Training in progress, step 481, 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 373077376
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d861b0c356377800f143df821f91f101184e3745994cd74adf5133106d79bde2
|
| 3 |
size 373077376
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 373225675
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:799ccb4c58e58c79e5f8070f3d59b69f01779859a514d0911cf768c6e286cb76
|
| 3 |
size 373225675
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1401
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a8b4903fb7be2c884aa62b335e9291d92240420b2bf52100c824048b850730b6
|
| 3 |
size 1401
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0
|
| 6 |
"eval_steps": 100,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -2840,6 +2840,573 @@
|
|
| 2840 |
"eval_samples_per_second": 22.501,
|
| 2841 |
"eval_steps_per_second": 2.813,
|
| 2842 |
"step": 400
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2843 |
}
|
| 2844 |
],
|
| 2845 |
"logging_steps": 1,
|
|
@@ -2854,12 +3421,12 @@
|
|
| 2854 |
"should_evaluate": false,
|
| 2855 |
"should_log": false,
|
| 2856 |
"should_save": true,
|
| 2857 |
-
"should_training_stop":
|
| 2858 |
},
|
| 2859 |
"attributes": {}
|
| 2860 |
}
|
| 2861 |
},
|
| 2862 |
-
"total_flos": 1.
|
| 2863 |
"train_batch_size": 32,
|
| 2864 |
"trial_name": null,
|
| 2865 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.0,
|
| 6 |
"eval_steps": 100,
|
| 7 |
+
"global_step": 481,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 2840 |
"eval_samples_per_second": 22.501,
|
| 2841 |
"eval_steps_per_second": 2.813,
|
| 2842 |
"step": 400
|
| 2843 |
+
},
|
| 2844 |
+
{
|
| 2845 |
+
"epoch": 0.8336798336798337,
|
| 2846 |
+
"grad_norm": 0.27734375,
|
| 2847 |
+
"learning_rate": 7.585430144121319e-05,
|
| 2848 |
+
"loss": 5.5678,
|
| 2849 |
+
"step": 401
|
| 2850 |
+
},
|
| 2851 |
+
{
|
| 2852 |
+
"epoch": 0.8357588357588358,
|
| 2853 |
+
"grad_norm": 0.275390625,
|
| 2854 |
+
"learning_rate": 7.404029558083653e-05,
|
| 2855 |
+
"loss": 5.4751,
|
| 2856 |
+
"step": 402
|
| 2857 |
+
},
|
| 2858 |
+
{
|
| 2859 |
+
"epoch": 0.8378378378378378,
|
| 2860 |
+
"grad_norm": 0.275390625,
|
| 2861 |
+
"learning_rate": 7.224650765840613e-05,
|
| 2862 |
+
"loss": 5.4895,
|
| 2863 |
+
"step": 403
|
| 2864 |
+
},
|
| 2865 |
+
{
|
| 2866 |
+
"epoch": 0.83991683991684,
|
| 2867 |
+
"grad_norm": 0.2890625,
|
| 2868 |
+
"learning_rate": 7.047302281505735e-05,
|
| 2869 |
+
"loss": 5.381,
|
| 2870 |
+
"step": 404
|
| 2871 |
+
},
|
| 2872 |
+
{
|
| 2873 |
+
"epoch": 0.841995841995842,
|
| 2874 |
+
"grad_norm": 0.64453125,
|
| 2875 |
+
"learning_rate": 6.871992522825182e-05,
|
| 2876 |
+
"loss": 5.3161,
|
| 2877 |
+
"step": 405
|
| 2878 |
+
},
|
| 2879 |
+
{
|
| 2880 |
+
"epoch": 0.8440748440748441,
|
| 2881 |
+
"grad_norm": 1.078125,
|
| 2882 |
+
"learning_rate": 6.698729810778065e-05,
|
| 2883 |
+
"loss": 5.3665,
|
| 2884 |
+
"step": 406
|
| 2885 |
+
},
|
| 2886 |
+
{
|
| 2887 |
+
"epoch": 0.8461538461538461,
|
| 2888 |
+
"grad_norm": 1.0859375,
|
| 2889 |
+
"learning_rate": 6.527522369181655e-05,
|
| 2890 |
+
"loss": 5.545,
|
| 2891 |
+
"step": 407
|
| 2892 |
+
},
|
| 2893 |
+
{
|
| 2894 |
+
"epoch": 0.8482328482328483,
|
| 2895 |
+
"grad_norm": 0.259765625,
|
| 2896 |
+
"learning_rate": 6.358378324300929e-05,
|
| 2897 |
+
"loss": 5.1788,
|
| 2898 |
+
"step": 408
|
| 2899 |
+
},
|
| 2900 |
+
{
|
| 2901 |
+
"epoch": 0.8503118503118503,
|
| 2902 |
+
"grad_norm": 0.26953125,
|
| 2903 |
+
"learning_rate": 6.191305704462896e-05,
|
| 2904 |
+
"loss": 5.2483,
|
| 2905 |
+
"step": 409
|
| 2906 |
+
},
|
| 2907 |
+
{
|
| 2908 |
+
"epoch": 0.8523908523908524,
|
| 2909 |
+
"grad_norm": 0.28125,
|
| 2910 |
+
"learning_rate": 6.026312439675552e-05,
|
| 2911 |
+
"loss": 4.8411,
|
| 2912 |
+
"step": 410
|
| 2913 |
+
},
|
| 2914 |
+
{
|
| 2915 |
+
"epoch": 0.8544698544698545,
|
| 2916 |
+
"grad_norm": 0.3359375,
|
| 2917 |
+
"learning_rate": 5.863406361251472e-05,
|
| 2918 |
+
"loss": 5.1144,
|
| 2919 |
+
"step": 411
|
| 2920 |
+
},
|
| 2921 |
+
{
|
| 2922 |
+
"epoch": 0.8565488565488566,
|
| 2923 |
+
"grad_norm": 1.0546875,
|
| 2924 |
+
"learning_rate": 5.7025952014361004e-05,
|
| 2925 |
+
"loss": 5.1504,
|
| 2926 |
+
"step": 412
|
| 2927 |
+
},
|
| 2928 |
+
{
|
| 2929 |
+
"epoch": 0.8586278586278586,
|
| 2930 |
+
"grad_norm": 0.337890625,
|
| 2931 |
+
"learning_rate": 5.543886593040737e-05,
|
| 2932 |
+
"loss": 5.2093,
|
| 2933 |
+
"step": 413
|
| 2934 |
+
},
|
| 2935 |
+
{
|
| 2936 |
+
"epoch": 0.8607068607068608,
|
| 2937 |
+
"grad_norm": 0.287109375,
|
| 2938 |
+
"learning_rate": 5.387288069080298e-05,
|
| 2939 |
+
"loss": 5.0402,
|
| 2940 |
+
"step": 414
|
| 2941 |
+
},
|
| 2942 |
+
{
|
| 2943 |
+
"epoch": 0.8627858627858628,
|
| 2944 |
+
"grad_norm": 0.32421875,
|
| 2945 |
+
"learning_rate": 5.23280706241569e-05,
|
| 2946 |
+
"loss": 4.7027,
|
| 2947 |
+
"step": 415
|
| 2948 |
+
},
|
| 2949 |
+
{
|
| 2950 |
+
"epoch": 0.8648648648648649,
|
| 2951 |
+
"grad_norm": 0.25,
|
| 2952 |
+
"learning_rate": 5.080450905401057e-05,
|
| 2953 |
+
"loss": 5.2766,
|
| 2954 |
+
"step": 416
|
| 2955 |
+
},
|
| 2956 |
+
{
|
| 2957 |
+
"epoch": 0.8669438669438669,
|
| 2958 |
+
"grad_norm": 0.28515625,
|
| 2959 |
+
"learning_rate": 4.930226829535767e-05,
|
| 2960 |
+
"loss": 5.3266,
|
| 2961 |
+
"step": 417
|
| 2962 |
+
},
|
| 2963 |
+
{
|
| 2964 |
+
"epoch": 0.8690228690228691,
|
| 2965 |
+
"grad_norm": 0.279296875,
|
| 2966 |
+
"learning_rate": 4.7821419651211284e-05,
|
| 2967 |
+
"loss": 5.1296,
|
| 2968 |
+
"step": 418
|
| 2969 |
+
},
|
| 2970 |
+
{
|
| 2971 |
+
"epoch": 0.8711018711018711,
|
| 2972 |
+
"grad_norm": 0.27734375,
|
| 2973 |
+
"learning_rate": 4.636203340922007e-05,
|
| 2974 |
+
"loss": 5.5194,
|
| 2975 |
+
"step": 419
|
| 2976 |
+
},
|
| 2977 |
+
{
|
| 2978 |
+
"epoch": 0.8731808731808732,
|
| 2979 |
+
"grad_norm": 0.2578125,
|
| 2980 |
+
"learning_rate": 4.492417883833155e-05,
|
| 2981 |
+
"loss": 4.968,
|
| 2982 |
+
"step": 420
|
| 2983 |
+
},
|
| 2984 |
+
{
|
| 2985 |
+
"epoch": 0.8752598752598753,
|
| 2986 |
+
"grad_norm": 0.2578125,
|
| 2987 |
+
"learning_rate": 4.350792418550509e-05,
|
| 2988 |
+
"loss": 5.6204,
|
| 2989 |
+
"step": 421
|
| 2990 |
+
},
|
| 2991 |
+
{
|
| 2992 |
+
"epoch": 0.8773388773388774,
|
| 2993 |
+
"grad_norm": 0.2451171875,
|
| 2994 |
+
"learning_rate": 4.211333667247125e-05,
|
| 2995 |
+
"loss": 5.2991,
|
| 2996 |
+
"step": 422
|
| 2997 |
+
},
|
| 2998 |
+
{
|
| 2999 |
+
"epoch": 0.8794178794178794,
|
| 3000 |
+
"grad_norm": 0.267578125,
|
| 3001 |
+
"learning_rate": 4.074048249254286e-05,
|
| 3002 |
+
"loss": 5.3253,
|
| 3003 |
+
"step": 423
|
| 3004 |
+
},
|
| 3005 |
+
{
|
| 3006 |
+
"epoch": 0.8814968814968815,
|
| 3007 |
+
"grad_norm": 1.0546875,
|
| 3008 |
+
"learning_rate": 3.938942680747176e-05,
|
| 3009 |
+
"loss": 4.9877,
|
| 3010 |
+
"step": 424
|
| 3011 |
+
},
|
| 3012 |
+
{
|
| 3013 |
+
"epoch": 0.8835758835758836,
|
| 3014 |
+
"grad_norm": 0.2373046875,
|
| 3015 |
+
"learning_rate": 3.806023374435663e-05,
|
| 3016 |
+
"loss": 5.5203,
|
| 3017 |
+
"step": 425
|
| 3018 |
+
},
|
| 3019 |
+
{
|
| 3020 |
+
"epoch": 0.8856548856548857,
|
| 3021 |
+
"grad_norm": 0.2890625,
|
| 3022 |
+
"learning_rate": 3.675296639259912e-05,
|
| 3023 |
+
"loss": 4.9162,
|
| 3024 |
+
"step": 426
|
| 3025 |
+
},
|
| 3026 |
+
{
|
| 3027 |
+
"epoch": 0.8877338877338877,
|
| 3028 |
+
"grad_norm": 0.267578125,
|
| 3029 |
+
"learning_rate": 3.546768680090934e-05,
|
| 3030 |
+
"loss": 5.3487,
|
| 3031 |
+
"step": 427
|
| 3032 |
+
},
|
| 3033 |
+
{
|
| 3034 |
+
"epoch": 0.8898128898128899,
|
| 3035 |
+
"grad_norm": 0.26171875,
|
| 3036 |
+
"learning_rate": 3.420445597436056e-05,
|
| 3037 |
+
"loss": 4.8628,
|
| 3038 |
+
"step": 428
|
| 3039 |
+
},
|
| 3040 |
+
{
|
| 3041 |
+
"epoch": 0.8918918918918919,
|
| 3042 |
+
"grad_norm": 0.271484375,
|
| 3043 |
+
"learning_rate": 3.296333387149392e-05,
|
| 3044 |
+
"loss": 5.2627,
|
| 3045 |
+
"step": 429
|
| 3046 |
+
},
|
| 3047 |
+
{
|
| 3048 |
+
"epoch": 0.893970893970894,
|
| 3049 |
+
"grad_norm": 0.27734375,
|
| 3050 |
+
"learning_rate": 3.174437940147268e-05,
|
| 3051 |
+
"loss": 5.3865,
|
| 3052 |
+
"step": 430
|
| 3053 |
+
},
|
| 3054 |
+
{
|
| 3055 |
+
"epoch": 0.896049896049896,
|
| 3056 |
+
"grad_norm": 0.296875,
|
| 3057 |
+
"learning_rate": 3.054765042128521e-05,
|
| 3058 |
+
"loss": 5.1347,
|
| 3059 |
+
"step": 431
|
| 3060 |
+
},
|
| 3061 |
+
{
|
| 3062 |
+
"epoch": 0.8981288981288982,
|
| 3063 |
+
"grad_norm": 0.296875,
|
| 3064 |
+
"learning_rate": 2.9373203733000232e-05,
|
| 3065 |
+
"loss": 5.308,
|
| 3066 |
+
"step": 432
|
| 3067 |
+
},
|
| 3068 |
+
{
|
| 3069 |
+
"epoch": 0.9002079002079002,
|
| 3070 |
+
"grad_norm": 0.26171875,
|
| 3071 |
+
"learning_rate": 2.8221095081069513e-05,
|
| 3072 |
+
"loss": 5.0021,
|
| 3073 |
+
"step": 433
|
| 3074 |
+
},
|
| 3075 |
+
{
|
| 3076 |
+
"epoch": 0.9022869022869023,
|
| 3077 |
+
"grad_norm": 1.046875,
|
| 3078 |
+
"learning_rate": 2.709137914968268e-05,
|
| 3079 |
+
"loss": 5.0882,
|
| 3080 |
+
"step": 434
|
| 3081 |
+
},
|
| 3082 |
+
{
|
| 3083 |
+
"epoch": 0.9043659043659044,
|
| 3084 |
+
"grad_norm": 1.0546875,
|
| 3085 |
+
"learning_rate": 2.5984109560171388e-05,
|
| 3086 |
+
"loss": 5.4556,
|
| 3087 |
+
"step": 435
|
| 3088 |
+
},
|
| 3089 |
+
{
|
| 3090 |
+
"epoch": 0.9064449064449065,
|
| 3091 |
+
"grad_norm": 0.248046875,
|
| 3092 |
+
"learning_rate": 2.4899338868464407e-05,
|
| 3093 |
+
"loss": 5.4887,
|
| 3094 |
+
"step": 436
|
| 3095 |
+
},
|
| 3096 |
+
{
|
| 3097 |
+
"epoch": 0.9085239085239085,
|
| 3098 |
+
"grad_norm": 0.2734375,
|
| 3099 |
+
"learning_rate": 2.3837118562592797e-05,
|
| 3100 |
+
"loss": 5.3905,
|
| 3101 |
+
"step": 437
|
| 3102 |
+
},
|
| 3103 |
+
{
|
| 3104 |
+
"epoch": 0.9106029106029107,
|
| 3105 |
+
"grad_norm": 0.283203125,
|
| 3106 |
+
"learning_rate": 2.2797499060246252e-05,
|
| 3107 |
+
"loss": 5.1305,
|
| 3108 |
+
"step": 438
|
| 3109 |
+
},
|
| 3110 |
+
{
|
| 3111 |
+
"epoch": 0.9126819126819127,
|
| 3112 |
+
"grad_norm": 0.2412109375,
|
| 3113 |
+
"learning_rate": 2.1780529706380336e-05,
|
| 3114 |
+
"loss": 5.3754,
|
| 3115 |
+
"step": 439
|
| 3116 |
+
},
|
| 3117 |
+
{
|
| 3118 |
+
"epoch": 0.9147609147609148,
|
| 3119 |
+
"grad_norm": 0.27734375,
|
| 3120 |
+
"learning_rate": 2.0786258770873646e-05,
|
| 3121 |
+
"loss": 5.2316,
|
| 3122 |
+
"step": 440
|
| 3123 |
+
},
|
| 3124 |
+
{
|
| 3125 |
+
"epoch": 0.9168399168399168,
|
| 3126 |
+
"grad_norm": 0.294921875,
|
| 3127 |
+
"learning_rate": 1.9814733446237355e-05,
|
| 3128 |
+
"loss": 5.37,
|
| 3129 |
+
"step": 441
|
| 3130 |
+
},
|
| 3131 |
+
{
|
| 3132 |
+
"epoch": 0.918918918918919,
|
| 3133 |
+
"grad_norm": 0.2392578125,
|
| 3134 |
+
"learning_rate": 1.886599984537479e-05,
|
| 3135 |
+
"loss": 5.4565,
|
| 3136 |
+
"step": 442
|
| 3137 |
+
},
|
| 3138 |
+
{
|
| 3139 |
+
"epoch": 0.920997920997921,
|
| 3140 |
+
"grad_norm": 0.31640625,
|
| 3141 |
+
"learning_rate": 1.7940102999393192e-05,
|
| 3142 |
+
"loss": 5.1685,
|
| 3143 |
+
"step": 443
|
| 3144 |
+
},
|
| 3145 |
+
{
|
| 3146 |
+
"epoch": 0.9230769230769231,
|
| 3147 |
+
"grad_norm": 0.275390625,
|
| 3148 |
+
"learning_rate": 1.70370868554659e-05,
|
| 3149 |
+
"loss": 5.2651,
|
| 3150 |
+
"step": 444
|
| 3151 |
+
},
|
| 3152 |
+
{
|
| 3153 |
+
"epoch": 0.9251559251559252,
|
| 3154 |
+
"grad_norm": 0.29296875,
|
| 3155 |
+
"learning_rate": 1.6156994274746485e-05,
|
| 3156 |
+
"loss": 5.1014,
|
| 3157 |
+
"step": 445
|
| 3158 |
+
},
|
| 3159 |
+
{
|
| 3160 |
+
"epoch": 0.9272349272349273,
|
| 3161 |
+
"grad_norm": 0.328125,
|
| 3162 |
+
"learning_rate": 1.5299867030334813e-05,
|
| 3163 |
+
"loss": 4.4473,
|
| 3164 |
+
"step": 446
|
| 3165 |
+
},
|
| 3166 |
+
{
|
| 3167 |
+
"epoch": 0.9293139293139293,
|
| 3168 |
+
"grad_norm": 0.275390625,
|
| 3169 |
+
"learning_rate": 1.4465745805293584e-05,
|
| 3170 |
+
"loss": 5.4957,
|
| 3171 |
+
"step": 447
|
| 3172 |
+
},
|
| 3173 |
+
{
|
| 3174 |
+
"epoch": 0.9313929313929314,
|
| 3175 |
+
"grad_norm": 0.259765625,
|
| 3176 |
+
"learning_rate": 1.3654670190718033e-05,
|
| 3177 |
+
"loss": 5.4438,
|
| 3178 |
+
"step": 448
|
| 3179 |
+
},
|
| 3180 |
+
{
|
| 3181 |
+
"epoch": 0.9334719334719335,
|
| 3182 |
+
"grad_norm": 0.263671875,
|
| 3183 |
+
"learning_rate": 1.286667868385627e-05,
|
| 3184 |
+
"loss": 5.1253,
|
| 3185 |
+
"step": 449
|
| 3186 |
+
},
|
| 3187 |
+
{
|
| 3188 |
+
"epoch": 0.9355509355509356,
|
| 3189 |
+
"grad_norm": 0.25390625,
|
| 3190 |
+
"learning_rate": 1.210180868628219e-05,
|
| 3191 |
+
"loss": 5.2587,
|
| 3192 |
+
"step": 450
|
| 3193 |
+
},
|
| 3194 |
+
{
|
| 3195 |
+
"epoch": 0.9376299376299376,
|
| 3196 |
+
"grad_norm": 0.28125,
|
| 3197 |
+
"learning_rate": 1.1360096502120388e-05,
|
| 3198 |
+
"loss": 5.2587,
|
| 3199 |
+
"step": 451
|
| 3200 |
+
},
|
| 3201 |
+
{
|
| 3202 |
+
"epoch": 0.9397089397089398,
|
| 3203 |
+
"grad_norm": 0.349609375,
|
| 3204 |
+
"learning_rate": 1.064157733632276e-05,
|
| 3205 |
+
"loss": 5.0973,
|
| 3206 |
+
"step": 452
|
| 3207 |
+
},
|
| 3208 |
+
{
|
| 3209 |
+
"epoch": 0.9417879417879418,
|
| 3210 |
+
"grad_norm": 0.263671875,
|
| 3211 |
+
"learning_rate": 9.94628529299768e-06,
|
| 3212 |
+
"loss": 5.1763,
|
| 3213 |
+
"step": 453
|
| 3214 |
+
},
|
| 3215 |
+
{
|
| 3216 |
+
"epoch": 0.9438669438669439,
|
| 3217 |
+
"grad_norm": 0.271484375,
|
| 3218 |
+
"learning_rate": 9.274253373791064e-06,
|
| 3219 |
+
"loss": 5.2711,
|
| 3220 |
+
"step": 454
|
| 3221 |
+
},
|
| 3222 |
+
{
|
| 3223 |
+
"epoch": 0.9459459459459459,
|
| 3224 |
+
"grad_norm": 0.279296875,
|
| 3225 |
+
"learning_rate": 8.62551347632029e-06,
|
| 3226 |
+
"loss": 5.1639,
|
| 3227 |
+
"step": 455
|
| 3228 |
+
},
|
| 3229 |
+
{
|
| 3230 |
+
"epoch": 0.9480249480249481,
|
| 3231 |
+
"grad_norm": 0.26171875,
|
| 3232 |
+
"learning_rate": 8.000096392660028e-06,
|
| 3233 |
+
"loss": 5.2604,
|
| 3234 |
+
"step": 456
|
| 3235 |
+
},
|
| 3236 |
+
{
|
| 3237 |
+
"epoch": 0.9501039501039501,
|
| 3238 |
+
"grad_norm": 0.7890625,
|
| 3239 |
+
"learning_rate": 7.398031807880457e-06,
|
| 3240 |
+
"loss": 4.9583,
|
| 3241 |
+
"step": 457
|
| 3242 |
+
},
|
| 3243 |
+
{
|
| 3244 |
+
"epoch": 0.9521829521829522,
|
| 3245 |
+
"grad_norm": 0.28125,
|
| 3246 |
+
"learning_rate": 6.819348298638839e-06,
|
| 3247 |
+
"loss": 5.5333,
|
| 3248 |
+
"step": 458
|
| 3249 |
+
},
|
| 3250 |
+
{
|
| 3251 |
+
"epoch": 0.9542619542619543,
|
| 3252 |
+
"grad_norm": 0.28515625,
|
| 3253 |
+
"learning_rate": 6.264073331822551e-06,
|
| 3254 |
+
"loss": 5.0472,
|
| 3255 |
+
"step": 459
|
| 3256 |
+
},
|
| 3257 |
+
{
|
| 3258 |
+
"epoch": 0.9563409563409564,
|
| 3259 |
+
"grad_norm": 0.373046875,
|
| 3260 |
+
"learning_rate": 5.732233263245845e-06,
|
| 3261 |
+
"loss": 4.8395,
|
| 3262 |
+
"step": 460
|
| 3263 |
+
},
|
| 3264 |
+
{
|
| 3265 |
+
"epoch": 0.9584199584199584,
|
| 3266 |
+
"grad_norm": 0.3046875,
|
| 3267 |
+
"learning_rate": 5.223853336398632e-06,
|
| 3268 |
+
"loss": 5.3617,
|
| 3269 |
+
"step": 461
|
| 3270 |
+
},
|
| 3271 |
+
{
|
| 3272 |
+
"epoch": 0.9604989604989606,
|
| 3273 |
+
"grad_norm": 0.380859375,
|
| 3274 |
+
"learning_rate": 4.738957681248379e-06,
|
| 3275 |
+
"loss": 4.9956,
|
| 3276 |
+
"step": 462
|
| 3277 |
+
},
|
| 3278 |
+
{
|
| 3279 |
+
"epoch": 0.9625779625779626,
|
| 3280 |
+
"grad_norm": 0.296875,
|
| 3281 |
+
"learning_rate": 4.277569313094809e-06,
|
| 3282 |
+
"loss": 5.0861,
|
| 3283 |
+
"step": 463
|
| 3284 |
+
},
|
| 3285 |
+
{
|
| 3286 |
+
"epoch": 0.9646569646569647,
|
| 3287 |
+
"grad_norm": 0.267578125,
|
| 3288 |
+
"learning_rate": 3.839710131477492e-06,
|
| 3289 |
+
"loss": 5.1291,
|
| 3290 |
+
"step": 464
|
| 3291 |
+
},
|
| 3292 |
+
{
|
| 3293 |
+
"epoch": 0.9667359667359667,
|
| 3294 |
+
"grad_norm": 0.26953125,
|
| 3295 |
+
"learning_rate": 3.4254009191363455e-06,
|
| 3296 |
+
"loss": 4.9485,
|
| 3297 |
+
"step": 465
|
| 3298 |
+
},
|
| 3299 |
+
{
|
| 3300 |
+
"epoch": 0.9688149688149689,
|
| 3301 |
+
"grad_norm": 0.263671875,
|
| 3302 |
+
"learning_rate": 3.034661341025258e-06,
|
| 3303 |
+
"loss": 5.0859,
|
| 3304 |
+
"step": 466
|
| 3305 |
+
},
|
| 3306 |
+
{
|
| 3307 |
+
"epoch": 0.9708939708939709,
|
| 3308 |
+
"grad_norm": 0.68359375,
|
| 3309 |
+
"learning_rate": 2.6675099433787208e-06,
|
| 3310 |
+
"loss": 5.1618,
|
| 3311 |
+
"step": 467
|
| 3312 |
+
},
|
| 3313 |
+
{
|
| 3314 |
+
"epoch": 0.972972972972973,
|
| 3315 |
+
"grad_norm": 0.263671875,
|
| 3316 |
+
"learning_rate": 2.323964152831426e-06,
|
| 3317 |
+
"loss": 5.5762,
|
| 3318 |
+
"step": 468
|
| 3319 |
+
},
|
| 3320 |
+
{
|
| 3321 |
+
"epoch": 0.975051975051975,
|
| 3322 |
+
"grad_norm": 0.36328125,
|
| 3323 |
+
"learning_rate": 2.0040402755912013e-06,
|
| 3324 |
+
"loss": 4.9616,
|
| 3325 |
+
"step": 469
|
| 3326 |
+
},
|
| 3327 |
+
{
|
| 3328 |
+
"epoch": 0.9771309771309772,
|
| 3329 |
+
"grad_norm": 0.283203125,
|
| 3330 |
+
"learning_rate": 1.7077534966650766e-06,
|
| 3331 |
+
"loss": 5.3332,
|
| 3332 |
+
"step": 470
|
| 3333 |
+
},
|
| 3334 |
+
{
|
| 3335 |
+
"epoch": 0.9792099792099792,
|
| 3336 |
+
"grad_norm": 0.26171875,
|
| 3337 |
+
"learning_rate": 1.43511787913847e-06,
|
| 3338 |
+
"loss": 4.9868,
|
| 3339 |
+
"step": 471
|
| 3340 |
+
},
|
| 3341 |
+
{
|
| 3342 |
+
"epoch": 0.9812889812889813,
|
| 3343 |
+
"grad_norm": 0.251953125,
|
| 3344 |
+
"learning_rate": 1.1861463635077786e-06,
|
| 3345 |
+
"loss": 5.3489,
|
| 3346 |
+
"step": 472
|
| 3347 |
+
},
|
| 3348 |
+
{
|
| 3349 |
+
"epoch": 0.9833679833679834,
|
| 3350 |
+
"grad_norm": 0.25390625,
|
| 3351 |
+
"learning_rate": 9.60850767065924e-07,
|
| 3352 |
+
"loss": 5.1146,
|
| 3353 |
+
"step": 473
|
| 3354 |
+
},
|
| 3355 |
+
{
|
| 3356 |
+
"epoch": 0.9854469854469855,
|
| 3357 |
+
"grad_norm": 0.66015625,
|
| 3358 |
+
"learning_rate": 7.592417833419129e-07,
|
| 3359 |
+
"loss": 5.0727,
|
| 3360 |
+
"step": 474
|
| 3361 |
+
},
|
| 3362 |
+
{
|
| 3363 |
+
"epoch": 0.9875259875259875,
|
| 3364 |
+
"grad_norm": 0.2421875,
|
| 3365 |
+
"learning_rate": 5.81328981592688e-07,
|
| 3366 |
+
"loss": 5.5923,
|
| 3367 |
+
"step": 475
|
| 3368 |
+
},
|
| 3369 |
+
{
|
| 3370 |
+
"epoch": 0.9896049896049897,
|
| 3371 |
+
"grad_norm": 0.259765625,
|
| 3372 |
+
"learning_rate": 4.2712080634949023e-07,
|
| 3373 |
+
"loss": 5.2152,
|
| 3374 |
+
"step": 476
|
| 3375 |
+
},
|
| 3376 |
+
{
|
| 3377 |
+
"epoch": 0.9916839916839917,
|
| 3378 |
+
"grad_norm": 0.25,
|
| 3379 |
+
"learning_rate": 2.966245770166243e-07,
|
| 3380 |
+
"loss": 5.2777,
|
| 3381 |
+
"step": 477
|
| 3382 |
+
},
|
| 3383 |
+
{
|
| 3384 |
+
"epoch": 0.9937629937629938,
|
| 3385 |
+
"grad_norm": 0.279296875,
|
| 3386 |
+
"learning_rate": 1.8984648752429223e-07,
|
| 3387 |
+
"loss": 5.2911,
|
| 3388 |
+
"step": 478
|
| 3389 |
+
},
|
| 3390 |
+
{
|
| 3391 |
+
"epoch": 0.9958419958419958,
|
| 3392 |
+
"grad_norm": 0.25390625,
|
| 3393 |
+
"learning_rate": 1.0679160603449533e-07,
|
| 3394 |
+
"loss": 5.1634,
|
| 3395 |
+
"step": 479
|
| 3396 |
+
},
|
| 3397 |
+
{
|
| 3398 |
+
"epoch": 0.997920997920998,
|
| 3399 |
+
"grad_norm": 0.2470703125,
|
| 3400 |
+
"learning_rate": 4.746387470044855e-08,
|
| 3401 |
+
"loss": 5.0967,
|
| 3402 |
+
"step": 480
|
| 3403 |
+
},
|
| 3404 |
+
{
|
| 3405 |
+
"epoch": 1.0,
|
| 3406 |
+
"grad_norm": 0.25390625,
|
| 3407 |
+
"learning_rate": 1.1866109479674591e-08,
|
| 3408 |
+
"loss": 5.2176,
|
| 3409 |
+
"step": 481
|
| 3410 |
}
|
| 3411 |
],
|
| 3412 |
"logging_steps": 1,
|
|
|
|
| 3421 |
"should_evaluate": false,
|
| 3422 |
"should_log": false,
|
| 3423 |
"should_save": true,
|
| 3424 |
+
"should_training_stop": true
|
| 3425 |
},
|
| 3426 |
"attributes": {}
|
| 3427 |
}
|
| 3428 |
},
|
| 3429 |
+
"total_flos": 1.2252207020978995e+17,
|
| 3430 |
"train_batch_size": 32,
|
| 3431 |
"trial_name": null,
|
| 3432 |
"trial_params": null
|