Model save
Browse files- all_results.json +9 -0
- train_results.json +9 -0
- trainer_state.json +967 -0
- training_args.bin +1 -1
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.0,
|
| 3 |
+
"total_flos": 1.1159654864928768e+16,
|
| 4 |
+
"train_loss": 1.1755467142439282,
|
| 5 |
+
"train_runtime": 8495.7942,
|
| 6 |
+
"train_samples": 397452,
|
| 7 |
+
"train_samples_per_second": 93.564,
|
| 8 |
+
"train_steps_per_second": 7.797
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.0,
|
| 3 |
+
"total_flos": 1.1159654864928768e+16,
|
| 4 |
+
"train_loss": 1.1755467142439282,
|
| 5 |
+
"train_runtime": 8495.7942,
|
| 6 |
+
"train_samples": 397452,
|
| 7 |
+
"train_samples_per_second": 93.564,
|
| 8 |
+
"train_steps_per_second": 7.797
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,967 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 66242,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.015096162555478397,
|
| 14 |
+
"grad_norm": 7.316836833953857,
|
| 15 |
+
"learning_rate": 9.924670148848163e-05,
|
| 16 |
+
"loss": 3.4381,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.030192325110956795,
|
| 21 |
+
"grad_norm": 7.617539882659912,
|
| 22 |
+
"learning_rate": 9.84918933607077e-05,
|
| 23 |
+
"loss": 3.0248,
|
| 24 |
+
"step": 1000
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.045288487666435194,
|
| 28 |
+
"grad_norm": 14.377625465393066,
|
| 29 |
+
"learning_rate": 9.773708523293379e-05,
|
| 30 |
+
"loss": 2.7697,
|
| 31 |
+
"step": 1500
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.06038465022191359,
|
| 35 |
+
"grad_norm": 8.488965034484863,
|
| 36 |
+
"learning_rate": 9.698227710515986e-05,
|
| 37 |
+
"loss": 2.5475,
|
| 38 |
+
"step": 2000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.07548081277739199,
|
| 42 |
+
"grad_norm": 10.39892292022705,
|
| 43 |
+
"learning_rate": 9.622746897738595e-05,
|
| 44 |
+
"loss": 2.4385,
|
| 45 |
+
"step": 2500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.09057697533287039,
|
| 49 |
+
"grad_norm": 11.845101356506348,
|
| 50 |
+
"learning_rate": 9.547266084961202e-05,
|
| 51 |
+
"loss": 2.2998,
|
| 52 |
+
"step": 3000
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.10567313788834878,
|
| 56 |
+
"grad_norm": 10.765714645385742,
|
| 57 |
+
"learning_rate": 9.471785272183812e-05,
|
| 58 |
+
"loss": 2.2303,
|
| 59 |
+
"step": 3500
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.12076930044382718,
|
| 63 |
+
"grad_norm": 13.092281341552734,
|
| 64 |
+
"learning_rate": 9.39630445940642e-05,
|
| 65 |
+
"loss": 2.1798,
|
| 66 |
+
"step": 4000
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.13586546299930558,
|
| 70 |
+
"grad_norm": 14.106061935424805,
|
| 71 |
+
"learning_rate": 9.320823646629028e-05,
|
| 72 |
+
"loss": 2.1111,
|
| 73 |
+
"step": 4500
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.15096162555478398,
|
| 77 |
+
"grad_norm": 10.392507553100586,
|
| 78 |
+
"learning_rate": 9.245342833851635e-05,
|
| 79 |
+
"loss": 2.1069,
|
| 80 |
+
"step": 5000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.16605778811026237,
|
| 84 |
+
"grad_norm": 9.839272499084473,
|
| 85 |
+
"learning_rate": 9.169862021074244e-05,
|
| 86 |
+
"loss": 2.0419,
|
| 87 |
+
"step": 5500
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.18115395066574078,
|
| 91 |
+
"grad_norm": 8.988909721374512,
|
| 92 |
+
"learning_rate": 9.094381208296851e-05,
|
| 93 |
+
"loss": 1.998,
|
| 94 |
+
"step": 6000
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.19625011322121916,
|
| 98 |
+
"grad_norm": 10.165138244628906,
|
| 99 |
+
"learning_rate": 9.01890039551946e-05,
|
| 100 |
+
"loss": 2.0117,
|
| 101 |
+
"step": 6500
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.21134627577669757,
|
| 105 |
+
"grad_norm": 6.843221187591553,
|
| 106 |
+
"learning_rate": 8.943419582742067e-05,
|
| 107 |
+
"loss": 1.9208,
|
| 108 |
+
"step": 7000
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.22644243833217595,
|
| 112 |
+
"grad_norm": 9.529773712158203,
|
| 113 |
+
"learning_rate": 8.867938769964676e-05,
|
| 114 |
+
"loss": 1.8786,
|
| 115 |
+
"step": 7500
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.24153860088765436,
|
| 119 |
+
"grad_norm": 10.685002326965332,
|
| 120 |
+
"learning_rate": 8.792457957187283e-05,
|
| 121 |
+
"loss": 1.8579,
|
| 122 |
+
"step": 8000
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.25663476344313274,
|
| 126 |
+
"grad_norm": 13.080573081970215,
|
| 127 |
+
"learning_rate": 8.716977144409891e-05,
|
| 128 |
+
"loss": 1.813,
|
| 129 |
+
"step": 8500
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.27173092599861115,
|
| 133 |
+
"grad_norm": 8.630938529968262,
|
| 134 |
+
"learning_rate": 8.641496331632499e-05,
|
| 135 |
+
"loss": 1.786,
|
| 136 |
+
"step": 9000
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.28682708855408956,
|
| 140 |
+
"grad_norm": 14.508255958557129,
|
| 141 |
+
"learning_rate": 8.566015518855107e-05,
|
| 142 |
+
"loss": 1.7593,
|
| 143 |
+
"step": 9500
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.30192325110956797,
|
| 147 |
+
"grad_norm": 16.67688751220703,
|
| 148 |
+
"learning_rate": 8.490534706077715e-05,
|
| 149 |
+
"loss": 1.7348,
|
| 150 |
+
"step": 10000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.3170194136650463,
|
| 154 |
+
"grad_norm": 9.229181289672852,
|
| 155 |
+
"learning_rate": 8.415053893300323e-05,
|
| 156 |
+
"loss": 1.7211,
|
| 157 |
+
"step": 10500
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.33211557622052473,
|
| 161 |
+
"grad_norm": 10.458921432495117,
|
| 162 |
+
"learning_rate": 8.339573080522932e-05,
|
| 163 |
+
"loss": 1.6878,
|
| 164 |
+
"step": 11000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.34721173877600314,
|
| 168 |
+
"grad_norm": 15.979294776916504,
|
| 169 |
+
"learning_rate": 8.264092267745539e-05,
|
| 170 |
+
"loss": 1.6594,
|
| 171 |
+
"step": 11500
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.36230790133148155,
|
| 175 |
+
"grad_norm": 10.298636436462402,
|
| 176 |
+
"learning_rate": 8.188611454968148e-05,
|
| 177 |
+
"loss": 1.6486,
|
| 178 |
+
"step": 12000
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.37740406388695996,
|
| 182 |
+
"grad_norm": 9.108405113220215,
|
| 183 |
+
"learning_rate": 8.113130642190756e-05,
|
| 184 |
+
"loss": 1.5902,
|
| 185 |
+
"step": 12500
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.3925002264424383,
|
| 189 |
+
"grad_norm": 10.374357223510742,
|
| 190 |
+
"learning_rate": 8.037649829413363e-05,
|
| 191 |
+
"loss": 1.5834,
|
| 192 |
+
"step": 13000
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.4075963889979167,
|
| 196 |
+
"grad_norm": 13.339021682739258,
|
| 197 |
+
"learning_rate": 7.962169016635972e-05,
|
| 198 |
+
"loss": 1.5589,
|
| 199 |
+
"step": 13500
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.42269255155339513,
|
| 203 |
+
"grad_norm": 10.74330997467041,
|
| 204 |
+
"learning_rate": 7.886688203858579e-05,
|
| 205 |
+
"loss": 1.5469,
|
| 206 |
+
"step": 14000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.43778871410887354,
|
| 210 |
+
"grad_norm": 10.266663551330566,
|
| 211 |
+
"learning_rate": 7.811207391081188e-05,
|
| 212 |
+
"loss": 1.5098,
|
| 213 |
+
"step": 14500
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.4528848766643519,
|
| 217 |
+
"grad_norm": 12.610342025756836,
|
| 218 |
+
"learning_rate": 7.735726578303795e-05,
|
| 219 |
+
"loss": 1.4979,
|
| 220 |
+
"step": 15000
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.4679810392198303,
|
| 224 |
+
"grad_norm": 11.99213695526123,
|
| 225 |
+
"learning_rate": 7.660245765526404e-05,
|
| 226 |
+
"loss": 1.4389,
|
| 227 |
+
"step": 15500
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.4830772017753087,
|
| 231 |
+
"grad_norm": 12.106096267700195,
|
| 232 |
+
"learning_rate": 7.584764952749012e-05,
|
| 233 |
+
"loss": 1.4559,
|
| 234 |
+
"step": 16000
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.4981733643307871,
|
| 238 |
+
"grad_norm": 16.475229263305664,
|
| 239 |
+
"learning_rate": 7.50928413997162e-05,
|
| 240 |
+
"loss": 1.3956,
|
| 241 |
+
"step": 16500
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.5132695268862655,
|
| 245 |
+
"grad_norm": 12.459781646728516,
|
| 246 |
+
"learning_rate": 7.433803327194228e-05,
|
| 247 |
+
"loss": 1.3987,
|
| 248 |
+
"step": 17000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.5283656894417439,
|
| 252 |
+
"grad_norm": 16.59553337097168,
|
| 253 |
+
"learning_rate": 7.358322514416835e-05,
|
| 254 |
+
"loss": 1.3747,
|
| 255 |
+
"step": 17500
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.5434618519972223,
|
| 259 |
+
"grad_norm": 12.136076927185059,
|
| 260 |
+
"learning_rate": 7.282841701639444e-05,
|
| 261 |
+
"loss": 1.3639,
|
| 262 |
+
"step": 18000
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.5585580145527007,
|
| 266 |
+
"grad_norm": 16.129152297973633,
|
| 267 |
+
"learning_rate": 7.207360888862051e-05,
|
| 268 |
+
"loss": 1.3646,
|
| 269 |
+
"step": 18500
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.5736541771081791,
|
| 273 |
+
"grad_norm": 11.556707382202148,
|
| 274 |
+
"learning_rate": 7.13188007608466e-05,
|
| 275 |
+
"loss": 1.3189,
|
| 276 |
+
"step": 19000
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.5887503396636575,
|
| 280 |
+
"grad_norm": 15.598160743713379,
|
| 281 |
+
"learning_rate": 7.056399263307267e-05,
|
| 282 |
+
"loss": 1.3361,
|
| 283 |
+
"step": 19500
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.6038465022191359,
|
| 287 |
+
"grad_norm": 15.071793556213379,
|
| 288 |
+
"learning_rate": 6.980918450529876e-05,
|
| 289 |
+
"loss": 1.3031,
|
| 290 |
+
"step": 20000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.6189426647746143,
|
| 294 |
+
"grad_norm": 16.57260513305664,
|
| 295 |
+
"learning_rate": 6.905437637752483e-05,
|
| 296 |
+
"loss": 1.3158,
|
| 297 |
+
"step": 20500
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.6340388273300926,
|
| 301 |
+
"grad_norm": 12.095833778381348,
|
| 302 |
+
"learning_rate": 6.829956824975092e-05,
|
| 303 |
+
"loss": 1.2785,
|
| 304 |
+
"step": 21000
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.6491349898855711,
|
| 308 |
+
"grad_norm": 15.220791816711426,
|
| 309 |
+
"learning_rate": 6.7544760121977e-05,
|
| 310 |
+
"loss": 1.2843,
|
| 311 |
+
"step": 21500
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.6642311524410495,
|
| 315 |
+
"grad_norm": 14.36874008178711,
|
| 316 |
+
"learning_rate": 6.678995199420309e-05,
|
| 317 |
+
"loss": 1.2265,
|
| 318 |
+
"step": 22000
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.6793273149965279,
|
| 322 |
+
"grad_norm": 7.681280612945557,
|
| 323 |
+
"learning_rate": 6.603514386642916e-05,
|
| 324 |
+
"loss": 1.2018,
|
| 325 |
+
"step": 22500
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.6944234775520063,
|
| 329 |
+
"grad_norm": 22.970266342163086,
|
| 330 |
+
"learning_rate": 6.528033573865525e-05,
|
| 331 |
+
"loss": 1.209,
|
| 332 |
+
"step": 23000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.7095196401074847,
|
| 336 |
+
"grad_norm": 16.40315055847168,
|
| 337 |
+
"learning_rate": 6.452552761088132e-05,
|
| 338 |
+
"loss": 1.196,
|
| 339 |
+
"step": 23500
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.7246158026629631,
|
| 343 |
+
"grad_norm": 9.259981155395508,
|
| 344 |
+
"learning_rate": 6.37707194831074e-05,
|
| 345 |
+
"loss": 1.2396,
|
| 346 |
+
"step": 24000
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.7397119652184415,
|
| 350 |
+
"grad_norm": 13.87867546081543,
|
| 351 |
+
"learning_rate": 6.301591135533348e-05,
|
| 352 |
+
"loss": 1.1755,
|
| 353 |
+
"step": 24500
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.7548081277739199,
|
| 357 |
+
"grad_norm": 19.097341537475586,
|
| 358 |
+
"learning_rate": 6.226110322755956e-05,
|
| 359 |
+
"loss": 1.1719,
|
| 360 |
+
"step": 25000
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.7699042903293982,
|
| 364 |
+
"grad_norm": 15.854406356811523,
|
| 365 |
+
"learning_rate": 6.150629509978564e-05,
|
| 366 |
+
"loss": 1.1554,
|
| 367 |
+
"step": 25500
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.7850004528848766,
|
| 371 |
+
"grad_norm": 11.38527774810791,
|
| 372 |
+
"learning_rate": 6.0751486972011714e-05,
|
| 373 |
+
"loss": 1.1181,
|
| 374 |
+
"step": 26000
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.800096615440355,
|
| 378 |
+
"grad_norm": 9.802115440368652,
|
| 379 |
+
"learning_rate": 5.9996678844237794e-05,
|
| 380 |
+
"loss": 1.1336,
|
| 381 |
+
"step": 26500
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.8151927779958335,
|
| 385 |
+
"grad_norm": 11.467633247375488,
|
| 386 |
+
"learning_rate": 5.924187071646388e-05,
|
| 387 |
+
"loss": 1.0586,
|
| 388 |
+
"step": 27000
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.8302889405513119,
|
| 392 |
+
"grad_norm": 7.793050289154053,
|
| 393 |
+
"learning_rate": 5.848706258868996e-05,
|
| 394 |
+
"loss": 1.1079,
|
| 395 |
+
"step": 27500
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.8453851031067903,
|
| 399 |
+
"grad_norm": 16.90721893310547,
|
| 400 |
+
"learning_rate": 5.773225446091604e-05,
|
| 401 |
+
"loss": 1.0793,
|
| 402 |
+
"step": 28000
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.8604812656622687,
|
| 406 |
+
"grad_norm": 8.10334587097168,
|
| 407 |
+
"learning_rate": 5.697744633314212e-05,
|
| 408 |
+
"loss": 1.0968,
|
| 409 |
+
"step": 28500
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.8755774282177471,
|
| 413 |
+
"grad_norm": 12.179634094238281,
|
| 414 |
+
"learning_rate": 5.6222638205368197e-05,
|
| 415 |
+
"loss": 1.0717,
|
| 416 |
+
"step": 29000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.8906735907732255,
|
| 420 |
+
"grad_norm": 10.544957160949707,
|
| 421 |
+
"learning_rate": 5.5467830077594276e-05,
|
| 422 |
+
"loss": 1.0724,
|
| 423 |
+
"step": 29500
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.9057697533287038,
|
| 427 |
+
"grad_norm": 16.116649627685547,
|
| 428 |
+
"learning_rate": 5.4713021949820355e-05,
|
| 429 |
+
"loss": 1.0402,
|
| 430 |
+
"step": 30000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.9208659158841822,
|
| 434 |
+
"grad_norm": 14.845929145812988,
|
| 435 |
+
"learning_rate": 5.395821382204644e-05,
|
| 436 |
+
"loss": 1.039,
|
| 437 |
+
"step": 30500
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.9359620784396606,
|
| 441 |
+
"grad_norm": 15.941644668579102,
|
| 442 |
+
"learning_rate": 5.320340569427252e-05,
|
| 443 |
+
"loss": 1.0279,
|
| 444 |
+
"step": 31000
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.951058240995139,
|
| 448 |
+
"grad_norm": 11.153121948242188,
|
| 449 |
+
"learning_rate": 5.2448597566498606e-05,
|
| 450 |
+
"loss": 1.0515,
|
| 451 |
+
"step": 31500
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.9661544035506174,
|
| 455 |
+
"grad_norm": 9.693687438964844,
|
| 456 |
+
"learning_rate": 5.1693789438724685e-05,
|
| 457 |
+
"loss": 1.0202,
|
| 458 |
+
"step": 32000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.9812505661060958,
|
| 462 |
+
"grad_norm": 11.340658187866211,
|
| 463 |
+
"learning_rate": 5.0938981310950765e-05,
|
| 464 |
+
"loss": 1.0226,
|
| 465 |
+
"step": 32500
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.9963467286615743,
|
| 469 |
+
"grad_norm": 15.900833129882812,
|
| 470 |
+
"learning_rate": 5.0184173183176844e-05,
|
| 471 |
+
"loss": 1.0472,
|
| 472 |
+
"step": 33000
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.0114428912170526,
|
| 476 |
+
"grad_norm": 12.715502738952637,
|
| 477 |
+
"learning_rate": 4.942936505540292e-05,
|
| 478 |
+
"loss": 0.9413,
|
| 479 |
+
"step": 33500
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.026539053772531,
|
| 483 |
+
"grad_norm": 19.717144012451172,
|
| 484 |
+
"learning_rate": 4.8674556927629e-05,
|
| 485 |
+
"loss": 0.941,
|
| 486 |
+
"step": 34000
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.0416352163280094,
|
| 490 |
+
"grad_norm": 16.49878692626953,
|
| 491 |
+
"learning_rate": 4.791974879985508e-05,
|
| 492 |
+
"loss": 0.9209,
|
| 493 |
+
"step": 34500
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.0567313788834878,
|
| 497 |
+
"grad_norm": 9.172051429748535,
|
| 498 |
+
"learning_rate": 4.716494067208116e-05,
|
| 499 |
+
"loss": 0.8879,
|
| 500 |
+
"step": 35000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.0718275414389662,
|
| 504 |
+
"grad_norm": 14.834227561950684,
|
| 505 |
+
"learning_rate": 4.641013254430724e-05,
|
| 506 |
+
"loss": 0.9343,
|
| 507 |
+
"step": 35500
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.0869237039944446,
|
| 511 |
+
"grad_norm": 15.549851417541504,
|
| 512 |
+
"learning_rate": 4.565532441653332e-05,
|
| 513 |
+
"loss": 0.9249,
|
| 514 |
+
"step": 36000
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.102019866549923,
|
| 518 |
+
"grad_norm": 8.137205123901367,
|
| 519 |
+
"learning_rate": 4.49005162887594e-05,
|
| 520 |
+
"loss": 0.8802,
|
| 521 |
+
"step": 36500
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.1171160291054014,
|
| 525 |
+
"grad_norm": 12.385357856750488,
|
| 526 |
+
"learning_rate": 4.414570816098548e-05,
|
| 527 |
+
"loss": 0.9158,
|
| 528 |
+
"step": 37000
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.1322121916608798,
|
| 532 |
+
"grad_norm": 14.191947937011719,
|
| 533 |
+
"learning_rate": 4.3390900033211564e-05,
|
| 534 |
+
"loss": 0.8878,
|
| 535 |
+
"step": 37500
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.1473083542163582,
|
| 539 |
+
"grad_norm": 12.700613975524902,
|
| 540 |
+
"learning_rate": 4.263609190543764e-05,
|
| 541 |
+
"loss": 0.8945,
|
| 542 |
+
"step": 38000
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.1624045167718366,
|
| 546 |
+
"grad_norm": 9.552591323852539,
|
| 547 |
+
"learning_rate": 4.188128377766372e-05,
|
| 548 |
+
"loss": 0.8681,
|
| 549 |
+
"step": 38500
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.177500679327315,
|
| 553 |
+
"grad_norm": 8.992091178894043,
|
| 554 |
+
"learning_rate": 4.11264756498898e-05,
|
| 555 |
+
"loss": 0.8636,
|
| 556 |
+
"step": 39000
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.1925968418827935,
|
| 560 |
+
"grad_norm": 9.592864990234375,
|
| 561 |
+
"learning_rate": 4.037166752211588e-05,
|
| 562 |
+
"loss": 0.8813,
|
| 563 |
+
"step": 39500
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.2076930044382719,
|
| 567 |
+
"grad_norm": 10.786945343017578,
|
| 568 |
+
"learning_rate": 3.961685939434196e-05,
|
| 569 |
+
"loss": 0.8586,
|
| 570 |
+
"step": 40000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.22278916699375,
|
| 574 |
+
"grad_norm": 14.606813430786133,
|
| 575 |
+
"learning_rate": 3.886205126656804e-05,
|
| 576 |
+
"loss": 0.8472,
|
| 577 |
+
"step": 40500
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.2378853295492287,
|
| 581 |
+
"grad_norm": 16.00104522705078,
|
| 582 |
+
"learning_rate": 3.810724313879412e-05,
|
| 583 |
+
"loss": 0.8492,
|
| 584 |
+
"step": 41000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.2529814921047069,
|
| 588 |
+
"grad_norm": 20.34152603149414,
|
| 589 |
+
"learning_rate": 3.73524350110202e-05,
|
| 590 |
+
"loss": 0.8528,
|
| 591 |
+
"step": 41500
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.2680776546601855,
|
| 595 |
+
"grad_norm": 13.854582786560059,
|
| 596 |
+
"learning_rate": 3.6597626883246284e-05,
|
| 597 |
+
"loss": 0.8315,
|
| 598 |
+
"step": 42000
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.2831738172156637,
|
| 602 |
+
"grad_norm": 8.319161415100098,
|
| 603 |
+
"learning_rate": 3.584281875547236e-05,
|
| 604 |
+
"loss": 0.8254,
|
| 605 |
+
"step": 42500
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.2982699797711421,
|
| 609 |
+
"grad_norm": 16.603830337524414,
|
| 610 |
+
"learning_rate": 3.508801062769844e-05,
|
| 611 |
+
"loss": 0.8311,
|
| 612 |
+
"step": 43000
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.3133661423266205,
|
| 616 |
+
"grad_norm": 32.95830535888672,
|
| 617 |
+
"learning_rate": 3.433320249992452e-05,
|
| 618 |
+
"loss": 0.816,
|
| 619 |
+
"step": 43500
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.328462304882099,
|
| 623 |
+
"grad_norm": 13.81757640838623,
|
| 624 |
+
"learning_rate": 3.35783943721506e-05,
|
| 625 |
+
"loss": 0.8198,
|
| 626 |
+
"step": 44000
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.3435584674375773,
|
| 630 |
+
"grad_norm": 6.295320987701416,
|
| 631 |
+
"learning_rate": 3.282358624437668e-05,
|
| 632 |
+
"loss": 0.8179,
|
| 633 |
+
"step": 44500
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.3586546299930558,
|
| 637 |
+
"grad_norm": 14.061367988586426,
|
| 638 |
+
"learning_rate": 3.206877811660276e-05,
|
| 639 |
+
"loss": 0.7924,
|
| 640 |
+
"step": 45000
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.3737507925485342,
|
| 644 |
+
"grad_norm": 19.793121337890625,
|
| 645 |
+
"learning_rate": 3.131396998882884e-05,
|
| 646 |
+
"loss": 0.8093,
|
| 647 |
+
"step": 45500
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.3888469551040126,
|
| 651 |
+
"grad_norm": 9.438281059265137,
|
| 652 |
+
"learning_rate": 3.055916186105492e-05,
|
| 653 |
+
"loss": 0.8255,
|
| 654 |
+
"step": 46000
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.403943117659491,
|
| 658 |
+
"grad_norm": 17.061111450195312,
|
| 659 |
+
"learning_rate": 2.9804353733281003e-05,
|
| 660 |
+
"loss": 0.7948,
|
| 661 |
+
"step": 46500
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.4190392802149694,
|
| 665 |
+
"grad_norm": 11.102389335632324,
|
| 666 |
+
"learning_rate": 2.9049545605507083e-05,
|
| 667 |
+
"loss": 0.785,
|
| 668 |
+
"step": 47000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.4341354427704478,
|
| 672 |
+
"grad_norm": 7.18344783782959,
|
| 673 |
+
"learning_rate": 2.8294737477733162e-05,
|
| 674 |
+
"loss": 0.7998,
|
| 675 |
+
"step": 47500
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.4492316053259262,
|
| 679 |
+
"grad_norm": 16.159982681274414,
|
| 680 |
+
"learning_rate": 2.753992934995924e-05,
|
| 681 |
+
"loss": 0.7842,
|
| 682 |
+
"step": 48000
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.4643277678814046,
|
| 686 |
+
"grad_norm": 21.598909378051758,
|
| 687 |
+
"learning_rate": 2.6785121222185324e-05,
|
| 688 |
+
"loss": 0.8171,
|
| 689 |
+
"step": 48500
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.4794239304368828,
|
| 693 |
+
"grad_norm": 13.887185096740723,
|
| 694 |
+
"learning_rate": 2.6030313094411403e-05,
|
| 695 |
+
"loss": 0.8031,
|
| 696 |
+
"step": 49000
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.4945200929923614,
|
| 700 |
+
"grad_norm": 5.625245094299316,
|
| 701 |
+
"learning_rate": 2.5275504966637482e-05,
|
| 702 |
+
"loss": 0.7824,
|
| 703 |
+
"step": 49500
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.5096162555478396,
|
| 707 |
+
"grad_norm": 12.34050178527832,
|
| 708 |
+
"learning_rate": 2.4520696838863565e-05,
|
| 709 |
+
"loss": 0.7807,
|
| 710 |
+
"step": 50000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.5247124181033183,
|
| 714 |
+
"grad_norm": 16.76266860961914,
|
| 715 |
+
"learning_rate": 2.3765888711089644e-05,
|
| 716 |
+
"loss": 0.7968,
|
| 717 |
+
"step": 50500
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.5398085806587964,
|
| 721 |
+
"grad_norm": 8.601690292358398,
|
| 722 |
+
"learning_rate": 2.3011080583315723e-05,
|
| 723 |
+
"loss": 0.7053,
|
| 724 |
+
"step": 51000
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.554904743214275,
|
| 728 |
+
"grad_norm": 10.708945274353027,
|
| 729 |
+
"learning_rate": 2.2256272455541802e-05,
|
| 730 |
+
"loss": 0.7548,
|
| 731 |
+
"step": 51500
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.5700009057697533,
|
| 735 |
+
"grad_norm": 12.558730125427246,
|
| 736 |
+
"learning_rate": 2.1501464327767882e-05,
|
| 737 |
+
"loss": 0.7547,
|
| 738 |
+
"step": 52000
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.5850970683252317,
|
| 742 |
+
"grad_norm": 13.777571678161621,
|
| 743 |
+
"learning_rate": 2.0746656199993964e-05,
|
| 744 |
+
"loss": 0.7463,
|
| 745 |
+
"step": 52500
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.60019323088071,
|
| 749 |
+
"grad_norm": 6.568544387817383,
|
| 750 |
+
"learning_rate": 1.9991848072220044e-05,
|
| 751 |
+
"loss": 0.7713,
|
| 752 |
+
"step": 53000
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.6152893934361885,
|
| 756 |
+
"grad_norm": 8.380231857299805,
|
| 757 |
+
"learning_rate": 1.9237039944446123e-05,
|
| 758 |
+
"loss": 0.7218,
|
| 759 |
+
"step": 53500
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.630385555991667,
|
| 763 |
+
"grad_norm": 15.468815803527832,
|
| 764 |
+
"learning_rate": 1.8482231816672202e-05,
|
| 765 |
+
"loss": 0.7452,
|
| 766 |
+
"step": 54000
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.6454817185471453,
|
| 770 |
+
"grad_norm": 9.33584976196289,
|
| 771 |
+
"learning_rate": 1.7727423688898285e-05,
|
| 772 |
+
"loss": 0.7433,
|
| 773 |
+
"step": 54500
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.6605778811026237,
|
| 777 |
+
"grad_norm": 25.546720504760742,
|
| 778 |
+
"learning_rate": 1.6972615561124364e-05,
|
| 779 |
+
"loss": 0.7679,
|
| 780 |
+
"step": 55000
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.6756740436581021,
|
| 784 |
+
"grad_norm": 25.5761661529541,
|
| 785 |
+
"learning_rate": 1.6217807433350443e-05,
|
| 786 |
+
"loss": 0.7717,
|
| 787 |
+
"step": 55500
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.6907702062135805,
|
| 791 |
+
"grad_norm": 13.344952583312988,
|
| 792 |
+
"learning_rate": 1.5462999305576522e-05,
|
| 793 |
+
"loss": 0.7313,
|
| 794 |
+
"step": 56000
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.705866368769059,
|
| 798 |
+
"grad_norm": 7.900672912597656,
|
| 799 |
+
"learning_rate": 1.4708191177802602e-05,
|
| 800 |
+
"loss": 0.7339,
|
| 801 |
+
"step": 56500
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 1.7209625313245374,
|
| 805 |
+
"grad_norm": 7.074102401733398,
|
| 806 |
+
"learning_rate": 1.3953383050028684e-05,
|
| 807 |
+
"loss": 0.7217,
|
| 808 |
+
"step": 57000
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.7360586938800155,
|
| 812 |
+
"grad_norm": 16.54216766357422,
|
| 813 |
+
"learning_rate": 1.3198574922254763e-05,
|
| 814 |
+
"loss": 0.7513,
|
| 815 |
+
"step": 57500
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.7511548564354942,
|
| 819 |
+
"grad_norm": 9.716275215148926,
|
| 820 |
+
"learning_rate": 1.2443766794480843e-05,
|
| 821 |
+
"loss": 0.7189,
|
| 822 |
+
"step": 58000
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.7662510189909724,
|
| 826 |
+
"grad_norm": 10.64577579498291,
|
| 827 |
+
"learning_rate": 1.1688958666706924e-05,
|
| 828 |
+
"loss": 0.7293,
|
| 829 |
+
"step": 58500
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.781347181546451,
|
| 833 |
+
"grad_norm": 18.039608001708984,
|
| 834 |
+
"learning_rate": 1.0934150538933004e-05,
|
| 835 |
+
"loss": 0.712,
|
| 836 |
+
"step": 59000
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.7964433441019292,
|
| 840 |
+
"grad_norm": 19.148468017578125,
|
| 841 |
+
"learning_rate": 1.0179342411159084e-05,
|
| 842 |
+
"loss": 0.7101,
|
| 843 |
+
"step": 59500
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 1.8115395066574078,
|
| 847 |
+
"grad_norm": 9.460704803466797,
|
| 848 |
+
"learning_rate": 9.424534283385165e-06,
|
| 849 |
+
"loss": 0.7298,
|
| 850 |
+
"step": 60000
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.826635669212886,
|
| 854 |
+
"grad_norm": 19.228601455688477,
|
| 855 |
+
"learning_rate": 8.669726155611244e-06,
|
| 856 |
+
"loss": 0.7265,
|
| 857 |
+
"step": 60500
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.8417318317683646,
|
| 861 |
+
"grad_norm": 32.19565200805664,
|
| 862 |
+
"learning_rate": 7.914918027837325e-06,
|
| 863 |
+
"loss": 0.7139,
|
| 864 |
+
"step": 61000
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.8568279943238428,
|
| 868 |
+
"grad_norm": 11.675516128540039,
|
| 869 |
+
"learning_rate": 7.160109900063404e-06,
|
| 870 |
+
"loss": 0.7163,
|
| 871 |
+
"step": 61500
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 1.8719241568793212,
|
| 875 |
+
"grad_norm": 16.321319580078125,
|
| 876 |
+
"learning_rate": 6.405301772289485e-06,
|
| 877 |
+
"loss": 0.7546,
|
| 878 |
+
"step": 62000
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.8870203194347996,
|
| 882 |
+
"grad_norm": 8.392330169677734,
|
| 883 |
+
"learning_rate": 5.650493644515564e-06,
|
| 884 |
+
"loss": 0.7005,
|
| 885 |
+
"step": 62500
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.902116481990278,
|
| 889 |
+
"grad_norm": 18.768943786621094,
|
| 890 |
+
"learning_rate": 4.895685516741644e-06,
|
| 891 |
+
"loss": 0.7305,
|
| 892 |
+
"step": 63000
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.9172126445457565,
|
| 896 |
+
"grad_norm": 26.690473556518555,
|
| 897 |
+
"learning_rate": 4.140877388967724e-06,
|
| 898 |
+
"loss": 0.7373,
|
| 899 |
+
"step": 63500
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 1.9323088071012349,
|
| 903 |
+
"grad_norm": 9.388639450073242,
|
| 904 |
+
"learning_rate": 3.386069261193805e-06,
|
| 905 |
+
"loss": 0.6859,
|
| 906 |
+
"step": 64000
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.9474049696567133,
|
| 910 |
+
"grad_norm": 10.29443645477295,
|
| 911 |
+
"learning_rate": 2.6312611334198844e-06,
|
| 912 |
+
"loss": 0.7192,
|
| 913 |
+
"step": 64500
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.9625011322121917,
|
| 917 |
+
"grad_norm": 20.20147705078125,
|
| 918 |
+
"learning_rate": 1.8764530056459647e-06,
|
| 919 |
+
"loss": 0.689,
|
| 920 |
+
"step": 65000
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.97759729476767,
|
| 924 |
+
"grad_norm": 7.911527633666992,
|
| 925 |
+
"learning_rate": 1.121644877872045e-06,
|
| 926 |
+
"loss": 0.6781,
|
| 927 |
+
"step": 65500
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 1.9926934573231485,
|
| 931 |
+
"grad_norm": 14.393226623535156,
|
| 932 |
+
"learning_rate": 3.6683675009812505e-07,
|
| 933 |
+
"loss": 0.6922,
|
| 934 |
+
"step": 66000
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 2.0,
|
| 938 |
+
"step": 66242,
|
| 939 |
+
"total_flos": 1.1159654864928768e+16,
|
| 940 |
+
"train_loss": 1.1755467142439282,
|
| 941 |
+
"train_runtime": 8495.7942,
|
| 942 |
+
"train_samples_per_second": 93.564,
|
| 943 |
+
"train_steps_per_second": 7.797
|
| 944 |
+
}
|
| 945 |
+
],
|
| 946 |
+
"logging_steps": 500,
|
| 947 |
+
"max_steps": 66242,
|
| 948 |
+
"num_input_tokens_seen": 0,
|
| 949 |
+
"num_train_epochs": 2,
|
| 950 |
+
"save_steps": 500,
|
| 951 |
+
"stateful_callbacks": {
|
| 952 |
+
"TrainerControl": {
|
| 953 |
+
"args": {
|
| 954 |
+
"should_epoch_stop": false,
|
| 955 |
+
"should_evaluate": false,
|
| 956 |
+
"should_log": false,
|
| 957 |
+
"should_save": false,
|
| 958 |
+
"should_training_stop": false
|
| 959 |
+
},
|
| 960 |
+
"attributes": {}
|
| 961 |
+
}
|
| 962 |
+
},
|
| 963 |
+
"total_flos": 1.1159654864928768e+16,
|
| 964 |
+
"train_batch_size": 12,
|
| 965 |
+
"trial_name": null,
|
| 966 |
+
"trial_params": null
|
| 967 |
+
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5841
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e38e4e1ac781d20530466f7cb3bfb8fae2ee9dfdc34f1279edd7cd4395fb9ded
|
| 3 |
size 5841
|