Training in progress, step 200, 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 191968
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8c7f760852553c885da7ac4651c37d05351dad144849dba33d67013f7dd0453e
|
| 3 |
size 191968
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 253144
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:79e6fe293e85b0f229b1a21ac73548863c9dfaf7ad7c010a558e69a5d604454b
|
| 3 |
size 253144
|
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 sha256:7a01750f036d9438514de7d76a47893cf8fe3739c2ce1fd668646188c049c326
|
| 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:d2d754412c61116546142914503e7369d0cc35d3c380a07e5218f595d76b6d96
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 10.
|
| 3 |
-
"best_model_checkpoint": "miner_id_24/checkpoint-
|
| 4 |
-
"epoch": 0.
|
| 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,
|
|
@@ -731,6 +731,722 @@
|
|
| 731 |
"eval_samples_per_second": 451.852,
|
| 732 |
"eval_steps_per_second": 112.963,
|
| 733 |
"step": 100
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 734 |
}
|
| 735 |
],
|
| 736 |
"logging_steps": 1,
|
|
@@ -754,12 +1470,12 @@
|
|
| 754 |
"should_evaluate": false,
|
| 755 |
"should_log": false,
|
| 756 |
"should_save": true,
|
| 757 |
-
"should_training_stop":
|
| 758 |
},
|
| 759 |
"attributes": {}
|
| 760 |
}
|
| 761 |
},
|
| 762 |
-
"total_flos":
|
| 763 |
"train_batch_size": 8,
|
| 764 |
"trial_name": null,
|
| 765 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 10.32220458984375,
|
| 3 |
+
"best_model_checkpoint": "miner_id_24/checkpoint-200",
|
| 4 |
+
"epoch": 0.642570281124498,
|
| 5 |
"eval_steps": 50,
|
| 6 |
+
"global_step": 200,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 731 |
"eval_samples_per_second": 451.852,
|
| 732 |
"eval_steps_per_second": 112.963,
|
| 733 |
"step": 100
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 0.3244979919678715,
|
| 737 |
+
"grad_norm": 0.13746659457683563,
|
| 738 |
+
"learning_rate": 5.330452921628497e-05,
|
| 739 |
+
"loss": 10.3277,
|
| 740 |
+
"step": 101
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 0.327710843373494,
|
| 744 |
+
"grad_norm": 0.14080485701560974,
|
| 745 |
+
"learning_rate": 5.247918773366112e-05,
|
| 746 |
+
"loss": 10.3229,
|
| 747 |
+
"step": 102
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 0.3309236947791165,
|
| 751 |
+
"grad_norm": 0.08412718772888184,
|
| 752 |
+
"learning_rate": 5.165316846586541e-05,
|
| 753 |
+
"loss": 10.3341,
|
| 754 |
+
"step": 103
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 0.334136546184739,
|
| 758 |
+
"grad_norm": 0.09817983210086823,
|
| 759 |
+
"learning_rate": 5.0826697238317935e-05,
|
| 760 |
+
"loss": 10.3305,
|
| 761 |
+
"step": 104
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.3373493975903614,
|
| 765 |
+
"grad_norm": 0.11353299766778946,
|
| 766 |
+
"learning_rate": 5e-05,
|
| 767 |
+
"loss": 10.3283,
|
| 768 |
+
"step": 105
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 0.3405622489959839,
|
| 772 |
+
"grad_norm": 0.1714153289794922,
|
| 773 |
+
"learning_rate": 4.917330276168208e-05,
|
| 774 |
+
"loss": 10.3298,
|
| 775 |
+
"step": 106
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 0.3437751004016064,
|
| 779 |
+
"grad_norm": 0.14416934549808502,
|
| 780 |
+
"learning_rate": 4.834683153413459e-05,
|
| 781 |
+
"loss": 10.3311,
|
| 782 |
+
"step": 107
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.3469879518072289,
|
| 786 |
+
"grad_norm": 0.12153627723455429,
|
| 787 |
+
"learning_rate": 4.7520812266338885e-05,
|
| 788 |
+
"loss": 10.3296,
|
| 789 |
+
"step": 108
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.3502008032128514,
|
| 793 |
+
"grad_norm": 0.09511595964431763,
|
| 794 |
+
"learning_rate": 4.669547078371504e-05,
|
| 795 |
+
"loss": 10.3231,
|
| 796 |
+
"step": 109
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 0.3534136546184739,
|
| 800 |
+
"grad_norm": 0.0920514166355133,
|
| 801 |
+
"learning_rate": 4.5871032726383386e-05,
|
| 802 |
+
"loss": 10.3294,
|
| 803 |
+
"step": 110
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 0.3566265060240964,
|
| 807 |
+
"grad_norm": 0.10794074833393097,
|
| 808 |
+
"learning_rate": 4.504772348747687e-05,
|
| 809 |
+
"loss": 10.3341,
|
| 810 |
+
"step": 111
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 0.3598393574297189,
|
| 814 |
+
"grad_norm": 0.10634323954582214,
|
| 815 |
+
"learning_rate": 4.4225768151520694e-05,
|
| 816 |
+
"loss": 10.3354,
|
| 817 |
+
"step": 112
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 0.36305220883534134,
|
| 821 |
+
"grad_norm": 0.11892643570899963,
|
| 822 |
+
"learning_rate": 4.3405391432896555e-05,
|
| 823 |
+
"loss": 10.3274,
|
| 824 |
+
"step": 113
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.36626506024096384,
|
| 828 |
+
"grad_norm": 0.14139418303966522,
|
| 829 |
+
"learning_rate": 4.2586817614407895e-05,
|
| 830 |
+
"loss": 10.3256,
|
| 831 |
+
"step": 114
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.36947791164658633,
|
| 835 |
+
"grad_norm": 0.13411003351211548,
|
| 836 |
+
"learning_rate": 4.17702704859633e-05,
|
| 837 |
+
"loss": 10.3273,
|
| 838 |
+
"step": 115
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 0.37269076305220883,
|
| 842 |
+
"grad_norm": 0.0930616557598114,
|
| 843 |
+
"learning_rate": 4.095597328339452e-05,
|
| 844 |
+
"loss": 10.3317,
|
| 845 |
+
"step": 116
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 0.3759036144578313,
|
| 849 |
+
"grad_norm": 0.08703553676605225,
|
| 850 |
+
"learning_rate": 4.0144148627425993e-05,
|
| 851 |
+
"loss": 10.3369,
|
| 852 |
+
"step": 117
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 0.3791164658634538,
|
| 856 |
+
"grad_norm": 0.08455926179885864,
|
| 857 |
+
"learning_rate": 3.933501846281267e-05,
|
| 858 |
+
"loss": 10.3318,
|
| 859 |
+
"step": 118
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 0.3823293172690763,
|
| 863 |
+
"grad_norm": 0.1004587933421135,
|
| 864 |
+
"learning_rate": 3.852880399766243e-05,
|
| 865 |
+
"loss": 10.3291,
|
| 866 |
+
"step": 119
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.3855421686746988,
|
| 870 |
+
"grad_norm": 0.09143741428852081,
|
| 871 |
+
"learning_rate": 3.772572564296005e-05,
|
| 872 |
+
"loss": 10.3216,
|
| 873 |
+
"step": 120
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.3887550200803213,
|
| 877 |
+
"grad_norm": 0.12408670037984848,
|
| 878 |
+
"learning_rate": 3.6926002952309016e-05,
|
| 879 |
+
"loss": 10.3266,
|
| 880 |
+
"step": 121
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 0.39196787148594375,
|
| 884 |
+
"grad_norm": 0.11100861430168152,
|
| 885 |
+
"learning_rate": 3.612985456190778e-05,
|
| 886 |
+
"loss": 10.3382,
|
| 887 |
+
"step": 122
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 0.39518072289156625,
|
| 891 |
+
"grad_norm": 0.09714003652334213,
|
| 892 |
+
"learning_rate": 3.533749813077677e-05,
|
| 893 |
+
"loss": 10.3232,
|
| 894 |
+
"step": 123
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 0.39839357429718875,
|
| 898 |
+
"grad_norm": 0.09558629989624023,
|
| 899 |
+
"learning_rate": 3.4549150281252636e-05,
|
| 900 |
+
"loss": 10.3278,
|
| 901 |
+
"step": 124
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 0.40160642570281124,
|
| 905 |
+
"grad_norm": 0.11661966890096664,
|
| 906 |
+
"learning_rate": 3.3765026539765834e-05,
|
| 907 |
+
"loss": 10.3297,
|
| 908 |
+
"step": 125
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 0.40481927710843374,
|
| 912 |
+
"grad_norm": 0.1436244696378708,
|
| 913 |
+
"learning_rate": 3.298534127791785e-05,
|
| 914 |
+
"loss": 10.3295,
|
| 915 |
+
"step": 126
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 0.40803212851405624,
|
| 919 |
+
"grad_norm": 0.13147498667240143,
|
| 920 |
+
"learning_rate": 3.221030765387417e-05,
|
| 921 |
+
"loss": 10.3174,
|
| 922 |
+
"step": 127
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 0.41124497991967873,
|
| 926 |
+
"grad_norm": 0.1136285737156868,
|
| 927 |
+
"learning_rate": 3.144013755408895e-05,
|
| 928 |
+
"loss": 10.3281,
|
| 929 |
+
"step": 128
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 0.41445783132530123,
|
| 933 |
+
"grad_norm": 0.1124458834528923,
|
| 934 |
+
"learning_rate": 3.0675041535377405e-05,
|
| 935 |
+
"loss": 10.3303,
|
| 936 |
+
"step": 129
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 0.41767068273092367,
|
| 940 |
+
"grad_norm": 0.16464786231517792,
|
| 941 |
+
"learning_rate": 2.991522876735154e-05,
|
| 942 |
+
"loss": 10.3337,
|
| 943 |
+
"step": 130
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 0.42088353413654617,
|
| 947 |
+
"grad_norm": 0.16290715336799622,
|
| 948 |
+
"learning_rate": 2.916090697523549e-05,
|
| 949 |
+
"loss": 10.3255,
|
| 950 |
+
"step": 131
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.42409638554216866,
|
| 954 |
+
"grad_norm": 0.09206510335206985,
|
| 955 |
+
"learning_rate": 2.8412282383075363e-05,
|
| 956 |
+
"loss": 10.3199,
|
| 957 |
+
"step": 132
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 0.42730923694779116,
|
| 961 |
+
"grad_norm": 0.11564340442419052,
|
| 962 |
+
"learning_rate": 2.766955965735968e-05,
|
| 963 |
+
"loss": 10.326,
|
| 964 |
+
"step": 133
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 0.43052208835341366,
|
| 968 |
+
"grad_norm": 0.10356107354164124,
|
| 969 |
+
"learning_rate": 2.693294185106562e-05,
|
| 970 |
+
"loss": 10.3294,
|
| 971 |
+
"step": 134
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 0.43373493975903615,
|
| 975 |
+
"grad_norm": 0.12882065773010254,
|
| 976 |
+
"learning_rate": 2.6202630348146324e-05,
|
| 977 |
+
"loss": 10.3299,
|
| 978 |
+
"step": 135
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 0.43694779116465865,
|
| 982 |
+
"grad_norm": 0.12127930670976639,
|
| 983 |
+
"learning_rate": 2.547882480847461e-05,
|
| 984 |
+
"loss": 10.3231,
|
| 985 |
+
"step": 136
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 0.44016064257028115,
|
| 989 |
+
"grad_norm": 0.10148213058710098,
|
| 990 |
+
"learning_rate": 2.476172311325783e-05,
|
| 991 |
+
"loss": 10.3299,
|
| 992 |
+
"step": 137
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.4433734939759036,
|
| 996 |
+
"grad_norm": 0.11116017401218414,
|
| 997 |
+
"learning_rate": 2.405152131093926e-05,
|
| 998 |
+
"loss": 10.3209,
|
| 999 |
+
"step": 138
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.4465863453815261,
|
| 1003 |
+
"grad_norm": 0.149327352643013,
|
| 1004 |
+
"learning_rate": 2.3348413563600325e-05,
|
| 1005 |
+
"loss": 10.3225,
|
| 1006 |
+
"step": 139
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 0.4497991967871486,
|
| 1010 |
+
"grad_norm": 0.13048338890075684,
|
| 1011 |
+
"learning_rate": 2.2652592093878666e-05,
|
| 1012 |
+
"loss": 10.3307,
|
| 1013 |
+
"step": 140
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 0.4530120481927711,
|
| 1017 |
+
"grad_norm": 0.09309875965118408,
|
| 1018 |
+
"learning_rate": 2.196424713241637e-05,
|
| 1019 |
+
"loss": 10.3179,
|
| 1020 |
+
"step": 141
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 0.4562248995983936,
|
| 1024 |
+
"grad_norm": 0.11135369539260864,
|
| 1025 |
+
"learning_rate": 2.128356686585282e-05,
|
| 1026 |
+
"loss": 10.318,
|
| 1027 |
+
"step": 142
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 0.45943775100401607,
|
| 1031 |
+
"grad_norm": 0.14271609485149384,
|
| 1032 |
+
"learning_rate": 2.061073738537635e-05,
|
| 1033 |
+
"loss": 10.3149,
|
| 1034 |
+
"step": 143
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 0.46265060240963857,
|
| 1038 |
+
"grad_norm": 0.12098371237516403,
|
| 1039 |
+
"learning_rate": 1.9945942635848748e-05,
|
| 1040 |
+
"loss": 10.3345,
|
| 1041 |
+
"step": 144
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 0.46586345381526106,
|
| 1045 |
+
"grad_norm": 0.09931446611881256,
|
| 1046 |
+
"learning_rate": 1.928936436551661e-05,
|
| 1047 |
+
"loss": 10.3217,
|
| 1048 |
+
"step": 145
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 0.46907630522088356,
|
| 1052 |
+
"grad_norm": 0.10187914222478867,
|
| 1053 |
+
"learning_rate": 1.8641182076323148e-05,
|
| 1054 |
+
"loss": 10.317,
|
| 1055 |
+
"step": 146
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 0.472289156626506,
|
| 1059 |
+
"grad_norm": 0.1341490000486374,
|
| 1060 |
+
"learning_rate": 1.800157297483417e-05,
|
| 1061 |
+
"loss": 10.3159,
|
| 1062 |
+
"step": 147
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 0.4755020080321285,
|
| 1066 |
+
"grad_norm": 0.13303139805793762,
|
| 1067 |
+
"learning_rate": 1.7370711923791567e-05,
|
| 1068 |
+
"loss": 10.3084,
|
| 1069 |
+
"step": 148
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 0.478714859437751,
|
| 1073 |
+
"grad_norm": 0.10958024859428406,
|
| 1074 |
+
"learning_rate": 1.6748771394307585e-05,
|
| 1075 |
+
"loss": 10.3196,
|
| 1076 |
+
"step": 149
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 0.4819277108433735,
|
| 1080 |
+
"grad_norm": 0.15488006174564362,
|
| 1081 |
+
"learning_rate": 1.6135921418712956e-05,
|
| 1082 |
+
"loss": 10.3069,
|
| 1083 |
+
"step": 150
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 0.4819277108433735,
|
| 1087 |
+
"eval_loss": 10.323417663574219,
|
| 1088 |
+
"eval_runtime": 1.1572,
|
| 1089 |
+
"eval_samples_per_second": 452.817,
|
| 1090 |
+
"eval_steps_per_second": 113.204,
|
| 1091 |
+
"step": 150
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 0.485140562248996,
|
| 1095 |
+
"grad_norm": 0.1342858374118805,
|
| 1096 |
+
"learning_rate": 1.553232954407171e-05,
|
| 1097 |
+
"loss": 10.3197,
|
| 1098 |
+
"step": 151
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 0.4883534136546185,
|
| 1102 |
+
"grad_norm": 0.08063612133264542,
|
| 1103 |
+
"learning_rate": 1.4938160786375572e-05,
|
| 1104 |
+
"loss": 10.3286,
|
| 1105 |
+
"step": 152
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 0.491566265060241,
|
| 1109 |
+
"grad_norm": 0.12848462164402008,
|
| 1110 |
+
"learning_rate": 1.435357758543015e-05,
|
| 1111 |
+
"loss": 10.3402,
|
| 1112 |
+
"step": 153
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 0.4947791164658635,
|
| 1116 |
+
"grad_norm": 0.10360779613256454,
|
| 1117 |
+
"learning_rate": 1.3778739760445552e-05,
|
| 1118 |
+
"loss": 10.327,
|
| 1119 |
+
"step": 154
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 0.4979919678714859,
|
| 1123 |
+
"grad_norm": 0.09004613012075424,
|
| 1124 |
+
"learning_rate": 1.3213804466343421e-05,
|
| 1125 |
+
"loss": 10.3244,
|
| 1126 |
+
"step": 155
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 0.5012048192771085,
|
| 1130 |
+
"grad_norm": 0.10969901084899902,
|
| 1131 |
+
"learning_rate": 1.2658926150792322e-05,
|
| 1132 |
+
"loss": 10.3339,
|
| 1133 |
+
"step": 156
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 0.5044176706827309,
|
| 1137 |
+
"grad_norm": 0.1162419244647026,
|
| 1138 |
+
"learning_rate": 1.2114256511983274e-05,
|
| 1139 |
+
"loss": 10.3268,
|
| 1140 |
+
"step": 157
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 0.5076305220883535,
|
| 1144 |
+
"grad_norm": 0.09138023853302002,
|
| 1145 |
+
"learning_rate": 1.157994445715706e-05,
|
| 1146 |
+
"loss": 10.3338,
|
| 1147 |
+
"step": 158
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 0.5108433734939759,
|
| 1151 |
+
"grad_norm": 0.09648050367832184,
|
| 1152 |
+
"learning_rate": 1.1056136061894384e-05,
|
| 1153 |
+
"loss": 10.3271,
|
| 1154 |
+
"step": 159
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.5140562248995983,
|
| 1158 |
+
"grad_norm": 0.08033082634210587,
|
| 1159 |
+
"learning_rate": 1.0542974530180327e-05,
|
| 1160 |
+
"loss": 10.328,
|
| 1161 |
+
"step": 160
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.5172690763052209,
|
| 1165 |
+
"grad_norm": 0.10192538797855377,
|
| 1166 |
+
"learning_rate": 1.0040600155253765e-05,
|
| 1167 |
+
"loss": 10.3286,
|
| 1168 |
+
"step": 161
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 0.5204819277108433,
|
| 1172 |
+
"grad_norm": 0.1149694174528122,
|
| 1173 |
+
"learning_rate": 9.549150281252633e-06,
|
| 1174 |
+
"loss": 10.3251,
|
| 1175 |
+
"step": 162
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 0.5236947791164659,
|
| 1179 |
+
"grad_norm": 0.11814501136541367,
|
| 1180 |
+
"learning_rate": 9.068759265665384e-06,
|
| 1181 |
+
"loss": 10.32,
|
| 1182 |
+
"step": 163
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 0.5269076305220883,
|
| 1186 |
+
"grad_norm": 0.09197898954153061,
|
| 1187 |
+
"learning_rate": 8.599558442598998e-06,
|
| 1188 |
+
"loss": 10.3235,
|
| 1189 |
+
"step": 164
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 0.5301204819277109,
|
| 1193 |
+
"grad_norm": 0.07296612858772278,
|
| 1194 |
+
"learning_rate": 8.141676086873572e-06,
|
| 1195 |
+
"loss": 10.3401,
|
| 1196 |
+
"step": 165
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.5333333333333333,
|
| 1200 |
+
"grad_norm": 0.09049750119447708,
|
| 1201 |
+
"learning_rate": 7.695237378953223e-06,
|
| 1202 |
+
"loss": 10.3267,
|
| 1203 |
+
"step": 166
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 0.5365461847389559,
|
| 1207 |
+
"grad_norm": 0.11494322121143341,
|
| 1208 |
+
"learning_rate": 7.260364370723044e-06,
|
| 1209 |
+
"loss": 10.3268,
|
| 1210 |
+
"step": 167
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 0.5397590361445783,
|
| 1214 |
+
"grad_norm": 0.10997875779867172,
|
| 1215 |
+
"learning_rate": 6.837175952121306e-06,
|
| 1216 |
+
"loss": 10.3252,
|
| 1217 |
+
"step": 168
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 0.5429718875502008,
|
| 1221 |
+
"grad_norm": 0.10923228412866592,
|
| 1222 |
+
"learning_rate": 6.425787818636131e-06,
|
| 1223 |
+
"loss": 10.3276,
|
| 1224 |
+
"step": 169
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 0.5461847389558233,
|
| 1228 |
+
"grad_norm": 0.09875035285949707,
|
| 1229 |
+
"learning_rate": 6.026312439675552e-06,
|
| 1230 |
+
"loss": 10.3228,
|
| 1231 |
+
"step": 170
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 0.5493975903614458,
|
| 1235 |
+
"grad_norm": 0.12667155265808105,
|
| 1236 |
+
"learning_rate": 5.6388590278194096e-06,
|
| 1237 |
+
"loss": 10.3328,
|
| 1238 |
+
"step": 171
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 0.5526104417670683,
|
| 1242 |
+
"grad_norm": 0.10388634353876114,
|
| 1243 |
+
"learning_rate": 5.263533508961827e-06,
|
| 1244 |
+
"loss": 10.3298,
|
| 1245 |
+
"step": 172
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 0.5558232931726907,
|
| 1249 |
+
"grad_norm": 0.12183015048503876,
|
| 1250 |
+
"learning_rate": 4.900438493352055e-06,
|
| 1251 |
+
"loss": 10.331,
|
| 1252 |
+
"step": 173
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 0.5590361445783133,
|
| 1256 |
+
"grad_norm": 0.12482339143753052,
|
| 1257 |
+
"learning_rate": 4.549673247541875e-06,
|
| 1258 |
+
"loss": 10.3244,
|
| 1259 |
+
"step": 174
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 0.5622489959839357,
|
| 1263 |
+
"grad_norm": 0.08300051093101501,
|
| 1264 |
+
"learning_rate": 4.2113336672471245e-06,
|
| 1265 |
+
"loss": 10.3323,
|
| 1266 |
+
"step": 175
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 0.5654618473895582,
|
| 1270 |
+
"grad_norm": 0.11860600858926773,
|
| 1271 |
+
"learning_rate": 3.885512251130763e-06,
|
| 1272 |
+
"loss": 10.3321,
|
| 1273 |
+
"step": 176
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 0.5686746987951807,
|
| 1277 |
+
"grad_norm": 0.11088336259126663,
|
| 1278 |
+
"learning_rate": 3.5722980755146517e-06,
|
| 1279 |
+
"loss": 10.3327,
|
| 1280 |
+
"step": 177
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 0.5718875502008032,
|
| 1284 |
+
"grad_norm": 0.08796709775924683,
|
| 1285 |
+
"learning_rate": 3.271776770026963e-06,
|
| 1286 |
+
"loss": 10.3308,
|
| 1287 |
+
"step": 178
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 0.5751004016064257,
|
| 1291 |
+
"grad_norm": 0.10811839252710342,
|
| 1292 |
+
"learning_rate": 2.9840304941919415e-06,
|
| 1293 |
+
"loss": 10.3314,
|
| 1294 |
+
"step": 179
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 0.5783132530120482,
|
| 1298 |
+
"grad_norm": 0.11419740319252014,
|
| 1299 |
+
"learning_rate": 2.7091379149682685e-06,
|
| 1300 |
+
"loss": 10.334,
|
| 1301 |
+
"step": 180
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 0.5815261044176707,
|
| 1305 |
+
"grad_norm": 0.09244905412197113,
|
| 1306 |
+
"learning_rate": 2.4471741852423237e-06,
|
| 1307 |
+
"loss": 10.325,
|
| 1308 |
+
"step": 181
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 0.5847389558232932,
|
| 1312 |
+
"grad_norm": 0.11072293668985367,
|
| 1313 |
+
"learning_rate": 2.1982109232821178e-06,
|
| 1314 |
+
"loss": 10.3272,
|
| 1315 |
+
"step": 182
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 0.5879518072289157,
|
| 1319 |
+
"grad_norm": 0.10999740660190582,
|
| 1320 |
+
"learning_rate": 1.962316193157593e-06,
|
| 1321 |
+
"loss": 10.3254,
|
| 1322 |
+
"step": 183
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 0.5911646586345382,
|
| 1326 |
+
"grad_norm": 0.09968981891870499,
|
| 1327 |
+
"learning_rate": 1.7395544861325718e-06,
|
| 1328 |
+
"loss": 10.3236,
|
| 1329 |
+
"step": 184
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 0.5943775100401606,
|
| 1333 |
+
"grad_norm": 0.12061024457216263,
|
| 1334 |
+
"learning_rate": 1.5299867030334814e-06,
|
| 1335 |
+
"loss": 10.3159,
|
| 1336 |
+
"step": 185
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 0.5975903614457831,
|
| 1340 |
+
"grad_norm": 0.07807968556880951,
|
| 1341 |
+
"learning_rate": 1.333670137599713e-06,
|
| 1342 |
+
"loss": 10.3207,
|
| 1343 |
+
"step": 186
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 0.6008032128514056,
|
| 1347 |
+
"grad_norm": 0.12807731330394745,
|
| 1348 |
+
"learning_rate": 1.1506584608200367e-06,
|
| 1349 |
+
"loss": 10.3216,
|
| 1350 |
+
"step": 187
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 0.6040160642570281,
|
| 1354 |
+
"grad_norm": 0.12024622410535812,
|
| 1355 |
+
"learning_rate": 9.810017062595322e-07,
|
| 1356 |
+
"loss": 10.3303,
|
| 1357 |
+
"step": 188
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 0.6072289156626506,
|
| 1361 |
+
"grad_norm": 0.12788936495780945,
|
| 1362 |
+
"learning_rate": 8.247462563808817e-07,
|
| 1363 |
+
"loss": 10.3188,
|
| 1364 |
+
"step": 189
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 0.6104417670682731,
|
| 1368 |
+
"grad_norm": 0.19885775446891785,
|
| 1369 |
+
"learning_rate": 6.819348298638839e-07,
|
| 1370 |
+
"loss": 10.3076,
|
| 1371 |
+
"step": 190
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 0.6136546184738956,
|
| 1375 |
+
"grad_norm": 0.11481358110904694,
|
| 1376 |
+
"learning_rate": 5.526064699265753e-07,
|
| 1377 |
+
"loss": 10.3277,
|
| 1378 |
+
"step": 191
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 0.6168674698795181,
|
| 1382 |
+
"grad_norm": 0.12053002417087555,
|
| 1383 |
+
"learning_rate": 4.367965336512403e-07,
|
| 1384 |
+
"loss": 10.3274,
|
| 1385 |
+
"step": 192
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 0.6200803212851406,
|
| 1389 |
+
"grad_norm": 0.10069674998521805,
|
| 1390 |
+
"learning_rate": 3.3453668231809286e-07,
|
| 1391 |
+
"loss": 10.3178,
|
| 1392 |
+
"step": 193
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 0.623293172690763,
|
| 1396 |
+
"grad_norm": 0.08938242495059967,
|
| 1397 |
+
"learning_rate": 2.458548727494292e-07,
|
| 1398 |
+
"loss": 10.3322,
|
| 1399 |
+
"step": 194
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 0.6265060240963856,
|
| 1403 |
+
"grad_norm": 0.13311991095542908,
|
| 1404 |
+
"learning_rate": 1.7077534966650766e-07,
|
| 1405 |
+
"loss": 10.3274,
|
| 1406 |
+
"step": 195
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 0.629718875502008,
|
| 1410 |
+
"grad_norm": 0.08758800476789474,
|
| 1411 |
+
"learning_rate": 1.0931863906127327e-07,
|
| 1412 |
+
"loss": 10.3239,
|
| 1413 |
+
"step": 196
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.6329317269076306,
|
| 1417 |
+
"grad_norm": 0.09972859919071198,
|
| 1418 |
+
"learning_rate": 6.150154258476315e-08,
|
| 1419 |
+
"loss": 10.3199,
|
| 1420 |
+
"step": 197
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 0.636144578313253,
|
| 1424 |
+
"grad_norm": 0.14015185832977295,
|
| 1425 |
+
"learning_rate": 2.7337132953697554e-08,
|
| 1426 |
+
"loss": 10.3094,
|
| 1427 |
+
"step": 198
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 0.6393574297188755,
|
| 1431 |
+
"grad_norm": 0.10132836550474167,
|
| 1432 |
+
"learning_rate": 6.834750376549792e-09,
|
| 1433 |
+
"loss": 10.3234,
|
| 1434 |
+
"step": 199
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.642570281124498,
|
| 1438 |
+
"grad_norm": 0.19862784445285797,
|
| 1439 |
+
"learning_rate": 0.0,
|
| 1440 |
+
"loss": 10.3101,
|
| 1441 |
+
"step": 200
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.642570281124498,
|
| 1445 |
+
"eval_loss": 10.32220458984375,
|
| 1446 |
+
"eval_runtime": 1.154,
|
| 1447 |
+
"eval_samples_per_second": 454.088,
|
| 1448 |
+
"eval_steps_per_second": 113.522,
|
| 1449 |
+
"step": 200
|
| 1450 |
}
|
| 1451 |
],
|
| 1452 |
"logging_steps": 1,
|
|
|
|
| 1470 |
"should_evaluate": false,
|
| 1471 |
"should_log": false,
|
| 1472 |
"should_save": true,
|
| 1473 |
+
"should_training_stop": true
|
| 1474 |
},
|
| 1475 |
"attributes": {}
|
| 1476 |
}
|
| 1477 |
},
|
| 1478 |
+
"total_flos": 22477458112512.0,
|
| 1479 |
"train_batch_size": 8,
|
| 1480 |
"trial_name": null,
|
| 1481 |
"trial_params": null
|