Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1267 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1267 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 14.723716381418093,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 765,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.097799511002445,
|
| 13 |
+
"grad_norm": 2.395451784133911,
|
| 14 |
+
"learning_rate": 1.9997891995035914e-05,
|
| 15 |
+
"loss": 0.8739,
|
| 16 |
+
"num_input_tokens_seen": 163840,
|
| 17 |
+
"step": 5
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.19559902200489,
|
| 21 |
+
"grad_norm": 1.3395614624023438,
|
| 22 |
+
"learning_rate": 1.999156886888064e-05,
|
| 23 |
+
"loss": 0.7957,
|
| 24 |
+
"num_input_tokens_seen": 327680,
|
| 25 |
+
"step": 10
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 0.293398533007335,
|
| 29 |
+
"grad_norm": 1.125549554824829,
|
| 30 |
+
"learning_rate": 1.9981033287370443e-05,
|
| 31 |
+
"loss": 0.7822,
|
| 32 |
+
"num_input_tokens_seen": 491520,
|
| 33 |
+
"step": 15
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.39119804400978,
|
| 37 |
+
"grad_norm": 1.1936776638031006,
|
| 38 |
+
"learning_rate": 1.9966289692316944e-05,
|
| 39 |
+
"loss": 0.7815,
|
| 40 |
+
"num_input_tokens_seen": 655360,
|
| 41 |
+
"step": 20
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.4889975550122249,
|
| 45 |
+
"grad_norm": 1.230684757232666,
|
| 46 |
+
"learning_rate": 1.9947344299634464e-05,
|
| 47 |
+
"loss": 0.7432,
|
| 48 |
+
"num_input_tokens_seen": 819200,
|
| 49 |
+
"step": 25
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.58679706601467,
|
| 53 |
+
"grad_norm": 1.0630358457565308,
|
| 54 |
+
"learning_rate": 1.992420509671936e-05,
|
| 55 |
+
"loss": 0.7389,
|
| 56 |
+
"num_input_tokens_seen": 982368,
|
| 57 |
+
"step": 30
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.684596577017115,
|
| 61 |
+
"grad_norm": 1.1900779008865356,
|
| 62 |
+
"learning_rate": 1.9896881839082554e-05,
|
| 63 |
+
"loss": 0.7654,
|
| 64 |
+
"num_input_tokens_seen": 1146208,
|
| 65 |
+
"step": 35
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.78239608801956,
|
| 69 |
+
"grad_norm": 1.0942089557647705,
|
| 70 |
+
"learning_rate": 1.9865386046236597e-05,
|
| 71 |
+
"loss": 0.7753,
|
| 72 |
+
"num_input_tokens_seen": 1310048,
|
| 73 |
+
"step": 40
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.8801955990220048,
|
| 77 |
+
"grad_norm": 1.0910882949829102,
|
| 78 |
+
"learning_rate": 1.982973099683902e-05,
|
| 79 |
+
"loss": 0.7256,
|
| 80 |
+
"num_input_tokens_seen": 1473888,
|
| 81 |
+
"step": 45
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.9779951100244498,
|
| 85 |
+
"grad_norm": 1.0902478694915771,
|
| 86 |
+
"learning_rate": 1.9789931723094046e-05,
|
| 87 |
+
"loss": 0.727,
|
| 88 |
+
"num_input_tokens_seen": 1637728,
|
| 89 |
+
"step": 50
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 1.058679706601467,
|
| 93 |
+
"grad_norm": 1.3777414560317993,
|
| 94 |
+
"learning_rate": 1.9746005004415004e-05,
|
| 95 |
+
"loss": 0.582,
|
| 96 |
+
"num_input_tokens_seen": 1770848,
|
| 97 |
+
"step": 55
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 1.156479217603912,
|
| 101 |
+
"grad_norm": 1.3062232732772827,
|
| 102 |
+
"learning_rate": 1.9697969360350098e-05,
|
| 103 |
+
"loss": 0.5628,
|
| 104 |
+
"num_input_tokens_seen": 1934688,
|
| 105 |
+
"step": 60
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.254278728606357,
|
| 109 |
+
"grad_norm": 1.239915132522583,
|
| 110 |
+
"learning_rate": 1.9645845042774555e-05,
|
| 111 |
+
"loss": 0.5561,
|
| 112 |
+
"num_input_tokens_seen": 2098528,
|
| 113 |
+
"step": 65
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 1.352078239608802,
|
| 117 |
+
"grad_norm": 1.1612083911895752,
|
| 118 |
+
"learning_rate": 1.9589654027352412e-05,
|
| 119 |
+
"loss": 0.5778,
|
| 120 |
+
"num_input_tokens_seen": 2262368,
|
| 121 |
+
"step": 70
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 1.449877750611247,
|
| 125 |
+
"grad_norm": 1.2168060541152954,
|
| 126 |
+
"learning_rate": 1.9529420004271568e-05,
|
| 127 |
+
"loss": 0.5207,
|
| 128 |
+
"num_input_tokens_seen": 2426128,
|
| 129 |
+
"step": 75
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.5476772616136918,
|
| 133 |
+
"grad_norm": 1.2806190252304077,
|
| 134 |
+
"learning_rate": 1.9465168368255946e-05,
|
| 135 |
+
"loss": 0.5452,
|
| 136 |
+
"num_input_tokens_seen": 2589968,
|
| 137 |
+
"step": 80
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.6454767726161368,
|
| 141 |
+
"grad_norm": 1.17153000831604,
|
| 142 |
+
"learning_rate": 1.9396926207859085e-05,
|
| 143 |
+
"loss": 0.5379,
|
| 144 |
+
"num_input_tokens_seen": 2753808,
|
| 145 |
+
"step": 85
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 1.7432762836185818,
|
| 149 |
+
"grad_norm": 1.1548662185668945,
|
| 150 |
+
"learning_rate": 1.932472229404356e-05,
|
| 151 |
+
"loss": 0.5116,
|
| 152 |
+
"num_input_tokens_seen": 2917648,
|
| 153 |
+
"step": 90
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 1.8410757946210268,
|
| 157 |
+
"grad_norm": 1.3172167539596558,
|
| 158 |
+
"learning_rate": 1.924858706805112e-05,
|
| 159 |
+
"loss": 0.5532,
|
| 160 |
+
"num_input_tokens_seen": 3081488,
|
| 161 |
+
"step": 95
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.9388753056234718,
|
| 165 |
+
"grad_norm": 1.1930724382400513,
|
| 166 |
+
"learning_rate": 1.9168552628568632e-05,
|
| 167 |
+
"loss": 0.5659,
|
| 168 |
+
"num_input_tokens_seen": 3245328,
|
| 169 |
+
"step": 100
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 2.019559902200489,
|
| 173 |
+
"grad_norm": 1.7079554796218872,
|
| 174 |
+
"learning_rate": 1.9084652718195237e-05,
|
| 175 |
+
"loss": 0.6028,
|
| 176 |
+
"num_input_tokens_seen": 3378448,
|
| 177 |
+
"step": 105
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 2.117359413202934,
|
| 181 |
+
"grad_norm": 1.7393510341644287,
|
| 182 |
+
"learning_rate": 1.8996922709216456e-05,
|
| 183 |
+
"loss": 0.4228,
|
| 184 |
+
"num_input_tokens_seen": 3542288,
|
| 185 |
+
"step": 110
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 2.215158924205379,
|
| 189 |
+
"grad_norm": 1.2822431325912476,
|
| 190 |
+
"learning_rate": 1.8905399588691165e-05,
|
| 191 |
+
"loss": 0.3648,
|
| 192 |
+
"num_input_tokens_seen": 3706128,
|
| 193 |
+
"step": 115
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 2.312958435207824,
|
| 197 |
+
"grad_norm": 1.3198189735412598,
|
| 198 |
+
"learning_rate": 1.8810121942857848e-05,
|
| 199 |
+
"loss": 0.3822,
|
| 200 |
+
"num_input_tokens_seen": 3869968,
|
| 201 |
+
"step": 120
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 2.410757946210269,
|
| 205 |
+
"grad_norm": 1.2831072807312012,
|
| 206 |
+
"learning_rate": 1.8711129940866577e-05,
|
| 207 |
+
"loss": 0.3907,
|
| 208 |
+
"num_input_tokens_seen": 4033808,
|
| 209 |
+
"step": 125
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 2.508557457212714,
|
| 213 |
+
"grad_norm": 1.4272454977035522,
|
| 214 |
+
"learning_rate": 1.860846531784368e-05,
|
| 215 |
+
"loss": 0.3936,
|
| 216 |
+
"num_input_tokens_seen": 4197648,
|
| 217 |
+
"step": 130
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 2.606356968215159,
|
| 221 |
+
"grad_norm": 1.3846728801727295,
|
| 222 |
+
"learning_rate": 1.8502171357296144e-05,
|
| 223 |
+
"loss": 0.376,
|
| 224 |
+
"num_input_tokens_seen": 4361488,
|
| 225 |
+
"step": 135
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 2.704156479217604,
|
| 229 |
+
"grad_norm": 1.1824907064437866,
|
| 230 |
+
"learning_rate": 1.839229287286327e-05,
|
| 231 |
+
"loss": 0.3735,
|
| 232 |
+
"num_input_tokens_seen": 4525328,
|
| 233 |
+
"step": 140
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 2.801955990220049,
|
| 237 |
+
"grad_norm": 1.2617613077163696,
|
| 238 |
+
"learning_rate": 1.827887618942318e-05,
|
| 239 |
+
"loss": 0.4114,
|
| 240 |
+
"num_input_tokens_seen": 4689168,
|
| 241 |
+
"step": 145
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 2.899755501222494,
|
| 245 |
+
"grad_norm": 1.1593624353408813,
|
| 246 |
+
"learning_rate": 1.816196912356222e-05,
|
| 247 |
+
"loss": 0.3696,
|
| 248 |
+
"num_input_tokens_seen": 4853008,
|
| 249 |
+
"step": 150
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 2.997555012224939,
|
| 253 |
+
"grad_norm": 1.1450086832046509,
|
| 254 |
+
"learning_rate": 1.8041620963415418e-05,
|
| 255 |
+
"loss": 0.3888,
|
| 256 |
+
"num_input_tokens_seen": 5016848,
|
| 257 |
+
"step": 155
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 3.078239608801956,
|
| 261 |
+
"grad_norm": 1.4760794639587402,
|
| 262 |
+
"learning_rate": 1.7917882447886585e-05,
|
| 263 |
+
"loss": 0.2605,
|
| 264 |
+
"num_input_tokens_seen": 5149968,
|
| 265 |
+
"step": 160
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 3.176039119804401,
|
| 269 |
+
"grad_norm": 1.520880103111267,
|
| 270 |
+
"learning_rate": 1.7790805745256703e-05,
|
| 271 |
+
"loss": 0.2475,
|
| 272 |
+
"num_input_tokens_seen": 5313808,
|
| 273 |
+
"step": 165
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 3.273838630806846,
|
| 277 |
+
"grad_norm": 1.4768394231796265,
|
| 278 |
+
"learning_rate": 1.766044443118978e-05,
|
| 279 |
+
"loss": 0.2344,
|
| 280 |
+
"num_input_tokens_seen": 5477648,
|
| 281 |
+
"step": 170
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 3.371638141809291,
|
| 285 |
+
"grad_norm": 1.4958367347717285,
|
| 286 |
+
"learning_rate": 1.7526853466145248e-05,
|
| 287 |
+
"loss": 0.2665,
|
| 288 |
+
"num_input_tokens_seen": 5641488,
|
| 289 |
+
"step": 175
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 3.469437652811736,
|
| 293 |
+
"grad_norm": 1.4445388317108154,
|
| 294 |
+
"learning_rate": 1.7390089172206594e-05,
|
| 295 |
+
"loss": 0.2477,
|
| 296 |
+
"num_input_tokens_seen": 5805328,
|
| 297 |
+
"step": 180
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 3.567237163814181,
|
| 301 |
+
"grad_norm": 1.4173372983932495,
|
| 302 |
+
"learning_rate": 1.725020920933593e-05,
|
| 303 |
+
"loss": 0.2679,
|
| 304 |
+
"num_input_tokens_seen": 5969168,
|
| 305 |
+
"step": 185
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 3.665036674816626,
|
| 309 |
+
"grad_norm": 1.6111114025115967,
|
| 310 |
+
"learning_rate": 1.710727255106447e-05,
|
| 311 |
+
"loss": 0.234,
|
| 312 |
+
"num_input_tokens_seen": 6133008,
|
| 313 |
+
"step": 190
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 3.762836185819071,
|
| 317 |
+
"grad_norm": 1.4121896028518677,
|
| 318 |
+
"learning_rate": 1.696133945962927e-05,
|
| 319 |
+
"loss": 0.2587,
|
| 320 |
+
"num_input_tokens_seen": 6296848,
|
| 321 |
+
"step": 195
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 3.860635696821516,
|
| 325 |
+
"grad_norm": 1.3143867254257202,
|
| 326 |
+
"learning_rate": 1.681247146056654e-05,
|
| 327 |
+
"loss": 0.2606,
|
| 328 |
+
"num_input_tokens_seen": 6460688,
|
| 329 |
+
"step": 200
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 3.958435207823961,
|
| 333 |
+
"grad_norm": 1.2710639238357544,
|
| 334 |
+
"learning_rate": 1.6660731316772503e-05,
|
| 335 |
+
"loss": 0.2543,
|
| 336 |
+
"num_input_tokens_seen": 6624528,
|
| 337 |
+
"step": 205
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 4.039119804400978,
|
| 341 |
+
"grad_norm": 1.284611463546753,
|
| 342 |
+
"learning_rate": 1.650618300204242e-05,
|
| 343 |
+
"loss": 0.1856,
|
| 344 |
+
"num_input_tokens_seen": 6757128,
|
| 345 |
+
"step": 210
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 4.136919315403423,
|
| 349 |
+
"grad_norm": 1.6113184690475464,
|
| 350 |
+
"learning_rate": 1.634889167409923e-05,
|
| 351 |
+
"loss": 0.1462,
|
| 352 |
+
"num_input_tokens_seen": 6920968,
|
| 353 |
+
"step": 215
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 4.234718826405868,
|
| 357 |
+
"grad_norm": 1.230480432510376,
|
| 358 |
+
"learning_rate": 1.6188923647122946e-05,
|
| 359 |
+
"loss": 0.17,
|
| 360 |
+
"num_input_tokens_seen": 7084808,
|
| 361 |
+
"step": 220
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 4.332518337408313,
|
| 365 |
+
"grad_norm": 1.388247013092041,
|
| 366 |
+
"learning_rate": 1.6026346363792565e-05,
|
| 367 |
+
"loss": 0.1427,
|
| 368 |
+
"num_input_tokens_seen": 7248648,
|
| 369 |
+
"step": 225
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 4.430317848410758,
|
| 373 |
+
"grad_norm": 1.3794785737991333,
|
| 374 |
+
"learning_rate": 1.5861228366852148e-05,
|
| 375 |
+
"loss": 0.15,
|
| 376 |
+
"num_input_tokens_seen": 7412488,
|
| 377 |
+
"step": 230
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 4.528117359413203,
|
| 381 |
+
"grad_norm": 1.5547810792922974,
|
| 382 |
+
"learning_rate": 1.5693639270213138e-05,
|
| 383 |
+
"loss": 0.1603,
|
| 384 |
+
"num_input_tokens_seen": 7576264,
|
| 385 |
+
"step": 235
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 4.625916870415648,
|
| 389 |
+
"grad_norm": 1.49036705493927,
|
| 390 |
+
"learning_rate": 1.552364972960506e-05,
|
| 391 |
+
"loss": 0.1536,
|
| 392 |
+
"num_input_tokens_seen": 7740104,
|
| 393 |
+
"step": 240
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 4.723716381418093,
|
| 397 |
+
"grad_norm": 1.2922449111938477,
|
| 398 |
+
"learning_rate": 1.5351331412787004e-05,
|
| 399 |
+
"loss": 0.1415,
|
| 400 |
+
"num_input_tokens_seen": 7903944,
|
| 401 |
+
"step": 245
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 4.821515892420538,
|
| 405 |
+
"grad_norm": 1.5458725690841675,
|
| 406 |
+
"learning_rate": 1.5176756969332428e-05,
|
| 407 |
+
"loss": 0.159,
|
| 408 |
+
"num_input_tokens_seen": 8067784,
|
| 409 |
+
"step": 250
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 4.919315403422983,
|
| 413 |
+
"grad_norm": 1.3060288429260254,
|
| 414 |
+
"learning_rate": 1.5000000000000002e-05,
|
| 415 |
+
"loss": 0.1699,
|
| 416 |
+
"num_input_tokens_seen": 8231624,
|
| 417 |
+
"step": 255
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 5.0,
|
| 421 |
+
"grad_norm": 4.3814568519592285,
|
| 422 |
+
"learning_rate": 1.4821135025703491e-05,
|
| 423 |
+
"loss": 0.1295,
|
| 424 |
+
"num_input_tokens_seen": 8364744,
|
| 425 |
+
"step": 260
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 5.097799511002445,
|
| 429 |
+
"grad_norm": 1.496863842010498,
|
| 430 |
+
"learning_rate": 1.4640237456093636e-05,
|
| 431 |
+
"loss": 0.1019,
|
| 432 |
+
"num_input_tokens_seen": 8528584,
|
| 433 |
+
"step": 265
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 5.19559902200489,
|
| 437 |
+
"grad_norm": 1.3016966581344604,
|
| 438 |
+
"learning_rate": 1.4457383557765385e-05,
|
| 439 |
+
"loss": 0.0916,
|
| 440 |
+
"num_input_tokens_seen": 8692424,
|
| 441 |
+
"step": 270
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 5.293398533007335,
|
| 445 |
+
"grad_norm": 1.3272321224212646,
|
| 446 |
+
"learning_rate": 1.427265042210381e-05,
|
| 447 |
+
"loss": 0.0935,
|
| 448 |
+
"num_input_tokens_seen": 8856264,
|
| 449 |
+
"step": 275
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 5.39119804400978,
|
| 453 |
+
"grad_norm": 1.2260195016860962,
|
| 454 |
+
"learning_rate": 1.4086115932782316e-05,
|
| 455 |
+
"loss": 0.0679,
|
| 456 |
+
"num_input_tokens_seen": 9020104,
|
| 457 |
+
"step": 280
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 5.488997555012225,
|
| 461 |
+
"grad_norm": 1.1941609382629395,
|
| 462 |
+
"learning_rate": 1.3897858732926794e-05,
|
| 463 |
+
"loss": 0.101,
|
| 464 |
+
"num_input_tokens_seen": 9183944,
|
| 465 |
+
"step": 285
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 5.58679706601467,
|
| 469 |
+
"grad_norm": 1.3348464965820312,
|
| 470 |
+
"learning_rate": 1.3707958191959609e-05,
|
| 471 |
+
"loss": 0.0802,
|
| 472 |
+
"num_input_tokens_seen": 9347784,
|
| 473 |
+
"step": 290
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 5.684596577017115,
|
| 477 |
+
"grad_norm": 1.3610597848892212,
|
| 478 |
+
"learning_rate": 1.3516494372137368e-05,
|
| 479 |
+
"loss": 0.0812,
|
| 480 |
+
"num_input_tokens_seen": 9511624,
|
| 481 |
+
"step": 295
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 5.78239608801956,
|
| 485 |
+
"grad_norm": 1.1276705265045166,
|
| 486 |
+
"learning_rate": 1.3323547994796597e-05,
|
| 487 |
+
"loss": 0.0812,
|
| 488 |
+
"num_input_tokens_seen": 9675464,
|
| 489 |
+
"step": 300
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 5.880195599022005,
|
| 493 |
+
"grad_norm": 1.4052499532699585,
|
| 494 |
+
"learning_rate": 1.3129200406321545e-05,
|
| 495 |
+
"loss": 0.0884,
|
| 496 |
+
"num_input_tokens_seen": 9839304,
|
| 497 |
+
"step": 305
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 5.97799511002445,
|
| 501 |
+
"grad_norm": 1.2434982061386108,
|
| 502 |
+
"learning_rate": 1.2933533543848462e-05,
|
| 503 |
+
"loss": 0.0906,
|
| 504 |
+
"num_input_tokens_seen": 10003144,
|
| 505 |
+
"step": 310
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 6.058679706601467,
|
| 509 |
+
"grad_norm": 1.19767427444458,
|
| 510 |
+
"learning_rate": 1.2736629900720832e-05,
|
| 511 |
+
"loss": 0.0616,
|
| 512 |
+
"num_input_tokens_seen": 10136264,
|
| 513 |
+
"step": 315
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 6.156479217603912,
|
| 517 |
+
"grad_norm": 1.2783230543136597,
|
| 518 |
+
"learning_rate": 1.2538572491710079e-05,
|
| 519 |
+
"loss": 0.0447,
|
| 520 |
+
"num_input_tokens_seen": 10300104,
|
| 521 |
+
"step": 320
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 6.254278728606357,
|
| 525 |
+
"grad_norm": 1.1590359210968018,
|
| 526 |
+
"learning_rate": 1.2339444818016488e-05,
|
| 527 |
+
"loss": 0.0528,
|
| 528 |
+
"num_input_tokens_seen": 10463944,
|
| 529 |
+
"step": 325
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 6.352078239608802,
|
| 533 |
+
"grad_norm": 1.2891935110092163,
|
| 534 |
+
"learning_rate": 1.2139330832064975e-05,
|
| 535 |
+
"loss": 0.0429,
|
| 536 |
+
"num_input_tokens_seen": 10627784,
|
| 537 |
+
"step": 330
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 6.449877750611247,
|
| 541 |
+
"grad_norm": 1.0882583856582642,
|
| 542 |
+
"learning_rate": 1.1938314902110701e-05,
|
| 543 |
+
"loss": 0.0442,
|
| 544 |
+
"num_input_tokens_seen": 10791624,
|
| 545 |
+
"step": 335
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 6.547677261613692,
|
| 549 |
+
"grad_norm": 1.1069742441177368,
|
| 550 |
+
"learning_rate": 1.1736481776669307e-05,
|
| 551 |
+
"loss": 0.047,
|
| 552 |
+
"num_input_tokens_seen": 10955464,
|
| 553 |
+
"step": 340
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 6.645476772616137,
|
| 557 |
+
"grad_norm": 1.0986578464508057,
|
| 558 |
+
"learning_rate": 1.1533916548786856e-05,
|
| 559 |
+
"loss": 0.0437,
|
| 560 |
+
"num_input_tokens_seen": 11119304,
|
| 561 |
+
"step": 345
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 6.743276283618582,
|
| 565 |
+
"grad_norm": 1.1824274063110352,
|
| 566 |
+
"learning_rate": 1.133070462016454e-05,
|
| 567 |
+
"loss": 0.0466,
|
| 568 |
+
"num_input_tokens_seen": 11282568,
|
| 569 |
+
"step": 350
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 6.841075794621027,
|
| 573 |
+
"grad_norm": 1.197988510131836,
|
| 574 |
+
"learning_rate": 1.1126931665153213e-05,
|
| 575 |
+
"loss": 0.0498,
|
| 576 |
+
"num_input_tokens_seen": 11446408,
|
| 577 |
+
"step": 355
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 6.938875305623472,
|
| 581 |
+
"grad_norm": 1.1721147298812866,
|
| 582 |
+
"learning_rate": 1.092268359463302e-05,
|
| 583 |
+
"loss": 0.0409,
|
| 584 |
+
"num_input_tokens_seen": 11610248,
|
| 585 |
+
"step": 360
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 7.019559902200489,
|
| 589 |
+
"grad_norm": 0.9280807971954346,
|
| 590 |
+
"learning_rate": 1.0718046519793276e-05,
|
| 591 |
+
"loss": 0.0444,
|
| 592 |
+
"num_input_tokens_seen": 11743368,
|
| 593 |
+
"step": 365
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 7.117359413202934,
|
| 597 |
+
"grad_norm": 0.8121551871299744,
|
| 598 |
+
"learning_rate": 1.0513106715827897e-05,
|
| 599 |
+
"loss": 0.0212,
|
| 600 |
+
"num_input_tokens_seen": 11906632,
|
| 601 |
+
"step": 370
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 7.215158924205379,
|
| 605 |
+
"grad_norm": 0.974238932132721,
|
| 606 |
+
"learning_rate": 1.0307950585561705e-05,
|
| 607 |
+
"loss": 0.0275,
|
| 608 |
+
"num_input_tokens_seen": 12070472,
|
| 609 |
+
"step": 375
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 7.312958435207824,
|
| 613 |
+
"grad_norm": 0.8143863081932068,
|
| 614 |
+
"learning_rate": 1.01026646230229e-05,
|
| 615 |
+
"loss": 0.0234,
|
| 616 |
+
"num_input_tokens_seen": 12234312,
|
| 617 |
+
"step": 380
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 7.410757946210269,
|
| 621 |
+
"grad_norm": 0.9007371664047241,
|
| 622 |
+
"learning_rate": 9.897335376977104e-06,
|
| 623 |
+
"loss": 0.0227,
|
| 624 |
+
"num_input_tokens_seen": 12398152,
|
| 625 |
+
"step": 385
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 7.508557457212714,
|
| 629 |
+
"grad_norm": 0.8347233533859253,
|
| 630 |
+
"learning_rate": 9.692049414438298e-06,
|
| 631 |
+
"loss": 0.0264,
|
| 632 |
+
"num_input_tokens_seen": 12561992,
|
| 633 |
+
"step": 390
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 7.606356968215159,
|
| 637 |
+
"grad_norm": 1.0170607566833496,
|
| 638 |
+
"learning_rate": 9.486893284172103e-06,
|
| 639 |
+
"loss": 0.0251,
|
| 640 |
+
"num_input_tokens_seen": 12725832,
|
| 641 |
+
"step": 395
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 7.704156479217604,
|
| 645 |
+
"grad_norm": 0.8994267582893372,
|
| 646 |
+
"learning_rate": 9.281953480206725e-06,
|
| 647 |
+
"loss": 0.0237,
|
| 648 |
+
"num_input_tokens_seen": 12889672,
|
| 649 |
+
"step": 400
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 7.801955990220049,
|
| 653 |
+
"grad_norm": 1.0272514820098877,
|
| 654 |
+
"learning_rate": 9.07731640536698e-06,
|
| 655 |
+
"loss": 0.0228,
|
| 656 |
+
"num_input_tokens_seen": 13053512,
|
| 657 |
+
"step": 405
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 7.899755501222494,
|
| 661 |
+
"grad_norm": 0.8432409167289734,
|
| 662 |
+
"learning_rate": 8.87306833484679e-06,
|
| 663 |
+
"loss": 0.0238,
|
| 664 |
+
"num_input_tokens_seen": 13217352,
|
| 665 |
+
"step": 410
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 7.997555012224939,
|
| 669 |
+
"grad_norm": 0.9295158386230469,
|
| 670 |
+
"learning_rate": 8.669295379835467e-06,
|
| 671 |
+
"loss": 0.0222,
|
| 672 |
+
"num_input_tokens_seen": 13381192,
|
| 673 |
+
"step": 415
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 8.078239608801956,
|
| 677 |
+
"grad_norm": 0.6758769750595093,
|
| 678 |
+
"learning_rate": 8.466083451213145e-06,
|
| 679 |
+
"loss": 0.0109,
|
| 680 |
+
"num_input_tokens_seen": 13514312,
|
| 681 |
+
"step": 420
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 8.1760391198044,
|
| 685 |
+
"grad_norm": 0.5933005213737488,
|
| 686 |
+
"learning_rate": 8.263518223330698e-06,
|
| 687 |
+
"loss": 0.0118,
|
| 688 |
+
"num_input_tokens_seen": 13677448,
|
| 689 |
+
"step": 425
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 8.273838630806846,
|
| 693 |
+
"grad_norm": 0.7031515836715698,
|
| 694 |
+
"learning_rate": 8.0616850978893e-06,
|
| 695 |
+
"loss": 0.01,
|
| 696 |
+
"num_input_tokens_seen": 13841288,
|
| 697 |
+
"step": 430
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 8.37163814180929,
|
| 701 |
+
"grad_norm": 0.8007884621620178,
|
| 702 |
+
"learning_rate": 7.860669167935028e-06,
|
| 703 |
+
"loss": 0.0112,
|
| 704 |
+
"num_input_tokens_seen": 14005128,
|
| 705 |
+
"step": 435
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 8.469437652811736,
|
| 709 |
+
"grad_norm": 0.6427900195121765,
|
| 710 |
+
"learning_rate": 7.660555181983517e-06,
|
| 711 |
+
"loss": 0.0133,
|
| 712 |
+
"num_input_tokens_seen": 14168968,
|
| 713 |
+
"step": 440
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 8.56723716381418,
|
| 717 |
+
"grad_norm": 0.6638085246086121,
|
| 718 |
+
"learning_rate": 7.461427508289922e-06,
|
| 719 |
+
"loss": 0.0113,
|
| 720 |
+
"num_input_tokens_seen": 14332808,
|
| 721 |
+
"step": 445
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 8.665036674816626,
|
| 725 |
+
"grad_norm": 0.6730697751045227,
|
| 726 |
+
"learning_rate": 7.263370099279173e-06,
|
| 727 |
+
"loss": 0.0102,
|
| 728 |
+
"num_input_tokens_seen": 14496648,
|
| 729 |
+
"step": 450
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 8.76283618581907,
|
| 733 |
+
"grad_norm": 0.7171938419342041,
|
| 734 |
+
"learning_rate": 7.066466456151541e-06,
|
| 735 |
+
"loss": 0.01,
|
| 736 |
+
"num_input_tokens_seen": 14660488,
|
| 737 |
+
"step": 455
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 8.860635696821516,
|
| 741 |
+
"grad_norm": 0.7203698754310608,
|
| 742 |
+
"learning_rate": 6.870799593678459e-06,
|
| 743 |
+
"loss": 0.0116,
|
| 744 |
+
"num_input_tokens_seen": 14824328,
|
| 745 |
+
"step": 460
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 8.95843520782396,
|
| 749 |
+
"grad_norm": 0.7533681392669678,
|
| 750 |
+
"learning_rate": 6.6764520052034054e-06,
|
| 751 |
+
"loss": 0.0111,
|
| 752 |
+
"num_input_tokens_seen": 14988168,
|
| 753 |
+
"step": 465
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 9.039119804400977,
|
| 757 |
+
"grad_norm": 0.22926057875156403,
|
| 758 |
+
"learning_rate": 6.483505627862632e-06,
|
| 759 |
+
"loss": 0.0071,
|
| 760 |
+
"num_input_tokens_seen": 15121288,
|
| 761 |
+
"step": 470
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 9.136919315403423,
|
| 765 |
+
"grad_norm": 0.4467085301876068,
|
| 766 |
+
"learning_rate": 6.292041808040393e-06,
|
| 767 |
+
"loss": 0.0054,
|
| 768 |
+
"num_input_tokens_seen": 15285128,
|
| 769 |
+
"step": 475
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 9.234718826405867,
|
| 773 |
+
"grad_norm": 0.5730965733528137,
|
| 774 |
+
"learning_rate": 6.102141267073207e-06,
|
| 775 |
+
"loss": 0.0059,
|
| 776 |
+
"num_input_tokens_seen": 15448968,
|
| 777 |
+
"step": 480
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 9.332518337408313,
|
| 781 |
+
"grad_norm": 0.3801887333393097,
|
| 782 |
+
"learning_rate": 5.913884067217686e-06,
|
| 783 |
+
"loss": 0.0045,
|
| 784 |
+
"num_input_tokens_seen": 15612552,
|
| 785 |
+
"step": 485
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 9.430317848410757,
|
| 789 |
+
"grad_norm": 0.4900866448879242,
|
| 790 |
+
"learning_rate": 5.727349577896194e-06,
|
| 791 |
+
"loss": 0.004,
|
| 792 |
+
"num_input_tokens_seen": 15776392,
|
| 793 |
+
"step": 490
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 9.528117359413203,
|
| 797 |
+
"grad_norm": 0.3590303957462311,
|
| 798 |
+
"learning_rate": 5.542616442234618e-06,
|
| 799 |
+
"loss": 0.0042,
|
| 800 |
+
"num_input_tokens_seen": 15940232,
|
| 801 |
+
"step": 495
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 9.625916870415647,
|
| 805 |
+
"grad_norm": 0.36999812722206116,
|
| 806 |
+
"learning_rate": 5.3597625439063685e-06,
|
| 807 |
+
"loss": 0.0049,
|
| 808 |
+
"num_input_tokens_seen": 16104072,
|
| 809 |
+
"step": 500
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 9.723716381418093,
|
| 813 |
+
"grad_norm": 0.3589613735675812,
|
| 814 |
+
"learning_rate": 5.178864974296511e-06,
|
| 815 |
+
"loss": 0.0045,
|
| 816 |
+
"num_input_tokens_seen": 16267912,
|
| 817 |
+
"step": 505
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 9.821515892420537,
|
| 821 |
+
"grad_norm": 0.30231136083602905,
|
| 822 |
+
"learning_rate": 5.000000000000003e-06,
|
| 823 |
+
"loss": 0.004,
|
| 824 |
+
"num_input_tokens_seen": 16431752,
|
| 825 |
+
"step": 510
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 9.919315403422983,
|
| 829 |
+
"grad_norm": 0.34575000405311584,
|
| 830 |
+
"learning_rate": 4.823243030667576e-06,
|
| 831 |
+
"loss": 0.0042,
|
| 832 |
+
"num_input_tokens_seen": 16595592,
|
| 833 |
+
"step": 515
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 10.0,
|
| 837 |
+
"grad_norm": 0.6714850068092346,
|
| 838 |
+
"learning_rate": 4.648668587212998e-06,
|
| 839 |
+
"loss": 0.0037,
|
| 840 |
+
"num_input_tokens_seen": 16728712,
|
| 841 |
+
"step": 520
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 10.097799511002446,
|
| 845 |
+
"grad_norm": 0.19611337780952454,
|
| 846 |
+
"learning_rate": 4.476350270394942e-06,
|
| 847 |
+
"loss": 0.0021,
|
| 848 |
+
"num_input_tokens_seen": 16892552,
|
| 849 |
+
"step": 525
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 10.19559902200489,
|
| 853 |
+
"grad_norm": 0.13641950488090515,
|
| 854 |
+
"learning_rate": 4.306360729786867e-06,
|
| 855 |
+
"loss": 0.002,
|
| 856 |
+
"num_input_tokens_seen": 17056392,
|
| 857 |
+
"step": 530
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 10.293398533007334,
|
| 861 |
+
"grad_norm": 0.16702738404273987,
|
| 862 |
+
"learning_rate": 4.138771633147856e-06,
|
| 863 |
+
"loss": 0.0023,
|
| 864 |
+
"num_input_tokens_seen": 17220232,
|
| 865 |
+
"step": 535
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 10.39119804400978,
|
| 869 |
+
"grad_norm": 0.2278226763010025,
|
| 870 |
+
"learning_rate": 3.973653636207437e-06,
|
| 871 |
+
"loss": 0.002,
|
| 872 |
+
"num_input_tokens_seen": 17384072,
|
| 873 |
+
"step": 540
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 10.488997555012224,
|
| 877 |
+
"grad_norm": 0.1814439296722412,
|
| 878 |
+
"learning_rate": 3.8110763528770543e-06,
|
| 879 |
+
"loss": 0.0021,
|
| 880 |
+
"num_input_tokens_seen": 17547912,
|
| 881 |
+
"step": 545
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 10.58679706601467,
|
| 885 |
+
"grad_norm": 0.2528248429298401,
|
| 886 |
+
"learning_rate": 3.651108325900773e-06,
|
| 887 |
+
"loss": 0.0025,
|
| 888 |
+
"num_input_tokens_seen": 17711752,
|
| 889 |
+
"step": 550
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 10.684596577017114,
|
| 893 |
+
"grad_norm": 0.165152445435524,
|
| 894 |
+
"learning_rate": 3.493816997957582e-06,
|
| 895 |
+
"loss": 0.0021,
|
| 896 |
+
"num_input_tokens_seen": 17875592,
|
| 897 |
+
"step": 555
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 10.78239608801956,
|
| 901 |
+
"grad_norm": 0.08041153103113174,
|
| 902 |
+
"learning_rate": 3.339268683227499e-06,
|
| 903 |
+
"loss": 0.0017,
|
| 904 |
+
"num_input_tokens_seen": 18039432,
|
| 905 |
+
"step": 560
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 10.880195599022004,
|
| 909 |
+
"grad_norm": 0.24567244946956635,
|
| 910 |
+
"learning_rate": 3.1875285394334575e-06,
|
| 911 |
+
"loss": 0.0018,
|
| 912 |
+
"num_input_tokens_seen": 18203272,
|
| 913 |
+
"step": 565
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 10.97799511002445,
|
| 917 |
+
"grad_norm": 0.21529506146907806,
|
| 918 |
+
"learning_rate": 3.0386605403707347e-06,
|
| 919 |
+
"loss": 0.0018,
|
| 920 |
+
"num_input_tokens_seen": 18367112,
|
| 921 |
+
"step": 570
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 11.058679706601467,
|
| 925 |
+
"grad_norm": 0.05531800910830498,
|
| 926 |
+
"learning_rate": 2.8927274489355296e-06,
|
| 927 |
+
"loss": 0.0014,
|
| 928 |
+
"num_input_tokens_seen": 18500232,
|
| 929 |
+
"step": 575
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 11.156479217603913,
|
| 933 |
+
"grad_norm": 0.04145563766360283,
|
| 934 |
+
"learning_rate": 2.749790790664074e-06,
|
| 935 |
+
"loss": 0.0013,
|
| 936 |
+
"num_input_tokens_seen": 18662984,
|
| 937 |
+
"step": 580
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 11.254278728606357,
|
| 941 |
+
"grad_norm": 0.07673907279968262,
|
| 942 |
+
"learning_rate": 2.6099108277934105e-06,
|
| 943 |
+
"loss": 0.0014,
|
| 944 |
+
"num_input_tokens_seen": 18826824,
|
| 945 |
+
"step": 585
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 11.352078239608803,
|
| 949 |
+
"grad_norm": 0.054488956928253174,
|
| 950 |
+
"learning_rate": 2.4731465338547556e-06,
|
| 951 |
+
"loss": 0.0013,
|
| 952 |
+
"num_input_tokens_seen": 18990664,
|
| 953 |
+
"step": 590
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 11.449877750611247,
|
| 957 |
+
"grad_norm": 0.11207219213247299,
|
| 958 |
+
"learning_rate": 2.339555568810221e-06,
|
| 959 |
+
"loss": 0.0013,
|
| 960 |
+
"num_input_tokens_seen": 19154504,
|
| 961 |
+
"step": 595
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 11.547677261613693,
|
| 965 |
+
"grad_norm": 0.062526635825634,
|
| 966 |
+
"learning_rate": 2.209194254743295e-06,
|
| 967 |
+
"loss": 0.0012,
|
| 968 |
+
"num_input_tokens_seen": 19318344,
|
| 969 |
+
"step": 600
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 11.645476772616137,
|
| 973 |
+
"grad_norm": 0.05114143341779709,
|
| 974 |
+
"learning_rate": 2.0821175521134208e-06,
|
| 975 |
+
"loss": 0.0013,
|
| 976 |
+
"num_input_tokens_seen": 19482184,
|
| 977 |
+
"step": 605
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 11.743276283618583,
|
| 981 |
+
"grad_norm": 0.05199455842375755,
|
| 982 |
+
"learning_rate": 1.9583790365845823e-06,
|
| 983 |
+
"loss": 0.0013,
|
| 984 |
+
"num_input_tokens_seen": 19646024,
|
| 985 |
+
"step": 610
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 11.841075794621027,
|
| 989 |
+
"grad_norm": 0.057281751185655594,
|
| 990 |
+
"learning_rate": 1.8380308764377841e-06,
|
| 991 |
+
"loss": 0.0014,
|
| 992 |
+
"num_input_tokens_seen": 19809864,
|
| 993 |
+
"step": 615
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 11.938875305623473,
|
| 997 |
+
"grad_norm": 0.04751597344875336,
|
| 998 |
+
"learning_rate": 1.7211238105768213e-06,
|
| 999 |
+
"loss": 0.0012,
|
| 1000 |
+
"num_input_tokens_seen": 19973704,
|
| 1001 |
+
"step": 620
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 12.01955990220049,
|
| 1005 |
+
"grad_norm": 0.04285968840122223,
|
| 1006 |
+
"learning_rate": 1.607707127136734e-06,
|
| 1007 |
+
"loss": 0.0014,
|
| 1008 |
+
"num_input_tokens_seen": 20106824,
|
| 1009 |
+
"step": 625
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 12.117359413202934,
|
| 1013 |
+
"grad_norm": 0.044368330389261246,
|
| 1014 |
+
"learning_rate": 1.4978286427038602e-06,
|
| 1015 |
+
"loss": 0.0011,
|
| 1016 |
+
"num_input_tokens_seen": 20270664,
|
| 1017 |
+
"step": 630
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 12.21515892420538,
|
| 1021 |
+
"grad_norm": 0.04820827767252922,
|
| 1022 |
+
"learning_rate": 1.3915346821563235e-06,
|
| 1023 |
+
"loss": 0.0012,
|
| 1024 |
+
"num_input_tokens_seen": 20434504,
|
| 1025 |
+
"step": 635
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 12.312958435207824,
|
| 1029 |
+
"grad_norm": 0.034025732427835464,
|
| 1030 |
+
"learning_rate": 1.2888700591334225e-06,
|
| 1031 |
+
"loss": 0.0011,
|
| 1032 |
+
"num_input_tokens_seen": 20598280,
|
| 1033 |
+
"step": 640
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 12.41075794621027,
|
| 1037 |
+
"grad_norm": 0.03853330388665199,
|
| 1038 |
+
"learning_rate": 1.1898780571421554e-06,
|
| 1039 |
+
"loss": 0.0011,
|
| 1040 |
+
"num_input_tokens_seen": 20762120,
|
| 1041 |
+
"step": 645
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 12.508557457212714,
|
| 1045 |
+
"grad_norm": 0.039888255298137665,
|
| 1046 |
+
"learning_rate": 1.0946004113088381e-06,
|
| 1047 |
+
"loss": 0.0012,
|
| 1048 |
+
"num_input_tokens_seen": 20925960,
|
| 1049 |
+
"step": 650
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 12.60635696821516,
|
| 1053 |
+
"grad_norm": 0.04366978630423546,
|
| 1054 |
+
"learning_rate": 1.0030772907835484e-06,
|
| 1055 |
+
"loss": 0.001,
|
| 1056 |
+
"num_input_tokens_seen": 21089800,
|
| 1057 |
+
"step": 655
|
| 1058 |
+
},
|
| 1059 |
+
{
|
| 1060 |
+
"epoch": 12.704156479217604,
|
| 1061 |
+
"grad_norm": 0.03416445106267929,
|
| 1062 |
+
"learning_rate": 9.153472818047627e-07,
|
| 1063 |
+
"loss": 0.001,
|
| 1064 |
+
"num_input_tokens_seen": 21253640,
|
| 1065 |
+
"step": 660
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 12.80195599022005,
|
| 1069 |
+
"grad_norm": 0.04817335307598114,
|
| 1070 |
+
"learning_rate": 8.31447371431372e-07,
|
| 1071 |
+
"loss": 0.0011,
|
| 1072 |
+
"num_input_tokens_seen": 21417480,
|
| 1073 |
+
"step": 665
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 12.899755501222494,
|
| 1077 |
+
"grad_norm": 0.03280309960246086,
|
| 1078 |
+
"learning_rate": 7.514129319488839e-07,
|
| 1079 |
+
"loss": 0.0011,
|
| 1080 |
+
"num_input_tokens_seen": 21581320,
|
| 1081 |
+
"step": 670
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 12.99755501222494,
|
| 1085 |
+
"grad_norm": 0.04426475614309311,
|
| 1086 |
+
"learning_rate": 6.752777059564431e-07,
|
| 1087 |
+
"loss": 0.0011,
|
| 1088 |
+
"num_input_tokens_seen": 21745160,
|
| 1089 |
+
"step": 675
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 13.078239608801956,
|
| 1093 |
+
"grad_norm": 0.03248964622616768,
|
| 1094 |
+
"learning_rate": 6.030737921409169e-07,
|
| 1095 |
+
"loss": 0.0011,
|
| 1096 |
+
"num_input_tokens_seen": 21877864,
|
| 1097 |
+
"step": 680
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 13.1760391198044,
|
| 1101 |
+
"grad_norm": 0.0347786545753479,
|
| 1102 |
+
"learning_rate": 5.348316317440549e-07,
|
| 1103 |
+
"loss": 0.001,
|
| 1104 |
+
"num_input_tokens_seen": 22041704,
|
| 1105 |
+
"step": 685
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 13.273838630806846,
|
| 1109 |
+
"grad_norm": 0.04258348420262337,
|
| 1110 |
+
"learning_rate": 4.7057999572843516e-07,
|
| 1111 |
+
"loss": 0.001,
|
| 1112 |
+
"num_input_tokens_seen": 22205544,
|
| 1113 |
+
"step": 690
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 13.37163814180929,
|
| 1117 |
+
"grad_norm": 0.0364152230322361,
|
| 1118 |
+
"learning_rate": 4.103459726475889e-07,
|
| 1119 |
+
"loss": 0.0011,
|
| 1120 |
+
"num_input_tokens_seen": 22369384,
|
| 1121 |
+
"step": 695
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 13.469437652811736,
|
| 1125 |
+
"grad_norm": 0.039297617971897125,
|
| 1126 |
+
"learning_rate": 3.541549572254488e-07,
|
| 1127 |
+
"loss": 0.0011,
|
| 1128 |
+
"num_input_tokens_seen": 22533224,
|
| 1129 |
+
"step": 700
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 13.56723716381418,
|
| 1133 |
+
"grad_norm": 0.037363260984420776,
|
| 1134 |
+
"learning_rate": 3.020306396499062e-07,
|
| 1135 |
+
"loss": 0.0009,
|
| 1136 |
+
"num_input_tokens_seen": 22697064,
|
| 1137 |
+
"step": 705
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 13.665036674816626,
|
| 1141 |
+
"grad_norm": 0.032989222556352615,
|
| 1142 |
+
"learning_rate": 2.539949955849985e-07,
|
| 1143 |
+
"loss": 0.0011,
|
| 1144 |
+
"num_input_tokens_seen": 22860904,
|
| 1145 |
+
"step": 710
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 13.76283618581907,
|
| 1149 |
+
"grad_norm": 0.03354249894618988,
|
| 1150 |
+
"learning_rate": 2.1006827690595478e-07,
|
| 1151 |
+
"loss": 0.001,
|
| 1152 |
+
"num_input_tokens_seen": 23024744,
|
| 1153 |
+
"step": 715
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 13.860635696821516,
|
| 1157 |
+
"grad_norm": 0.03322712704539299,
|
| 1158 |
+
"learning_rate": 1.7026900316098217e-07,
|
| 1159 |
+
"loss": 0.001,
|
| 1160 |
+
"num_input_tokens_seen": 23188584,
|
| 1161 |
+
"step": 720
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 13.95843520782396,
|
| 1165 |
+
"grad_norm": 0.03775469958782196,
|
| 1166 |
+
"learning_rate": 1.3461395376340502e-07,
|
| 1167 |
+
"loss": 0.001,
|
| 1168 |
+
"num_input_tokens_seen": 23352424,
|
| 1169 |
+
"step": 725
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"epoch": 14.039119804400977,
|
| 1173 |
+
"grad_norm": 0.039018385112285614,
|
| 1174 |
+
"learning_rate": 1.0311816091744698e-07,
|
| 1175 |
+
"loss": 0.001,
|
| 1176 |
+
"num_input_tokens_seen": 23485296,
|
| 1177 |
+
"step": 730
|
| 1178 |
+
},
|
| 1179 |
+
{
|
| 1180 |
+
"epoch": 14.136919315403423,
|
| 1181 |
+
"grad_norm": 0.03440910577774048,
|
| 1182 |
+
"learning_rate": 7.579490328064265e-08,
|
| 1183 |
+
"loss": 0.001,
|
| 1184 |
+
"num_input_tokens_seen": 23649136,
|
| 1185 |
+
"step": 735
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 14.234718826405867,
|
| 1189 |
+
"grad_norm": 0.03226885199546814,
|
| 1190 |
+
"learning_rate": 5.265570036553813e-08,
|
| 1191 |
+
"loss": 0.001,
|
| 1192 |
+
"num_input_tokens_seen": 23812976,
|
| 1193 |
+
"step": 740
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 14.332518337408313,
|
| 1197 |
+
"grad_norm": 0.03353099152445793,
|
| 1198 |
+
"learning_rate": 3.371030768305583e-08,
|
| 1199 |
+
"loss": 0.0009,
|
| 1200 |
+
"num_input_tokens_seen": 23976816,
|
| 1201 |
+
"step": 745
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 14.430317848410757,
|
| 1205 |
+
"grad_norm": 0.04043235257267952,
|
| 1206 |
+
"learning_rate": 1.896671262955896e-08,
|
| 1207 |
+
"loss": 0.001,
|
| 1208 |
+
"num_input_tokens_seen": 24140656,
|
| 1209 |
+
"step": 750
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 14.528117359413203,
|
| 1213 |
+
"grad_norm": 0.0331251285970211,
|
| 1214 |
+
"learning_rate": 8.431131119361891e-09,
|
| 1215 |
+
"loss": 0.001,
|
| 1216 |
+
"num_input_tokens_seen": 24304496,
|
| 1217 |
+
"step": 755
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 14.625916870415647,
|
| 1221 |
+
"grad_norm": 0.03793644160032272,
|
| 1222 |
+
"learning_rate": 2.108004964086474e-09,
|
| 1223 |
+
"loss": 0.001,
|
| 1224 |
+
"num_input_tokens_seen": 24467760,
|
| 1225 |
+
"step": 760
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 14.723716381418093,
|
| 1229 |
+
"grad_norm": 0.03213175758719444,
|
| 1230 |
+
"learning_rate": 0.0,
|
| 1231 |
+
"loss": 0.001,
|
| 1232 |
+
"num_input_tokens_seen": 24631600,
|
| 1233 |
+
"step": 765
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 14.723716381418093,
|
| 1237 |
+
"num_input_tokens_seen": 24631600,
|
| 1238 |
+
"step": 765,
|
| 1239 |
+
"total_flos": 1.936547143200768e+17,
|
| 1240 |
+
"train_loss": 0.1555636612302776,
|
| 1241 |
+
"train_runtime": 3426.6896,
|
| 1242 |
+
"train_samples_per_second": 3.576,
|
| 1243 |
+
"train_steps_per_second": 0.223
|
| 1244 |
+
}
|
| 1245 |
+
],
|
| 1246 |
+
"logging_steps": 5,
|
| 1247 |
+
"max_steps": 765,
|
| 1248 |
+
"num_input_tokens_seen": 24631600,
|
| 1249 |
+
"num_train_epochs": 15,
|
| 1250 |
+
"save_steps": 100,
|
| 1251 |
+
"stateful_callbacks": {
|
| 1252 |
+
"TrainerControl": {
|
| 1253 |
+
"args": {
|
| 1254 |
+
"should_epoch_stop": false,
|
| 1255 |
+
"should_evaluate": false,
|
| 1256 |
+
"should_log": false,
|
| 1257 |
+
"should_save": true,
|
| 1258 |
+
"should_training_stop": true
|
| 1259 |
+
},
|
| 1260 |
+
"attributes": {}
|
| 1261 |
+
}
|
| 1262 |
+
},
|
| 1263 |
+
"total_flos": 1.936547143200768e+17,
|
| 1264 |
+
"train_batch_size": 2,
|
| 1265 |
+
"trial_name": null,
|
| 1266 |
+
"trial_params": null
|
| 1267 |
+
}
|