Upload trainer_state.json
Browse files- trainer_state.json +1232 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1232 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.5798903703689575,
|
| 3 |
+
"best_model_checkpoint": "models/checkpoints/checkpoint-16000",
|
| 4 |
+
"epoch": 14.47952951820855,
|
| 5 |
+
"global_step": 16000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.09,
|
| 12 |
+
"learning_rate": 1.9879336349924587e-05,
|
| 13 |
+
"loss": 1.3189,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.18,
|
| 18 |
+
"learning_rate": 1.9758672699849173e-05,
|
| 19 |
+
"loss": 1.1007,
|
| 20 |
+
"step": 200
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.27,
|
| 24 |
+
"learning_rate": 1.9638009049773755e-05,
|
| 25 |
+
"loss": 1.0933,
|
| 26 |
+
"step": 300
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.36,
|
| 30 |
+
"learning_rate": 1.951734539969834e-05,
|
| 31 |
+
"loss": 1.0187,
|
| 32 |
+
"step": 400
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.45,
|
| 36 |
+
"learning_rate": 1.9396681749622927e-05,
|
| 37 |
+
"loss": 1.018,
|
| 38 |
+
"step": 500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.45,
|
| 42 |
+
"eval_loss": 0.9501454830169678,
|
| 43 |
+
"eval_runtime": 52.6762,
|
| 44 |
+
"eval_samples_per_second": 149.194,
|
| 45 |
+
"eval_steps_per_second": 9.34,
|
| 46 |
+
"step": 500
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.54,
|
| 50 |
+
"learning_rate": 1.9276018099547512e-05,
|
| 51 |
+
"loss": 1.0293,
|
| 52 |
+
"step": 600
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.63,
|
| 56 |
+
"learning_rate": 1.9155354449472098e-05,
|
| 57 |
+
"loss": 0.9875,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.72,
|
| 62 |
+
"learning_rate": 1.9034690799396684e-05,
|
| 63 |
+
"loss": 0.9752,
|
| 64 |
+
"step": 800
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 0.81,
|
| 68 |
+
"learning_rate": 1.891402714932127e-05,
|
| 69 |
+
"loss": 0.979,
|
| 70 |
+
"step": 900
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 0.9,
|
| 74 |
+
"learning_rate": 1.8793363499245855e-05,
|
| 75 |
+
"loss": 0.9513,
|
| 76 |
+
"step": 1000
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.9,
|
| 80 |
+
"eval_loss": 0.8483895063400269,
|
| 81 |
+
"eval_runtime": 32.7078,
|
| 82 |
+
"eval_samples_per_second": 240.279,
|
| 83 |
+
"eval_steps_per_second": 15.042,
|
| 84 |
+
"step": 1000
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.0,
|
| 88 |
+
"learning_rate": 1.867269984917044e-05,
|
| 89 |
+
"loss": 0.9691,
|
| 90 |
+
"step": 1100
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 1.09,
|
| 94 |
+
"learning_rate": 1.8552036199095026e-05,
|
| 95 |
+
"loss": 0.8995,
|
| 96 |
+
"step": 1200
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 1.18,
|
| 100 |
+
"learning_rate": 1.843137254901961e-05,
|
| 101 |
+
"loss": 0.9392,
|
| 102 |
+
"step": 1300
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.27,
|
| 106 |
+
"learning_rate": 1.8310708898944194e-05,
|
| 107 |
+
"loss": 0.9156,
|
| 108 |
+
"step": 1400
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 1.36,
|
| 112 |
+
"learning_rate": 1.819004524886878e-05,
|
| 113 |
+
"loss": 0.9142,
|
| 114 |
+
"step": 1500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 1.36,
|
| 118 |
+
"eval_loss": 0.8270823359489441,
|
| 119 |
+
"eval_runtime": 30.0805,
|
| 120 |
+
"eval_samples_per_second": 261.266,
|
| 121 |
+
"eval_steps_per_second": 16.356,
|
| 122 |
+
"step": 1500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 1.45,
|
| 126 |
+
"learning_rate": 1.8069381598793365e-05,
|
| 127 |
+
"loss": 0.8579,
|
| 128 |
+
"step": 1600
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 1.54,
|
| 132 |
+
"learning_rate": 1.794871794871795e-05,
|
| 133 |
+
"loss": 0.8649,
|
| 134 |
+
"step": 1700
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 1.63,
|
| 138 |
+
"learning_rate": 1.7828054298642537e-05,
|
| 139 |
+
"loss": 0.8935,
|
| 140 |
+
"step": 1800
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.72,
|
| 144 |
+
"learning_rate": 1.770739064856712e-05,
|
| 145 |
+
"loss": 0.8534,
|
| 146 |
+
"step": 1900
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 1.81,
|
| 150 |
+
"learning_rate": 1.7586726998491705e-05,
|
| 151 |
+
"loss": 0.8427,
|
| 152 |
+
"step": 2000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 1.81,
|
| 156 |
+
"eval_loss": 0.8167164921760559,
|
| 157 |
+
"eval_runtime": 29.67,
|
| 158 |
+
"eval_samples_per_second": 264.88,
|
| 159 |
+
"eval_steps_per_second": 16.582,
|
| 160 |
+
"step": 2000
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 1.9,
|
| 164 |
+
"learning_rate": 1.746606334841629e-05,
|
| 165 |
+
"loss": 0.8634,
|
| 166 |
+
"step": 2100
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 1.99,
|
| 170 |
+
"learning_rate": 1.7345399698340876e-05,
|
| 171 |
+
"loss": 0.8735,
|
| 172 |
+
"step": 2200
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.08,
|
| 176 |
+
"learning_rate": 1.7224736048265462e-05,
|
| 177 |
+
"loss": 0.8401,
|
| 178 |
+
"step": 2300
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 2.17,
|
| 182 |
+
"learning_rate": 1.7104072398190047e-05,
|
| 183 |
+
"loss": 0.8403,
|
| 184 |
+
"step": 2400
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 2.26,
|
| 188 |
+
"learning_rate": 1.6983408748114633e-05,
|
| 189 |
+
"loss": 0.8212,
|
| 190 |
+
"step": 2500
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 2.26,
|
| 194 |
+
"eval_loss": 0.8272273540496826,
|
| 195 |
+
"eval_runtime": 30.3829,
|
| 196 |
+
"eval_samples_per_second": 258.665,
|
| 197 |
+
"eval_steps_per_second": 16.193,
|
| 198 |
+
"step": 2500
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 2.35,
|
| 202 |
+
"learning_rate": 1.686274509803922e-05,
|
| 203 |
+
"loss": 0.8301,
|
| 204 |
+
"step": 2600
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 2.44,
|
| 208 |
+
"learning_rate": 1.6742081447963804e-05,
|
| 209 |
+
"loss": 0.8396,
|
| 210 |
+
"step": 2700
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 2.53,
|
| 214 |
+
"learning_rate": 1.662141779788839e-05,
|
| 215 |
+
"loss": 0.8067,
|
| 216 |
+
"step": 2800
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 2.62,
|
| 220 |
+
"learning_rate": 1.6500754147812972e-05,
|
| 221 |
+
"loss": 0.8231,
|
| 222 |
+
"step": 2900
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 2.71,
|
| 226 |
+
"learning_rate": 1.6380090497737558e-05,
|
| 227 |
+
"loss": 0.8125,
|
| 228 |
+
"step": 3000
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 2.71,
|
| 232 |
+
"eval_loss": 0.7901880741119385,
|
| 233 |
+
"eval_runtime": 30.2297,
|
| 234 |
+
"eval_samples_per_second": 259.976,
|
| 235 |
+
"eval_steps_per_second": 16.275,
|
| 236 |
+
"step": 3000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 2.81,
|
| 240 |
+
"learning_rate": 1.6259426847662144e-05,
|
| 241 |
+
"loss": 0.808,
|
| 242 |
+
"step": 3100
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 2.9,
|
| 246 |
+
"learning_rate": 1.613876319758673e-05,
|
| 247 |
+
"loss": 0.8359,
|
| 248 |
+
"step": 3200
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 2.99,
|
| 252 |
+
"learning_rate": 1.6018099547511315e-05,
|
| 253 |
+
"loss": 0.8017,
|
| 254 |
+
"step": 3300
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 3.08,
|
| 258 |
+
"learning_rate": 1.5897435897435897e-05,
|
| 259 |
+
"loss": 0.8226,
|
| 260 |
+
"step": 3400
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 3.17,
|
| 264 |
+
"learning_rate": 1.5776772247360483e-05,
|
| 265 |
+
"loss": 0.8121,
|
| 266 |
+
"step": 3500
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 3.17,
|
| 270 |
+
"eval_loss": 0.7621824741363525,
|
| 271 |
+
"eval_runtime": 30.3432,
|
| 272 |
+
"eval_samples_per_second": 259.004,
|
| 273 |
+
"eval_steps_per_second": 16.215,
|
| 274 |
+
"step": 3500
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 3.26,
|
| 278 |
+
"learning_rate": 1.565610859728507e-05,
|
| 279 |
+
"loss": 0.7861,
|
| 280 |
+
"step": 3600
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 3.35,
|
| 284 |
+
"learning_rate": 1.5535444947209654e-05,
|
| 285 |
+
"loss": 0.7634,
|
| 286 |
+
"step": 3700
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 3.44,
|
| 290 |
+
"learning_rate": 1.541478129713424e-05,
|
| 291 |
+
"loss": 0.7941,
|
| 292 |
+
"step": 3800
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 3.53,
|
| 296 |
+
"learning_rate": 1.5294117647058822e-05,
|
| 297 |
+
"loss": 0.7628,
|
| 298 |
+
"step": 3900
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 3.62,
|
| 302 |
+
"learning_rate": 1.517345399698341e-05,
|
| 303 |
+
"loss": 0.7572,
|
| 304 |
+
"step": 4000
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 3.62,
|
| 308 |
+
"eval_loss": 0.7411866188049316,
|
| 309 |
+
"eval_runtime": 29.8313,
|
| 310 |
+
"eval_samples_per_second": 263.448,
|
| 311 |
+
"eval_steps_per_second": 16.493,
|
| 312 |
+
"step": 4000
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 3.71,
|
| 316 |
+
"learning_rate": 1.5052790346907995e-05,
|
| 317 |
+
"loss": 0.7975,
|
| 318 |
+
"step": 4100
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 3.8,
|
| 322 |
+
"learning_rate": 1.4932126696832581e-05,
|
| 323 |
+
"loss": 0.7626,
|
| 324 |
+
"step": 4200
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 3.89,
|
| 328 |
+
"learning_rate": 1.4811463046757167e-05,
|
| 329 |
+
"loss": 0.7467,
|
| 330 |
+
"step": 4300
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 3.98,
|
| 334 |
+
"learning_rate": 1.4690799396681752e-05,
|
| 335 |
+
"loss": 0.7689,
|
| 336 |
+
"step": 4400
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 4.07,
|
| 340 |
+
"learning_rate": 1.4570135746606336e-05,
|
| 341 |
+
"loss": 0.7392,
|
| 342 |
+
"step": 4500
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 4.07,
|
| 346 |
+
"eval_loss": 0.7343409061431885,
|
| 347 |
+
"eval_runtime": 30.138,
|
| 348 |
+
"eval_samples_per_second": 260.767,
|
| 349 |
+
"eval_steps_per_second": 16.325,
|
| 350 |
+
"step": 4500
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 4.16,
|
| 354 |
+
"learning_rate": 1.444947209653092e-05,
|
| 355 |
+
"loss": 0.7636,
|
| 356 |
+
"step": 4600
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 4.25,
|
| 360 |
+
"learning_rate": 1.4328808446455506e-05,
|
| 361 |
+
"loss": 0.768,
|
| 362 |
+
"step": 4700
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 4.34,
|
| 366 |
+
"learning_rate": 1.4208144796380091e-05,
|
| 367 |
+
"loss": 0.7234,
|
| 368 |
+
"step": 4800
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 4.43,
|
| 372 |
+
"learning_rate": 1.4087481146304677e-05,
|
| 373 |
+
"loss": 0.7868,
|
| 374 |
+
"step": 4900
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 4.52,
|
| 378 |
+
"learning_rate": 1.3966817496229261e-05,
|
| 379 |
+
"loss": 0.7538,
|
| 380 |
+
"step": 5000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 4.52,
|
| 384 |
+
"eval_loss": 0.7230245471000671,
|
| 385 |
+
"eval_runtime": 29.8163,
|
| 386 |
+
"eval_samples_per_second": 263.581,
|
| 387 |
+
"eval_steps_per_second": 16.501,
|
| 388 |
+
"step": 5000
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 4.62,
|
| 392 |
+
"learning_rate": 1.3846153846153847e-05,
|
| 393 |
+
"loss": 0.7261,
|
| 394 |
+
"step": 5100
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 4.71,
|
| 398 |
+
"learning_rate": 1.3725490196078432e-05,
|
| 399 |
+
"loss": 0.7462,
|
| 400 |
+
"step": 5200
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 4.8,
|
| 404 |
+
"learning_rate": 1.3604826546003018e-05,
|
| 405 |
+
"loss": 0.7169,
|
| 406 |
+
"step": 5300
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 4.89,
|
| 410 |
+
"learning_rate": 1.3484162895927604e-05,
|
| 411 |
+
"loss": 0.7542,
|
| 412 |
+
"step": 5400
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 4.98,
|
| 416 |
+
"learning_rate": 1.3363499245852188e-05,
|
| 417 |
+
"loss": 0.71,
|
| 418 |
+
"step": 5500
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 4.98,
|
| 422 |
+
"eval_loss": 0.7268965840339661,
|
| 423 |
+
"eval_runtime": 29.6638,
|
| 424 |
+
"eval_samples_per_second": 264.936,
|
| 425 |
+
"eval_steps_per_second": 16.586,
|
| 426 |
+
"step": 5500
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 5.07,
|
| 430 |
+
"learning_rate": 1.3242835595776773e-05,
|
| 431 |
+
"loss": 0.7217,
|
| 432 |
+
"step": 5600
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 5.16,
|
| 436 |
+
"learning_rate": 1.3122171945701359e-05,
|
| 437 |
+
"loss": 0.7226,
|
| 438 |
+
"step": 5700
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 5.25,
|
| 442 |
+
"learning_rate": 1.3001508295625945e-05,
|
| 443 |
+
"loss": 0.7177,
|
| 444 |
+
"step": 5800
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 5.34,
|
| 448 |
+
"learning_rate": 1.288084464555053e-05,
|
| 449 |
+
"loss": 0.7038,
|
| 450 |
+
"step": 5900
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 5.43,
|
| 454 |
+
"learning_rate": 1.2760180995475113e-05,
|
| 455 |
+
"loss": 0.6966,
|
| 456 |
+
"step": 6000
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 5.43,
|
| 460 |
+
"eval_loss": 0.7119535803794861,
|
| 461 |
+
"eval_runtime": 29.7811,
|
| 462 |
+
"eval_samples_per_second": 263.892,
|
| 463 |
+
"eval_steps_per_second": 16.521,
|
| 464 |
+
"step": 6000
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 5.52,
|
| 468 |
+
"learning_rate": 1.2639517345399698e-05,
|
| 469 |
+
"loss": 0.7167,
|
| 470 |
+
"step": 6100
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 5.61,
|
| 474 |
+
"learning_rate": 1.2518853695324284e-05,
|
| 475 |
+
"loss": 0.6907,
|
| 476 |
+
"step": 6200
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 5.7,
|
| 480 |
+
"learning_rate": 1.239819004524887e-05,
|
| 481 |
+
"loss": 0.6992,
|
| 482 |
+
"step": 6300
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 5.79,
|
| 486 |
+
"learning_rate": 1.2277526395173455e-05,
|
| 487 |
+
"loss": 0.7038,
|
| 488 |
+
"step": 6400
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 5.88,
|
| 492 |
+
"learning_rate": 1.215686274509804e-05,
|
| 493 |
+
"loss": 0.6995,
|
| 494 |
+
"step": 6500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 5.88,
|
| 498 |
+
"eval_loss": 0.7289978861808777,
|
| 499 |
+
"eval_runtime": 30.2132,
|
| 500 |
+
"eval_samples_per_second": 260.118,
|
| 501 |
+
"eval_steps_per_second": 16.284,
|
| 502 |
+
"step": 6500
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 5.97,
|
| 506 |
+
"learning_rate": 1.2036199095022625e-05,
|
| 507 |
+
"loss": 0.6841,
|
| 508 |
+
"step": 6600
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 6.06,
|
| 512 |
+
"learning_rate": 1.191553544494721e-05,
|
| 513 |
+
"loss": 0.6827,
|
| 514 |
+
"step": 6700
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 6.15,
|
| 518 |
+
"learning_rate": 1.1794871794871796e-05,
|
| 519 |
+
"loss": 0.6664,
|
| 520 |
+
"step": 6800
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 6.24,
|
| 524 |
+
"learning_rate": 1.1674208144796382e-05,
|
| 525 |
+
"loss": 0.6911,
|
| 526 |
+
"step": 6900
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 6.33,
|
| 530 |
+
"learning_rate": 1.1553544494720966e-05,
|
| 531 |
+
"loss": 0.6719,
|
| 532 |
+
"step": 7000
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 6.33,
|
| 536 |
+
"eval_loss": 0.7228043675422668,
|
| 537 |
+
"eval_runtime": 31.0054,
|
| 538 |
+
"eval_samples_per_second": 253.472,
|
| 539 |
+
"eval_steps_per_second": 15.868,
|
| 540 |
+
"step": 7000
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 6.43,
|
| 544 |
+
"learning_rate": 1.1432880844645552e-05,
|
| 545 |
+
"loss": 0.6785,
|
| 546 |
+
"step": 7100
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 6.52,
|
| 550 |
+
"learning_rate": 1.1312217194570137e-05,
|
| 551 |
+
"loss": 0.6908,
|
| 552 |
+
"step": 7200
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 6.61,
|
| 556 |
+
"learning_rate": 1.1191553544494723e-05,
|
| 557 |
+
"loss": 0.6775,
|
| 558 |
+
"step": 7300
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 6.7,
|
| 562 |
+
"learning_rate": 1.1070889894419309e-05,
|
| 563 |
+
"loss": 0.6566,
|
| 564 |
+
"step": 7400
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 6.79,
|
| 568 |
+
"learning_rate": 1.0950226244343893e-05,
|
| 569 |
+
"loss": 0.6862,
|
| 570 |
+
"step": 7500
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 6.79,
|
| 574 |
+
"eval_loss": 0.6712033152580261,
|
| 575 |
+
"eval_runtime": 29.8128,
|
| 576 |
+
"eval_samples_per_second": 263.612,
|
| 577 |
+
"eval_steps_per_second": 16.503,
|
| 578 |
+
"step": 7500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 6.88,
|
| 582 |
+
"learning_rate": 1.0829562594268476e-05,
|
| 583 |
+
"loss": 0.6965,
|
| 584 |
+
"step": 7600
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 6.97,
|
| 588 |
+
"learning_rate": 1.0708898944193062e-05,
|
| 589 |
+
"loss": 0.6836,
|
| 590 |
+
"step": 7700
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 7.06,
|
| 594 |
+
"learning_rate": 1.0588235294117648e-05,
|
| 595 |
+
"loss": 0.6684,
|
| 596 |
+
"step": 7800
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 7.15,
|
| 600 |
+
"learning_rate": 1.0467571644042233e-05,
|
| 601 |
+
"loss": 0.6767,
|
| 602 |
+
"step": 7900
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 7.24,
|
| 606 |
+
"learning_rate": 1.0346907993966819e-05,
|
| 607 |
+
"loss": 0.6775,
|
| 608 |
+
"step": 8000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 7.24,
|
| 612 |
+
"eval_loss": 0.6799022555351257,
|
| 613 |
+
"eval_runtime": 30.3013,
|
| 614 |
+
"eval_samples_per_second": 259.362,
|
| 615 |
+
"eval_steps_per_second": 16.237,
|
| 616 |
+
"step": 8000
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 7.33,
|
| 620 |
+
"learning_rate": 1.0226244343891403e-05,
|
| 621 |
+
"loss": 0.6371,
|
| 622 |
+
"step": 8100
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 7.42,
|
| 626 |
+
"learning_rate": 1.0105580693815989e-05,
|
| 627 |
+
"loss": 0.6623,
|
| 628 |
+
"step": 8200
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 7.51,
|
| 632 |
+
"learning_rate": 9.984917043740574e-06,
|
| 633 |
+
"loss": 0.6499,
|
| 634 |
+
"step": 8300
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 7.6,
|
| 638 |
+
"learning_rate": 9.86425339366516e-06,
|
| 639 |
+
"loss": 0.6457,
|
| 640 |
+
"step": 8400
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 7.69,
|
| 644 |
+
"learning_rate": 9.743589743589744e-06,
|
| 645 |
+
"loss": 0.6283,
|
| 646 |
+
"step": 8500
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 7.69,
|
| 650 |
+
"eval_loss": 0.6853426098823547,
|
| 651 |
+
"eval_runtime": 29.9239,
|
| 652 |
+
"eval_samples_per_second": 262.633,
|
| 653 |
+
"eval_steps_per_second": 16.442,
|
| 654 |
+
"step": 8500
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 7.78,
|
| 658 |
+
"learning_rate": 9.62292609351433e-06,
|
| 659 |
+
"loss": 0.6151,
|
| 660 |
+
"step": 8600
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 7.87,
|
| 664 |
+
"learning_rate": 9.502262443438914e-06,
|
| 665 |
+
"loss": 0.6431,
|
| 666 |
+
"step": 8700
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 7.96,
|
| 670 |
+
"learning_rate": 9.3815987933635e-06,
|
| 671 |
+
"loss": 0.6293,
|
| 672 |
+
"step": 8800
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 8.05,
|
| 676 |
+
"learning_rate": 9.260935143288085e-06,
|
| 677 |
+
"loss": 0.6266,
|
| 678 |
+
"step": 8900
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 8.14,
|
| 682 |
+
"learning_rate": 9.14027149321267e-06,
|
| 683 |
+
"loss": 0.6028,
|
| 684 |
+
"step": 9000
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 8.14,
|
| 688 |
+
"eval_loss": 0.6904362440109253,
|
| 689 |
+
"eval_runtime": 31.3612,
|
| 690 |
+
"eval_samples_per_second": 250.596,
|
| 691 |
+
"eval_steps_per_second": 15.688,
|
| 692 |
+
"step": 9000
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 8.24,
|
| 696 |
+
"learning_rate": 9.019607843137256e-06,
|
| 697 |
+
"loss": 0.6494,
|
| 698 |
+
"step": 9100
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 8.33,
|
| 702 |
+
"learning_rate": 8.89894419306184e-06,
|
| 703 |
+
"loss": 0.6014,
|
| 704 |
+
"step": 9200
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 8.42,
|
| 708 |
+
"learning_rate": 8.778280542986426e-06,
|
| 709 |
+
"loss": 0.6208,
|
| 710 |
+
"step": 9300
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 8.51,
|
| 714 |
+
"learning_rate": 8.657616892911012e-06,
|
| 715 |
+
"loss": 0.6461,
|
| 716 |
+
"step": 9400
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 8.6,
|
| 720 |
+
"learning_rate": 8.536953242835596e-06,
|
| 721 |
+
"loss": 0.6291,
|
| 722 |
+
"step": 9500
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 8.6,
|
| 726 |
+
"eval_loss": 0.6408899426460266,
|
| 727 |
+
"eval_runtime": 31.2717,
|
| 728 |
+
"eval_samples_per_second": 251.314,
|
| 729 |
+
"eval_steps_per_second": 15.733,
|
| 730 |
+
"step": 9500
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 8.69,
|
| 734 |
+
"learning_rate": 8.416289592760181e-06,
|
| 735 |
+
"loss": 0.6056,
|
| 736 |
+
"step": 9600
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 8.78,
|
| 740 |
+
"learning_rate": 8.295625942684767e-06,
|
| 741 |
+
"loss": 0.6052,
|
| 742 |
+
"step": 9700
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 8.87,
|
| 746 |
+
"learning_rate": 8.174962292609353e-06,
|
| 747 |
+
"loss": 0.6128,
|
| 748 |
+
"step": 9800
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 8.96,
|
| 752 |
+
"learning_rate": 8.054298642533938e-06,
|
| 753 |
+
"loss": 0.5895,
|
| 754 |
+
"step": 9900
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 9.05,
|
| 758 |
+
"learning_rate": 7.933634992458522e-06,
|
| 759 |
+
"loss": 0.6304,
|
| 760 |
+
"step": 10000
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 9.05,
|
| 764 |
+
"eval_loss": 0.6174328327178955,
|
| 765 |
+
"eval_runtime": 30.7682,
|
| 766 |
+
"eval_samples_per_second": 255.426,
|
| 767 |
+
"eval_steps_per_second": 15.991,
|
| 768 |
+
"step": 10000
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 9.14,
|
| 772 |
+
"learning_rate": 7.812971342383108e-06,
|
| 773 |
+
"loss": 0.5882,
|
| 774 |
+
"step": 10100
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 9.23,
|
| 778 |
+
"learning_rate": 7.692307692307694e-06,
|
| 779 |
+
"loss": 0.6304,
|
| 780 |
+
"step": 10200
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 9.32,
|
| 784 |
+
"learning_rate": 7.5716440422322776e-06,
|
| 785 |
+
"loss": 0.6103,
|
| 786 |
+
"step": 10300
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 9.41,
|
| 790 |
+
"learning_rate": 7.450980392156863e-06,
|
| 791 |
+
"loss": 0.5952,
|
| 792 |
+
"step": 10400
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 9.5,
|
| 796 |
+
"learning_rate": 7.330316742081448e-06,
|
| 797 |
+
"loss": 0.6039,
|
| 798 |
+
"step": 10500
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 9.5,
|
| 802 |
+
"eval_loss": 0.6438130140304565,
|
| 803 |
+
"eval_runtime": 31.9863,
|
| 804 |
+
"eval_samples_per_second": 245.699,
|
| 805 |
+
"eval_steps_per_second": 15.382,
|
| 806 |
+
"step": 10500
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 9.59,
|
| 810 |
+
"learning_rate": 7.209653092006034e-06,
|
| 811 |
+
"loss": 0.6079,
|
| 812 |
+
"step": 10600
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 9.68,
|
| 816 |
+
"learning_rate": 7.088989441930619e-06,
|
| 817 |
+
"loss": 0.6093,
|
| 818 |
+
"step": 10700
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 9.77,
|
| 822 |
+
"learning_rate": 6.968325791855204e-06,
|
| 823 |
+
"loss": 0.5787,
|
| 824 |
+
"step": 10800
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 9.86,
|
| 828 |
+
"learning_rate": 6.84766214177979e-06,
|
| 829 |
+
"loss": 0.6138,
|
| 830 |
+
"step": 10900
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 9.95,
|
| 834 |
+
"learning_rate": 6.7269984917043755e-06,
|
| 835 |
+
"loss": 0.6011,
|
| 836 |
+
"step": 11000
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 9.95,
|
| 840 |
+
"eval_loss": 0.6626730561256409,
|
| 841 |
+
"eval_runtime": 30.1966,
|
| 842 |
+
"eval_samples_per_second": 260.261,
|
| 843 |
+
"eval_steps_per_second": 16.293,
|
| 844 |
+
"step": 11000
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 10.05,
|
| 848 |
+
"learning_rate": 6.6063348416289595e-06,
|
| 849 |
+
"loss": 0.6087,
|
| 850 |
+
"step": 11100
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 10.14,
|
| 854 |
+
"learning_rate": 6.485671191553545e-06,
|
| 855 |
+
"loss": 0.5929,
|
| 856 |
+
"step": 11200
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 10.23,
|
| 860 |
+
"learning_rate": 6.36500754147813e-06,
|
| 861 |
+
"loss": 0.5849,
|
| 862 |
+
"step": 11300
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 10.32,
|
| 866 |
+
"learning_rate": 6.244343891402716e-06,
|
| 867 |
+
"loss": 0.6042,
|
| 868 |
+
"step": 11400
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 10.41,
|
| 872 |
+
"learning_rate": 6.123680241327301e-06,
|
| 873 |
+
"loss": 0.637,
|
| 874 |
+
"step": 11500
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 10.41,
|
| 878 |
+
"eval_loss": 0.6357372403144836,
|
| 879 |
+
"eval_runtime": 29.8866,
|
| 880 |
+
"eval_samples_per_second": 262.961,
|
| 881 |
+
"eval_steps_per_second": 16.462,
|
| 882 |
+
"step": 11500
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 10.5,
|
| 886 |
+
"learning_rate": 6.003016591251885e-06,
|
| 887 |
+
"loss": 0.5783,
|
| 888 |
+
"step": 11600
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 10.59,
|
| 892 |
+
"learning_rate": 5.882352941176471e-06,
|
| 893 |
+
"loss": 0.5739,
|
| 894 |
+
"step": 11700
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 10.68,
|
| 898 |
+
"learning_rate": 5.761689291101056e-06,
|
| 899 |
+
"loss": 0.5617,
|
| 900 |
+
"step": 11800
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 10.77,
|
| 904 |
+
"learning_rate": 5.641025641025641e-06,
|
| 905 |
+
"loss": 0.5724,
|
| 906 |
+
"step": 11900
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 10.86,
|
| 910 |
+
"learning_rate": 5.520361990950227e-06,
|
| 911 |
+
"loss": 0.5635,
|
| 912 |
+
"step": 12000
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 10.86,
|
| 916 |
+
"eval_loss": 0.6293562054634094,
|
| 917 |
+
"eval_runtime": 42.9677,
|
| 918 |
+
"eval_samples_per_second": 182.905,
|
| 919 |
+
"eval_steps_per_second": 11.45,
|
| 920 |
+
"step": 12000
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 10.95,
|
| 924 |
+
"learning_rate": 5.399698340874812e-06,
|
| 925 |
+
"loss": 0.6074,
|
| 926 |
+
"step": 12100
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 11.04,
|
| 930 |
+
"learning_rate": 5.2790346907993975e-06,
|
| 931 |
+
"loss": 0.5755,
|
| 932 |
+
"step": 12200
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 11.13,
|
| 936 |
+
"learning_rate": 5.158371040723983e-06,
|
| 937 |
+
"loss": 0.5653,
|
| 938 |
+
"step": 12300
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 11.22,
|
| 942 |
+
"learning_rate": 5.037707390648567e-06,
|
| 943 |
+
"loss": 0.5685,
|
| 944 |
+
"step": 12400
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 11.31,
|
| 948 |
+
"learning_rate": 4.917043740573153e-06,
|
| 949 |
+
"loss": 0.596,
|
| 950 |
+
"step": 12500
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 11.31,
|
| 954 |
+
"eval_loss": 0.619467556476593,
|
| 955 |
+
"eval_runtime": 30.8886,
|
| 956 |
+
"eval_samples_per_second": 254.431,
|
| 957 |
+
"eval_steps_per_second": 15.928,
|
| 958 |
+
"step": 12500
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 11.4,
|
| 962 |
+
"learning_rate": 4.7963800904977385e-06,
|
| 963 |
+
"loss": 0.5765,
|
| 964 |
+
"step": 12600
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 11.49,
|
| 968 |
+
"learning_rate": 4.675716440422323e-06,
|
| 969 |
+
"loss": 0.563,
|
| 970 |
+
"step": 12700
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 11.58,
|
| 974 |
+
"learning_rate": 4.555052790346908e-06,
|
| 975 |
+
"loss": 0.5871,
|
| 976 |
+
"step": 12800
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 11.67,
|
| 980 |
+
"learning_rate": 4.434389140271493e-06,
|
| 981 |
+
"loss": 0.5345,
|
| 982 |
+
"step": 12900
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 11.76,
|
| 986 |
+
"learning_rate": 4.313725490196079e-06,
|
| 987 |
+
"loss": 0.5647,
|
| 988 |
+
"step": 13000
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 11.76,
|
| 992 |
+
"eval_loss": 0.6173378825187683,
|
| 993 |
+
"eval_runtime": 30.5561,
|
| 994 |
+
"eval_samples_per_second": 257.199,
|
| 995 |
+
"eval_steps_per_second": 16.102,
|
| 996 |
+
"step": 13000
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 11.86,
|
| 1000 |
+
"learning_rate": 4.193061840120664e-06,
|
| 1001 |
+
"loss": 0.5798,
|
| 1002 |
+
"step": 13100
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 11.95,
|
| 1006 |
+
"learning_rate": 4.072398190045249e-06,
|
| 1007 |
+
"loss": 0.5553,
|
| 1008 |
+
"step": 13200
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 12.04,
|
| 1012 |
+
"learning_rate": 3.951734539969834e-06,
|
| 1013 |
+
"loss": 0.5372,
|
| 1014 |
+
"step": 13300
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 12.13,
|
| 1018 |
+
"learning_rate": 3.83107088989442e-06,
|
| 1019 |
+
"loss": 0.5432,
|
| 1020 |
+
"step": 13400
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 12.22,
|
| 1024 |
+
"learning_rate": 3.710407239819005e-06,
|
| 1025 |
+
"loss": 0.5304,
|
| 1026 |
+
"step": 13500
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 12.22,
|
| 1030 |
+
"eval_loss": 0.6232503056526184,
|
| 1031 |
+
"eval_runtime": 30.2832,
|
| 1032 |
+
"eval_samples_per_second": 259.517,
|
| 1033 |
+
"eval_steps_per_second": 16.247,
|
| 1034 |
+
"step": 13500
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 12.31,
|
| 1038 |
+
"learning_rate": 3.58974358974359e-06,
|
| 1039 |
+
"loss": 0.5423,
|
| 1040 |
+
"step": 13600
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 12.4,
|
| 1044 |
+
"learning_rate": 3.4690799396681753e-06,
|
| 1045 |
+
"loss": 0.546,
|
| 1046 |
+
"step": 13700
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 12.49,
|
| 1050 |
+
"learning_rate": 3.34841628959276e-06,
|
| 1051 |
+
"loss": 0.5932,
|
| 1052 |
+
"step": 13800
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 12.58,
|
| 1056 |
+
"learning_rate": 3.2277526395173458e-06,
|
| 1057 |
+
"loss": 0.5721,
|
| 1058 |
+
"step": 13900
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 12.67,
|
| 1062 |
+
"learning_rate": 3.107088989441931e-06,
|
| 1063 |
+
"loss": 0.5815,
|
| 1064 |
+
"step": 14000
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 12.67,
|
| 1068 |
+
"eval_loss": 0.6265811324119568,
|
| 1069 |
+
"eval_runtime": 31.274,
|
| 1070 |
+
"eval_samples_per_second": 251.295,
|
| 1071 |
+
"eval_steps_per_second": 15.732,
|
| 1072 |
+
"step": 14000
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 12.76,
|
| 1076 |
+
"learning_rate": 2.986425339366516e-06,
|
| 1077 |
+
"loss": 0.5646,
|
| 1078 |
+
"step": 14100
|
| 1079 |
+
},
|
| 1080 |
+
{
|
| 1081 |
+
"epoch": 12.85,
|
| 1082 |
+
"learning_rate": 2.865761689291101e-06,
|
| 1083 |
+
"loss": 0.5706,
|
| 1084 |
+
"step": 14200
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 12.94,
|
| 1088 |
+
"learning_rate": 2.7450980392156867e-06,
|
| 1089 |
+
"loss": 0.5561,
|
| 1090 |
+
"step": 14300
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 13.03,
|
| 1094 |
+
"learning_rate": 2.624434389140272e-06,
|
| 1095 |
+
"loss": 0.5668,
|
| 1096 |
+
"step": 14400
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 13.12,
|
| 1100 |
+
"learning_rate": 2.503770739064857e-06,
|
| 1101 |
+
"loss": 0.5555,
|
| 1102 |
+
"step": 14500
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 13.12,
|
| 1106 |
+
"eval_loss": 0.6143834590911865,
|
| 1107 |
+
"eval_runtime": 30.6444,
|
| 1108 |
+
"eval_samples_per_second": 256.458,
|
| 1109 |
+
"eval_steps_per_second": 16.055,
|
| 1110 |
+
"step": 14500
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 13.21,
|
| 1114 |
+
"learning_rate": 2.3831070889894425e-06,
|
| 1115 |
+
"loss": 0.5327,
|
| 1116 |
+
"step": 14600
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 13.3,
|
| 1120 |
+
"learning_rate": 2.2624434389140273e-06,
|
| 1121 |
+
"loss": 0.5491,
|
| 1122 |
+
"step": 14700
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 13.39,
|
| 1126 |
+
"learning_rate": 2.1417797888386125e-06,
|
| 1127 |
+
"loss": 0.5633,
|
| 1128 |
+
"step": 14800
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 13.48,
|
| 1132 |
+
"learning_rate": 2.0211161387631978e-06,
|
| 1133 |
+
"loss": 0.5612,
|
| 1134 |
+
"step": 14900
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 13.57,
|
| 1138 |
+
"learning_rate": 1.9004524886877828e-06,
|
| 1139 |
+
"loss": 0.5725,
|
| 1140 |
+
"step": 15000
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 13.57,
|
| 1144 |
+
"eval_loss": 0.6175289750099182,
|
| 1145 |
+
"eval_runtime": 30.1245,
|
| 1146 |
+
"eval_samples_per_second": 260.884,
|
| 1147 |
+
"eval_steps_per_second": 16.332,
|
| 1148 |
+
"step": 15000
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 13.67,
|
| 1152 |
+
"learning_rate": 1.7797888386123682e-06,
|
| 1153 |
+
"loss": 0.5502,
|
| 1154 |
+
"step": 15100
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 13.76,
|
| 1158 |
+
"learning_rate": 1.6591251885369533e-06,
|
| 1159 |
+
"loss": 0.5529,
|
| 1160 |
+
"step": 15200
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 13.85,
|
| 1164 |
+
"learning_rate": 1.5384615384615387e-06,
|
| 1165 |
+
"loss": 0.5351,
|
| 1166 |
+
"step": 15300
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 13.94,
|
| 1170 |
+
"learning_rate": 1.4177978883861237e-06,
|
| 1171 |
+
"loss": 0.5563,
|
| 1172 |
+
"step": 15400
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 14.03,
|
| 1176 |
+
"learning_rate": 1.2971342383107092e-06,
|
| 1177 |
+
"loss": 0.5616,
|
| 1178 |
+
"step": 15500
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 14.03,
|
| 1182 |
+
"eval_loss": 0.6149749755859375,
|
| 1183 |
+
"eval_runtime": 31.0977,
|
| 1184 |
+
"eval_samples_per_second": 252.719,
|
| 1185 |
+
"eval_steps_per_second": 15.821,
|
| 1186 |
+
"step": 15500
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 14.12,
|
| 1190 |
+
"learning_rate": 1.1764705882352942e-06,
|
| 1191 |
+
"loss": 0.5404,
|
| 1192 |
+
"step": 15600
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 14.21,
|
| 1196 |
+
"learning_rate": 1.0558069381598795e-06,
|
| 1197 |
+
"loss": 0.5461,
|
| 1198 |
+
"step": 15700
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"epoch": 14.3,
|
| 1202 |
+
"learning_rate": 9.351432880844646e-07,
|
| 1203 |
+
"loss": 0.5449,
|
| 1204 |
+
"step": 15800
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"epoch": 14.39,
|
| 1208 |
+
"learning_rate": 8.144796380090498e-07,
|
| 1209 |
+
"loss": 0.5296,
|
| 1210 |
+
"step": 15900
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 14.48,
|
| 1214 |
+
"learning_rate": 6.938159879336351e-07,
|
| 1215 |
+
"loss": 0.5471,
|
| 1216 |
+
"step": 16000
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 14.48,
|
| 1220 |
+
"eval_loss": 0.5798903703689575,
|
| 1221 |
+
"eval_runtime": 29.2689,
|
| 1222 |
+
"eval_samples_per_second": 268.51,
|
| 1223 |
+
"eval_steps_per_second": 16.81,
|
| 1224 |
+
"step": 16000
|
| 1225 |
+
}
|
| 1226 |
+
],
|
| 1227 |
+
"max_steps": 16575,
|
| 1228 |
+
"num_train_epochs": 15,
|
| 1229 |
+
"total_flos": 5.59327922919324e+16,
|
| 1230 |
+
"trial_name": null,
|
| 1231 |
+
"trial_params": null
|
| 1232 |
+
}
|