Training in progress, step 1500, 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 527048968
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:be8a45329c3e8e3a2f45c77d7ac9080fd0aad7fa0e5966556ccab7ad8fa2f098
|
| 3 |
size 527048968
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1054135994
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1178946cf0771c850e2a43f81c79255d308890814174c3df8070c8810c0eba4c
|
| 3 |
size 1054135994
|
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 sha256:86a4d2549b1634b5162bba6559970b4387f95d4c5153179e15ae2066cc09b884
|
| 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:d00fd7759971df004f86f26240bbcec291b25e581e60ea022fb56951da465e4f
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": 0.7166430950164795,
|
| 3 |
"best_model_checkpoint": "./output/checkpoint-450",
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 150,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -732,6 +732,731 @@
|
|
| 732 |
"EMA_steps_per_second": 24.839,
|
| 733 |
"epoch": 32.608695652173914,
|
| 734 |
"step": 750
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 735 |
}
|
| 736 |
],
|
| 737 |
"logging_steps": 10,
|
|
@@ -751,7 +1476,7 @@
|
|
| 751 |
"attributes": {}
|
| 752 |
}
|
| 753 |
},
|
| 754 |
-
"total_flos":
|
| 755 |
"train_batch_size": 4,
|
| 756 |
"trial_name": null,
|
| 757 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": 0.7166430950164795,
|
| 3 |
"best_model_checkpoint": "./output/checkpoint-450",
|
| 4 |
+
"epoch": 65.21739130434783,
|
| 5 |
"eval_steps": 150,
|
| 6 |
+
"global_step": 1500,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 732 |
"EMA_steps_per_second": 24.839,
|
| 733 |
"epoch": 32.608695652173914,
|
| 734 |
"step": 750
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 33.04347826086956,
|
| 738 |
+
"grad_norm": 1.4949195384979248,
|
| 739 |
+
"learning_rate": 2.998705251845287e-05,
|
| 740 |
+
"loss": 0.43,
|
| 741 |
+
"step": 760
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 33.47826086956522,
|
| 745 |
+
"grad_norm": 1.6518038511276245,
|
| 746 |
+
"learning_rate": 2.9986657256812e-05,
|
| 747 |
+
"loss": 0.4301,
|
| 748 |
+
"step": 770
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 33.91304347826087,
|
| 752 |
+
"grad_norm": 1.2894669771194458,
|
| 753 |
+
"learning_rate": 2.9986256054939022e-05,
|
| 754 |
+
"loss": 0.408,
|
| 755 |
+
"step": 780
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 34.34782608695652,
|
| 759 |
+
"grad_norm": 1.4762516021728516,
|
| 760 |
+
"learning_rate": 2.9985848912992956e-05,
|
| 761 |
+
"loss": 0.4029,
|
| 762 |
+
"step": 790
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 34.78260869565217,
|
| 766 |
+
"grad_norm": 1.5660409927368164,
|
| 767 |
+
"learning_rate": 2.9985435831135184e-05,
|
| 768 |
+
"loss": 0.3832,
|
| 769 |
+
"step": 800
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 35.21739130434783,
|
| 773 |
+
"grad_norm": 1.3075863122940063,
|
| 774 |
+
"learning_rate": 2.9985016809529437e-05,
|
| 775 |
+
"loss": 0.4744,
|
| 776 |
+
"step": 810
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 35.65217391304348,
|
| 780 |
+
"grad_norm": 1.3411126136779785,
|
| 781 |
+
"learning_rate": 2.9984591848341806e-05,
|
| 782 |
+
"loss": 0.403,
|
| 783 |
+
"step": 820
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 36.08695652173913,
|
| 787 |
+
"grad_norm": 1.0828583240509033,
|
| 788 |
+
"learning_rate": 2.9984160947740723e-05,
|
| 789 |
+
"loss": 0.4181,
|
| 790 |
+
"step": 830
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 36.52173913043478,
|
| 794 |
+
"grad_norm": 1.1622037887573242,
|
| 795 |
+
"learning_rate": 2.9983724107896993e-05,
|
| 796 |
+
"loss": 0.3806,
|
| 797 |
+
"step": 840
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 36.95652173913044,
|
| 801 |
+
"grad_norm": 1.4791110754013062,
|
| 802 |
+
"learning_rate": 2.9983281328983757e-05,
|
| 803 |
+
"loss": 0.4499,
|
| 804 |
+
"step": 850
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 37.391304347826086,
|
| 808 |
+
"grad_norm": 1.8963046073913574,
|
| 809 |
+
"learning_rate": 2.9982832611176523e-05,
|
| 810 |
+
"loss": 0.4181,
|
| 811 |
+
"step": 860
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 37.82608695652174,
|
| 815 |
+
"grad_norm": 1.270815372467041,
|
| 816 |
+
"learning_rate": 2.998237795465315e-05,
|
| 817 |
+
"loss": 0.3714,
|
| 818 |
+
"step": 870
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 38.26086956521739,
|
| 822 |
+
"grad_norm": 1.264829397201538,
|
| 823 |
+
"learning_rate": 2.9981917359593843e-05,
|
| 824 |
+
"loss": 0.4013,
|
| 825 |
+
"step": 880
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 38.69565217391305,
|
| 829 |
+
"grad_norm": 1.4431074857711792,
|
| 830 |
+
"learning_rate": 2.9981450826181172e-05,
|
| 831 |
+
"loss": 0.3552,
|
| 832 |
+
"step": 890
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 39.130434782608695,
|
| 836 |
+
"grad_norm": 1.9556941986083984,
|
| 837 |
+
"learning_rate": 2.9980978354600057e-05,
|
| 838 |
+
"loss": 0.463,
|
| 839 |
+
"step": 900
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 39.130434782608695,
|
| 843 |
+
"eval_loss": 0.7511647343635559,
|
| 844 |
+
"eval_runtime": 0.4145,
|
| 845 |
+
"eval_samples_per_second": 24.126,
|
| 846 |
+
"eval_steps_per_second": 24.126,
|
| 847 |
+
"step": 900
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"Start_State_loss": 0.8601926565170288,
|
| 851 |
+
"Start_State_runtime": 0.4115,
|
| 852 |
+
"Start_State_samples_per_second": 24.303,
|
| 853 |
+
"Start_State_steps_per_second": 24.303,
|
| 854 |
+
"epoch": 39.130434782608695,
|
| 855 |
+
"step": 900
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"Raw_Model_loss": 0.7511647343635559,
|
| 859 |
+
"Raw_Model_runtime": 0.399,
|
| 860 |
+
"Raw_Model_samples_per_second": 25.063,
|
| 861 |
+
"Raw_Model_steps_per_second": 25.063,
|
| 862 |
+
"epoch": 39.130434782608695,
|
| 863 |
+
"step": 900
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"SWA_loss": 0.7235903739929199,
|
| 867 |
+
"SWA_runtime": 0.3941,
|
| 868 |
+
"SWA_samples_per_second": 25.377,
|
| 869 |
+
"SWA_steps_per_second": 25.377,
|
| 870 |
+
"epoch": 39.130434782608695,
|
| 871 |
+
"step": 900
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"EMA_loss": 0.8609917759895325,
|
| 875 |
+
"EMA_runtime": 0.3995,
|
| 876 |
+
"EMA_samples_per_second": 25.033,
|
| 877 |
+
"EMA_steps_per_second": 25.033,
|
| 878 |
+
"epoch": 39.130434782608695,
|
| 879 |
+
"step": 900
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 39.56521739130435,
|
| 883 |
+
"grad_norm": 1.542538046836853,
|
| 884 |
+
"learning_rate": 2.9980499945037765e-05,
|
| 885 |
+
"loss": 0.3835,
|
| 886 |
+
"step": 910
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 40.0,
|
| 890 |
+
"grad_norm": 3.0124218463897705,
|
| 891 |
+
"learning_rate": 2.998001559768393e-05,
|
| 892 |
+
"loss": 0.3867,
|
| 893 |
+
"step": 920
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 40.43478260869565,
|
| 897 |
+
"grad_norm": 1.5339196920394897,
|
| 898 |
+
"learning_rate": 2.9979525312730525e-05,
|
| 899 |
+
"loss": 0.4492,
|
| 900 |
+
"step": 930
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 40.869565217391305,
|
| 904 |
+
"grad_norm": 1.6727086305618286,
|
| 905 |
+
"learning_rate": 2.9979029090371885e-05,
|
| 906 |
+
"loss": 0.3412,
|
| 907 |
+
"step": 940
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 41.30434782608695,
|
| 911 |
+
"grad_norm": 2.2182319164276123,
|
| 912 |
+
"learning_rate": 2.99785269308047e-05,
|
| 913 |
+
"loss": 0.3413,
|
| 914 |
+
"step": 950
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 41.73913043478261,
|
| 918 |
+
"grad_norm": 1.5122953653335571,
|
| 919 |
+
"learning_rate": 2.9978018834228007e-05,
|
| 920 |
+
"loss": 0.365,
|
| 921 |
+
"step": 960
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 42.17391304347826,
|
| 925 |
+
"grad_norm": 1.5070980787277222,
|
| 926 |
+
"learning_rate": 2.9977504800843197e-05,
|
| 927 |
+
"loss": 0.4346,
|
| 928 |
+
"step": 970
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 42.608695652173914,
|
| 932 |
+
"grad_norm": 1.5313963890075684,
|
| 933 |
+
"learning_rate": 2.9976984830854022e-05,
|
| 934 |
+
"loss": 0.3752,
|
| 935 |
+
"step": 980
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 43.04347826086956,
|
| 939 |
+
"grad_norm": 1.653640866279602,
|
| 940 |
+
"learning_rate": 2.997645892446658e-05,
|
| 941 |
+
"loss": 0.367,
|
| 942 |
+
"step": 990
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 43.47826086956522,
|
| 946 |
+
"grad_norm": 1.4292306900024414,
|
| 947 |
+
"learning_rate": 2.9975927081889322e-05,
|
| 948 |
+
"loss": 0.391,
|
| 949 |
+
"step": 1000
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 43.91304347826087,
|
| 953 |
+
"grad_norm": 1.1838629245758057,
|
| 954 |
+
"learning_rate": 2.9975389303333047e-05,
|
| 955 |
+
"loss": 0.3456,
|
| 956 |
+
"step": 1010
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 44.34782608695652,
|
| 960 |
+
"grad_norm": 2.111812114715576,
|
| 961 |
+
"learning_rate": 2.997484558901093e-05,
|
| 962 |
+
"loss": 0.3922,
|
| 963 |
+
"step": 1020
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 44.78260869565217,
|
| 967 |
+
"grad_norm": 1.6915301084518433,
|
| 968 |
+
"learning_rate": 2.9974295939138465e-05,
|
| 969 |
+
"loss": 0.3804,
|
| 970 |
+
"step": 1030
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 45.21739130434783,
|
| 974 |
+
"grad_norm": 1.2465533018112183,
|
| 975 |
+
"learning_rate": 2.9973740353933523e-05,
|
| 976 |
+
"loss": 0.2648,
|
| 977 |
+
"step": 1040
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 45.65217391304348,
|
| 981 |
+
"grad_norm": 1.68025541305542,
|
| 982 |
+
"learning_rate": 2.997317883361632e-05,
|
| 983 |
+
"loss": 0.3611,
|
| 984 |
+
"step": 1050
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 45.65217391304348,
|
| 988 |
+
"eval_loss": 0.7759392261505127,
|
| 989 |
+
"eval_runtime": 0.5365,
|
| 990 |
+
"eval_samples_per_second": 18.64,
|
| 991 |
+
"eval_steps_per_second": 18.64,
|
| 992 |
+
"step": 1050
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"Start_State_loss": 0.8601926565170288,
|
| 996 |
+
"Start_State_runtime": 0.5037,
|
| 997 |
+
"Start_State_samples_per_second": 19.854,
|
| 998 |
+
"Start_State_steps_per_second": 19.854,
|
| 999 |
+
"epoch": 45.65217391304348,
|
| 1000 |
+
"step": 1050
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"Raw_Model_loss": 0.7759392261505127,
|
| 1004 |
+
"Raw_Model_runtime": 0.4675,
|
| 1005 |
+
"Raw_Model_samples_per_second": 21.389,
|
| 1006 |
+
"Raw_Model_steps_per_second": 21.389,
|
| 1007 |
+
"epoch": 45.65217391304348,
|
| 1008 |
+
"step": 1050
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"SWA_loss": 0.7227687835693359,
|
| 1012 |
+
"SWA_runtime": 0.4756,
|
| 1013 |
+
"SWA_samples_per_second": 21.025,
|
| 1014 |
+
"SWA_steps_per_second": 21.025,
|
| 1015 |
+
"epoch": 45.65217391304348,
|
| 1016 |
+
"step": 1050
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"EMA_loss": 0.8605559468269348,
|
| 1020 |
+
"EMA_runtime": 0.4881,
|
| 1021 |
+
"EMA_samples_per_second": 20.489,
|
| 1022 |
+
"EMA_steps_per_second": 20.489,
|
| 1023 |
+
"epoch": 45.65217391304348,
|
| 1024 |
+
"step": 1050
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 46.08695652173913,
|
| 1028 |
+
"grad_norm": 1.7922283411026,
|
| 1029 |
+
"learning_rate": 2.997261137840943e-05,
|
| 1030 |
+
"loss": 0.4104,
|
| 1031 |
+
"step": 1060
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 46.52173913043478,
|
| 1035 |
+
"grad_norm": 2.145780324935913,
|
| 1036 |
+
"learning_rate": 2.9972037988537758e-05,
|
| 1037 |
+
"loss": 0.3784,
|
| 1038 |
+
"step": 1070
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 46.95652173913044,
|
| 1042 |
+
"grad_norm": 1.9540642499923706,
|
| 1043 |
+
"learning_rate": 2.9971458664228595e-05,
|
| 1044 |
+
"loss": 0.3325,
|
| 1045 |
+
"step": 1080
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 47.391304347826086,
|
| 1049 |
+
"grad_norm": 2.150164842605591,
|
| 1050 |
+
"learning_rate": 2.997087340571156e-05,
|
| 1051 |
+
"loss": 0.3369,
|
| 1052 |
+
"step": 1090
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 47.82608695652174,
|
| 1056 |
+
"grad_norm": 1.539474606513977,
|
| 1057 |
+
"learning_rate": 2.997028221321863e-05,
|
| 1058 |
+
"loss": 0.3564,
|
| 1059 |
+
"step": 1100
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 48.26086956521739,
|
| 1063 |
+
"grad_norm": 2.3236191272735596,
|
| 1064 |
+
"learning_rate": 2.9969685086984132e-05,
|
| 1065 |
+
"loss": 0.3736,
|
| 1066 |
+
"step": 1110
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 48.69565217391305,
|
| 1070 |
+
"grad_norm": 1.6481757164001465,
|
| 1071 |
+
"learning_rate": 2.9969082027244755e-05,
|
| 1072 |
+
"loss": 0.2999,
|
| 1073 |
+
"step": 1120
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 49.130434782608695,
|
| 1077 |
+
"grad_norm": 1.8113096952438354,
|
| 1078 |
+
"learning_rate": 2.996847303423953e-05,
|
| 1079 |
+
"loss": 0.4149,
|
| 1080 |
+
"step": 1130
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 49.56521739130435,
|
| 1084 |
+
"grad_norm": 1.3106703758239746,
|
| 1085 |
+
"learning_rate": 2.9967858108209838e-05,
|
| 1086 |
+
"loss": 0.3714,
|
| 1087 |
+
"step": 1140
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 50.0,
|
| 1091 |
+
"grad_norm": 2.9416587352752686,
|
| 1092 |
+
"learning_rate": 2.9967237249399417e-05,
|
| 1093 |
+
"loss": 0.292,
|
| 1094 |
+
"step": 1150
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 50.43478260869565,
|
| 1098 |
+
"grad_norm": 1.5631065368652344,
|
| 1099 |
+
"learning_rate": 2.996661045805436e-05,
|
| 1100 |
+
"loss": 0.2963,
|
| 1101 |
+
"step": 1160
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 50.869565217391305,
|
| 1105 |
+
"grad_norm": 1.8589760065078735,
|
| 1106 |
+
"learning_rate": 2.9965977734423106e-05,
|
| 1107 |
+
"loss": 0.3415,
|
| 1108 |
+
"step": 1170
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 51.30434782608695,
|
| 1112 |
+
"grad_norm": 1.9185295104980469,
|
| 1113 |
+
"learning_rate": 2.9965339078756445e-05,
|
| 1114 |
+
"loss": 0.3539,
|
| 1115 |
+
"step": 1180
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 51.73913043478261,
|
| 1119 |
+
"grad_norm": 1.1838868856430054,
|
| 1120 |
+
"learning_rate": 2.9964694491307514e-05,
|
| 1121 |
+
"loss": 0.2803,
|
| 1122 |
+
"step": 1190
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 52.17391304347826,
|
| 1126 |
+
"grad_norm": 2.4929492473602295,
|
| 1127 |
+
"learning_rate": 2.996404397233182e-05,
|
| 1128 |
+
"loss": 0.4083,
|
| 1129 |
+
"step": 1200
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 52.17391304347826,
|
| 1133 |
+
"eval_loss": 0.8023056983947754,
|
| 1134 |
+
"eval_runtime": 0.4133,
|
| 1135 |
+
"eval_samples_per_second": 24.196,
|
| 1136 |
+
"eval_steps_per_second": 24.196,
|
| 1137 |
+
"step": 1200
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"Start_State_loss": 0.8601926565170288,
|
| 1141 |
+
"Start_State_runtime": 0.407,
|
| 1142 |
+
"Start_State_samples_per_second": 24.569,
|
| 1143 |
+
"Start_State_steps_per_second": 24.569,
|
| 1144 |
+
"epoch": 52.17391304347826,
|
| 1145 |
+
"step": 1200
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"Raw_Model_loss": 0.8023056983947754,
|
| 1149 |
+
"Raw_Model_runtime": 0.3981,
|
| 1150 |
+
"Raw_Model_samples_per_second": 25.122,
|
| 1151 |
+
"Raw_Model_steps_per_second": 25.122,
|
| 1152 |
+
"epoch": 52.17391304347826,
|
| 1153 |
+
"step": 1200
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"SWA_loss": 0.7237697243690491,
|
| 1157 |
+
"SWA_runtime": 0.4024,
|
| 1158 |
+
"SWA_samples_per_second": 24.852,
|
| 1159 |
+
"SWA_steps_per_second": 24.852,
|
| 1160 |
+
"epoch": 52.17391304347826,
|
| 1161 |
+
"step": 1200
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"EMA_loss": 0.8598647117614746,
|
| 1165 |
+
"EMA_runtime": 0.4326,
|
| 1166 |
+
"EMA_samples_per_second": 23.115,
|
| 1167 |
+
"EMA_steps_per_second": 23.115,
|
| 1168 |
+
"epoch": 52.17391304347826,
|
| 1169 |
+
"step": 1200
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"epoch": 52.608695652173914,
|
| 1173 |
+
"grad_norm": 1.6113795042037964,
|
| 1174 |
+
"learning_rate": 1.4982021986165911e-06,
|
| 1175 |
+
"loss": 0.2866,
|
| 1176 |
+
"step": 1210
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 53.04347826086956,
|
| 1180 |
+
"grad_norm": 1.7770823240280151,
|
| 1181 |
+
"learning_rate": 2.9964043972331822e-06,
|
| 1182 |
+
"loss": 0.365,
|
| 1183 |
+
"step": 1220
|
| 1184 |
+
},
|
| 1185 |
+
{
|
| 1186 |
+
"epoch": 53.47826086956522,
|
| 1187 |
+
"grad_norm": 1.4132719039916992,
|
| 1188 |
+
"learning_rate": 4.494606595849773e-06,
|
| 1189 |
+
"loss": 0.2718,
|
| 1190 |
+
"step": 1230
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 53.91304347826087,
|
| 1194 |
+
"grad_norm": 1.9334650039672852,
|
| 1195 |
+
"learning_rate": 5.9928087944663644e-06,
|
| 1196 |
+
"loss": 0.3338,
|
| 1197 |
+
"step": 1240
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 54.34782608695652,
|
| 1201 |
+
"grad_norm": 1.9728986024856567,
|
| 1202 |
+
"learning_rate": 7.491010993082955e-06,
|
| 1203 |
+
"loss": 0.3853,
|
| 1204 |
+
"step": 1250
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"epoch": 54.78260869565217,
|
| 1208 |
+
"grad_norm": 1.1599531173706055,
|
| 1209 |
+
"learning_rate": 8.989213191699545e-06,
|
| 1210 |
+
"loss": 0.2838,
|
| 1211 |
+
"step": 1260
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 55.21739130434783,
|
| 1215 |
+
"grad_norm": 1.558973789215088,
|
| 1216 |
+
"learning_rate": 1.0487415390316136e-05,
|
| 1217 |
+
"loss": 0.311,
|
| 1218 |
+
"step": 1270
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 55.65217391304348,
|
| 1222 |
+
"grad_norm": 1.7310874462127686,
|
| 1223 |
+
"learning_rate": 1.1985617588932729e-05,
|
| 1224 |
+
"loss": 0.3553,
|
| 1225 |
+
"step": 1280
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 56.08695652173913,
|
| 1229 |
+
"grad_norm": 2.2715365886688232,
|
| 1230 |
+
"learning_rate": 1.348381978754932e-05,
|
| 1231 |
+
"loss": 0.2844,
|
| 1232 |
+
"step": 1290
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 56.52173913043478,
|
| 1236 |
+
"grad_norm": 1.467916488647461,
|
| 1237 |
+
"learning_rate": 1.498202198616591e-05,
|
| 1238 |
+
"loss": 0.3391,
|
| 1239 |
+
"step": 1300
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 56.95652173913044,
|
| 1243 |
+
"grad_norm": 1.974404215812683,
|
| 1244 |
+
"learning_rate": 1.4982020501567203e-05,
|
| 1245 |
+
"loss": 0.3314,
|
| 1246 |
+
"step": 1310
|
| 1247 |
+
},
|
| 1248 |
+
{
|
| 1249 |
+
"epoch": 57.391304347826086,
|
| 1250 |
+
"grad_norm": 1.4068485498428345,
|
| 1251 |
+
"learning_rate": 1.4982016047771664e-05,
|
| 1252 |
+
"loss": 0.3113,
|
| 1253 |
+
"step": 1320
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 57.82608695652174,
|
| 1257 |
+
"grad_norm": 2.7793936729431152,
|
| 1258 |
+
"learning_rate": 1.4982008624781062e-05,
|
| 1259 |
+
"loss": 0.3372,
|
| 1260 |
+
"step": 1330
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 58.26086956521739,
|
| 1264 |
+
"grad_norm": 1.4399445056915283,
|
| 1265 |
+
"learning_rate": 1.4981998232598337e-05,
|
| 1266 |
+
"loss": 0.3301,
|
| 1267 |
+
"step": 1340
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"epoch": 58.69565217391305,
|
| 1271 |
+
"grad_norm": 1.8218740224838257,
|
| 1272 |
+
"learning_rate": 1.4981984871227611e-05,
|
| 1273 |
+
"loss": 0.3077,
|
| 1274 |
+
"step": 1350
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 58.69565217391305,
|
| 1278 |
+
"eval_loss": 0.8209422826766968,
|
| 1279 |
+
"eval_runtime": 0.4642,
|
| 1280 |
+
"eval_samples_per_second": 21.544,
|
| 1281 |
+
"eval_steps_per_second": 21.544,
|
| 1282 |
+
"step": 1350
|
| 1283 |
+
},
|
| 1284 |
+
{
|
| 1285 |
+
"Start_State_loss": 0.8601926565170288,
|
| 1286 |
+
"Start_State_runtime": 0.4587,
|
| 1287 |
+
"Start_State_samples_per_second": 21.801,
|
| 1288 |
+
"Start_State_steps_per_second": 21.801,
|
| 1289 |
+
"epoch": 58.69565217391305,
|
| 1290 |
+
"step": 1350
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"Raw_Model_loss": 0.8209422826766968,
|
| 1294 |
+
"Raw_Model_runtime": 0.5144,
|
| 1295 |
+
"Raw_Model_samples_per_second": 19.438,
|
| 1296 |
+
"Raw_Model_steps_per_second": 19.438,
|
| 1297 |
+
"epoch": 58.69565217391305,
|
| 1298 |
+
"step": 1350
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"SWA_loss": 0.7251114249229431,
|
| 1302 |
+
"SWA_runtime": 0.4605,
|
| 1303 |
+
"SWA_samples_per_second": 21.718,
|
| 1304 |
+
"SWA_steps_per_second": 21.718,
|
| 1305 |
+
"epoch": 58.69565217391305,
|
| 1306 |
+
"step": 1350
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"EMA_loss": 0.8608489036560059,
|
| 1310 |
+
"EMA_runtime": 0.4317,
|
| 1311 |
+
"EMA_samples_per_second": 23.166,
|
| 1312 |
+
"EMA_steps_per_second": 23.166,
|
| 1313 |
+
"epoch": 58.69565217391305,
|
| 1314 |
+
"step": 1350
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 59.130434782608695,
|
| 1318 |
+
"grad_norm": 1.5807944536209106,
|
| 1319 |
+
"learning_rate": 1.4981968540674177e-05,
|
| 1320 |
+
"loss": 0.3206,
|
| 1321 |
+
"step": 1360
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"epoch": 59.56521739130435,
|
| 1325 |
+
"grad_norm": 1.40355384349823,
|
| 1326 |
+
"learning_rate": 1.4981949240944509e-05,
|
| 1327 |
+
"loss": 0.3012,
|
| 1328 |
+
"step": 1370
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 60.0,
|
| 1332 |
+
"grad_norm": 1.6165056228637695,
|
| 1333 |
+
"learning_rate": 1.4981926972046258e-05,
|
| 1334 |
+
"loss": 0.3098,
|
| 1335 |
+
"step": 1380
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 60.43478260869565,
|
| 1339 |
+
"grad_norm": 1.9167027473449707,
|
| 1340 |
+
"learning_rate": 1.498190173398825e-05,
|
| 1341 |
+
"loss": 0.3171,
|
| 1342 |
+
"step": 1390
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 60.869565217391305,
|
| 1346 |
+
"grad_norm": 1.539297342300415,
|
| 1347 |
+
"learning_rate": 1.4981873526780487e-05,
|
| 1348 |
+
"loss": 0.3049,
|
| 1349 |
+
"step": 1400
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 61.30434782608695,
|
| 1353 |
+
"grad_norm": 1.4211211204528809,
|
| 1354 |
+
"learning_rate": 1.4981842350434152e-05,
|
| 1355 |
+
"loss": 0.3045,
|
| 1356 |
+
"step": 1410
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 61.73913043478261,
|
| 1360 |
+
"grad_norm": 1.4864341020584106,
|
| 1361 |
+
"learning_rate": 1.49818082049616e-05,
|
| 1362 |
+
"loss": 0.3207,
|
| 1363 |
+
"step": 1420
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 62.17391304347826,
|
| 1367 |
+
"grad_norm": 2.1776299476623535,
|
| 1368 |
+
"learning_rate": 1.4981771090376367e-05,
|
| 1369 |
+
"loss": 0.2862,
|
| 1370 |
+
"step": 1430
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 62.608695652173914,
|
| 1374 |
+
"grad_norm": 1.8853501081466675,
|
| 1375 |
+
"learning_rate": 1.4981731006693164e-05,
|
| 1376 |
+
"loss": 0.3212,
|
| 1377 |
+
"step": 1440
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 63.04347826086956,
|
| 1381 |
+
"grad_norm": 1.3142286539077759,
|
| 1382 |
+
"learning_rate": 1.4981687953927875e-05,
|
| 1383 |
+
"loss": 0.3127,
|
| 1384 |
+
"step": 1450
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 63.47826086956522,
|
| 1388 |
+
"grad_norm": 1.9734851121902466,
|
| 1389 |
+
"learning_rate": 1.498164193209757e-05,
|
| 1390 |
+
"loss": 0.3447,
|
| 1391 |
+
"step": 1460
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 63.91304347826087,
|
| 1395 |
+
"grad_norm": 1.655447006225586,
|
| 1396 |
+
"learning_rate": 1.498159294122049e-05,
|
| 1397 |
+
"loss": 0.2921,
|
| 1398 |
+
"step": 1470
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 64.34782608695652,
|
| 1402 |
+
"grad_norm": 1.7767964601516724,
|
| 1403 |
+
"learning_rate": 1.4981540981316052e-05,
|
| 1404 |
+
"loss": 0.269,
|
| 1405 |
+
"step": 1480
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 64.78260869565217,
|
| 1409 |
+
"grad_norm": 1.5196256637573242,
|
| 1410 |
+
"learning_rate": 1.4981486052404848e-05,
|
| 1411 |
+
"loss": 0.3583,
|
| 1412 |
+
"step": 1490
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 65.21739130434783,
|
| 1416 |
+
"grad_norm": 1.4027047157287598,
|
| 1417 |
+
"learning_rate": 1.4981428154508652e-05,
|
| 1418 |
+
"loss": 0.2693,
|
| 1419 |
+
"step": 1500
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 65.21739130434783,
|
| 1423 |
+
"eval_loss": 0.832839846611023,
|
| 1424 |
+
"eval_runtime": 0.4275,
|
| 1425 |
+
"eval_samples_per_second": 23.391,
|
| 1426 |
+
"eval_steps_per_second": 23.391,
|
| 1427 |
+
"step": 1500
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"Start_State_loss": 0.8601926565170288,
|
| 1431 |
+
"Start_State_runtime": 0.4059,
|
| 1432 |
+
"Start_State_samples_per_second": 24.634,
|
| 1433 |
+
"Start_State_steps_per_second": 24.634,
|
| 1434 |
+
"epoch": 65.21739130434783,
|
| 1435 |
+
"step": 1500
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"Raw_Model_loss": 0.832839846611023,
|
| 1439 |
+
"Raw_Model_runtime": 0.3946,
|
| 1440 |
+
"Raw_Model_samples_per_second": 25.341,
|
| 1441 |
+
"Raw_Model_steps_per_second": 25.341,
|
| 1442 |
+
"epoch": 65.21739130434783,
|
| 1443 |
+
"step": 1500
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"SWA_loss": 0.7298181056976318,
|
| 1447 |
+
"SWA_runtime": 0.3986,
|
| 1448 |
+
"SWA_samples_per_second": 25.087,
|
| 1449 |
+
"SWA_steps_per_second": 25.087,
|
| 1450 |
+
"epoch": 65.21739130434783,
|
| 1451 |
+
"step": 1500
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"EMA_loss": 0.8607869148254395,
|
| 1455 |
+
"EMA_runtime": 0.4036,
|
| 1456 |
+
"EMA_samples_per_second": 24.774,
|
| 1457 |
+
"EMA_steps_per_second": 24.774,
|
| 1458 |
+
"epoch": 65.21739130434783,
|
| 1459 |
+
"step": 1500
|
| 1460 |
}
|
| 1461 |
],
|
| 1462 |
"logging_steps": 10,
|
|
|
|
| 1476 |
"attributes": {}
|
| 1477 |
}
|
| 1478 |
},
|
| 1479 |
+
"total_flos": 3.854430872108237e+16,
|
| 1480 |
"train_batch_size": 4,
|
| 1481 |
"trial_name": null,
|
| 1482 |
"trial_params": null
|