Uploaded checkpoint-5000
Browse files- model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +711 -3
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2692969128
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4d140c6a5fd0e6a420d01533a63679c0ba91e292e3f59614b27aca6eedf09f1e
|
| 3 |
size 2692969128
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5386075202
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9867b2e3def1ca7c140948427f730843b2d79132f56cd449e28c21e0bdcffb8a
|
| 3 |
size 5386075202
|
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 sha256:4ce3f1fd7866b93bfea3f328fd35124d6c8818c1dd3b24d163380d4a576714e9
|
| 3 |
size 14244
|
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 sha256:2be67f6aac8e482bb2022409709d8774ffb125292c0c9cf025c0ae747f3a6d57
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": 0.01529085636138916,
|
| 3 |
"best_model_checkpoint": "runs/deepseek_CMU-AIR2/math-deepseek-FULL-ArithHard-30k_20240424-195522/checkpoint-3000",
|
| 4 |
-
"epoch": 1.
|
| 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,
|
|
@@ -2839,6 +2839,714 @@
|
|
| 2839 |
"eval_samples_per_second": 18.667,
|
| 2840 |
"eval_steps_per_second": 18.667,
|
| 2841 |
"step": 4000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2842 |
}
|
| 2843 |
],
|
| 2844 |
"logging_steps": 10,
|
|
@@ -2846,7 +3554,7 @@
|
|
| 2846 |
"num_input_tokens_seen": 0,
|
| 2847 |
"num_train_epochs": 2,
|
| 2848 |
"save_steps": 1000,
|
| 2849 |
-
"total_flos":
|
| 2850 |
"train_batch_size": 1,
|
| 2851 |
"trial_name": null,
|
| 2852 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": 0.01529085636138916,
|
| 3 |
"best_model_checkpoint": "runs/deepseek_CMU-AIR2/math-deepseek-FULL-ArithHard-30k_20240424-195522/checkpoint-3000",
|
| 4 |
+
"epoch": 1.5665387326701654,
|
| 5 |
"eval_steps": 1000,
|
| 6 |
+
"global_step": 5000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 2839 |
"eval_samples_per_second": 18.667,
|
| 2840 |
"eval_steps_per_second": 18.667,
|
| 2841 |
"step": 4000
|
| 2842 |
+
},
|
| 2843 |
+
{
|
| 2844 |
+
"epoch": 1.26,
|
| 2845 |
+
"grad_norm": 0.0050048828125,
|
| 2846 |
+
"learning_rate": 4.4e-06,
|
| 2847 |
+
"loss": 0.0287,
|
| 2848 |
+
"step": 4010
|
| 2849 |
+
},
|
| 2850 |
+
{
|
| 2851 |
+
"epoch": 1.26,
|
| 2852 |
+
"grad_norm": 0.84375,
|
| 2853 |
+
"learning_rate": 4.3555555555555555e-06,
|
| 2854 |
+
"loss": 0.018,
|
| 2855 |
+
"step": 4020
|
| 2856 |
+
},
|
| 2857 |
+
{
|
| 2858 |
+
"epoch": 1.26,
|
| 2859 |
+
"grad_norm": 30.75,
|
| 2860 |
+
"learning_rate": 4.3111111111111115e-06,
|
| 2861 |
+
"loss": 0.0626,
|
| 2862 |
+
"step": 4030
|
| 2863 |
+
},
|
| 2864 |
+
{
|
| 2865 |
+
"epoch": 1.27,
|
| 2866 |
+
"grad_norm": 0.00146484375,
|
| 2867 |
+
"learning_rate": 4.266666666666668e-06,
|
| 2868 |
+
"loss": 0.0084,
|
| 2869 |
+
"step": 4040
|
| 2870 |
+
},
|
| 2871 |
+
{
|
| 2872 |
+
"epoch": 1.27,
|
| 2873 |
+
"grad_norm": 0.004638671875,
|
| 2874 |
+
"learning_rate": 4.222222222222223e-06,
|
| 2875 |
+
"loss": 0.1006,
|
| 2876 |
+
"step": 4050
|
| 2877 |
+
},
|
| 2878 |
+
{
|
| 2879 |
+
"epoch": 1.27,
|
| 2880 |
+
"grad_norm": 4.875,
|
| 2881 |
+
"learning_rate": 4.177777777777778e-06,
|
| 2882 |
+
"loss": 0.0239,
|
| 2883 |
+
"step": 4060
|
| 2884 |
+
},
|
| 2885 |
+
{
|
| 2886 |
+
"epoch": 1.28,
|
| 2887 |
+
"grad_norm": 0.050537109375,
|
| 2888 |
+
"learning_rate": 4.133333333333333e-06,
|
| 2889 |
+
"loss": 0.0034,
|
| 2890 |
+
"step": 4070
|
| 2891 |
+
},
|
| 2892 |
+
{
|
| 2893 |
+
"epoch": 1.28,
|
| 2894 |
+
"grad_norm": 7.1875,
|
| 2895 |
+
"learning_rate": 4.088888888888889e-06,
|
| 2896 |
+
"loss": 0.0413,
|
| 2897 |
+
"step": 4080
|
| 2898 |
+
},
|
| 2899 |
+
{
|
| 2900 |
+
"epoch": 1.28,
|
| 2901 |
+
"grad_norm": 0.0019683837890625,
|
| 2902 |
+
"learning_rate": 4.044444444444445e-06,
|
| 2903 |
+
"loss": 0.0236,
|
| 2904 |
+
"step": 4090
|
| 2905 |
+
},
|
| 2906 |
+
{
|
| 2907 |
+
"epoch": 1.28,
|
| 2908 |
+
"grad_norm": 0.0152587890625,
|
| 2909 |
+
"learning_rate": 4.000000000000001e-06,
|
| 2910 |
+
"loss": 0.0404,
|
| 2911 |
+
"step": 4100
|
| 2912 |
+
},
|
| 2913 |
+
{
|
| 2914 |
+
"epoch": 1.29,
|
| 2915 |
+
"grad_norm": 0.00201416015625,
|
| 2916 |
+
"learning_rate": 3.955555555555556e-06,
|
| 2917 |
+
"loss": 0.0219,
|
| 2918 |
+
"step": 4110
|
| 2919 |
+
},
|
| 2920 |
+
{
|
| 2921 |
+
"epoch": 1.29,
|
| 2922 |
+
"grad_norm": 10.6875,
|
| 2923 |
+
"learning_rate": 3.911111111111112e-06,
|
| 2924 |
+
"loss": 0.0543,
|
| 2925 |
+
"step": 4120
|
| 2926 |
+
},
|
| 2927 |
+
{
|
| 2928 |
+
"epoch": 1.29,
|
| 2929 |
+
"grad_norm": 6.9375,
|
| 2930 |
+
"learning_rate": 3.866666666666667e-06,
|
| 2931 |
+
"loss": 0.0308,
|
| 2932 |
+
"step": 4130
|
| 2933 |
+
},
|
| 2934 |
+
{
|
| 2935 |
+
"epoch": 1.3,
|
| 2936 |
+
"grad_norm": 0.05615234375,
|
| 2937 |
+
"learning_rate": 3.8222222222222224e-06,
|
| 2938 |
+
"loss": 0.019,
|
| 2939 |
+
"step": 4140
|
| 2940 |
+
},
|
| 2941 |
+
{
|
| 2942 |
+
"epoch": 1.3,
|
| 2943 |
+
"grad_norm": 0.00099945068359375,
|
| 2944 |
+
"learning_rate": 3.777777777777778e-06,
|
| 2945 |
+
"loss": 0.0747,
|
| 2946 |
+
"step": 4150
|
| 2947 |
+
},
|
| 2948 |
+
{
|
| 2949 |
+
"epoch": 1.3,
|
| 2950 |
+
"grad_norm": 7.75,
|
| 2951 |
+
"learning_rate": 3.7333333333333337e-06,
|
| 2952 |
+
"loss": 0.0319,
|
| 2953 |
+
"step": 4160
|
| 2954 |
+
},
|
| 2955 |
+
{
|
| 2956 |
+
"epoch": 1.31,
|
| 2957 |
+
"grad_norm": 4.21875,
|
| 2958 |
+
"learning_rate": 3.688888888888889e-06,
|
| 2959 |
+
"loss": 0.009,
|
| 2960 |
+
"step": 4170
|
| 2961 |
+
},
|
| 2962 |
+
{
|
| 2963 |
+
"epoch": 1.31,
|
| 2964 |
+
"grad_norm": 0.001708984375,
|
| 2965 |
+
"learning_rate": 3.644444444444445e-06,
|
| 2966 |
+
"loss": 0.0174,
|
| 2967 |
+
"step": 4180
|
| 2968 |
+
},
|
| 2969 |
+
{
|
| 2970 |
+
"epoch": 1.31,
|
| 2971 |
+
"grad_norm": 0.000949859619140625,
|
| 2972 |
+
"learning_rate": 3.6000000000000003e-06,
|
| 2973 |
+
"loss": 0.0062,
|
| 2974 |
+
"step": 4190
|
| 2975 |
+
},
|
| 2976 |
+
{
|
| 2977 |
+
"epoch": 1.32,
|
| 2978 |
+
"grad_norm": 0.0011138916015625,
|
| 2979 |
+
"learning_rate": 3.555555555555556e-06,
|
| 2980 |
+
"loss": 0.0063,
|
| 2981 |
+
"step": 4200
|
| 2982 |
+
},
|
| 2983 |
+
{
|
| 2984 |
+
"epoch": 1.32,
|
| 2985 |
+
"grad_norm": 15.0,
|
| 2986 |
+
"learning_rate": 3.511111111111111e-06,
|
| 2987 |
+
"loss": 0.016,
|
| 2988 |
+
"step": 4210
|
| 2989 |
+
},
|
| 2990 |
+
{
|
| 2991 |
+
"epoch": 1.32,
|
| 2992 |
+
"grad_norm": 0.0059814453125,
|
| 2993 |
+
"learning_rate": 3.4666666666666672e-06,
|
| 2994 |
+
"loss": 0.017,
|
| 2995 |
+
"step": 4220
|
| 2996 |
+
},
|
| 2997 |
+
{
|
| 2998 |
+
"epoch": 1.33,
|
| 2999 |
+
"grad_norm": 0.005126953125,
|
| 3000 |
+
"learning_rate": 3.4222222222222224e-06,
|
| 3001 |
+
"loss": 0.0445,
|
| 3002 |
+
"step": 4230
|
| 3003 |
+
},
|
| 3004 |
+
{
|
| 3005 |
+
"epoch": 1.33,
|
| 3006 |
+
"grad_norm": 0.00115203857421875,
|
| 3007 |
+
"learning_rate": 3.377777777777778e-06,
|
| 3008 |
+
"loss": 0.0146,
|
| 3009 |
+
"step": 4240
|
| 3010 |
+
},
|
| 3011 |
+
{
|
| 3012 |
+
"epoch": 1.33,
|
| 3013 |
+
"grad_norm": 0.00102996826171875,
|
| 3014 |
+
"learning_rate": 3.3333333333333333e-06,
|
| 3015 |
+
"loss": 0.0121,
|
| 3016 |
+
"step": 4250
|
| 3017 |
+
},
|
| 3018 |
+
{
|
| 3019 |
+
"epoch": 1.33,
|
| 3020 |
+
"grad_norm": 0.0018157958984375,
|
| 3021 |
+
"learning_rate": 3.2888888888888894e-06,
|
| 3022 |
+
"loss": 0.015,
|
| 3023 |
+
"step": 4260
|
| 3024 |
+
},
|
| 3025 |
+
{
|
| 3026 |
+
"epoch": 1.34,
|
| 3027 |
+
"grad_norm": 0.0048828125,
|
| 3028 |
+
"learning_rate": 3.2444444444444446e-06,
|
| 3029 |
+
"loss": 0.0258,
|
| 3030 |
+
"step": 4270
|
| 3031 |
+
},
|
| 3032 |
+
{
|
| 3033 |
+
"epoch": 1.34,
|
| 3034 |
+
"grad_norm": 6.8125,
|
| 3035 |
+
"learning_rate": 3.2000000000000003e-06,
|
| 3036 |
+
"loss": 0.1234,
|
| 3037 |
+
"step": 4280
|
| 3038 |
+
},
|
| 3039 |
+
{
|
| 3040 |
+
"epoch": 1.34,
|
| 3041 |
+
"grad_norm": 0.0011138916015625,
|
| 3042 |
+
"learning_rate": 3.1555555555555555e-06,
|
| 3043 |
+
"loss": 0.0093,
|
| 3044 |
+
"step": 4290
|
| 3045 |
+
},
|
| 3046 |
+
{
|
| 3047 |
+
"epoch": 1.35,
|
| 3048 |
+
"grad_norm": 0.0012969970703125,
|
| 3049 |
+
"learning_rate": 3.1111111111111116e-06,
|
| 3050 |
+
"loss": 0.007,
|
| 3051 |
+
"step": 4300
|
| 3052 |
+
},
|
| 3053 |
+
{
|
| 3054 |
+
"epoch": 1.35,
|
| 3055 |
+
"grad_norm": 0.00946044921875,
|
| 3056 |
+
"learning_rate": 3.066666666666667e-06,
|
| 3057 |
+
"loss": 0.0176,
|
| 3058 |
+
"step": 4310
|
| 3059 |
+
},
|
| 3060 |
+
{
|
| 3061 |
+
"epoch": 1.35,
|
| 3062 |
+
"grad_norm": 0.00457763671875,
|
| 3063 |
+
"learning_rate": 3.0222222222222225e-06,
|
| 3064 |
+
"loss": 0.0416,
|
| 3065 |
+
"step": 4320
|
| 3066 |
+
},
|
| 3067 |
+
{
|
| 3068 |
+
"epoch": 1.36,
|
| 3069 |
+
"grad_norm": 0.0016326904296875,
|
| 3070 |
+
"learning_rate": 2.9777777777777777e-06,
|
| 3071 |
+
"loss": 0.0243,
|
| 3072 |
+
"step": 4330
|
| 3073 |
+
},
|
| 3074 |
+
{
|
| 3075 |
+
"epoch": 1.36,
|
| 3076 |
+
"grad_norm": 0.0013275146484375,
|
| 3077 |
+
"learning_rate": 2.9333333333333338e-06,
|
| 3078 |
+
"loss": 0.047,
|
| 3079 |
+
"step": 4340
|
| 3080 |
+
},
|
| 3081 |
+
{
|
| 3082 |
+
"epoch": 1.36,
|
| 3083 |
+
"grad_norm": 0.004638671875,
|
| 3084 |
+
"learning_rate": 2.888888888888889e-06,
|
| 3085 |
+
"loss": 0.0327,
|
| 3086 |
+
"step": 4350
|
| 3087 |
+
},
|
| 3088 |
+
{
|
| 3089 |
+
"epoch": 1.37,
|
| 3090 |
+
"grad_norm": 6.34375,
|
| 3091 |
+
"learning_rate": 2.8444444444444446e-06,
|
| 3092 |
+
"loss": 0.0559,
|
| 3093 |
+
"step": 4360
|
| 3094 |
+
},
|
| 3095 |
+
{
|
| 3096 |
+
"epoch": 1.37,
|
| 3097 |
+
"grad_norm": 13.625,
|
| 3098 |
+
"learning_rate": 2.8000000000000003e-06,
|
| 3099 |
+
"loss": 0.0301,
|
| 3100 |
+
"step": 4370
|
| 3101 |
+
},
|
| 3102 |
+
{
|
| 3103 |
+
"epoch": 1.37,
|
| 3104 |
+
"grad_norm": 6.28125,
|
| 3105 |
+
"learning_rate": 2.755555555555556e-06,
|
| 3106 |
+
"loss": 0.0494,
|
| 3107 |
+
"step": 4380
|
| 3108 |
+
},
|
| 3109 |
+
{
|
| 3110 |
+
"epoch": 1.38,
|
| 3111 |
+
"grad_norm": 0.00408935546875,
|
| 3112 |
+
"learning_rate": 2.7111111111111116e-06,
|
| 3113 |
+
"loss": 0.0182,
|
| 3114 |
+
"step": 4390
|
| 3115 |
+
},
|
| 3116 |
+
{
|
| 3117 |
+
"epoch": 1.38,
|
| 3118 |
+
"grad_norm": 14.125,
|
| 3119 |
+
"learning_rate": 2.666666666666667e-06,
|
| 3120 |
+
"loss": 0.045,
|
| 3121 |
+
"step": 4400
|
| 3122 |
+
},
|
| 3123 |
+
{
|
| 3124 |
+
"epoch": 1.38,
|
| 3125 |
+
"grad_norm": 0.00121307373046875,
|
| 3126 |
+
"learning_rate": 2.6222222222222225e-06,
|
| 3127 |
+
"loss": 0.0493,
|
| 3128 |
+
"step": 4410
|
| 3129 |
+
},
|
| 3130 |
+
{
|
| 3131 |
+
"epoch": 1.38,
|
| 3132 |
+
"grad_norm": 6.53125,
|
| 3133 |
+
"learning_rate": 2.577777777777778e-06,
|
| 3134 |
+
"loss": 0.0232,
|
| 3135 |
+
"step": 4420
|
| 3136 |
+
},
|
| 3137 |
+
{
|
| 3138 |
+
"epoch": 1.39,
|
| 3139 |
+
"grad_norm": 0.00131988525390625,
|
| 3140 |
+
"learning_rate": 2.5333333333333338e-06,
|
| 3141 |
+
"loss": 0.0207,
|
| 3142 |
+
"step": 4430
|
| 3143 |
+
},
|
| 3144 |
+
{
|
| 3145 |
+
"epoch": 1.39,
|
| 3146 |
+
"grad_norm": 2.984375,
|
| 3147 |
+
"learning_rate": 2.488888888888889e-06,
|
| 3148 |
+
"loss": 0.0627,
|
| 3149 |
+
"step": 4440
|
| 3150 |
+
},
|
| 3151 |
+
{
|
| 3152 |
+
"epoch": 1.39,
|
| 3153 |
+
"grad_norm": 0.00109100341796875,
|
| 3154 |
+
"learning_rate": 2.4444444444444447e-06,
|
| 3155 |
+
"loss": 0.0272,
|
| 3156 |
+
"step": 4450
|
| 3157 |
+
},
|
| 3158 |
+
{
|
| 3159 |
+
"epoch": 1.4,
|
| 3160 |
+
"grad_norm": 0.00457763671875,
|
| 3161 |
+
"learning_rate": 2.4000000000000003e-06,
|
| 3162 |
+
"loss": 0.0139,
|
| 3163 |
+
"step": 4460
|
| 3164 |
+
},
|
| 3165 |
+
{
|
| 3166 |
+
"epoch": 1.4,
|
| 3167 |
+
"grad_norm": 5.375,
|
| 3168 |
+
"learning_rate": 2.3555555555555555e-06,
|
| 3169 |
+
"loss": 0.0355,
|
| 3170 |
+
"step": 4470
|
| 3171 |
+
},
|
| 3172 |
+
{
|
| 3173 |
+
"epoch": 1.4,
|
| 3174 |
+
"grad_norm": 13.375,
|
| 3175 |
+
"learning_rate": 2.311111111111111e-06,
|
| 3176 |
+
"loss": 0.0465,
|
| 3177 |
+
"step": 4480
|
| 3178 |
+
},
|
| 3179 |
+
{
|
| 3180 |
+
"epoch": 1.41,
|
| 3181 |
+
"grad_norm": 0.045654296875,
|
| 3182 |
+
"learning_rate": 2.266666666666667e-06,
|
| 3183 |
+
"loss": 0.0154,
|
| 3184 |
+
"step": 4490
|
| 3185 |
+
},
|
| 3186 |
+
{
|
| 3187 |
+
"epoch": 1.41,
|
| 3188 |
+
"grad_norm": 0.004638671875,
|
| 3189 |
+
"learning_rate": 2.222222222222222e-06,
|
| 3190 |
+
"loss": 0.0128,
|
| 3191 |
+
"step": 4500
|
| 3192 |
+
},
|
| 3193 |
+
{
|
| 3194 |
+
"epoch": 1.41,
|
| 3195 |
+
"grad_norm": 0.00168609619140625,
|
| 3196 |
+
"learning_rate": 2.1777777777777777e-06,
|
| 3197 |
+
"loss": 0.0219,
|
| 3198 |
+
"step": 4510
|
| 3199 |
+
},
|
| 3200 |
+
{
|
| 3201 |
+
"epoch": 1.42,
|
| 3202 |
+
"grad_norm": 3.53125,
|
| 3203 |
+
"learning_rate": 2.133333333333334e-06,
|
| 3204 |
+
"loss": 0.0216,
|
| 3205 |
+
"step": 4520
|
| 3206 |
+
},
|
| 3207 |
+
{
|
| 3208 |
+
"epoch": 1.42,
|
| 3209 |
+
"grad_norm": 0.000942230224609375,
|
| 3210 |
+
"learning_rate": 2.088888888888889e-06,
|
| 3211 |
+
"loss": 0.0253,
|
| 3212 |
+
"step": 4530
|
| 3213 |
+
},
|
| 3214 |
+
{
|
| 3215 |
+
"epoch": 1.42,
|
| 3216 |
+
"grad_norm": 0.001617431640625,
|
| 3217 |
+
"learning_rate": 2.0444444444444447e-06,
|
| 3218 |
+
"loss": 0.0129,
|
| 3219 |
+
"step": 4540
|
| 3220 |
+
},
|
| 3221 |
+
{
|
| 3222 |
+
"epoch": 1.43,
|
| 3223 |
+
"grad_norm": 0.0012664794921875,
|
| 3224 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 3225 |
+
"loss": 0.0101,
|
| 3226 |
+
"step": 4550
|
| 3227 |
+
},
|
| 3228 |
+
{
|
| 3229 |
+
"epoch": 1.43,
|
| 3230 |
+
"grad_norm": 0.00170135498046875,
|
| 3231 |
+
"learning_rate": 1.955555555555556e-06,
|
| 3232 |
+
"loss": 0.0086,
|
| 3233 |
+
"step": 4560
|
| 3234 |
+
},
|
| 3235 |
+
{
|
| 3236 |
+
"epoch": 1.43,
|
| 3237 |
+
"grad_norm": 0.000946044921875,
|
| 3238 |
+
"learning_rate": 1.9111111111111112e-06,
|
| 3239 |
+
"loss": 0.0114,
|
| 3240 |
+
"step": 4570
|
| 3241 |
+
},
|
| 3242 |
+
{
|
| 3243 |
+
"epoch": 1.43,
|
| 3244 |
+
"grad_norm": 0.00153350830078125,
|
| 3245 |
+
"learning_rate": 1.8666666666666669e-06,
|
| 3246 |
+
"loss": 0.0088,
|
| 3247 |
+
"step": 4580
|
| 3248 |
+
},
|
| 3249 |
+
{
|
| 3250 |
+
"epoch": 1.44,
|
| 3251 |
+
"grad_norm": 0.004669189453125,
|
| 3252 |
+
"learning_rate": 1.8222222222222225e-06,
|
| 3253 |
+
"loss": 0.0502,
|
| 3254 |
+
"step": 4590
|
| 3255 |
+
},
|
| 3256 |
+
{
|
| 3257 |
+
"epoch": 1.44,
|
| 3258 |
+
"grad_norm": 0.004638671875,
|
| 3259 |
+
"learning_rate": 1.777777777777778e-06,
|
| 3260 |
+
"loss": 0.0345,
|
| 3261 |
+
"step": 4600
|
| 3262 |
+
},
|
| 3263 |
+
{
|
| 3264 |
+
"epoch": 1.44,
|
| 3265 |
+
"grad_norm": 0.00092315673828125,
|
| 3266 |
+
"learning_rate": 1.7333333333333336e-06,
|
| 3267 |
+
"loss": 0.0224,
|
| 3268 |
+
"step": 4610
|
| 3269 |
+
},
|
| 3270 |
+
{
|
| 3271 |
+
"epoch": 1.45,
|
| 3272 |
+
"grad_norm": 0.00113677978515625,
|
| 3273 |
+
"learning_rate": 1.688888888888889e-06,
|
| 3274 |
+
"loss": 0.025,
|
| 3275 |
+
"step": 4620
|
| 3276 |
+
},
|
| 3277 |
+
{
|
| 3278 |
+
"epoch": 1.45,
|
| 3279 |
+
"grad_norm": 5.71875,
|
| 3280 |
+
"learning_rate": 1.6444444444444447e-06,
|
| 3281 |
+
"loss": 0.066,
|
| 3282 |
+
"step": 4630
|
| 3283 |
+
},
|
| 3284 |
+
{
|
| 3285 |
+
"epoch": 1.45,
|
| 3286 |
+
"grad_norm": 0.0013427734375,
|
| 3287 |
+
"learning_rate": 1.6000000000000001e-06,
|
| 3288 |
+
"loss": 0.0,
|
| 3289 |
+
"step": 4640
|
| 3290 |
+
},
|
| 3291 |
+
{
|
| 3292 |
+
"epoch": 1.46,
|
| 3293 |
+
"grad_norm": 5.875,
|
| 3294 |
+
"learning_rate": 1.5555555555555558e-06,
|
| 3295 |
+
"loss": 0.0285,
|
| 3296 |
+
"step": 4650
|
| 3297 |
+
},
|
| 3298 |
+
{
|
| 3299 |
+
"epoch": 1.46,
|
| 3300 |
+
"grad_norm": 0.0703125,
|
| 3301 |
+
"learning_rate": 1.5111111111111112e-06,
|
| 3302 |
+
"loss": 0.0825,
|
| 3303 |
+
"step": 4660
|
| 3304 |
+
},
|
| 3305 |
+
{
|
| 3306 |
+
"epoch": 1.46,
|
| 3307 |
+
"grad_norm": 0.00113677978515625,
|
| 3308 |
+
"learning_rate": 1.4666666666666669e-06,
|
| 3309 |
+
"loss": 0.0107,
|
| 3310 |
+
"step": 4670
|
| 3311 |
+
},
|
| 3312 |
+
{
|
| 3313 |
+
"epoch": 1.47,
|
| 3314 |
+
"grad_norm": 0.0009918212890625,
|
| 3315 |
+
"learning_rate": 1.4222222222222223e-06,
|
| 3316 |
+
"loss": 0.0052,
|
| 3317 |
+
"step": 4680
|
| 3318 |
+
},
|
| 3319 |
+
{
|
| 3320 |
+
"epoch": 1.47,
|
| 3321 |
+
"grad_norm": 4.0625,
|
| 3322 |
+
"learning_rate": 1.377777777777778e-06,
|
| 3323 |
+
"loss": 0.0596,
|
| 3324 |
+
"step": 4690
|
| 3325 |
+
},
|
| 3326 |
+
{
|
| 3327 |
+
"epoch": 1.47,
|
| 3328 |
+
"grad_norm": 3.78125,
|
| 3329 |
+
"learning_rate": 1.3333333333333334e-06,
|
| 3330 |
+
"loss": 0.0114,
|
| 3331 |
+
"step": 4700
|
| 3332 |
+
},
|
| 3333 |
+
{
|
| 3334 |
+
"epoch": 1.48,
|
| 3335 |
+
"grad_norm": 0.0018310546875,
|
| 3336 |
+
"learning_rate": 1.288888888888889e-06,
|
| 3337 |
+
"loss": 0.0101,
|
| 3338 |
+
"step": 4710
|
| 3339 |
+
},
|
| 3340 |
+
{
|
| 3341 |
+
"epoch": 1.48,
|
| 3342 |
+
"grad_norm": 0.00299072265625,
|
| 3343 |
+
"learning_rate": 1.2444444444444445e-06,
|
| 3344 |
+
"loss": 0.014,
|
| 3345 |
+
"step": 4720
|
| 3346 |
+
},
|
| 3347 |
+
{
|
| 3348 |
+
"epoch": 1.48,
|
| 3349 |
+
"grad_norm": 1.859375,
|
| 3350 |
+
"learning_rate": 1.2000000000000002e-06,
|
| 3351 |
+
"loss": 0.0429,
|
| 3352 |
+
"step": 4730
|
| 3353 |
+
},
|
| 3354 |
+
{
|
| 3355 |
+
"epoch": 1.49,
|
| 3356 |
+
"grad_norm": 12.0625,
|
| 3357 |
+
"learning_rate": 1.1555555555555556e-06,
|
| 3358 |
+
"loss": 0.0624,
|
| 3359 |
+
"step": 4740
|
| 3360 |
+
},
|
| 3361 |
+
{
|
| 3362 |
+
"epoch": 1.49,
|
| 3363 |
+
"grad_norm": 0.00177001953125,
|
| 3364 |
+
"learning_rate": 1.111111111111111e-06,
|
| 3365 |
+
"loss": 0.0485,
|
| 3366 |
+
"step": 4750
|
| 3367 |
+
},
|
| 3368 |
+
{
|
| 3369 |
+
"epoch": 1.49,
|
| 3370 |
+
"grad_norm": 0.00482177734375,
|
| 3371 |
+
"learning_rate": 1.066666666666667e-06,
|
| 3372 |
+
"loss": 0.0042,
|
| 3373 |
+
"step": 4760
|
| 3374 |
+
},
|
| 3375 |
+
{
|
| 3376 |
+
"epoch": 1.49,
|
| 3377 |
+
"grad_norm": 3.671875,
|
| 3378 |
+
"learning_rate": 1.0222222222222223e-06,
|
| 3379 |
+
"loss": 0.032,
|
| 3380 |
+
"step": 4770
|
| 3381 |
+
},
|
| 3382 |
+
{
|
| 3383 |
+
"epoch": 1.5,
|
| 3384 |
+
"grad_norm": 0.003570556640625,
|
| 3385 |
+
"learning_rate": 9.77777777777778e-07,
|
| 3386 |
+
"loss": 0.0588,
|
| 3387 |
+
"step": 4780
|
| 3388 |
+
},
|
| 3389 |
+
{
|
| 3390 |
+
"epoch": 1.5,
|
| 3391 |
+
"grad_norm": 0.004608154296875,
|
| 3392 |
+
"learning_rate": 9.333333333333334e-07,
|
| 3393 |
+
"loss": 0.034,
|
| 3394 |
+
"step": 4790
|
| 3395 |
+
},
|
| 3396 |
+
{
|
| 3397 |
+
"epoch": 1.5,
|
| 3398 |
+
"grad_norm": 10.5,
|
| 3399 |
+
"learning_rate": 8.88888888888889e-07,
|
| 3400 |
+
"loss": 0.1101,
|
| 3401 |
+
"step": 4800
|
| 3402 |
+
},
|
| 3403 |
+
{
|
| 3404 |
+
"epoch": 1.51,
|
| 3405 |
+
"grad_norm": 0.0017547607421875,
|
| 3406 |
+
"learning_rate": 8.444444444444445e-07,
|
| 3407 |
+
"loss": 0.0062,
|
| 3408 |
+
"step": 4810
|
| 3409 |
+
},
|
| 3410 |
+
{
|
| 3411 |
+
"epoch": 1.51,
|
| 3412 |
+
"grad_norm": 0.004669189453125,
|
| 3413 |
+
"learning_rate": 8.000000000000001e-07,
|
| 3414 |
+
"loss": 0.042,
|
| 3415 |
+
"step": 4820
|
| 3416 |
+
},
|
| 3417 |
+
{
|
| 3418 |
+
"epoch": 1.51,
|
| 3419 |
+
"grad_norm": 9.1875,
|
| 3420 |
+
"learning_rate": 7.555555555555556e-07,
|
| 3421 |
+
"loss": 0.0495,
|
| 3422 |
+
"step": 4830
|
| 3423 |
+
},
|
| 3424 |
+
{
|
| 3425 |
+
"epoch": 1.52,
|
| 3426 |
+
"grad_norm": 0.1708984375,
|
| 3427 |
+
"learning_rate": 7.111111111111112e-07,
|
| 3428 |
+
"loss": 0.0104,
|
| 3429 |
+
"step": 4840
|
| 3430 |
+
},
|
| 3431 |
+
{
|
| 3432 |
+
"epoch": 1.52,
|
| 3433 |
+
"grad_norm": 0.004608154296875,
|
| 3434 |
+
"learning_rate": 6.666666666666667e-07,
|
| 3435 |
+
"loss": 0.0433,
|
| 3436 |
+
"step": 4850
|
| 3437 |
+
},
|
| 3438 |
+
{
|
| 3439 |
+
"epoch": 1.52,
|
| 3440 |
+
"grad_norm": 0.001251220703125,
|
| 3441 |
+
"learning_rate": 6.222222222222223e-07,
|
| 3442 |
+
"loss": 0.003,
|
| 3443 |
+
"step": 4860
|
| 3444 |
+
},
|
| 3445 |
+
{
|
| 3446 |
+
"epoch": 1.53,
|
| 3447 |
+
"grad_norm": 0.001007080078125,
|
| 3448 |
+
"learning_rate": 5.777777777777778e-07,
|
| 3449 |
+
"loss": 0.0561,
|
| 3450 |
+
"step": 4870
|
| 3451 |
+
},
|
| 3452 |
+
{
|
| 3453 |
+
"epoch": 1.53,
|
| 3454 |
+
"grad_norm": 5.46875,
|
| 3455 |
+
"learning_rate": 5.333333333333335e-07,
|
| 3456 |
+
"loss": 0.0171,
|
| 3457 |
+
"step": 4880
|
| 3458 |
+
},
|
| 3459 |
+
{
|
| 3460 |
+
"epoch": 1.53,
|
| 3461 |
+
"grad_norm": 16.625,
|
| 3462 |
+
"learning_rate": 4.88888888888889e-07,
|
| 3463 |
+
"loss": 0.0468,
|
| 3464 |
+
"step": 4890
|
| 3465 |
+
},
|
| 3466 |
+
{
|
| 3467 |
+
"epoch": 1.54,
|
| 3468 |
+
"grad_norm": 0.00116729736328125,
|
| 3469 |
+
"learning_rate": 4.444444444444445e-07,
|
| 3470 |
+
"loss": 0.024,
|
| 3471 |
+
"step": 4900
|
| 3472 |
+
},
|
| 3473 |
+
{
|
| 3474 |
+
"epoch": 1.54,
|
| 3475 |
+
"grad_norm": 0.0010528564453125,
|
| 3476 |
+
"learning_rate": 4.0000000000000003e-07,
|
| 3477 |
+
"loss": 0.0727,
|
| 3478 |
+
"step": 4910
|
| 3479 |
+
},
|
| 3480 |
+
{
|
| 3481 |
+
"epoch": 1.54,
|
| 3482 |
+
"grad_norm": 0.02197265625,
|
| 3483 |
+
"learning_rate": 3.555555555555556e-07,
|
| 3484 |
+
"loss": 0.0079,
|
| 3485 |
+
"step": 4920
|
| 3486 |
+
},
|
| 3487 |
+
{
|
| 3488 |
+
"epoch": 1.54,
|
| 3489 |
+
"grad_norm": 0.0010528564453125,
|
| 3490 |
+
"learning_rate": 3.111111111111111e-07,
|
| 3491 |
+
"loss": 0.0251,
|
| 3492 |
+
"step": 4930
|
| 3493 |
+
},
|
| 3494 |
+
{
|
| 3495 |
+
"epoch": 1.55,
|
| 3496 |
+
"grad_norm": 0.0011444091796875,
|
| 3497 |
+
"learning_rate": 2.666666666666667e-07,
|
| 3498 |
+
"loss": 0.0053,
|
| 3499 |
+
"step": 4940
|
| 3500 |
+
},
|
| 3501 |
+
{
|
| 3502 |
+
"epoch": 1.55,
|
| 3503 |
+
"grad_norm": 0.001190185546875,
|
| 3504 |
+
"learning_rate": 2.2222222222222224e-07,
|
| 3505 |
+
"loss": 0.016,
|
| 3506 |
+
"step": 4950
|
| 3507 |
+
},
|
| 3508 |
+
{
|
| 3509 |
+
"epoch": 1.55,
|
| 3510 |
+
"grad_norm": 0.0018310546875,
|
| 3511 |
+
"learning_rate": 1.777777777777778e-07,
|
| 3512 |
+
"loss": 0.0398,
|
| 3513 |
+
"step": 4960
|
| 3514 |
+
},
|
| 3515 |
+
{
|
| 3516 |
+
"epoch": 1.56,
|
| 3517 |
+
"grad_norm": 0.00189971923828125,
|
| 3518 |
+
"learning_rate": 1.3333333333333336e-07,
|
| 3519 |
+
"loss": 0.0438,
|
| 3520 |
+
"step": 4970
|
| 3521 |
+
},
|
| 3522 |
+
{
|
| 3523 |
+
"epoch": 1.56,
|
| 3524 |
+
"grad_norm": 0.00121307373046875,
|
| 3525 |
+
"learning_rate": 8.88888888888889e-08,
|
| 3526 |
+
"loss": 0.02,
|
| 3527 |
+
"step": 4980
|
| 3528 |
+
},
|
| 3529 |
+
{
|
| 3530 |
+
"epoch": 1.56,
|
| 3531 |
+
"grad_norm": 0.0034027099609375,
|
| 3532 |
+
"learning_rate": 4.444444444444445e-08,
|
| 3533 |
+
"loss": 0.0511,
|
| 3534 |
+
"step": 4990
|
| 3535 |
+
},
|
| 3536 |
+
{
|
| 3537 |
+
"epoch": 1.57,
|
| 3538 |
+
"grad_norm": 0.00183868408203125,
|
| 3539 |
+
"learning_rate": 0.0,
|
| 3540 |
+
"loss": 0.0318,
|
| 3541 |
+
"step": 5000
|
| 3542 |
+
},
|
| 3543 |
+
{
|
| 3544 |
+
"epoch": 1.57,
|
| 3545 |
+
"eval_loss": 0.02059042826294899,
|
| 3546 |
+
"eval_runtime": 53.6829,
|
| 3547 |
+
"eval_samples_per_second": 18.628,
|
| 3548 |
+
"eval_steps_per_second": 18.628,
|
| 3549 |
+
"step": 5000
|
| 3550 |
}
|
| 3551 |
],
|
| 3552 |
"logging_steps": 10,
|
|
|
|
| 3554 |
"num_input_tokens_seen": 0,
|
| 3555 |
"num_train_epochs": 2,
|
| 3556 |
"save_steps": 1000,
|
| 3557 |
+
"total_flos": 7.866849165312e+16,
|
| 3558 |
"train_batch_size": 1,
|
| 3559 |
"trial_name": null,
|
| 3560 |
"trial_params": null
|