Training in progress, step 2282, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 791781368
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bc189396d250cd02b58eccc2bb364409eaee0c8425639afa61aa38ebc2574ad2
|
| 3 |
size 791781368
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2375487866
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f3a0e5ae6c5f327214099f137a7cc8083d4122fae307d1b2016880d7e75c10cd
|
| 3 |
size 2375487866
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1000
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d305aeb5f66edef533fe495bedd6ae8dca8a6220560f62265310b5e9dbe7ba24
|
| 3 |
size 1000
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 1000,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -2823,6 +2823,398 @@
|
|
| 2823 |
"eval_samples_per_second": 1013.774,
|
| 2824 |
"eval_steps_per_second": 31.682,
|
| 2825 |
"step": 2000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2826 |
}
|
| 2827 |
],
|
| 2828 |
"logging_steps": 5,
|
|
@@ -2837,12 +3229,12 @@
|
|
| 2837 |
"should_evaluate": false,
|
| 2838 |
"should_log": false,
|
| 2839 |
"should_save": true,
|
| 2840 |
-
"should_training_stop":
|
| 2841 |
},
|
| 2842 |
"attributes": {}
|
| 2843 |
}
|
| 2844 |
},
|
| 2845 |
-
"total_flos": 1.
|
| 2846 |
"train_batch_size": 8,
|
| 2847 |
"trial_name": null,
|
| 2848 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9997672540080229,
|
| 5 |
"eval_steps": 1000,
|
| 6 |
+
"global_step": 2282,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 2823 |
"eval_samples_per_second": 1013.774,
|
| 2824 |
"eval_steps_per_second": 31.682,
|
| 2825 |
"step": 2000
|
| 2826 |
+
},
|
| 2827 |
+
{
|
| 2828 |
+
"epoch": 0.8784107556030175,
|
| 2829 |
+
"grad_norm": 20.109375,
|
| 2830 |
+
"learning_rate": 2.698490014612762e-06,
|
| 2831 |
+
"loss": 25.8034,
|
| 2832 |
+
"step": 2005
|
| 2833 |
+
},
|
| 2834 |
+
{
|
| 2835 |
+
"epoch": 0.8806013061157432,
|
| 2836 |
+
"grad_norm": 10.953125,
|
| 2837 |
+
"learning_rate": 2.6497808085728203e-06,
|
| 2838 |
+
"loss": 25.0685,
|
| 2839 |
+
"step": 2010
|
| 2840 |
+
},
|
| 2841 |
+
{
|
| 2842 |
+
"epoch": 0.8827918566284689,
|
| 2843 |
+
"grad_norm": 31.0625,
|
| 2844 |
+
"learning_rate": 2.601071602532879e-06,
|
| 2845 |
+
"loss": 25.7096,
|
| 2846 |
+
"step": 2015
|
| 2847 |
+
},
|
| 2848 |
+
{
|
| 2849 |
+
"epoch": 0.8849824071411947,
|
| 2850 |
+
"grad_norm": 24.046875,
|
| 2851 |
+
"learning_rate": 2.5523623964929375e-06,
|
| 2852 |
+
"loss": 25.3376,
|
| 2853 |
+
"step": 2020
|
| 2854 |
+
},
|
| 2855 |
+
{
|
| 2856 |
+
"epoch": 0.8871729576539203,
|
| 2857 |
+
"grad_norm": 27.65625,
|
| 2858 |
+
"learning_rate": 2.503653190452996e-06,
|
| 2859 |
+
"loss": 24.6723,
|
| 2860 |
+
"step": 2025
|
| 2861 |
+
},
|
| 2862 |
+
{
|
| 2863 |
+
"epoch": 0.8893635081666461,
|
| 2864 |
+
"grad_norm": 12.5078125,
|
| 2865 |
+
"learning_rate": 2.4549439844130542e-06,
|
| 2866 |
+
"loss": 25.1757,
|
| 2867 |
+
"step": 2030
|
| 2868 |
+
},
|
| 2869 |
+
{
|
| 2870 |
+
"epoch": 0.8915540586793719,
|
| 2871 |
+
"grad_norm": 15.1015625,
|
| 2872 |
+
"learning_rate": 2.4062347783731126e-06,
|
| 2873 |
+
"loss": 25.6934,
|
| 2874 |
+
"step": 2035
|
| 2875 |
+
},
|
| 2876 |
+
{
|
| 2877 |
+
"epoch": 0.8937446091920976,
|
| 2878 |
+
"grad_norm": 28.75,
|
| 2879 |
+
"learning_rate": 2.357525572333171e-06,
|
| 2880 |
+
"loss": 25.4336,
|
| 2881 |
+
"step": 2040
|
| 2882 |
+
},
|
| 2883 |
+
{
|
| 2884 |
+
"epoch": 0.8959351597048233,
|
| 2885 |
+
"grad_norm": 18.53125,
|
| 2886 |
+
"learning_rate": 2.3088163662932294e-06,
|
| 2887 |
+
"loss": 24.4317,
|
| 2888 |
+
"step": 2045
|
| 2889 |
+
},
|
| 2890 |
+
{
|
| 2891 |
+
"epoch": 0.8981257102175491,
|
| 2892 |
+
"grad_norm": 15.265625,
|
| 2893 |
+
"learning_rate": 2.260107160253288e-06,
|
| 2894 |
+
"loss": 24.6006,
|
| 2895 |
+
"step": 2050
|
| 2896 |
+
},
|
| 2897 |
+
{
|
| 2898 |
+
"epoch": 0.9003162607302748,
|
| 2899 |
+
"grad_norm": 12.1796875,
|
| 2900 |
+
"learning_rate": 2.2113979542133465e-06,
|
| 2901 |
+
"loss": 24.7936,
|
| 2902 |
+
"step": 2055
|
| 2903 |
+
},
|
| 2904 |
+
{
|
| 2905 |
+
"epoch": 0.9025068112430005,
|
| 2906 |
+
"grad_norm": 12.21875,
|
| 2907 |
+
"learning_rate": 2.162688748173405e-06,
|
| 2908 |
+
"loss": 26.129,
|
| 2909 |
+
"step": 2060
|
| 2910 |
+
},
|
| 2911 |
+
{
|
| 2912 |
+
"epoch": 0.9046973617557262,
|
| 2913 |
+
"grad_norm": 12.984375,
|
| 2914 |
+
"learning_rate": 2.1139795421334633e-06,
|
| 2915 |
+
"loss": 24.6225,
|
| 2916 |
+
"step": 2065
|
| 2917 |
+
},
|
| 2918 |
+
{
|
| 2919 |
+
"epoch": 0.906887912268452,
|
| 2920 |
+
"grad_norm": 15.65625,
|
| 2921 |
+
"learning_rate": 2.065270336093522e-06,
|
| 2922 |
+
"loss": 24.7391,
|
| 2923 |
+
"step": 2070
|
| 2924 |
+
},
|
| 2925 |
+
{
|
| 2926 |
+
"epoch": 0.9090784627811777,
|
| 2927 |
+
"grad_norm": 13.3046875,
|
| 2928 |
+
"learning_rate": 2.0165611300535805e-06,
|
| 2929 |
+
"loss": 24.7855,
|
| 2930 |
+
"step": 2075
|
| 2931 |
+
},
|
| 2932 |
+
{
|
| 2933 |
+
"epoch": 0.9112690132939034,
|
| 2934 |
+
"grad_norm": 15.015625,
|
| 2935 |
+
"learning_rate": 1.967851924013639e-06,
|
| 2936 |
+
"loss": 24.7362,
|
| 2937 |
+
"step": 2080
|
| 2938 |
+
},
|
| 2939 |
+
{
|
| 2940 |
+
"epoch": 0.9134595638066292,
|
| 2941 |
+
"grad_norm": 18.53125,
|
| 2942 |
+
"learning_rate": 1.9191427179736972e-06,
|
| 2943 |
+
"loss": 24.8633,
|
| 2944 |
+
"step": 2085
|
| 2945 |
+
},
|
| 2946 |
+
{
|
| 2947 |
+
"epoch": 0.9156501143193548,
|
| 2948 |
+
"grad_norm": 19.0625,
|
| 2949 |
+
"learning_rate": 1.8704335119337556e-06,
|
| 2950 |
+
"loss": 24.7622,
|
| 2951 |
+
"step": 2090
|
| 2952 |
+
},
|
| 2953 |
+
{
|
| 2954 |
+
"epoch": 0.9178406648320806,
|
| 2955 |
+
"grad_norm": 22.96875,
|
| 2956 |
+
"learning_rate": 1.8217243058938142e-06,
|
| 2957 |
+
"loss": 24.8469,
|
| 2958 |
+
"step": 2095
|
| 2959 |
+
},
|
| 2960 |
+
{
|
| 2961 |
+
"epoch": 0.9200312153448064,
|
| 2962 |
+
"grad_norm": 16.546875,
|
| 2963 |
+
"learning_rate": 1.7730150998538726e-06,
|
| 2964 |
+
"loss": 24.4148,
|
| 2965 |
+
"step": 2100
|
| 2966 |
+
},
|
| 2967 |
+
{
|
| 2968 |
+
"epoch": 0.922221765857532,
|
| 2969 |
+
"grad_norm": 17.15625,
|
| 2970 |
+
"learning_rate": 1.724305893813931e-06,
|
| 2971 |
+
"loss": 24.5911,
|
| 2972 |
+
"step": 2105
|
| 2973 |
+
},
|
| 2974 |
+
{
|
| 2975 |
+
"epoch": 0.9244123163702578,
|
| 2976 |
+
"grad_norm": 32.375,
|
| 2977 |
+
"learning_rate": 1.6755966877739893e-06,
|
| 2978 |
+
"loss": 24.3472,
|
| 2979 |
+
"step": 2110
|
| 2980 |
+
},
|
| 2981 |
+
{
|
| 2982 |
+
"epoch": 0.9266028668829835,
|
| 2983 |
+
"grad_norm": 11.1796875,
|
| 2984 |
+
"learning_rate": 1.626887481734048e-06,
|
| 2985 |
+
"loss": 24.272,
|
| 2986 |
+
"step": 2115
|
| 2987 |
+
},
|
| 2988 |
+
{
|
| 2989 |
+
"epoch": 0.9287934173957093,
|
| 2990 |
+
"grad_norm": 15.828125,
|
| 2991 |
+
"learning_rate": 1.5781782756941063e-06,
|
| 2992 |
+
"loss": 24.1789,
|
| 2993 |
+
"step": 2120
|
| 2994 |
+
},
|
| 2995 |
+
{
|
| 2996 |
+
"epoch": 0.930983967908435,
|
| 2997 |
+
"grad_norm": 13.609375,
|
| 2998 |
+
"learning_rate": 1.5294690696541647e-06,
|
| 2999 |
+
"loss": 24.5268,
|
| 3000 |
+
"step": 2125
|
| 3001 |
+
},
|
| 3002 |
+
{
|
| 3003 |
+
"epoch": 0.9331745184211607,
|
| 3004 |
+
"grad_norm": 11.7890625,
|
| 3005 |
+
"learning_rate": 1.480759863614223e-06,
|
| 3006 |
+
"loss": 24.4547,
|
| 3007 |
+
"step": 2130
|
| 3008 |
+
},
|
| 3009 |
+
{
|
| 3010 |
+
"epoch": 0.9353650689338865,
|
| 3011 |
+
"grad_norm": 14.46875,
|
| 3012 |
+
"learning_rate": 1.4320506575742814e-06,
|
| 3013 |
+
"loss": 24.5342,
|
| 3014 |
+
"step": 2135
|
| 3015 |
+
},
|
| 3016 |
+
{
|
| 3017 |
+
"epoch": 0.9375556194466121,
|
| 3018 |
+
"grad_norm": 12.9765625,
|
| 3019 |
+
"learning_rate": 1.3833414515343402e-06,
|
| 3020 |
+
"loss": 24.1581,
|
| 3021 |
+
"step": 2140
|
| 3022 |
+
},
|
| 3023 |
+
{
|
| 3024 |
+
"epoch": 0.9397461699593379,
|
| 3025 |
+
"grad_norm": 13.4140625,
|
| 3026 |
+
"learning_rate": 1.3346322454943986e-06,
|
| 3027 |
+
"loss": 24.0952,
|
| 3028 |
+
"step": 2145
|
| 3029 |
+
},
|
| 3030 |
+
{
|
| 3031 |
+
"epoch": 0.9419367204720637,
|
| 3032 |
+
"grad_norm": 11.921875,
|
| 3033 |
+
"learning_rate": 1.285923039454457e-06,
|
| 3034 |
+
"loss": 24.6826,
|
| 3035 |
+
"step": 2150
|
| 3036 |
+
},
|
| 3037 |
+
{
|
| 3038 |
+
"epoch": 0.9441272709847893,
|
| 3039 |
+
"grad_norm": 13.7734375,
|
| 3040 |
+
"learning_rate": 1.2372138334145156e-06,
|
| 3041 |
+
"loss": 23.6338,
|
| 3042 |
+
"step": 2155
|
| 3043 |
+
},
|
| 3044 |
+
{
|
| 3045 |
+
"epoch": 0.9463178214975151,
|
| 3046 |
+
"grad_norm": 12.1015625,
|
| 3047 |
+
"learning_rate": 1.188504627374574e-06,
|
| 3048 |
+
"loss": 24.3637,
|
| 3049 |
+
"step": 2160
|
| 3050 |
+
},
|
| 3051 |
+
{
|
| 3052 |
+
"epoch": 0.9485083720102409,
|
| 3053 |
+
"grad_norm": 14.578125,
|
| 3054 |
+
"learning_rate": 1.1397954213346323e-06,
|
| 3055 |
+
"loss": 24.6624,
|
| 3056 |
+
"step": 2165
|
| 3057 |
+
},
|
| 3058 |
+
{
|
| 3059 |
+
"epoch": 0.9506989225229665,
|
| 3060 |
+
"grad_norm": 12.7578125,
|
| 3061 |
+
"learning_rate": 1.0910862152946907e-06,
|
| 3062 |
+
"loss": 23.9978,
|
| 3063 |
+
"step": 2170
|
| 3064 |
+
},
|
| 3065 |
+
{
|
| 3066 |
+
"epoch": 0.9528894730356923,
|
| 3067 |
+
"grad_norm": 13.2265625,
|
| 3068 |
+
"learning_rate": 1.0423770092547493e-06,
|
| 3069 |
+
"loss": 24.4468,
|
| 3070 |
+
"step": 2175
|
| 3071 |
+
},
|
| 3072 |
+
{
|
| 3073 |
+
"epoch": 0.955080023548418,
|
| 3074 |
+
"grad_norm": 11.7890625,
|
| 3075 |
+
"learning_rate": 9.936678032148077e-07,
|
| 3076 |
+
"loss": 24.303,
|
| 3077 |
+
"step": 2180
|
| 3078 |
+
},
|
| 3079 |
+
{
|
| 3080 |
+
"epoch": 0.9572705740611437,
|
| 3081 |
+
"grad_norm": 16.25,
|
| 3082 |
+
"learning_rate": 9.44958597174866e-07,
|
| 3083 |
+
"loss": 24.5248,
|
| 3084 |
+
"step": 2185
|
| 3085 |
+
},
|
| 3086 |
+
{
|
| 3087 |
+
"epoch": 0.9594611245738695,
|
| 3088 |
+
"grad_norm": 18.859375,
|
| 3089 |
+
"learning_rate": 8.962493911349246e-07,
|
| 3090 |
+
"loss": 24.4316,
|
| 3091 |
+
"step": 2190
|
| 3092 |
+
},
|
| 3093 |
+
{
|
| 3094 |
+
"epoch": 0.9616516750865952,
|
| 3095 |
+
"grad_norm": 10.6796875,
|
| 3096 |
+
"learning_rate": 8.47540185094983e-07,
|
| 3097 |
+
"loss": 23.8851,
|
| 3098 |
+
"step": 2195
|
| 3099 |
+
},
|
| 3100 |
+
{
|
| 3101 |
+
"epoch": 0.963842225599321,
|
| 3102 |
+
"grad_norm": 12.140625,
|
| 3103 |
+
"learning_rate": 7.988309790550415e-07,
|
| 3104 |
+
"loss": 24.2927,
|
| 3105 |
+
"step": 2200
|
| 3106 |
+
},
|
| 3107 |
+
{
|
| 3108 |
+
"epoch": 0.9660327761120466,
|
| 3109 |
+
"grad_norm": 18.078125,
|
| 3110 |
+
"learning_rate": 7.501217730150999e-07,
|
| 3111 |
+
"loss": 24.1125,
|
| 3112 |
+
"step": 2205
|
| 3113 |
+
},
|
| 3114 |
+
{
|
| 3115 |
+
"epoch": 0.9682233266247724,
|
| 3116 |
+
"grad_norm": 9.7109375,
|
| 3117 |
+
"learning_rate": 7.014125669751585e-07,
|
| 3118 |
+
"loss": 24.1763,
|
| 3119 |
+
"step": 2210
|
| 3120 |
+
},
|
| 3121 |
+
{
|
| 3122 |
+
"epoch": 0.9704138771374982,
|
| 3123 |
+
"grad_norm": 12.6953125,
|
| 3124 |
+
"learning_rate": 6.527033609352168e-07,
|
| 3125 |
+
"loss": 23.1948,
|
| 3126 |
+
"step": 2215
|
| 3127 |
+
},
|
| 3128 |
+
{
|
| 3129 |
+
"epoch": 0.9726044276502238,
|
| 3130 |
+
"grad_norm": 18.671875,
|
| 3131 |
+
"learning_rate": 6.039941548952752e-07,
|
| 3132 |
+
"loss": 24.2813,
|
| 3133 |
+
"step": 2220
|
| 3134 |
+
},
|
| 3135 |
+
{
|
| 3136 |
+
"epoch": 0.9747949781629496,
|
| 3137 |
+
"grad_norm": 9.96875,
|
| 3138 |
+
"learning_rate": 5.552849488553337e-07,
|
| 3139 |
+
"loss": 23.7533,
|
| 3140 |
+
"step": 2225
|
| 3141 |
+
},
|
| 3142 |
+
{
|
| 3143 |
+
"epoch": 0.9769855286756753,
|
| 3144 |
+
"grad_norm": 13.71875,
|
| 3145 |
+
"learning_rate": 5.065757428153922e-07,
|
| 3146 |
+
"loss": 24.5382,
|
| 3147 |
+
"step": 2230
|
| 3148 |
+
},
|
| 3149 |
+
{
|
| 3150 |
+
"epoch": 0.979176079188401,
|
| 3151 |
+
"grad_norm": 13.4296875,
|
| 3152 |
+
"learning_rate": 4.578665367754506e-07,
|
| 3153 |
+
"loss": 23.7636,
|
| 3154 |
+
"step": 2235
|
| 3155 |
+
},
|
| 3156 |
+
{
|
| 3157 |
+
"epoch": 0.9813666297011268,
|
| 3158 |
+
"grad_norm": 13.640625,
|
| 3159 |
+
"learning_rate": 4.091573307355091e-07,
|
| 3160 |
+
"loss": 23.785,
|
| 3161 |
+
"step": 2240
|
| 3162 |
+
},
|
| 3163 |
+
{
|
| 3164 |
+
"epoch": 0.9835571802138525,
|
| 3165 |
+
"grad_norm": 27.078125,
|
| 3166 |
+
"learning_rate": 3.6044812469556747e-07,
|
| 3167 |
+
"loss": 23.8269,
|
| 3168 |
+
"step": 2245
|
| 3169 |
+
},
|
| 3170 |
+
{
|
| 3171 |
+
"epoch": 0.9857477307265782,
|
| 3172 |
+
"grad_norm": 13.4609375,
|
| 3173 |
+
"learning_rate": 3.1173891865562595e-07,
|
| 3174 |
+
"loss": 24.0428,
|
| 3175 |
+
"step": 2250
|
| 3176 |
+
},
|
| 3177 |
+
{
|
| 3178 |
+
"epoch": 0.9879382812393039,
|
| 3179 |
+
"grad_norm": 12.109375,
|
| 3180 |
+
"learning_rate": 2.630297126156844e-07,
|
| 3181 |
+
"loss": 24.3293,
|
| 3182 |
+
"step": 2255
|
| 3183 |
+
},
|
| 3184 |
+
{
|
| 3185 |
+
"epoch": 0.9901288317520297,
|
| 3186 |
+
"grad_norm": 20.84375,
|
| 3187 |
+
"learning_rate": 2.1432050657574284e-07,
|
| 3188 |
+
"loss": 24.0019,
|
| 3189 |
+
"step": 2260
|
| 3190 |
+
},
|
| 3191 |
+
{
|
| 3192 |
+
"epoch": 0.9923193822647554,
|
| 3193 |
+
"grad_norm": 16.265625,
|
| 3194 |
+
"learning_rate": 1.6561130053580127e-07,
|
| 3195 |
+
"loss": 24.3604,
|
| 3196 |
+
"step": 2265
|
| 3197 |
+
},
|
| 3198 |
+
{
|
| 3199 |
+
"epoch": 0.9945099327774811,
|
| 3200 |
+
"grad_norm": 19.96875,
|
| 3201 |
+
"learning_rate": 1.1690209449585972e-07,
|
| 3202 |
+
"loss": 24.1925,
|
| 3203 |
+
"step": 2270
|
| 3204 |
+
},
|
| 3205 |
+
{
|
| 3206 |
+
"epoch": 0.9967004832902069,
|
| 3207 |
+
"grad_norm": 16.953125,
|
| 3208 |
+
"learning_rate": 6.819288845591817e-08,
|
| 3209 |
+
"loss": 23.6996,
|
| 3210 |
+
"step": 2275
|
| 3211 |
+
},
|
| 3212 |
+
{
|
| 3213 |
+
"epoch": 0.9988910338029326,
|
| 3214 |
+
"grad_norm": 19.6875,
|
| 3215 |
+
"learning_rate": 1.9483682415976622e-08,
|
| 3216 |
+
"loss": 23.9507,
|
| 3217 |
+
"step": 2280
|
| 3218 |
}
|
| 3219 |
],
|
| 3220 |
"logging_steps": 5,
|
|
|
|
| 3229 |
"should_evaluate": false,
|
| 3230 |
"should_log": false,
|
| 3231 |
"should_save": true,
|
| 3232 |
+
"should_training_stop": true
|
| 3233 |
},
|
| 3234 |
"attributes": {}
|
| 3235 |
}
|
| 3236 |
},
|
| 3237 |
+
"total_flos": 1.9772879918220706e+19,
|
| 3238 |
"train_batch_size": 8,
|
| 3239 |
"trial_name": null,
|
| 3240 |
"trial_params": null
|