End of training
Browse files- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +1240 -0
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 1.4134330246261506e+19,
|
| 4 |
+
"train_loss": 0.10345972847079468,
|
| 5 |
+
"train_runtime": 4872.3509,
|
| 6 |
+
"train_samples_per_second": 11.138,
|
| 7 |
+
"train_steps_per_second": 0.348
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 1.4134330246261506e+19,
|
| 4 |
+
"train_loss": 0.10345972847079468,
|
| 5 |
+
"train_runtime": 4872.3509,
|
| 6 |
+
"train_samples_per_second": 11.138,
|
| 7 |
+
"train_steps_per_second": 0.348
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1240 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9951154529307282,
|
| 3 |
+
"best_model_checkpoint": "vit-hybrid-base-bit-384-finetuned-ibird/checkpoint-1131",
|
| 4 |
+
"epoch": 2.9973474801061006,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1695,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.02,
|
| 13 |
+
"grad_norm": 10.60638427734375,
|
| 14 |
+
"learning_rate": 2.9411764705882355e-06,
|
| 15 |
+
"loss": 3.3842,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.04,
|
| 20 |
+
"grad_norm": 10.456100463867188,
|
| 21 |
+
"learning_rate": 5.882352941176471e-06,
|
| 22 |
+
"loss": 3.1282,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.05,
|
| 27 |
+
"grad_norm": 9.986790657043457,
|
| 28 |
+
"learning_rate": 8.823529411764707e-06,
|
| 29 |
+
"loss": 2.8815,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.07,
|
| 34 |
+
"grad_norm": 10.580012321472168,
|
| 35 |
+
"learning_rate": 1.1764705882352942e-05,
|
| 36 |
+
"loss": 2.3087,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.09,
|
| 41 |
+
"grad_norm": 7.9744133949279785,
|
| 42 |
+
"learning_rate": 1.4705882352941177e-05,
|
| 43 |
+
"loss": 1.5822,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.11,
|
| 48 |
+
"grad_norm": 6.688446044921875,
|
| 49 |
+
"learning_rate": 1.7647058823529414e-05,
|
| 50 |
+
"loss": 0.9381,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.12,
|
| 55 |
+
"grad_norm": 4.64487361907959,
|
| 56 |
+
"learning_rate": 2.058823529411765e-05,
|
| 57 |
+
"loss": 0.4115,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.14,
|
| 62 |
+
"grad_norm": 5.120860576629639,
|
| 63 |
+
"learning_rate": 2.3529411764705884e-05,
|
| 64 |
+
"loss": 0.1915,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.16,
|
| 69 |
+
"grad_norm": 2.6903035640716553,
|
| 70 |
+
"learning_rate": 2.647058823529412e-05,
|
| 71 |
+
"loss": 0.0935,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.18,
|
| 76 |
+
"grad_norm": 2.5217204093933105,
|
| 77 |
+
"learning_rate": 2.9411764705882354e-05,
|
| 78 |
+
"loss": 0.0949,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.19,
|
| 83 |
+
"grad_norm": 0.7664793133735657,
|
| 84 |
+
"learning_rate": 3.235294117647059e-05,
|
| 85 |
+
"loss": 0.0545,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.21,
|
| 90 |
+
"grad_norm": 4.868194580078125,
|
| 91 |
+
"learning_rate": 3.529411764705883e-05,
|
| 92 |
+
"loss": 0.0495,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.23,
|
| 97 |
+
"grad_norm": 3.152761459350586,
|
| 98 |
+
"learning_rate": 3.8235294117647055e-05,
|
| 99 |
+
"loss": 0.0566,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.25,
|
| 104 |
+
"grad_norm": 3.1319901943206787,
|
| 105 |
+
"learning_rate": 4.11764705882353e-05,
|
| 106 |
+
"loss": 0.0522,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.27,
|
| 111 |
+
"grad_norm": 2.599245309829712,
|
| 112 |
+
"learning_rate": 4.411764705882353e-05,
|
| 113 |
+
"loss": 0.0605,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.28,
|
| 118 |
+
"grad_norm": 0.845979630947113,
|
| 119 |
+
"learning_rate": 4.705882352941177e-05,
|
| 120 |
+
"loss": 0.0521,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.3,
|
| 125 |
+
"grad_norm": 6.3421759605407715,
|
| 126 |
+
"learning_rate": 5e-05,
|
| 127 |
+
"loss": 0.0849,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.32,
|
| 132 |
+
"grad_norm": 2.421968698501587,
|
| 133 |
+
"learning_rate": 4.967213114754098e-05,
|
| 134 |
+
"loss": 0.0502,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.34,
|
| 139 |
+
"grad_norm": 0.6349968910217285,
|
| 140 |
+
"learning_rate": 4.934426229508197e-05,
|
| 141 |
+
"loss": 0.0161,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.35,
|
| 146 |
+
"grad_norm": 0.3493717908859253,
|
| 147 |
+
"learning_rate": 4.9016393442622957e-05,
|
| 148 |
+
"loss": 0.027,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.37,
|
| 153 |
+
"grad_norm": 0.24897021055221558,
|
| 154 |
+
"learning_rate": 4.868852459016394e-05,
|
| 155 |
+
"loss": 0.0481,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.39,
|
| 160 |
+
"grad_norm": 0.10974353551864624,
|
| 161 |
+
"learning_rate": 4.836065573770492e-05,
|
| 162 |
+
"loss": 0.017,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.41,
|
| 167 |
+
"grad_norm": 2.431795358657837,
|
| 168 |
+
"learning_rate": 4.8032786885245904e-05,
|
| 169 |
+
"loss": 0.0494,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.42,
|
| 174 |
+
"grad_norm": 3.4805123805999756,
|
| 175 |
+
"learning_rate": 4.770491803278689e-05,
|
| 176 |
+
"loss": 0.0388,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.44,
|
| 181 |
+
"grad_norm": 2.0460708141326904,
|
| 182 |
+
"learning_rate": 4.737704918032787e-05,
|
| 183 |
+
"loss": 0.0507,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.46,
|
| 188 |
+
"grad_norm": 0.7923760414123535,
|
| 189 |
+
"learning_rate": 4.704918032786885e-05,
|
| 190 |
+
"loss": 0.0781,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.48,
|
| 195 |
+
"grad_norm": 2.829329252243042,
|
| 196 |
+
"learning_rate": 4.672131147540984e-05,
|
| 197 |
+
"loss": 0.0446,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.5,
|
| 202 |
+
"grad_norm": 0.8546658158302307,
|
| 203 |
+
"learning_rate": 4.6393442622950825e-05,
|
| 204 |
+
"loss": 0.0543,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.51,
|
| 209 |
+
"grad_norm": 0.2265116274356842,
|
| 210 |
+
"learning_rate": 4.6065573770491805e-05,
|
| 211 |
+
"loss": 0.0528,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.53,
|
| 216 |
+
"grad_norm": 2.7868409156799316,
|
| 217 |
+
"learning_rate": 4.5737704918032786e-05,
|
| 218 |
+
"loss": 0.0442,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.55,
|
| 223 |
+
"grad_norm": 0.05122404918074608,
|
| 224 |
+
"learning_rate": 4.540983606557377e-05,
|
| 225 |
+
"loss": 0.0473,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.57,
|
| 230 |
+
"grad_norm": 4.265463829040527,
|
| 231 |
+
"learning_rate": 4.508196721311476e-05,
|
| 232 |
+
"loss": 0.0358,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.58,
|
| 237 |
+
"grad_norm": 0.3292916417121887,
|
| 238 |
+
"learning_rate": 4.475409836065574e-05,
|
| 239 |
+
"loss": 0.0291,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.6,
|
| 244 |
+
"grad_norm": 1.2873409986495972,
|
| 245 |
+
"learning_rate": 4.442622950819673e-05,
|
| 246 |
+
"loss": 0.0681,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.62,
|
| 251 |
+
"grad_norm": 0.5806503891944885,
|
| 252 |
+
"learning_rate": 4.409836065573771e-05,
|
| 253 |
+
"loss": 0.0838,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.64,
|
| 258 |
+
"grad_norm": 4.996669769287109,
|
| 259 |
+
"learning_rate": 4.377049180327869e-05,
|
| 260 |
+
"loss": 0.0182,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.65,
|
| 265 |
+
"grad_norm": 0.058223385363817215,
|
| 266 |
+
"learning_rate": 4.3442622950819674e-05,
|
| 267 |
+
"loss": 0.0454,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.67,
|
| 272 |
+
"grad_norm": 0.5088571906089783,
|
| 273 |
+
"learning_rate": 4.311475409836066e-05,
|
| 274 |
+
"loss": 0.0444,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.69,
|
| 279 |
+
"grad_norm": 6.739996433258057,
|
| 280 |
+
"learning_rate": 4.278688524590164e-05,
|
| 281 |
+
"loss": 0.0445,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.71,
|
| 286 |
+
"grad_norm": 0.22746062278747559,
|
| 287 |
+
"learning_rate": 4.245901639344262e-05,
|
| 288 |
+
"loss": 0.0258,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.73,
|
| 293 |
+
"grad_norm": 0.04663905128836632,
|
| 294 |
+
"learning_rate": 4.213114754098361e-05,
|
| 295 |
+
"loss": 0.0023,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.74,
|
| 300 |
+
"grad_norm": 0.3145718574523926,
|
| 301 |
+
"learning_rate": 4.1803278688524595e-05,
|
| 302 |
+
"loss": 0.0205,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.76,
|
| 307 |
+
"grad_norm": 0.06759347021579742,
|
| 308 |
+
"learning_rate": 4.1475409836065575e-05,
|
| 309 |
+
"loss": 0.0779,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.78,
|
| 314 |
+
"grad_norm": 0.7269498705863953,
|
| 315 |
+
"learning_rate": 4.1147540983606556e-05,
|
| 316 |
+
"loss": 0.0194,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.8,
|
| 321 |
+
"grad_norm": 0.6663650274276733,
|
| 322 |
+
"learning_rate": 4.081967213114754e-05,
|
| 323 |
+
"loss": 0.0495,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.81,
|
| 328 |
+
"grad_norm": 0.015244757756590843,
|
| 329 |
+
"learning_rate": 4.049180327868853e-05,
|
| 330 |
+
"loss": 0.0217,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.83,
|
| 335 |
+
"grad_norm": 0.05687205493450165,
|
| 336 |
+
"learning_rate": 4.016393442622951e-05,
|
| 337 |
+
"loss": 0.0157,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.85,
|
| 342 |
+
"grad_norm": 0.06807345896959305,
|
| 343 |
+
"learning_rate": 3.983606557377049e-05,
|
| 344 |
+
"loss": 0.0021,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.87,
|
| 349 |
+
"grad_norm": 6.138463497161865,
|
| 350 |
+
"learning_rate": 3.950819672131148e-05,
|
| 351 |
+
"loss": 0.0473,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.88,
|
| 356 |
+
"grad_norm": 0.23685503005981445,
|
| 357 |
+
"learning_rate": 3.9180327868852464e-05,
|
| 358 |
+
"loss": 0.0289,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.9,
|
| 363 |
+
"grad_norm": 0.19919660687446594,
|
| 364 |
+
"learning_rate": 3.8852459016393444e-05,
|
| 365 |
+
"loss": 0.0659,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.92,
|
| 370 |
+
"grad_norm": 2.63227915763855,
|
| 371 |
+
"learning_rate": 3.8524590163934424e-05,
|
| 372 |
+
"loss": 0.0618,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.94,
|
| 377 |
+
"grad_norm": 0.18995435535907745,
|
| 378 |
+
"learning_rate": 3.819672131147541e-05,
|
| 379 |
+
"loss": 0.0151,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.95,
|
| 384 |
+
"grad_norm": 0.05399240553379059,
|
| 385 |
+
"learning_rate": 3.78688524590164e-05,
|
| 386 |
+
"loss": 0.0168,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.97,
|
| 391 |
+
"grad_norm": 0.13513301312923431,
|
| 392 |
+
"learning_rate": 3.754098360655738e-05,
|
| 393 |
+
"loss": 0.0492,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.99,
|
| 398 |
+
"grad_norm": 8.097681999206543,
|
| 399 |
+
"learning_rate": 3.721311475409836e-05,
|
| 400 |
+
"loss": 0.0997,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.0,
|
| 405 |
+
"eval_accuracy": 0.9906749555950266,
|
| 406 |
+
"eval_loss": 0.031034117564558983,
|
| 407 |
+
"eval_runtime": 124.3045,
|
| 408 |
+
"eval_samples_per_second": 18.117,
|
| 409 |
+
"eval_steps_per_second": 2.269,
|
| 410 |
+
"step": 565
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.01,
|
| 414 |
+
"grad_norm": 0.08757388591766357,
|
| 415 |
+
"learning_rate": 3.6885245901639346e-05,
|
| 416 |
+
"loss": 0.0243,
|
| 417 |
+
"step": 570
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.03,
|
| 421 |
+
"grad_norm": 0.03581409156322479,
|
| 422 |
+
"learning_rate": 3.655737704918033e-05,
|
| 423 |
+
"loss": 0.0101,
|
| 424 |
+
"step": 580
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.04,
|
| 428 |
+
"grad_norm": 0.01882335916161537,
|
| 429 |
+
"learning_rate": 3.622950819672131e-05,
|
| 430 |
+
"loss": 0.0034,
|
| 431 |
+
"step": 590
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.06,
|
| 435 |
+
"grad_norm": 0.12845346331596375,
|
| 436 |
+
"learning_rate": 3.590163934426229e-05,
|
| 437 |
+
"loss": 0.006,
|
| 438 |
+
"step": 600
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.08,
|
| 442 |
+
"grad_norm": 6.836170196533203,
|
| 443 |
+
"learning_rate": 3.557377049180328e-05,
|
| 444 |
+
"loss": 0.0054,
|
| 445 |
+
"step": 610
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 1.1,
|
| 449 |
+
"grad_norm": 0.014669831842184067,
|
| 450 |
+
"learning_rate": 3.524590163934427e-05,
|
| 451 |
+
"loss": 0.0007,
|
| 452 |
+
"step": 620
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.11,
|
| 456 |
+
"grad_norm": 0.0303161833435297,
|
| 457 |
+
"learning_rate": 3.491803278688525e-05,
|
| 458 |
+
"loss": 0.0272,
|
| 459 |
+
"step": 630
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.13,
|
| 463 |
+
"grad_norm": 0.010958039201796055,
|
| 464 |
+
"learning_rate": 3.459016393442623e-05,
|
| 465 |
+
"loss": 0.0016,
|
| 466 |
+
"step": 640
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.15,
|
| 470 |
+
"grad_norm": 0.10876976698637009,
|
| 471 |
+
"learning_rate": 3.4262295081967214e-05,
|
| 472 |
+
"loss": 0.0113,
|
| 473 |
+
"step": 650
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.17,
|
| 477 |
+
"grad_norm": 0.024475092068314552,
|
| 478 |
+
"learning_rate": 3.39344262295082e-05,
|
| 479 |
+
"loss": 0.0199,
|
| 480 |
+
"step": 660
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.18,
|
| 484 |
+
"grad_norm": 3.7424721717834473,
|
| 485 |
+
"learning_rate": 3.360655737704918e-05,
|
| 486 |
+
"loss": 0.0153,
|
| 487 |
+
"step": 670
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.2,
|
| 491 |
+
"grad_norm": 3.548234224319458,
|
| 492 |
+
"learning_rate": 3.327868852459017e-05,
|
| 493 |
+
"loss": 0.0124,
|
| 494 |
+
"step": 680
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.22,
|
| 498 |
+
"grad_norm": 0.018901199102401733,
|
| 499 |
+
"learning_rate": 3.295081967213115e-05,
|
| 500 |
+
"loss": 0.0059,
|
| 501 |
+
"step": 690
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 1.24,
|
| 505 |
+
"grad_norm": 0.01598968915641308,
|
| 506 |
+
"learning_rate": 3.2622950819672136e-05,
|
| 507 |
+
"loss": 0.0272,
|
| 508 |
+
"step": 700
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 1.26,
|
| 512 |
+
"grad_norm": 0.01613859459757805,
|
| 513 |
+
"learning_rate": 3.2295081967213116e-05,
|
| 514 |
+
"loss": 0.013,
|
| 515 |
+
"step": 710
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.27,
|
| 519 |
+
"grad_norm": 0.030256085097789764,
|
| 520 |
+
"learning_rate": 3.19672131147541e-05,
|
| 521 |
+
"loss": 0.0061,
|
| 522 |
+
"step": 720
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 1.29,
|
| 526 |
+
"grad_norm": 0.08354157209396362,
|
| 527 |
+
"learning_rate": 3.163934426229508e-05,
|
| 528 |
+
"loss": 0.0008,
|
| 529 |
+
"step": 730
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 1.31,
|
| 533 |
+
"grad_norm": 0.1932225227355957,
|
| 534 |
+
"learning_rate": 3.131147540983606e-05,
|
| 535 |
+
"loss": 0.0019,
|
| 536 |
+
"step": 740
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.33,
|
| 540 |
+
"grad_norm": 0.4999471604824066,
|
| 541 |
+
"learning_rate": 3.098360655737705e-05,
|
| 542 |
+
"loss": 0.012,
|
| 543 |
+
"step": 750
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 1.34,
|
| 547 |
+
"grad_norm": 0.016076169908046722,
|
| 548 |
+
"learning_rate": 3.065573770491804e-05,
|
| 549 |
+
"loss": 0.0006,
|
| 550 |
+
"step": 760
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.36,
|
| 554 |
+
"grad_norm": 0.010563875548541546,
|
| 555 |
+
"learning_rate": 3.0327868852459017e-05,
|
| 556 |
+
"loss": 0.0034,
|
| 557 |
+
"step": 770
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 1.38,
|
| 561 |
+
"grad_norm": 0.020218759775161743,
|
| 562 |
+
"learning_rate": 3e-05,
|
| 563 |
+
"loss": 0.0016,
|
| 564 |
+
"step": 780
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 1.4,
|
| 568 |
+
"grad_norm": 0.016868168488144875,
|
| 569 |
+
"learning_rate": 2.967213114754098e-05,
|
| 570 |
+
"loss": 0.0033,
|
| 571 |
+
"step": 790
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.41,
|
| 575 |
+
"grad_norm": 0.013298319652676582,
|
| 576 |
+
"learning_rate": 2.934426229508197e-05,
|
| 577 |
+
"loss": 0.0037,
|
| 578 |
+
"step": 800
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.43,
|
| 582 |
+
"grad_norm": 0.062481291592121124,
|
| 583 |
+
"learning_rate": 2.901639344262295e-05,
|
| 584 |
+
"loss": 0.0118,
|
| 585 |
+
"step": 810
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.45,
|
| 589 |
+
"grad_norm": 1.4717094898223877,
|
| 590 |
+
"learning_rate": 2.8688524590163935e-05,
|
| 591 |
+
"loss": 0.0104,
|
| 592 |
+
"step": 820
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.47,
|
| 596 |
+
"grad_norm": 1.5437506437301636,
|
| 597 |
+
"learning_rate": 2.8360655737704922e-05,
|
| 598 |
+
"loss": 0.0062,
|
| 599 |
+
"step": 830
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.49,
|
| 603 |
+
"grad_norm": 0.03338282182812691,
|
| 604 |
+
"learning_rate": 2.8032786885245906e-05,
|
| 605 |
+
"loss": 0.0025,
|
| 606 |
+
"step": 840
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 1.5,
|
| 610 |
+
"grad_norm": 1.8838506937026978,
|
| 611 |
+
"learning_rate": 2.7704918032786886e-05,
|
| 612 |
+
"loss": 0.0193,
|
| 613 |
+
"step": 850
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 1.52,
|
| 617 |
+
"grad_norm": 0.007599662058055401,
|
| 618 |
+
"learning_rate": 2.737704918032787e-05,
|
| 619 |
+
"loss": 0.0083,
|
| 620 |
+
"step": 860
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.54,
|
| 624 |
+
"grad_norm": 0.0383157841861248,
|
| 625 |
+
"learning_rate": 2.7049180327868856e-05,
|
| 626 |
+
"loss": 0.0059,
|
| 627 |
+
"step": 870
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 1.56,
|
| 631 |
+
"grad_norm": 0.004423599224537611,
|
| 632 |
+
"learning_rate": 2.6721311475409837e-05,
|
| 633 |
+
"loss": 0.0019,
|
| 634 |
+
"step": 880
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 1.57,
|
| 638 |
+
"grad_norm": 7.753259658813477,
|
| 639 |
+
"learning_rate": 2.639344262295082e-05,
|
| 640 |
+
"loss": 0.0197,
|
| 641 |
+
"step": 890
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.59,
|
| 645 |
+
"grad_norm": 0.0051222569309175014,
|
| 646 |
+
"learning_rate": 2.6065573770491804e-05,
|
| 647 |
+
"loss": 0.0128,
|
| 648 |
+
"step": 900
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.61,
|
| 652 |
+
"grad_norm": 0.006367386784404516,
|
| 653 |
+
"learning_rate": 2.573770491803279e-05,
|
| 654 |
+
"loss": 0.0097,
|
| 655 |
+
"step": 910
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 1.63,
|
| 659 |
+
"grad_norm": 0.016590917482972145,
|
| 660 |
+
"learning_rate": 2.540983606557377e-05,
|
| 661 |
+
"loss": 0.0163,
|
| 662 |
+
"step": 920
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 1.64,
|
| 666 |
+
"grad_norm": 0.006148540414869785,
|
| 667 |
+
"learning_rate": 2.5081967213114754e-05,
|
| 668 |
+
"loss": 0.0107,
|
| 669 |
+
"step": 930
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 1.66,
|
| 673 |
+
"grad_norm": 0.021444685757160187,
|
| 674 |
+
"learning_rate": 2.4754098360655738e-05,
|
| 675 |
+
"loss": 0.008,
|
| 676 |
+
"step": 940
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 1.68,
|
| 680 |
+
"grad_norm": 0.27239102125167847,
|
| 681 |
+
"learning_rate": 2.442622950819672e-05,
|
| 682 |
+
"loss": 0.0009,
|
| 683 |
+
"step": 950
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 1.7,
|
| 687 |
+
"grad_norm": 2.461951971054077,
|
| 688 |
+
"learning_rate": 2.4098360655737705e-05,
|
| 689 |
+
"loss": 0.0224,
|
| 690 |
+
"step": 960
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 1.72,
|
| 694 |
+
"grad_norm": 0.009482527151703835,
|
| 695 |
+
"learning_rate": 2.377049180327869e-05,
|
| 696 |
+
"loss": 0.0032,
|
| 697 |
+
"step": 970
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 1.73,
|
| 701 |
+
"grad_norm": 0.0694417804479599,
|
| 702 |
+
"learning_rate": 2.3442622950819672e-05,
|
| 703 |
+
"loss": 0.001,
|
| 704 |
+
"step": 980
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.75,
|
| 708 |
+
"grad_norm": 0.008940008468925953,
|
| 709 |
+
"learning_rate": 2.311475409836066e-05,
|
| 710 |
+
"loss": 0.0063,
|
| 711 |
+
"step": 990
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 1.77,
|
| 715 |
+
"grad_norm": 0.008408569730818272,
|
| 716 |
+
"learning_rate": 2.278688524590164e-05,
|
| 717 |
+
"loss": 0.0026,
|
| 718 |
+
"step": 1000
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 1.79,
|
| 722 |
+
"grad_norm": 0.00473896786570549,
|
| 723 |
+
"learning_rate": 2.2459016393442626e-05,
|
| 724 |
+
"loss": 0.0162,
|
| 725 |
+
"step": 1010
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.8,
|
| 729 |
+
"grad_norm": 0.0444784052670002,
|
| 730 |
+
"learning_rate": 2.2131147540983607e-05,
|
| 731 |
+
"loss": 0.0004,
|
| 732 |
+
"step": 1020
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.82,
|
| 736 |
+
"grad_norm": 0.002952872309833765,
|
| 737 |
+
"learning_rate": 2.1803278688524594e-05,
|
| 738 |
+
"loss": 0.0079,
|
| 739 |
+
"step": 1030
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 1.84,
|
| 743 |
+
"grad_norm": 0.009093045257031918,
|
| 744 |
+
"learning_rate": 2.1475409836065574e-05,
|
| 745 |
+
"loss": 0.0049,
|
| 746 |
+
"step": 1040
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.86,
|
| 750 |
+
"grad_norm": 5.8584818840026855,
|
| 751 |
+
"learning_rate": 2.114754098360656e-05,
|
| 752 |
+
"loss": 0.0048,
|
| 753 |
+
"step": 1050
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 1.87,
|
| 757 |
+
"grad_norm": 0.025469429790973663,
|
| 758 |
+
"learning_rate": 2.081967213114754e-05,
|
| 759 |
+
"loss": 0.0118,
|
| 760 |
+
"step": 1060
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 1.89,
|
| 764 |
+
"grad_norm": 0.005341151263564825,
|
| 765 |
+
"learning_rate": 2.0491803278688525e-05,
|
| 766 |
+
"loss": 0.0016,
|
| 767 |
+
"step": 1070
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.91,
|
| 771 |
+
"grad_norm": 0.006102704908698797,
|
| 772 |
+
"learning_rate": 2.0163934426229508e-05,
|
| 773 |
+
"loss": 0.001,
|
| 774 |
+
"step": 1080
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.93,
|
| 778 |
+
"grad_norm": 0.009585659019649029,
|
| 779 |
+
"learning_rate": 1.9836065573770492e-05,
|
| 780 |
+
"loss": 0.0003,
|
| 781 |
+
"step": 1090
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 1.95,
|
| 785 |
+
"grad_norm": 0.003977091982960701,
|
| 786 |
+
"learning_rate": 1.9508196721311475e-05,
|
| 787 |
+
"loss": 0.0034,
|
| 788 |
+
"step": 1100
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.96,
|
| 792 |
+
"grad_norm": 0.008223881945014,
|
| 793 |
+
"learning_rate": 1.918032786885246e-05,
|
| 794 |
+
"loss": 0.0016,
|
| 795 |
+
"step": 1110
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 1.98,
|
| 799 |
+
"grad_norm": 0.014407954178750515,
|
| 800 |
+
"learning_rate": 1.8852459016393442e-05,
|
| 801 |
+
"loss": 0.0013,
|
| 802 |
+
"step": 1120
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 2.0,
|
| 806 |
+
"grad_norm": 0.3026231825351715,
|
| 807 |
+
"learning_rate": 1.8524590163934426e-05,
|
| 808 |
+
"loss": 0.0005,
|
| 809 |
+
"step": 1130
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 2.0,
|
| 813 |
+
"eval_accuracy": 0.9951154529307282,
|
| 814 |
+
"eval_loss": 0.01963997818529606,
|
| 815 |
+
"eval_runtime": 122.5197,
|
| 816 |
+
"eval_samples_per_second": 18.381,
|
| 817 |
+
"eval_steps_per_second": 2.302,
|
| 818 |
+
"step": 1131
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 2.02,
|
| 822 |
+
"grad_norm": 0.005760945845395327,
|
| 823 |
+
"learning_rate": 1.8196721311475413e-05,
|
| 824 |
+
"loss": 0.0028,
|
| 825 |
+
"step": 1140
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 2.03,
|
| 829 |
+
"grad_norm": 0.00677720969542861,
|
| 830 |
+
"learning_rate": 1.7868852459016393e-05,
|
| 831 |
+
"loss": 0.0008,
|
| 832 |
+
"step": 1150
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 2.05,
|
| 836 |
+
"grad_norm": 0.004843315575271845,
|
| 837 |
+
"learning_rate": 1.754098360655738e-05,
|
| 838 |
+
"loss": 0.002,
|
| 839 |
+
"step": 1160
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 2.07,
|
| 843 |
+
"grad_norm": 0.003445906564593315,
|
| 844 |
+
"learning_rate": 1.721311475409836e-05,
|
| 845 |
+
"loss": 0.0002,
|
| 846 |
+
"step": 1170
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 2.09,
|
| 850 |
+
"grad_norm": 0.0024833856150507927,
|
| 851 |
+
"learning_rate": 1.6885245901639347e-05,
|
| 852 |
+
"loss": 0.0028,
|
| 853 |
+
"step": 1180
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 2.1,
|
| 857 |
+
"grad_norm": 0.024947261437773705,
|
| 858 |
+
"learning_rate": 1.6557377049180328e-05,
|
| 859 |
+
"loss": 0.0003,
|
| 860 |
+
"step": 1190
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 2.12,
|
| 864 |
+
"grad_norm": 0.0022959029302001,
|
| 865 |
+
"learning_rate": 1.6229508196721314e-05,
|
| 866 |
+
"loss": 0.0002,
|
| 867 |
+
"step": 1200
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 2.14,
|
| 871 |
+
"grad_norm": 0.01967487297952175,
|
| 872 |
+
"learning_rate": 1.5901639344262295e-05,
|
| 873 |
+
"loss": 0.0003,
|
| 874 |
+
"step": 1210
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 2.16,
|
| 878 |
+
"grad_norm": 0.357111394405365,
|
| 879 |
+
"learning_rate": 1.557377049180328e-05,
|
| 880 |
+
"loss": 0.0005,
|
| 881 |
+
"step": 1220
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 2.18,
|
| 885 |
+
"grad_norm": 0.0058966344222426414,
|
| 886 |
+
"learning_rate": 1.5245901639344262e-05,
|
| 887 |
+
"loss": 0.0001,
|
| 888 |
+
"step": 1230
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 2.19,
|
| 892 |
+
"grad_norm": 0.0033721905201673508,
|
| 893 |
+
"learning_rate": 1.4918032786885247e-05,
|
| 894 |
+
"loss": 0.0002,
|
| 895 |
+
"step": 1240
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 2.21,
|
| 899 |
+
"grad_norm": 0.0022145358379930258,
|
| 900 |
+
"learning_rate": 1.4590163934426229e-05,
|
| 901 |
+
"loss": 0.0002,
|
| 902 |
+
"step": 1250
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 2.23,
|
| 906 |
+
"grad_norm": 0.035282738506793976,
|
| 907 |
+
"learning_rate": 1.4262295081967214e-05,
|
| 908 |
+
"loss": 0.0002,
|
| 909 |
+
"step": 1260
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 2.25,
|
| 913 |
+
"grad_norm": 0.0015229666605591774,
|
| 914 |
+
"learning_rate": 1.3934426229508196e-05,
|
| 915 |
+
"loss": 0.0002,
|
| 916 |
+
"step": 1270
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 2.26,
|
| 920 |
+
"grad_norm": 0.005502559244632721,
|
| 921 |
+
"learning_rate": 1.3606557377049181e-05,
|
| 922 |
+
"loss": 0.0021,
|
| 923 |
+
"step": 1280
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 2.28,
|
| 927 |
+
"grad_norm": 0.0072832051664590836,
|
| 928 |
+
"learning_rate": 1.3278688524590163e-05,
|
| 929 |
+
"loss": 0.0001,
|
| 930 |
+
"step": 1290
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 2.3,
|
| 934 |
+
"grad_norm": 0.001554520451463759,
|
| 935 |
+
"learning_rate": 1.2950819672131149e-05,
|
| 936 |
+
"loss": 0.0002,
|
| 937 |
+
"step": 1300
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 2.32,
|
| 941 |
+
"grad_norm": 0.0054123131558299065,
|
| 942 |
+
"learning_rate": 1.2622950819672132e-05,
|
| 943 |
+
"loss": 0.0002,
|
| 944 |
+
"step": 1310
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 2.33,
|
| 948 |
+
"grad_norm": 0.013851379975676537,
|
| 949 |
+
"learning_rate": 1.2295081967213116e-05,
|
| 950 |
+
"loss": 0.0066,
|
| 951 |
+
"step": 1320
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 2.35,
|
| 955 |
+
"grad_norm": 0.0056920950300991535,
|
| 956 |
+
"learning_rate": 1.19672131147541e-05,
|
| 957 |
+
"loss": 0.0002,
|
| 958 |
+
"step": 1330
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 2.37,
|
| 962 |
+
"grad_norm": 0.00168787338770926,
|
| 963 |
+
"learning_rate": 1.1639344262295083e-05,
|
| 964 |
+
"loss": 0.0002,
|
| 965 |
+
"step": 1340
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 2.39,
|
| 969 |
+
"grad_norm": 0.005322947632521391,
|
| 970 |
+
"learning_rate": 1.1311475409836065e-05,
|
| 971 |
+
"loss": 0.0002,
|
| 972 |
+
"step": 1350
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 2.4,
|
| 976 |
+
"grad_norm": 0.002933880779892206,
|
| 977 |
+
"learning_rate": 1.0983606557377048e-05,
|
| 978 |
+
"loss": 0.0002,
|
| 979 |
+
"step": 1360
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 2.42,
|
| 983 |
+
"grad_norm": 0.0015972905093804002,
|
| 984 |
+
"learning_rate": 1.0655737704918032e-05,
|
| 985 |
+
"loss": 0.0003,
|
| 986 |
+
"step": 1370
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 2.44,
|
| 990 |
+
"grad_norm": 0.0015742258401587605,
|
| 991 |
+
"learning_rate": 1.0327868852459017e-05,
|
| 992 |
+
"loss": 0.0003,
|
| 993 |
+
"step": 1380
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 2.46,
|
| 997 |
+
"grad_norm": 0.003087802790105343,
|
| 998 |
+
"learning_rate": 1e-05,
|
| 999 |
+
"loss": 0.0002,
|
| 1000 |
+
"step": 1390
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 2.48,
|
| 1004 |
+
"grad_norm": 0.002106766914948821,
|
| 1005 |
+
"learning_rate": 9.672131147540984e-06,
|
| 1006 |
+
"loss": 0.0008,
|
| 1007 |
+
"step": 1400
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 2.49,
|
| 1011 |
+
"grad_norm": 0.058892298489809036,
|
| 1012 |
+
"learning_rate": 9.344262295081968e-06,
|
| 1013 |
+
"loss": 0.0003,
|
| 1014 |
+
"step": 1410
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 2.51,
|
| 1018 |
+
"grad_norm": 0.0024353417102247477,
|
| 1019 |
+
"learning_rate": 9.016393442622952e-06,
|
| 1020 |
+
"loss": 0.0002,
|
| 1021 |
+
"step": 1420
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 2.53,
|
| 1025 |
+
"grad_norm": 0.003295810893177986,
|
| 1026 |
+
"learning_rate": 8.688524590163935e-06,
|
| 1027 |
+
"loss": 0.0028,
|
| 1028 |
+
"step": 1430
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 2.55,
|
| 1032 |
+
"grad_norm": 0.006564658600836992,
|
| 1033 |
+
"learning_rate": 8.360655737704919e-06,
|
| 1034 |
+
"loss": 0.0001,
|
| 1035 |
+
"step": 1440
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 2.56,
|
| 1039 |
+
"grad_norm": 0.17080430686473846,
|
| 1040 |
+
"learning_rate": 8.032786885245902e-06,
|
| 1041 |
+
"loss": 0.0005,
|
| 1042 |
+
"step": 1450
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 2.58,
|
| 1046 |
+
"grad_norm": 0.012207315303385258,
|
| 1047 |
+
"learning_rate": 7.704918032786886e-06,
|
| 1048 |
+
"loss": 0.0008,
|
| 1049 |
+
"step": 1460
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 2.6,
|
| 1053 |
+
"grad_norm": 0.07983198761940002,
|
| 1054 |
+
"learning_rate": 7.3770491803278695e-06,
|
| 1055 |
+
"loss": 0.0003,
|
| 1056 |
+
"step": 1470
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 2.62,
|
| 1060 |
+
"grad_norm": 0.0021029170602560043,
|
| 1061 |
+
"learning_rate": 7.049180327868852e-06,
|
| 1062 |
+
"loss": 0.0002,
|
| 1063 |
+
"step": 1480
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"epoch": 2.63,
|
| 1067 |
+
"grad_norm": 0.001665195683017373,
|
| 1068 |
+
"learning_rate": 6.721311475409836e-06,
|
| 1069 |
+
"loss": 0.0002,
|
| 1070 |
+
"step": 1490
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 2.65,
|
| 1074 |
+
"grad_norm": 0.0021821956615895033,
|
| 1075 |
+
"learning_rate": 6.393442622950819e-06,
|
| 1076 |
+
"loss": 0.0003,
|
| 1077 |
+
"step": 1500
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 2.67,
|
| 1081 |
+
"grad_norm": 0.0023076001089066267,
|
| 1082 |
+
"learning_rate": 6.065573770491804e-06,
|
| 1083 |
+
"loss": 0.0007,
|
| 1084 |
+
"step": 1510
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 2.69,
|
| 1088 |
+
"grad_norm": 0.002858547493815422,
|
| 1089 |
+
"learning_rate": 5.737704918032787e-06,
|
| 1090 |
+
"loss": 0.0002,
|
| 1091 |
+
"step": 1520
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 2.71,
|
| 1095 |
+
"grad_norm": 0.014774510636925697,
|
| 1096 |
+
"learning_rate": 5.409836065573771e-06,
|
| 1097 |
+
"loss": 0.0002,
|
| 1098 |
+
"step": 1530
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 2.72,
|
| 1102 |
+
"grad_norm": 0.0014713788405060768,
|
| 1103 |
+
"learning_rate": 5.0819672131147545e-06,
|
| 1104 |
+
"loss": 0.0001,
|
| 1105 |
+
"step": 1540
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 2.74,
|
| 1109 |
+
"grad_norm": 0.0027496495749801397,
|
| 1110 |
+
"learning_rate": 4.754098360655738e-06,
|
| 1111 |
+
"loss": 0.0002,
|
| 1112 |
+
"step": 1550
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 2.76,
|
| 1116 |
+
"grad_norm": 0.002773697255179286,
|
| 1117 |
+
"learning_rate": 4.426229508196722e-06,
|
| 1118 |
+
"loss": 0.0002,
|
| 1119 |
+
"step": 1560
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 2.78,
|
| 1123 |
+
"grad_norm": 0.002084126928821206,
|
| 1124 |
+
"learning_rate": 4.098360655737704e-06,
|
| 1125 |
+
"loss": 0.0001,
|
| 1126 |
+
"step": 1570
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 2.79,
|
| 1130 |
+
"grad_norm": 0.0016491630813106894,
|
| 1131 |
+
"learning_rate": 3.770491803278689e-06,
|
| 1132 |
+
"loss": 0.0002,
|
| 1133 |
+
"step": 1580
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 2.81,
|
| 1137 |
+
"grad_norm": 0.003069406608119607,
|
| 1138 |
+
"learning_rate": 3.4426229508196724e-06,
|
| 1139 |
+
"loss": 0.0074,
|
| 1140 |
+
"step": 1590
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 2.83,
|
| 1144 |
+
"grad_norm": 0.001330082886852324,
|
| 1145 |
+
"learning_rate": 3.114754098360656e-06,
|
| 1146 |
+
"loss": 0.0003,
|
| 1147 |
+
"step": 1600
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 2.85,
|
| 1151 |
+
"grad_norm": 0.9783486127853394,
|
| 1152 |
+
"learning_rate": 2.7868852459016396e-06,
|
| 1153 |
+
"loss": 0.0007,
|
| 1154 |
+
"step": 1610
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 2.86,
|
| 1158 |
+
"grad_norm": 0.001621302799321711,
|
| 1159 |
+
"learning_rate": 2.459016393442623e-06,
|
| 1160 |
+
"loss": 0.0001,
|
| 1161 |
+
"step": 1620
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 2.88,
|
| 1165 |
+
"grad_norm": 0.05432431399822235,
|
| 1166 |
+
"learning_rate": 2.1311475409836067e-06,
|
| 1167 |
+
"loss": 0.0002,
|
| 1168 |
+
"step": 1630
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 2.9,
|
| 1172 |
+
"grad_norm": 0.001987756695598364,
|
| 1173 |
+
"learning_rate": 1.8032786885245903e-06,
|
| 1174 |
+
"loss": 0.0003,
|
| 1175 |
+
"step": 1640
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 2.92,
|
| 1179 |
+
"grad_norm": 0.0034532160498201847,
|
| 1180 |
+
"learning_rate": 1.4754098360655739e-06,
|
| 1181 |
+
"loss": 0.0001,
|
| 1182 |
+
"step": 1650
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 2.94,
|
| 1186 |
+
"grad_norm": 0.0017541700508445501,
|
| 1187 |
+
"learning_rate": 1.1475409836065575e-06,
|
| 1188 |
+
"loss": 0.0112,
|
| 1189 |
+
"step": 1660
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 2.95,
|
| 1193 |
+
"grad_norm": 0.02740754559636116,
|
| 1194 |
+
"learning_rate": 8.19672131147541e-07,
|
| 1195 |
+
"loss": 0.0002,
|
| 1196 |
+
"step": 1670
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 2.97,
|
| 1200 |
+
"grad_norm": 0.0015147700905799866,
|
| 1201 |
+
"learning_rate": 4.918032786885246e-07,
|
| 1202 |
+
"loss": 0.0001,
|
| 1203 |
+
"step": 1680
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 2.99,
|
| 1207 |
+
"grad_norm": 2.0458450317382812,
|
| 1208 |
+
"learning_rate": 1.639344262295082e-07,
|
| 1209 |
+
"loss": 0.0014,
|
| 1210 |
+
"step": 1690
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 3.0,
|
| 1214 |
+
"eval_accuracy": 0.9942273534635879,
|
| 1215 |
+
"eval_loss": 0.02003500424325466,
|
| 1216 |
+
"eval_runtime": 123.0739,
|
| 1217 |
+
"eval_samples_per_second": 18.298,
|
| 1218 |
+
"eval_steps_per_second": 2.291,
|
| 1219 |
+
"step": 1695
|
| 1220 |
+
},
|
| 1221 |
+
{
|
| 1222 |
+
"epoch": 3.0,
|
| 1223 |
+
"step": 1695,
|
| 1224 |
+
"total_flos": 1.4134330246261506e+19,
|
| 1225 |
+
"train_loss": 0.10345972847079468,
|
| 1226 |
+
"train_runtime": 4872.3509,
|
| 1227 |
+
"train_samples_per_second": 11.138,
|
| 1228 |
+
"train_steps_per_second": 0.348
|
| 1229 |
+
}
|
| 1230 |
+
],
|
| 1231 |
+
"logging_steps": 10,
|
| 1232 |
+
"max_steps": 1695,
|
| 1233 |
+
"num_input_tokens_seen": 0,
|
| 1234 |
+
"num_train_epochs": 3,
|
| 1235 |
+
"save_steps": 500,
|
| 1236 |
+
"total_flos": 1.4134330246261506e+19,
|
| 1237 |
+
"train_batch_size": 8,
|
| 1238 |
+
"trial_name": null,
|
| 1239 |
+
"trial_params": null
|
| 1240 |
+
}
|