Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1021 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1021 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 50,
|
| 6 |
+
"global_step": 485,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.008247422680412371,
|
| 13 |
+
"grad_norm": 16.375,
|
| 14 |
+
"learning_rate": 6.666666666666667e-06,
|
| 15 |
+
"loss": 0.8858,
|
| 16 |
+
"num_input_tokens_seen": 1413808,
|
| 17 |
+
"step": 4
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.016494845360824743,
|
| 21 |
+
"grad_norm": 4.1875,
|
| 22 |
+
"learning_rate": 1.3333333333333333e-05,
|
| 23 |
+
"loss": 0.4088,
|
| 24 |
+
"num_input_tokens_seen": 2866496,
|
| 25 |
+
"step": 8
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 0.024742268041237112,
|
| 29 |
+
"grad_norm": 3.03125,
|
| 30 |
+
"learning_rate": 2e-05,
|
| 31 |
+
"loss": 0.2731,
|
| 32 |
+
"num_input_tokens_seen": 4305104,
|
| 33 |
+
"step": 12
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.032989690721649485,
|
| 37 |
+
"grad_norm": 1.6484375,
|
| 38 |
+
"learning_rate": 2.6666666666666667e-05,
|
| 39 |
+
"loss": 0.2563,
|
| 40 |
+
"num_input_tokens_seen": 5594128,
|
| 41 |
+
"step": 16
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.041237113402061855,
|
| 45 |
+
"grad_norm": 1.46875,
|
| 46 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 47 |
+
"loss": 0.2536,
|
| 48 |
+
"num_input_tokens_seen": 6683376,
|
| 49 |
+
"step": 20
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.049484536082474224,
|
| 53 |
+
"grad_norm": 1.2421875,
|
| 54 |
+
"learning_rate": 4e-05,
|
| 55 |
+
"loss": 0.229,
|
| 56 |
+
"num_input_tokens_seen": 8030336,
|
| 57 |
+
"step": 24
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.0577319587628866,
|
| 61 |
+
"grad_norm": 1.2578125,
|
| 62 |
+
"learning_rate": 3.9992569962849926e-05,
|
| 63 |
+
"loss": 0.2212,
|
| 64 |
+
"num_input_tokens_seen": 9395728,
|
| 65 |
+
"step": 28
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.06597938144329897,
|
| 69 |
+
"grad_norm": 1.3671875,
|
| 70 |
+
"learning_rate": 3.99702853719449e-05,
|
| 71 |
+
"loss": 0.2275,
|
| 72 |
+
"num_input_tokens_seen": 10689344,
|
| 73 |
+
"step": 32
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.07422680412371134,
|
| 77 |
+
"grad_norm": 1.328125,
|
| 78 |
+
"learning_rate": 3.9933162784818745e-05,
|
| 79 |
+
"loss": 0.2262,
|
| 80 |
+
"num_input_tokens_seen": 11936704,
|
| 81 |
+
"step": 36
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.08247422680412371,
|
| 85 |
+
"grad_norm": 1.484375,
|
| 86 |
+
"learning_rate": 3.988122978369162e-05,
|
| 87 |
+
"loss": 0.2254,
|
| 88 |
+
"num_input_tokens_seen": 13217248,
|
| 89 |
+
"step": 40
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.09072164948453608,
|
| 93 |
+
"grad_norm": 1.15625,
|
| 94 |
+
"learning_rate": 3.981452495497628e-05,
|
| 95 |
+
"loss": 0.2186,
|
| 96 |
+
"num_input_tokens_seen": 14587328,
|
| 97 |
+
"step": 44
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.09896907216494845,
|
| 101 |
+
"grad_norm": 1.109375,
|
| 102 |
+
"learning_rate": 3.973309786060829e-05,
|
| 103 |
+
"loss": 0.1971,
|
| 104 |
+
"num_input_tokens_seen": 15976464,
|
| 105 |
+
"step": 48
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.10721649484536082,
|
| 109 |
+
"grad_norm": 1.125,
|
| 110 |
+
"learning_rate": 3.963700900122124e-05,
|
| 111 |
+
"loss": 0.2231,
|
| 112 |
+
"num_input_tokens_seen": 17262576,
|
| 113 |
+
"step": 52
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 0.1154639175257732,
|
| 117 |
+
"grad_norm": 0.9765625,
|
| 118 |
+
"learning_rate": 3.952632977119465e-05,
|
| 119 |
+
"loss": 0.2029,
|
| 120 |
+
"num_input_tokens_seen": 18801264,
|
| 121 |
+
"step": 56
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.12371134020618557,
|
| 125 |
+
"grad_norm": 1.0625,
|
| 126 |
+
"learning_rate": 3.9401142405607594e-05,
|
| 127 |
+
"loss": 0.2033,
|
| 128 |
+
"num_input_tokens_seen": 20158000,
|
| 129 |
+
"step": 60
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.13195876288659794,
|
| 133 |
+
"grad_norm": 1.09375,
|
| 134 |
+
"learning_rate": 3.9261539919137776e-05,
|
| 135 |
+
"loss": 0.2278,
|
| 136 |
+
"num_input_tokens_seen": 21322240,
|
| 137 |
+
"step": 64
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.1402061855670103,
|
| 141 |
+
"grad_norm": 1.0,
|
| 142 |
+
"learning_rate": 3.9107626036951266e-05,
|
| 143 |
+
"loss": 0.1998,
|
| 144 |
+
"num_input_tokens_seen": 22631360,
|
| 145 |
+
"step": 68
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.14845360824742268,
|
| 149 |
+
"grad_norm": 1.109375,
|
| 150 |
+
"learning_rate": 3.8939515117634326e-05,
|
| 151 |
+
"loss": 0.2148,
|
| 152 |
+
"num_input_tokens_seen": 23848496,
|
| 153 |
+
"step": 72
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.15670103092783505,
|
| 157 |
+
"grad_norm": 1.1484375,
|
| 158 |
+
"learning_rate": 3.875733206822452e-05,
|
| 159 |
+
"loss": 0.2246,
|
| 160 |
+
"num_input_tokens_seen": 25148336,
|
| 161 |
+
"step": 76
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.16494845360824742,
|
| 165 |
+
"grad_norm": 1.203125,
|
| 166 |
+
"learning_rate": 3.8561212251404406e-05,
|
| 167 |
+
"loss": 0.2056,
|
| 168 |
+
"num_input_tokens_seen": 26427264,
|
| 169 |
+
"step": 80
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.1731958762886598,
|
| 173 |
+
"grad_norm": 1.1328125,
|
| 174 |
+
"learning_rate": 3.835130138492644e-05,
|
| 175 |
+
"loss": 0.203,
|
| 176 |
+
"num_input_tokens_seen": 27833024,
|
| 177 |
+
"step": 84
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.18144329896907216,
|
| 181 |
+
"grad_norm": 1.1015625,
|
| 182 |
+
"learning_rate": 3.812775543334425e-05,
|
| 183 |
+
"loss": 0.1912,
|
| 184 |
+
"num_input_tokens_seen": 29273008,
|
| 185 |
+
"step": 88
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.18969072164948453,
|
| 189 |
+
"grad_norm": 1.2265625,
|
| 190 |
+
"learning_rate": 3.789074049213033e-05,
|
| 191 |
+
"loss": 0.2182,
|
| 192 |
+
"num_input_tokens_seen": 30624112,
|
| 193 |
+
"step": 92
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 0.1979381443298969,
|
| 197 |
+
"grad_norm": 1.1796875,
|
| 198 |
+
"learning_rate": 3.7640432664266514e-05,
|
| 199 |
+
"loss": 0.216,
|
| 200 |
+
"num_input_tokens_seen": 31857552,
|
| 201 |
+
"step": 96
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.20618556701030927,
|
| 205 |
+
"grad_norm": 1.125,
|
| 206 |
+
"learning_rate": 3.737701792939881e-05,
|
| 207 |
+
"loss": 0.2065,
|
| 208 |
+
"num_input_tokens_seen": 33116768,
|
| 209 |
+
"step": 100
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.21443298969072164,
|
| 213 |
+
"grad_norm": 1.0,
|
| 214 |
+
"learning_rate": 3.7100692005653796e-05,
|
| 215 |
+
"loss": 0.206,
|
| 216 |
+
"num_input_tokens_seen": 34461024,
|
| 217 |
+
"step": 104
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 0.22268041237113403,
|
| 221 |
+
"grad_norm": 1.0703125,
|
| 222 |
+
"learning_rate": 3.681166020421938e-05,
|
| 223 |
+
"loss": 0.1912,
|
| 224 |
+
"num_input_tokens_seen": 35915264,
|
| 225 |
+
"step": 108
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.2309278350515464,
|
| 229 |
+
"grad_norm": 1.0625,
|
| 230 |
+
"learning_rate": 3.6510137276797786e-05,
|
| 231 |
+
"loss": 0.1952,
|
| 232 |
+
"num_input_tokens_seen": 37264080,
|
| 233 |
+
"step": 112
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.23917525773195877,
|
| 237 |
+
"grad_norm": 1.109375,
|
| 238 |
+
"learning_rate": 3.6196347256044236e-05,
|
| 239 |
+
"loss": 0.2273,
|
| 240 |
+
"num_input_tokens_seen": 38539072,
|
| 241 |
+
"step": 116
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.24742268041237114,
|
| 245 |
+
"grad_norm": 1.109375,
|
| 246 |
+
"learning_rate": 3.5870523289109886e-05,
|
| 247 |
+
"loss": 0.2041,
|
| 248 |
+
"num_input_tokens_seen": 39930480,
|
| 249 |
+
"step": 120
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.2556701030927835,
|
| 253 |
+
"grad_norm": 1.15625,
|
| 254 |
+
"learning_rate": 3.553290746441261e-05,
|
| 255 |
+
"loss": 0.2065,
|
| 256 |
+
"num_input_tokens_seen": 41066544,
|
| 257 |
+
"step": 124
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.2639175257731959,
|
| 261 |
+
"grad_norm": 1.0390625,
|
| 262 |
+
"learning_rate": 3.5183750631764406e-05,
|
| 263 |
+
"loss": 0.1979,
|
| 264 |
+
"num_input_tokens_seen": 42372160,
|
| 265 |
+
"step": 128
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.2721649484536082,
|
| 269 |
+
"grad_norm": 1.0703125,
|
| 270 |
+
"learning_rate": 3.4823312215989046e-05,
|
| 271 |
+
"loss": 0.2079,
|
| 272 |
+
"num_input_tokens_seen": 43644832,
|
| 273 |
+
"step": 132
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.2804123711340206,
|
| 277 |
+
"grad_norm": 1.0078125,
|
| 278 |
+
"learning_rate": 3.445186002416849e-05,
|
| 279 |
+
"loss": 0.2058,
|
| 280 |
+
"num_input_tokens_seen": 44948816,
|
| 281 |
+
"step": 136
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 0.28865979381443296,
|
| 285 |
+
"grad_norm": 1.09375,
|
| 286 |
+
"learning_rate": 3.4069670046661197e-05,
|
| 287 |
+
"loss": 0.1857,
|
| 288 |
+
"num_input_tokens_seen": 46404048,
|
| 289 |
+
"step": 140
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.29690721649484536,
|
| 293 |
+
"grad_norm": 1.1015625,
|
| 294 |
+
"learning_rate": 3.3677026252040306e-05,
|
| 295 |
+
"loss": 0.212,
|
| 296 |
+
"num_input_tokens_seen": 47646208,
|
| 297 |
+
"step": 144
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.30515463917525776,
|
| 301 |
+
"grad_norm": 1.0234375,
|
| 302 |
+
"learning_rate": 3.327422037610389e-05,
|
| 303 |
+
"loss": 0.1983,
|
| 304 |
+
"num_input_tokens_seen": 49010928,
|
| 305 |
+
"step": 148
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 0.3134020618556701,
|
| 309 |
+
"grad_norm": 0.9375,
|
| 310 |
+
"learning_rate": 3.286155170511419e-05,
|
| 311 |
+
"loss": 0.197,
|
| 312 |
+
"num_input_tokens_seen": 50440128,
|
| 313 |
+
"step": 152
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.3216494845360825,
|
| 317 |
+
"grad_norm": 1.140625,
|
| 318 |
+
"learning_rate": 3.2439326853426824e-05,
|
| 319 |
+
"loss": 0.2028,
|
| 320 |
+
"num_input_tokens_seen": 51797840,
|
| 321 |
+
"step": 156
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.32989690721649484,
|
| 325 |
+
"grad_norm": 0.984375,
|
| 326 |
+
"learning_rate": 3.200785953567517e-05,
|
| 327 |
+
"loss": 0.196,
|
| 328 |
+
"num_input_tokens_seen": 53109456,
|
| 329 |
+
"step": 160
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 0.33814432989690724,
|
| 333 |
+
"grad_norm": 1.0390625,
|
| 334 |
+
"learning_rate": 3.156747033367922e-05,
|
| 335 |
+
"loss": 0.2016,
|
| 336 |
+
"num_input_tokens_seen": 54440768,
|
| 337 |
+
"step": 164
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 0.3463917525773196,
|
| 341 |
+
"grad_norm": 0.97265625,
|
| 342 |
+
"learning_rate": 3.1118486458252094e-05,
|
| 343 |
+
"loss": 0.1975,
|
| 344 |
+
"num_input_tokens_seen": 55879424,
|
| 345 |
+
"step": 168
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.354639175257732,
|
| 349 |
+
"grad_norm": 1.0234375,
|
| 350 |
+
"learning_rate": 3.0661241506081236e-05,
|
| 351 |
+
"loss": 0.1965,
|
| 352 |
+
"num_input_tokens_seen": 57154384,
|
| 353 |
+
"step": 172
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.3628865979381443,
|
| 357 |
+
"grad_norm": 0.95703125,
|
| 358 |
+
"learning_rate": 3.019607521186475e-05,
|
| 359 |
+
"loss": 0.2078,
|
| 360 |
+
"num_input_tokens_seen": 58470672,
|
| 361 |
+
"step": 176
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.3711340206185567,
|
| 365 |
+
"grad_norm": 1.0625,
|
| 366 |
+
"learning_rate": 2.972333319588736e-05,
|
| 367 |
+
"loss": 0.2092,
|
| 368 |
+
"num_input_tokens_seen": 59684416,
|
| 369 |
+
"step": 180
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.37938144329896906,
|
| 373 |
+
"grad_norm": 0.96484375,
|
| 374 |
+
"learning_rate": 2.9243366707223165e-05,
|
| 375 |
+
"loss": 0.2018,
|
| 376 |
+
"num_input_tokens_seen": 61002832,
|
| 377 |
+
"step": 184
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.38762886597938145,
|
| 381 |
+
"grad_norm": 1.0546875,
|
| 382 |
+
"learning_rate": 2.875653236275632e-05,
|
| 383 |
+
"loss": 0.2072,
|
| 384 |
+
"num_input_tokens_seen": 62262064,
|
| 385 |
+
"step": 188
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 0.3958762886597938,
|
| 389 |
+
"grad_norm": 0.92578125,
|
| 390 |
+
"learning_rate": 2.8263191882213362e-05,
|
| 391 |
+
"loss": 0.1936,
|
| 392 |
+
"num_input_tokens_seen": 63678896,
|
| 393 |
+
"step": 192
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.4041237113402062,
|
| 397 |
+
"grad_norm": 0.9609375,
|
| 398 |
+
"learning_rate": 2.7763711819404098e-05,
|
| 399 |
+
"loss": 0.2069,
|
| 400 |
+
"num_input_tokens_seen": 64844672,
|
| 401 |
+
"step": 196
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.41237113402061853,
|
| 405 |
+
"grad_norm": 1.0546875,
|
| 406 |
+
"learning_rate": 2.7258463289870764e-05,
|
| 407 |
+
"loss": 0.1924,
|
| 408 |
+
"num_input_tokens_seen": 66274544,
|
| 409 |
+
"step": 200
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.42061855670103093,
|
| 413 |
+
"grad_norm": 0.88671875,
|
| 414 |
+
"learning_rate": 2.6747821695147806e-05,
|
| 415 |
+
"loss": 0.1949,
|
| 416 |
+
"num_input_tokens_seen": 67683072,
|
| 417 |
+
"step": 204
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.4288659793814433,
|
| 421 |
+
"grad_norm": 1.125,
|
| 422 |
+
"learning_rate": 2.623216644383715e-05,
|
| 423 |
+
"loss": 0.2092,
|
| 424 |
+
"num_input_tokens_seen": 68860288,
|
| 425 |
+
"step": 208
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.43711340206185567,
|
| 429 |
+
"grad_norm": 1.078125,
|
| 430 |
+
"learning_rate": 2.5711880669706172e-05,
|
| 431 |
+
"loss": 0.1959,
|
| 432 |
+
"num_input_tokens_seen": 70182736,
|
| 433 |
+
"step": 212
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 0.44536082474226807,
|
| 437 |
+
"grad_norm": 0.9296875,
|
| 438 |
+
"learning_rate": 2.5187350947017918e-05,
|
| 439 |
+
"loss": 0.2101,
|
| 440 |
+
"num_input_tokens_seen": 71494624,
|
| 441 |
+
"step": 216
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.4536082474226804,
|
| 445 |
+
"grad_norm": 0.86328125,
|
| 446 |
+
"learning_rate": 2.4658967003304986e-05,
|
| 447 |
+
"loss": 0.1925,
|
| 448 |
+
"num_input_tokens_seen": 72877248,
|
| 449 |
+
"step": 220
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.4618556701030928,
|
| 453 |
+
"grad_norm": 1.0078125,
|
| 454 |
+
"learning_rate": 2.4127121429800498e-05,
|
| 455 |
+
"loss": 0.1841,
|
| 456 |
+
"num_input_tokens_seen": 74118560,
|
| 457 |
+
"step": 224
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.47010309278350515,
|
| 461 |
+
"grad_norm": 0.88671875,
|
| 462 |
+
"learning_rate": 2.3592209389741372e-05,
|
| 463 |
+
"loss": 0.174,
|
| 464 |
+
"num_input_tokens_seen": 75598912,
|
| 465 |
+
"step": 228
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.47835051546391755,
|
| 469 |
+
"grad_norm": 1.0234375,
|
| 470 |
+
"learning_rate": 2.30546283247606e-05,
|
| 471 |
+
"loss": 0.207,
|
| 472 |
+
"num_input_tokens_seen": 76742752,
|
| 473 |
+
"step": 232
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.4865979381443299,
|
| 477 |
+
"grad_norm": 1.015625,
|
| 478 |
+
"learning_rate": 2.251477765958655e-05,
|
| 479 |
+
"loss": 0.1932,
|
| 480 |
+
"num_input_tokens_seen": 78206256,
|
| 481 |
+
"step": 236
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 0.4948453608247423,
|
| 485 |
+
"grad_norm": 1.0546875,
|
| 486 |
+
"learning_rate": 2.1973058505269007e-05,
|
| 487 |
+
"loss": 0.1946,
|
| 488 |
+
"num_input_tokens_seen": 79491408,
|
| 489 |
+
"step": 240
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.5030927835051546,
|
| 493 |
+
"grad_norm": 1.0546875,
|
| 494 |
+
"learning_rate": 2.1429873361152124e-05,
|
| 495 |
+
"loss": 0.1975,
|
| 496 |
+
"num_input_tokens_seen": 80718320,
|
| 497 |
+
"step": 244
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 0.511340206185567,
|
| 501 |
+
"grad_norm": 0.91015625,
|
| 502 |
+
"learning_rate": 2.088562581581592e-05,
|
| 503 |
+
"loss": 0.1964,
|
| 504 |
+
"num_input_tokens_seen": 81915456,
|
| 505 |
+
"step": 248
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.5195876288659794,
|
| 509 |
+
"grad_norm": 1.1015625,
|
| 510 |
+
"learning_rate": 2.0340720247208447e-05,
|
| 511 |
+
"loss": 0.191,
|
| 512 |
+
"num_input_tokens_seen": 83180624,
|
| 513 |
+
"step": 252
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.5278350515463918,
|
| 517 |
+
"grad_norm": 0.90234375,
|
| 518 |
+
"learning_rate": 1.9795561522191523e-05,
|
| 519 |
+
"loss": 0.1832,
|
| 520 |
+
"num_input_tokens_seen": 84571536,
|
| 521 |
+
"step": 256
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.5360824742268041,
|
| 525 |
+
"grad_norm": 1.046875,
|
| 526 |
+
"learning_rate": 1.9250554695723107e-05,
|
| 527 |
+
"loss": 0.1964,
|
| 528 |
+
"num_input_tokens_seen": 85841328,
|
| 529 |
+
"step": 260
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 0.5443298969072164,
|
| 533 |
+
"grad_norm": 0.953125,
|
| 534 |
+
"learning_rate": 1.8706104709899964e-05,
|
| 535 |
+
"loss": 0.1875,
|
| 536 |
+
"num_input_tokens_seen": 87241616,
|
| 537 |
+
"step": 264
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.5525773195876289,
|
| 541 |
+
"grad_norm": 0.92578125,
|
| 542 |
+
"learning_rate": 1.816261609308419e-05,
|
| 543 |
+
"loss": 0.1809,
|
| 544 |
+
"num_input_tokens_seen": 88600352,
|
| 545 |
+
"step": 268
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 0.5608247422680412,
|
| 549 |
+
"grad_norm": 0.87109375,
|
| 550 |
+
"learning_rate": 1.7620492659337155e-05,
|
| 551 |
+
"loss": 0.1793,
|
| 552 |
+
"num_input_tokens_seen": 90051376,
|
| 553 |
+
"step": 272
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 0.5690721649484536,
|
| 557 |
+
"grad_norm": 1.0390625,
|
| 558 |
+
"learning_rate": 1.7080137208384122e-05,
|
| 559 |
+
"loss": 0.1865,
|
| 560 |
+
"num_input_tokens_seen": 91429472,
|
| 561 |
+
"step": 276
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.5773195876288659,
|
| 565 |
+
"grad_norm": 0.9140625,
|
| 566 |
+
"learning_rate": 1.6541951226332565e-05,
|
| 567 |
+
"loss": 0.1745,
|
| 568 |
+
"num_input_tokens_seen": 92791856,
|
| 569 |
+
"step": 280
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.5855670103092784,
|
| 573 |
+
"grad_norm": 0.875,
|
| 574 |
+
"learning_rate": 1.600633458736653e-05,
|
| 575 |
+
"loss": 0.1925,
|
| 576 |
+
"num_input_tokens_seen": 94068304,
|
| 577 |
+
"step": 284
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.5938144329896907,
|
| 581 |
+
"grad_norm": 0.98828125,
|
| 582 |
+
"learning_rate": 1.5473685256638572e-05,
|
| 583 |
+
"loss": 0.1903,
|
| 584 |
+
"num_input_tokens_seen": 95338656,
|
| 585 |
+
"step": 288
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.6020618556701031,
|
| 589 |
+
"grad_norm": 0.90625,
|
| 590 |
+
"learning_rate": 1.4944398994580232e-05,
|
| 591 |
+
"loss": 0.1834,
|
| 592 |
+
"num_input_tokens_seen": 96565872,
|
| 593 |
+
"step": 292
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.6103092783505155,
|
| 597 |
+
"grad_norm": 1.0,
|
| 598 |
+
"learning_rate": 1.4418869062850514e-05,
|
| 599 |
+
"loss": 0.211,
|
| 600 |
+
"num_input_tokens_seen": 97845776,
|
| 601 |
+
"step": 296
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 0.6185567010309279,
|
| 605 |
+
"grad_norm": 0.92578125,
|
| 606 |
+
"learning_rate": 1.3897485932141042e-05,
|
| 607 |
+
"loss": 0.1872,
|
| 608 |
+
"num_input_tokens_seen": 99080048,
|
| 609 |
+
"step": 300
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 0.6268041237113402,
|
| 613 |
+
"grad_norm": 0.88671875,
|
| 614 |
+
"learning_rate": 1.3380636992054878e-05,
|
| 615 |
+
"loss": 0.17,
|
| 616 |
+
"num_input_tokens_seen": 100563184,
|
| 617 |
+
"step": 304
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 0.6350515463917525,
|
| 621 |
+
"grad_norm": 0.9375,
|
| 622 |
+
"learning_rate": 1.2868706263274602e-05,
|
| 623 |
+
"loss": 0.1935,
|
| 624 |
+
"num_input_tokens_seen": 101820432,
|
| 625 |
+
"step": 308
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.643298969072165,
|
| 629 |
+
"grad_norm": 0.88671875,
|
| 630 |
+
"learning_rate": 1.236207411223353e-05,
|
| 631 |
+
"loss": 0.1833,
|
| 632 |
+
"num_input_tokens_seen": 103280736,
|
| 633 |
+
"step": 312
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.6515463917525773,
|
| 637 |
+
"grad_norm": 0.88671875,
|
| 638 |
+
"learning_rate": 1.1861116968502015e-05,
|
| 639 |
+
"loss": 0.1815,
|
| 640 |
+
"num_input_tokens_seen": 104563920,
|
| 641 |
+
"step": 316
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 0.6597938144329897,
|
| 645 |
+
"grad_norm": 0.875,
|
| 646 |
+
"learning_rate": 1.136620704509892e-05,
|
| 647 |
+
"loss": 0.1816,
|
| 648 |
+
"num_input_tokens_seen": 105869408,
|
| 649 |
+
"step": 320
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 0.668041237113402,
|
| 653 |
+
"grad_norm": 0.9296875,
|
| 654 |
+
"learning_rate": 1.087771206193593e-05,
|
| 655 |
+
"loss": 0.1837,
|
| 656 |
+
"num_input_tokens_seen": 107213792,
|
| 657 |
+
"step": 324
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 0.6762886597938145,
|
| 661 |
+
"grad_norm": 0.96484375,
|
| 662 |
+
"learning_rate": 1.0395994972600285e-05,
|
| 663 |
+
"loss": 0.1775,
|
| 664 |
+
"num_input_tokens_seen": 108623536,
|
| 665 |
+
"step": 328
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 0.6845360824742268,
|
| 669 |
+
"grad_norm": 0.97265625,
|
| 670 |
+
"learning_rate": 9.921413694678959e-06,
|
| 671 |
+
"loss": 0.2035,
|
| 672 |
+
"num_input_tokens_seen": 109750560,
|
| 673 |
+
"step": 332
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 0.6927835051546392,
|
| 677 |
+
"grad_norm": 0.921875,
|
| 678 |
+
"learning_rate": 9.454320843824512e-06,
|
| 679 |
+
"loss": 0.1862,
|
| 680 |
+
"num_input_tokens_seen": 111023152,
|
| 681 |
+
"step": 336
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.7010309278350515,
|
| 685 |
+
"grad_norm": 0.95703125,
|
| 686 |
+
"learning_rate": 8.995063471760377e-06,
|
| 687 |
+
"loss": 0.1927,
|
| 688 |
+
"num_input_tokens_seen": 112284320,
|
| 689 |
+
"step": 340
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.709278350515464,
|
| 693 |
+
"grad_norm": 0.96484375,
|
| 694 |
+
"learning_rate": 8.543982808420156e-06,
|
| 695 |
+
"loss": 0.1856,
|
| 696 |
+
"num_input_tokens_seen": 113630688,
|
| 697 |
+
"step": 344
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 0.7175257731958763,
|
| 701 |
+
"grad_norm": 0.953125,
|
| 702 |
+
"learning_rate": 8.101414008412469e-06,
|
| 703 |
+
"loss": 0.1792,
|
| 704 |
+
"num_input_tokens_seen": 114946320,
|
| 705 |
+
"step": 348
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.7257731958762886,
|
| 709 |
+
"grad_norm": 0.94921875,
|
| 710 |
+
"learning_rate": 7.667685901999875e-06,
|
| 711 |
+
"loss": 0.1891,
|
| 712 |
+
"num_input_tokens_seen": 116220208,
|
| 713 |
+
"step": 352
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.734020618556701,
|
| 717 |
+
"grad_norm": 0.8671875,
|
| 718 |
+
"learning_rate": 7.24312075077674e-06,
|
| 719 |
+
"loss": 0.1891,
|
| 720 |
+
"num_input_tokens_seen": 117614672,
|
| 721 |
+
"step": 356
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 0.7422680412371134,
|
| 725 |
+
"grad_norm": 1.0859375,
|
| 726 |
+
"learning_rate": 6.828034008227678e-06,
|
| 727 |
+
"loss": 0.1714,
|
| 728 |
+
"num_input_tokens_seen": 118996816,
|
| 729 |
+
"step": 360
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.7505154639175258,
|
| 733 |
+
"grad_norm": 0.90234375,
|
| 734 |
+
"learning_rate": 6.422734085344464e-06,
|
| 735 |
+
"loss": 0.1871,
|
| 736 |
+
"num_input_tokens_seen": 120229232,
|
| 737 |
+
"step": 364
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.7587628865979381,
|
| 741 |
+
"grad_norm": 0.8203125,
|
| 742 |
+
"learning_rate": 6.027522121475482e-06,
|
| 743 |
+
"loss": 0.1795,
|
| 744 |
+
"num_input_tokens_seen": 121495936,
|
| 745 |
+
"step": 368
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.7670103092783506,
|
| 749 |
+
"grad_norm": 0.95703125,
|
| 750 |
+
"learning_rate": 5.642691760578116e-06,
|
| 751 |
+
"loss": 0.1833,
|
| 752 |
+
"num_input_tokens_seen": 122787872,
|
| 753 |
+
"step": 372
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.7752577319587629,
|
| 757 |
+
"grad_norm": 0.83984375,
|
| 758 |
+
"learning_rate": 5.268528933040147e-06,
|
| 759 |
+
"loss": 0.1673,
|
| 760 |
+
"num_input_tokens_seen": 124257600,
|
| 761 |
+
"step": 376
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.7835051546391752,
|
| 765 |
+
"grad_norm": 0.875,
|
| 766 |
+
"learning_rate": 4.905311643232464e-06,
|
| 767 |
+
"loss": 0.1763,
|
| 768 |
+
"num_input_tokens_seen": 125705408,
|
| 769 |
+
"step": 380
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 0.7917525773195876,
|
| 773 |
+
"grad_norm": 0.9765625,
|
| 774 |
+
"learning_rate": 4.553309762950739e-06,
|
| 775 |
+
"loss": 0.1877,
|
| 776 |
+
"num_input_tokens_seen": 126862272,
|
| 777 |
+
"step": 384
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 0.8,
|
| 781 |
+
"grad_norm": 0.90625,
|
| 782 |
+
"learning_rate": 4.212784830899725e-06,
|
| 783 |
+
"loss": 0.1795,
|
| 784 |
+
"num_input_tokens_seen": 128153600,
|
| 785 |
+
"step": 388
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 0.8082474226804124,
|
| 789 |
+
"grad_norm": 0.828125,
|
| 790 |
+
"learning_rate": 3.8839898583689725e-06,
|
| 791 |
+
"loss": 0.1803,
|
| 792 |
+
"num_input_tokens_seen": 129461872,
|
| 793 |
+
"step": 392
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.8164948453608247,
|
| 797 |
+
"grad_norm": 0.88671875,
|
| 798 |
+
"learning_rate": 3.567169141244562e-06,
|
| 799 |
+
"loss": 0.179,
|
| 800 |
+
"num_input_tokens_seen": 130662064,
|
| 801 |
+
"step": 396
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.8247422680412371,
|
| 805 |
+
"grad_norm": 0.8359375,
|
| 806 |
+
"learning_rate": 3.262558078496301e-06,
|
| 807 |
+
"loss": 0.1679,
|
| 808 |
+
"num_input_tokens_seen": 131997568,
|
| 809 |
+
"step": 400
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 0.8329896907216495,
|
| 813 |
+
"grad_norm": 0.94921875,
|
| 814 |
+
"learning_rate": 2.9703829972754407e-06,
|
| 815 |
+
"loss": 0.1974,
|
| 816 |
+
"num_input_tokens_seen": 133415744,
|
| 817 |
+
"step": 404
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 0.8412371134020619,
|
| 821 |
+
"grad_norm": 0.828125,
|
| 822 |
+
"learning_rate": 2.69086098475277e-06,
|
| 823 |
+
"loss": 0.1699,
|
| 824 |
+
"num_input_tokens_seen": 134815840,
|
| 825 |
+
"step": 408
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 0.8494845360824742,
|
| 829 |
+
"grad_norm": 0.9921875,
|
| 830 |
+
"learning_rate": 2.4241997268220096e-06,
|
| 831 |
+
"loss": 0.1815,
|
| 832 |
+
"num_input_tokens_seen": 136262128,
|
| 833 |
+
"step": 412
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 0.8577319587628865,
|
| 837 |
+
"grad_norm": 0.96484375,
|
| 838 |
+
"learning_rate": 2.1705973537884615e-06,
|
| 839 |
+
"loss": 0.1781,
|
| 840 |
+
"num_input_tokens_seen": 137430160,
|
| 841 |
+
"step": 416
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 0.865979381443299,
|
| 845 |
+
"grad_norm": 0.95703125,
|
| 846 |
+
"learning_rate": 1.9302422931574183e-06,
|
| 847 |
+
"loss": 0.1899,
|
| 848 |
+
"num_input_tokens_seen": 138709200,
|
| 849 |
+
"step": 420
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.8742268041237113,
|
| 853 |
+
"grad_norm": 0.83984375,
|
| 854 |
+
"learning_rate": 1.7033131296318473e-06,
|
| 855 |
+
"loss": 0.1795,
|
| 856 |
+
"num_input_tokens_seen": 140033680,
|
| 857 |
+
"step": 424
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.8824742268041237,
|
| 861 |
+
"grad_norm": 0.8203125,
|
| 862 |
+
"learning_rate": 1.4899784724232968e-06,
|
| 863 |
+
"loss": 0.1749,
|
| 864 |
+
"num_input_tokens_seen": 141348848,
|
| 865 |
+
"step": 428
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 0.8907216494845361,
|
| 869 |
+
"grad_norm": 0.83984375,
|
| 870 |
+
"learning_rate": 1.2903968299746094e-06,
|
| 871 |
+
"loss": 0.171,
|
| 872 |
+
"num_input_tokens_seen": 142797664,
|
| 873 |
+
"step": 432
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.8989690721649485,
|
| 877 |
+
"grad_norm": 0.89453125,
|
| 878 |
+
"learning_rate": 1.104716492187574e-06,
|
| 879 |
+
"loss": 0.1812,
|
| 880 |
+
"num_input_tokens_seen": 144154208,
|
| 881 |
+
"step": 436
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 0.9072164948453608,
|
| 885 |
+
"grad_norm": 1.03125,
|
| 886 |
+
"learning_rate": 9.330754202429726e-07,
|
| 887 |
+
"loss": 0.1891,
|
| 888 |
+
"num_input_tokens_seen": 145332560,
|
| 889 |
+
"step": 440
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 0.9154639175257732,
|
| 893 |
+
"grad_norm": 0.98046875,
|
| 894 |
+
"learning_rate": 7.756011440948996e-07,
|
| 895 |
+
"loss": 0.1902,
|
| 896 |
+
"num_input_tokens_seen": 146527344,
|
| 897 |
+
"step": 444
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.9237113402061856,
|
| 901 |
+
"grad_norm": 1.0,
|
| 902 |
+
"learning_rate": 6.324106677155573e-07,
|
| 903 |
+
"loss": 0.1793,
|
| 904 |
+
"num_input_tokens_seen": 147821568,
|
| 905 |
+
"step": 448
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.931958762886598,
|
| 909 |
+
"grad_norm": 0.88671875,
|
| 910 |
+
"learning_rate": 5.036103821608485e-07,
|
| 911 |
+
"loss": 0.1844,
|
| 912 |
+
"num_input_tokens_seen": 149191664,
|
| 913 |
+
"step": 452
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.9402061855670103,
|
| 917 |
+
"grad_norm": 0.8515625,
|
| 918 |
+
"learning_rate": 3.892959865214363e-07,
|
| 919 |
+
"loss": 0.1795,
|
| 920 |
+
"num_input_tokens_seen": 150526864,
|
| 921 |
+
"step": 456
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 0.9484536082474226,
|
| 925 |
+
"grad_norm": 0.9453125,
|
| 926 |
+
"learning_rate": 2.8955241681795534e-07,
|
| 927 |
+
"loss": 0.1859,
|
| 928 |
+
"num_input_tokens_seen": 151861952,
|
| 929 |
+
"step": 460
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 0.9567010309278351,
|
| 933 |
+
"grad_norm": 0.71875,
|
| 934 |
+
"learning_rate": 2.044537828932458e-07,
|
| 935 |
+
"loss": 0.1787,
|
| 936 |
+
"num_input_tokens_seen": 153201488,
|
| 937 |
+
"step": 464
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 0.9649484536082474,
|
| 941 |
+
"grad_norm": 0.87109375,
|
| 942 |
+
"learning_rate": 1.3406331334845813e-07,
|
| 943 |
+
"loss": 0.1884,
|
| 944 |
+
"num_input_tokens_seen": 154511184,
|
| 945 |
+
"step": 468
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 0.9731958762886598,
|
| 949 |
+
"grad_norm": 0.796875,
|
| 950 |
+
"learning_rate": 7.843330856396103e-08,
|
| 951 |
+
"loss": 0.1858,
|
| 952 |
+
"num_input_tokens_seen": 155737200,
|
| 953 |
+
"step": 472
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 0.9814432989690721,
|
| 957 |
+
"grad_norm": 0.8671875,
|
| 958 |
+
"learning_rate": 3.760510183997701e-08,
|
| 959 |
+
"loss": 0.183,
|
| 960 |
+
"num_input_tokens_seen": 157084960,
|
| 961 |
+
"step": 476
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.9896907216494846,
|
| 965 |
+
"grad_norm": 0.93359375,
|
| 966 |
+
"learning_rate": 1.160902868577951e-08,
|
| 967 |
+
"loss": 0.1908,
|
| 968 |
+
"num_input_tokens_seen": 158349904,
|
| 969 |
+
"step": 480
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.9979381443298969,
|
| 973 |
+
"grad_norm": 0.8984375,
|
| 974 |
+
"learning_rate": 4.64404280295927e-10,
|
| 975 |
+
"loss": 0.1764,
|
| 976 |
+
"num_input_tokens_seen": 159669152,
|
| 977 |
+
"step": 484
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 1.0,
|
| 981 |
+
"eval_loss": 0.0952233299612999,
|
| 982 |
+
"eval_runtime": 83.4314,
|
| 983 |
+
"eval_samples_per_second": 12.453,
|
| 984 |
+
"eval_steps_per_second": 0.396,
|
| 985 |
+
"num_input_tokens_seen": 160041824,
|
| 986 |
+
"step": 485
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 1.0,
|
| 990 |
+
"num_input_tokens_seen": 160041824,
|
| 991 |
+
"step": 485,
|
| 992 |
+
"total_flos": 9.01334912177537e+17,
|
| 993 |
+
"train_loss": 0.20323730206366666,
|
| 994 |
+
"train_runtime": 14397.8817,
|
| 995 |
+
"train_samples_per_second": 4.306,
|
| 996 |
+
"train_steps_per_second": 0.034,
|
| 997 |
+
"train_tokens_per_second": 1386.226
|
| 998 |
+
}
|
| 999 |
+
],
|
| 1000 |
+
"logging_steps": 4,
|
| 1001 |
+
"max_steps": 485,
|
| 1002 |
+
"num_input_tokens_seen": 160041824,
|
| 1003 |
+
"num_train_epochs": 1,
|
| 1004 |
+
"save_steps": 0,
|
| 1005 |
+
"stateful_callbacks": {
|
| 1006 |
+
"TrainerControl": {
|
| 1007 |
+
"args": {
|
| 1008 |
+
"should_epoch_stop": false,
|
| 1009 |
+
"should_evaluate": false,
|
| 1010 |
+
"should_log": false,
|
| 1011 |
+
"should_save": false,
|
| 1012 |
+
"should_training_stop": false
|
| 1013 |
+
},
|
| 1014 |
+
"attributes": {}
|
| 1015 |
+
}
|
| 1016 |
+
},
|
| 1017 |
+
"total_flos": 9.01334912177537e+17,
|
| 1018 |
+
"train_batch_size": 16,
|
| 1019 |
+
"trial_name": null,
|
| 1020 |
+
"trial_params": null
|
| 1021 |
+
}
|