should_log": false,
|
| 2280 |
"should_save": true,
|
| 2281 |
+
"should_training_stop": true
|
| 2282 |
},
|
| 2283 |
"attributes": {}
|
| 2284 |
}
|
| 2285 |
},
|
| 2286 |
+
"total_flos": 2.696474000149709e+16,
|
| 2287 |
"train_batch_size": 2,
|
| 2288 |
"trial_name": null,
|
| 2289 |
"trial_params": null
|