Training in progress, step 500, 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 83945296
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:940c67a8075faf9c1464020b3e68da45b00542da4d7cfcd4f84c3589e2f2c400
|
3 |
size 83945296
|
last-checkpoint/optimizer.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 42546196
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:37dd2877b24b58af0380515e2e777fd635667bff222f1ed18423409b53a91d15
|
3 |
size 42546196
|
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 sha256:3a1c514f7fd02541ee711c621db702524520ab84a034f9ac00217461e361da06
|
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 sha256:0fce62aa16b9baa5f49dced04db0ff1567e194dc4daa50ed8d0d1098cd28a3b6
|
3 |
size 1064
|
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,
|
@@ -2857,13 +2857,163 @@
|
|
2857 |
"learning_rate": 1.129032258064516e-05,
|
2858 |
"loss": 1.536,
|
2859 |
"step": 475
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2860 |
}
|
2861 |
],
|
2862 |
"logging_steps": 1,
|
2863 |
"max_steps": 501,
|
2864 |
"num_train_epochs": 1,
|
2865 |
"save_steps": 25,
|
2866 |
-
"total_flos":
|
2867 |
"trial_name": null,
|
2868 |
"trial_params": null
|
2869 |
}
|
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
+
"epoch": 0.09452689290103035,
|
5 |
"eval_steps": 1000,
|
6 |
+
"global_step": 500,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
|
|
2857 |
"learning_rate": 1.129032258064516e-05,
|
2858 |
"loss": 1.536,
|
2859 |
"step": 475
|
2860 |
+
},
|
2861 |
+
{
|
2862 |
+
"epoch": 0.09,
|
2863 |
+
"learning_rate": 1.0887096774193549e-05,
|
2864 |
+
"loss": 1.7395,
|
2865 |
+
"step": 476
|
2866 |
+
},
|
2867 |
+
{
|
2868 |
+
"epoch": 0.09,
|
2869 |
+
"learning_rate": 1.0483870967741936e-05,
|
2870 |
+
"loss": 1.4813,
|
2871 |
+
"step": 477
|
2872 |
+
},
|
2873 |
+
{
|
2874 |
+
"epoch": 0.09,
|
2875 |
+
"learning_rate": 1.0080645161290323e-05,
|
2876 |
+
"loss": 1.4942,
|
2877 |
+
"step": 478
|
2878 |
+
},
|
2879 |
+
{
|
2880 |
+
"epoch": 0.09,
|
2881 |
+
"learning_rate": 9.67741935483871e-06,
|
2882 |
+
"loss": 1.5411,
|
2883 |
+
"step": 479
|
2884 |
+
},
|
2885 |
+
{
|
2886 |
+
"epoch": 0.09,
|
2887 |
+
"learning_rate": 9.274193548387097e-06,
|
2888 |
+
"loss": 1.6977,
|
2889 |
+
"step": 480
|
2890 |
+
},
|
2891 |
+
{
|
2892 |
+
"epoch": 0.09,
|
2893 |
+
"learning_rate": 8.870967741935484e-06,
|
2894 |
+
"loss": 1.5042,
|
2895 |
+
"step": 481
|
2896 |
+
},
|
2897 |
+
{
|
2898 |
+
"epoch": 0.09,
|
2899 |
+
"learning_rate": 8.46774193548387e-06,
|
2900 |
+
"loss": 1.9896,
|
2901 |
+
"step": 482
|
2902 |
+
},
|
2903 |
+
{
|
2904 |
+
"epoch": 0.09,
|
2905 |
+
"learning_rate": 8.064516129032258e-06,
|
2906 |
+
"loss": 1.4309,
|
2907 |
+
"step": 483
|
2908 |
+
},
|
2909 |
+
{
|
2910 |
+
"epoch": 0.09,
|
2911 |
+
"learning_rate": 7.661290322580646e-06,
|
2912 |
+
"loss": 1.6371,
|
2913 |
+
"step": 484
|
2914 |
+
},
|
2915 |
+
{
|
2916 |
+
"epoch": 0.09,
|
2917 |
+
"learning_rate": 7.258064516129033e-06,
|
2918 |
+
"loss": 1.3355,
|
2919 |
+
"step": 485
|
2920 |
+
},
|
2921 |
+
{
|
2922 |
+
"epoch": 0.09,
|
2923 |
+
"learning_rate": 6.854838709677419e-06,
|
2924 |
+
"loss": 1.7125,
|
2925 |
+
"step": 486
|
2926 |
+
},
|
2927 |
+
{
|
2928 |
+
"epoch": 0.09,
|
2929 |
+
"learning_rate": 6.451612903225806e-06,
|
2930 |
+
"loss": 1.5899,
|
2931 |
+
"step": 487
|
2932 |
+
},
|
2933 |
+
{
|
2934 |
+
"epoch": 0.09,
|
2935 |
+
"learning_rate": 6.048387096774194e-06,
|
2936 |
+
"loss": 1.3346,
|
2937 |
+
"step": 488
|
2938 |
+
},
|
2939 |
+
{
|
2940 |
+
"epoch": 0.09,
|
2941 |
+
"learning_rate": 5.64516129032258e-06,
|
2942 |
+
"loss": 1.5943,
|
2943 |
+
"step": 489
|
2944 |
+
},
|
2945 |
+
{
|
2946 |
+
"epoch": 0.09,
|
2947 |
+
"learning_rate": 5.241935483870968e-06,
|
2948 |
+
"loss": 1.6136,
|
2949 |
+
"step": 490
|
2950 |
+
},
|
2951 |
+
{
|
2952 |
+
"epoch": 0.09,
|
2953 |
+
"learning_rate": 4.838709677419355e-06,
|
2954 |
+
"loss": 1.8624,
|
2955 |
+
"step": 491
|
2956 |
+
},
|
2957 |
+
{
|
2958 |
+
"epoch": 0.09,
|
2959 |
+
"learning_rate": 4.435483870967742e-06,
|
2960 |
+
"loss": 1.5963,
|
2961 |
+
"step": 492
|
2962 |
+
},
|
2963 |
+
{
|
2964 |
+
"epoch": 0.09,
|
2965 |
+
"learning_rate": 4.032258064516129e-06,
|
2966 |
+
"loss": 1.5722,
|
2967 |
+
"step": 493
|
2968 |
+
},
|
2969 |
+
{
|
2970 |
+
"epoch": 0.09,
|
2971 |
+
"learning_rate": 3.6290322580645166e-06,
|
2972 |
+
"loss": 1.7267,
|
2973 |
+
"step": 494
|
2974 |
+
},
|
2975 |
+
{
|
2976 |
+
"epoch": 0.09,
|
2977 |
+
"learning_rate": 3.225806451612903e-06,
|
2978 |
+
"loss": 1.4455,
|
2979 |
+
"step": 495
|
2980 |
+
},
|
2981 |
+
{
|
2982 |
+
"epoch": 0.09,
|
2983 |
+
"learning_rate": 2.82258064516129e-06,
|
2984 |
+
"loss": 1.6525,
|
2985 |
+
"step": 496
|
2986 |
+
},
|
2987 |
+
{
|
2988 |
+
"epoch": 0.09,
|
2989 |
+
"learning_rate": 2.4193548387096776e-06,
|
2990 |
+
"loss": 1.5588,
|
2991 |
+
"step": 497
|
2992 |
+
},
|
2993 |
+
{
|
2994 |
+
"epoch": 0.09,
|
2995 |
+
"learning_rate": 2.0161290322580646e-06,
|
2996 |
+
"loss": 1.7389,
|
2997 |
+
"step": 498
|
2998 |
+
},
|
2999 |
+
{
|
3000 |
+
"epoch": 0.09,
|
3001 |
+
"learning_rate": 1.6129032258064516e-06,
|
3002 |
+
"loss": 1.4058,
|
3003 |
+
"step": 499
|
3004 |
+
},
|
3005 |
+
{
|
3006 |
+
"epoch": 0.09,
|
3007 |
+
"learning_rate": 1.2096774193548388e-06,
|
3008 |
+
"loss": 1.518,
|
3009 |
+
"step": 500
|
3010 |
}
|
3011 |
],
|
3012 |
"logging_steps": 1,
|
3013 |
"max_steps": 501,
|
3014 |
"num_train_epochs": 1,
|
3015 |
"save_steps": 25,
|
3016 |
+
"total_flos": 3.0694656581369856e+16,
|
3017 |
"trial_name": null,
|
3018 |
"trial_params": null
|
3019 |
}
|