Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1258 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1258 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 50000,
|
| 3 |
+
"best_metric": 0.5881,
|
| 4 |
+
"best_model_checkpoint": "models/NED/SPACCC_full_upsampled_tfidf/Meta-Llama-3-8B-Instruct/checkpoint-50000",
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 2000,
|
| 7 |
+
"global_step": 103965,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"entropy": 1.771104082763195,
|
| 14 |
+
"epoch": 0.057711729909104025,
|
| 15 |
+
"grad_norm": 8.875,
|
| 16 |
+
"learning_rate": 1.9227316447579353e-05,
|
| 17 |
+
"loss": 0.807,
|
| 18 |
+
"mean_token_accuracy": 0.8383643639683723,
|
| 19 |
+
"num_tokens": 15534479.0,
|
| 20 |
+
"step": 2000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.057711729909104025,
|
| 24 |
+
"eval_entropy": 1.7064778925563795,
|
| 25 |
+
"eval_loss": 0.5752137899398804,
|
| 26 |
+
"eval_mean_token_accuracy": 0.8736604764762166,
|
| 27 |
+
"eval_num_gold": 908,
|
| 28 |
+
"eval_num_guess": 908,
|
| 29 |
+
"eval_num_tokens": 15534479.0,
|
| 30 |
+
"eval_recall": 0.4857,
|
| 31 |
+
"eval_runtime": 19.5541,
|
| 32 |
+
"eval_samples_per_second": 46.435,
|
| 33 |
+
"eval_steps_per_second": 46.435,
|
| 34 |
+
"step": 2000
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"entropy": 1.7855026668310165,
|
| 38 |
+
"epoch": 0.11542345981820805,
|
| 39 |
+
"grad_norm": 4.9375,
|
| 40 |
+
"learning_rate": 2.973821470360748e-05,
|
| 41 |
+
"loss": 0.4284,
|
| 42 |
+
"mean_token_accuracy": 0.9000137696564198,
|
| 43 |
+
"num_tokens": 31091376.0,
|
| 44 |
+
"step": 4000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.11542345981820805,
|
| 48 |
+
"eval_entropy": 1.9107416086391205,
|
| 49 |
+
"eval_loss": 0.6141767501831055,
|
| 50 |
+
"eval_mean_token_accuracy": 0.8813558152921924,
|
| 51 |
+
"eval_num_gold": 908,
|
| 52 |
+
"eval_num_guess": 908,
|
| 53 |
+
"eval_num_tokens": 31091376.0,
|
| 54 |
+
"eval_recall": 0.5132,
|
| 55 |
+
"eval_runtime": 18.6743,
|
| 56 |
+
"eval_samples_per_second": 48.623,
|
| 57 |
+
"eval_steps_per_second": 48.623,
|
| 58 |
+
"step": 4000
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"entropy": 1.7454647228717803,
|
| 62 |
+
"epoch": 0.17313518972731207,
|
| 63 |
+
"grad_norm": 6.40625,
|
| 64 |
+
"learning_rate": 2.914324812089721e-05,
|
| 65 |
+
"loss": 0.3478,
|
| 66 |
+
"mean_token_accuracy": 0.9183464118242264,
|
| 67 |
+
"num_tokens": 46695519.0,
|
| 68 |
+
"step": 6000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.17313518972731207,
|
| 72 |
+
"eval_entropy": 1.6889207995279245,
|
| 73 |
+
"eval_loss": 0.7172139286994934,
|
| 74 |
+
"eval_mean_token_accuracy": 0.8881947658627831,
|
| 75 |
+
"eval_num_gold": 908,
|
| 76 |
+
"eval_num_guess": 908,
|
| 77 |
+
"eval_num_tokens": 46695519.0,
|
| 78 |
+
"eval_recall": 0.5551,
|
| 79 |
+
"eval_runtime": 18.7505,
|
| 80 |
+
"eval_samples_per_second": 48.425,
|
| 81 |
+
"eval_steps_per_second": 48.425,
|
| 82 |
+
"step": 6000
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"entropy": 1.686599359869957,
|
| 86 |
+
"epoch": 0.2308469196364161,
|
| 87 |
+
"grad_norm": 5.0,
|
| 88 |
+
"learning_rate": 2.8548281538186937e-05,
|
| 89 |
+
"loss": 0.2911,
|
| 90 |
+
"mean_token_accuracy": 0.9297748121023178,
|
| 91 |
+
"num_tokens": 62201311.0,
|
| 92 |
+
"step": 8000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.2308469196364161,
|
| 96 |
+
"eval_entropy": 1.765830583837589,
|
| 97 |
+
"eval_loss": 0.6870580315589905,
|
| 98 |
+
"eval_mean_token_accuracy": 0.8954090231184392,
|
| 99 |
+
"eval_num_gold": 908,
|
| 100 |
+
"eval_num_guess": 908,
|
| 101 |
+
"eval_num_tokens": 62201311.0,
|
| 102 |
+
"eval_recall": 0.554,
|
| 103 |
+
"eval_runtime": 18.2183,
|
| 104 |
+
"eval_samples_per_second": 49.84,
|
| 105 |
+
"eval_steps_per_second": 49.84,
|
| 106 |
+
"step": 8000
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"entropy": 1.6513705806136132,
|
| 110 |
+
"epoch": 0.28855864954552013,
|
| 111 |
+
"grad_norm": 2.890625,
|
| 112 |
+
"learning_rate": 2.795331495547667e-05,
|
| 113 |
+
"loss": 0.2723,
|
| 114 |
+
"mean_token_accuracy": 0.9330688781142235,
|
| 115 |
+
"num_tokens": 77723725.0,
|
| 116 |
+
"step": 10000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.28855864954552013,
|
| 120 |
+
"eval_entropy": 1.7299244680199855,
|
| 121 |
+
"eval_loss": 0.671196460723877,
|
| 122 |
+
"eval_mean_token_accuracy": 0.8996821519003828,
|
| 123 |
+
"eval_num_gold": 908,
|
| 124 |
+
"eval_num_guess": 908,
|
| 125 |
+
"eval_num_tokens": 77723725.0,
|
| 126 |
+
"eval_recall": 0.5837,
|
| 127 |
+
"eval_runtime": 18.1693,
|
| 128 |
+
"eval_samples_per_second": 49.975,
|
| 129 |
+
"eval_steps_per_second": 49.975,
|
| 130 |
+
"step": 10000
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"entropy": 1.6137347612977029,
|
| 134 |
+
"epoch": 0.34627037945462413,
|
| 135 |
+
"grad_norm": 4.15625,
|
| 136 |
+
"learning_rate": 2.7358348372766396e-05,
|
| 137 |
+
"loss": 0.247,
|
| 138 |
+
"mean_token_accuracy": 0.938195524007082,
|
| 139 |
+
"num_tokens": 93268948.0,
|
| 140 |
+
"step": 12000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.34627037945462413,
|
| 144 |
+
"eval_entropy": 1.642808350631844,
|
| 145 |
+
"eval_loss": 0.7069945335388184,
|
| 146 |
+
"eval_mean_token_accuracy": 0.9005183333998735,
|
| 147 |
+
"eval_num_gold": 908,
|
| 148 |
+
"eval_num_guess": 908,
|
| 149 |
+
"eval_num_tokens": 93268948.0,
|
| 150 |
+
"eval_recall": 0.5683,
|
| 151 |
+
"eval_runtime": 18.2098,
|
| 152 |
+
"eval_samples_per_second": 49.863,
|
| 153 |
+
"eval_steps_per_second": 49.863,
|
| 154 |
+
"step": 12000
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"entropy": 1.5417808018922805,
|
| 158 |
+
"epoch": 0.4039821093637282,
|
| 159 |
+
"grad_norm": 6.15625,
|
| 160 |
+
"learning_rate": 2.6763381790056127e-05,
|
| 161 |
+
"loss": 0.2369,
|
| 162 |
+
"mean_token_accuracy": 0.9400165711343288,
|
| 163 |
+
"num_tokens": 108712209.0,
|
| 164 |
+
"step": 14000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.4039821093637282,
|
| 168 |
+
"eval_entropy": 1.6070995822483223,
|
| 169 |
+
"eval_loss": 0.7336843013763428,
|
| 170 |
+
"eval_mean_token_accuracy": 0.9035924484026064,
|
| 171 |
+
"eval_num_gold": 908,
|
| 172 |
+
"eval_num_guess": 908,
|
| 173 |
+
"eval_num_tokens": 108712209.0,
|
| 174 |
+
"eval_recall": 0.5749,
|
| 175 |
+
"eval_runtime": 18.1147,
|
| 176 |
+
"eval_samples_per_second": 50.125,
|
| 177 |
+
"eval_steps_per_second": 50.125,
|
| 178 |
+
"step": 14000
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"entropy": 1.5307720832824707,
|
| 182 |
+
"epoch": 0.4616938392728322,
|
| 183 |
+
"grad_norm": 1.3203125,
|
| 184 |
+
"learning_rate": 2.6168415207345855e-05,
|
| 185 |
+
"loss": 0.224,
|
| 186 |
+
"mean_token_accuracy": 0.9427571404874325,
|
| 187 |
+
"num_tokens": 124237300.0,
|
| 188 |
+
"step": 16000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.4616938392728322,
|
| 192 |
+
"eval_entropy": 1.6007142412242386,
|
| 193 |
+
"eval_loss": 0.723686933517456,
|
| 194 |
+
"eval_mean_token_accuracy": 0.9036177859891878,
|
| 195 |
+
"eval_num_gold": 908,
|
| 196 |
+
"eval_num_guess": 908,
|
| 197 |
+
"eval_num_tokens": 124237300.0,
|
| 198 |
+
"eval_recall": 0.5738,
|
| 199 |
+
"eval_runtime": 18.0095,
|
| 200 |
+
"eval_samples_per_second": 50.418,
|
| 201 |
+
"eval_steps_per_second": 50.418,
|
| 202 |
+
"step": 16000
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"entropy": 1.5197161840200424,
|
| 206 |
+
"epoch": 0.5194055691819363,
|
| 207 |
+
"grad_norm": 4.375,
|
| 208 |
+
"learning_rate": 2.5573448624635583e-05,
|
| 209 |
+
"loss": 0.2127,
|
| 210 |
+
"mean_token_accuracy": 0.9449084457457065,
|
| 211 |
+
"num_tokens": 139804083.0,
|
| 212 |
+
"step": 18000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.5194055691819363,
|
| 216 |
+
"eval_entropy": 1.5377285603909765,
|
| 217 |
+
"eval_loss": 0.7464824318885803,
|
| 218 |
+
"eval_mean_token_accuracy": 0.9009800682968505,
|
| 219 |
+
"eval_num_gold": 908,
|
| 220 |
+
"eval_num_guess": 908,
|
| 221 |
+
"eval_num_tokens": 139804083.0,
|
| 222 |
+
"eval_recall": 0.5694,
|
| 223 |
+
"eval_runtime": 17.3142,
|
| 224 |
+
"eval_samples_per_second": 52.443,
|
| 225 |
+
"eval_steps_per_second": 52.443,
|
| 226 |
+
"step": 18000
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"entropy": 1.4599213127493857,
|
| 230 |
+
"epoch": 0.5771172990910403,
|
| 231 |
+
"grad_norm": 4.96875,
|
| 232 |
+
"learning_rate": 2.4978482041925314e-05,
|
| 233 |
+
"loss": 0.2026,
|
| 234 |
+
"mean_token_accuracy": 0.9471077627837657,
|
| 235 |
+
"num_tokens": 155379953.0,
|
| 236 |
+
"step": 20000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.5771172990910403,
|
| 240 |
+
"eval_entropy": 1.475688523001608,
|
| 241 |
+
"eval_loss": 0.762593150138855,
|
| 242 |
+
"eval_mean_token_accuracy": 0.9033181490089399,
|
| 243 |
+
"eval_num_gold": 908,
|
| 244 |
+
"eval_num_guess": 908,
|
| 245 |
+
"eval_num_tokens": 155379953.0,
|
| 246 |
+
"eval_recall": 0.5705,
|
| 247 |
+
"eval_runtime": 17.8659,
|
| 248 |
+
"eval_samples_per_second": 50.823,
|
| 249 |
+
"eval_steps_per_second": 50.823,
|
| 250 |
+
"step": 20000
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"entropy": 1.4277473657727242,
|
| 254 |
+
"epoch": 0.6348290290001443,
|
| 255 |
+
"grad_norm": 5.15625,
|
| 256 |
+
"learning_rate": 2.4383515459215042e-05,
|
| 257 |
+
"loss": 0.1907,
|
| 258 |
+
"mean_token_accuracy": 0.9498722539842128,
|
| 259 |
+
"num_tokens": 170982670.0,
|
| 260 |
+
"step": 22000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.6348290290001443,
|
| 264 |
+
"eval_entropy": 1.5171283333837198,
|
| 265 |
+
"eval_loss": 0.7434535026550293,
|
| 266 |
+
"eval_mean_token_accuracy": 0.9021543205965983,
|
| 267 |
+
"eval_num_gold": 908,
|
| 268 |
+
"eval_num_guess": 908,
|
| 269 |
+
"eval_num_tokens": 170982670.0,
|
| 270 |
+
"eval_recall": 0.5727,
|
| 271 |
+
"eval_runtime": 17.4411,
|
| 272 |
+
"eval_samples_per_second": 52.061,
|
| 273 |
+
"eval_steps_per_second": 52.061,
|
| 274 |
+
"step": 22000
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"entropy": 1.4187169399261474,
|
| 278 |
+
"epoch": 0.6925407589092483,
|
| 279 |
+
"grad_norm": 2.9375,
|
| 280 |
+
"learning_rate": 2.378854887650477e-05,
|
| 281 |
+
"loss": 0.1905,
|
| 282 |
+
"mean_token_accuracy": 0.9493927232325077,
|
| 283 |
+
"num_tokens": 186663193.0,
|
| 284 |
+
"step": 24000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.6925407589092483,
|
| 288 |
+
"eval_entropy": 1.4800235901503835,
|
| 289 |
+
"eval_loss": 0.7626135945320129,
|
| 290 |
+
"eval_mean_token_accuracy": 0.9030716799298047,
|
| 291 |
+
"eval_num_gold": 908,
|
| 292 |
+
"eval_num_guess": 908,
|
| 293 |
+
"eval_num_tokens": 186663193.0,
|
| 294 |
+
"eval_recall": 0.5771,
|
| 295 |
+
"eval_runtime": 17.6742,
|
| 296 |
+
"eval_samples_per_second": 51.374,
|
| 297 |
+
"eval_steps_per_second": 51.374,
|
| 298 |
+
"step": 24000
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"entropy": 1.3858990859389304,
|
| 302 |
+
"epoch": 0.7502524888183524,
|
| 303 |
+
"grad_norm": 5.4375,
|
| 304 |
+
"learning_rate": 2.31935822937945e-05,
|
| 305 |
+
"loss": 0.1847,
|
| 306 |
+
"mean_token_accuracy": 0.9506744608581066,
|
| 307 |
+
"num_tokens": 202382574.0,
|
| 308 |
+
"step": 26000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.7502524888183524,
|
| 312 |
+
"eval_entropy": 1.4523340987476483,
|
| 313 |
+
"eval_loss": 0.7447758316993713,
|
| 314 |
+
"eval_mean_token_accuracy": 0.9053517018777159,
|
| 315 |
+
"eval_num_gold": 908,
|
| 316 |
+
"eval_num_guess": 908,
|
| 317 |
+
"eval_num_tokens": 202382574.0,
|
| 318 |
+
"eval_recall": 0.5837,
|
| 319 |
+
"eval_runtime": 17.5503,
|
| 320 |
+
"eval_samples_per_second": 51.737,
|
| 321 |
+
"eval_steps_per_second": 51.737,
|
| 322 |
+
"step": 26000
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"entropy": 1.371607663989067,
|
| 326 |
+
"epoch": 0.8079642187274564,
|
| 327 |
+
"grad_norm": 6.46875,
|
| 328 |
+
"learning_rate": 2.259861571108423e-05,
|
| 329 |
+
"loss": 0.1747,
|
| 330 |
+
"mean_token_accuracy": 0.9534294557571411,
|
| 331 |
+
"num_tokens": 217894933.0,
|
| 332 |
+
"step": 28000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.8079642187274564,
|
| 336 |
+
"eval_entropy": 1.449135780728336,
|
| 337 |
+
"eval_loss": 0.726492166519165,
|
| 338 |
+
"eval_mean_token_accuracy": 0.9063065528607054,
|
| 339 |
+
"eval_num_gold": 908,
|
| 340 |
+
"eval_num_guess": 908,
|
| 341 |
+
"eval_num_tokens": 217894933.0,
|
| 342 |
+
"eval_recall": 0.5782,
|
| 343 |
+
"eval_runtime": 17.4046,
|
| 344 |
+
"eval_samples_per_second": 52.17,
|
| 345 |
+
"eval_steps_per_second": 52.17,
|
| 346 |
+
"step": 28000
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"entropy": 1.3474767149090767,
|
| 350 |
+
"epoch": 0.8656759486365604,
|
| 351 |
+
"grad_norm": 1.15625,
|
| 352 |
+
"learning_rate": 2.2003649128373957e-05,
|
| 353 |
+
"loss": 0.1662,
|
| 354 |
+
"mean_token_accuracy": 0.9552805411219597,
|
| 355 |
+
"num_tokens": 233442927.0,
|
| 356 |
+
"step": 30000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.8656759486365604,
|
| 360 |
+
"eval_entropy": 1.4218564205495272,
|
| 361 |
+
"eval_loss": 0.7416213750839233,
|
| 362 |
+
"eval_mean_token_accuracy": 0.9069001922129535,
|
| 363 |
+
"eval_num_gold": 908,
|
| 364 |
+
"eval_num_guess": 908,
|
| 365 |
+
"eval_num_tokens": 233442927.0,
|
| 366 |
+
"eval_recall": 0.587,
|
| 367 |
+
"eval_runtime": 17.6827,
|
| 368 |
+
"eval_samples_per_second": 51.35,
|
| 369 |
+
"eval_steps_per_second": 51.35,
|
| 370 |
+
"step": 30000
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"entropy": 1.337467650592327,
|
| 374 |
+
"epoch": 0.9233876785456644,
|
| 375 |
+
"grad_norm": 4.375,
|
| 376 |
+
"learning_rate": 2.1408682545663684e-05,
|
| 377 |
+
"loss": 0.1637,
|
| 378 |
+
"mean_token_accuracy": 0.9557446602284908,
|
| 379 |
+
"num_tokens": 249071659.0,
|
| 380 |
+
"step": 32000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.9233876785456644,
|
| 384 |
+
"eval_entropy": 1.3889587524835234,
|
| 385 |
+
"eval_loss": 0.7390624284744263,
|
| 386 |
+
"eval_mean_token_accuracy": 0.9073679444876537,
|
| 387 |
+
"eval_num_gold": 908,
|
| 388 |
+
"eval_num_guess": 908,
|
| 389 |
+
"eval_num_tokens": 249071659.0,
|
| 390 |
+
"eval_recall": 0.5848,
|
| 391 |
+
"eval_runtime": 17.5032,
|
| 392 |
+
"eval_samples_per_second": 51.876,
|
| 393 |
+
"eval_steps_per_second": 51.876,
|
| 394 |
+
"step": 32000
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"entropy": 1.3001156712770463,
|
| 398 |
+
"epoch": 0.9810994084547684,
|
| 399 |
+
"grad_norm": 3.890625,
|
| 400 |
+
"learning_rate": 2.0813715962953412e-05,
|
| 401 |
+
"loss": 0.1557,
|
| 402 |
+
"mean_token_accuracy": 0.9573536138236522,
|
| 403 |
+
"num_tokens": 264672169.0,
|
| 404 |
+
"step": 34000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.9810994084547684,
|
| 408 |
+
"eval_entropy": 1.4234498099214705,
|
| 409 |
+
"eval_loss": 0.7484801411628723,
|
| 410 |
+
"eval_mean_token_accuracy": 0.9075630426012997,
|
| 411 |
+
"eval_num_gold": 908,
|
| 412 |
+
"eval_num_guess": 908,
|
| 413 |
+
"eval_num_tokens": 264672169.0,
|
| 414 |
+
"eval_recall": 0.5859,
|
| 415 |
+
"eval_runtime": 16.2686,
|
| 416 |
+
"eval_samples_per_second": 55.813,
|
| 417 |
+
"eval_steps_per_second": 55.813,
|
| 418 |
+
"step": 34000
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"entropy": 1.2362971892952919,
|
| 422 |
+
"epoch": 1.0388111383638725,
|
| 423 |
+
"grad_norm": 4.96875,
|
| 424 |
+
"learning_rate": 2.0218749380243143e-05,
|
| 425 |
+
"loss": 0.1179,
|
| 426 |
+
"mean_token_accuracy": 0.9659893708825111,
|
| 427 |
+
"num_tokens": 280288735.0,
|
| 428 |
+
"step": 36000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 1.0388111383638725,
|
| 432 |
+
"eval_entropy": 1.268248316540592,
|
| 433 |
+
"eval_loss": 0.7719414830207825,
|
| 434 |
+
"eval_mean_token_accuracy": 0.9055645169796923,
|
| 435 |
+
"eval_num_gold": 908,
|
| 436 |
+
"eval_num_guess": 908,
|
| 437 |
+
"eval_num_tokens": 280288735.0,
|
| 438 |
+
"eval_recall": 0.5771,
|
| 439 |
+
"eval_runtime": 16.2819,
|
| 440 |
+
"eval_samples_per_second": 55.767,
|
| 441 |
+
"eval_steps_per_second": 55.767,
|
| 442 |
+
"step": 36000
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"entropy": 1.1861163977086544,
|
| 446 |
+
"epoch": 1.0965228682729764,
|
| 447 |
+
"grad_norm": 2.703125,
|
| 448 |
+
"learning_rate": 1.962378279753287e-05,
|
| 449 |
+
"loss": 0.1001,
|
| 450 |
+
"mean_token_accuracy": 0.9700573923885822,
|
| 451 |
+
"num_tokens": 295889252.0,
|
| 452 |
+
"step": 38000
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.0965228682729764,
|
| 456 |
+
"eval_entropy": 1.2859099207339308,
|
| 457 |
+
"eval_loss": 0.7740228176116943,
|
| 458 |
+
"eval_mean_token_accuracy": 0.9053704237425905,
|
| 459 |
+
"eval_num_gold": 908,
|
| 460 |
+
"eval_num_guess": 908,
|
| 461 |
+
"eval_num_tokens": 295889252.0,
|
| 462 |
+
"eval_recall": 0.5815,
|
| 463 |
+
"eval_runtime": 17.2227,
|
| 464 |
+
"eval_samples_per_second": 52.721,
|
| 465 |
+
"eval_steps_per_second": 52.721,
|
| 466 |
+
"step": 38000
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"entropy": 1.172551353752613,
|
| 470 |
+
"epoch": 1.1542345981820805,
|
| 471 |
+
"grad_norm": 5.09375,
|
| 472 |
+
"learning_rate": 1.90288162148226e-05,
|
| 473 |
+
"loss": 0.0999,
|
| 474 |
+
"mean_token_accuracy": 0.9701334120929241,
|
| 475 |
+
"num_tokens": 311460812.0,
|
| 476 |
+
"step": 40000
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 1.1542345981820805,
|
| 480 |
+
"eval_entropy": 1.2553664291613953,
|
| 481 |
+
"eval_loss": 0.7721803784370422,
|
| 482 |
+
"eval_mean_token_accuracy": 0.9035136323537071,
|
| 483 |
+
"eval_num_gold": 908,
|
| 484 |
+
"eval_num_guess": 908,
|
| 485 |
+
"eval_num_tokens": 311460812.0,
|
| 486 |
+
"eval_recall": 0.576,
|
| 487 |
+
"eval_runtime": 16.7852,
|
| 488 |
+
"eval_samples_per_second": 54.095,
|
| 489 |
+
"eval_steps_per_second": 54.095,
|
| 490 |
+
"step": 40000
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"entropy": 1.1619984501898288,
|
| 494 |
+
"epoch": 1.2119463280911846,
|
| 495 |
+
"grad_norm": 2.21875,
|
| 496 |
+
"learning_rate": 1.843384963211233e-05,
|
| 497 |
+
"loss": 0.0965,
|
| 498 |
+
"mean_token_accuracy": 0.9710877353549003,
|
| 499 |
+
"num_tokens": 327135410.0,
|
| 500 |
+
"step": 42000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.2119463280911846,
|
| 504 |
+
"eval_entropy": 1.2579897131426219,
|
| 505 |
+
"eval_loss": 0.7692885398864746,
|
| 506 |
+
"eval_mean_token_accuracy": 0.9031982754033042,
|
| 507 |
+
"eval_num_gold": 908,
|
| 508 |
+
"eval_num_guess": 908,
|
| 509 |
+
"eval_num_tokens": 327135410.0,
|
| 510 |
+
"eval_recall": 0.576,
|
| 511 |
+
"eval_runtime": 17.2218,
|
| 512 |
+
"eval_samples_per_second": 52.724,
|
| 513 |
+
"eval_steps_per_second": 52.724,
|
| 514 |
+
"step": 42000
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"entropy": 1.1411213338077069,
|
| 518 |
+
"epoch": 1.2696580580002885,
|
| 519 |
+
"grad_norm": 3.578125,
|
| 520 |
+
"learning_rate": 1.7838883049402058e-05,
|
| 521 |
+
"loss": 0.0944,
|
| 522 |
+
"mean_token_accuracy": 0.9715765230953693,
|
| 523 |
+
"num_tokens": 342774177.0,
|
| 524 |
+
"step": 44000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 1.2696580580002885,
|
| 528 |
+
"eval_entropy": 1.253242742319464,
|
| 529 |
+
"eval_loss": 0.7519774436950684,
|
| 530 |
+
"eval_mean_token_accuracy": 0.9045354708175827,
|
| 531 |
+
"eval_num_gold": 908,
|
| 532 |
+
"eval_num_guess": 908,
|
| 533 |
+
"eval_num_tokens": 342774177.0,
|
| 534 |
+
"eval_recall": 0.576,
|
| 535 |
+
"eval_runtime": 17.1793,
|
| 536 |
+
"eval_samples_per_second": 52.854,
|
| 537 |
+
"eval_steps_per_second": 52.854,
|
| 538 |
+
"step": 44000
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"entropy": 1.1595853001475334,
|
| 542 |
+
"epoch": 1.3273697879093926,
|
| 543 |
+
"grad_norm": 5.5,
|
| 544 |
+
"learning_rate": 1.724391646669179e-05,
|
| 545 |
+
"loss": 0.0963,
|
| 546 |
+
"mean_token_accuracy": 0.9711391851603984,
|
| 547 |
+
"num_tokens": 358312922.0,
|
| 548 |
+
"step": 46000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.3273697879093926,
|
| 552 |
+
"eval_entropy": 1.2453804171689282,
|
| 553 |
+
"eval_loss": 0.7676454186439514,
|
| 554 |
+
"eval_mean_token_accuracy": 0.9064169454876547,
|
| 555 |
+
"eval_num_gold": 908,
|
| 556 |
+
"eval_num_guess": 908,
|
| 557 |
+
"eval_num_tokens": 358312922.0,
|
| 558 |
+
"eval_recall": 0.5837,
|
| 559 |
+
"eval_runtime": 16.8037,
|
| 560 |
+
"eval_samples_per_second": 54.036,
|
| 561 |
+
"eval_steps_per_second": 54.036,
|
| 562 |
+
"step": 46000
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"entropy": 1.1609133576154709,
|
| 566 |
+
"epoch": 1.3850815178184965,
|
| 567 |
+
"grad_norm": 4.03125,
|
| 568 |
+
"learning_rate": 1.6648949883981517e-05,
|
| 569 |
+
"loss": 0.0922,
|
| 570 |
+
"mean_token_accuracy": 0.9723608312606812,
|
| 571 |
+
"num_tokens": 373752333.0,
|
| 572 |
+
"step": 48000
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 1.3850815178184965,
|
| 576 |
+
"eval_entropy": 1.2345776492934921,
|
| 577 |
+
"eval_loss": 0.7665285468101501,
|
| 578 |
+
"eval_mean_token_accuracy": 0.9063460667687365,
|
| 579 |
+
"eval_num_gold": 908,
|
| 580 |
+
"eval_num_guess": 908,
|
| 581 |
+
"eval_num_tokens": 373752333.0,
|
| 582 |
+
"eval_recall": 0.5793,
|
| 583 |
+
"eval_runtime": 16.7578,
|
| 584 |
+
"eval_samples_per_second": 54.184,
|
| 585 |
+
"eval_steps_per_second": 54.184,
|
| 586 |
+
"step": 48000
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"entropy": 1.1655547478497028,
|
| 590 |
+
"epoch": 1.4427932477276006,
|
| 591 |
+
"grad_norm": 5.59375,
|
| 592 |
+
"learning_rate": 1.6053983301271245e-05,
|
| 593 |
+
"loss": 0.094,
|
| 594 |
+
"mean_token_accuracy": 0.9717481000125409,
|
| 595 |
+
"num_tokens": 389447345.0,
|
| 596 |
+
"step": 50000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 1.4427932477276006,
|
| 600 |
+
"eval_entropy": 1.2292915042407713,
|
| 601 |
+
"eval_loss": 0.7735024094581604,
|
| 602 |
+
"eval_mean_token_accuracy": 0.907910385517822,
|
| 603 |
+
"eval_num_gold": 908,
|
| 604 |
+
"eval_num_guess": 908,
|
| 605 |
+
"eval_num_tokens": 389447345.0,
|
| 606 |
+
"eval_recall": 0.5881,
|
| 607 |
+
"eval_runtime": 17.3697,
|
| 608 |
+
"eval_samples_per_second": 52.275,
|
| 609 |
+
"eval_steps_per_second": 52.275,
|
| 610 |
+
"step": 50000
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"entropy": 1.1552352701127528,
|
| 614 |
+
"epoch": 1.5005049776367048,
|
| 615 |
+
"grad_norm": 3.0,
|
| 616 |
+
"learning_rate": 1.5459016718560976e-05,
|
| 617 |
+
"loss": 0.091,
|
| 618 |
+
"mean_token_accuracy": 0.9726284679472447,
|
| 619 |
+
"num_tokens": 404935652.0,
|
| 620 |
+
"step": 52000
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.5005049776367048,
|
| 624 |
+
"eval_entropy": 1.2490241264325406,
|
| 625 |
+
"eval_loss": 0.7779573202133179,
|
| 626 |
+
"eval_mean_token_accuracy": 0.9046718338053132,
|
| 627 |
+
"eval_num_gold": 908,
|
| 628 |
+
"eval_num_guess": 908,
|
| 629 |
+
"eval_num_tokens": 404935652.0,
|
| 630 |
+
"eval_recall": 0.5848,
|
| 631 |
+
"eval_runtime": 17.2173,
|
| 632 |
+
"eval_samples_per_second": 52.738,
|
| 633 |
+
"eval_steps_per_second": 52.738,
|
| 634 |
+
"step": 52000
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"entropy": 1.1580015743076801,
|
| 638 |
+
"epoch": 1.5582167075458087,
|
| 639 |
+
"grad_norm": 0.0019989013671875,
|
| 640 |
+
"learning_rate": 1.4864050135850704e-05,
|
| 641 |
+
"loss": 0.0674,
|
| 642 |
+
"mean_token_accuracy": 0.979576114565134,
|
| 643 |
+
"num_tokens": 15533221.0,
|
| 644 |
+
"step": 54000
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 1.5582167075458087,
|
| 648 |
+
"eval_entropy": 1.204221866138706,
|
| 649 |
+
"eval_loss": 0.8085830211639404,
|
| 650 |
+
"eval_mean_token_accuracy": 0.9048162211668124,
|
| 651 |
+
"eval_num_gold": 908,
|
| 652 |
+
"eval_num_guess": 908,
|
| 653 |
+
"eval_num_tokens": 15533221.0,
|
| 654 |
+
"eval_recall": 0.5738,
|
| 655 |
+
"eval_runtime": 17.1435,
|
| 656 |
+
"eval_samples_per_second": 52.965,
|
| 657 |
+
"eval_steps_per_second": 52.965,
|
| 658 |
+
"step": 54000
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"entropy": 1.144241349697113,
|
| 662 |
+
"epoch": 1.6159284374549125,
|
| 663 |
+
"grad_norm": 3.765625,
|
| 664 |
+
"learning_rate": 1.4269083553140432e-05,
|
| 665 |
+
"loss": 0.0633,
|
| 666 |
+
"mean_token_accuracy": 0.9807874869704246,
|
| 667 |
+
"num_tokens": 31150685.0,
|
| 668 |
+
"step": 56000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.6159284374549125,
|
| 672 |
+
"eval_entropy": 1.2577752770306256,
|
| 673 |
+
"eval_loss": 0.8108322024345398,
|
| 674 |
+
"eval_mean_token_accuracy": 0.904205797002179,
|
| 675 |
+
"eval_num_gold": 908,
|
| 676 |
+
"eval_num_guess": 908,
|
| 677 |
+
"eval_num_tokens": 31150685.0,
|
| 678 |
+
"eval_recall": 0.5738,
|
| 679 |
+
"eval_runtime": 16.4274,
|
| 680 |
+
"eval_samples_per_second": 55.274,
|
| 681 |
+
"eval_steps_per_second": 55.274,
|
| 682 |
+
"step": 56000
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"entropy": 1.162702257514,
|
| 686 |
+
"epoch": 1.6736401673640167,
|
| 687 |
+
"grad_norm": 2.8125,
|
| 688 |
+
"learning_rate": 1.3674116970430161e-05,
|
| 689 |
+
"loss": 0.0665,
|
| 690 |
+
"mean_token_accuracy": 0.9797295650243759,
|
| 691 |
+
"num_tokens": 46832332.0,
|
| 692 |
+
"step": 58000
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 1.6736401673640167,
|
| 696 |
+
"eval_entropy": 1.268515376989537,
|
| 697 |
+
"eval_loss": 0.814584493637085,
|
| 698 |
+
"eval_mean_token_accuracy": 0.904229478295154,
|
| 699 |
+
"eval_num_gold": 908,
|
| 700 |
+
"eval_num_guess": 908,
|
| 701 |
+
"eval_num_tokens": 46832332.0,
|
| 702 |
+
"eval_recall": 0.5716,
|
| 703 |
+
"eval_runtime": 16.4853,
|
| 704 |
+
"eval_samples_per_second": 55.079,
|
| 705 |
+
"eval_steps_per_second": 55.079,
|
| 706 |
+
"step": 58000
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"entropy": 1.1661596206724645,
|
| 710 |
+
"epoch": 1.7313518972731208,
|
| 711 |
+
"grad_norm": 5.96875,
|
| 712 |
+
"learning_rate": 1.3079150387719889e-05,
|
| 713 |
+
"loss": 0.0672,
|
| 714 |
+
"mean_token_accuracy": 0.9796462517380714,
|
| 715 |
+
"num_tokens": 62558817.0,
|
| 716 |
+
"step": 60000
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.7313518972731208,
|
| 720 |
+
"eval_entropy": 1.2622852631996382,
|
| 721 |
+
"eval_loss": 0.8227198123931885,
|
| 722 |
+
"eval_mean_token_accuracy": 0.9038923141846048,
|
| 723 |
+
"eval_num_gold": 908,
|
| 724 |
+
"eval_num_guess": 908,
|
| 725 |
+
"eval_num_tokens": 62558817.0,
|
| 726 |
+
"eval_recall": 0.5716,
|
| 727 |
+
"eval_runtime": 16.9948,
|
| 728 |
+
"eval_samples_per_second": 53.428,
|
| 729 |
+
"eval_steps_per_second": 53.428,
|
| 730 |
+
"step": 60000
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"entropy": 1.1797457176148891,
|
| 734 |
+
"epoch": 1.789063627182225,
|
| 735 |
+
"grad_norm": 0.5546875,
|
| 736 |
+
"learning_rate": 1.2484183805009618e-05,
|
| 737 |
+
"loss": 0.0657,
|
| 738 |
+
"mean_token_accuracy": 0.980204150468111,
|
| 739 |
+
"num_tokens": 78074806.0,
|
| 740 |
+
"step": 62000
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 1.789063627182225,
|
| 744 |
+
"eval_entropy": 1.2418163208052975,
|
| 745 |
+
"eval_loss": 0.8185028433799744,
|
| 746 |
+
"eval_mean_token_accuracy": 0.9041991046740621,
|
| 747 |
+
"eval_num_gold": 908,
|
| 748 |
+
"eval_num_guess": 908,
|
| 749 |
+
"eval_num_tokens": 78074806.0,
|
| 750 |
+
"eval_recall": 0.5705,
|
| 751 |
+
"eval_runtime": 17.1144,
|
| 752 |
+
"eval_samples_per_second": 53.055,
|
| 753 |
+
"eval_steps_per_second": 53.055,
|
| 754 |
+
"step": 62000
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"entropy": 1.1774089051187038,
|
| 758 |
+
"epoch": 1.8467753570913288,
|
| 759 |
+
"grad_norm": 5.5625,
|
| 760 |
+
"learning_rate": 1.1889217222299348e-05,
|
| 761 |
+
"loss": 0.0649,
|
| 762 |
+
"mean_token_accuracy": 0.9804997465908527,
|
| 763 |
+
"num_tokens": 93602629.0,
|
| 764 |
+
"step": 64000
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 1.8467753570913288,
|
| 768 |
+
"eval_entropy": 1.2988805646807087,
|
| 769 |
+
"eval_loss": 0.8260899782180786,
|
| 770 |
+
"eval_mean_token_accuracy": 0.9030656689523601,
|
| 771 |
+
"eval_num_gold": 908,
|
| 772 |
+
"eval_num_guess": 908,
|
| 773 |
+
"eval_num_tokens": 93602629.0,
|
| 774 |
+
"eval_recall": 0.576,
|
| 775 |
+
"eval_runtime": 16.1643,
|
| 776 |
+
"eval_samples_per_second": 56.173,
|
| 777 |
+
"eval_steps_per_second": 56.173,
|
| 778 |
+
"step": 64000
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"entropy": 1.174987347126007,
|
| 782 |
+
"epoch": 1.9044870870004327,
|
| 783 |
+
"grad_norm": 5.53125,
|
| 784 |
+
"learning_rate": 1.1294250639589077e-05,
|
| 785 |
+
"loss": 0.064,
|
| 786 |
+
"mean_token_accuracy": 0.9806980607807636,
|
| 787 |
+
"num_tokens": 109249414.0,
|
| 788 |
+
"step": 66000
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.9044870870004327,
|
| 792 |
+
"eval_entropy": 1.2433809736489199,
|
| 793 |
+
"eval_loss": 0.8272661566734314,
|
| 794 |
+
"eval_mean_token_accuracy": 0.9028221254569319,
|
| 795 |
+
"eval_num_gold": 908,
|
| 796 |
+
"eval_num_guess": 908,
|
| 797 |
+
"eval_num_tokens": 109249414.0,
|
| 798 |
+
"eval_recall": 0.5727,
|
| 799 |
+
"eval_runtime": 16.3988,
|
| 800 |
+
"eval_samples_per_second": 55.37,
|
| 801 |
+
"eval_steps_per_second": 55.37,
|
| 802 |
+
"step": 66000
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"entropy": 1.1633582679629326,
|
| 806 |
+
"epoch": 1.9621988169095368,
|
| 807 |
+
"grad_norm": 5.34375,
|
| 808 |
+
"learning_rate": 1.0699284056878807e-05,
|
| 809 |
+
"loss": 0.0643,
|
| 810 |
+
"mean_token_accuracy": 0.9805754337012768,
|
| 811 |
+
"num_tokens": 124880720.0,
|
| 812 |
+
"step": 68000
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 1.9621988169095368,
|
| 816 |
+
"eval_entropy": 1.224490842367584,
|
| 817 |
+
"eval_loss": 0.8288715481758118,
|
| 818 |
+
"eval_mean_token_accuracy": 0.9034351931991557,
|
| 819 |
+
"eval_num_gold": 908,
|
| 820 |
+
"eval_num_guess": 908,
|
| 821 |
+
"eval_num_tokens": 124880720.0,
|
| 822 |
+
"eval_recall": 0.5738,
|
| 823 |
+
"eval_runtime": 16.4997,
|
| 824 |
+
"eval_samples_per_second": 55.031,
|
| 825 |
+
"eval_steps_per_second": 55.031,
|
| 826 |
+
"step": 68000
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"entropy": 1.1513627296090125,
|
| 830 |
+
"epoch": 2.019910546818641,
|
| 831 |
+
"grad_norm": 0.79296875,
|
| 832 |
+
"learning_rate": 1.0104317474168535e-05,
|
| 833 |
+
"loss": 0.0633,
|
| 834 |
+
"mean_token_accuracy": 0.9811660476624966,
|
| 835 |
+
"num_tokens": 140499220.0,
|
| 836 |
+
"step": 70000
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 2.019910546818641,
|
| 840 |
+
"eval_entropy": 1.2267822175561593,
|
| 841 |
+
"eval_loss": 0.8458257913589478,
|
| 842 |
+
"eval_mean_token_accuracy": 0.9038964834572986,
|
| 843 |
+
"eval_num_gold": 908,
|
| 844 |
+
"eval_num_guess": 908,
|
| 845 |
+
"eval_num_tokens": 140499220.0,
|
| 846 |
+
"eval_recall": 0.5705,
|
| 847 |
+
"eval_runtime": 16.4967,
|
| 848 |
+
"eval_samples_per_second": 55.041,
|
| 849 |
+
"eval_steps_per_second": 55.041,
|
| 850 |
+
"step": 70000
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"entropy": 1.143776093840599,
|
| 854 |
+
"epoch": 2.077622276727745,
|
| 855 |
+
"grad_norm": 7.9375,
|
| 856 |
+
"learning_rate": 9.509350891458264e-06,
|
| 857 |
+
"loss": 0.0597,
|
| 858 |
+
"mean_token_accuracy": 0.9825106913745403,
|
| 859 |
+
"num_tokens": 156048918.0,
|
| 860 |
+
"step": 72000
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 2.077622276727745,
|
| 864 |
+
"eval_entropy": 1.2148328015195116,
|
| 865 |
+
"eval_loss": 0.8337165713310242,
|
| 866 |
+
"eval_mean_token_accuracy": 0.9035390550475814,
|
| 867 |
+
"eval_num_gold": 908,
|
| 868 |
+
"eval_num_guess": 908,
|
| 869 |
+
"eval_num_tokens": 156048918.0,
|
| 870 |
+
"eval_recall": 0.5705,
|
| 871 |
+
"eval_runtime": 16.506,
|
| 872 |
+
"eval_samples_per_second": 55.01,
|
| 873 |
+
"eval_steps_per_second": 55.01,
|
| 874 |
+
"step": 72000
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"entropy": 1.1460822140574456,
|
| 878 |
+
"epoch": 2.135334006636849,
|
| 879 |
+
"grad_norm": 12.375,
|
| 880 |
+
"learning_rate": 8.914384308747992e-06,
|
| 881 |
+
"loss": 0.0596,
|
| 882 |
+
"mean_token_accuracy": 0.98244061678648,
|
| 883 |
+
"num_tokens": 171653895.0,
|
| 884 |
+
"step": 74000
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 2.135334006636849,
|
| 888 |
+
"eval_entropy": 1.2635613490175046,
|
| 889 |
+
"eval_loss": 0.8348618745803833,
|
| 890 |
+
"eval_mean_token_accuracy": 0.9038379774285308,
|
| 891 |
+
"eval_num_gold": 908,
|
| 892 |
+
"eval_num_guess": 908,
|
| 893 |
+
"eval_num_tokens": 171653895.0,
|
| 894 |
+
"eval_recall": 0.5694,
|
| 895 |
+
"eval_runtime": 16.4822,
|
| 896 |
+
"eval_samples_per_second": 55.09,
|
| 897 |
+
"eval_steps_per_second": 55.09,
|
| 898 |
+
"step": 74000
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"entropy": 1.1560133908391,
|
| 902 |
+
"epoch": 2.193045736545953,
|
| 903 |
+
"grad_norm": 7.625,
|
| 904 |
+
"learning_rate": 8.319417726037721e-06,
|
| 905 |
+
"loss": 0.06,
|
| 906 |
+
"mean_token_accuracy": 0.9822552761137485,
|
| 907 |
+
"num_tokens": 187228261.0,
|
| 908 |
+
"step": 76000
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 2.193045736545953,
|
| 912 |
+
"eval_entropy": 1.2220293277554575,
|
| 913 |
+
"eval_loss": 0.8315507769584656,
|
| 914 |
+
"eval_mean_token_accuracy": 0.9036543207809263,
|
| 915 |
+
"eval_num_gold": 908,
|
| 916 |
+
"eval_num_guess": 908,
|
| 917 |
+
"eval_num_tokens": 187228261.0,
|
| 918 |
+
"eval_recall": 0.5716,
|
| 919 |
+
"eval_runtime": 16.703,
|
| 920 |
+
"eval_samples_per_second": 54.361,
|
| 921 |
+
"eval_steps_per_second": 54.361,
|
| 922 |
+
"step": 76000
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"entropy": 1.1676500248610973,
|
| 926 |
+
"epoch": 2.250757466455057,
|
| 927 |
+
"grad_norm": 4.84375,
|
| 928 |
+
"learning_rate": 7.72445114332745e-06,
|
| 929 |
+
"loss": 0.0611,
|
| 930 |
+
"mean_token_accuracy": 0.9819406977891922,
|
| 931 |
+
"num_tokens": 202699683.0,
|
| 932 |
+
"step": 78000
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 2.250757466455057,
|
| 936 |
+
"eval_entropy": 1.2448954319638827,
|
| 937 |
+
"eval_loss": 0.8309385776519775,
|
| 938 |
+
"eval_mean_token_accuracy": 0.9030922418255113,
|
| 939 |
+
"eval_num_gold": 908,
|
| 940 |
+
"eval_num_guess": 908,
|
| 941 |
+
"eval_num_tokens": 202699683.0,
|
| 942 |
+
"eval_recall": 0.5694,
|
| 943 |
+
"eval_runtime": 16.603,
|
| 944 |
+
"eval_samples_per_second": 54.689,
|
| 945 |
+
"eval_steps_per_second": 54.689,
|
| 946 |
+
"step": 78000
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"entropy": 1.1656713368594647,
|
| 950 |
+
"epoch": 2.308469196364161,
|
| 951 |
+
"grad_norm": 6.53125,
|
| 952 |
+
"learning_rate": 7.129484560617179e-06,
|
| 953 |
+
"loss": 0.0618,
|
| 954 |
+
"mean_token_accuracy": 0.9817487963140011,
|
| 955 |
+
"num_tokens": 218284466.0,
|
| 956 |
+
"step": 80000
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 2.308469196364161,
|
| 960 |
+
"eval_entropy": 1.255102663181952,
|
| 961 |
+
"eval_loss": 0.8435425162315369,
|
| 962 |
+
"eval_mean_token_accuracy": 0.902260869642974,
|
| 963 |
+
"eval_num_gold": 908,
|
| 964 |
+
"eval_num_guess": 908,
|
| 965 |
+
"eval_num_tokens": 218284466.0,
|
| 966 |
+
"eval_recall": 0.5661,
|
| 967 |
+
"eval_runtime": 16.8256,
|
| 968 |
+
"eval_samples_per_second": 53.965,
|
| 969 |
+
"eval_steps_per_second": 53.965,
|
| 970 |
+
"step": 80000
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"entropy": 1.1597592905461789,
|
| 974 |
+
"epoch": 2.366180926273265,
|
| 975 |
+
"grad_norm": 2.5,
|
| 976 |
+
"learning_rate": 6.534517977906908e-06,
|
| 977 |
+
"loss": 0.0602,
|
| 978 |
+
"mean_token_accuracy": 0.9821576415896416,
|
| 979 |
+
"num_tokens": 233928452.0,
|
| 980 |
+
"step": 82000
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 2.366180926273265,
|
| 984 |
+
"eval_entropy": 1.2422783964924875,
|
| 985 |
+
"eval_loss": 0.8390738368034363,
|
| 986 |
+
"eval_mean_token_accuracy": 0.9032785006950605,
|
| 987 |
+
"eval_num_gold": 908,
|
| 988 |
+
"eval_num_guess": 908,
|
| 989 |
+
"eval_num_tokens": 233928452.0,
|
| 990 |
+
"eval_recall": 0.5683,
|
| 991 |
+
"eval_runtime": 16.7577,
|
| 992 |
+
"eval_samples_per_second": 54.184,
|
| 993 |
+
"eval_steps_per_second": 54.184,
|
| 994 |
+
"step": 82000
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"entropy": 1.17008468157053,
|
| 998 |
+
"epoch": 2.4238926561823693,
|
| 999 |
+
"grad_norm": 0.0400390625,
|
| 1000 |
+
"learning_rate": 5.939551395196637e-06,
|
| 1001 |
+
"loss": 0.0591,
|
| 1002 |
+
"mean_token_accuracy": 0.9825585896968841,
|
| 1003 |
+
"num_tokens": 249419664.0,
|
| 1004 |
+
"step": 84000
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 2.4238926561823693,
|
| 1008 |
+
"eval_entropy": 1.2469606770424067,
|
| 1009 |
+
"eval_loss": 0.8383654356002808,
|
| 1010 |
+
"eval_mean_token_accuracy": 0.9040639832418921,
|
| 1011 |
+
"eval_num_gold": 908,
|
| 1012 |
+
"eval_num_guess": 908,
|
| 1013 |
+
"eval_num_tokens": 249419664.0,
|
| 1014 |
+
"eval_recall": 0.5705,
|
| 1015 |
+
"eval_runtime": 16.6392,
|
| 1016 |
+
"eval_samples_per_second": 54.57,
|
| 1017 |
+
"eval_steps_per_second": 54.57,
|
| 1018 |
+
"step": 84000
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"entropy": 1.163529093414545,
|
| 1022 |
+
"epoch": 2.481604386091473,
|
| 1023 |
+
"grad_norm": 6.28125,
|
| 1024 |
+
"learning_rate": 5.3445848124863655e-06,
|
| 1025 |
+
"loss": 0.0568,
|
| 1026 |
+
"mean_token_accuracy": 0.9832313210368157,
|
| 1027 |
+
"num_tokens": 264982654.0,
|
| 1028 |
+
"step": 86000
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 2.481604386091473,
|
| 1032 |
+
"eval_entropy": 1.236849331908289,
|
| 1033 |
+
"eval_loss": 0.8381890058517456,
|
| 1034 |
+
"eval_mean_token_accuracy": 0.9027883698630438,
|
| 1035 |
+
"eval_num_gold": 908,
|
| 1036 |
+
"eval_num_guess": 908,
|
| 1037 |
+
"eval_num_tokens": 264982654.0,
|
| 1038 |
+
"eval_recall": 0.5672,
|
| 1039 |
+
"eval_runtime": 16.5964,
|
| 1040 |
+
"eval_samples_per_second": 54.711,
|
| 1041 |
+
"eval_steps_per_second": 54.711,
|
| 1042 |
+
"step": 86000
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"entropy": 1.1701532056927682,
|
| 1046 |
+
"epoch": 2.539316116000577,
|
| 1047 |
+
"grad_norm": 0.87109375,
|
| 1048 |
+
"learning_rate": 4.749618229776094e-06,
|
| 1049 |
+
"loss": 0.0574,
|
| 1050 |
+
"mean_token_accuracy": 0.9830155865848065,
|
| 1051 |
+
"num_tokens": 280520807.0,
|
| 1052 |
+
"step": 88000
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 2.539316116000577,
|
| 1056 |
+
"eval_entropy": 1.2492524392673097,
|
| 1057 |
+
"eval_loss": 0.839518666267395,
|
| 1058 |
+
"eval_mean_token_accuracy": 0.9025986767681685,
|
| 1059 |
+
"eval_num_gold": 908,
|
| 1060 |
+
"eval_num_guess": 908,
|
| 1061 |
+
"eval_num_tokens": 280520807.0,
|
| 1062 |
+
"eval_recall": 0.5661,
|
| 1063 |
+
"eval_runtime": 16.5519,
|
| 1064 |
+
"eval_samples_per_second": 54.858,
|
| 1065 |
+
"eval_steps_per_second": 54.858,
|
| 1066 |
+
"step": 88000
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"entropy": 1.167941878914833,
|
| 1070 |
+
"epoch": 2.597027845909681,
|
| 1071 |
+
"grad_norm": 0.451171875,
|
| 1072 |
+
"learning_rate": 4.154651647065824e-06,
|
| 1073 |
+
"loss": 0.0602,
|
| 1074 |
+
"mean_token_accuracy": 0.9820802296400071,
|
| 1075 |
+
"num_tokens": 296146535.0,
|
| 1076 |
+
"step": 90000
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 2.597027845909681,
|
| 1080 |
+
"eval_entropy": 1.2443812186234848,
|
| 1081 |
+
"eval_loss": 0.8395401835441589,
|
| 1082 |
+
"eval_mean_token_accuracy": 0.9034286766981764,
|
| 1083 |
+
"eval_num_gold": 908,
|
| 1084 |
+
"eval_num_guess": 908,
|
| 1085 |
+
"eval_num_tokens": 296146535.0,
|
| 1086 |
+
"eval_recall": 0.5683,
|
| 1087 |
+
"eval_runtime": 16.788,
|
| 1088 |
+
"eval_samples_per_second": 54.086,
|
| 1089 |
+
"eval_steps_per_second": 54.086,
|
| 1090 |
+
"step": 90000
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"entropy": 1.1601335457861424,
|
| 1094 |
+
"epoch": 2.6547395758187853,
|
| 1095 |
+
"grad_norm": 4.6875,
|
| 1096 |
+
"learning_rate": 3.559685064355552e-06,
|
| 1097 |
+
"loss": 0.0584,
|
| 1098 |
+
"mean_token_accuracy": 0.9827592859268188,
|
| 1099 |
+
"num_tokens": 311778551.0,
|
| 1100 |
+
"step": 92000
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 2.6547395758187853,
|
| 1104 |
+
"eval_entropy": 1.2437387075324415,
|
| 1105 |
+
"eval_loss": 0.836577296257019,
|
| 1106 |
+
"eval_mean_token_accuracy": 0.9039776291419231,
|
| 1107 |
+
"eval_num_gold": 908,
|
| 1108 |
+
"eval_num_guess": 908,
|
| 1109 |
+
"eval_num_tokens": 311778551.0,
|
| 1110 |
+
"eval_recall": 0.5705,
|
| 1111 |
+
"eval_runtime": 16.2272,
|
| 1112 |
+
"eval_samples_per_second": 55.956,
|
| 1113 |
+
"eval_steps_per_second": 55.956,
|
| 1114 |
+
"step": 92000
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"entropy": 1.1733056641221047,
|
| 1118 |
+
"epoch": 2.712451305727889,
|
| 1119 |
+
"grad_norm": 5.03125,
|
| 1120 |
+
"learning_rate": 2.964718481645281e-06,
|
| 1121 |
+
"loss": 0.0564,
|
| 1122 |
+
"mean_token_accuracy": 0.9832823853492737,
|
| 1123 |
+
"num_tokens": 327170479.0,
|
| 1124 |
+
"step": 94000
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 2.712451305727889,
|
| 1128 |
+
"eval_entropy": 1.2440849004337966,
|
| 1129 |
+
"eval_loss": 0.8399211168289185,
|
| 1130 |
+
"eval_mean_token_accuracy": 0.9033104040155326,
|
| 1131 |
+
"eval_num_gold": 908,
|
| 1132 |
+
"eval_num_guess": 908,
|
| 1133 |
+
"eval_num_tokens": 327170479.0,
|
| 1134 |
+
"eval_recall": 0.5683,
|
| 1135 |
+
"eval_runtime": 16.2233,
|
| 1136 |
+
"eval_samples_per_second": 55.969,
|
| 1137 |
+
"eval_steps_per_second": 55.969,
|
| 1138 |
+
"step": 94000
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"entropy": 1.1586334483027458,
|
| 1142 |
+
"epoch": 2.770163035636993,
|
| 1143 |
+
"grad_norm": 3.953125,
|
| 1144 |
+
"learning_rate": 2.36975189893501e-06,
|
| 1145 |
+
"loss": 0.0585,
|
| 1146 |
+
"mean_token_accuracy": 0.9826480825543403,
|
| 1147 |
+
"num_tokens": 342791353.0,
|
| 1148 |
+
"step": 96000
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 2.770163035636993,
|
| 1152 |
+
"eval_entropy": 1.2412338042180444,
|
| 1153 |
+
"eval_loss": 0.8378188610076904,
|
| 1154 |
+
"eval_mean_token_accuracy": 0.9035194405572005,
|
| 1155 |
+
"eval_num_gold": 908,
|
| 1156 |
+
"eval_num_guess": 908,
|
| 1157 |
+
"eval_num_tokens": 342791353.0,
|
| 1158 |
+
"eval_recall": 0.5672,
|
| 1159 |
+
"eval_runtime": 16.2077,
|
| 1160 |
+
"eval_samples_per_second": 56.023,
|
| 1161 |
+
"eval_steps_per_second": 56.023,
|
| 1162 |
+
"step": 96000
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"entropy": 1.1629991734027862,
|
| 1166 |
+
"epoch": 2.827874765546097,
|
| 1167 |
+
"grad_norm": 6.4375,
|
| 1168 |
+
"learning_rate": 1.7747853162247388e-06,
|
| 1169 |
+
"loss": 0.0608,
|
| 1170 |
+
"mean_token_accuracy": 0.9821404512822628,
|
| 1171 |
+
"num_tokens": 358436354.0,
|
| 1172 |
+
"step": 98000
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 2.827874765546097,
|
| 1176 |
+
"eval_entropy": 1.2435034370369848,
|
| 1177 |
+
"eval_loss": 0.8380420207977295,
|
| 1178 |
+
"eval_mean_token_accuracy": 0.9037704004327631,
|
| 1179 |
+
"eval_num_gold": 908,
|
| 1180 |
+
"eval_num_guess": 908,
|
| 1181 |
+
"eval_num_tokens": 358436354.0,
|
| 1182 |
+
"eval_recall": 0.5694,
|
| 1183 |
+
"eval_runtime": 16.2198,
|
| 1184 |
+
"eval_samples_per_second": 55.981,
|
| 1185 |
+
"eval_steps_per_second": 55.981,
|
| 1186 |
+
"step": 98000
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"entropy": 1.1640874392092229,
|
| 1190 |
+
"epoch": 2.8855864954552013,
|
| 1191 |
+
"grad_norm": 0.8203125,
|
| 1192 |
+
"learning_rate": 1.1798187335144677e-06,
|
| 1193 |
+
"loss": 0.0574,
|
| 1194 |
+
"mean_token_accuracy": 0.9829988768994808,
|
| 1195 |
+
"num_tokens": 374029027.0,
|
| 1196 |
+
"step": 100000
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 2.8855864954552013,
|
| 1200 |
+
"eval_entropy": 1.2438825091207606,
|
| 1201 |
+
"eval_loss": 0.8370459079742432,
|
| 1202 |
+
"eval_mean_token_accuracy": 0.9030486140810445,
|
| 1203 |
+
"eval_num_gold": 908,
|
| 1204 |
+
"eval_num_guess": 908,
|
| 1205 |
+
"eval_num_tokens": 374029027.0,
|
| 1206 |
+
"eval_recall": 0.5661,
|
| 1207 |
+
"eval_runtime": 17.3712,
|
| 1208 |
+
"eval_samples_per_second": 52.271,
|
| 1209 |
+
"eval_steps_per_second": 52.271,
|
| 1210 |
+
"step": 100000
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"entropy": 1.1683570961356162,
|
| 1214 |
+
"epoch": 2.9432982253643054,
|
| 1215 |
+
"grad_norm": 6.65625,
|
| 1216 |
+
"learning_rate": 5.848521508041964e-07,
|
| 1217 |
+
"loss": 0.0583,
|
| 1218 |
+
"mean_token_accuracy": 0.982835016399622,
|
| 1219 |
+
"num_tokens": 389554889.0,
|
| 1220 |
+
"step": 102000
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 2.9432982253643054,
|
| 1224 |
+
"eval_entropy": 1.2444104566710636,
|
| 1225 |
+
"eval_loss": 0.8376456499099731,
|
| 1226 |
+
"eval_mean_token_accuracy": 0.9037148623608282,
|
| 1227 |
+
"eval_num_gold": 908,
|
| 1228 |
+
"eval_num_guess": 908,
|
| 1229 |
+
"eval_num_tokens": 389554889.0,
|
| 1230 |
+
"eval_recall": 0.5683,
|
| 1231 |
+
"eval_runtime": 16.2408,
|
| 1232 |
+
"eval_samples_per_second": 55.909,
|
| 1233 |
+
"eval_steps_per_second": 55.909,
|
| 1234 |
+
"step": 102000
|
| 1235 |
+
}
|
| 1236 |
+
],
|
| 1237 |
+
"logging_steps": 2000,
|
| 1238 |
+
"max_steps": 103965,
|
| 1239 |
+
"num_input_tokens_seen": 0,
|
| 1240 |
+
"num_train_epochs": 3,
|
| 1241 |
+
"save_steps": 2000,
|
| 1242 |
+
"stateful_callbacks": {
|
| 1243 |
+
"TrainerControl": {
|
| 1244 |
+
"args": {
|
| 1245 |
+
"should_epoch_stop": false,
|
| 1246 |
+
"should_evaluate": false,
|
| 1247 |
+
"should_log": false,
|
| 1248 |
+
"should_save": true,
|
| 1249 |
+
"should_training_stop": true
|
| 1250 |
+
},
|
| 1251 |
+
"attributes": {}
|
| 1252 |
+
}
|
| 1253 |
+
},
|
| 1254 |
+
"total_flos": 3.6466469785747587e+19,
|
| 1255 |
+
"train_batch_size": 1,
|
| 1256 |
+
"trial_name": null,
|
| 1257 |
+
"trial_params": null
|
| 1258 |
+
}
|