End of training
Browse files- all_results.json +15 -0
- eval_results.json +10 -0
- logs/events.out.tfevents.1686455083.serv-9224.1583679.1 +3 -0
- train_results.json +8 -0
- trainer_state.json +1057 -0
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.66,
|
| 3 |
+
"eval_accuracy": 0.15114683656073724,
|
| 4 |
+
"eval_loss": 6.007903099060059,
|
| 5 |
+
"eval_runtime": 2023.6869,
|
| 6 |
+
"eval_samples": 308356,
|
| 7 |
+
"eval_samples_per_second": 152.373,
|
| 8 |
+
"eval_steps_per_second": 3.175,
|
| 9 |
+
"perplexity": 406.62976334153257,
|
| 10 |
+
"train_loss": 6.2887065917574985,
|
| 11 |
+
"train_runtime": 86523.957,
|
| 12 |
+
"train_samples": 5858758,
|
| 13 |
+
"train_samples_per_second": 338.563,
|
| 14 |
+
"train_steps_per_second": 7.053
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.66,
|
| 3 |
+
"eval_accuracy": 0.15114683656073724,
|
| 4 |
+
"eval_loss": 6.007903099060059,
|
| 5 |
+
"eval_runtime": 2023.6869,
|
| 6 |
+
"eval_samples": 308356,
|
| 7 |
+
"eval_samples_per_second": 152.373,
|
| 8 |
+
"eval_steps_per_second": 3.175,
|
| 9 |
+
"perplexity": 406.62976334153257
|
| 10 |
+
}
|
logs/events.out.tfevents.1686455083.serv-9224.1583679.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b990f45810f466c2c787d43beb83be5016bbd0713c7bb17bac871fa83d8e5b79
|
| 3 |
+
size 369
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.66,
|
| 3 |
+
"train_loss": 6.2887065917574985,
|
| 4 |
+
"train_runtime": 86523.957,
|
| 5 |
+
"train_samples": 5858758,
|
| 6 |
+
"train_samples_per_second": 338.563,
|
| 7 |
+
"train_steps_per_second": 7.053
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1057 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 6.008976459503174,
|
| 3 |
+
"best_model_checkpoint": "bert_base_24/checkpoint-80000",
|
| 4 |
+
"epoch": 0.6554343017254093,
|
| 5 |
+
"global_step": 80001,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.0,
|
| 12 |
+
"learning_rate": 5.000000000000001e-07,
|
| 13 |
+
"loss": 9.9822,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.01,
|
| 18 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 19 |
+
"loss": 9.2026,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.01,
|
| 24 |
+
"learning_rate": 1.5e-06,
|
| 25 |
+
"loss": 8.7769,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.02,
|
| 30 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 31 |
+
"loss": 8.297,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.02,
|
| 36 |
+
"learning_rate": 2.5e-06,
|
| 37 |
+
"loss": 7.7815,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.02,
|
| 42 |
+
"learning_rate": 3e-06,
|
| 43 |
+
"loss": 7.3833,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.03,
|
| 48 |
+
"learning_rate": 3.5e-06,
|
| 49 |
+
"loss": 7.1337,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.03,
|
| 54 |
+
"learning_rate": 4.000000000000001e-06,
|
| 55 |
+
"loss": 7.0008,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.04,
|
| 60 |
+
"learning_rate": 4.5e-06,
|
| 61 |
+
"loss": 6.8954,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.04,
|
| 66 |
+
"learning_rate": 5e-06,
|
| 67 |
+
"loss": 6.8278,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.05,
|
| 72 |
+
"learning_rate": 5.500000000000001e-06,
|
| 73 |
+
"loss": 6.763,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.05,
|
| 78 |
+
"learning_rate": 6e-06,
|
| 79 |
+
"loss": 6.7141,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.05,
|
| 84 |
+
"learning_rate": 6.5000000000000004e-06,
|
| 85 |
+
"loss": 6.6796,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.06,
|
| 90 |
+
"learning_rate": 7e-06,
|
| 91 |
+
"loss": 6.6416,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.06,
|
| 96 |
+
"learning_rate": 7.500000000000001e-06,
|
| 97 |
+
"loss": 6.6118,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.07,
|
| 102 |
+
"learning_rate": 8.000000000000001e-06,
|
| 103 |
+
"loss": 6.5762,
|
| 104 |
+
"step": 8000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.07,
|
| 108 |
+
"learning_rate": 8.5e-06,
|
| 109 |
+
"loss": 6.5481,
|
| 110 |
+
"step": 8500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.07,
|
| 114 |
+
"learning_rate": 9e-06,
|
| 115 |
+
"loss": 6.5293,
|
| 116 |
+
"step": 9000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.08,
|
| 120 |
+
"learning_rate": 9.5e-06,
|
| 121 |
+
"loss": 6.506,
|
| 122 |
+
"step": 9500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.08,
|
| 126 |
+
"learning_rate": 1e-05,
|
| 127 |
+
"loss": 6.4917,
|
| 128 |
+
"step": 10000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.08,
|
| 132 |
+
"eval_accuracy": 0.1406164011008725,
|
| 133 |
+
"eval_loss": 6.442233562469482,
|
| 134 |
+
"eval_runtime": 2026.7135,
|
| 135 |
+
"eval_samples_per_second": 152.146,
|
| 136 |
+
"eval_steps_per_second": 3.17,
|
| 137 |
+
"step": 10000
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.09,
|
| 141 |
+
"learning_rate": 9.991670692498626e-06,
|
| 142 |
+
"loss": 6.4711,
|
| 143 |
+
"step": 10500
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.09,
|
| 147 |
+
"learning_rate": 9.983341384997253e-06,
|
| 148 |
+
"loss": 6.4602,
|
| 149 |
+
"step": 11000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.09,
|
| 153 |
+
"learning_rate": 9.975012077495878e-06,
|
| 154 |
+
"loss": 6.4399,
|
| 155 |
+
"step": 11500
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.1,
|
| 159 |
+
"learning_rate": 9.966682769994504e-06,
|
| 160 |
+
"loss": 6.428,
|
| 161 |
+
"step": 12000
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.1,
|
| 165 |
+
"learning_rate": 9.95835346249313e-06,
|
| 166 |
+
"loss": 6.4067,
|
| 167 |
+
"step": 12500
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 0.11,
|
| 171 |
+
"learning_rate": 9.950024154991756e-06,
|
| 172 |
+
"loss": 6.3975,
|
| 173 |
+
"step": 13000
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 0.11,
|
| 177 |
+
"learning_rate": 9.941694847490381e-06,
|
| 178 |
+
"loss": 6.3889,
|
| 179 |
+
"step": 13500
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.11,
|
| 183 |
+
"learning_rate": 9.933365539989006e-06,
|
| 184 |
+
"loss": 6.3798,
|
| 185 |
+
"step": 14000
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.12,
|
| 189 |
+
"learning_rate": 9.925036232487631e-06,
|
| 190 |
+
"loss": 6.363,
|
| 191 |
+
"step": 14500
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.12,
|
| 195 |
+
"learning_rate": 9.916706924986258e-06,
|
| 196 |
+
"loss": 6.3547,
|
| 197 |
+
"step": 15000
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.13,
|
| 201 |
+
"learning_rate": 9.908377617484883e-06,
|
| 202 |
+
"loss": 6.3391,
|
| 203 |
+
"step": 15500
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 0.13,
|
| 207 |
+
"learning_rate": 9.90004830998351e-06,
|
| 208 |
+
"loss": 6.3392,
|
| 209 |
+
"step": 16000
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.14,
|
| 213 |
+
"learning_rate": 9.891719002482134e-06,
|
| 214 |
+
"loss": 6.3375,
|
| 215 |
+
"step": 16500
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.14,
|
| 219 |
+
"learning_rate": 9.883389694980761e-06,
|
| 220 |
+
"loss": 6.319,
|
| 221 |
+
"step": 17000
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.14,
|
| 225 |
+
"learning_rate": 9.875060387479386e-06,
|
| 226 |
+
"loss": 6.3203,
|
| 227 |
+
"step": 17500
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.15,
|
| 231 |
+
"learning_rate": 9.866731079978011e-06,
|
| 232 |
+
"loss": 6.3119,
|
| 233 |
+
"step": 18000
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.15,
|
| 237 |
+
"learning_rate": 9.858401772476636e-06,
|
| 238 |
+
"loss": 6.311,
|
| 239 |
+
"step": 18500
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.16,
|
| 243 |
+
"learning_rate": 9.850072464975263e-06,
|
| 244 |
+
"loss": 6.296,
|
| 245 |
+
"step": 19000
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 0.16,
|
| 249 |
+
"learning_rate": 9.84174315747389e-06,
|
| 250 |
+
"loss": 6.2894,
|
| 251 |
+
"step": 19500
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.16,
|
| 255 |
+
"learning_rate": 9.833413849972514e-06,
|
| 256 |
+
"loss": 6.2848,
|
| 257 |
+
"step": 20000
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.16,
|
| 261 |
+
"eval_accuracy": 0.14784890360476463,
|
| 262 |
+
"eval_loss": 6.264363765716553,
|
| 263 |
+
"eval_runtime": 2026.7257,
|
| 264 |
+
"eval_samples_per_second": 152.145,
|
| 265 |
+
"eval_steps_per_second": 3.17,
|
| 266 |
+
"step": 20000
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.17,
|
| 270 |
+
"learning_rate": 9.82508454247114e-06,
|
| 271 |
+
"loss": 6.279,
|
| 272 |
+
"step": 20500
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.17,
|
| 276 |
+
"learning_rate": 9.816755234969766e-06,
|
| 277 |
+
"loss": 6.2699,
|
| 278 |
+
"step": 21000
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.18,
|
| 282 |
+
"learning_rate": 9.808425927468391e-06,
|
| 283 |
+
"loss": 6.2756,
|
| 284 |
+
"step": 21500
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.18,
|
| 288 |
+
"learning_rate": 9.800096619967016e-06,
|
| 289 |
+
"loss": 6.2597,
|
| 290 |
+
"step": 22000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.18,
|
| 294 |
+
"learning_rate": 9.791767312465643e-06,
|
| 295 |
+
"loss": 6.2627,
|
| 296 |
+
"step": 22500
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.19,
|
| 300 |
+
"learning_rate": 9.783438004964268e-06,
|
| 301 |
+
"loss": 6.2457,
|
| 302 |
+
"step": 23000
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.19,
|
| 306 |
+
"learning_rate": 9.775108697462895e-06,
|
| 307 |
+
"loss": 6.2424,
|
| 308 |
+
"step": 23500
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.2,
|
| 312 |
+
"learning_rate": 9.76677938996152e-06,
|
| 313 |
+
"loss": 6.2365,
|
| 314 |
+
"step": 24000
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.2,
|
| 318 |
+
"learning_rate": 9.758450082460145e-06,
|
| 319 |
+
"loss": 6.2395,
|
| 320 |
+
"step": 24500
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.2,
|
| 324 |
+
"learning_rate": 9.750120774958771e-06,
|
| 325 |
+
"loss": 6.2347,
|
| 326 |
+
"step": 25000
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.21,
|
| 330 |
+
"learning_rate": 9.741791467457396e-06,
|
| 331 |
+
"loss": 6.2256,
|
| 332 |
+
"step": 25500
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.21,
|
| 336 |
+
"learning_rate": 9.733462159956021e-06,
|
| 337 |
+
"loss": 6.2239,
|
| 338 |
+
"step": 26000
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.22,
|
| 342 |
+
"learning_rate": 9.725132852454648e-06,
|
| 343 |
+
"loss": 6.2282,
|
| 344 |
+
"step": 26500
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.22,
|
| 348 |
+
"learning_rate": 9.716803544953275e-06,
|
| 349 |
+
"loss": 6.2222,
|
| 350 |
+
"step": 27000
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.23,
|
| 354 |
+
"learning_rate": 9.7084742374519e-06,
|
| 355 |
+
"loss": 6.213,
|
| 356 |
+
"step": 27500
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.23,
|
| 360 |
+
"learning_rate": 9.700144929950525e-06,
|
| 361 |
+
"loss": 6.2109,
|
| 362 |
+
"step": 28000
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.23,
|
| 366 |
+
"learning_rate": 9.69181562244915e-06,
|
| 367 |
+
"loss": 6.2101,
|
| 368 |
+
"step": 28500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.24,
|
| 372 |
+
"learning_rate": 9.683486314947776e-06,
|
| 373 |
+
"loss": 6.2061,
|
| 374 |
+
"step": 29000
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.24,
|
| 378 |
+
"learning_rate": 9.675157007446401e-06,
|
| 379 |
+
"loss": 6.199,
|
| 380 |
+
"step": 29500
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.25,
|
| 384 |
+
"learning_rate": 9.666827699945028e-06,
|
| 385 |
+
"loss": 6.1988,
|
| 386 |
+
"step": 30000
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.25,
|
| 390 |
+
"eval_accuracy": 0.1493027153779554,
|
| 391 |
+
"eval_loss": 6.185213088989258,
|
| 392 |
+
"eval_runtime": 2025.4966,
|
| 393 |
+
"eval_samples_per_second": 152.237,
|
| 394 |
+
"eval_steps_per_second": 3.172,
|
| 395 |
+
"step": 30000
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.25,
|
| 399 |
+
"learning_rate": 9.658498392443653e-06,
|
| 400 |
+
"loss": 6.1916,
|
| 401 |
+
"step": 30500
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.25,
|
| 405 |
+
"learning_rate": 9.65016908494228e-06,
|
| 406 |
+
"loss": 6.1918,
|
| 407 |
+
"step": 31000
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 0.26,
|
| 411 |
+
"learning_rate": 9.641839777440905e-06,
|
| 412 |
+
"loss": 6.1931,
|
| 413 |
+
"step": 31500
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.26,
|
| 417 |
+
"learning_rate": 9.63351046993953e-06,
|
| 418 |
+
"loss": 6.1847,
|
| 419 |
+
"step": 32000
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 0.27,
|
| 423 |
+
"learning_rate": 9.625181162438155e-06,
|
| 424 |
+
"loss": 6.1801,
|
| 425 |
+
"step": 32500
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.27,
|
| 429 |
+
"learning_rate": 9.616851854936781e-06,
|
| 430 |
+
"loss": 6.178,
|
| 431 |
+
"step": 33000
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 0.27,
|
| 435 |
+
"learning_rate": 9.608522547435406e-06,
|
| 436 |
+
"loss": 6.1729,
|
| 437 |
+
"step": 33500
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.28,
|
| 441 |
+
"learning_rate": 9.600193239934033e-06,
|
| 442 |
+
"loss": 6.1726,
|
| 443 |
+
"step": 34000
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.28,
|
| 447 |
+
"learning_rate": 9.591863932432658e-06,
|
| 448 |
+
"loss": 6.1659,
|
| 449 |
+
"step": 34500
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.29,
|
| 453 |
+
"learning_rate": 9.583534624931285e-06,
|
| 454 |
+
"loss": 6.1674,
|
| 455 |
+
"step": 35000
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.29,
|
| 459 |
+
"learning_rate": 9.57520531742991e-06,
|
| 460 |
+
"loss": 6.1718,
|
| 461 |
+
"step": 35500
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 0.29,
|
| 465 |
+
"learning_rate": 9.566876009928535e-06,
|
| 466 |
+
"loss": 6.1651,
|
| 467 |
+
"step": 36000
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 0.3,
|
| 471 |
+
"learning_rate": 9.55854670242716e-06,
|
| 472 |
+
"loss": 6.1618,
|
| 473 |
+
"step": 36500
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.3,
|
| 477 |
+
"learning_rate": 9.550217394925787e-06,
|
| 478 |
+
"loss": 6.1585,
|
| 479 |
+
"step": 37000
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.31,
|
| 483 |
+
"learning_rate": 9.541888087424413e-06,
|
| 484 |
+
"loss": 6.1559,
|
| 485 |
+
"step": 37500
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.31,
|
| 489 |
+
"learning_rate": 9.533558779923038e-06,
|
| 490 |
+
"loss": 6.1451,
|
| 491 |
+
"step": 38000
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 0.32,
|
| 495 |
+
"learning_rate": 9.525229472421663e-06,
|
| 496 |
+
"loss": 6.1481,
|
| 497 |
+
"step": 38500
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 0.32,
|
| 501 |
+
"learning_rate": 9.51690016492029e-06,
|
| 502 |
+
"loss": 6.1411,
|
| 503 |
+
"step": 39000
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 0.32,
|
| 507 |
+
"learning_rate": 9.508570857418915e-06,
|
| 508 |
+
"loss": 6.1441,
|
| 509 |
+
"step": 39500
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 0.33,
|
| 513 |
+
"learning_rate": 9.50024154991754e-06,
|
| 514 |
+
"loss": 6.148,
|
| 515 |
+
"step": 40000
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.33,
|
| 519 |
+
"eval_accuracy": 0.15008841811581047,
|
| 520 |
+
"eval_loss": 6.128693580627441,
|
| 521 |
+
"eval_runtime": 2030.9811,
|
| 522 |
+
"eval_samples_per_second": 151.826,
|
| 523 |
+
"eval_steps_per_second": 3.163,
|
| 524 |
+
"step": 40000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.33,
|
| 528 |
+
"learning_rate": 9.491912242416167e-06,
|
| 529 |
+
"loss": 6.1396,
|
| 530 |
+
"step": 40500
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.34,
|
| 534 |
+
"learning_rate": 9.483582934914792e-06,
|
| 535 |
+
"loss": 6.1376,
|
| 536 |
+
"step": 41000
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.34,
|
| 540 |
+
"learning_rate": 9.475253627413418e-06,
|
| 541 |
+
"loss": 6.1338,
|
| 542 |
+
"step": 41500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.34,
|
| 546 |
+
"learning_rate": 9.466924319912043e-06,
|
| 547 |
+
"loss": 6.1375,
|
| 548 |
+
"step": 42000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.35,
|
| 552 |
+
"learning_rate": 9.458595012410668e-06,
|
| 553 |
+
"loss": 6.1254,
|
| 554 |
+
"step": 42500
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.35,
|
| 558 |
+
"learning_rate": 9.450265704909295e-06,
|
| 559 |
+
"loss": 6.1252,
|
| 560 |
+
"step": 43000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.36,
|
| 564 |
+
"learning_rate": 9.44193639740792e-06,
|
| 565 |
+
"loss": 6.1209,
|
| 566 |
+
"step": 43500
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.36,
|
| 570 |
+
"learning_rate": 9.433607089906545e-06,
|
| 571 |
+
"loss": 6.1252,
|
| 572 |
+
"step": 44000
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.36,
|
| 576 |
+
"learning_rate": 9.425277782405172e-06,
|
| 577 |
+
"loss": 6.1205,
|
| 578 |
+
"step": 44500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.37,
|
| 582 |
+
"learning_rate": 9.416948474903798e-06,
|
| 583 |
+
"loss": 6.1177,
|
| 584 |
+
"step": 45000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.37,
|
| 588 |
+
"learning_rate": 9.408619167402423e-06,
|
| 589 |
+
"loss": 6.1204,
|
| 590 |
+
"step": 45500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.38,
|
| 594 |
+
"learning_rate": 9.400289859901048e-06,
|
| 595 |
+
"loss": 6.1101,
|
| 596 |
+
"step": 46000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.38,
|
| 600 |
+
"learning_rate": 9.391960552399673e-06,
|
| 601 |
+
"loss": 6.1094,
|
| 602 |
+
"step": 46500
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.39,
|
| 606 |
+
"learning_rate": 9.3836312448983e-06,
|
| 607 |
+
"loss": 6.1038,
|
| 608 |
+
"step": 47000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.39,
|
| 612 |
+
"learning_rate": 9.375301937396925e-06,
|
| 613 |
+
"loss": 6.1052,
|
| 614 |
+
"step": 47500
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.39,
|
| 618 |
+
"learning_rate": 9.366972629895552e-06,
|
| 619 |
+
"loss": 6.1056,
|
| 620 |
+
"step": 48000
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.4,
|
| 624 |
+
"learning_rate": 9.358643322394177e-06,
|
| 625 |
+
"loss": 6.1009,
|
| 626 |
+
"step": 48500
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.4,
|
| 630 |
+
"learning_rate": 9.350314014892804e-06,
|
| 631 |
+
"loss": 6.1006,
|
| 632 |
+
"step": 49000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.41,
|
| 636 |
+
"learning_rate": 9.341984707391429e-06,
|
| 637 |
+
"loss": 6.1009,
|
| 638 |
+
"step": 49500
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 0.41,
|
| 642 |
+
"learning_rate": 9.333655399890054e-06,
|
| 643 |
+
"loss": 6.1007,
|
| 644 |
+
"step": 50000
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 0.41,
|
| 648 |
+
"eval_accuracy": 0.1500883878026566,
|
| 649 |
+
"eval_loss": 6.088829040527344,
|
| 650 |
+
"eval_runtime": 2025.8132,
|
| 651 |
+
"eval_samples_per_second": 152.213,
|
| 652 |
+
"eval_steps_per_second": 3.172,
|
| 653 |
+
"step": 50000
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.41,
|
| 657 |
+
"learning_rate": 9.325326092388679e-06,
|
| 658 |
+
"loss": 6.0929,
|
| 659 |
+
"step": 50500
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 0.42,
|
| 663 |
+
"learning_rate": 9.316996784887305e-06,
|
| 664 |
+
"loss": 6.0942,
|
| 665 |
+
"step": 51000
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 0.42,
|
| 669 |
+
"learning_rate": 9.30866747738593e-06,
|
| 670 |
+
"loss": 6.097,
|
| 671 |
+
"step": 51500
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 0.43,
|
| 675 |
+
"learning_rate": 9.300338169884557e-06,
|
| 676 |
+
"loss": 6.0922,
|
| 677 |
+
"step": 52000
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 0.43,
|
| 681 |
+
"learning_rate": 9.292008862383182e-06,
|
| 682 |
+
"loss": 6.088,
|
| 683 |
+
"step": 52500
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.43,
|
| 687 |
+
"learning_rate": 9.283679554881809e-06,
|
| 688 |
+
"loss": 6.0922,
|
| 689 |
+
"step": 53000
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.44,
|
| 693 |
+
"learning_rate": 9.275350247380434e-06,
|
| 694 |
+
"loss": 6.0847,
|
| 695 |
+
"step": 53500
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.44,
|
| 699 |
+
"learning_rate": 9.267020939879059e-06,
|
| 700 |
+
"loss": 6.0872,
|
| 701 |
+
"step": 54000
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 0.45,
|
| 705 |
+
"learning_rate": 9.258691632377685e-06,
|
| 706 |
+
"loss": 6.0861,
|
| 707 |
+
"step": 54500
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 0.45,
|
| 711 |
+
"learning_rate": 9.25036232487631e-06,
|
| 712 |
+
"loss": 6.0777,
|
| 713 |
+
"step": 55000
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.45,
|
| 717 |
+
"learning_rate": 9.242033017374937e-06,
|
| 718 |
+
"loss": 6.0849,
|
| 719 |
+
"step": 55500
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 0.46,
|
| 723 |
+
"learning_rate": 9.233703709873562e-06,
|
| 724 |
+
"loss": 6.0792,
|
| 725 |
+
"step": 56000
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 0.46,
|
| 729 |
+
"learning_rate": 9.225374402372189e-06,
|
| 730 |
+
"loss": 6.0751,
|
| 731 |
+
"step": 56500
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.47,
|
| 735 |
+
"learning_rate": 9.217045094870814e-06,
|
| 736 |
+
"loss": 6.066,
|
| 737 |
+
"step": 57000
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.47,
|
| 741 |
+
"learning_rate": 9.208715787369439e-06,
|
| 742 |
+
"loss": 6.0767,
|
| 743 |
+
"step": 57500
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 0.48,
|
| 747 |
+
"learning_rate": 9.200386479868064e-06,
|
| 748 |
+
"loss": 6.0771,
|
| 749 |
+
"step": 58000
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 0.48,
|
| 753 |
+
"learning_rate": 9.19205717236669e-06,
|
| 754 |
+
"loss": 6.0733,
|
| 755 |
+
"step": 58500
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 0.48,
|
| 759 |
+
"learning_rate": 9.183727864865315e-06,
|
| 760 |
+
"loss": 6.0693,
|
| 761 |
+
"step": 59000
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.49,
|
| 765 |
+
"learning_rate": 9.175398557363942e-06,
|
| 766 |
+
"loss": 6.0739,
|
| 767 |
+
"step": 59500
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.49,
|
| 771 |
+
"learning_rate": 9.167069249862567e-06,
|
| 772 |
+
"loss": 6.0721,
|
| 773 |
+
"step": 60000
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.49,
|
| 777 |
+
"eval_accuracy": 0.14991843929852017,
|
| 778 |
+
"eval_loss": 6.0554890632629395,
|
| 779 |
+
"eval_runtime": 2030.6269,
|
| 780 |
+
"eval_samples_per_second": 151.853,
|
| 781 |
+
"eval_steps_per_second": 3.164,
|
| 782 |
+
"step": 60000
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.5,
|
| 786 |
+
"learning_rate": 9.158739942361194e-06,
|
| 787 |
+
"loss": 6.062,
|
| 788 |
+
"step": 60500
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.5,
|
| 792 |
+
"learning_rate": 9.150410634859819e-06,
|
| 793 |
+
"loss": 6.0598,
|
| 794 |
+
"step": 61000
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.5,
|
| 798 |
+
"learning_rate": 9.142081327358444e-06,
|
| 799 |
+
"loss": 6.0644,
|
| 800 |
+
"step": 61500
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.51,
|
| 804 |
+
"learning_rate": 9.133752019857069e-06,
|
| 805 |
+
"loss": 6.0561,
|
| 806 |
+
"step": 62000
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 0.51,
|
| 810 |
+
"learning_rate": 9.125422712355696e-06,
|
| 811 |
+
"loss": 6.0606,
|
| 812 |
+
"step": 62500
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.52,
|
| 816 |
+
"learning_rate": 9.11709340485432e-06,
|
| 817 |
+
"loss": 6.0611,
|
| 818 |
+
"step": 63000
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 0.52,
|
| 822 |
+
"learning_rate": 9.108764097352947e-06,
|
| 823 |
+
"loss": 6.0494,
|
| 824 |
+
"step": 63500
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.52,
|
| 828 |
+
"learning_rate": 9.100434789851572e-06,
|
| 829 |
+
"loss": 6.0507,
|
| 830 |
+
"step": 64000
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.53,
|
| 834 |
+
"learning_rate": 9.092105482350199e-06,
|
| 835 |
+
"loss": 6.0536,
|
| 836 |
+
"step": 64500
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.53,
|
| 840 |
+
"learning_rate": 9.083776174848824e-06,
|
| 841 |
+
"loss": 6.0523,
|
| 842 |
+
"step": 65000
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.54,
|
| 846 |
+
"learning_rate": 9.075446867347449e-06,
|
| 847 |
+
"loss": 6.0545,
|
| 848 |
+
"step": 65500
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.54,
|
| 852 |
+
"learning_rate": 9.067117559846076e-06,
|
| 853 |
+
"loss": 6.0559,
|
| 854 |
+
"step": 66000
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.54,
|
| 858 |
+
"learning_rate": 9.0587882523447e-06,
|
| 859 |
+
"loss": 6.0449,
|
| 860 |
+
"step": 66500
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 0.55,
|
| 864 |
+
"learning_rate": 9.050458944843327e-06,
|
| 865 |
+
"loss": 6.0449,
|
| 866 |
+
"step": 67000
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.55,
|
| 870 |
+
"learning_rate": 9.042129637341952e-06,
|
| 871 |
+
"loss": 6.0459,
|
| 872 |
+
"step": 67500
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.56,
|
| 876 |
+
"learning_rate": 9.033800329840577e-06,
|
| 877 |
+
"loss": 6.0498,
|
| 878 |
+
"step": 68000
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.56,
|
| 882 |
+
"learning_rate": 9.025471022339204e-06,
|
| 883 |
+
"loss": 6.0462,
|
| 884 |
+
"step": 68500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.57,
|
| 888 |
+
"learning_rate": 9.017141714837829e-06,
|
| 889 |
+
"loss": 6.043,
|
| 890 |
+
"step": 69000
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 0.57,
|
| 894 |
+
"learning_rate": 9.008812407336454e-06,
|
| 895 |
+
"loss": 6.0461,
|
| 896 |
+
"step": 69500
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.57,
|
| 900 |
+
"learning_rate": 9.00048309983508e-06,
|
| 901 |
+
"loss": 6.0414,
|
| 902 |
+
"step": 70000
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 0.57,
|
| 906 |
+
"eval_accuracy": 0.15135291944253934,
|
| 907 |
+
"eval_loss": 6.027411937713623,
|
| 908 |
+
"eval_runtime": 2027.398,
|
| 909 |
+
"eval_samples_per_second": 152.094,
|
| 910 |
+
"eval_steps_per_second": 3.169,
|
| 911 |
+
"step": 70000
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 0.58,
|
| 915 |
+
"learning_rate": 8.992153792333706e-06,
|
| 916 |
+
"loss": 6.0374,
|
| 917 |
+
"step": 70500
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 0.58,
|
| 921 |
+
"learning_rate": 8.983824484832332e-06,
|
| 922 |
+
"loss": 6.0307,
|
| 923 |
+
"step": 71000
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 0.59,
|
| 927 |
+
"learning_rate": 8.975495177330957e-06,
|
| 928 |
+
"loss": 6.0401,
|
| 929 |
+
"step": 71500
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 0.59,
|
| 933 |
+
"learning_rate": 8.967165869829582e-06,
|
| 934 |
+
"loss": 6.0376,
|
| 935 |
+
"step": 72000
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 0.59,
|
| 939 |
+
"learning_rate": 8.958836562328209e-06,
|
| 940 |
+
"loss": 6.0373,
|
| 941 |
+
"step": 72500
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.6,
|
| 945 |
+
"learning_rate": 8.950507254826834e-06,
|
| 946 |
+
"loss": 6.0303,
|
| 947 |
+
"step": 73000
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.6,
|
| 951 |
+
"learning_rate": 8.942177947325459e-06,
|
| 952 |
+
"loss": 6.0365,
|
| 953 |
+
"step": 73500
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 0.61,
|
| 957 |
+
"learning_rate": 8.933848639824086e-06,
|
| 958 |
+
"loss": 6.0295,
|
| 959 |
+
"step": 74000
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 0.61,
|
| 963 |
+
"learning_rate": 8.925519332322713e-06,
|
| 964 |
+
"loss": 6.0279,
|
| 965 |
+
"step": 74500
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 0.61,
|
| 969 |
+
"learning_rate": 8.917190024821338e-06,
|
| 970 |
+
"loss": 6.0316,
|
| 971 |
+
"step": 75000
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 0.62,
|
| 975 |
+
"learning_rate": 8.908860717319963e-06,
|
| 976 |
+
"loss": 6.0249,
|
| 977 |
+
"step": 75500
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 0.62,
|
| 981 |
+
"learning_rate": 8.900531409818588e-06,
|
| 982 |
+
"loss": 6.0282,
|
| 983 |
+
"step": 76000
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.63,
|
| 987 |
+
"learning_rate": 8.892202102317214e-06,
|
| 988 |
+
"loss": 6.0228,
|
| 989 |
+
"step": 76500
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.63,
|
| 993 |
+
"learning_rate": 8.88387279481584e-06,
|
| 994 |
+
"loss": 6.0238,
|
| 995 |
+
"step": 77000
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 0.63,
|
| 999 |
+
"learning_rate": 8.875543487314466e-06,
|
| 1000 |
+
"loss": 6.0215,
|
| 1001 |
+
"step": 77500
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 0.64,
|
| 1005 |
+
"learning_rate": 8.867214179813091e-06,
|
| 1006 |
+
"loss": 6.0204,
|
| 1007 |
+
"step": 78000
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 0.64,
|
| 1011 |
+
"learning_rate": 8.858884872311718e-06,
|
| 1012 |
+
"loss": 6.0249,
|
| 1013 |
+
"step": 78500
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 0.65,
|
| 1017 |
+
"learning_rate": 8.850555564810343e-06,
|
| 1018 |
+
"loss": 6.0199,
|
| 1019 |
+
"step": 79000
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 0.65,
|
| 1023 |
+
"learning_rate": 8.842226257308968e-06,
|
| 1024 |
+
"loss": 6.0208,
|
| 1025 |
+
"step": 79500
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.66,
|
| 1029 |
+
"learning_rate": 8.833896949807593e-06,
|
| 1030 |
+
"loss": 6.0229,
|
| 1031 |
+
"step": 80000
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.66,
|
| 1035 |
+
"eval_accuracy": 0.1511706409765356,
|
| 1036 |
+
"eval_loss": 6.008976459503174,
|
| 1037 |
+
"eval_runtime": 2027.8147,
|
| 1038 |
+
"eval_samples_per_second": 152.063,
|
| 1039 |
+
"eval_steps_per_second": 3.168,
|
| 1040 |
+
"step": 80000
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.66,
|
| 1044 |
+
"step": 80001,
|
| 1045 |
+
"total_flos": 1.0107191126130688e+18,
|
| 1046 |
+
"train_loss": 6.2887065917574985,
|
| 1047 |
+
"train_runtime": 86523.957,
|
| 1048 |
+
"train_samples_per_second": 338.563,
|
| 1049 |
+
"train_steps_per_second": 7.053
|
| 1050 |
+
}
|
| 1051 |
+
],
|
| 1052 |
+
"max_steps": 610290,
|
| 1053 |
+
"num_train_epochs": 5,
|
| 1054 |
+
"total_flos": 1.0107191126130688e+18,
|
| 1055 |
+
"trial_name": null,
|
| 1056 |
+
"trial_params": null
|
| 1057 |
+
}
|