Upload 4 files
Browse files- optimizer.pt +3 -0
- scheduler.pt +3 -0
- trainer_state.json +945 -0
- training_args.bin +3 -0
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e62dc2d191b49cee9fd815ec78acb9c7a579b2a37b2b1d7f8aa6375ca3a0e23a
|
| 3 |
+
size 997370106
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:61062d9a7fb887f32150bfb8d43225ea263194b767e580db272ea50e0e292cb0
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,945 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1000,
|
| 3 |
+
"best_metric": 0.8438986049887841,
|
| 4 |
+
"best_model_checkpoint": "./results/checkpoint-1000",
|
| 5 |
+
"epoch": 0.029189526267256648,
|
| 6 |
+
"eval_steps": 100,
|
| 7 |
+
"global_step": 1100,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.00026535932970233315,
|
| 14 |
+
"grad_norm": 4.446621894836426,
|
| 15 |
+
"learning_rate": 3.6e-07,
|
| 16 |
+
"loss": 1.855,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0005307186594046663,
|
| 21 |
+
"grad_norm": 6.435311317443848,
|
| 22 |
+
"learning_rate": 7.6e-07,
|
| 23 |
+
"loss": 1.8131,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.0007960779891069995,
|
| 28 |
+
"grad_norm": 3.5201945304870605,
|
| 29 |
+
"learning_rate": 1.1600000000000001e-06,
|
| 30 |
+
"loss": 1.7821,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.0010614373188093326,
|
| 35 |
+
"grad_norm": 5.47129487991333,
|
| 36 |
+
"learning_rate": 1.56e-06,
|
| 37 |
+
"loss": 1.8187,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.001326796648511666,
|
| 42 |
+
"grad_norm": 7.340343475341797,
|
| 43 |
+
"learning_rate": 1.9600000000000003e-06,
|
| 44 |
+
"loss": 1.7791,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.001592155978213999,
|
| 49 |
+
"grad_norm": 5.010741233825684,
|
| 50 |
+
"learning_rate": 2.3600000000000003e-06,
|
| 51 |
+
"loss": 1.7705,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.0018575153079163323,
|
| 56 |
+
"grad_norm": 4.706456184387207,
|
| 57 |
+
"learning_rate": 2.7600000000000003e-06,
|
| 58 |
+
"loss": 1.7616,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.002122874637618665,
|
| 63 |
+
"grad_norm": 3.865488290786743,
|
| 64 |
+
"learning_rate": 3.1600000000000002e-06,
|
| 65 |
+
"loss": 1.7265,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.0023882339673209985,
|
| 70 |
+
"grad_norm": 6.390440464019775,
|
| 71 |
+
"learning_rate": 3.5600000000000002e-06,
|
| 72 |
+
"loss": 1.7143,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.002653593297023332,
|
| 77 |
+
"grad_norm": 5.747302532196045,
|
| 78 |
+
"learning_rate": 3.96e-06,
|
| 79 |
+
"loss": 1.6675,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.002653593297023332,
|
| 84 |
+
"eval_accuracy": 0.3978160782284494,
|
| 85 |
+
"eval_f1": 0.22667861021073907,
|
| 86 |
+
"eval_loss": 1.7215721607208252,
|
| 87 |
+
"eval_precision": 0.3277163462999534,
|
| 88 |
+
"eval_recall": 0.3978160782284494,
|
| 89 |
+
"eval_runtime": 1141.6234,
|
| 90 |
+
"eval_samples_per_second": 66.019,
|
| 91 |
+
"eval_steps_per_second": 8.253,
|
| 92 |
+
"step": 100
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.0029189526267256647,
|
| 96 |
+
"grad_norm": 5.64985466003418,
|
| 97 |
+
"learning_rate": 4.360000000000001e-06,
|
| 98 |
+
"loss": 1.6701,
|
| 99 |
+
"step": 110
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.003184311956427998,
|
| 103 |
+
"grad_norm": 7.48480224609375,
|
| 104 |
+
"learning_rate": 4.76e-06,
|
| 105 |
+
"loss": 1.6665,
|
| 106 |
+
"step": 120
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 0.0034496712861303313,
|
| 110 |
+
"grad_norm": 6.59235954284668,
|
| 111 |
+
"learning_rate": 5.1600000000000006e-06,
|
| 112 |
+
"loss": 1.6049,
|
| 113 |
+
"step": 130
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 0.0037150306158326646,
|
| 117 |
+
"grad_norm": 5.440073013305664,
|
| 118 |
+
"learning_rate": 5.560000000000001e-06,
|
| 119 |
+
"loss": 1.6174,
|
| 120 |
+
"step": 140
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 0.0039803899455349975,
|
| 124 |
+
"grad_norm": 5.735574245452881,
|
| 125 |
+
"learning_rate": 5.9600000000000005e-06,
|
| 126 |
+
"loss": 1.5654,
|
| 127 |
+
"step": 150
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"epoch": 0.00424574927523733,
|
| 131 |
+
"grad_norm": 5.891671180725098,
|
| 132 |
+
"learning_rate": 6.360000000000001e-06,
|
| 133 |
+
"loss": 1.5445,
|
| 134 |
+
"step": 160
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.004511108604939664,
|
| 138 |
+
"grad_norm": 7.2471089363098145,
|
| 139 |
+
"learning_rate": 6.760000000000001e-06,
|
| 140 |
+
"loss": 1.5545,
|
| 141 |
+
"step": 170
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.004776467934641997,
|
| 145 |
+
"grad_norm": 8.074533462524414,
|
| 146 |
+
"learning_rate": 7.16e-06,
|
| 147 |
+
"loss": 1.5121,
|
| 148 |
+
"step": 180
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.00504182726434433,
|
| 152 |
+
"grad_norm": 6.5745768547058105,
|
| 153 |
+
"learning_rate": 7.5600000000000005e-06,
|
| 154 |
+
"loss": 1.4951,
|
| 155 |
+
"step": 190
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.005307186594046664,
|
| 159 |
+
"grad_norm": 7.1547112464904785,
|
| 160 |
+
"learning_rate": 7.960000000000002e-06,
|
| 161 |
+
"loss": 1.5225,
|
| 162 |
+
"step": 200
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 0.005307186594046664,
|
| 166 |
+
"eval_accuracy": 0.5399965503058286,
|
| 167 |
+
"eval_f1": 0.4954655997286682,
|
| 168 |
+
"eval_loss": 1.4961189031600952,
|
| 169 |
+
"eval_precision": 0.5230960318752841,
|
| 170 |
+
"eval_recall": 0.5399965503058286,
|
| 171 |
+
"eval_runtime": 1146.418,
|
| 172 |
+
"eval_samples_per_second": 65.743,
|
| 173 |
+
"eval_steps_per_second": 8.219,
|
| 174 |
+
"step": 200
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 0.0055725459237489965,
|
| 178 |
+
"grad_norm": 10.55716609954834,
|
| 179 |
+
"learning_rate": 8.36e-06,
|
| 180 |
+
"loss": 1.486,
|
| 181 |
+
"step": 210
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 0.005837905253451329,
|
| 185 |
+
"grad_norm": 10.896256446838379,
|
| 186 |
+
"learning_rate": 8.76e-06,
|
| 187 |
+
"loss": 1.3551,
|
| 188 |
+
"step": 220
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.006103264583153663,
|
| 192 |
+
"grad_norm": 10.593475341796875,
|
| 193 |
+
"learning_rate": 9.16e-06,
|
| 194 |
+
"loss": 1.3457,
|
| 195 |
+
"step": 230
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.006368623912855996,
|
| 199 |
+
"grad_norm": 12.37448787689209,
|
| 200 |
+
"learning_rate": 9.56e-06,
|
| 201 |
+
"loss": 1.4264,
|
| 202 |
+
"step": 240
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 0.006633983242558329,
|
| 206 |
+
"grad_norm": 15.805830001831055,
|
| 207 |
+
"learning_rate": 9.960000000000001e-06,
|
| 208 |
+
"loss": 1.1405,
|
| 209 |
+
"step": 250
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.006899342572260663,
|
| 213 |
+
"grad_norm": 7.878324031829834,
|
| 214 |
+
"learning_rate": 1.036e-05,
|
| 215 |
+
"loss": 1.3623,
|
| 216 |
+
"step": 260
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 0.0071647019019629955,
|
| 220 |
+
"grad_norm": 9.763978004455566,
|
| 221 |
+
"learning_rate": 1.0760000000000002e-05,
|
| 222 |
+
"loss": 1.1918,
|
| 223 |
+
"step": 270
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 0.007430061231665329,
|
| 227 |
+
"grad_norm": 8.449042320251465,
|
| 228 |
+
"learning_rate": 1.1160000000000002e-05,
|
| 229 |
+
"loss": 1.1437,
|
| 230 |
+
"step": 280
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.007695420561367662,
|
| 234 |
+
"grad_norm": 10.577520370483398,
|
| 235 |
+
"learning_rate": 1.156e-05,
|
| 236 |
+
"loss": 1.0793,
|
| 237 |
+
"step": 290
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.007960779891069995,
|
| 241 |
+
"grad_norm": 11.707535743713379,
|
| 242 |
+
"learning_rate": 1.196e-05,
|
| 243 |
+
"loss": 1.1895,
|
| 244 |
+
"step": 300
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 0.007960779891069995,
|
| 248 |
+
"eval_accuracy": 0.6815534238214651,
|
| 249 |
+
"eval_f1": 0.6624035468309406,
|
| 250 |
+
"eval_loss": 1.228468894958496,
|
| 251 |
+
"eval_precision": 0.6548079319665443,
|
| 252 |
+
"eval_recall": 0.6815534238214651,
|
| 253 |
+
"eval_runtime": 1147.8213,
|
| 254 |
+
"eval_samples_per_second": 65.663,
|
| 255 |
+
"eval_steps_per_second": 8.209,
|
| 256 |
+
"step": 300
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.008226139220772328,
|
| 260 |
+
"grad_norm": 7.825327396392822,
|
| 261 |
+
"learning_rate": 1.236e-05,
|
| 262 |
+
"loss": 1.0,
|
| 263 |
+
"step": 310
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 0.00849149855047466,
|
| 267 |
+
"grad_norm": 9.7120943069458,
|
| 268 |
+
"learning_rate": 1.2760000000000001e-05,
|
| 269 |
+
"loss": 1.2152,
|
| 270 |
+
"step": 320
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 0.008756857880176995,
|
| 274 |
+
"grad_norm": 10.450387954711914,
|
| 275 |
+
"learning_rate": 1.3160000000000001e-05,
|
| 276 |
+
"loss": 1.2838,
|
| 277 |
+
"step": 330
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 0.009022217209879328,
|
| 281 |
+
"grad_norm": 6.901065826416016,
|
| 282 |
+
"learning_rate": 1.3560000000000002e-05,
|
| 283 |
+
"loss": 1.1989,
|
| 284 |
+
"step": 340
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.009287576539581661,
|
| 288 |
+
"grad_norm": 7.151400089263916,
|
| 289 |
+
"learning_rate": 1.396e-05,
|
| 290 |
+
"loss": 1.0571,
|
| 291 |
+
"step": 350
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 0.009552935869283994,
|
| 295 |
+
"grad_norm": 7.359131813049316,
|
| 296 |
+
"learning_rate": 1.4360000000000001e-05,
|
| 297 |
+
"loss": 1.2923,
|
| 298 |
+
"step": 360
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 0.009818295198986327,
|
| 302 |
+
"grad_norm": 6.762883186340332,
|
| 303 |
+
"learning_rate": 1.4760000000000001e-05,
|
| 304 |
+
"loss": 1.2465,
|
| 305 |
+
"step": 370
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 0.01008365452868866,
|
| 309 |
+
"grad_norm": 7.525689125061035,
|
| 310 |
+
"learning_rate": 1.516e-05,
|
| 311 |
+
"loss": 0.889,
|
| 312 |
+
"step": 380
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 0.010349013858390994,
|
| 316 |
+
"grad_norm": 12.986421585083008,
|
| 317 |
+
"learning_rate": 1.556e-05,
|
| 318 |
+
"loss": 1.2117,
|
| 319 |
+
"step": 390
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 0.010614373188093327,
|
| 323 |
+
"grad_norm": 10.896193504333496,
|
| 324 |
+
"learning_rate": 1.5960000000000003e-05,
|
| 325 |
+
"loss": 1.0,
|
| 326 |
+
"step": 400
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.010614373188093327,
|
| 330 |
+
"eval_accuracy": 0.7040295081532195,
|
| 331 |
+
"eval_f1": 0.7118047953474329,
|
| 332 |
+
"eval_loss": 1.004681944847107,
|
| 333 |
+
"eval_precision": 0.7430775529469658,
|
| 334 |
+
"eval_recall": 0.7040295081532195,
|
| 335 |
+
"eval_runtime": 1149.1166,
|
| 336 |
+
"eval_samples_per_second": 65.589,
|
| 337 |
+
"eval_steps_per_second": 8.199,
|
| 338 |
+
"step": 400
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.01087973251779566,
|
| 342 |
+
"grad_norm": 10.948391914367676,
|
| 343 |
+
"learning_rate": 1.636e-05,
|
| 344 |
+
"loss": 1.1142,
|
| 345 |
+
"step": 410
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.011145091847497993,
|
| 349 |
+
"grad_norm": 9.782318115234375,
|
| 350 |
+
"learning_rate": 1.6760000000000002e-05,
|
| 351 |
+
"loss": 1.0815,
|
| 352 |
+
"step": 420
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.011410451177200326,
|
| 356 |
+
"grad_norm": 15.20601749420166,
|
| 357 |
+
"learning_rate": 1.7160000000000002e-05,
|
| 358 |
+
"loss": 1.1052,
|
| 359 |
+
"step": 430
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.011675810506902659,
|
| 363 |
+
"grad_norm": 13.752850532531738,
|
| 364 |
+
"learning_rate": 1.756e-05,
|
| 365 |
+
"loss": 0.9449,
|
| 366 |
+
"step": 440
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.011941169836604993,
|
| 370 |
+
"grad_norm": 14.253933906555176,
|
| 371 |
+
"learning_rate": 1.796e-05,
|
| 372 |
+
"loss": 0.8107,
|
| 373 |
+
"step": 450
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.012206529166307326,
|
| 377 |
+
"grad_norm": 10.736539840698242,
|
| 378 |
+
"learning_rate": 1.8360000000000004e-05,
|
| 379 |
+
"loss": 0.8507,
|
| 380 |
+
"step": 460
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.012471888496009659,
|
| 384 |
+
"grad_norm": 19.90713119506836,
|
| 385 |
+
"learning_rate": 1.876e-05,
|
| 386 |
+
"loss": 0.8218,
|
| 387 |
+
"step": 470
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.012737247825711992,
|
| 391 |
+
"grad_norm": 5.8942084312438965,
|
| 392 |
+
"learning_rate": 1.916e-05,
|
| 393 |
+
"loss": 0.8322,
|
| 394 |
+
"step": 480
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.013002607155414325,
|
| 398 |
+
"grad_norm": 5.852909088134766,
|
| 399 |
+
"learning_rate": 1.9560000000000002e-05,
|
| 400 |
+
"loss": 0.72,
|
| 401 |
+
"step": 490
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.013267966485116658,
|
| 405 |
+
"grad_norm": 8.721864700317383,
|
| 406 |
+
"learning_rate": 1.9960000000000002e-05,
|
| 407 |
+
"loss": 0.8702,
|
| 408 |
+
"step": 500
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.013267966485116658,
|
| 412 |
+
"eval_accuracy": 0.7919303692499569,
|
| 413 |
+
"eval_f1": 0.7952185444780413,
|
| 414 |
+
"eval_loss": 0.7821776270866394,
|
| 415 |
+
"eval_precision": 0.8109789187273935,
|
| 416 |
+
"eval_recall": 0.7919303692499569,
|
| 417 |
+
"eval_runtime": 1149.5055,
|
| 418 |
+
"eval_samples_per_second": 65.566,
|
| 419 |
+
"eval_steps_per_second": 8.197,
|
| 420 |
+
"step": 500
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.013533325814818992,
|
| 424 |
+
"grad_norm": 7.045879364013672,
|
| 425 |
+
"learning_rate": 1.9999042145593872e-05,
|
| 426 |
+
"loss": 0.8162,
|
| 427 |
+
"step": 510
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 0.013798685144521325,
|
| 431 |
+
"grad_norm": 12.590984344482422,
|
| 432 |
+
"learning_rate": 1.9997977862920393e-05,
|
| 433 |
+
"loss": 0.8281,
|
| 434 |
+
"step": 520
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.014064044474223658,
|
| 438 |
+
"grad_norm": 4.697333812713623,
|
| 439 |
+
"learning_rate": 1.9996913580246914e-05,
|
| 440 |
+
"loss": 0.8597,
|
| 441 |
+
"step": 530
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.014329403803925991,
|
| 445 |
+
"grad_norm": 7.0437188148498535,
|
| 446 |
+
"learning_rate": 1.999584929757344e-05,
|
| 447 |
+
"loss": 0.6957,
|
| 448 |
+
"step": 540
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.014594763133628324,
|
| 452 |
+
"grad_norm": 12.778396606445312,
|
| 453 |
+
"learning_rate": 1.999478501489996e-05,
|
| 454 |
+
"loss": 0.8493,
|
| 455 |
+
"step": 550
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.014860122463330658,
|
| 459 |
+
"grad_norm": 30.98528480529785,
|
| 460 |
+
"learning_rate": 1.999372073222648e-05,
|
| 461 |
+
"loss": 0.8598,
|
| 462 |
+
"step": 560
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 0.015125481793032991,
|
| 466 |
+
"grad_norm": 6.842329502105713,
|
| 467 |
+
"learning_rate": 1.9992656449553e-05,
|
| 468 |
+
"loss": 0.7805,
|
| 469 |
+
"step": 570
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 0.015390841122735324,
|
| 473 |
+
"grad_norm": 7.865843772888184,
|
| 474 |
+
"learning_rate": 1.9991592166879526e-05,
|
| 475 |
+
"loss": 0.6858,
|
| 476 |
+
"step": 580
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.01565620045243766,
|
| 480 |
+
"grad_norm": 7.990331172943115,
|
| 481 |
+
"learning_rate": 1.9990527884206047e-05,
|
| 482 |
+
"loss": 0.8487,
|
| 483 |
+
"step": 590
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 0.01592155978213999,
|
| 487 |
+
"grad_norm": 7.9941935539245605,
|
| 488 |
+
"learning_rate": 1.9989463601532568e-05,
|
| 489 |
+
"loss": 0.7483,
|
| 490 |
+
"step": 600
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 0.01592155978213999,
|
| 494 |
+
"eval_accuracy": 0.8143533813636906,
|
| 495 |
+
"eval_f1": 0.8091828053145604,
|
| 496 |
+
"eval_loss": 0.7799807190895081,
|
| 497 |
+
"eval_precision": 0.8135668963695762,
|
| 498 |
+
"eval_recall": 0.8143533813636906,
|
| 499 |
+
"eval_runtime": 1149.6535,
|
| 500 |
+
"eval_samples_per_second": 65.558,
|
| 501 |
+
"eval_steps_per_second": 8.196,
|
| 502 |
+
"step": 600
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 0.016186919111842325,
|
| 506 |
+
"grad_norm": 7.86228609085083,
|
| 507 |
+
"learning_rate": 1.9988399318859092e-05,
|
| 508 |
+
"loss": 0.8267,
|
| 509 |
+
"step": 610
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 0.016452278441544656,
|
| 513 |
+
"grad_norm": 17.805816650390625,
|
| 514 |
+
"learning_rate": 1.9987335036185613e-05,
|
| 515 |
+
"loss": 0.7449,
|
| 516 |
+
"step": 620
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 0.01671763777124699,
|
| 520 |
+
"grad_norm": 20.419509887695312,
|
| 521 |
+
"learning_rate": 1.9986270753512134e-05,
|
| 522 |
+
"loss": 0.806,
|
| 523 |
+
"step": 630
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 0.01698299710094932,
|
| 527 |
+
"grad_norm": 8.158143043518066,
|
| 528 |
+
"learning_rate": 1.9985206470838655e-05,
|
| 529 |
+
"loss": 0.5455,
|
| 530 |
+
"step": 640
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.017248356430651656,
|
| 534 |
+
"grad_norm": 19.537343978881836,
|
| 535 |
+
"learning_rate": 1.998414218816518e-05,
|
| 536 |
+
"loss": 0.6334,
|
| 537 |
+
"step": 650
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.01751371576035399,
|
| 541 |
+
"grad_norm": 13.77968692779541,
|
| 542 |
+
"learning_rate": 1.99830779054917e-05,
|
| 543 |
+
"loss": 0.6848,
|
| 544 |
+
"step": 660
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 0.017779075090056322,
|
| 548 |
+
"grad_norm": 13.80045223236084,
|
| 549 |
+
"learning_rate": 1.9982013622818222e-05,
|
| 550 |
+
"loss": 0.8403,
|
| 551 |
+
"step": 670
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 0.018044434419758656,
|
| 555 |
+
"grad_norm": 7.564460277557373,
|
| 556 |
+
"learning_rate": 1.9980949340144743e-05,
|
| 557 |
+
"loss": 0.8387,
|
| 558 |
+
"step": 680
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 0.018309793749460988,
|
| 562 |
+
"grad_norm": 9.932668685913086,
|
| 563 |
+
"learning_rate": 1.9979885057471267e-05,
|
| 564 |
+
"loss": 0.6252,
|
| 565 |
+
"step": 690
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 0.018575153079163322,
|
| 569 |
+
"grad_norm": 3.894618034362793,
|
| 570 |
+
"learning_rate": 1.997882077479779e-05,
|
| 571 |
+
"loss": 0.6915,
|
| 572 |
+
"step": 700
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.018575153079163322,
|
| 576 |
+
"eval_accuracy": 0.8227255237564516,
|
| 577 |
+
"eval_f1": 0.8164082076886476,
|
| 578 |
+
"eval_loss": 0.7442639470100403,
|
| 579 |
+
"eval_precision": 0.8239897054500639,
|
| 580 |
+
"eval_recall": 0.8227255237564516,
|
| 581 |
+
"eval_runtime": 1149.8124,
|
| 582 |
+
"eval_samples_per_second": 65.549,
|
| 583 |
+
"eval_steps_per_second": 8.194,
|
| 584 |
+
"step": 700
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.018840512408865657,
|
| 588 |
+
"grad_norm": 8.771450996398926,
|
| 589 |
+
"learning_rate": 1.997775649212431e-05,
|
| 590 |
+
"loss": 0.5164,
|
| 591 |
+
"step": 710
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.019105871738567988,
|
| 595 |
+
"grad_norm": 13.675606727600098,
|
| 596 |
+
"learning_rate": 1.997669220945083e-05,
|
| 597 |
+
"loss": 0.7917,
|
| 598 |
+
"step": 720
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.019371231068270323,
|
| 602 |
+
"grad_norm": 7.2742462158203125,
|
| 603 |
+
"learning_rate": 1.9975627926777355e-05,
|
| 604 |
+
"loss": 0.5966,
|
| 605 |
+
"step": 730
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.019636590397972654,
|
| 609 |
+
"grad_norm": 8.653703689575195,
|
| 610 |
+
"learning_rate": 1.9974563644103876e-05,
|
| 611 |
+
"loss": 0.7009,
|
| 612 |
+
"step": 740
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.01990194972767499,
|
| 616 |
+
"grad_norm": 4.837522983551025,
|
| 617 |
+
"learning_rate": 1.9973499361430397e-05,
|
| 618 |
+
"loss": 0.6471,
|
| 619 |
+
"step": 750
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.02016730905737732,
|
| 623 |
+
"grad_norm": 51.784481048583984,
|
| 624 |
+
"learning_rate": 1.997243507875692e-05,
|
| 625 |
+
"loss": 0.5877,
|
| 626 |
+
"step": 760
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.020432668387079654,
|
| 630 |
+
"grad_norm": 10.597892761230469,
|
| 631 |
+
"learning_rate": 1.9971370796083442e-05,
|
| 632 |
+
"loss": 0.5925,
|
| 633 |
+
"step": 770
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.02069802771678199,
|
| 637 |
+
"grad_norm": 13.237262725830078,
|
| 638 |
+
"learning_rate": 1.9970306513409963e-05,
|
| 639 |
+
"loss": 0.6298,
|
| 640 |
+
"step": 780
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.02096338704648432,
|
| 644 |
+
"grad_norm": 9.751429557800293,
|
| 645 |
+
"learning_rate": 1.9969242230736484e-05,
|
| 646 |
+
"loss": 0.5414,
|
| 647 |
+
"step": 790
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.021228746376186654,
|
| 651 |
+
"grad_norm": 15.811433792114258,
|
| 652 |
+
"learning_rate": 1.996817794806301e-05,
|
| 653 |
+
"loss": 0.7402,
|
| 654 |
+
"step": 800
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.021228746376186654,
|
| 658 |
+
"eval_accuracy": 0.8336318645596996,
|
| 659 |
+
"eval_f1": 0.8326606774493011,
|
| 660 |
+
"eval_loss": 0.6281165480613708,
|
| 661 |
+
"eval_precision": 0.8348071025847044,
|
| 662 |
+
"eval_recall": 0.8336318645596996,
|
| 663 |
+
"eval_runtime": 1150.0313,
|
| 664 |
+
"eval_samples_per_second": 65.536,
|
| 665 |
+
"eval_steps_per_second": 8.193,
|
| 666 |
+
"step": 800
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 0.021494105705888986,
|
| 670 |
+
"grad_norm": 6.421304225921631,
|
| 671 |
+
"learning_rate": 1.9967113665389526e-05,
|
| 672 |
+
"loss": 0.5246,
|
| 673 |
+
"step": 810
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 0.02175946503559132,
|
| 677 |
+
"grad_norm": 10.94848918914795,
|
| 678 |
+
"learning_rate": 1.996604938271605e-05,
|
| 679 |
+
"loss": 0.4998,
|
| 680 |
+
"step": 820
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 0.022024824365293655,
|
| 684 |
+
"grad_norm": 18.42159080505371,
|
| 685 |
+
"learning_rate": 1.9964985100042572e-05,
|
| 686 |
+
"loss": 0.3855,
|
| 687 |
+
"step": 830
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 0.022290183694995986,
|
| 691 |
+
"grad_norm": 6.594385623931885,
|
| 692 |
+
"learning_rate": 1.9963920817369096e-05,
|
| 693 |
+
"loss": 0.6205,
|
| 694 |
+
"step": 840
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 0.02255554302469832,
|
| 698 |
+
"grad_norm": 15.90577220916748,
|
| 699 |
+
"learning_rate": 1.9962856534695617e-05,
|
| 700 |
+
"loss": 0.5837,
|
| 701 |
+
"step": 850
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 0.022820902354400652,
|
| 705 |
+
"grad_norm": 13.169767379760742,
|
| 706 |
+
"learning_rate": 1.996179225202214e-05,
|
| 707 |
+
"loss": 0.6141,
|
| 708 |
+
"step": 860
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 0.023086261684102986,
|
| 712 |
+
"grad_norm": 18.30284881591797,
|
| 713 |
+
"learning_rate": 1.9960727969348663e-05,
|
| 714 |
+
"loss": 0.5903,
|
| 715 |
+
"step": 870
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 0.023351621013805317,
|
| 719 |
+
"grad_norm": 18.575279235839844,
|
| 720 |
+
"learning_rate": 1.9959663686675184e-05,
|
| 721 |
+
"loss": 0.6436,
|
| 722 |
+
"step": 880
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 0.023616980343507652,
|
| 726 |
+
"grad_norm": 7.2494587898254395,
|
| 727 |
+
"learning_rate": 1.9958599404001705e-05,
|
| 728 |
+
"loss": 0.5912,
|
| 729 |
+
"step": 890
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.023882339673209987,
|
| 733 |
+
"grad_norm": 27.128528594970703,
|
| 734 |
+
"learning_rate": 1.9957535121328226e-05,
|
| 735 |
+
"loss": 0.6165,
|
| 736 |
+
"step": 900
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 0.023882339673209987,
|
| 740 |
+
"eval_accuracy": 0.8347198450291233,
|
| 741 |
+
"eval_f1": 0.8366928520828808,
|
| 742 |
+
"eval_loss": 0.6065123081207275,
|
| 743 |
+
"eval_precision": 0.8420416556337017,
|
| 744 |
+
"eval_recall": 0.8347198450291233,
|
| 745 |
+
"eval_runtime": 1150.012,
|
| 746 |
+
"eval_samples_per_second": 65.538,
|
| 747 |
+
"eval_steps_per_second": 8.193,
|
| 748 |
+
"step": 900
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 0.024147699002912318,
|
| 752 |
+
"grad_norm": 11.817070960998535,
|
| 753 |
+
"learning_rate": 1.995647083865475e-05,
|
| 754 |
+
"loss": 0.6235,
|
| 755 |
+
"step": 910
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 0.024413058332614652,
|
| 759 |
+
"grad_norm": 13.93792724609375,
|
| 760 |
+
"learning_rate": 1.9955406555981268e-05,
|
| 761 |
+
"loss": 0.5442,
|
| 762 |
+
"step": 920
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 0.024678417662316984,
|
| 766 |
+
"grad_norm": 12.457510948181152,
|
| 767 |
+
"learning_rate": 1.9954342273307792e-05,
|
| 768 |
+
"loss": 0.5218,
|
| 769 |
+
"step": 930
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 0.024943776992019318,
|
| 773 |
+
"grad_norm": 19.165714263916016,
|
| 774 |
+
"learning_rate": 1.9953277990634313e-05,
|
| 775 |
+
"loss": 0.6517,
|
| 776 |
+
"step": 940
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 0.025209136321721653,
|
| 780 |
+
"grad_norm": 16.57741355895996,
|
| 781 |
+
"learning_rate": 1.9952213707960838e-05,
|
| 782 |
+
"loss": 0.6778,
|
| 783 |
+
"step": 950
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 0.025474495651423984,
|
| 787 |
+
"grad_norm": 10.39408016204834,
|
| 788 |
+
"learning_rate": 1.995114942528736e-05,
|
| 789 |
+
"loss": 0.7138,
|
| 790 |
+
"step": 960
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 0.02573985498112632,
|
| 794 |
+
"grad_norm": 9.648213386535645,
|
| 795 |
+
"learning_rate": 1.995008514261388e-05,
|
| 796 |
+
"loss": 0.63,
|
| 797 |
+
"step": 970
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 0.02600521431082865,
|
| 801 |
+
"grad_norm": 18.695945739746094,
|
| 802 |
+
"learning_rate": 1.99490208599404e-05,
|
| 803 |
+
"loss": 0.6557,
|
| 804 |
+
"step": 980
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 0.026270573640530984,
|
| 808 |
+
"grad_norm": 16.400314331054688,
|
| 809 |
+
"learning_rate": 1.9947956577266925e-05,
|
| 810 |
+
"loss": 0.7228,
|
| 811 |
+
"step": 990
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 0.026535932970233315,
|
| 815 |
+
"grad_norm": 6.129698753356934,
|
| 816 |
+
"learning_rate": 1.9946892294593446e-05,
|
| 817 |
+
"loss": 0.4626,
|
| 818 |
+
"step": 1000
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 0.026535932970233315,
|
| 822 |
+
"eval_accuracy": 0.846382464939166,
|
| 823 |
+
"eval_f1": 0.8438986049887841,
|
| 824 |
+
"eval_loss": 0.6415271759033203,
|
| 825 |
+
"eval_precision": 0.8464411216655519,
|
| 826 |
+
"eval_recall": 0.846382464939166,
|
| 827 |
+
"eval_runtime": 1149.903,
|
| 828 |
+
"eval_samples_per_second": 65.544,
|
| 829 |
+
"eval_steps_per_second": 8.194,
|
| 830 |
+
"step": 1000
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.02680129229993565,
|
| 834 |
+
"grad_norm": 9.027094841003418,
|
| 835 |
+
"learning_rate": 1.9945828011919967e-05,
|
| 836 |
+
"loss": 0.7249,
|
| 837 |
+
"step": 1010
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.027066651629637985,
|
| 841 |
+
"grad_norm": 14.91533374786377,
|
| 842 |
+
"learning_rate": 1.9944763729246492e-05,
|
| 843 |
+
"loss": 0.6765,
|
| 844 |
+
"step": 1020
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 0.027332010959340316,
|
| 848 |
+
"grad_norm": 3.9888482093811035,
|
| 849 |
+
"learning_rate": 1.994369944657301e-05,
|
| 850 |
+
"loss": 0.7675,
|
| 851 |
+
"step": 1030
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 0.02759737028904265,
|
| 855 |
+
"grad_norm": 7.831120491027832,
|
| 856 |
+
"learning_rate": 1.9942635163899534e-05,
|
| 857 |
+
"loss": 0.5764,
|
| 858 |
+
"step": 1040
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 0.02786272961874498,
|
| 862 |
+
"grad_norm": 7.661794185638428,
|
| 863 |
+
"learning_rate": 1.9941570881226055e-05,
|
| 864 |
+
"loss": 0.6496,
|
| 865 |
+
"step": 1050
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 0.028128088948447316,
|
| 869 |
+
"grad_norm": 5.446905136108398,
|
| 870 |
+
"learning_rate": 1.994050659855258e-05,
|
| 871 |
+
"loss": 0.7055,
|
| 872 |
+
"step": 1060
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.02839344827814965,
|
| 876 |
+
"grad_norm": 11.230253219604492,
|
| 877 |
+
"learning_rate": 1.9939442315879097e-05,
|
| 878 |
+
"loss": 0.6096,
|
| 879 |
+
"step": 1070
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.028658807607851982,
|
| 883 |
+
"grad_norm": 4.4712982177734375,
|
| 884 |
+
"learning_rate": 1.993837803320562e-05,
|
| 885 |
+
"loss": 0.603,
|
| 886 |
+
"step": 1080
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 0.028924166937554317,
|
| 890 |
+
"grad_norm": 6.14396858215332,
|
| 891 |
+
"learning_rate": 1.9937313750532142e-05,
|
| 892 |
+
"loss": 0.7755,
|
| 893 |
+
"step": 1090
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 0.029189526267256648,
|
| 897 |
+
"grad_norm": 15.914993286132812,
|
| 898 |
+
"learning_rate": 1.9936249467858667e-05,
|
| 899 |
+
"loss": 0.4619,
|
| 900 |
+
"step": 1100
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 0.029189526267256648,
|
| 904 |
+
"eval_accuracy": 0.846621289920259,
|
| 905 |
+
"eval_f1": 0.842526016669277,
|
| 906 |
+
"eval_loss": 0.6318368911743164,
|
| 907 |
+
"eval_precision": 0.8510012460868045,
|
| 908 |
+
"eval_recall": 0.846621289920259,
|
| 909 |
+
"eval_runtime": 1149.9706,
|
| 910 |
+
"eval_samples_per_second": 65.54,
|
| 911 |
+
"eval_steps_per_second": 8.193,
|
| 912 |
+
"step": 1100
|
| 913 |
+
}
|
| 914 |
+
],
|
| 915 |
+
"logging_steps": 10,
|
| 916 |
+
"max_steps": 188420,
|
| 917 |
+
"num_input_tokens_seen": 0,
|
| 918 |
+
"num_train_epochs": 5,
|
| 919 |
+
"save_steps": 100,
|
| 920 |
+
"stateful_callbacks": {
|
| 921 |
+
"EarlyStoppingCallback": {
|
| 922 |
+
"args": {
|
| 923 |
+
"early_stopping_patience": 3,
|
| 924 |
+
"early_stopping_threshold": 0.01
|
| 925 |
+
},
|
| 926 |
+
"attributes": {
|
| 927 |
+
"early_stopping_patience_counter": 3
|
| 928 |
+
}
|
| 929 |
+
},
|
| 930 |
+
"TrainerControl": {
|
| 931 |
+
"args": {
|
| 932 |
+
"should_epoch_stop": false,
|
| 933 |
+
"should_evaluate": false,
|
| 934 |
+
"should_log": false,
|
| 935 |
+
"should_save": true,
|
| 936 |
+
"should_training_stop": true
|
| 937 |
+
},
|
| 938 |
+
"attributes": {}
|
| 939 |
+
}
|
| 940 |
+
},
|
| 941 |
+
"total_flos": 4630920885043200.0,
|
| 942 |
+
"train_batch_size": 4,
|
| 943 |
+
"trial_name": null,
|
| 944 |
+
"trial_params": null
|
| 945 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:63651316137ef8a43b7b68fffe153c9c989f961f0e3702d3b3c782028c356ff2
|
| 3 |
+
size 5240
|