Training in progress, step 150, checkpoint
Browse files- last-checkpoint/adapter_model.safetensors +1 -1
- last-checkpoint/optimizer.pt +1 -1
- last-checkpoint/rng_state_0.pth +1 -1
- last-checkpoint/rng_state_1.pth +1 -1
- last-checkpoint/rng_state_2.pth +1 -1
- last-checkpoint/rng_state_3.pth +1 -1
- last-checkpoint/rng_state_4.pth +1 -1
- last-checkpoint/rng_state_5.pth +1 -1
- last-checkpoint/rng_state_6.pth +1 -1
- last-checkpoint/rng_state_7.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +363 -5
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1157746040
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d6478c44dcfe47c66e4bdb2ba94b92fae5c70ae71b5c5a0d8ea6b7672d5e6e71
|
| 3 |
size 1157746040
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 588699796
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cd29317c2b5752c34c30b11c987e6a9a73a267623ac7dd1f34dea83250278bb0
|
| 3 |
size 588699796
|
last-checkpoint/rng_state_0.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:10aa2f7f27f33adc50a21abe7d03eb50cd7ff3458bacfe21c4799b75bb717e18
|
| 3 |
size 15984
|
last-checkpoint/rng_state_1.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c72bcbbd26c8fcfe452930e80a6a47167cb37dbf0bdd0200ff327d7e1132715e
|
| 3 |
size 15984
|
last-checkpoint/rng_state_2.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cc8184544394f1a57802d6333ead772da158364e2fdd23f212103e4adc4129fb
|
| 3 |
size 15984
|
last-checkpoint/rng_state_3.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5b0c20ad89e1dade716473de79b6a1b09484cafc267937e7b2b3559110ba1465
|
| 3 |
size 15984
|
last-checkpoint/rng_state_4.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:570729bd73d3b7473398d29397a9604c4a007633b21fa95116088a959e2aba23
|
| 3 |
size 15984
|
last-checkpoint/rng_state_5.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:03c5626167b0c70752e1312fcf7e67d6398f540bf389d7c2280409358563f97e
|
| 3 |
size 15984
|
last-checkpoint/rng_state_6.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ac9b7b84707a25e39a97b7cfc050664d7ad683356a7793e16bcfebddae502898
|
| 3 |
size 15984
|
last-checkpoint/rng_state_7.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:98b5bf0d3c51757e3c2768a4f754c048cee95ca5ec2b8320316be70a556963ec
|
| 3 |
size 15984
|
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 sha256:7d606eeb1aa97b417de3c30d0a970be83ac979e2c7cc0fa41135c63d459909e5
|
| 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": 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 +731,364 @@
|
|
| 731 |
"eval_samples_per_second": 61.624,
|
| 732 |
"eval_steps_per_second": 1.927,
|
| 733 |
"step": 100
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 734 |
}
|
| 735 |
],
|
| 736 |
"logging_steps": 1,
|
|
@@ -759,7 +1117,7 @@
|
|
| 759 |
"attributes": {}
|
| 760 |
}
|
| 761 |
},
|
| 762 |
-
"total_flos":
|
| 763 |
"train_batch_size": 8,
|
| 764 |
"trial_name": null,
|
| 765 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.40253087878227234,
|
| 3 |
+
"best_model_checkpoint": "miner_id_24/checkpoint-150",
|
| 4 |
+
"epoch": 0.20060180541624875,
|
| 5 |
"eval_steps": 50,
|
| 6 |
+
"global_step": 150,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 731 |
"eval_samples_per_second": 61.624,
|
| 732 |
"eval_steps_per_second": 1.927,
|
| 733 |
"step": 100
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 0.13507188231360748,
|
| 737 |
+
"grad_norm": 0.5436654090881348,
|
| 738 |
+
"learning_rate": 5.262631578947368e-05,
|
| 739 |
+
"loss": 0.4414,
|
| 740 |
+
"step": 101
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 0.13640922768304914,
|
| 744 |
+
"grad_norm": 0.500196635723114,
|
| 745 |
+
"learning_rate": 5.209473684210527e-05,
|
| 746 |
+
"loss": 0.4313,
|
| 747 |
+
"step": 102
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 0.1377465730524908,
|
| 751 |
+
"grad_norm": 0.5394172668457031,
|
| 752 |
+
"learning_rate": 5.1563157894736844e-05,
|
| 753 |
+
"loss": 0.3968,
|
| 754 |
+
"step": 103
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 0.13908391842193246,
|
| 758 |
+
"grad_norm": 0.5281717777252197,
|
| 759 |
+
"learning_rate": 5.1031578947368426e-05,
|
| 760 |
+
"loss": 0.4193,
|
| 761 |
+
"step": 104
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.14042126379137412,
|
| 765 |
+
"grad_norm": 0.5725506544113159,
|
| 766 |
+
"learning_rate": 5.05e-05,
|
| 767 |
+
"loss": 0.4368,
|
| 768 |
+
"step": 105
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 0.14175860916081578,
|
| 772 |
+
"grad_norm": 0.614000678062439,
|
| 773 |
+
"learning_rate": 4.9968421052631576e-05,
|
| 774 |
+
"loss": 0.4904,
|
| 775 |
+
"step": 106
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 0.14309595453025745,
|
| 779 |
+
"grad_norm": 0.5707330107688904,
|
| 780 |
+
"learning_rate": 4.943684210526316e-05,
|
| 781 |
+
"loss": 0.4335,
|
| 782 |
+
"step": 107
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.1444332998996991,
|
| 786 |
+
"grad_norm": 0.8218697905540466,
|
| 787 |
+
"learning_rate": 4.890526315789474e-05,
|
| 788 |
+
"loss": 0.4607,
|
| 789 |
+
"step": 108
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.14577064526914077,
|
| 793 |
+
"grad_norm": 0.49518197774887085,
|
| 794 |
+
"learning_rate": 4.8373684210526316e-05,
|
| 795 |
+
"loss": 0.4167,
|
| 796 |
+
"step": 109
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 0.1471079906385824,
|
| 800 |
+
"grad_norm": 0.5105622410774231,
|
| 801 |
+
"learning_rate": 4.784210526315789e-05,
|
| 802 |
+
"loss": 0.3983,
|
| 803 |
+
"step": 110
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 0.14844533600802406,
|
| 807 |
+
"grad_norm": 0.559273898601532,
|
| 808 |
+
"learning_rate": 4.731052631578947e-05,
|
| 809 |
+
"loss": 0.4263,
|
| 810 |
+
"step": 111
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 0.14978268137746573,
|
| 814 |
+
"grad_norm": 0.634856641292572,
|
| 815 |
+
"learning_rate": 4.6778947368421055e-05,
|
| 816 |
+
"loss": 0.4869,
|
| 817 |
+
"step": 112
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 0.1511200267469074,
|
| 821 |
+
"grad_norm": 0.557166337966919,
|
| 822 |
+
"learning_rate": 4.624736842105263e-05,
|
| 823 |
+
"loss": 0.4447,
|
| 824 |
+
"step": 113
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.15245737211634905,
|
| 828 |
+
"grad_norm": 0.4786439836025238,
|
| 829 |
+
"learning_rate": 4.571578947368421e-05,
|
| 830 |
+
"loss": 0.4436,
|
| 831 |
+
"step": 114
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.1537947174857907,
|
| 835 |
+
"grad_norm": 0.47446924448013306,
|
| 836 |
+
"learning_rate": 4.518421052631579e-05,
|
| 837 |
+
"loss": 0.3959,
|
| 838 |
+
"step": 115
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 0.15513206285523237,
|
| 842 |
+
"grad_norm": 0.5089144110679626,
|
| 843 |
+
"learning_rate": 4.465263157894737e-05,
|
| 844 |
+
"loss": 0.4078,
|
| 845 |
+
"step": 116
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 0.15646940822467403,
|
| 849 |
+
"grad_norm": 0.6025608777999878,
|
| 850 |
+
"learning_rate": 4.412105263157895e-05,
|
| 851 |
+
"loss": 0.4322,
|
| 852 |
+
"step": 117
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 0.15780675359411567,
|
| 856 |
+
"grad_norm": 0.6016778945922852,
|
| 857 |
+
"learning_rate": 4.358947368421053e-05,
|
| 858 |
+
"loss": 0.4408,
|
| 859 |
+
"step": 118
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 0.15914409896355733,
|
| 863 |
+
"grad_norm": 0.6104781031608582,
|
| 864 |
+
"learning_rate": 4.30578947368421e-05,
|
| 865 |
+
"loss": 0.4705,
|
| 866 |
+
"step": 119
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.160481444332999,
|
| 870 |
+
"grad_norm": 0.49418070912361145,
|
| 871 |
+
"learning_rate": 4.2526315789473685e-05,
|
| 872 |
+
"loss": 0.4095,
|
| 873 |
+
"step": 120
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.16181878970244065,
|
| 877 |
+
"grad_norm": 0.49163714051246643,
|
| 878 |
+
"learning_rate": 4.199473684210527e-05,
|
| 879 |
+
"loss": 0.3926,
|
| 880 |
+
"step": 121
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 0.1631561350718823,
|
| 884 |
+
"grad_norm": 0.4370480179786682,
|
| 885 |
+
"learning_rate": 4.146315789473684e-05,
|
| 886 |
+
"loss": 0.3847,
|
| 887 |
+
"step": 122
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 0.16449348044132397,
|
| 891 |
+
"grad_norm": 0.5089737772941589,
|
| 892 |
+
"learning_rate": 4.093157894736842e-05,
|
| 893 |
+
"loss": 0.3866,
|
| 894 |
+
"step": 123
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 0.16583082581076564,
|
| 898 |
+
"grad_norm": 0.5304917097091675,
|
| 899 |
+
"learning_rate": 4.0400000000000006e-05,
|
| 900 |
+
"loss": 0.4045,
|
| 901 |
+
"step": 124
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 0.1671681711802073,
|
| 905 |
+
"grad_norm": 0.6393993496894836,
|
| 906 |
+
"learning_rate": 3.986842105263158e-05,
|
| 907 |
+
"loss": 0.5036,
|
| 908 |
+
"step": 125
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 0.16850551654964896,
|
| 912 |
+
"grad_norm": 0.544893741607666,
|
| 913 |
+
"learning_rate": 3.933684210526316e-05,
|
| 914 |
+
"loss": 0.3761,
|
| 915 |
+
"step": 126
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 0.1698428619190906,
|
| 919 |
+
"grad_norm": 0.5071455240249634,
|
| 920 |
+
"learning_rate": 3.880526315789473e-05,
|
| 921 |
+
"loss": 0.4161,
|
| 922 |
+
"step": 127
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 0.17118020728853225,
|
| 926 |
+
"grad_norm": 0.48931699991226196,
|
| 927 |
+
"learning_rate": 3.827368421052632e-05,
|
| 928 |
+
"loss": 0.384,
|
| 929 |
+
"step": 128
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 0.17251755265797392,
|
| 933 |
+
"grad_norm": 0.5249499678611755,
|
| 934 |
+
"learning_rate": 3.7742105263157896e-05,
|
| 935 |
+
"loss": 0.3637,
|
| 936 |
+
"step": 129
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 0.17385489802741558,
|
| 940 |
+
"grad_norm": 0.6430131196975708,
|
| 941 |
+
"learning_rate": 3.721052631578947e-05,
|
| 942 |
+
"loss": 0.4372,
|
| 943 |
+
"step": 130
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 0.17519224339685724,
|
| 947 |
+
"grad_norm": 0.5788580775260925,
|
| 948 |
+
"learning_rate": 3.6678947368421054e-05,
|
| 949 |
+
"loss": 0.4384,
|
| 950 |
+
"step": 131
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.1765295887662989,
|
| 954 |
+
"grad_norm": 0.525442361831665,
|
| 955 |
+
"learning_rate": 3.6147368421052636e-05,
|
| 956 |
+
"loss": 0.4069,
|
| 957 |
+
"step": 132
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 0.17786693413574056,
|
| 961 |
+
"grad_norm": 0.491260826587677,
|
| 962 |
+
"learning_rate": 3.561578947368421e-05,
|
| 963 |
+
"loss": 0.4497,
|
| 964 |
+
"step": 133
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 0.17920427950518222,
|
| 968 |
+
"grad_norm": 0.46042758226394653,
|
| 969 |
+
"learning_rate": 3.508421052631579e-05,
|
| 970 |
+
"loss": 0.3671,
|
| 971 |
+
"step": 134
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 0.18054162487462388,
|
| 975 |
+
"grad_norm": 0.5297501087188721,
|
| 976 |
+
"learning_rate": 3.455263157894737e-05,
|
| 977 |
+
"loss": 0.3714,
|
| 978 |
+
"step": 135
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 0.18187897024406552,
|
| 982 |
+
"grad_norm": 0.5078780651092529,
|
| 983 |
+
"learning_rate": 3.402105263157895e-05,
|
| 984 |
+
"loss": 0.3958,
|
| 985 |
+
"step": 136
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 0.18321631561350718,
|
| 989 |
+
"grad_norm": 0.6146392226219177,
|
| 990 |
+
"learning_rate": 3.3489473684210526e-05,
|
| 991 |
+
"loss": 0.4246,
|
| 992 |
+
"step": 137
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.18455366098294884,
|
| 996 |
+
"grad_norm": 0.6537202596664429,
|
| 997 |
+
"learning_rate": 3.295789473684211e-05,
|
| 998 |
+
"loss": 0.4329,
|
| 999 |
+
"step": 138
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.1858910063523905,
|
| 1003 |
+
"grad_norm": 0.5298619270324707,
|
| 1004 |
+
"learning_rate": 3.242631578947368e-05,
|
| 1005 |
+
"loss": 0.4508,
|
| 1006 |
+
"step": 139
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 0.18722835172183216,
|
| 1010 |
+
"grad_norm": 0.4924841523170471,
|
| 1011 |
+
"learning_rate": 3.1894736842105265e-05,
|
| 1012 |
+
"loss": 0.3845,
|
| 1013 |
+
"step": 140
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 0.18856569709127383,
|
| 1017 |
+
"grad_norm": 0.4037409722805023,
|
| 1018 |
+
"learning_rate": 3.136315789473685e-05,
|
| 1019 |
+
"loss": 0.3572,
|
| 1020 |
+
"step": 141
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 0.1899030424607155,
|
| 1024 |
+
"grad_norm": 0.5435509085655212,
|
| 1025 |
+
"learning_rate": 3.083157894736842e-05,
|
| 1026 |
+
"loss": 0.3659,
|
| 1027 |
+
"step": 142
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 0.19124038783015715,
|
| 1031 |
+
"grad_norm": 0.5061467289924622,
|
| 1032 |
+
"learning_rate": 3.0299999999999998e-05,
|
| 1033 |
+
"loss": 0.4057,
|
| 1034 |
+
"step": 143
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 0.19257773319959878,
|
| 1038 |
+
"grad_norm": 0.5949498414993286,
|
| 1039 |
+
"learning_rate": 2.9768421052631577e-05,
|
| 1040 |
+
"loss": 0.4538,
|
| 1041 |
+
"step": 144
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 0.19391507856904044,
|
| 1045 |
+
"grad_norm": 0.4520629346370697,
|
| 1046 |
+
"learning_rate": 2.923684210526316e-05,
|
| 1047 |
+
"loss": 0.4127,
|
| 1048 |
+
"step": 145
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 0.1952524239384821,
|
| 1052 |
+
"grad_norm": 0.4914908707141876,
|
| 1053 |
+
"learning_rate": 2.8705263157894737e-05,
|
| 1054 |
+
"loss": 0.4048,
|
| 1055 |
+
"step": 146
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 0.19658976930792377,
|
| 1059 |
+
"grad_norm": 0.5175070762634277,
|
| 1060 |
+
"learning_rate": 2.8173684210526313e-05,
|
| 1061 |
+
"loss": 0.4261,
|
| 1062 |
+
"step": 147
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 0.19792711467736543,
|
| 1066 |
+
"grad_norm": 0.4606378376483917,
|
| 1067 |
+
"learning_rate": 2.7642105263157898e-05,
|
| 1068 |
+
"loss": 0.3737,
|
| 1069 |
+
"step": 148
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 0.1992644600468071,
|
| 1073 |
+
"grad_norm": 0.5699568390846252,
|
| 1074 |
+
"learning_rate": 2.7110526315789473e-05,
|
| 1075 |
+
"loss": 0.4222,
|
| 1076 |
+
"step": 149
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 0.20060180541624875,
|
| 1080 |
+
"grad_norm": 0.5272060632705688,
|
| 1081 |
+
"learning_rate": 2.6578947368421052e-05,
|
| 1082 |
+
"loss": 0.4438,
|
| 1083 |
+
"step": 150
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 0.20060180541624875,
|
| 1087 |
+
"eval_loss": 0.40253087878227234,
|
| 1088 |
+
"eval_runtime": 163.1972,
|
| 1089 |
+
"eval_samples_per_second": 61.735,
|
| 1090 |
+
"eval_steps_per_second": 1.93,
|
| 1091 |
+
"step": 150
|
| 1092 |
}
|
| 1093 |
],
|
| 1094 |
"logging_steps": 1,
|
|
|
|
| 1117 |
"attributes": {}
|
| 1118 |
}
|
| 1119 |
},
|
| 1120 |
+
"total_flos": 3.172246609849221e+18,
|
| 1121 |
"train_batch_size": 8,
|
| 1122 |
"trial_name": null,
|
| 1123 |
"trial_params": null
|