Model save
Browse files- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +1812 -0
all_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 2.9987959060806744,
|
3 |
+
"total_flos": 1.3193601009550295e+19,
|
4 |
+
"train_loss": 0.06868853225451277,
|
5 |
+
"train_runtime": 5741.3098,
|
6 |
+
"train_samples_per_second": 55.537,
|
7 |
+
"train_steps_per_second": 0.434
|
8 |
+
}
|
train_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 2.9987959060806744,
|
3 |
+
"total_flos": 1.3193601009550295e+19,
|
4 |
+
"train_loss": 0.06868853225451277,
|
5 |
+
"train_runtime": 5741.3098,
|
6 |
+
"train_samples_per_second": 55.537,
|
7 |
+
"train_steps_per_second": 0.434
|
8 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,1812 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.9916771315123634,
|
3 |
+
"best_model_checkpoint": "convnext-tiny-224-finetuned-eurosat/checkpoint-2490",
|
4 |
+
"epoch": 2.9987959060806744,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 2490,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.012040939193257074,
|
13 |
+
"grad_norm": 13.281408309936523,
|
14 |
+
"learning_rate": 2.0080321285140564e-06,
|
15 |
+
"loss": 1.4682,
|
16 |
+
"step": 10
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.024081878386514148,
|
20 |
+
"grad_norm": 11.210328102111816,
|
21 |
+
"learning_rate": 4.016064257028113e-06,
|
22 |
+
"loss": 1.4426,
|
23 |
+
"step": 20
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.036122817579771226,
|
27 |
+
"grad_norm": 14.131393432617188,
|
28 |
+
"learning_rate": 6.024096385542169e-06,
|
29 |
+
"loss": 1.3941,
|
30 |
+
"step": 30
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.048163756773028296,
|
34 |
+
"grad_norm": 12.672858238220215,
|
35 |
+
"learning_rate": 8.032128514056226e-06,
|
36 |
+
"loss": 1.3162,
|
37 |
+
"step": 40
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.060204695966285374,
|
41 |
+
"grad_norm": 12.453862190246582,
|
42 |
+
"learning_rate": 1.0040160642570281e-05,
|
43 |
+
"loss": 1.1984,
|
44 |
+
"step": 50
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.07224563515954245,
|
48 |
+
"grad_norm": 18.517236709594727,
|
49 |
+
"learning_rate": 1.2048192771084338e-05,
|
50 |
+
"loss": 1.0524,
|
51 |
+
"step": 60
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.08428657435279951,
|
55 |
+
"grad_norm": 16.96401023864746,
|
56 |
+
"learning_rate": 1.4056224899598394e-05,
|
57 |
+
"loss": 0.8876,
|
58 |
+
"step": 70
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.09632751354605659,
|
62 |
+
"grad_norm": 41.16529083251953,
|
63 |
+
"learning_rate": 1.606425702811245e-05,
|
64 |
+
"loss": 0.7224,
|
65 |
+
"step": 80
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.10836845273931367,
|
69 |
+
"grad_norm": 48.28220748901367,
|
70 |
+
"learning_rate": 1.8072289156626505e-05,
|
71 |
+
"loss": 0.6089,
|
72 |
+
"step": 90
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.12040939193257075,
|
76 |
+
"grad_norm": 37.04490661621094,
|
77 |
+
"learning_rate": 2.0080321285140562e-05,
|
78 |
+
"loss": 0.5196,
|
79 |
+
"step": 100
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.13245033112582782,
|
83 |
+
"grad_norm": 32.357425689697266,
|
84 |
+
"learning_rate": 2.208835341365462e-05,
|
85 |
+
"loss": 0.4334,
|
86 |
+
"step": 110
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.1444912703190849,
|
90 |
+
"grad_norm": 41.277889251708984,
|
91 |
+
"learning_rate": 2.4096385542168677e-05,
|
92 |
+
"loss": 0.3515,
|
93 |
+
"step": 120
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.15653220951234195,
|
97 |
+
"grad_norm": 36.72848892211914,
|
98 |
+
"learning_rate": 2.6104417670682734e-05,
|
99 |
+
"loss": 0.2723,
|
100 |
+
"step": 130
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.16857314870559903,
|
104 |
+
"grad_norm": 48.640384674072266,
|
105 |
+
"learning_rate": 2.8112449799196788e-05,
|
106 |
+
"loss": 0.2383,
|
107 |
+
"step": 140
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.1806140878988561,
|
111 |
+
"grad_norm": 41.72344970703125,
|
112 |
+
"learning_rate": 3.012048192771085e-05,
|
113 |
+
"loss": 0.1954,
|
114 |
+
"step": 150
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.19265502709211318,
|
118 |
+
"grad_norm": 26.390892028808594,
|
119 |
+
"learning_rate": 3.21285140562249e-05,
|
120 |
+
"loss": 0.1937,
|
121 |
+
"step": 160
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.20469596628537026,
|
125 |
+
"grad_norm": 31.522932052612305,
|
126 |
+
"learning_rate": 3.413654618473896e-05,
|
127 |
+
"loss": 0.1657,
|
128 |
+
"step": 170
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.21673690547862734,
|
132 |
+
"grad_norm": 34.83334732055664,
|
133 |
+
"learning_rate": 3.614457831325301e-05,
|
134 |
+
"loss": 0.1678,
|
135 |
+
"step": 180
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.22877784467188442,
|
139 |
+
"grad_norm": 17.641130447387695,
|
140 |
+
"learning_rate": 3.815261044176707e-05,
|
141 |
+
"loss": 0.1367,
|
142 |
+
"step": 190
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.2408187838651415,
|
146 |
+
"grad_norm": 12.848413467407227,
|
147 |
+
"learning_rate": 4.0160642570281125e-05,
|
148 |
+
"loss": 0.1183,
|
149 |
+
"step": 200
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.25285972305839854,
|
153 |
+
"grad_norm": 12.54123306274414,
|
154 |
+
"learning_rate": 4.2168674698795186e-05,
|
155 |
+
"loss": 0.1026,
|
156 |
+
"step": 210
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.26490066225165565,
|
160 |
+
"grad_norm": 14.553470611572266,
|
161 |
+
"learning_rate": 4.417670682730924e-05,
|
162 |
+
"loss": 0.0932,
|
163 |
+
"step": 220
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.2769416014449127,
|
167 |
+
"grad_norm": 13.884803771972656,
|
168 |
+
"learning_rate": 4.61847389558233e-05,
|
169 |
+
"loss": 0.0697,
|
170 |
+
"step": 230
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.2889825406381698,
|
174 |
+
"grad_norm": 19.946374893188477,
|
175 |
+
"learning_rate": 4.8192771084337354e-05,
|
176 |
+
"loss": 0.0845,
|
177 |
+
"step": 240
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.30102347983142685,
|
181 |
+
"grad_norm": 25.795095443725586,
|
182 |
+
"learning_rate": 4.9977688531905406e-05,
|
183 |
+
"loss": 0.0927,
|
184 |
+
"step": 250
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.3130644190246839,
|
188 |
+
"grad_norm": 17.59208869934082,
|
189 |
+
"learning_rate": 4.97545738509594e-05,
|
190 |
+
"loss": 0.0621,
|
191 |
+
"step": 260
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.325105358217941,
|
195 |
+
"grad_norm": 14.659806251525879,
|
196 |
+
"learning_rate": 4.953145917001339e-05,
|
197 |
+
"loss": 0.0644,
|
198 |
+
"step": 270
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 0.33714629741119806,
|
202 |
+
"grad_norm": 15.293780326843262,
|
203 |
+
"learning_rate": 4.930834448906738e-05,
|
204 |
+
"loss": 0.0882,
|
205 |
+
"step": 280
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 0.34918723660445516,
|
209 |
+
"grad_norm": 11.717020988464355,
|
210 |
+
"learning_rate": 4.908522980812137e-05,
|
211 |
+
"loss": 0.0594,
|
212 |
+
"step": 290
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.3612281757977122,
|
216 |
+
"grad_norm": 29.984556198120117,
|
217 |
+
"learning_rate": 4.886211512717537e-05,
|
218 |
+
"loss": 0.0854,
|
219 |
+
"step": 300
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.3732691149909693,
|
223 |
+
"grad_norm": 10.837246894836426,
|
224 |
+
"learning_rate": 4.8639000446229364e-05,
|
225 |
+
"loss": 0.0782,
|
226 |
+
"step": 310
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 0.38531005418422637,
|
230 |
+
"grad_norm": 11.394307136535645,
|
231 |
+
"learning_rate": 4.8415885765283355e-05,
|
232 |
+
"loss": 0.0666,
|
233 |
+
"step": 320
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 0.3973509933774834,
|
237 |
+
"grad_norm": 31.305824279785156,
|
238 |
+
"learning_rate": 4.8192771084337354e-05,
|
239 |
+
"loss": 0.0529,
|
240 |
+
"step": 330
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 0.4093919325707405,
|
244 |
+
"grad_norm": 31.709115982055664,
|
245 |
+
"learning_rate": 4.7969656403391346e-05,
|
246 |
+
"loss": 0.0504,
|
247 |
+
"step": 340
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 0.4214328717639976,
|
251 |
+
"grad_norm": 3.85288143157959,
|
252 |
+
"learning_rate": 4.774654172244534e-05,
|
253 |
+
"loss": 0.0498,
|
254 |
+
"step": 350
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 0.4334738109572547,
|
258 |
+
"grad_norm": 11.564428329467773,
|
259 |
+
"learning_rate": 4.7523427041499336e-05,
|
260 |
+
"loss": 0.0635,
|
261 |
+
"step": 360
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 0.44551475015051173,
|
265 |
+
"grad_norm": 35.95602798461914,
|
266 |
+
"learning_rate": 4.730031236055333e-05,
|
267 |
+
"loss": 0.0464,
|
268 |
+
"step": 370
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 0.45755568934376883,
|
272 |
+
"grad_norm": 16.262226104736328,
|
273 |
+
"learning_rate": 4.707719767960732e-05,
|
274 |
+
"loss": 0.0404,
|
275 |
+
"step": 380
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 0.4695966285370259,
|
279 |
+
"grad_norm": 15.375874519348145,
|
280 |
+
"learning_rate": 4.685408299866131e-05,
|
281 |
+
"loss": 0.0399,
|
282 |
+
"step": 390
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 0.481637567730283,
|
286 |
+
"grad_norm": 5.388984680175781,
|
287 |
+
"learning_rate": 4.663096831771531e-05,
|
288 |
+
"loss": 0.0381,
|
289 |
+
"step": 400
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 0.49367850692354004,
|
293 |
+
"grad_norm": 1.3834246397018433,
|
294 |
+
"learning_rate": 4.64078536367693e-05,
|
295 |
+
"loss": 0.046,
|
296 |
+
"step": 410
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 0.5057194461167971,
|
300 |
+
"grad_norm": 29.666913986206055,
|
301 |
+
"learning_rate": 4.61847389558233e-05,
|
302 |
+
"loss": 0.0549,
|
303 |
+
"step": 420
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 0.5177603853100542,
|
307 |
+
"grad_norm": 11.310073852539062,
|
308 |
+
"learning_rate": 4.596162427487729e-05,
|
309 |
+
"loss": 0.0428,
|
310 |
+
"step": 430
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 0.5298013245033113,
|
314 |
+
"grad_norm": 14.786700248718262,
|
315 |
+
"learning_rate": 4.5738509593931284e-05,
|
316 |
+
"loss": 0.0376,
|
317 |
+
"step": 440
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 0.5418422636965683,
|
321 |
+
"grad_norm": 28.993574142456055,
|
322 |
+
"learning_rate": 4.5515394912985275e-05,
|
323 |
+
"loss": 0.0409,
|
324 |
+
"step": 450
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 0.5538832028898254,
|
328 |
+
"grad_norm": 12.82999038696289,
|
329 |
+
"learning_rate": 4.529228023203927e-05,
|
330 |
+
"loss": 0.0576,
|
331 |
+
"step": 460
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 0.5659241420830825,
|
335 |
+
"grad_norm": 25.034116744995117,
|
336 |
+
"learning_rate": 4.506916555109326e-05,
|
337 |
+
"loss": 0.0384,
|
338 |
+
"step": 470
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 0.5779650812763396,
|
342 |
+
"grad_norm": 10.026959419250488,
|
343 |
+
"learning_rate": 4.484605087014726e-05,
|
344 |
+
"loss": 0.0352,
|
345 |
+
"step": 480
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 0.5900060204695966,
|
349 |
+
"grad_norm": 4.99667501449585,
|
350 |
+
"learning_rate": 4.4622936189201256e-05,
|
351 |
+
"loss": 0.0333,
|
352 |
+
"step": 490
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 0.6020469596628537,
|
356 |
+
"grad_norm": 29.965518951416016,
|
357 |
+
"learning_rate": 4.439982150825525e-05,
|
358 |
+
"loss": 0.0425,
|
359 |
+
"step": 500
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 0.6140878988561108,
|
363 |
+
"grad_norm": 15.231363296508789,
|
364 |
+
"learning_rate": 4.417670682730924e-05,
|
365 |
+
"loss": 0.0414,
|
366 |
+
"step": 510
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 0.6261288380493678,
|
370 |
+
"grad_norm": 34.312477111816406,
|
371 |
+
"learning_rate": 4.395359214636323e-05,
|
372 |
+
"loss": 0.0411,
|
373 |
+
"step": 520
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 0.6381697772426249,
|
377 |
+
"grad_norm": 60.11940383911133,
|
378 |
+
"learning_rate": 4.373047746541722e-05,
|
379 |
+
"loss": 0.0436,
|
380 |
+
"step": 530
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 0.650210716435882,
|
384 |
+
"grad_norm": 2.1523265838623047,
|
385 |
+
"learning_rate": 4.350736278447122e-05,
|
386 |
+
"loss": 0.0364,
|
387 |
+
"step": 540
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 0.6622516556291391,
|
391 |
+
"grad_norm": 22.606645584106445,
|
392 |
+
"learning_rate": 4.328424810352521e-05,
|
393 |
+
"loss": 0.0348,
|
394 |
+
"step": 550
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 0.6742925948223961,
|
398 |
+
"grad_norm": 15.017701148986816,
|
399 |
+
"learning_rate": 4.306113342257921e-05,
|
400 |
+
"loss": 0.0301,
|
401 |
+
"step": 560
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 0.6863335340156532,
|
405 |
+
"grad_norm": 8.95254898071289,
|
406 |
+
"learning_rate": 4.2838018741633203e-05,
|
407 |
+
"loss": 0.0322,
|
408 |
+
"step": 570
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 0.6983744732089103,
|
412 |
+
"grad_norm": 12.393402099609375,
|
413 |
+
"learning_rate": 4.2614904060687195e-05,
|
414 |
+
"loss": 0.0207,
|
415 |
+
"step": 580
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 0.7104154124021673,
|
419 |
+
"grad_norm": 4.132020950317383,
|
420 |
+
"learning_rate": 4.239178937974119e-05,
|
421 |
+
"loss": 0.0479,
|
422 |
+
"step": 590
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 0.7224563515954244,
|
426 |
+
"grad_norm": 11.15093994140625,
|
427 |
+
"learning_rate": 4.2168674698795186e-05,
|
428 |
+
"loss": 0.0195,
|
429 |
+
"step": 600
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 0.7344972907886815,
|
433 |
+
"grad_norm": 19.265222549438477,
|
434 |
+
"learning_rate": 4.194556001784918e-05,
|
435 |
+
"loss": 0.0229,
|
436 |
+
"step": 610
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 0.7465382299819386,
|
440 |
+
"grad_norm": 14.732630729675293,
|
441 |
+
"learning_rate": 4.172244533690317e-05,
|
442 |
+
"loss": 0.0243,
|
443 |
+
"step": 620
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 0.7585791691751956,
|
447 |
+
"grad_norm": 2.9074995517730713,
|
448 |
+
"learning_rate": 4.149933065595716e-05,
|
449 |
+
"loss": 0.0324,
|
450 |
+
"step": 630
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 0.7706201083684527,
|
454 |
+
"grad_norm": 12.28730583190918,
|
455 |
+
"learning_rate": 4.127621597501116e-05,
|
456 |
+
"loss": 0.0341,
|
457 |
+
"step": 640
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 0.7826610475617098,
|
461 |
+
"grad_norm": 5.263113975524902,
|
462 |
+
"learning_rate": 4.105310129406515e-05,
|
463 |
+
"loss": 0.0345,
|
464 |
+
"step": 650
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 0.7947019867549668,
|
468 |
+
"grad_norm": 20.13115119934082,
|
469 |
+
"learning_rate": 4.082998661311915e-05,
|
470 |
+
"loss": 0.0371,
|
471 |
+
"step": 660
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 0.8067429259482239,
|
475 |
+
"grad_norm": 6.125397205352783,
|
476 |
+
"learning_rate": 4.060687193217314e-05,
|
477 |
+
"loss": 0.0218,
|
478 |
+
"step": 670
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 0.818783865141481,
|
482 |
+
"grad_norm": 4.126222133636475,
|
483 |
+
"learning_rate": 4.038375725122713e-05,
|
484 |
+
"loss": 0.0266,
|
485 |
+
"step": 680
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 0.8308248043347382,
|
489 |
+
"grad_norm": 4.383463382720947,
|
490 |
+
"learning_rate": 4.0160642570281125e-05,
|
491 |
+
"loss": 0.0472,
|
492 |
+
"step": 690
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 0.8428657435279951,
|
496 |
+
"grad_norm": 1.4781919717788696,
|
497 |
+
"learning_rate": 3.993752788933512e-05,
|
498 |
+
"loss": 0.0265,
|
499 |
+
"step": 700
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 0.8549066827212523,
|
503 |
+
"grad_norm": 7.968236446380615,
|
504 |
+
"learning_rate": 3.9714413208389115e-05,
|
505 |
+
"loss": 0.0326,
|
506 |
+
"step": 710
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 0.8669476219145094,
|
510 |
+
"grad_norm": 15.246943473815918,
|
511 |
+
"learning_rate": 3.949129852744311e-05,
|
512 |
+
"loss": 0.0256,
|
513 |
+
"step": 720
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 0.8789885611077664,
|
517 |
+
"grad_norm": 16.984224319458008,
|
518 |
+
"learning_rate": 3.9268183846497105e-05,
|
519 |
+
"loss": 0.0128,
|
520 |
+
"step": 730
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 0.8910295003010235,
|
524 |
+
"grad_norm": 15.068581581115723,
|
525 |
+
"learning_rate": 3.90450691655511e-05,
|
526 |
+
"loss": 0.0245,
|
527 |
+
"step": 740
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 0.9030704394942806,
|
531 |
+
"grad_norm": 0.4837843179702759,
|
532 |
+
"learning_rate": 3.882195448460509e-05,
|
533 |
+
"loss": 0.0091,
|
534 |
+
"step": 750
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 0.9151113786875377,
|
538 |
+
"grad_norm": 21.087526321411133,
|
539 |
+
"learning_rate": 3.859883980365908e-05,
|
540 |
+
"loss": 0.036,
|
541 |
+
"step": 760
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 0.9271523178807947,
|
545 |
+
"grad_norm": 13.440828323364258,
|
546 |
+
"learning_rate": 3.837572512271307e-05,
|
547 |
+
"loss": 0.0137,
|
548 |
+
"step": 770
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 0.9391932570740518,
|
552 |
+
"grad_norm": 4.101161956787109,
|
553 |
+
"learning_rate": 3.815261044176707e-05,
|
554 |
+
"loss": 0.0207,
|
555 |
+
"step": 780
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 0.9512341962673089,
|
559 |
+
"grad_norm": 9.98717975616455,
|
560 |
+
"learning_rate": 3.792949576082106e-05,
|
561 |
+
"loss": 0.032,
|
562 |
+
"step": 790
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 0.963275135460566,
|
566 |
+
"grad_norm": 20.295673370361328,
|
567 |
+
"learning_rate": 3.770638107987506e-05,
|
568 |
+
"loss": 0.0288,
|
569 |
+
"step": 800
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 0.975316074653823,
|
573 |
+
"grad_norm": 4.261445999145508,
|
574 |
+
"learning_rate": 3.748326639892905e-05,
|
575 |
+
"loss": 0.0231,
|
576 |
+
"step": 810
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 0.9873570138470801,
|
580 |
+
"grad_norm": 19.60774803161621,
|
581 |
+
"learning_rate": 3.7260151717983045e-05,
|
582 |
+
"loss": 0.0376,
|
583 |
+
"step": 820
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 0.9993979530403372,
|
587 |
+
"grad_norm": 7.224467754364014,
|
588 |
+
"learning_rate": 3.7037037037037037e-05,
|
589 |
+
"loss": 0.0369,
|
590 |
+
"step": 830
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 0.9993979530403372,
|
594 |
+
"eval_f1": 0.9770145182151784,
|
595 |
+
"eval_loss": 0.03777425363659859,
|
596 |
+
"eval_runtime": 102.228,
|
597 |
+
"eval_samples_per_second": 115.526,
|
598 |
+
"eval_steps_per_second": 3.619,
|
599 |
+
"step": 830
|
600 |
+
},
|
601 |
+
{
|
602 |
+
"epoch": 1.0117399157134257,
|
603 |
+
"grad_norm": 15.622239112854004,
|
604 |
+
"learning_rate": 3.6813922356091035e-05,
|
605 |
+
"loss": 0.0179,
|
606 |
+
"step": 840
|
607 |
+
},
|
608 |
+
{
|
609 |
+
"epoch": 1.0237808549066827,
|
610 |
+
"grad_norm": 0.41096824407577515,
|
611 |
+
"learning_rate": 3.659080767514503e-05,
|
612 |
+
"loss": 0.0076,
|
613 |
+
"step": 850
|
614 |
+
},
|
615 |
+
{
|
616 |
+
"epoch": 1.03582179409994,
|
617 |
+
"grad_norm": 18.979677200317383,
|
618 |
+
"learning_rate": 3.636769299419902e-05,
|
619 |
+
"loss": 0.024,
|
620 |
+
"step": 860
|
621 |
+
},
|
622 |
+
{
|
623 |
+
"epoch": 1.047862733293197,
|
624 |
+
"grad_norm": 9.764080047607422,
|
625 |
+
"learning_rate": 3.614457831325301e-05,
|
626 |
+
"loss": 0.0142,
|
627 |
+
"step": 870
|
628 |
+
},
|
629 |
+
{
|
630 |
+
"epoch": 1.059903672486454,
|
631 |
+
"grad_norm": 2.4317643642425537,
|
632 |
+
"learning_rate": 3.592146363230701e-05,
|
633 |
+
"loss": 0.0265,
|
634 |
+
"step": 880
|
635 |
+
},
|
636 |
+
{
|
637 |
+
"epoch": 1.0719446116797111,
|
638 |
+
"grad_norm": 15.766806602478027,
|
639 |
+
"learning_rate": 3.5698348951361e-05,
|
640 |
+
"loss": 0.0234,
|
641 |
+
"step": 890
|
642 |
+
},
|
643 |
+
{
|
644 |
+
"epoch": 1.083985550872968,
|
645 |
+
"grad_norm": 11.236108779907227,
|
646 |
+
"learning_rate": 3.5475234270415e-05,
|
647 |
+
"loss": 0.0142,
|
648 |
+
"step": 900
|
649 |
+
},
|
650 |
+
{
|
651 |
+
"epoch": 1.096026490066225,
|
652 |
+
"grad_norm": 2.210698127746582,
|
653 |
+
"learning_rate": 3.525211958946899e-05,
|
654 |
+
"loss": 0.0034,
|
655 |
+
"step": 910
|
656 |
+
},
|
657 |
+
{
|
658 |
+
"epoch": 1.1080674292594823,
|
659 |
+
"grad_norm": 0.30994459986686707,
|
660 |
+
"learning_rate": 3.502900490852298e-05,
|
661 |
+
"loss": 0.0038,
|
662 |
+
"step": 920
|
663 |
+
},
|
664 |
+
{
|
665 |
+
"epoch": 1.1201083684527393,
|
666 |
+
"grad_norm": 7.224539279937744,
|
667 |
+
"learning_rate": 3.4805890227576974e-05,
|
668 |
+
"loss": 0.0246,
|
669 |
+
"step": 930
|
670 |
+
},
|
671 |
+
{
|
672 |
+
"epoch": 1.1321493076459963,
|
673 |
+
"grad_norm": 0.3623636066913605,
|
674 |
+
"learning_rate": 3.4582775546630966e-05,
|
675 |
+
"loss": 0.006,
|
676 |
+
"step": 940
|
677 |
+
},
|
678 |
+
{
|
679 |
+
"epoch": 1.1441902468392535,
|
680 |
+
"grad_norm": 3.8137762546539307,
|
681 |
+
"learning_rate": 3.4359660865684965e-05,
|
682 |
+
"loss": 0.0176,
|
683 |
+
"step": 950
|
684 |
+
},
|
685 |
+
{
|
686 |
+
"epoch": 1.1562311860325105,
|
687 |
+
"grad_norm": 0.8698212504386902,
|
688 |
+
"learning_rate": 3.413654618473896e-05,
|
689 |
+
"loss": 0.0166,
|
690 |
+
"step": 960
|
691 |
+
},
|
692 |
+
{
|
693 |
+
"epoch": 1.1682721252257675,
|
694 |
+
"grad_norm": 14.512166976928711,
|
695 |
+
"learning_rate": 3.3913431503792955e-05,
|
696 |
+
"loss": 0.0179,
|
697 |
+
"step": 970
|
698 |
+
},
|
699 |
+
{
|
700 |
+
"epoch": 1.1803130644190247,
|
701 |
+
"grad_norm": 23.775272369384766,
|
702 |
+
"learning_rate": 3.369031682284695e-05,
|
703 |
+
"loss": 0.0247,
|
704 |
+
"step": 980
|
705 |
+
},
|
706 |
+
{
|
707 |
+
"epoch": 1.1923540036122817,
|
708 |
+
"grad_norm": 9.087154388427734,
|
709 |
+
"learning_rate": 3.346720214190094e-05,
|
710 |
+
"loss": 0.0212,
|
711 |
+
"step": 990
|
712 |
+
},
|
713 |
+
{
|
714 |
+
"epoch": 1.2043949428055387,
|
715 |
+
"grad_norm": 23.993961334228516,
|
716 |
+
"learning_rate": 3.324408746095493e-05,
|
717 |
+
"loss": 0.0198,
|
718 |
+
"step": 1000
|
719 |
+
},
|
720 |
+
{
|
721 |
+
"epoch": 1.216435881998796,
|
722 |
+
"grad_norm": 5.462221622467041,
|
723 |
+
"learning_rate": 3.302097278000892e-05,
|
724 |
+
"loss": 0.0077,
|
725 |
+
"step": 1010
|
726 |
+
},
|
727 |
+
{
|
728 |
+
"epoch": 1.228476821192053,
|
729 |
+
"grad_norm": 0.17719963192939758,
|
730 |
+
"learning_rate": 3.279785809906292e-05,
|
731 |
+
"loss": 0.0107,
|
732 |
+
"step": 1020
|
733 |
+
},
|
734 |
+
{
|
735 |
+
"epoch": 1.2405177603853101,
|
736 |
+
"grad_norm": 19.494287490844727,
|
737 |
+
"learning_rate": 3.257474341811691e-05,
|
738 |
+
"loss": 0.0125,
|
739 |
+
"step": 1030
|
740 |
+
},
|
741 |
+
{
|
742 |
+
"epoch": 1.2525586995785671,
|
743 |
+
"grad_norm": 7.95407772064209,
|
744 |
+
"learning_rate": 3.235162873717091e-05,
|
745 |
+
"loss": 0.0105,
|
746 |
+
"step": 1040
|
747 |
+
},
|
748 |
+
{
|
749 |
+
"epoch": 1.2645996387718241,
|
750 |
+
"grad_norm": 5.586263656616211,
|
751 |
+
"learning_rate": 3.21285140562249e-05,
|
752 |
+
"loss": 0.0178,
|
753 |
+
"step": 1050
|
754 |
+
},
|
755 |
+
{
|
756 |
+
"epoch": 1.2766405779650813,
|
757 |
+
"grad_norm": 0.2848265469074249,
|
758 |
+
"learning_rate": 3.1905399375278894e-05,
|
759 |
+
"loss": 0.0095,
|
760 |
+
"step": 1060
|
761 |
+
},
|
762 |
+
{
|
763 |
+
"epoch": 1.2886815171583383,
|
764 |
+
"grad_norm": 1.2747416496276855,
|
765 |
+
"learning_rate": 3.1682284694332886e-05,
|
766 |
+
"loss": 0.0046,
|
767 |
+
"step": 1070
|
768 |
+
},
|
769 |
+
{
|
770 |
+
"epoch": 1.3007224563515956,
|
771 |
+
"grad_norm": 0.1546485871076584,
|
772 |
+
"learning_rate": 3.1459170013386885e-05,
|
773 |
+
"loss": 0.0176,
|
774 |
+
"step": 1080
|
775 |
+
},
|
776 |
+
{
|
777 |
+
"epoch": 1.3127633955448526,
|
778 |
+
"grad_norm": 0.07813813537359238,
|
779 |
+
"learning_rate": 3.1236055332440876e-05,
|
780 |
+
"loss": 0.0153,
|
781 |
+
"step": 1090
|
782 |
+
},
|
783 |
+
{
|
784 |
+
"epoch": 1.3248043347381095,
|
785 |
+
"grad_norm": 0.5948330163955688,
|
786 |
+
"learning_rate": 3.101294065149487e-05,
|
787 |
+
"loss": 0.0133,
|
788 |
+
"step": 1100
|
789 |
+
},
|
790 |
+
{
|
791 |
+
"epoch": 1.3368452739313668,
|
792 |
+
"grad_norm": 26.891010284423828,
|
793 |
+
"learning_rate": 3.078982597054887e-05,
|
794 |
+
"loss": 0.026,
|
795 |
+
"step": 1110
|
796 |
+
},
|
797 |
+
{
|
798 |
+
"epoch": 1.3488862131246238,
|
799 |
+
"grad_norm": 5.57611083984375,
|
800 |
+
"learning_rate": 3.056671128960286e-05,
|
801 |
+
"loss": 0.0129,
|
802 |
+
"step": 1120
|
803 |
+
},
|
804 |
+
{
|
805 |
+
"epoch": 1.3609271523178808,
|
806 |
+
"grad_norm": 29.697946548461914,
|
807 |
+
"learning_rate": 3.034359660865685e-05,
|
808 |
+
"loss": 0.0136,
|
809 |
+
"step": 1130
|
810 |
+
},
|
811 |
+
{
|
812 |
+
"epoch": 1.372968091511138,
|
813 |
+
"grad_norm": 0.9194643497467041,
|
814 |
+
"learning_rate": 3.012048192771085e-05,
|
815 |
+
"loss": 0.0018,
|
816 |
+
"step": 1140
|
817 |
+
},
|
818 |
+
{
|
819 |
+
"epoch": 1.385009030704395,
|
820 |
+
"grad_norm": 0.1039569079875946,
|
821 |
+
"learning_rate": 2.989736724676484e-05,
|
822 |
+
"loss": 0.0158,
|
823 |
+
"step": 1150
|
824 |
+
},
|
825 |
+
{
|
826 |
+
"epoch": 1.397049969897652,
|
827 |
+
"grad_norm": 32.51161193847656,
|
828 |
+
"learning_rate": 2.9674252565818832e-05,
|
829 |
+
"loss": 0.0085,
|
830 |
+
"step": 1160
|
831 |
+
},
|
832 |
+
{
|
833 |
+
"epoch": 1.4090909090909092,
|
834 |
+
"grad_norm": 9.65622329711914,
|
835 |
+
"learning_rate": 2.9451137884872827e-05,
|
836 |
+
"loss": 0.0142,
|
837 |
+
"step": 1170
|
838 |
+
},
|
839 |
+
{
|
840 |
+
"epoch": 1.4211318482841662,
|
841 |
+
"grad_norm": 0.4631725251674652,
|
842 |
+
"learning_rate": 2.922802320392682e-05,
|
843 |
+
"loss": 0.0096,
|
844 |
+
"step": 1180
|
845 |
+
},
|
846 |
+
{
|
847 |
+
"epoch": 1.4331727874774232,
|
848 |
+
"grad_norm": 9.818026542663574,
|
849 |
+
"learning_rate": 2.900490852298081e-05,
|
850 |
+
"loss": 0.0055,
|
851 |
+
"step": 1190
|
852 |
+
},
|
853 |
+
{
|
854 |
+
"epoch": 1.4452137266706804,
|
855 |
+
"grad_norm": 5.643733978271484,
|
856 |
+
"learning_rate": 2.878179384203481e-05,
|
857 |
+
"loss": 0.0146,
|
858 |
+
"step": 1200
|
859 |
+
},
|
860 |
+
{
|
861 |
+
"epoch": 1.4572546658639374,
|
862 |
+
"grad_norm": 5.967852592468262,
|
863 |
+
"learning_rate": 2.85586791610888e-05,
|
864 |
+
"loss": 0.0228,
|
865 |
+
"step": 1210
|
866 |
+
},
|
867 |
+
{
|
868 |
+
"epoch": 1.4692956050571944,
|
869 |
+
"grad_norm": 0.1257736086845398,
|
870 |
+
"learning_rate": 2.8335564480142796e-05,
|
871 |
+
"loss": 0.003,
|
872 |
+
"step": 1220
|
873 |
+
},
|
874 |
+
{
|
875 |
+
"epoch": 1.4813365442504516,
|
876 |
+
"grad_norm": 14.535481452941895,
|
877 |
+
"learning_rate": 2.8112449799196788e-05,
|
878 |
+
"loss": 0.0156,
|
879 |
+
"step": 1230
|
880 |
+
},
|
881 |
+
{
|
882 |
+
"epoch": 1.4933774834437086,
|
883 |
+
"grad_norm": 1.1483675241470337,
|
884 |
+
"learning_rate": 2.788933511825078e-05,
|
885 |
+
"loss": 0.0051,
|
886 |
+
"step": 1240
|
887 |
+
},
|
888 |
+
{
|
889 |
+
"epoch": 1.5054184226369656,
|
890 |
+
"grad_norm": 0.11251900345087051,
|
891 |
+
"learning_rate": 2.7666220437304775e-05,
|
892 |
+
"loss": 0.0063,
|
893 |
+
"step": 1250
|
894 |
+
},
|
895 |
+
{
|
896 |
+
"epoch": 1.5174593618302228,
|
897 |
+
"grad_norm": 0.16821934282779694,
|
898 |
+
"learning_rate": 2.7443105756358774e-05,
|
899 |
+
"loss": 0.0066,
|
900 |
+
"step": 1260
|
901 |
+
},
|
902 |
+
{
|
903 |
+
"epoch": 1.5295003010234798,
|
904 |
+
"grad_norm": 1.6054164171218872,
|
905 |
+
"learning_rate": 2.7219991075412765e-05,
|
906 |
+
"loss": 0.0092,
|
907 |
+
"step": 1270
|
908 |
+
},
|
909 |
+
{
|
910 |
+
"epoch": 1.5415412402167368,
|
911 |
+
"grad_norm": 1.5014593601226807,
|
912 |
+
"learning_rate": 2.6996876394466757e-05,
|
913 |
+
"loss": 0.008,
|
914 |
+
"step": 1280
|
915 |
+
},
|
916 |
+
{
|
917 |
+
"epoch": 1.553582179409994,
|
918 |
+
"grad_norm": 27.260873794555664,
|
919 |
+
"learning_rate": 2.6773761713520752e-05,
|
920 |
+
"loss": 0.0277,
|
921 |
+
"step": 1290
|
922 |
+
},
|
923 |
+
{
|
924 |
+
"epoch": 1.5656231186032512,
|
925 |
+
"grad_norm": 6.338423728942871,
|
926 |
+
"learning_rate": 2.6550647032574744e-05,
|
927 |
+
"loss": 0.0202,
|
928 |
+
"step": 1300
|
929 |
+
},
|
930 |
+
{
|
931 |
+
"epoch": 1.577664057796508,
|
932 |
+
"grad_norm": 18.155574798583984,
|
933 |
+
"learning_rate": 2.6327532351628736e-05,
|
934 |
+
"loss": 0.0069,
|
935 |
+
"step": 1310
|
936 |
+
},
|
937 |
+
{
|
938 |
+
"epoch": 1.5897049969897652,
|
939 |
+
"grad_norm": 0.7306817770004272,
|
940 |
+
"learning_rate": 2.6104417670682734e-05,
|
941 |
+
"loss": 0.0074,
|
942 |
+
"step": 1320
|
943 |
+
},
|
944 |
+
{
|
945 |
+
"epoch": 1.6017459361830224,
|
946 |
+
"grad_norm": 6.356385231018066,
|
947 |
+
"learning_rate": 2.5881302989736726e-05,
|
948 |
+
"loss": 0.0125,
|
949 |
+
"step": 1330
|
950 |
+
},
|
951 |
+
{
|
952 |
+
"epoch": 1.6137868753762792,
|
953 |
+
"grad_norm": 16.603483200073242,
|
954 |
+
"learning_rate": 2.565818830879072e-05,
|
955 |
+
"loss": 0.0264,
|
956 |
+
"step": 1340
|
957 |
+
},
|
958 |
+
{
|
959 |
+
"epoch": 1.6258278145695364,
|
960 |
+
"grad_norm": 10.486541748046875,
|
961 |
+
"learning_rate": 2.5435073627844713e-05,
|
962 |
+
"loss": 0.0148,
|
963 |
+
"step": 1350
|
964 |
+
},
|
965 |
+
{
|
966 |
+
"epoch": 1.6378687537627936,
|
967 |
+
"grad_norm": 0.5101543664932251,
|
968 |
+
"learning_rate": 2.5211958946898705e-05,
|
969 |
+
"loss": 0.0158,
|
970 |
+
"step": 1360
|
971 |
+
},
|
972 |
+
{
|
973 |
+
"epoch": 1.6499096929560506,
|
974 |
+
"grad_norm": 19.748857498168945,
|
975 |
+
"learning_rate": 2.4988844265952703e-05,
|
976 |
+
"loss": 0.0132,
|
977 |
+
"step": 1370
|
978 |
+
},
|
979 |
+
{
|
980 |
+
"epoch": 1.6619506321493076,
|
981 |
+
"grad_norm": 0.3710501194000244,
|
982 |
+
"learning_rate": 2.4765729585006695e-05,
|
983 |
+
"loss": 0.0153,
|
984 |
+
"step": 1380
|
985 |
+
},
|
986 |
+
{
|
987 |
+
"epoch": 1.6739915713425648,
|
988 |
+
"grad_norm": 0.0527074858546257,
|
989 |
+
"learning_rate": 2.4542614904060687e-05,
|
990 |
+
"loss": 0.0119,
|
991 |
+
"step": 1390
|
992 |
+
},
|
993 |
+
{
|
994 |
+
"epoch": 1.6860325105358218,
|
995 |
+
"grad_norm": 1.91316556930542,
|
996 |
+
"learning_rate": 2.4319500223114682e-05,
|
997 |
+
"loss": 0.0104,
|
998 |
+
"step": 1400
|
999 |
+
},
|
1000 |
+
{
|
1001 |
+
"epoch": 1.6980734497290788,
|
1002 |
+
"grad_norm": 2.1063218116760254,
|
1003 |
+
"learning_rate": 2.4096385542168677e-05,
|
1004 |
+
"loss": 0.0055,
|
1005 |
+
"step": 1410
|
1006 |
+
},
|
1007 |
+
{
|
1008 |
+
"epoch": 1.710114388922336,
|
1009 |
+
"grad_norm": 9.150762557983398,
|
1010 |
+
"learning_rate": 2.387327086122267e-05,
|
1011 |
+
"loss": 0.0104,
|
1012 |
+
"step": 1420
|
1013 |
+
},
|
1014 |
+
{
|
1015 |
+
"epoch": 1.722155328115593,
|
1016 |
+
"grad_norm": 0.08365010470151901,
|
1017 |
+
"learning_rate": 2.3650156180276664e-05,
|
1018 |
+
"loss": 0.0107,
|
1019 |
+
"step": 1430
|
1020 |
+
},
|
1021 |
+
{
|
1022 |
+
"epoch": 1.73419626730885,
|
1023 |
+
"grad_norm": 8.636667251586914,
|
1024 |
+
"learning_rate": 2.3427041499330656e-05,
|
1025 |
+
"loss": 0.0099,
|
1026 |
+
"step": 1440
|
1027 |
+
},
|
1028 |
+
{
|
1029 |
+
"epoch": 1.7462372065021072,
|
1030 |
+
"grad_norm": 1.0073809623718262,
|
1031 |
+
"learning_rate": 2.320392681838465e-05,
|
1032 |
+
"loss": 0.0171,
|
1033 |
+
"step": 1450
|
1034 |
+
},
|
1035 |
+
{
|
1036 |
+
"epoch": 1.7582781456953642,
|
1037 |
+
"grad_norm": 5.5500969886779785,
|
1038 |
+
"learning_rate": 2.2980812137438646e-05,
|
1039 |
+
"loss": 0.0031,
|
1040 |
+
"step": 1460
|
1041 |
+
},
|
1042 |
+
{
|
1043 |
+
"epoch": 1.7703190848886212,
|
1044 |
+
"grad_norm": 9.417837142944336,
|
1045 |
+
"learning_rate": 2.2757697456492638e-05,
|
1046 |
+
"loss": 0.01,
|
1047 |
+
"step": 1470
|
1048 |
+
},
|
1049 |
+
{
|
1050 |
+
"epoch": 1.7823600240818784,
|
1051 |
+
"grad_norm": 18.310321807861328,
|
1052 |
+
"learning_rate": 2.253458277554663e-05,
|
1053 |
+
"loss": 0.0113,
|
1054 |
+
"step": 1480
|
1055 |
+
},
|
1056 |
+
{
|
1057 |
+
"epoch": 1.7944009632751354,
|
1058 |
+
"grad_norm": 1.6077951192855835,
|
1059 |
+
"learning_rate": 2.2311468094600628e-05,
|
1060 |
+
"loss": 0.0136,
|
1061 |
+
"step": 1490
|
1062 |
+
},
|
1063 |
+
{
|
1064 |
+
"epoch": 1.8064419024683924,
|
1065 |
+
"grad_norm": 2.8965089321136475,
|
1066 |
+
"learning_rate": 2.208835341365462e-05,
|
1067 |
+
"loss": 0.0109,
|
1068 |
+
"step": 1500
|
1069 |
+
},
|
1070 |
+
{
|
1071 |
+
"epoch": 1.8184828416616496,
|
1072 |
+
"grad_norm": 1.5971770286560059,
|
1073 |
+
"learning_rate": 2.186523873270861e-05,
|
1074 |
+
"loss": 0.0016,
|
1075 |
+
"step": 1510
|
1076 |
+
},
|
1077 |
+
{
|
1078 |
+
"epoch": 1.8305237808549066,
|
1079 |
+
"grad_norm": 5.2004170417785645,
|
1080 |
+
"learning_rate": 2.1642124051762607e-05,
|
1081 |
+
"loss": 0.0035,
|
1082 |
+
"step": 1520
|
1083 |
+
},
|
1084 |
+
{
|
1085 |
+
"epoch": 1.8425647200481636,
|
1086 |
+
"grad_norm": 0.644066572189331,
|
1087 |
+
"learning_rate": 2.1419009370816602e-05,
|
1088 |
+
"loss": 0.0108,
|
1089 |
+
"step": 1530
|
1090 |
+
},
|
1091 |
+
{
|
1092 |
+
"epoch": 1.8546056592414208,
|
1093 |
+
"grad_norm": 7.4833760261535645,
|
1094 |
+
"learning_rate": 2.1195894689870593e-05,
|
1095 |
+
"loss": 0.0047,
|
1096 |
+
"step": 1540
|
1097 |
+
},
|
1098 |
+
{
|
1099 |
+
"epoch": 1.866646598434678,
|
1100 |
+
"grad_norm": 1.3332066535949707,
|
1101 |
+
"learning_rate": 2.097278000892459e-05,
|
1102 |
+
"loss": 0.0114,
|
1103 |
+
"step": 1550
|
1104 |
+
},
|
1105 |
+
{
|
1106 |
+
"epoch": 1.8786875376279348,
|
1107 |
+
"grad_norm": 22.227462768554688,
|
1108 |
+
"learning_rate": 2.074966532797858e-05,
|
1109 |
+
"loss": 0.007,
|
1110 |
+
"step": 1560
|
1111 |
+
},
|
1112 |
+
{
|
1113 |
+
"epoch": 1.890728476821192,
|
1114 |
+
"grad_norm": 0.06626732647418976,
|
1115 |
+
"learning_rate": 2.0526550647032576e-05,
|
1116 |
+
"loss": 0.0034,
|
1117 |
+
"step": 1570
|
1118 |
+
},
|
1119 |
+
{
|
1120 |
+
"epoch": 1.9027694160144493,
|
1121 |
+
"grad_norm": 0.06219644472002983,
|
1122 |
+
"learning_rate": 2.030343596608657e-05,
|
1123 |
+
"loss": 0.0161,
|
1124 |
+
"step": 1580
|
1125 |
+
},
|
1126 |
+
{
|
1127 |
+
"epoch": 1.914810355207706,
|
1128 |
+
"grad_norm": 14.115644454956055,
|
1129 |
+
"learning_rate": 2.0080321285140562e-05,
|
1130 |
+
"loss": 0.0213,
|
1131 |
+
"step": 1590
|
1132 |
+
},
|
1133 |
+
{
|
1134 |
+
"epoch": 1.9268512944009633,
|
1135 |
+
"grad_norm": 0.06290453672409058,
|
1136 |
+
"learning_rate": 1.9857206604194558e-05,
|
1137 |
+
"loss": 0.0196,
|
1138 |
+
"step": 1600
|
1139 |
+
},
|
1140 |
+
{
|
1141 |
+
"epoch": 1.9388922335942205,
|
1142 |
+
"grad_norm": 0.5245229601860046,
|
1143 |
+
"learning_rate": 1.9634091923248553e-05,
|
1144 |
+
"loss": 0.0067,
|
1145 |
+
"step": 1610
|
1146 |
+
},
|
1147 |
+
{
|
1148 |
+
"epoch": 1.9509331727874775,
|
1149 |
+
"grad_norm": 2.8521981239318848,
|
1150 |
+
"learning_rate": 1.9410977242302544e-05,
|
1151 |
+
"loss": 0.0193,
|
1152 |
+
"step": 1620
|
1153 |
+
},
|
1154 |
+
{
|
1155 |
+
"epoch": 1.9629741119807345,
|
1156 |
+
"grad_norm": 0.47254133224487305,
|
1157 |
+
"learning_rate": 1.9187862561356536e-05,
|
1158 |
+
"loss": 0.0199,
|
1159 |
+
"step": 1630
|
1160 |
+
},
|
1161 |
+
{
|
1162 |
+
"epoch": 1.9750150511739917,
|
1163 |
+
"grad_norm": 0.11533841490745544,
|
1164 |
+
"learning_rate": 1.896474788041053e-05,
|
1165 |
+
"loss": 0.0031,
|
1166 |
+
"step": 1640
|
1167 |
+
},
|
1168 |
+
{
|
1169 |
+
"epoch": 1.9870559903672487,
|
1170 |
+
"grad_norm": 12.329407691955566,
|
1171 |
+
"learning_rate": 1.8741633199464527e-05,
|
1172 |
+
"loss": 0.012,
|
1173 |
+
"step": 1650
|
1174 |
+
},
|
1175 |
+
{
|
1176 |
+
"epoch": 1.9990969295605057,
|
1177 |
+
"grad_norm": 4.933974742889404,
|
1178 |
+
"learning_rate": 1.8518518518518518e-05,
|
1179 |
+
"loss": 0.0152,
|
1180 |
+
"step": 1660
|
1181 |
+
},
|
1182 |
+
{
|
1183 |
+
"epoch": 1.9990969295605057,
|
1184 |
+
"eval_f1": 0.9903039700741034,
|
1185 |
+
"eval_loss": 0.020218143239617348,
|
1186 |
+
"eval_runtime": 101.1159,
|
1187 |
+
"eval_samples_per_second": 116.797,
|
1188 |
+
"eval_steps_per_second": 3.659,
|
1189 |
+
"step": 1660
|
1190 |
+
},
|
1191 |
+
{
|
1192 |
+
"epoch": 2.011438892233594,
|
1193 |
+
"grad_norm": 0.7306141257286072,
|
1194 |
+
"learning_rate": 1.8295403837572513e-05,
|
1195 |
+
"loss": 0.0078,
|
1196 |
+
"step": 1670
|
1197 |
+
},
|
1198 |
+
{
|
1199 |
+
"epoch": 2.0234798314268514,
|
1200 |
+
"grad_norm": 0.0547499842941761,
|
1201 |
+
"learning_rate": 1.8072289156626505e-05,
|
1202 |
+
"loss": 0.0102,
|
1203 |
+
"step": 1680
|
1204 |
+
},
|
1205 |
+
{
|
1206 |
+
"epoch": 2.035520770620108,
|
1207 |
+
"grad_norm": 1.1440118551254272,
|
1208 |
+
"learning_rate": 1.78491744756805e-05,
|
1209 |
+
"loss": 0.0011,
|
1210 |
+
"step": 1690
|
1211 |
+
},
|
1212 |
+
{
|
1213 |
+
"epoch": 2.0475617098133654,
|
1214 |
+
"grad_norm": 0.14842332899570465,
|
1215 |
+
"learning_rate": 1.7626059794734495e-05,
|
1216 |
+
"loss": 0.0087,
|
1217 |
+
"step": 1700
|
1218 |
+
},
|
1219 |
+
{
|
1220 |
+
"epoch": 2.0596026490066226,
|
1221 |
+
"grad_norm": 0.18685872852802277,
|
1222 |
+
"learning_rate": 1.7402945113788487e-05,
|
1223 |
+
"loss": 0.002,
|
1224 |
+
"step": 1710
|
1225 |
+
},
|
1226 |
+
{
|
1227 |
+
"epoch": 2.07164358819988,
|
1228 |
+
"grad_norm": 13.721619606018066,
|
1229 |
+
"learning_rate": 1.7179830432842482e-05,
|
1230 |
+
"loss": 0.0052,
|
1231 |
+
"step": 1720
|
1232 |
+
},
|
1233 |
+
{
|
1234 |
+
"epoch": 2.0836845273931366,
|
1235 |
+
"grad_norm": 1.7337596416473389,
|
1236 |
+
"learning_rate": 1.6956715751896478e-05,
|
1237 |
+
"loss": 0.0007,
|
1238 |
+
"step": 1730
|
1239 |
+
},
|
1240 |
+
{
|
1241 |
+
"epoch": 2.095725466586394,
|
1242 |
+
"grad_norm": 0.023796595633029938,
|
1243 |
+
"learning_rate": 1.673360107095047e-05,
|
1244 |
+
"loss": 0.0006,
|
1245 |
+
"step": 1740
|
1246 |
+
},
|
1247 |
+
{
|
1248 |
+
"epoch": 2.107766405779651,
|
1249 |
+
"grad_norm": 0.21926841139793396,
|
1250 |
+
"learning_rate": 1.651048639000446e-05,
|
1251 |
+
"loss": 0.0002,
|
1252 |
+
"step": 1750
|
1253 |
+
},
|
1254 |
+
{
|
1255 |
+
"epoch": 2.119807344972908,
|
1256 |
+
"grad_norm": 0.08207955211400986,
|
1257 |
+
"learning_rate": 1.6287371709058456e-05,
|
1258 |
+
"loss": 0.0006,
|
1259 |
+
"step": 1760
|
1260 |
+
},
|
1261 |
+
{
|
1262 |
+
"epoch": 2.131848284166165,
|
1263 |
+
"grad_norm": 0.06928452104330063,
|
1264 |
+
"learning_rate": 1.606425702811245e-05,
|
1265 |
+
"loss": 0.0004,
|
1266 |
+
"step": 1770
|
1267 |
+
},
|
1268 |
+
{
|
1269 |
+
"epoch": 2.1438892233594222,
|
1270 |
+
"grad_norm": 0.028248103335499763,
|
1271 |
+
"learning_rate": 1.5841142347166443e-05,
|
1272 |
+
"loss": 0.0007,
|
1273 |
+
"step": 1780
|
1274 |
+
},
|
1275 |
+
{
|
1276 |
+
"epoch": 2.155930162552679,
|
1277 |
+
"grad_norm": 0.13549260795116425,
|
1278 |
+
"learning_rate": 1.5618027666220438e-05,
|
1279 |
+
"loss": 0.0012,
|
1280 |
+
"step": 1790
|
1281 |
+
},
|
1282 |
+
{
|
1283 |
+
"epoch": 2.167971101745936,
|
1284 |
+
"grad_norm": 0.24678213894367218,
|
1285 |
+
"learning_rate": 1.5394912985274433e-05,
|
1286 |
+
"loss": 0.0002,
|
1287 |
+
"step": 1800
|
1288 |
+
},
|
1289 |
+
{
|
1290 |
+
"epoch": 2.1800120409391934,
|
1291 |
+
"grad_norm": 0.043063148856163025,
|
1292 |
+
"learning_rate": 1.5171798304328425e-05,
|
1293 |
+
"loss": 0.0103,
|
1294 |
+
"step": 1810
|
1295 |
+
},
|
1296 |
+
{
|
1297 |
+
"epoch": 2.19205298013245,
|
1298 |
+
"grad_norm": 7.088438510894775,
|
1299 |
+
"learning_rate": 1.494868362338242e-05,
|
1300 |
+
"loss": 0.0161,
|
1301 |
+
"step": 1820
|
1302 |
+
},
|
1303 |
+
{
|
1304 |
+
"epoch": 2.2040939193257074,
|
1305 |
+
"grad_norm": 2.1872127056121826,
|
1306 |
+
"learning_rate": 1.4725568942436414e-05,
|
1307 |
+
"loss": 0.0127,
|
1308 |
+
"step": 1830
|
1309 |
+
},
|
1310 |
+
{
|
1311 |
+
"epoch": 2.2161348585189646,
|
1312 |
+
"grad_norm": 2.7211828231811523,
|
1313 |
+
"learning_rate": 1.4502454261490405e-05,
|
1314 |
+
"loss": 0.0133,
|
1315 |
+
"step": 1840
|
1316 |
+
},
|
1317 |
+
{
|
1318 |
+
"epoch": 2.2281757977122214,
|
1319 |
+
"grad_norm": 0.6726425290107727,
|
1320 |
+
"learning_rate": 1.42793395805444e-05,
|
1321 |
+
"loss": 0.0004,
|
1322 |
+
"step": 1850
|
1323 |
+
},
|
1324 |
+
{
|
1325 |
+
"epoch": 2.2402167369054786,
|
1326 |
+
"grad_norm": 0.05588046833872795,
|
1327 |
+
"learning_rate": 1.4056224899598394e-05,
|
1328 |
+
"loss": 0.0003,
|
1329 |
+
"step": 1860
|
1330 |
+
},
|
1331 |
+
{
|
1332 |
+
"epoch": 2.252257676098736,
|
1333 |
+
"grad_norm": 0.2764362394809723,
|
1334 |
+
"learning_rate": 1.3833110218652387e-05,
|
1335 |
+
"loss": 0.0008,
|
1336 |
+
"step": 1870
|
1337 |
+
},
|
1338 |
+
{
|
1339 |
+
"epoch": 2.2642986152919926,
|
1340 |
+
"grad_norm": 0.05939013138413429,
|
1341 |
+
"learning_rate": 1.3609995537706383e-05,
|
1342 |
+
"loss": 0.0017,
|
1343 |
+
"step": 1880
|
1344 |
+
},
|
1345 |
+
{
|
1346 |
+
"epoch": 2.27633955448525,
|
1347 |
+
"grad_norm": 43.3293342590332,
|
1348 |
+
"learning_rate": 1.3386880856760376e-05,
|
1349 |
+
"loss": 0.0015,
|
1350 |
+
"step": 1890
|
1351 |
+
},
|
1352 |
+
{
|
1353 |
+
"epoch": 2.288380493678507,
|
1354 |
+
"grad_norm": 0.021860118955373764,
|
1355 |
+
"learning_rate": 1.3163766175814368e-05,
|
1356 |
+
"loss": 0.0093,
|
1357 |
+
"step": 1900
|
1358 |
+
},
|
1359 |
+
{
|
1360 |
+
"epoch": 2.300421432871764,
|
1361 |
+
"grad_norm": 0.5367532968521118,
|
1362 |
+
"learning_rate": 1.2940651494868363e-05,
|
1363 |
+
"loss": 0.0049,
|
1364 |
+
"step": 1910
|
1365 |
+
},
|
1366 |
+
{
|
1367 |
+
"epoch": 2.312462372065021,
|
1368 |
+
"grad_norm": 0.046667639166116714,
|
1369 |
+
"learning_rate": 1.2717536813922356e-05,
|
1370 |
+
"loss": 0.001,
|
1371 |
+
"step": 1920
|
1372 |
+
},
|
1373 |
+
{
|
1374 |
+
"epoch": 2.3245033112582782,
|
1375 |
+
"grad_norm": 0.019831106066703796,
|
1376 |
+
"learning_rate": 1.2494422132976352e-05,
|
1377 |
+
"loss": 0.0063,
|
1378 |
+
"step": 1930
|
1379 |
+
},
|
1380 |
+
{
|
1381 |
+
"epoch": 2.336544250451535,
|
1382 |
+
"grad_norm": 0.6081582903862,
|
1383 |
+
"learning_rate": 1.2271307452030343e-05,
|
1384 |
+
"loss": 0.006,
|
1385 |
+
"step": 1940
|
1386 |
+
},
|
1387 |
+
{
|
1388 |
+
"epoch": 2.3485851896447922,
|
1389 |
+
"grad_norm": 18.08721351623535,
|
1390 |
+
"learning_rate": 1.2048192771084338e-05,
|
1391 |
+
"loss": 0.0008,
|
1392 |
+
"step": 1950
|
1393 |
+
},
|
1394 |
+
{
|
1395 |
+
"epoch": 2.3606261288380495,
|
1396 |
+
"grad_norm": 4.2922892570495605,
|
1397 |
+
"learning_rate": 1.1825078090138332e-05,
|
1398 |
+
"loss": 0.0051,
|
1399 |
+
"step": 1960
|
1400 |
+
},
|
1401 |
+
{
|
1402 |
+
"epoch": 2.3726670680313067,
|
1403 |
+
"grad_norm": 0.017288248986005783,
|
1404 |
+
"learning_rate": 1.1601963409192325e-05,
|
1405 |
+
"loss": 0.0002,
|
1406 |
+
"step": 1970
|
1407 |
+
},
|
1408 |
+
{
|
1409 |
+
"epoch": 2.3847080072245634,
|
1410 |
+
"grad_norm": 0.011165632866322994,
|
1411 |
+
"learning_rate": 1.1378848728246319e-05,
|
1412 |
+
"loss": 0.0006,
|
1413 |
+
"step": 1980
|
1414 |
+
},
|
1415 |
+
{
|
1416 |
+
"epoch": 2.3967489464178207,
|
1417 |
+
"grad_norm": 0.038627851754426956,
|
1418 |
+
"learning_rate": 1.1155734047300314e-05,
|
1419 |
+
"loss": 0.0048,
|
1420 |
+
"step": 1990
|
1421 |
+
},
|
1422 |
+
{
|
1423 |
+
"epoch": 2.4087898856110774,
|
1424 |
+
"grad_norm": 0.03750608116388321,
|
1425 |
+
"learning_rate": 1.0932619366354306e-05,
|
1426 |
+
"loss": 0.0024,
|
1427 |
+
"step": 2000
|
1428 |
+
},
|
1429 |
+
{
|
1430 |
+
"epoch": 2.4208308248043346,
|
1431 |
+
"grad_norm": 0.14723530411720276,
|
1432 |
+
"learning_rate": 1.0709504685408301e-05,
|
1433 |
+
"loss": 0.0004,
|
1434 |
+
"step": 2010
|
1435 |
+
},
|
1436 |
+
{
|
1437 |
+
"epoch": 2.432871763997592,
|
1438 |
+
"grad_norm": 0.020417124032974243,
|
1439 |
+
"learning_rate": 1.0486390004462294e-05,
|
1440 |
+
"loss": 0.0006,
|
1441 |
+
"step": 2020
|
1442 |
+
},
|
1443 |
+
{
|
1444 |
+
"epoch": 2.444912703190849,
|
1445 |
+
"grad_norm": 0.12044423818588257,
|
1446 |
+
"learning_rate": 1.0263275323516288e-05,
|
1447 |
+
"loss": 0.0279,
|
1448 |
+
"step": 2030
|
1449 |
+
},
|
1450 |
+
{
|
1451 |
+
"epoch": 2.456953642384106,
|
1452 |
+
"grad_norm": 11.781253814697266,
|
1453 |
+
"learning_rate": 1.0040160642570281e-05,
|
1454 |
+
"loss": 0.0117,
|
1455 |
+
"step": 2040
|
1456 |
+
},
|
1457 |
+
{
|
1458 |
+
"epoch": 2.468994581577363,
|
1459 |
+
"grad_norm": 3.7121262550354004,
|
1460 |
+
"learning_rate": 9.817045961624276e-06,
|
1461 |
+
"loss": 0.0065,
|
1462 |
+
"step": 2050
|
1463 |
+
},
|
1464 |
+
{
|
1465 |
+
"epoch": 2.4810355207706203,
|
1466 |
+
"grad_norm": 0.06580865383148193,
|
1467 |
+
"learning_rate": 9.593931280678268e-06,
|
1468 |
+
"loss": 0.0033,
|
1469 |
+
"step": 2060
|
1470 |
+
},
|
1471 |
+
{
|
1472 |
+
"epoch": 2.493076459963877,
|
1473 |
+
"grad_norm": 0.04538366571068764,
|
1474 |
+
"learning_rate": 9.370816599732263e-06,
|
1475 |
+
"loss": 0.0013,
|
1476 |
+
"step": 2070
|
1477 |
+
},
|
1478 |
+
{
|
1479 |
+
"epoch": 2.5051173991571343,
|
1480 |
+
"grad_norm": 0.024890244007110596,
|
1481 |
+
"learning_rate": 9.147701918786257e-06,
|
1482 |
+
"loss": 0.0104,
|
1483 |
+
"step": 2080
|
1484 |
+
},
|
1485 |
+
{
|
1486 |
+
"epoch": 2.5171583383503915,
|
1487 |
+
"grad_norm": 0.15436908602714539,
|
1488 |
+
"learning_rate": 8.92458723784025e-06,
|
1489 |
+
"loss": 0.012,
|
1490 |
+
"step": 2090
|
1491 |
+
},
|
1492 |
+
{
|
1493 |
+
"epoch": 2.5291992775436483,
|
1494 |
+
"grad_norm": 0.18890848755836487,
|
1495 |
+
"learning_rate": 8.701472556894244e-06,
|
1496 |
+
"loss": 0.0006,
|
1497 |
+
"step": 2100
|
1498 |
+
},
|
1499 |
+
{
|
1500 |
+
"epoch": 2.5412402167369055,
|
1501 |
+
"grad_norm": 0.014392940327525139,
|
1502 |
+
"learning_rate": 8.478357875948239e-06,
|
1503 |
+
"loss": 0.0084,
|
1504 |
+
"step": 2110
|
1505 |
+
},
|
1506 |
+
{
|
1507 |
+
"epoch": 2.5532811559301627,
|
1508 |
+
"grad_norm": 0.05412567779421806,
|
1509 |
+
"learning_rate": 8.25524319500223e-06,
|
1510 |
+
"loss": 0.0082,
|
1511 |
+
"step": 2120
|
1512 |
+
},
|
1513 |
+
{
|
1514 |
+
"epoch": 2.5653220951234195,
|
1515 |
+
"grad_norm": 0.08901382237672806,
|
1516 |
+
"learning_rate": 8.032128514056226e-06,
|
1517 |
+
"loss": 0.0131,
|
1518 |
+
"step": 2130
|
1519 |
+
},
|
1520 |
+
{
|
1521 |
+
"epoch": 2.5773630343166767,
|
1522 |
+
"grad_norm": 0.03455715253949165,
|
1523 |
+
"learning_rate": 7.809013833110219e-06,
|
1524 |
+
"loss": 0.0002,
|
1525 |
+
"step": 2140
|
1526 |
+
},
|
1527 |
+
{
|
1528 |
+
"epoch": 2.589403973509934,
|
1529 |
+
"grad_norm": 11.11595630645752,
|
1530 |
+
"learning_rate": 7.5858991521642126e-06,
|
1531 |
+
"loss": 0.0113,
|
1532 |
+
"step": 2150
|
1533 |
+
},
|
1534 |
+
{
|
1535 |
+
"epoch": 2.601444912703191,
|
1536 |
+
"grad_norm": 3.2087442874908447,
|
1537 |
+
"learning_rate": 7.362784471218207e-06,
|
1538 |
+
"loss": 0.0094,
|
1539 |
+
"step": 2160
|
1540 |
+
},
|
1541 |
+
{
|
1542 |
+
"epoch": 2.613485851896448,
|
1543 |
+
"grad_norm": 26.5289306640625,
|
1544 |
+
"learning_rate": 7.1396697902722e-06,
|
1545 |
+
"loss": 0.0035,
|
1546 |
+
"step": 2170
|
1547 |
+
},
|
1548 |
+
{
|
1549 |
+
"epoch": 2.625526791089705,
|
1550 |
+
"grad_norm": 0.03234562277793884,
|
1551 |
+
"learning_rate": 6.916555109326194e-06,
|
1552 |
+
"loss": 0.0028,
|
1553 |
+
"step": 2180
|
1554 |
+
},
|
1555 |
+
{
|
1556 |
+
"epoch": 2.637567730282962,
|
1557 |
+
"grad_norm": 0.032831065356731415,
|
1558 |
+
"learning_rate": 6.693440428380188e-06,
|
1559 |
+
"loss": 0.0025,
|
1560 |
+
"step": 2190
|
1561 |
+
},
|
1562 |
+
{
|
1563 |
+
"epoch": 2.649608669476219,
|
1564 |
+
"grad_norm": 1.0490331649780273,
|
1565 |
+
"learning_rate": 6.4703257474341815e-06,
|
1566 |
+
"loss": 0.0024,
|
1567 |
+
"step": 2200
|
1568 |
+
},
|
1569 |
+
{
|
1570 |
+
"epoch": 2.6616496086694763,
|
1571 |
+
"grad_norm": 0.14360417425632477,
|
1572 |
+
"learning_rate": 6.247211066488176e-06,
|
1573 |
+
"loss": 0.0015,
|
1574 |
+
"step": 2210
|
1575 |
+
},
|
1576 |
+
{
|
1577 |
+
"epoch": 2.6736905478627335,
|
1578 |
+
"grad_norm": 5.024721622467041,
|
1579 |
+
"learning_rate": 6.024096385542169e-06,
|
1580 |
+
"loss": 0.0042,
|
1581 |
+
"step": 2220
|
1582 |
+
},
|
1583 |
+
{
|
1584 |
+
"epoch": 2.6857314870559903,
|
1585 |
+
"grad_norm": 27.92305564880371,
|
1586 |
+
"learning_rate": 5.800981704596163e-06,
|
1587 |
+
"loss": 0.0163,
|
1588 |
+
"step": 2230
|
1589 |
+
},
|
1590 |
+
{
|
1591 |
+
"epoch": 2.6977724262492475,
|
1592 |
+
"grad_norm": 0.07593529671430588,
|
1593 |
+
"learning_rate": 5.577867023650157e-06,
|
1594 |
+
"loss": 0.0015,
|
1595 |
+
"step": 2240
|
1596 |
+
},
|
1597 |
+
{
|
1598 |
+
"epoch": 2.7098133654425043,
|
1599 |
+
"grad_norm": 0.033851731568574905,
|
1600 |
+
"learning_rate": 5.3547523427041504e-06,
|
1601 |
+
"loss": 0.0002,
|
1602 |
+
"step": 2250
|
1603 |
+
},
|
1604 |
+
{
|
1605 |
+
"epoch": 2.7218543046357615,
|
1606 |
+
"grad_norm": 0.08031222969293594,
|
1607 |
+
"learning_rate": 5.131637661758144e-06,
|
1608 |
+
"loss": 0.0013,
|
1609 |
+
"step": 2260
|
1610 |
+
},
|
1611 |
+
{
|
1612 |
+
"epoch": 2.7338952438290187,
|
1613 |
+
"grad_norm": 4.381354808807373,
|
1614 |
+
"learning_rate": 4.908522980812138e-06,
|
1615 |
+
"loss": 0.008,
|
1616 |
+
"step": 2270
|
1617 |
+
},
|
1618 |
+
{
|
1619 |
+
"epoch": 2.745936183022276,
|
1620 |
+
"grad_norm": 0.03481106460094452,
|
1621 |
+
"learning_rate": 4.685408299866132e-06,
|
1622 |
+
"loss": 0.0015,
|
1623 |
+
"step": 2280
|
1624 |
+
},
|
1625 |
+
{
|
1626 |
+
"epoch": 2.7579771222155327,
|
1627 |
+
"grad_norm": 20.805809020996094,
|
1628 |
+
"learning_rate": 4.462293618920125e-06,
|
1629 |
+
"loss": 0.0034,
|
1630 |
+
"step": 2290
|
1631 |
+
},
|
1632 |
+
{
|
1633 |
+
"epoch": 2.77001806140879,
|
1634 |
+
"grad_norm": 0.6730156540870667,
|
1635 |
+
"learning_rate": 4.239178937974119e-06,
|
1636 |
+
"loss": 0.0005,
|
1637 |
+
"step": 2300
|
1638 |
+
},
|
1639 |
+
{
|
1640 |
+
"epoch": 2.7820590006020467,
|
1641 |
+
"grad_norm": 0.05091029778122902,
|
1642 |
+
"learning_rate": 4.016064257028113e-06,
|
1643 |
+
"loss": 0.0083,
|
1644 |
+
"step": 2310
|
1645 |
+
},
|
1646 |
+
{
|
1647 |
+
"epoch": 2.794099939795304,
|
1648 |
+
"grad_norm": 10.03028392791748,
|
1649 |
+
"learning_rate": 3.7929495760821063e-06,
|
1650 |
+
"loss": 0.0087,
|
1651 |
+
"step": 2320
|
1652 |
+
},
|
1653 |
+
{
|
1654 |
+
"epoch": 2.806140878988561,
|
1655 |
+
"grad_norm": 0.1488943099975586,
|
1656 |
+
"learning_rate": 3.5698348951361e-06,
|
1657 |
+
"loss": 0.0083,
|
1658 |
+
"step": 2330
|
1659 |
+
},
|
1660 |
+
{
|
1661 |
+
"epoch": 2.8181818181818183,
|
1662 |
+
"grad_norm": 0.9400956034660339,
|
1663 |
+
"learning_rate": 3.346720214190094e-06,
|
1664 |
+
"loss": 0.0005,
|
1665 |
+
"step": 2340
|
1666 |
+
},
|
1667 |
+
{
|
1668 |
+
"epoch": 2.830222757375075,
|
1669 |
+
"grad_norm": 0.054990120232105255,
|
1670 |
+
"learning_rate": 3.123605533244088e-06,
|
1671 |
+
"loss": 0.0022,
|
1672 |
+
"step": 2350
|
1673 |
+
},
|
1674 |
+
{
|
1675 |
+
"epoch": 2.8422636965683323,
|
1676 |
+
"grad_norm": 0.18718010187149048,
|
1677 |
+
"learning_rate": 2.9004908522980813e-06,
|
1678 |
+
"loss": 0.0059,
|
1679 |
+
"step": 2360
|
1680 |
+
},
|
1681 |
+
{
|
1682 |
+
"epoch": 2.8543046357615895,
|
1683 |
+
"grad_norm": 0.06312959641218185,
|
1684 |
+
"learning_rate": 2.6773761713520752e-06,
|
1685 |
+
"loss": 0.007,
|
1686 |
+
"step": 2370
|
1687 |
+
},
|
1688 |
+
{
|
1689 |
+
"epoch": 2.8663455749548463,
|
1690 |
+
"grad_norm": 0.41855642199516296,
|
1691 |
+
"learning_rate": 2.454261490406069e-06,
|
1692 |
+
"loss": 0.0062,
|
1693 |
+
"step": 2380
|
1694 |
+
},
|
1695 |
+
{
|
1696 |
+
"epoch": 2.8783865141481035,
|
1697 |
+
"grad_norm": 0.06557010114192963,
|
1698 |
+
"learning_rate": 2.2311468094600625e-06,
|
1699 |
+
"loss": 0.0047,
|
1700 |
+
"step": 2390
|
1701 |
+
},
|
1702 |
+
{
|
1703 |
+
"epoch": 2.8904274533413608,
|
1704 |
+
"grad_norm": 0.06863750517368317,
|
1705 |
+
"learning_rate": 2.0080321285140564e-06,
|
1706 |
+
"loss": 0.0003,
|
1707 |
+
"step": 2400
|
1708 |
+
},
|
1709 |
+
{
|
1710 |
+
"epoch": 2.902468392534618,
|
1711 |
+
"grad_norm": 0.14578412473201752,
|
1712 |
+
"learning_rate": 1.78491744756805e-06,
|
1713 |
+
"loss": 0.0072,
|
1714 |
+
"step": 2410
|
1715 |
+
},
|
1716 |
+
{
|
1717 |
+
"epoch": 2.9145093317278747,
|
1718 |
+
"grad_norm": 0.020962489768862724,
|
1719 |
+
"learning_rate": 1.561802766622044e-06,
|
1720 |
+
"loss": 0.0032,
|
1721 |
+
"step": 2420
|
1722 |
+
},
|
1723 |
+
{
|
1724 |
+
"epoch": 2.926550270921132,
|
1725 |
+
"grad_norm": 0.07511858642101288,
|
1726 |
+
"learning_rate": 1.3386880856760376e-06,
|
1727 |
+
"loss": 0.0161,
|
1728 |
+
"step": 2430
|
1729 |
+
},
|
1730 |
+
{
|
1731 |
+
"epoch": 2.9385912101143887,
|
1732 |
+
"grad_norm": 12.450850486755371,
|
1733 |
+
"learning_rate": 1.1155734047300313e-06,
|
1734 |
+
"loss": 0.0074,
|
1735 |
+
"step": 2440
|
1736 |
+
},
|
1737 |
+
{
|
1738 |
+
"epoch": 2.950632149307646,
|
1739 |
+
"grad_norm": 0.02944479137659073,
|
1740 |
+
"learning_rate": 8.92458723784025e-07,
|
1741 |
+
"loss": 0.0002,
|
1742 |
+
"step": 2450
|
1743 |
+
},
|
1744 |
+
{
|
1745 |
+
"epoch": 2.962673088500903,
|
1746 |
+
"grad_norm": 0.387787789106369,
|
1747 |
+
"learning_rate": 6.693440428380188e-07,
|
1748 |
+
"loss": 0.0018,
|
1749 |
+
"step": 2460
|
1750 |
+
},
|
1751 |
+
{
|
1752 |
+
"epoch": 2.9747140276941604,
|
1753 |
+
"grad_norm": 0.2808152139186859,
|
1754 |
+
"learning_rate": 4.462293618920125e-07,
|
1755 |
+
"loss": 0.0055,
|
1756 |
+
"step": 2470
|
1757 |
+
},
|
1758 |
+
{
|
1759 |
+
"epoch": 2.986754966887417,
|
1760 |
+
"grad_norm": 0.053598176687955856,
|
1761 |
+
"learning_rate": 2.2311468094600626e-07,
|
1762 |
+
"loss": 0.0051,
|
1763 |
+
"step": 2480
|
1764 |
+
},
|
1765 |
+
{
|
1766 |
+
"epoch": 2.9987959060806744,
|
1767 |
+
"grad_norm": 0.3022330403327942,
|
1768 |
+
"learning_rate": 0.0,
|
1769 |
+
"loss": 0.0003,
|
1770 |
+
"step": 2490
|
1771 |
+
},
|
1772 |
+
{
|
1773 |
+
"epoch": 2.9987959060806744,
|
1774 |
+
"eval_f1": 0.9916771315123634,
|
1775 |
+
"eval_loss": 0.011577575467526913,
|
1776 |
+
"eval_runtime": 101.2234,
|
1777 |
+
"eval_samples_per_second": 116.673,
|
1778 |
+
"eval_steps_per_second": 3.655,
|
1779 |
+
"step": 2490
|
1780 |
+
},
|
1781 |
+
{
|
1782 |
+
"epoch": 2.9987959060806744,
|
1783 |
+
"step": 2490,
|
1784 |
+
"total_flos": 1.3193601009550295e+19,
|
1785 |
+
"train_loss": 0.06868853225451277,
|
1786 |
+
"train_runtime": 5741.3098,
|
1787 |
+
"train_samples_per_second": 55.537,
|
1788 |
+
"train_steps_per_second": 0.434
|
1789 |
+
}
|
1790 |
+
],
|
1791 |
+
"logging_steps": 10,
|
1792 |
+
"max_steps": 2490,
|
1793 |
+
"num_input_tokens_seen": 0,
|
1794 |
+
"num_train_epochs": 3,
|
1795 |
+
"save_steps": 500,
|
1796 |
+
"stateful_callbacks": {
|
1797 |
+
"TrainerControl": {
|
1798 |
+
"args": {
|
1799 |
+
"should_epoch_stop": false,
|
1800 |
+
"should_evaluate": false,
|
1801 |
+
"should_log": false,
|
1802 |
+
"should_save": true,
|
1803 |
+
"should_training_stop": true
|
1804 |
+
},
|
1805 |
+
"attributes": {}
|
1806 |
+
}
|
1807 |
+
},
|
1808 |
+
"total_flos": 1.3193601009550295e+19,
|
1809 |
+
"train_batch_size": 32,
|
1810 |
+
"trial_name": null,
|
1811 |
+
"trial_params": null
|
1812 |
+
}
|