🍻 cheers
Browse files- all_results.json +10 -5
- eval_results.json +8 -0
- train_results.json +5 -5
- trainer_state.json +1099 -0
- validation_results.json +8 -0
all_results.json
CHANGED
|
@@ -1,8 +1,13 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 2.0,
|
| 3 |
-
"
|
| 4 |
-
"
|
| 5 |
-
"
|
| 6 |
-
"
|
| 7 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 2.0,
|
| 3 |
+
"eval_accuracy": 0.43737166324435317,
|
| 4 |
+
"eval_loss": 2.7726495265960693,
|
| 5 |
+
"eval_runtime": 6.4303,
|
| 6 |
+
"eval_samples_per_second": 75.735,
|
| 7 |
+
"eval_steps_per_second": 9.486,
|
| 8 |
+
"total_flos": 1.1959742538326016e+17,
|
| 9 |
+
"train_loss": 3.1177605087162594,
|
| 10 |
+
"train_runtime": 412.2299,
|
| 11 |
+
"train_samples_per_second": 7.549,
|
| 12 |
+
"train_steps_per_second": 3.775
|
| 13 |
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.0,
|
| 3 |
+
"eval_accuracy": 0.43737166324435317,
|
| 4 |
+
"eval_loss": 2.7726495265960693,
|
| 5 |
+
"eval_runtime": 6.4303,
|
| 6 |
+
"eval_samples_per_second": 75.735,
|
| 7 |
+
"eval_steps_per_second": 9.486
|
| 8 |
+
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 2.0,
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss":
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 2.0,
|
| 3 |
+
"total_flos": 1.1959742538326016e+17,
|
| 4 |
+
"train_loss": 3.1177605087162594,
|
| 5 |
+
"train_runtime": 412.2299,
|
| 6 |
+
"train_samples_per_second": 7.549,
|
| 7 |
+
"train_steps_per_second": 3.775
|
| 8 |
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1099 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 2.783235788345337,
|
| 3 |
+
"best_model_checkpoint": "/content/Train/checkpoint-1400",
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
+
"global_step": 1556,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.01,
|
| 12 |
+
"learning_rate": 0.0001994858611825193,
|
| 13 |
+
"loss": 3.2724,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.03,
|
| 18 |
+
"learning_rate": 0.00019820051413881748,
|
| 19 |
+
"loss": 3.2838,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.04,
|
| 24 |
+
"learning_rate": 0.0001969151670951157,
|
| 25 |
+
"loss": 3.3858,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.05,
|
| 30 |
+
"learning_rate": 0.00019562982005141388,
|
| 31 |
+
"loss": 3.1207,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.06,
|
| 36 |
+
"learning_rate": 0.0001943444730077121,
|
| 37 |
+
"loss": 3.4414,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.08,
|
| 42 |
+
"learning_rate": 0.0001930591259640103,
|
| 43 |
+
"loss": 3.2139,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.09,
|
| 48 |
+
"learning_rate": 0.0001917737789203085,
|
| 49 |
+
"loss": 3.1755,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.1,
|
| 54 |
+
"learning_rate": 0.0001904884318766067,
|
| 55 |
+
"loss": 3.7297,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.12,
|
| 60 |
+
"learning_rate": 0.0001892030848329049,
|
| 61 |
+
"loss": 3.4405,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.13,
|
| 66 |
+
"learning_rate": 0.0001879177377892031,
|
| 67 |
+
"loss": 3.1529,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.13,
|
| 72 |
+
"eval_accuracy": 0.05897435897435897,
|
| 73 |
+
"eval_loss": 3.211608648300171,
|
| 74 |
+
"eval_runtime": 5.3959,
|
| 75 |
+
"eval_samples_per_second": 72.278,
|
| 76 |
+
"eval_steps_per_second": 9.081,
|
| 77 |
+
"step": 100
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 0.14,
|
| 81 |
+
"learning_rate": 0.0001866323907455013,
|
| 82 |
+
"loss": 3.1837,
|
| 83 |
+
"step": 110
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 0.15,
|
| 87 |
+
"learning_rate": 0.0001853470437017995,
|
| 88 |
+
"loss": 3.2418,
|
| 89 |
+
"step": 120
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.17,
|
| 93 |
+
"learning_rate": 0.0001840616966580977,
|
| 94 |
+
"loss": 3.2269,
|
| 95 |
+
"step": 130
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 0.18,
|
| 99 |
+
"learning_rate": 0.0001827763496143959,
|
| 100 |
+
"loss": 3.1065,
|
| 101 |
+
"step": 140
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.19,
|
| 105 |
+
"learning_rate": 0.00018149100257069408,
|
| 106 |
+
"loss": 3.3563,
|
| 107 |
+
"step": 150
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.21,
|
| 111 |
+
"learning_rate": 0.0001803341902313625,
|
| 112 |
+
"loss": 3.2934,
|
| 113 |
+
"step": 160
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 0.22,
|
| 117 |
+
"learning_rate": 0.00017904884318766068,
|
| 118 |
+
"loss": 3.4139,
|
| 119 |
+
"step": 170
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 0.23,
|
| 123 |
+
"learning_rate": 0.0001777634961439589,
|
| 124 |
+
"loss": 3.2532,
|
| 125 |
+
"step": 180
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 0.24,
|
| 129 |
+
"learning_rate": 0.00017647814910025708,
|
| 130 |
+
"loss": 3.2196,
|
| 131 |
+
"step": 190
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.26,
|
| 135 |
+
"learning_rate": 0.00017519280205655527,
|
| 136 |
+
"loss": 3.1644,
|
| 137 |
+
"step": 200
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.26,
|
| 141 |
+
"eval_accuracy": 0.06923076923076923,
|
| 142 |
+
"eval_loss": 3.209580421447754,
|
| 143 |
+
"eval_runtime": 4.8409,
|
| 144 |
+
"eval_samples_per_second": 80.564,
|
| 145 |
+
"eval_steps_per_second": 10.122,
|
| 146 |
+
"step": 200
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.27,
|
| 150 |
+
"learning_rate": 0.00017390745501285349,
|
| 151 |
+
"loss": 3.3494,
|
| 152 |
+
"step": 210
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.28,
|
| 156 |
+
"learning_rate": 0.00017262210796915167,
|
| 157 |
+
"loss": 3.336,
|
| 158 |
+
"step": 220
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.3,
|
| 162 |
+
"learning_rate": 0.0001713367609254499,
|
| 163 |
+
"loss": 3.2151,
|
| 164 |
+
"step": 230
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.31,
|
| 168 |
+
"learning_rate": 0.00017005141388174808,
|
| 169 |
+
"loss": 3.2835,
|
| 170 |
+
"step": 240
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.32,
|
| 174 |
+
"learning_rate": 0.0001687660668380463,
|
| 175 |
+
"loss": 3.4358,
|
| 176 |
+
"step": 250
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.33,
|
| 180 |
+
"learning_rate": 0.00016748071979434448,
|
| 181 |
+
"loss": 3.303,
|
| 182 |
+
"step": 260
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.35,
|
| 186 |
+
"learning_rate": 0.00016619537275064267,
|
| 187 |
+
"loss": 3.2475,
|
| 188 |
+
"step": 270
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.36,
|
| 192 |
+
"learning_rate": 0.00016491002570694088,
|
| 193 |
+
"loss": 3.1522,
|
| 194 |
+
"step": 280
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.37,
|
| 198 |
+
"learning_rate": 0.00016362467866323907,
|
| 199 |
+
"loss": 3.2906,
|
| 200 |
+
"step": 290
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.39,
|
| 204 |
+
"learning_rate": 0.00016233933161953728,
|
| 205 |
+
"loss": 3.1549,
|
| 206 |
+
"step": 300
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.39,
|
| 210 |
+
"eval_accuracy": 0.06923076923076923,
|
| 211 |
+
"eval_loss": 3.1985976696014404,
|
| 212 |
+
"eval_runtime": 4.6899,
|
| 213 |
+
"eval_samples_per_second": 83.157,
|
| 214 |
+
"eval_steps_per_second": 10.448,
|
| 215 |
+
"step": 300
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.4,
|
| 219 |
+
"learning_rate": 0.0001610539845758355,
|
| 220 |
+
"loss": 3.4558,
|
| 221 |
+
"step": 310
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.41,
|
| 225 |
+
"learning_rate": 0.00015976863753213369,
|
| 226 |
+
"loss": 3.2304,
|
| 227 |
+
"step": 320
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.42,
|
| 231 |
+
"learning_rate": 0.0001584832904884319,
|
| 232 |
+
"loss": 3.2686,
|
| 233 |
+
"step": 330
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.44,
|
| 237 |
+
"learning_rate": 0.0001571979434447301,
|
| 238 |
+
"loss": 3.3832,
|
| 239 |
+
"step": 340
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.45,
|
| 243 |
+
"learning_rate": 0.00015591259640102828,
|
| 244 |
+
"loss": 3.3443,
|
| 245 |
+
"step": 350
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 0.46,
|
| 249 |
+
"learning_rate": 0.0001546272493573265,
|
| 250 |
+
"loss": 3.2797,
|
| 251 |
+
"step": 360
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.48,
|
| 255 |
+
"learning_rate": 0.00015334190231362468,
|
| 256 |
+
"loss": 3.2673,
|
| 257 |
+
"step": 370
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.49,
|
| 261 |
+
"learning_rate": 0.0001520565552699229,
|
| 262 |
+
"loss": 3.1919,
|
| 263 |
+
"step": 380
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 0.5,
|
| 267 |
+
"learning_rate": 0.00015077120822622108,
|
| 268 |
+
"loss": 3.1238,
|
| 269 |
+
"step": 390
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.51,
|
| 273 |
+
"learning_rate": 0.0001494858611825193,
|
| 274 |
+
"loss": 3.2998,
|
| 275 |
+
"step": 400
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.51,
|
| 279 |
+
"eval_accuracy": 0.1076923076923077,
|
| 280 |
+
"eval_loss": 3.1967546939849854,
|
| 281 |
+
"eval_runtime": 4.6763,
|
| 282 |
+
"eval_samples_per_second": 83.399,
|
| 283 |
+
"eval_steps_per_second": 10.478,
|
| 284 |
+
"step": 400
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.53,
|
| 288 |
+
"learning_rate": 0.00014820051413881748,
|
| 289 |
+
"loss": 3.237,
|
| 290 |
+
"step": 410
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.54,
|
| 294 |
+
"learning_rate": 0.00014691516709511567,
|
| 295 |
+
"loss": 3.1511,
|
| 296 |
+
"step": 420
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.55,
|
| 300 |
+
"learning_rate": 0.00014562982005141388,
|
| 301 |
+
"loss": 3.1972,
|
| 302 |
+
"step": 430
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.57,
|
| 306 |
+
"learning_rate": 0.0001443444730077121,
|
| 307 |
+
"loss": 3.3029,
|
| 308 |
+
"step": 440
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.58,
|
| 312 |
+
"learning_rate": 0.0001430591259640103,
|
| 313 |
+
"loss": 3.2639,
|
| 314 |
+
"step": 450
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.59,
|
| 318 |
+
"learning_rate": 0.0001417737789203085,
|
| 319 |
+
"loss": 3.2602,
|
| 320 |
+
"step": 460
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.6,
|
| 324 |
+
"learning_rate": 0.0001404884318766067,
|
| 325 |
+
"loss": 3.1096,
|
| 326 |
+
"step": 470
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.62,
|
| 330 |
+
"learning_rate": 0.0001392030848329049,
|
| 331 |
+
"loss": 3.2692,
|
| 332 |
+
"step": 480
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.63,
|
| 336 |
+
"learning_rate": 0.0001379177377892031,
|
| 337 |
+
"loss": 3.3299,
|
| 338 |
+
"step": 490
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.64,
|
| 342 |
+
"learning_rate": 0.0001366323907455013,
|
| 343 |
+
"loss": 3.1344,
|
| 344 |
+
"step": 500
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.64,
|
| 348 |
+
"eval_accuracy": 0.1717948717948718,
|
| 349 |
+
"eval_loss": 3.1666767597198486,
|
| 350 |
+
"eval_runtime": 4.8058,
|
| 351 |
+
"eval_samples_per_second": 81.152,
|
| 352 |
+
"eval_steps_per_second": 10.196,
|
| 353 |
+
"step": 500
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.66,
|
| 357 |
+
"learning_rate": 0.0001353470437017995,
|
| 358 |
+
"loss": 3.0552,
|
| 359 |
+
"step": 510
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.67,
|
| 363 |
+
"learning_rate": 0.00013406169665809768,
|
| 364 |
+
"loss": 3.3452,
|
| 365 |
+
"step": 520
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 0.68,
|
| 369 |
+
"learning_rate": 0.0001327763496143959,
|
| 370 |
+
"loss": 3.3108,
|
| 371 |
+
"step": 530
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 0.69,
|
| 375 |
+
"learning_rate": 0.00013149100257069408,
|
| 376 |
+
"loss": 3.0806,
|
| 377 |
+
"step": 540
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.71,
|
| 381 |
+
"learning_rate": 0.0001302056555269923,
|
| 382 |
+
"loss": 3.4034,
|
| 383 |
+
"step": 550
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 0.72,
|
| 387 |
+
"learning_rate": 0.00012892030848329049,
|
| 388 |
+
"loss": 3.0853,
|
| 389 |
+
"step": 560
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 0.73,
|
| 393 |
+
"learning_rate": 0.00012763496143958867,
|
| 394 |
+
"loss": 3.1718,
|
| 395 |
+
"step": 570
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.75,
|
| 399 |
+
"learning_rate": 0.00012634961439588692,
|
| 400 |
+
"loss": 3.1265,
|
| 401 |
+
"step": 580
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.76,
|
| 405 |
+
"learning_rate": 0.0001250642673521851,
|
| 406 |
+
"loss": 3.183,
|
| 407 |
+
"step": 590
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 0.77,
|
| 411 |
+
"learning_rate": 0.0001237789203084833,
|
| 412 |
+
"loss": 3.3638,
|
| 413 |
+
"step": 600
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.77,
|
| 417 |
+
"eval_accuracy": 0.17435897435897435,
|
| 418 |
+
"eval_loss": 3.1383864879608154,
|
| 419 |
+
"eval_runtime": 4.7872,
|
| 420 |
+
"eval_samples_per_second": 81.466,
|
| 421 |
+
"eval_steps_per_second": 10.236,
|
| 422 |
+
"step": 600
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.78,
|
| 426 |
+
"learning_rate": 0.0001224935732647815,
|
| 427 |
+
"loss": 3.1725,
|
| 428 |
+
"step": 610
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.8,
|
| 432 |
+
"learning_rate": 0.0001212082262210797,
|
| 433 |
+
"loss": 3.5329,
|
| 434 |
+
"step": 620
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.81,
|
| 438 |
+
"learning_rate": 0.0001199228791773779,
|
| 439 |
+
"loss": 3.1399,
|
| 440 |
+
"step": 630
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.82,
|
| 444 |
+
"learning_rate": 0.0001186375321336761,
|
| 445 |
+
"loss": 3.1998,
|
| 446 |
+
"step": 640
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.84,
|
| 450 |
+
"learning_rate": 0.0001173521850899743,
|
| 451 |
+
"loss": 3.1562,
|
| 452 |
+
"step": 650
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.85,
|
| 456 |
+
"learning_rate": 0.0001160668380462725,
|
| 457 |
+
"loss": 3.3428,
|
| 458 |
+
"step": 660
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.86,
|
| 462 |
+
"learning_rate": 0.0001147814910025707,
|
| 463 |
+
"loss": 3.0586,
|
| 464 |
+
"step": 670
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.87,
|
| 468 |
+
"learning_rate": 0.0001134961439588689,
|
| 469 |
+
"loss": 3.0464,
|
| 470 |
+
"step": 680
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.89,
|
| 474 |
+
"learning_rate": 0.00011221079691516709,
|
| 475 |
+
"loss": 3.1097,
|
| 476 |
+
"step": 690
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.9,
|
| 480 |
+
"learning_rate": 0.00011092544987146529,
|
| 481 |
+
"loss": 3.1482,
|
| 482 |
+
"step": 700
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.9,
|
| 486 |
+
"eval_accuracy": 0.20256410256410257,
|
| 487 |
+
"eval_loss": 3.096599578857422,
|
| 488 |
+
"eval_runtime": 4.7875,
|
| 489 |
+
"eval_samples_per_second": 81.462,
|
| 490 |
+
"eval_steps_per_second": 10.235,
|
| 491 |
+
"step": 700
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 0.91,
|
| 495 |
+
"learning_rate": 0.00010964010282776349,
|
| 496 |
+
"loss": 3.1459,
|
| 497 |
+
"step": 710
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 0.93,
|
| 501 |
+
"learning_rate": 0.00010835475578406172,
|
| 502 |
+
"loss": 3.1789,
|
| 503 |
+
"step": 720
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 0.94,
|
| 507 |
+
"learning_rate": 0.0001070694087403599,
|
| 508 |
+
"loss": 3.2216,
|
| 509 |
+
"step": 730
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 0.95,
|
| 513 |
+
"learning_rate": 0.0001057840616966581,
|
| 514 |
+
"loss": 2.8712,
|
| 515 |
+
"step": 740
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.96,
|
| 519 |
+
"learning_rate": 0.00010449871465295631,
|
| 520 |
+
"loss": 3.1489,
|
| 521 |
+
"step": 750
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.98,
|
| 525 |
+
"learning_rate": 0.00010321336760925451,
|
| 526 |
+
"loss": 3.438,
|
| 527 |
+
"step": 760
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.99,
|
| 531 |
+
"learning_rate": 0.00010192802056555271,
|
| 532 |
+
"loss": 3.23,
|
| 533 |
+
"step": 770
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 1.0,
|
| 537 |
+
"learning_rate": 0.00010064267352185091,
|
| 538 |
+
"loss": 3.1966,
|
| 539 |
+
"step": 780
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 1.02,
|
| 543 |
+
"learning_rate": 9.93573264781491e-05,
|
| 544 |
+
"loss": 3.0066,
|
| 545 |
+
"step": 790
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 1.03,
|
| 549 |
+
"learning_rate": 9.80719794344473e-05,
|
| 550 |
+
"loss": 3.1366,
|
| 551 |
+
"step": 800
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 1.03,
|
| 555 |
+
"eval_accuracy": 0.18974358974358974,
|
| 556 |
+
"eval_loss": 3.0483875274658203,
|
| 557 |
+
"eval_runtime": 4.6499,
|
| 558 |
+
"eval_samples_per_second": 83.872,
|
| 559 |
+
"eval_steps_per_second": 10.538,
|
| 560 |
+
"step": 800
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 1.04,
|
| 564 |
+
"learning_rate": 9.67866323907455e-05,
|
| 565 |
+
"loss": 3.1106,
|
| 566 |
+
"step": 810
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 1.05,
|
| 570 |
+
"learning_rate": 9.550128534704372e-05,
|
| 571 |
+
"loss": 2.9235,
|
| 572 |
+
"step": 820
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 1.07,
|
| 576 |
+
"learning_rate": 9.421593830334192e-05,
|
| 577 |
+
"loss": 3.0565,
|
| 578 |
+
"step": 830
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.08,
|
| 582 |
+
"learning_rate": 9.29305912596401e-05,
|
| 583 |
+
"loss": 3.0503,
|
| 584 |
+
"step": 840
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.09,
|
| 588 |
+
"learning_rate": 9.16452442159383e-05,
|
| 589 |
+
"loss": 3.2726,
|
| 590 |
+
"step": 850
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.11,
|
| 594 |
+
"learning_rate": 9.03598971722365e-05,
|
| 595 |
+
"loss": 3.0256,
|
| 596 |
+
"step": 860
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 1.12,
|
| 600 |
+
"learning_rate": 8.907455012853471e-05,
|
| 601 |
+
"loss": 3.0892,
|
| 602 |
+
"step": 870
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 1.13,
|
| 606 |
+
"learning_rate": 8.778920308483291e-05,
|
| 607 |
+
"loss": 3.1357,
|
| 608 |
+
"step": 880
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 1.14,
|
| 612 |
+
"learning_rate": 8.650385604113111e-05,
|
| 613 |
+
"loss": 3.1402,
|
| 614 |
+
"step": 890
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 1.16,
|
| 618 |
+
"learning_rate": 8.521850899742931e-05,
|
| 619 |
+
"loss": 3.0206,
|
| 620 |
+
"step": 900
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.16,
|
| 624 |
+
"eval_accuracy": 0.30256410256410254,
|
| 625 |
+
"eval_loss": 3.016371250152588,
|
| 626 |
+
"eval_runtime": 4.6485,
|
| 627 |
+
"eval_samples_per_second": 83.899,
|
| 628 |
+
"eval_steps_per_second": 10.541,
|
| 629 |
+
"step": 900
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 1.17,
|
| 633 |
+
"learning_rate": 8.393316195372751e-05,
|
| 634 |
+
"loss": 3.0057,
|
| 635 |
+
"step": 910
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 1.18,
|
| 639 |
+
"learning_rate": 8.264781491002571e-05,
|
| 640 |
+
"loss": 3.1709,
|
| 641 |
+
"step": 920
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.2,
|
| 645 |
+
"learning_rate": 8.136246786632391e-05,
|
| 646 |
+
"loss": 3.2243,
|
| 647 |
+
"step": 930
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.21,
|
| 651 |
+
"learning_rate": 8.007712082262212e-05,
|
| 652 |
+
"loss": 3.19,
|
| 653 |
+
"step": 940
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.22,
|
| 657 |
+
"learning_rate": 7.87917737789203e-05,
|
| 658 |
+
"loss": 3.2785,
|
| 659 |
+
"step": 950
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 1.23,
|
| 663 |
+
"learning_rate": 7.750642673521852e-05,
|
| 664 |
+
"loss": 2.8424,
|
| 665 |
+
"step": 960
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 1.25,
|
| 669 |
+
"learning_rate": 7.622107969151672e-05,
|
| 670 |
+
"loss": 3.2575,
|
| 671 |
+
"step": 970
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 1.26,
|
| 675 |
+
"learning_rate": 7.493573264781492e-05,
|
| 676 |
+
"loss": 2.9012,
|
| 677 |
+
"step": 980
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 1.27,
|
| 681 |
+
"learning_rate": 7.365038560411311e-05,
|
| 682 |
+
"loss": 3.1516,
|
| 683 |
+
"step": 990
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 1.29,
|
| 687 |
+
"learning_rate": 7.236503856041131e-05,
|
| 688 |
+
"loss": 2.921,
|
| 689 |
+
"step": 1000
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.29,
|
| 693 |
+
"eval_accuracy": 0.3230769230769231,
|
| 694 |
+
"eval_loss": 2.984600305557251,
|
| 695 |
+
"eval_runtime": 4.9539,
|
| 696 |
+
"eval_samples_per_second": 78.726,
|
| 697 |
+
"eval_steps_per_second": 9.891,
|
| 698 |
+
"step": 1000
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 1.3,
|
| 702 |
+
"learning_rate": 7.107969151670951e-05,
|
| 703 |
+
"loss": 2.9828,
|
| 704 |
+
"step": 1010
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.31,
|
| 708 |
+
"learning_rate": 6.979434447300771e-05,
|
| 709 |
+
"loss": 3.0413,
|
| 710 |
+
"step": 1020
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.32,
|
| 714 |
+
"learning_rate": 6.850899742930593e-05,
|
| 715 |
+
"loss": 2.9759,
|
| 716 |
+
"step": 1030
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.34,
|
| 720 |
+
"learning_rate": 6.722365038560411e-05,
|
| 721 |
+
"loss": 2.9434,
|
| 722 |
+
"step": 1040
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 1.35,
|
| 726 |
+
"learning_rate": 6.593830334190231e-05,
|
| 727 |
+
"loss": 3.0893,
|
| 728 |
+
"step": 1050
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 1.36,
|
| 732 |
+
"learning_rate": 6.465295629820052e-05,
|
| 733 |
+
"loss": 3.0797,
|
| 734 |
+
"step": 1060
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 1.38,
|
| 738 |
+
"learning_rate": 6.336760925449872e-05,
|
| 739 |
+
"loss": 2.9291,
|
| 740 |
+
"step": 1070
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 1.39,
|
| 744 |
+
"learning_rate": 6.208226221079692e-05,
|
| 745 |
+
"loss": 3.0357,
|
| 746 |
+
"step": 1080
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.4,
|
| 750 |
+
"learning_rate": 6.079691516709511e-05,
|
| 751 |
+
"loss": 3.0109,
|
| 752 |
+
"step": 1090
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.41,
|
| 756 |
+
"learning_rate": 5.951156812339333e-05,
|
| 757 |
+
"loss": 3.0027,
|
| 758 |
+
"step": 1100
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.41,
|
| 762 |
+
"eval_accuracy": 0.33589743589743587,
|
| 763 |
+
"eval_loss": 2.933824062347412,
|
| 764 |
+
"eval_runtime": 4.8337,
|
| 765 |
+
"eval_samples_per_second": 80.684,
|
| 766 |
+
"eval_steps_per_second": 10.137,
|
| 767 |
+
"step": 1100
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.43,
|
| 771 |
+
"learning_rate": 5.822622107969152e-05,
|
| 772 |
+
"loss": 3.2462,
|
| 773 |
+
"step": 1110
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.44,
|
| 777 |
+
"learning_rate": 5.694087403598972e-05,
|
| 778 |
+
"loss": 2.9376,
|
| 779 |
+
"step": 1120
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.45,
|
| 783 |
+
"learning_rate": 5.5655526992287924e-05,
|
| 784 |
+
"loss": 2.9242,
|
| 785 |
+
"step": 1130
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 1.47,
|
| 789 |
+
"learning_rate": 5.437017994858612e-05,
|
| 790 |
+
"loss": 2.862,
|
| 791 |
+
"step": 1140
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 1.48,
|
| 795 |
+
"learning_rate": 5.308483290488432e-05,
|
| 796 |
+
"loss": 3.1005,
|
| 797 |
+
"step": 1150
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 1.49,
|
| 801 |
+
"learning_rate": 5.1799485861182514e-05,
|
| 802 |
+
"loss": 3.1811,
|
| 803 |
+
"step": 1160
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 1.5,
|
| 807 |
+
"learning_rate": 5.051413881748073e-05,
|
| 808 |
+
"loss": 3.1254,
|
| 809 |
+
"step": 1170
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 1.52,
|
| 813 |
+
"learning_rate": 4.922879177377892e-05,
|
| 814 |
+
"loss": 2.797,
|
| 815 |
+
"step": 1180
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.53,
|
| 819 |
+
"learning_rate": 4.7943444730077124e-05,
|
| 820 |
+
"loss": 3.1899,
|
| 821 |
+
"step": 1190
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.54,
|
| 825 |
+
"learning_rate": 4.6658097686375325e-05,
|
| 826 |
+
"loss": 2.9047,
|
| 827 |
+
"step": 1200
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 1.54,
|
| 831 |
+
"eval_accuracy": 0.34615384615384615,
|
| 832 |
+
"eval_loss": 2.8916842937469482,
|
| 833 |
+
"eval_runtime": 4.6342,
|
| 834 |
+
"eval_samples_per_second": 84.157,
|
| 835 |
+
"eval_steps_per_second": 10.574,
|
| 836 |
+
"step": 1200
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.56,
|
| 840 |
+
"learning_rate": 4.537275064267352e-05,
|
| 841 |
+
"loss": 3.1712,
|
| 842 |
+
"step": 1210
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.57,
|
| 846 |
+
"learning_rate": 4.408740359897173e-05,
|
| 847 |
+
"loss": 2.9201,
|
| 848 |
+
"step": 1220
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 1.58,
|
| 852 |
+
"learning_rate": 4.280205655526993e-05,
|
| 853 |
+
"loss": 2.9952,
|
| 854 |
+
"step": 1230
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 1.59,
|
| 858 |
+
"learning_rate": 4.151670951156812e-05,
|
| 859 |
+
"loss": 2.8225,
|
| 860 |
+
"step": 1240
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 1.61,
|
| 864 |
+
"learning_rate": 4.0231362467866324e-05,
|
| 865 |
+
"loss": 2.9507,
|
| 866 |
+
"step": 1250
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 1.62,
|
| 870 |
+
"learning_rate": 3.8946015424164526e-05,
|
| 871 |
+
"loss": 2.96,
|
| 872 |
+
"step": 1260
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.63,
|
| 876 |
+
"learning_rate": 3.766066838046273e-05,
|
| 877 |
+
"loss": 3.0775,
|
| 878 |
+
"step": 1270
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.65,
|
| 882 |
+
"learning_rate": 3.637532133676093e-05,
|
| 883 |
+
"loss": 2.9242,
|
| 884 |
+
"step": 1280
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.66,
|
| 888 |
+
"learning_rate": 3.508997429305913e-05,
|
| 889 |
+
"loss": 2.8814,
|
| 890 |
+
"step": 1290
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 1.67,
|
| 894 |
+
"learning_rate": 3.380462724935733e-05,
|
| 895 |
+
"loss": 2.8579,
|
| 896 |
+
"step": 1300
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 1.67,
|
| 900 |
+
"eval_accuracy": 0.4025641025641026,
|
| 901 |
+
"eval_loss": 2.8616135120391846,
|
| 902 |
+
"eval_runtime": 4.7031,
|
| 903 |
+
"eval_samples_per_second": 82.924,
|
| 904 |
+
"eval_steps_per_second": 10.419,
|
| 905 |
+
"step": 1300
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.68,
|
| 909 |
+
"learning_rate": 3.251928020565553e-05,
|
| 910 |
+
"loss": 2.6972,
|
| 911 |
+
"step": 1310
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 1.7,
|
| 915 |
+
"learning_rate": 3.1233933161953726e-05,
|
| 916 |
+
"loss": 2.8446,
|
| 917 |
+
"step": 1320
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 1.71,
|
| 921 |
+
"learning_rate": 2.994858611825193e-05,
|
| 922 |
+
"loss": 2.915,
|
| 923 |
+
"step": 1330
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 1.72,
|
| 927 |
+
"learning_rate": 2.866323907455013e-05,
|
| 928 |
+
"loss": 3.1146,
|
| 929 |
+
"step": 1340
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 1.74,
|
| 933 |
+
"learning_rate": 2.737789203084833e-05,
|
| 934 |
+
"loss": 2.9168,
|
| 935 |
+
"step": 1350
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 1.75,
|
| 939 |
+
"learning_rate": 2.6092544987146534e-05,
|
| 940 |
+
"loss": 2.8762,
|
| 941 |
+
"step": 1360
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.76,
|
| 945 |
+
"learning_rate": 2.480719794344473e-05,
|
| 946 |
+
"loss": 2.9499,
|
| 947 |
+
"step": 1370
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.77,
|
| 951 |
+
"learning_rate": 2.3521850899742933e-05,
|
| 952 |
+
"loss": 2.907,
|
| 953 |
+
"step": 1380
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 1.79,
|
| 957 |
+
"learning_rate": 2.2236503856041134e-05,
|
| 958 |
+
"loss": 3.0447,
|
| 959 |
+
"step": 1390
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 1.8,
|
| 963 |
+
"learning_rate": 2.095115681233933e-05,
|
| 964 |
+
"loss": 2.988,
|
| 965 |
+
"step": 1400
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 1.8,
|
| 969 |
+
"eval_accuracy": 0.4076923076923077,
|
| 970 |
+
"eval_loss": 2.783235788345337,
|
| 971 |
+
"eval_runtime": 4.64,
|
| 972 |
+
"eval_samples_per_second": 84.052,
|
| 973 |
+
"eval_steps_per_second": 10.56,
|
| 974 |
+
"step": 1400
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 1.81,
|
| 978 |
+
"learning_rate": 1.9665809768637533e-05,
|
| 979 |
+
"loss": 2.8267,
|
| 980 |
+
"step": 1410
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 1.83,
|
| 984 |
+
"learning_rate": 1.8380462724935734e-05,
|
| 985 |
+
"loss": 2.7718,
|
| 986 |
+
"step": 1420
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 1.84,
|
| 990 |
+
"learning_rate": 1.7095115681233935e-05,
|
| 991 |
+
"loss": 2.8327,
|
| 992 |
+
"step": 1430
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 1.85,
|
| 996 |
+
"learning_rate": 1.5809768637532136e-05,
|
| 997 |
+
"loss": 2.9942,
|
| 998 |
+
"step": 1440
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 1.86,
|
| 1002 |
+
"learning_rate": 1.4524421593830334e-05,
|
| 1003 |
+
"loss": 3.0647,
|
| 1004 |
+
"step": 1450
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.88,
|
| 1008 |
+
"learning_rate": 1.3239074550128535e-05,
|
| 1009 |
+
"loss": 2.8267,
|
| 1010 |
+
"step": 1460
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.89,
|
| 1014 |
+
"learning_rate": 1.1953727506426736e-05,
|
| 1015 |
+
"loss": 2.8312,
|
| 1016 |
+
"step": 1470
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 1.9,
|
| 1020 |
+
"learning_rate": 1.0668380462724936e-05,
|
| 1021 |
+
"loss": 3.1317,
|
| 1022 |
+
"step": 1480
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 1.92,
|
| 1026 |
+
"learning_rate": 9.383033419023137e-06,
|
| 1027 |
+
"loss": 3.1244,
|
| 1028 |
+
"step": 1490
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 1.93,
|
| 1032 |
+
"learning_rate": 8.097686375321336e-06,
|
| 1033 |
+
"loss": 2.8553,
|
| 1034 |
+
"step": 1500
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 1.93,
|
| 1038 |
+
"eval_accuracy": 0.3871794871794872,
|
| 1039 |
+
"eval_loss": 2.821709632873535,
|
| 1040 |
+
"eval_runtime": 4.6901,
|
| 1041 |
+
"eval_samples_per_second": 83.154,
|
| 1042 |
+
"eval_steps_per_second": 10.448,
|
| 1043 |
+
"step": 1500
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 1.94,
|
| 1047 |
+
"learning_rate": 6.812339331619537e-06,
|
| 1048 |
+
"loss": 2.6944,
|
| 1049 |
+
"step": 1510
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 1.95,
|
| 1053 |
+
"learning_rate": 5.526992287917738e-06,
|
| 1054 |
+
"loss": 3.1573,
|
| 1055 |
+
"step": 1520
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 1.97,
|
| 1059 |
+
"learning_rate": 4.241645244215939e-06,
|
| 1060 |
+
"loss": 2.8412,
|
| 1061 |
+
"step": 1530
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 1.98,
|
| 1065 |
+
"learning_rate": 2.956298200514139e-06,
|
| 1066 |
+
"loss": 2.6937,
|
| 1067 |
+
"step": 1540
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 1.99,
|
| 1071 |
+
"learning_rate": 1.6709511568123394e-06,
|
| 1072 |
+
"loss": 2.65,
|
| 1073 |
+
"step": 1550
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 2.0,
|
| 1077 |
+
"step": 1556,
|
| 1078 |
+
"total_flos": 1.1959742538326016e+17,
|
| 1079 |
+
"train_loss": 3.1177605087162594,
|
| 1080 |
+
"train_runtime": 412.2299,
|
| 1081 |
+
"train_samples_per_second": 7.549,
|
| 1082 |
+
"train_steps_per_second": 3.775
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 2.0,
|
| 1086 |
+
"eval_accuracy": 0.4076923076923077,
|
| 1087 |
+
"eval_loss": 2.783235788345337,
|
| 1088 |
+
"eval_runtime": 5.3396,
|
| 1089 |
+
"eval_samples_per_second": 73.039,
|
| 1090 |
+
"eval_steps_per_second": 9.177,
|
| 1091 |
+
"step": 1556
|
| 1092 |
+
}
|
| 1093 |
+
],
|
| 1094 |
+
"max_steps": 1556,
|
| 1095 |
+
"num_train_epochs": 2,
|
| 1096 |
+
"total_flos": 1.1959742538326016e+17,
|
| 1097 |
+
"trial_name": null,
|
| 1098 |
+
"trial_params": null
|
| 1099 |
+
}
|
validation_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.0,
|
| 3 |
+
"eval_accuracy": 0.4076923076923077,
|
| 4 |
+
"eval_loss": 2.783235788345337,
|
| 5 |
+
"eval_runtime": 5.3396,
|
| 6 |
+
"eval_samples_per_second": 73.039,
|
| 7 |
+
"eval_steps_per_second": 9.177
|
| 8 |
+
}
|