Farouk
commited on
Commit
·
bb0c7cf
1
Parent(s):
3fe2bb9
commit files to HF hub
Browse files- adapter_model.bin +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +576 -3
adapter_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 319977229
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:698f7f4394137c9f9a027f61094a137c19b351bca2776822ea338ffdc3b048db
|
| 3 |
size 319977229
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1279539973
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2e9b62486989ca29406a18af203ed5b33e5c27bed51df27f20aa9786e46628ff
|
| 3 |
size 1279539973
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14511
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c027db1dc669a948797121e90c1a982d60d14a868af0519ca5f007f6183bbdd0
|
| 3 |
size 14511
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 627
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d1351871b6e069a7a4437616d40b913152ecc2cd046e996fcfeb300a3cf60638
|
| 3 |
size 627
|
trainer_state.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": 0.6964578628540039,
|
| 3 |
"best_model_checkpoint": "./output_v2/7b_cluster06_Nous-Hermes-llama-2-7b_partitioned_v3_standardized_06/checkpoint-800",
|
| 4 |
-
"epoch":
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
@@ -770,11 +770,584 @@
|
|
| 770 |
"mmlu_eval_accuracy_world_religions": 0.7368421052631579,
|
| 771 |
"mmlu_loss": 1.069855675548704,
|
| 772 |
"step": 800
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 773 |
}
|
| 774 |
],
|
| 775 |
"max_steps": 5000,
|
| 776 |
"num_train_epochs": 12,
|
| 777 |
-
"total_flos":
|
| 778 |
"trial_name": null,
|
| 779 |
"trial_params": null
|
| 780 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": 0.6964578628540039,
|
| 3 |
"best_model_checkpoint": "./output_v2/7b_cluster06_Nous-Hermes-llama-2-7b_partitioned_v3_standardized_06/checkpoint-800",
|
| 4 |
+
"epoch": 3.218390804597701,
|
| 5 |
+
"global_step": 1400,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
|
|
| 770 |
"mmlu_eval_accuracy_world_religions": 0.7368421052631579,
|
| 771 |
"mmlu_loss": 1.069855675548704,
|
| 772 |
"step": 800
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.86,
|
| 776 |
+
"learning_rate": 0.0002,
|
| 777 |
+
"loss": 0.6798,
|
| 778 |
+
"step": 810
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 1.89,
|
| 782 |
+
"learning_rate": 0.0002,
|
| 783 |
+
"loss": 0.6485,
|
| 784 |
+
"step": 820
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 1.91,
|
| 788 |
+
"learning_rate": 0.0002,
|
| 789 |
+
"loss": 0.6419,
|
| 790 |
+
"step": 830
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 1.93,
|
| 794 |
+
"learning_rate": 0.0002,
|
| 795 |
+
"loss": 0.6528,
|
| 796 |
+
"step": 840
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 1.95,
|
| 800 |
+
"learning_rate": 0.0002,
|
| 801 |
+
"loss": 0.6674,
|
| 802 |
+
"step": 850
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 1.98,
|
| 806 |
+
"learning_rate": 0.0002,
|
| 807 |
+
"loss": 0.6487,
|
| 808 |
+
"step": 860
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 2.0,
|
| 812 |
+
"learning_rate": 0.0002,
|
| 813 |
+
"loss": 0.6742,
|
| 814 |
+
"step": 870
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 2.02,
|
| 818 |
+
"learning_rate": 0.0002,
|
| 819 |
+
"loss": 0.5303,
|
| 820 |
+
"step": 880
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 2.05,
|
| 824 |
+
"learning_rate": 0.0002,
|
| 825 |
+
"loss": 0.5264,
|
| 826 |
+
"step": 890
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 2.07,
|
| 830 |
+
"learning_rate": 0.0002,
|
| 831 |
+
"loss": 0.5578,
|
| 832 |
+
"step": 900
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 2.09,
|
| 836 |
+
"learning_rate": 0.0002,
|
| 837 |
+
"loss": 0.5399,
|
| 838 |
+
"step": 910
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 2.11,
|
| 842 |
+
"learning_rate": 0.0002,
|
| 843 |
+
"loss": 0.6028,
|
| 844 |
+
"step": 920
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 2.14,
|
| 848 |
+
"learning_rate": 0.0002,
|
| 849 |
+
"loss": 0.5292,
|
| 850 |
+
"step": 930
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 2.16,
|
| 854 |
+
"learning_rate": 0.0002,
|
| 855 |
+
"loss": 0.5715,
|
| 856 |
+
"step": 940
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 2.18,
|
| 860 |
+
"learning_rate": 0.0002,
|
| 861 |
+
"loss": 0.514,
|
| 862 |
+
"step": 950
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 2.21,
|
| 866 |
+
"learning_rate": 0.0002,
|
| 867 |
+
"loss": 0.5381,
|
| 868 |
+
"step": 960
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 2.23,
|
| 872 |
+
"learning_rate": 0.0002,
|
| 873 |
+
"loss": 0.5259,
|
| 874 |
+
"step": 970
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 2.25,
|
| 878 |
+
"learning_rate": 0.0002,
|
| 879 |
+
"loss": 0.5476,
|
| 880 |
+
"step": 980
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 2.28,
|
| 884 |
+
"learning_rate": 0.0002,
|
| 885 |
+
"loss": 0.5369,
|
| 886 |
+
"step": 990
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 2.3,
|
| 890 |
+
"learning_rate": 0.0002,
|
| 891 |
+
"loss": 0.5541,
|
| 892 |
+
"step": 1000
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 2.3,
|
| 896 |
+
"eval_loss": 0.7225061058998108,
|
| 897 |
+
"eval_runtime": 248.2663,
|
| 898 |
+
"eval_samples_per_second": 4.028,
|
| 899 |
+
"eval_steps_per_second": 2.014,
|
| 900 |
+
"step": 1000
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 2.3,
|
| 904 |
+
"mmlu_eval_accuracy": 0.46504233740978407,
|
| 905 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
|
| 906 |
+
"mmlu_eval_accuracy_anatomy": 0.5714285714285714,
|
| 907 |
+
"mmlu_eval_accuracy_astronomy": 0.375,
|
| 908 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 909 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.4482758620689655,
|
| 910 |
+
"mmlu_eval_accuracy_college_biology": 0.4375,
|
| 911 |
+
"mmlu_eval_accuracy_college_chemistry": 0.125,
|
| 912 |
+
"mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
|
| 913 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 914 |
+
"mmlu_eval_accuracy_college_medicine": 0.36363636363636365,
|
| 915 |
+
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
|
| 916 |
+
"mmlu_eval_accuracy_computer_security": 0.2727272727272727,
|
| 917 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.3076923076923077,
|
| 918 |
+
"mmlu_eval_accuracy_econometrics": 0.16666666666666666,
|
| 919 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.5,
|
| 920 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
|
| 921 |
+
"mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
|
| 922 |
+
"mmlu_eval_accuracy_global_facts": 0.4,
|
| 923 |
+
"mmlu_eval_accuracy_high_school_biology": 0.34375,
|
| 924 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
|
| 925 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
|
| 926 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
|
| 927 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
|
| 928 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
|
| 929 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.32558139534883723,
|
| 930 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724,
|
| 931 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.38461538461538464,
|
| 932 |
+
"mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
|
| 933 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.7,
|
| 934 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
|
| 935 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
|
| 936 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
|
| 937 |
+
"mmlu_eval_accuracy_human_aging": 0.6521739130434783,
|
| 938 |
+
"mmlu_eval_accuracy_human_sexuality": 0.5,
|
| 939 |
+
"mmlu_eval_accuracy_international_law": 0.9230769230769231,
|
| 940 |
+
"mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
|
| 941 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.5555555555555556,
|
| 942 |
+
"mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
|
| 943 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 944 |
+
"mmlu_eval_accuracy_marketing": 0.72,
|
| 945 |
+
"mmlu_eval_accuracy_medical_genetics": 0.6363636363636364,
|
| 946 |
+
"mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
|
| 947 |
+
"mmlu_eval_accuracy_moral_disputes": 0.4473684210526316,
|
| 948 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.24,
|
| 949 |
+
"mmlu_eval_accuracy_nutrition": 0.6060606060606061,
|
| 950 |
+
"mmlu_eval_accuracy_philosophy": 0.35294117647058826,
|
| 951 |
+
"mmlu_eval_accuracy_prehistory": 0.42857142857142855,
|
| 952 |
+
"mmlu_eval_accuracy_professional_accounting": 0.25806451612903225,
|
| 953 |
+
"mmlu_eval_accuracy_professional_law": 0.3588235294117647,
|
| 954 |
+
"mmlu_eval_accuracy_professional_medicine": 0.41935483870967744,
|
| 955 |
+
"mmlu_eval_accuracy_professional_psychology": 0.37681159420289856,
|
| 956 |
+
"mmlu_eval_accuracy_public_relations": 0.5,
|
| 957 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 958 |
+
"mmlu_eval_accuracy_sociology": 0.6818181818181818,
|
| 959 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
|
| 960 |
+
"mmlu_eval_accuracy_virology": 0.4444444444444444,
|
| 961 |
+
"mmlu_eval_accuracy_world_religions": 0.7894736842105263,
|
| 962 |
+
"mmlu_loss": 1.1883746445879302,
|
| 963 |
+
"step": 1000
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 2.32,
|
| 967 |
+
"learning_rate": 0.0002,
|
| 968 |
+
"loss": 0.5264,
|
| 969 |
+
"step": 1010
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 2.34,
|
| 973 |
+
"learning_rate": 0.0002,
|
| 974 |
+
"loss": 0.5494,
|
| 975 |
+
"step": 1020
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 2.37,
|
| 979 |
+
"learning_rate": 0.0002,
|
| 980 |
+
"loss": 0.5728,
|
| 981 |
+
"step": 1030
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 2.39,
|
| 985 |
+
"learning_rate": 0.0002,
|
| 986 |
+
"loss": 0.5291,
|
| 987 |
+
"step": 1040
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 2.41,
|
| 991 |
+
"learning_rate": 0.0002,
|
| 992 |
+
"loss": 0.523,
|
| 993 |
+
"step": 1050
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 2.44,
|
| 997 |
+
"learning_rate": 0.0002,
|
| 998 |
+
"loss": 0.5893,
|
| 999 |
+
"step": 1060
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 2.46,
|
| 1003 |
+
"learning_rate": 0.0002,
|
| 1004 |
+
"loss": 0.5839,
|
| 1005 |
+
"step": 1070
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 2.48,
|
| 1009 |
+
"learning_rate": 0.0002,
|
| 1010 |
+
"loss": 0.5653,
|
| 1011 |
+
"step": 1080
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 2.51,
|
| 1015 |
+
"learning_rate": 0.0002,
|
| 1016 |
+
"loss": 0.5518,
|
| 1017 |
+
"step": 1090
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 2.53,
|
| 1021 |
+
"learning_rate": 0.0002,
|
| 1022 |
+
"loss": 0.5497,
|
| 1023 |
+
"step": 1100
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 2.55,
|
| 1027 |
+
"learning_rate": 0.0002,
|
| 1028 |
+
"loss": 0.5789,
|
| 1029 |
+
"step": 1110
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 2.57,
|
| 1033 |
+
"learning_rate": 0.0002,
|
| 1034 |
+
"loss": 0.5358,
|
| 1035 |
+
"step": 1120
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 2.6,
|
| 1039 |
+
"learning_rate": 0.0002,
|
| 1040 |
+
"loss": 0.5576,
|
| 1041 |
+
"step": 1130
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 2.62,
|
| 1045 |
+
"learning_rate": 0.0002,
|
| 1046 |
+
"loss": 0.5015,
|
| 1047 |
+
"step": 1140
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 2.64,
|
| 1051 |
+
"learning_rate": 0.0002,
|
| 1052 |
+
"loss": 0.5494,
|
| 1053 |
+
"step": 1150
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 2.67,
|
| 1057 |
+
"learning_rate": 0.0002,
|
| 1058 |
+
"loss": 0.5482,
|
| 1059 |
+
"step": 1160
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 2.69,
|
| 1063 |
+
"learning_rate": 0.0002,
|
| 1064 |
+
"loss": 0.5882,
|
| 1065 |
+
"step": 1170
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 2.71,
|
| 1069 |
+
"learning_rate": 0.0002,
|
| 1070 |
+
"loss": 0.5525,
|
| 1071 |
+
"step": 1180
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 2.74,
|
| 1075 |
+
"learning_rate": 0.0002,
|
| 1076 |
+
"loss": 0.5455,
|
| 1077 |
+
"step": 1190
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 2.76,
|
| 1081 |
+
"learning_rate": 0.0002,
|
| 1082 |
+
"loss": 0.5813,
|
| 1083 |
+
"step": 1200
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 2.76,
|
| 1087 |
+
"eval_loss": 0.7202425003051758,
|
| 1088 |
+
"eval_runtime": 248.3045,
|
| 1089 |
+
"eval_samples_per_second": 4.027,
|
| 1090 |
+
"eval_steps_per_second": 2.014,
|
| 1091 |
+
"step": 1200
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 2.76,
|
| 1095 |
+
"mmlu_eval_accuracy": 0.4683330314670407,
|
| 1096 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
|
| 1097 |
+
"mmlu_eval_accuracy_anatomy": 0.5714285714285714,
|
| 1098 |
+
"mmlu_eval_accuracy_astronomy": 0.3125,
|
| 1099 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 1100 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
|
| 1101 |
+
"mmlu_eval_accuracy_college_biology": 0.3125,
|
| 1102 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 1103 |
+
"mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
|
| 1104 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 1105 |
+
"mmlu_eval_accuracy_college_medicine": 0.5,
|
| 1106 |
+
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
|
| 1107 |
+
"mmlu_eval_accuracy_computer_security": 0.45454545454545453,
|
| 1108 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
|
| 1109 |
+
"mmlu_eval_accuracy_econometrics": 0.16666666666666666,
|
| 1110 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.4375,
|
| 1111 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.2682926829268293,
|
| 1112 |
+
"mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
|
| 1113 |
+
"mmlu_eval_accuracy_global_facts": 0.4,
|
| 1114 |
+
"mmlu_eval_accuracy_high_school_biology": 0.375,
|
| 1115 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
|
| 1116 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
|
| 1117 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
|
| 1118 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
|
| 1119 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
|
| 1120 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.32558139534883723,
|
| 1121 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
|
| 1122 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.38461538461538464,
|
| 1123 |
+
"mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
|
| 1124 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8,
|
| 1125 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
|
| 1126 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
|
| 1127 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.5769230769230769,
|
| 1128 |
+
"mmlu_eval_accuracy_human_aging": 0.6086956521739131,
|
| 1129 |
+
"mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
|
| 1130 |
+
"mmlu_eval_accuracy_international_law": 0.9230769230769231,
|
| 1131 |
+
"mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
|
| 1132 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.5555555555555556,
|
| 1133 |
+
"mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
|
| 1134 |
+
"mmlu_eval_accuracy_management": 0.7272727272727273,
|
| 1135 |
+
"mmlu_eval_accuracy_marketing": 0.72,
|
| 1136 |
+
"mmlu_eval_accuracy_medical_genetics": 0.6363636363636364,
|
| 1137 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
|
| 1138 |
+
"mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
|
| 1139 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.24,
|
| 1140 |
+
"mmlu_eval_accuracy_nutrition": 0.6363636363636364,
|
| 1141 |
+
"mmlu_eval_accuracy_philosophy": 0.4117647058823529,
|
| 1142 |
+
"mmlu_eval_accuracy_prehistory": 0.4857142857142857,
|
| 1143 |
+
"mmlu_eval_accuracy_professional_accounting": 0.1935483870967742,
|
| 1144 |
+
"mmlu_eval_accuracy_professional_law": 0.3058823529411765,
|
| 1145 |
+
"mmlu_eval_accuracy_professional_medicine": 0.41935483870967744,
|
| 1146 |
+
"mmlu_eval_accuracy_professional_psychology": 0.4057971014492754,
|
| 1147 |
+
"mmlu_eval_accuracy_public_relations": 0.4166666666666667,
|
| 1148 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 1149 |
+
"mmlu_eval_accuracy_sociology": 0.7272727272727273,
|
| 1150 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
|
| 1151 |
+
"mmlu_eval_accuracy_virology": 0.5,
|
| 1152 |
+
"mmlu_eval_accuracy_world_religions": 0.7368421052631579,
|
| 1153 |
+
"mmlu_loss": 1.1609547404528597,
|
| 1154 |
+
"step": 1200
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 2.78,
|
| 1158 |
+
"learning_rate": 0.0002,
|
| 1159 |
+
"loss": 0.5819,
|
| 1160 |
+
"step": 1210
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 2.8,
|
| 1164 |
+
"learning_rate": 0.0002,
|
| 1165 |
+
"loss": 0.5756,
|
| 1166 |
+
"step": 1220
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 2.83,
|
| 1170 |
+
"learning_rate": 0.0002,
|
| 1171 |
+
"loss": 0.5345,
|
| 1172 |
+
"step": 1230
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 2.85,
|
| 1176 |
+
"learning_rate": 0.0002,
|
| 1177 |
+
"loss": 0.5642,
|
| 1178 |
+
"step": 1240
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 2.87,
|
| 1182 |
+
"learning_rate": 0.0002,
|
| 1183 |
+
"loss": 0.5226,
|
| 1184 |
+
"step": 1250
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 2.9,
|
| 1188 |
+
"learning_rate": 0.0002,
|
| 1189 |
+
"loss": 0.5812,
|
| 1190 |
+
"step": 1260
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 2.92,
|
| 1194 |
+
"learning_rate": 0.0002,
|
| 1195 |
+
"loss": 0.5701,
|
| 1196 |
+
"step": 1270
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 2.94,
|
| 1200 |
+
"learning_rate": 0.0002,
|
| 1201 |
+
"loss": 0.5534,
|
| 1202 |
+
"step": 1280
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 2.97,
|
| 1206 |
+
"learning_rate": 0.0002,
|
| 1207 |
+
"loss": 0.526,
|
| 1208 |
+
"step": 1290
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 2.99,
|
| 1212 |
+
"learning_rate": 0.0002,
|
| 1213 |
+
"loss": 0.5366,
|
| 1214 |
+
"step": 1300
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 3.01,
|
| 1218 |
+
"learning_rate": 0.0002,
|
| 1219 |
+
"loss": 0.4988,
|
| 1220 |
+
"step": 1310
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 3.03,
|
| 1224 |
+
"learning_rate": 0.0002,
|
| 1225 |
+
"loss": 0.4567,
|
| 1226 |
+
"step": 1320
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 3.06,
|
| 1230 |
+
"learning_rate": 0.0002,
|
| 1231 |
+
"loss": 0.4366,
|
| 1232 |
+
"step": 1330
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 3.08,
|
| 1236 |
+
"learning_rate": 0.0002,
|
| 1237 |
+
"loss": 0.4387,
|
| 1238 |
+
"step": 1340
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 3.1,
|
| 1242 |
+
"learning_rate": 0.0002,
|
| 1243 |
+
"loss": 0.3887,
|
| 1244 |
+
"step": 1350
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 3.13,
|
| 1248 |
+
"learning_rate": 0.0002,
|
| 1249 |
+
"loss": 0.4344,
|
| 1250 |
+
"step": 1360
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 3.15,
|
| 1254 |
+
"learning_rate": 0.0002,
|
| 1255 |
+
"loss": 0.4543,
|
| 1256 |
+
"step": 1370
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 3.17,
|
| 1260 |
+
"learning_rate": 0.0002,
|
| 1261 |
+
"loss": 0.4104,
|
| 1262 |
+
"step": 1380
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 3.2,
|
| 1266 |
+
"learning_rate": 0.0002,
|
| 1267 |
+
"loss": 0.4564,
|
| 1268 |
+
"step": 1390
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 3.22,
|
| 1272 |
+
"learning_rate": 0.0002,
|
| 1273 |
+
"loss": 0.4217,
|
| 1274 |
+
"step": 1400
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 3.22,
|
| 1278 |
+
"eval_loss": 0.7734031081199646,
|
| 1279 |
+
"eval_runtime": 248.2207,
|
| 1280 |
+
"eval_samples_per_second": 4.029,
|
| 1281 |
+
"eval_steps_per_second": 2.014,
|
| 1282 |
+
"step": 1400
|
| 1283 |
+
},
|
| 1284 |
+
{
|
| 1285 |
+
"epoch": 3.22,
|
| 1286 |
+
"mmlu_eval_accuracy": 0.4660104484440945,
|
| 1287 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
|
| 1288 |
+
"mmlu_eval_accuracy_anatomy": 0.5714285714285714,
|
| 1289 |
+
"mmlu_eval_accuracy_astronomy": 0.375,
|
| 1290 |
+
"mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
|
| 1291 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.4482758620689655,
|
| 1292 |
+
"mmlu_eval_accuracy_college_biology": 0.375,
|
| 1293 |
+
"mmlu_eval_accuracy_college_chemistry": 0.125,
|
| 1294 |
+
"mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
|
| 1295 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 1296 |
+
"mmlu_eval_accuracy_college_medicine": 0.45454545454545453,
|
| 1297 |
+
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
|
| 1298 |
+
"mmlu_eval_accuracy_computer_security": 0.2727272727272727,
|
| 1299 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.3076923076923077,
|
| 1300 |
+
"mmlu_eval_accuracy_econometrics": 0.16666666666666666,
|
| 1301 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.4375,
|
| 1302 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
|
| 1303 |
+
"mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
|
| 1304 |
+
"mmlu_eval_accuracy_global_facts": 0.4,
|
| 1305 |
+
"mmlu_eval_accuracy_high_school_biology": 0.34375,
|
| 1306 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453,
|
| 1307 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 1308 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
|
| 1309 |
+
"mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
|
| 1310 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
|
| 1311 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.27906976744186046,
|
| 1312 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966,
|
| 1313 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.4230769230769231,
|
| 1314 |
+
"mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
|
| 1315 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.7333333333333333,
|
| 1316 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
|
| 1317 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
|
| 1318 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
|
| 1319 |
+
"mmlu_eval_accuracy_human_aging": 0.6956521739130435,
|
| 1320 |
+
"mmlu_eval_accuracy_human_sexuality": 0.5,
|
| 1321 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 1322 |
+
"mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
|
| 1323 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.5555555555555556,
|
| 1324 |
+
"mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
|
| 1325 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 1326 |
+
"mmlu_eval_accuracy_marketing": 0.72,
|
| 1327 |
+
"mmlu_eval_accuracy_medical_genetics": 0.6363636363636364,
|
| 1328 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6976744186046512,
|
| 1329 |
+
"mmlu_eval_accuracy_moral_disputes": 0.4473684210526316,
|
| 1330 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.24,
|
| 1331 |
+
"mmlu_eval_accuracy_nutrition": 0.6060606060606061,
|
| 1332 |
+
"mmlu_eval_accuracy_philosophy": 0.38235294117647056,
|
| 1333 |
+
"mmlu_eval_accuracy_prehistory": 0.4,
|
| 1334 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
|
| 1335 |
+
"mmlu_eval_accuracy_professional_law": 0.3411764705882353,
|
| 1336 |
+
"mmlu_eval_accuracy_professional_medicine": 0.41935483870967744,
|
| 1337 |
+
"mmlu_eval_accuracy_professional_psychology": 0.4057971014492754,
|
| 1338 |
+
"mmlu_eval_accuracy_public_relations": 0.5,
|
| 1339 |
+
"mmlu_eval_accuracy_security_studies": 0.4074074074074074,
|
| 1340 |
+
"mmlu_eval_accuracy_sociology": 0.6818181818181818,
|
| 1341 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
|
| 1342 |
+
"mmlu_eval_accuracy_virology": 0.5,
|
| 1343 |
+
"mmlu_eval_accuracy_world_religions": 0.7894736842105263,
|
| 1344 |
+
"mmlu_loss": 1.0775621755841351,
|
| 1345 |
+
"step": 1400
|
| 1346 |
}
|
| 1347 |
],
|
| 1348 |
"max_steps": 5000,
|
| 1349 |
"num_train_epochs": 12,
|
| 1350 |
+
"total_flos": 3.3098129260712755e+17,
|
| 1351 |
"trial_name": null,
|
| 1352 |
"trial_params": null
|
| 1353 |
}
|