🍻 cheers
Browse files- all_results.json +12 -0
- eval_results.json +8 -0
- runs/Jun27_00-20-05_bernard/events.out.tfevents.1656283610.bernard.1177787.2 +3 -0
- train_results.json +7 -0
- trainer_state.json +901 -0
all_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.0,
|
| 3 |
+
"eval_accuracy": 0.9067077872012336,
|
| 4 |
+
"eval_loss": 0.3908935785293579,
|
| 5 |
+
"eval_runtime": 42.6956,
|
| 6 |
+
"eval_samples_per_second": 60.756,
|
| 7 |
+
"eval_steps_per_second": 7.612,
|
| 8 |
+
"train_loss": 1.0863585265353322,
|
| 9 |
+
"train_runtime": 1526.0713,
|
| 10 |
+
"train_samples_per_second": 26.762,
|
| 11 |
+
"train_steps_per_second": 0.839
|
| 12 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.0,
|
| 3 |
+
"eval_accuracy": 0.9067077872012336,
|
| 4 |
+
"eval_loss": 0.3908935785293579,
|
| 5 |
+
"eval_runtime": 42.6956,
|
| 6 |
+
"eval_samples_per_second": 60.756,
|
| 7 |
+
"eval_steps_per_second": 7.612
|
| 8 |
+
}
|
runs/Jun27_00-20-05_bernard/events.out.tfevents.1656283610.bernard.1177787.2
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:28f74ae319ab0e71eea1cb91fe0523c05d894201e351057b6b4e9daa4f295b45
|
| 3 |
+
size 363
|
train_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.0,
|
| 3 |
+
"train_loss": 1.0863585265353322,
|
| 4 |
+
"train_runtime": 1526.0713,
|
| 5 |
+
"train_samples_per_second": 26.762,
|
| 6 |
+
"train_steps_per_second": 0.839
|
| 7 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,901 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.3908935785293579,
|
| 3 |
+
"best_model_checkpoint": "./exper_batch_32_e4/checkpoint-1200",
|
| 4 |
+
"epoch": 4.0,
|
| 5 |
+
"global_step": 1280,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.03,
|
| 12 |
+
"learning_rate": 0.00019843750000000002,
|
| 13 |
+
"loss": 4.9283,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.06,
|
| 18 |
+
"learning_rate": 0.000196875,
|
| 19 |
+
"loss": 4.7121,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.09,
|
| 24 |
+
"learning_rate": 0.0001953125,
|
| 25 |
+
"loss": 4.4878,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.12,
|
| 30 |
+
"learning_rate": 0.00019375000000000002,
|
| 31 |
+
"loss": 4.3163,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.16,
|
| 36 |
+
"learning_rate": 0.0001921875,
|
| 37 |
+
"loss": 4.1709,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.19,
|
| 42 |
+
"learning_rate": 0.000190625,
|
| 43 |
+
"loss": 4.0397,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.22,
|
| 48 |
+
"learning_rate": 0.00018906250000000002,
|
| 49 |
+
"loss": 3.8221,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.25,
|
| 54 |
+
"learning_rate": 0.0001875,
|
| 55 |
+
"loss": 3.7326,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.28,
|
| 60 |
+
"learning_rate": 0.0001859375,
|
| 61 |
+
"loss": 3.541,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.31,
|
| 66 |
+
"learning_rate": 0.000184375,
|
| 67 |
+
"loss": 3.4295,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.31,
|
| 72 |
+
"eval_accuracy": 0.28373168851195063,
|
| 73 |
+
"eval_loss": 3.4026525020599365,
|
| 74 |
+
"eval_runtime": 43.01,
|
| 75 |
+
"eval_samples_per_second": 60.312,
|
| 76 |
+
"eval_steps_per_second": 7.556,
|
| 77 |
+
"step": 100
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 0.34,
|
| 81 |
+
"learning_rate": 0.0001828125,
|
| 82 |
+
"loss": 3.2807,
|
| 83 |
+
"step": 110
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 0.38,
|
| 87 |
+
"learning_rate": 0.00018125000000000001,
|
| 88 |
+
"loss": 3.2503,
|
| 89 |
+
"step": 120
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.41,
|
| 93 |
+
"learning_rate": 0.0001796875,
|
| 94 |
+
"loss": 3.0753,
|
| 95 |
+
"step": 130
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 0.44,
|
| 99 |
+
"learning_rate": 0.000178125,
|
| 100 |
+
"loss": 3.0426,
|
| 101 |
+
"step": 140
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.47,
|
| 105 |
+
"learning_rate": 0.00017656250000000002,
|
| 106 |
+
"loss": 2.8475,
|
| 107 |
+
"step": 150
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.5,
|
| 111 |
+
"learning_rate": 0.000175,
|
| 112 |
+
"loss": 2.8298,
|
| 113 |
+
"step": 160
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 0.53,
|
| 117 |
+
"learning_rate": 0.0001734375,
|
| 118 |
+
"loss": 2.7558,
|
| 119 |
+
"step": 170
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 0.56,
|
| 123 |
+
"learning_rate": 0.00017187500000000002,
|
| 124 |
+
"loss": 2.5849,
|
| 125 |
+
"step": 180
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 0.59,
|
| 129 |
+
"learning_rate": 0.0001703125,
|
| 130 |
+
"loss": 2.5335,
|
| 131 |
+
"step": 190
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.62,
|
| 135 |
+
"learning_rate": 0.00016875,
|
| 136 |
+
"loss": 2.5035,
|
| 137 |
+
"step": 200
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.62,
|
| 141 |
+
"eval_accuracy": 0.5246723207401697,
|
| 142 |
+
"eval_loss": 2.433885097503662,
|
| 143 |
+
"eval_runtime": 43.0144,
|
| 144 |
+
"eval_samples_per_second": 60.305,
|
| 145 |
+
"eval_steps_per_second": 7.556,
|
| 146 |
+
"step": 200
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.66,
|
| 150 |
+
"learning_rate": 0.00016718750000000002,
|
| 151 |
+
"loss": 2.4416,
|
| 152 |
+
"step": 210
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.69,
|
| 156 |
+
"learning_rate": 0.000165625,
|
| 157 |
+
"loss": 2.2758,
|
| 158 |
+
"step": 220
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.72,
|
| 162 |
+
"learning_rate": 0.0001640625,
|
| 163 |
+
"loss": 2.2845,
|
| 164 |
+
"step": 230
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.75,
|
| 168 |
+
"learning_rate": 0.00016250000000000002,
|
| 169 |
+
"loss": 2.2053,
|
| 170 |
+
"step": 240
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.78,
|
| 174 |
+
"learning_rate": 0.0001609375,
|
| 175 |
+
"loss": 2.165,
|
| 176 |
+
"step": 250
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.81,
|
| 180 |
+
"learning_rate": 0.000159375,
|
| 181 |
+
"loss": 1.9497,
|
| 182 |
+
"step": 260
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.84,
|
| 186 |
+
"learning_rate": 0.00015781250000000002,
|
| 187 |
+
"loss": 1.9581,
|
| 188 |
+
"step": 270
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.88,
|
| 192 |
+
"learning_rate": 0.00015625,
|
| 193 |
+
"loss": 1.8753,
|
| 194 |
+
"step": 280
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.91,
|
| 198 |
+
"learning_rate": 0.0001546875,
|
| 199 |
+
"loss": 1.7976,
|
| 200 |
+
"step": 290
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.94,
|
| 204 |
+
"learning_rate": 0.000153125,
|
| 205 |
+
"loss": 1.6542,
|
| 206 |
+
"step": 300
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.94,
|
| 210 |
+
"eval_accuracy": 0.6387818041634541,
|
| 211 |
+
"eval_loss": 1.7689646482467651,
|
| 212 |
+
"eval_runtime": 42.9608,
|
| 213 |
+
"eval_samples_per_second": 60.381,
|
| 214 |
+
"eval_steps_per_second": 7.565,
|
| 215 |
+
"step": 300
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.97,
|
| 219 |
+
"learning_rate": 0.0001515625,
|
| 220 |
+
"loss": 1.7968,
|
| 221 |
+
"step": 310
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 1.0,
|
| 225 |
+
"learning_rate": 0.00015000000000000001,
|
| 226 |
+
"loss": 1.84,
|
| 227 |
+
"step": 320
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 1.03,
|
| 231 |
+
"learning_rate": 0.0001484375,
|
| 232 |
+
"loss": 1.5215,
|
| 233 |
+
"step": 330
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.06,
|
| 237 |
+
"learning_rate": 0.000146875,
|
| 238 |
+
"loss": 1.3148,
|
| 239 |
+
"step": 340
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 1.09,
|
| 243 |
+
"learning_rate": 0.00014531250000000002,
|
| 244 |
+
"loss": 1.3781,
|
| 245 |
+
"step": 350
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 1.12,
|
| 249 |
+
"learning_rate": 0.00014375,
|
| 250 |
+
"loss": 1.3547,
|
| 251 |
+
"step": 360
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 1.16,
|
| 255 |
+
"learning_rate": 0.0001421875,
|
| 256 |
+
"loss": 1.2466,
|
| 257 |
+
"step": 370
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 1.19,
|
| 261 |
+
"learning_rate": 0.00014062500000000002,
|
| 262 |
+
"loss": 1.1866,
|
| 263 |
+
"step": 380
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 1.22,
|
| 267 |
+
"learning_rate": 0.0001390625,
|
| 268 |
+
"loss": 1.187,
|
| 269 |
+
"step": 390
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 1.25,
|
| 273 |
+
"learning_rate": 0.0001375,
|
| 274 |
+
"loss": 1.1589,
|
| 275 |
+
"step": 400
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.25,
|
| 279 |
+
"eval_accuracy": 0.7459521973785659,
|
| 280 |
+
"eval_loss": 1.3106426000595093,
|
| 281 |
+
"eval_runtime": 43.0294,
|
| 282 |
+
"eval_samples_per_second": 60.284,
|
| 283 |
+
"eval_steps_per_second": 7.553,
|
| 284 |
+
"step": 400
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.28,
|
| 288 |
+
"learning_rate": 0.00013593750000000002,
|
| 289 |
+
"loss": 1.0901,
|
| 290 |
+
"step": 410
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 1.31,
|
| 294 |
+
"learning_rate": 0.000134375,
|
| 295 |
+
"loss": 1.0998,
|
| 296 |
+
"step": 420
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.34,
|
| 300 |
+
"learning_rate": 0.0001328125,
|
| 301 |
+
"loss": 1.028,
|
| 302 |
+
"step": 430
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 1.38,
|
| 306 |
+
"learning_rate": 0.00013125000000000002,
|
| 307 |
+
"loss": 1.1139,
|
| 308 |
+
"step": 440
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.41,
|
| 312 |
+
"learning_rate": 0.0001296875,
|
| 313 |
+
"loss": 0.9936,
|
| 314 |
+
"step": 450
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 1.44,
|
| 318 |
+
"learning_rate": 0.000128125,
|
| 319 |
+
"loss": 0.9392,
|
| 320 |
+
"step": 460
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 1.47,
|
| 324 |
+
"learning_rate": 0.0001265625,
|
| 325 |
+
"loss": 0.8785,
|
| 326 |
+
"step": 470
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.5,
|
| 330 |
+
"learning_rate": 0.000125,
|
| 331 |
+
"loss": 0.8851,
|
| 332 |
+
"step": 480
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.53,
|
| 336 |
+
"learning_rate": 0.0001234375,
|
| 337 |
+
"loss": 0.8302,
|
| 338 |
+
"step": 490
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.56,
|
| 342 |
+
"learning_rate": 0.00012187500000000001,
|
| 343 |
+
"loss": 0.9363,
|
| 344 |
+
"step": 500
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 1.56,
|
| 348 |
+
"eval_accuracy": 0.7802621434078643,
|
| 349 |
+
"eval_loss": 0.997726321220398,
|
| 350 |
+
"eval_runtime": 42.8785,
|
| 351 |
+
"eval_samples_per_second": 60.496,
|
| 352 |
+
"eval_steps_per_second": 7.58,
|
| 353 |
+
"step": 500
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 1.59,
|
| 357 |
+
"learning_rate": 0.0001203125,
|
| 358 |
+
"loss": 0.826,
|
| 359 |
+
"step": 510
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.62,
|
| 363 |
+
"learning_rate": 0.00011875,
|
| 364 |
+
"loss": 0.8231,
|
| 365 |
+
"step": 520
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 1.66,
|
| 369 |
+
"learning_rate": 0.00011718750000000001,
|
| 370 |
+
"loss": 0.8144,
|
| 371 |
+
"step": 530
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 1.69,
|
| 375 |
+
"learning_rate": 0.000115625,
|
| 376 |
+
"loss": 0.8248,
|
| 377 |
+
"step": 540
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 1.72,
|
| 381 |
+
"learning_rate": 0.0001140625,
|
| 382 |
+
"loss": 0.8278,
|
| 383 |
+
"step": 550
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 1.75,
|
| 387 |
+
"learning_rate": 0.00011250000000000001,
|
| 388 |
+
"loss": 0.8477,
|
| 389 |
+
"step": 560
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.78,
|
| 393 |
+
"learning_rate": 0.0001109375,
|
| 394 |
+
"loss": 0.714,
|
| 395 |
+
"step": 570
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.81,
|
| 399 |
+
"learning_rate": 0.000109375,
|
| 400 |
+
"loss": 0.7819,
|
| 401 |
+
"step": 580
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.84,
|
| 405 |
+
"learning_rate": 0.00010781250000000001,
|
| 406 |
+
"loss": 0.7466,
|
| 407 |
+
"step": 590
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 1.88,
|
| 411 |
+
"learning_rate": 0.00010625000000000001,
|
| 412 |
+
"loss": 0.6946,
|
| 413 |
+
"step": 600
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 1.88,
|
| 417 |
+
"eval_accuracy": 0.8207401696222051,
|
| 418 |
+
"eval_loss": 0.8138005137443542,
|
| 419 |
+
"eval_runtime": 42.9052,
|
| 420 |
+
"eval_samples_per_second": 60.459,
|
| 421 |
+
"eval_steps_per_second": 7.575,
|
| 422 |
+
"step": 600
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.91,
|
| 426 |
+
"learning_rate": 0.0001046875,
|
| 427 |
+
"loss": 0.699,
|
| 428 |
+
"step": 610
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 1.94,
|
| 432 |
+
"learning_rate": 0.000103125,
|
| 433 |
+
"loss": 0.5999,
|
| 434 |
+
"step": 620
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.97,
|
| 438 |
+
"learning_rate": 0.00010156250000000001,
|
| 439 |
+
"loss": 0.5976,
|
| 440 |
+
"step": 630
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 2.0,
|
| 444 |
+
"learning_rate": 0.0001,
|
| 445 |
+
"loss": 0.5776,
|
| 446 |
+
"step": 640
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 2.03,
|
| 450 |
+
"learning_rate": 9.84375e-05,
|
| 451 |
+
"loss": 0.4293,
|
| 452 |
+
"step": 650
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 2.06,
|
| 456 |
+
"learning_rate": 9.687500000000001e-05,
|
| 457 |
+
"loss": 0.4565,
|
| 458 |
+
"step": 660
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 2.09,
|
| 462 |
+
"learning_rate": 9.53125e-05,
|
| 463 |
+
"loss": 0.4259,
|
| 464 |
+
"step": 670
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.12,
|
| 468 |
+
"learning_rate": 9.375e-05,
|
| 469 |
+
"loss": 0.3683,
|
| 470 |
+
"step": 680
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 2.16,
|
| 474 |
+
"learning_rate": 9.21875e-05,
|
| 475 |
+
"loss": 0.4605,
|
| 476 |
+
"step": 690
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 2.19,
|
| 480 |
+
"learning_rate": 9.062500000000001e-05,
|
| 481 |
+
"loss": 0.3488,
|
| 482 |
+
"step": 700
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 2.19,
|
| 486 |
+
"eval_accuracy": 0.848882035466461,
|
| 487 |
+
"eval_loss": 0.6592601537704468,
|
| 488 |
+
"eval_runtime": 42.672,
|
| 489 |
+
"eval_samples_per_second": 60.789,
|
| 490 |
+
"eval_steps_per_second": 7.616,
|
| 491 |
+
"step": 700
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 2.22,
|
| 495 |
+
"learning_rate": 8.90625e-05,
|
| 496 |
+
"loss": 0.4594,
|
| 497 |
+
"step": 710
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 2.25,
|
| 501 |
+
"learning_rate": 8.75e-05,
|
| 502 |
+
"loss": 0.3706,
|
| 503 |
+
"step": 720
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 2.28,
|
| 507 |
+
"learning_rate": 8.593750000000001e-05,
|
| 508 |
+
"loss": 0.3855,
|
| 509 |
+
"step": 730
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 2.31,
|
| 513 |
+
"learning_rate": 8.4375e-05,
|
| 514 |
+
"loss": 0.4005,
|
| 515 |
+
"step": 740
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 2.34,
|
| 519 |
+
"learning_rate": 8.28125e-05,
|
| 520 |
+
"loss": 0.3648,
|
| 521 |
+
"step": 750
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 2.38,
|
| 525 |
+
"learning_rate": 8.125000000000001e-05,
|
| 526 |
+
"loss": 0.3759,
|
| 527 |
+
"step": 760
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.41,
|
| 531 |
+
"learning_rate": 7.96875e-05,
|
| 532 |
+
"loss": 0.3622,
|
| 533 |
+
"step": 770
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 2.44,
|
| 537 |
+
"learning_rate": 7.8125e-05,
|
| 538 |
+
"loss": 0.3209,
|
| 539 |
+
"step": 780
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 2.47,
|
| 543 |
+
"learning_rate": 7.65625e-05,
|
| 544 |
+
"loss": 0.2794,
|
| 545 |
+
"step": 790
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 2.5,
|
| 549 |
+
"learning_rate": 7.500000000000001e-05,
|
| 550 |
+
"loss": 0.2935,
|
| 551 |
+
"step": 800
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 2.5,
|
| 555 |
+
"eval_accuracy": 0.8662297609868929,
|
| 556 |
+
"eval_loss": 0.5725119113922119,
|
| 557 |
+
"eval_runtime": 42.6907,
|
| 558 |
+
"eval_samples_per_second": 60.763,
|
| 559 |
+
"eval_steps_per_second": 7.613,
|
| 560 |
+
"step": 800
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 2.53,
|
| 564 |
+
"learning_rate": 7.34375e-05,
|
| 565 |
+
"loss": 0.3629,
|
| 566 |
+
"step": 810
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 2.56,
|
| 570 |
+
"learning_rate": 7.1875e-05,
|
| 571 |
+
"loss": 0.3006,
|
| 572 |
+
"step": 820
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 2.59,
|
| 576 |
+
"learning_rate": 7.031250000000001e-05,
|
| 577 |
+
"loss": 0.2876,
|
| 578 |
+
"step": 830
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 2.62,
|
| 582 |
+
"learning_rate": 6.875e-05,
|
| 583 |
+
"loss": 0.2612,
|
| 584 |
+
"step": 840
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 2.66,
|
| 588 |
+
"learning_rate": 6.71875e-05,
|
| 589 |
+
"loss": 0.3033,
|
| 590 |
+
"step": 850
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.69,
|
| 594 |
+
"learning_rate": 6.562500000000001e-05,
|
| 595 |
+
"loss": 0.2857,
|
| 596 |
+
"step": 860
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 2.72,
|
| 600 |
+
"learning_rate": 6.40625e-05,
|
| 601 |
+
"loss": 0.2877,
|
| 602 |
+
"step": 870
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 2.75,
|
| 606 |
+
"learning_rate": 6.25e-05,
|
| 607 |
+
"loss": 0.2429,
|
| 608 |
+
"step": 880
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 2.78,
|
| 612 |
+
"learning_rate": 6.0937500000000004e-05,
|
| 613 |
+
"loss": 0.2133,
|
| 614 |
+
"step": 890
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 2.81,
|
| 618 |
+
"learning_rate": 5.9375e-05,
|
| 619 |
+
"loss": 0.2557,
|
| 620 |
+
"step": 900
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 2.81,
|
| 624 |
+
"eval_accuracy": 0.8855050115651504,
|
| 625 |
+
"eval_loss": 0.5088278651237488,
|
| 626 |
+
"eval_runtime": 43.0972,
|
| 627 |
+
"eval_samples_per_second": 60.189,
|
| 628 |
+
"eval_steps_per_second": 7.541,
|
| 629 |
+
"step": 900
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 2.84,
|
| 633 |
+
"learning_rate": 5.78125e-05,
|
| 634 |
+
"loss": 0.3011,
|
| 635 |
+
"step": 910
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 2.88,
|
| 639 |
+
"learning_rate": 5.6250000000000005e-05,
|
| 640 |
+
"loss": 0.2739,
|
| 641 |
+
"step": 920
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 2.91,
|
| 645 |
+
"learning_rate": 5.46875e-05,
|
| 646 |
+
"loss": 0.215,
|
| 647 |
+
"step": 930
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 2.94,
|
| 651 |
+
"learning_rate": 5.3125000000000004e-05,
|
| 652 |
+
"loss": 0.2399,
|
| 653 |
+
"step": 940
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.97,
|
| 657 |
+
"learning_rate": 5.15625e-05,
|
| 658 |
+
"loss": 0.2356,
|
| 659 |
+
"step": 950
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 3.0,
|
| 663 |
+
"learning_rate": 5e-05,
|
| 664 |
+
"loss": 0.2304,
|
| 665 |
+
"step": 960
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 3.03,
|
| 669 |
+
"learning_rate": 4.8437500000000005e-05,
|
| 670 |
+
"loss": 0.21,
|
| 671 |
+
"step": 970
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 3.06,
|
| 675 |
+
"learning_rate": 4.6875e-05,
|
| 676 |
+
"loss": 0.1801,
|
| 677 |
+
"step": 980
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 3.09,
|
| 681 |
+
"learning_rate": 4.5312500000000004e-05,
|
| 682 |
+
"loss": 0.12,
|
| 683 |
+
"step": 990
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 3.12,
|
| 687 |
+
"learning_rate": 4.375e-05,
|
| 688 |
+
"loss": 0.1509,
|
| 689 |
+
"step": 1000
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 3.12,
|
| 693 |
+
"eval_accuracy": 0.8970701619121049,
|
| 694 |
+
"eval_loss": 0.4571980834007263,
|
| 695 |
+
"eval_runtime": 42.5823,
|
| 696 |
+
"eval_samples_per_second": 60.917,
|
| 697 |
+
"eval_steps_per_second": 7.632,
|
| 698 |
+
"step": 1000
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 3.16,
|
| 702 |
+
"learning_rate": 4.21875e-05,
|
| 703 |
+
"loss": 0.1448,
|
| 704 |
+
"step": 1010
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 3.19,
|
| 708 |
+
"learning_rate": 4.0625000000000005e-05,
|
| 709 |
+
"loss": 0.1272,
|
| 710 |
+
"step": 1020
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 3.22,
|
| 714 |
+
"learning_rate": 3.90625e-05,
|
| 715 |
+
"loss": 0.1215,
|
| 716 |
+
"step": 1030
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.25,
|
| 720 |
+
"learning_rate": 3.7500000000000003e-05,
|
| 721 |
+
"loss": 0.1236,
|
| 722 |
+
"step": 1040
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 3.28,
|
| 726 |
+
"learning_rate": 3.59375e-05,
|
| 727 |
+
"loss": 0.1364,
|
| 728 |
+
"step": 1050
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 3.31,
|
| 732 |
+
"learning_rate": 3.4375e-05,
|
| 733 |
+
"loss": 0.1074,
|
| 734 |
+
"step": 1060
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 3.34,
|
| 738 |
+
"learning_rate": 3.2812500000000005e-05,
|
| 739 |
+
"loss": 0.1383,
|
| 740 |
+
"step": 1070
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 3.38,
|
| 744 |
+
"learning_rate": 3.125e-05,
|
| 745 |
+
"loss": 0.1304,
|
| 746 |
+
"step": 1080
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 3.41,
|
| 750 |
+
"learning_rate": 2.96875e-05,
|
| 751 |
+
"loss": 0.1431,
|
| 752 |
+
"step": 1090
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 3.44,
|
| 756 |
+
"learning_rate": 2.8125000000000003e-05,
|
| 757 |
+
"loss": 0.1367,
|
| 758 |
+
"step": 1100
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.44,
|
| 762 |
+
"eval_accuracy": 0.9090208172706246,
|
| 763 |
+
"eval_loss": 0.4128749370574951,
|
| 764 |
+
"eval_runtime": 42.4657,
|
| 765 |
+
"eval_samples_per_second": 61.085,
|
| 766 |
+
"eval_steps_per_second": 7.653,
|
| 767 |
+
"step": 1100
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 3.47,
|
| 771 |
+
"learning_rate": 2.6562500000000002e-05,
|
| 772 |
+
"loss": 0.1202,
|
| 773 |
+
"step": 1110
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 3.5,
|
| 777 |
+
"learning_rate": 2.5e-05,
|
| 778 |
+
"loss": 0.1309,
|
| 779 |
+
"step": 1120
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.53,
|
| 783 |
+
"learning_rate": 2.34375e-05,
|
| 784 |
+
"loss": 0.1081,
|
| 785 |
+
"step": 1130
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 3.56,
|
| 789 |
+
"learning_rate": 2.1875e-05,
|
| 790 |
+
"loss": 0.144,
|
| 791 |
+
"step": 1140
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 3.59,
|
| 795 |
+
"learning_rate": 2.0312500000000002e-05,
|
| 796 |
+
"loss": 0.1104,
|
| 797 |
+
"step": 1150
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 3.62,
|
| 801 |
+
"learning_rate": 1.8750000000000002e-05,
|
| 802 |
+
"loss": 0.0947,
|
| 803 |
+
"step": 1160
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 3.66,
|
| 807 |
+
"learning_rate": 1.71875e-05,
|
| 808 |
+
"loss": 0.1008,
|
| 809 |
+
"step": 1170
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 3.69,
|
| 813 |
+
"learning_rate": 1.5625e-05,
|
| 814 |
+
"loss": 0.1177,
|
| 815 |
+
"step": 1180
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 3.72,
|
| 819 |
+
"learning_rate": 1.4062500000000001e-05,
|
| 820 |
+
"loss": 0.1347,
|
| 821 |
+
"step": 1190
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.75,
|
| 825 |
+
"learning_rate": 1.25e-05,
|
| 826 |
+
"loss": 0.1078,
|
| 827 |
+
"step": 1200
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 3.75,
|
| 831 |
+
"eval_accuracy": 0.9067077872012336,
|
| 832 |
+
"eval_loss": 0.3908935785293579,
|
| 833 |
+
"eval_runtime": 42.3296,
|
| 834 |
+
"eval_samples_per_second": 61.281,
|
| 835 |
+
"eval_steps_per_second": 7.678,
|
| 836 |
+
"step": 1200
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 3.78,
|
| 840 |
+
"learning_rate": 1.09375e-05,
|
| 841 |
+
"loss": 0.1007,
|
| 842 |
+
"step": 1210
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.81,
|
| 846 |
+
"learning_rate": 9.375000000000001e-06,
|
| 847 |
+
"loss": 0.1205,
|
| 848 |
+
"step": 1220
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 3.84,
|
| 852 |
+
"learning_rate": 7.8125e-06,
|
| 853 |
+
"loss": 0.1124,
|
| 854 |
+
"step": 1230
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 3.88,
|
| 858 |
+
"learning_rate": 6.25e-06,
|
| 859 |
+
"loss": 0.1064,
|
| 860 |
+
"step": 1240
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 3.91,
|
| 864 |
+
"learning_rate": 4.6875000000000004e-06,
|
| 865 |
+
"loss": 0.1117,
|
| 866 |
+
"step": 1250
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 3.94,
|
| 870 |
+
"learning_rate": 3.125e-06,
|
| 871 |
+
"loss": 0.1043,
|
| 872 |
+
"step": 1260
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 3.97,
|
| 876 |
+
"learning_rate": 1.5625e-06,
|
| 877 |
+
"loss": 0.1033,
|
| 878 |
+
"step": 1270
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 4.0,
|
| 882 |
+
"learning_rate": 0.0,
|
| 883 |
+
"loss": 0.1101,
|
| 884 |
+
"step": 1280
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 4.0,
|
| 888 |
+
"step": 1280,
|
| 889 |
+
"total_flos": 3.1689424899978854e+18,
|
| 890 |
+
"train_loss": 1.0863585265353322,
|
| 891 |
+
"train_runtime": 1526.0713,
|
| 892 |
+
"train_samples_per_second": 26.762,
|
| 893 |
+
"train_steps_per_second": 0.839
|
| 894 |
+
}
|
| 895 |
+
],
|
| 896 |
+
"max_steps": 1280,
|
| 897 |
+
"num_train_epochs": 4,
|
| 898 |
+
"total_flos": 3.1689424899978854e+18,
|
| 899 |
+
"trial_name": null,
|
| 900 |
+
"trial_params": null
|
| 901 |
+
}
|