End of training
Browse files- all_results.json +17 -0
- eval_results.json +12 -0
- runs/May23_11-27-41_4b659dec4dc1/events.out.tfevents.1684847604.4b659dec4dc1.6139.2 +3 -0
- train_results.json +8 -0
- trainer_state.json +1615 -0
all_results.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_bleu": 78.4172,
|
| 4 |
+
"eval_em": 0.08,
|
| 5 |
+
"eval_gen_len": 48.92,
|
| 6 |
+
"eval_loss": 0.9001668691635132,
|
| 7 |
+
"eval_rm": 0.9565,
|
| 8 |
+
"eval_runtime": 324.9714,
|
| 9 |
+
"eval_samples": 50,
|
| 10 |
+
"eval_samples_per_second": 0.154,
|
| 11 |
+
"eval_steps_per_second": 0.022,
|
| 12 |
+
"train_loss": 0.1476078914309933,
|
| 13 |
+
"train_runtime": 5095.148,
|
| 14 |
+
"train_samples": 19276,
|
| 15 |
+
"train_samples_per_second": 37.832,
|
| 16 |
+
"train_steps_per_second": 4.73
|
| 17 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_bleu": 78.4172,
|
| 4 |
+
"eval_em": 0.08,
|
| 5 |
+
"eval_gen_len": 48.92,
|
| 6 |
+
"eval_loss": 0.9001668691635132,
|
| 7 |
+
"eval_rm": 0.9565,
|
| 8 |
+
"eval_runtime": 324.9714,
|
| 9 |
+
"eval_samples": 50,
|
| 10 |
+
"eval_samples_per_second": 0.154,
|
| 11 |
+
"eval_steps_per_second": 0.022
|
| 12 |
+
}
|
runs/May23_11-27-41_4b659dec4dc1/events.out.tfevents.1684847604.4b659dec4dc1.6139.2
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a0ccf530a16bbe9125b97d3e49f8c02a128fb6a60c8b59a0e42d6bbd8eaa542d
|
| 3 |
+
size 559
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"train_loss": 0.1476078914309933,
|
| 4 |
+
"train_runtime": 5095.148,
|
| 5 |
+
"train_samples": 19276,
|
| 6 |
+
"train_samples_per_second": 37.832,
|
| 7 |
+
"train_steps_per_second": 4.73
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1615 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"global_step": 24100,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.04,
|
| 12 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 13 |
+
"loss": 92.256,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.08,
|
| 18 |
+
"learning_rate": 4.000000000000001e-06,
|
| 19 |
+
"loss": 52.0764,
|
| 20 |
+
"step": 200
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.12,
|
| 24 |
+
"learning_rate": 6e-06,
|
| 25 |
+
"loss": 41.3627,
|
| 26 |
+
"step": 300
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.17,
|
| 30 |
+
"learning_rate": 8.000000000000001e-06,
|
| 31 |
+
"loss": 37.237,
|
| 32 |
+
"step": 400
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.21,
|
| 36 |
+
"learning_rate": 1e-05,
|
| 37 |
+
"loss": 34.3031,
|
| 38 |
+
"step": 500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.25,
|
| 42 |
+
"learning_rate": 1.2e-05,
|
| 43 |
+
"loss": 31.7147,
|
| 44 |
+
"step": 600
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.29,
|
| 48 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 49 |
+
"loss": 29.5475,
|
| 50 |
+
"step": 700
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.33,
|
| 54 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 55 |
+
"loss": 27.914,
|
| 56 |
+
"step": 800
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.37,
|
| 60 |
+
"learning_rate": 1.8e-05,
|
| 61 |
+
"loss": 26.1499,
|
| 62 |
+
"step": 900
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.41,
|
| 66 |
+
"learning_rate": 2e-05,
|
| 67 |
+
"loss": 24.1629,
|
| 68 |
+
"step": 1000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.46,
|
| 72 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 73 |
+
"loss": 22.1829,
|
| 74 |
+
"step": 1100
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.5,
|
| 78 |
+
"learning_rate": 2.4e-05,
|
| 79 |
+
"loss": 20.2111,
|
| 80 |
+
"step": 1200
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.54,
|
| 84 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 85 |
+
"loss": 18.6642,
|
| 86 |
+
"step": 1300
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.58,
|
| 90 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 91 |
+
"loss": 16.7946,
|
| 92 |
+
"step": 1400
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.62,
|
| 96 |
+
"learning_rate": 3e-05,
|
| 97 |
+
"loss": 15.0783,
|
| 98 |
+
"step": 1500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.66,
|
| 102 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 103 |
+
"loss": 13.366,
|
| 104 |
+
"step": 1600
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.71,
|
| 108 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 109 |
+
"loss": 11.706,
|
| 110 |
+
"step": 1700
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.75,
|
| 114 |
+
"learning_rate": 3.6e-05,
|
| 115 |
+
"loss": 10.2113,
|
| 116 |
+
"step": 1800
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.79,
|
| 120 |
+
"learning_rate": 3.8e-05,
|
| 121 |
+
"loss": 8.5921,
|
| 122 |
+
"step": 1900
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.83,
|
| 126 |
+
"learning_rate": 4e-05,
|
| 127 |
+
"loss": 7.2445,
|
| 128 |
+
"step": 2000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.83,
|
| 132 |
+
"eval_bleu": 1.523,
|
| 133 |
+
"eval_em": 0.0,
|
| 134 |
+
"eval_gen_len": 204.48,
|
| 135 |
+
"eval_loss": 7.561939239501953,
|
| 136 |
+
"eval_rm": NaN,
|
| 137 |
+
"eval_runtime": 100.5427,
|
| 138 |
+
"eval_samples_per_second": 0.497,
|
| 139 |
+
"eval_steps_per_second": 0.07,
|
| 140 |
+
"step": 2000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.87,
|
| 144 |
+
"learning_rate": 4.2e-05,
|
| 145 |
+
"loss": 6.8461,
|
| 146 |
+
"step": 2100
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.91,
|
| 150 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 151 |
+
"loss": 6.223,
|
| 152 |
+
"step": 2200
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.95,
|
| 156 |
+
"learning_rate": 4.600000000000001e-05,
|
| 157 |
+
"loss": 6.1167,
|
| 158 |
+
"step": 2300
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.0,
|
| 162 |
+
"learning_rate": 4.8e-05,
|
| 163 |
+
"loss": 5.7623,
|
| 164 |
+
"step": 2400
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.04,
|
| 168 |
+
"learning_rate": 5e-05,
|
| 169 |
+
"loss": 5.5056,
|
| 170 |
+
"step": 2500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.08,
|
| 174 |
+
"learning_rate": 4.976851851851852e-05,
|
| 175 |
+
"loss": 5.2979,
|
| 176 |
+
"step": 2600
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 1.12,
|
| 180 |
+
"learning_rate": 4.9537037037037035e-05,
|
| 181 |
+
"loss": 4.9765,
|
| 182 |
+
"step": 2700
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.16,
|
| 186 |
+
"learning_rate": 4.930555555555556e-05,
|
| 187 |
+
"loss": 4.7721,
|
| 188 |
+
"step": 2800
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 1.2,
|
| 192 |
+
"learning_rate": 4.9074074074074075e-05,
|
| 193 |
+
"loss": 4.6957,
|
| 194 |
+
"step": 2900
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 1.24,
|
| 198 |
+
"learning_rate": 4.8842592592592595e-05,
|
| 199 |
+
"loss": 4.7368,
|
| 200 |
+
"step": 3000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.29,
|
| 204 |
+
"learning_rate": 4.8611111111111115e-05,
|
| 205 |
+
"loss": 4.6074,
|
| 206 |
+
"step": 3100
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.33,
|
| 210 |
+
"learning_rate": 4.837962962962963e-05,
|
| 211 |
+
"loss": 4.3678,
|
| 212 |
+
"step": 3200
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 1.37,
|
| 216 |
+
"learning_rate": 4.814814814814815e-05,
|
| 217 |
+
"loss": 4.19,
|
| 218 |
+
"step": 3300
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 1.41,
|
| 222 |
+
"learning_rate": 4.791666666666667e-05,
|
| 223 |
+
"loss": 4.0959,
|
| 224 |
+
"step": 3400
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 1.45,
|
| 228 |
+
"learning_rate": 4.768518518518519e-05,
|
| 229 |
+
"loss": 4.0323,
|
| 230 |
+
"step": 3500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 1.49,
|
| 234 |
+
"learning_rate": 4.745370370370371e-05,
|
| 235 |
+
"loss": 3.7777,
|
| 236 |
+
"step": 3600
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 1.54,
|
| 240 |
+
"learning_rate": 4.722222222222222e-05,
|
| 241 |
+
"loss": 3.8014,
|
| 242 |
+
"step": 3700
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.58,
|
| 246 |
+
"learning_rate": 4.699074074074074e-05,
|
| 247 |
+
"loss": 3.6346,
|
| 248 |
+
"step": 3800
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.62,
|
| 252 |
+
"learning_rate": 4.675925925925926e-05,
|
| 253 |
+
"loss": 3.4876,
|
| 254 |
+
"step": 3900
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.66,
|
| 258 |
+
"learning_rate": 4.652777777777778e-05,
|
| 259 |
+
"loss": 3.4458,
|
| 260 |
+
"step": 4000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 1.66,
|
| 264 |
+
"eval_bleu": 16.0372,
|
| 265 |
+
"eval_em": 0.0,
|
| 266 |
+
"eval_gen_len": 76.36,
|
| 267 |
+
"eval_loss": 3.4201648235321045,
|
| 268 |
+
"eval_rm": NaN,
|
| 269 |
+
"eval_runtime": 91.7678,
|
| 270 |
+
"eval_samples_per_second": 0.545,
|
| 271 |
+
"eval_steps_per_second": 0.076,
|
| 272 |
+
"step": 4000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 1.7,
|
| 276 |
+
"learning_rate": 4.62962962962963e-05,
|
| 277 |
+
"loss": 3.4425,
|
| 278 |
+
"step": 4100
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 1.74,
|
| 282 |
+
"learning_rate": 4.6064814814814814e-05,
|
| 283 |
+
"loss": 3.3346,
|
| 284 |
+
"step": 4200
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.78,
|
| 288 |
+
"learning_rate": 4.5833333333333334e-05,
|
| 289 |
+
"loss": 3.2582,
|
| 290 |
+
"step": 4300
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 1.83,
|
| 294 |
+
"learning_rate": 4.5601851851851854e-05,
|
| 295 |
+
"loss": 3.1174,
|
| 296 |
+
"step": 4400
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.87,
|
| 300 |
+
"learning_rate": 4.5370370370370374e-05,
|
| 301 |
+
"loss": 2.9873,
|
| 302 |
+
"step": 4500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 1.91,
|
| 306 |
+
"learning_rate": 4.5138888888888894e-05,
|
| 307 |
+
"loss": 3.1041,
|
| 308 |
+
"step": 4600
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.95,
|
| 312 |
+
"learning_rate": 4.490740740740741e-05,
|
| 313 |
+
"loss": 2.9365,
|
| 314 |
+
"step": 4700
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 1.99,
|
| 318 |
+
"learning_rate": 4.467592592592593e-05,
|
| 319 |
+
"loss": 2.9494,
|
| 320 |
+
"step": 4800
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 2.03,
|
| 324 |
+
"learning_rate": 4.4444444444444447e-05,
|
| 325 |
+
"loss": 2.8309,
|
| 326 |
+
"step": 4900
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 2.07,
|
| 330 |
+
"learning_rate": 4.4212962962962966e-05,
|
| 331 |
+
"loss": 2.6814,
|
| 332 |
+
"step": 5000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 2.12,
|
| 336 |
+
"learning_rate": 4.3981481481481486e-05,
|
| 337 |
+
"loss": 2.6331,
|
| 338 |
+
"step": 5100
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 2.16,
|
| 342 |
+
"learning_rate": 4.375e-05,
|
| 343 |
+
"loss": 2.599,
|
| 344 |
+
"step": 5200
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 2.2,
|
| 348 |
+
"learning_rate": 4.351851851851852e-05,
|
| 349 |
+
"loss": 2.5576,
|
| 350 |
+
"step": 5300
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 2.24,
|
| 354 |
+
"learning_rate": 4.328703703703704e-05,
|
| 355 |
+
"loss": 2.4838,
|
| 356 |
+
"step": 5400
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 2.28,
|
| 360 |
+
"learning_rate": 4.305555555555556e-05,
|
| 361 |
+
"loss": 2.4532,
|
| 362 |
+
"step": 5500
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 2.32,
|
| 366 |
+
"learning_rate": 4.282407407407408e-05,
|
| 367 |
+
"loss": 2.3583,
|
| 368 |
+
"step": 5600
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.37,
|
| 372 |
+
"learning_rate": 4.259259259259259e-05,
|
| 373 |
+
"loss": 2.4245,
|
| 374 |
+
"step": 5700
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 2.41,
|
| 378 |
+
"learning_rate": 4.236111111111111e-05,
|
| 379 |
+
"loss": 2.3121,
|
| 380 |
+
"step": 5800
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 2.45,
|
| 384 |
+
"learning_rate": 4.212962962962963e-05,
|
| 385 |
+
"loss": 2.3254,
|
| 386 |
+
"step": 5900
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 2.49,
|
| 390 |
+
"learning_rate": 4.1898148148148145e-05,
|
| 391 |
+
"loss": 2.2571,
|
| 392 |
+
"step": 6000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 2.49,
|
| 396 |
+
"eval_bleu": 34.5826,
|
| 397 |
+
"eval_em": 0.0,
|
| 398 |
+
"eval_gen_len": 48.98,
|
| 399 |
+
"eval_loss": 2.197143316268921,
|
| 400 |
+
"eval_rm": 0.8125,
|
| 401 |
+
"eval_runtime": 79.1278,
|
| 402 |
+
"eval_samples_per_second": 0.632,
|
| 403 |
+
"eval_steps_per_second": 0.088,
|
| 404 |
+
"step": 6000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 2.53,
|
| 408 |
+
"learning_rate": 4.166666666666667e-05,
|
| 409 |
+
"loss": 2.2655,
|
| 410 |
+
"step": 6100
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 2.57,
|
| 414 |
+
"learning_rate": 4.1435185185185185e-05,
|
| 415 |
+
"loss": 2.1366,
|
| 416 |
+
"step": 6200
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 2.61,
|
| 420 |
+
"learning_rate": 4.1203703703703705e-05,
|
| 421 |
+
"loss": 2.1777,
|
| 422 |
+
"step": 6300
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 2.66,
|
| 426 |
+
"learning_rate": 4.0972222222222225e-05,
|
| 427 |
+
"loss": 2.1166,
|
| 428 |
+
"step": 6400
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 2.7,
|
| 432 |
+
"learning_rate": 4.074074074074074e-05,
|
| 433 |
+
"loss": 2.0804,
|
| 434 |
+
"step": 6500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 2.74,
|
| 438 |
+
"learning_rate": 4.0509259259259265e-05,
|
| 439 |
+
"loss": 2.1024,
|
| 440 |
+
"step": 6600
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 2.78,
|
| 444 |
+
"learning_rate": 4.027777777777778e-05,
|
| 445 |
+
"loss": 2.0787,
|
| 446 |
+
"step": 6700
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 2.82,
|
| 450 |
+
"learning_rate": 4.00462962962963e-05,
|
| 451 |
+
"loss": 1.9875,
|
| 452 |
+
"step": 6800
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 2.86,
|
| 456 |
+
"learning_rate": 3.981481481481482e-05,
|
| 457 |
+
"loss": 1.9653,
|
| 458 |
+
"step": 6900
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 2.9,
|
| 462 |
+
"learning_rate": 3.958333333333333e-05,
|
| 463 |
+
"loss": 1.9254,
|
| 464 |
+
"step": 7000
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.95,
|
| 468 |
+
"learning_rate": 3.935185185185186e-05,
|
| 469 |
+
"loss": 1.9232,
|
| 470 |
+
"step": 7100
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 2.99,
|
| 474 |
+
"learning_rate": 3.912037037037037e-05,
|
| 475 |
+
"loss": 1.8927,
|
| 476 |
+
"step": 7200
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 3.03,
|
| 480 |
+
"learning_rate": 3.888888888888889e-05,
|
| 481 |
+
"loss": 1.7833,
|
| 482 |
+
"step": 7300
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 3.07,
|
| 486 |
+
"learning_rate": 3.865740740740741e-05,
|
| 487 |
+
"loss": 1.6375,
|
| 488 |
+
"step": 7400
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 3.11,
|
| 492 |
+
"learning_rate": 3.8425925925925924e-05,
|
| 493 |
+
"loss": 1.7233,
|
| 494 |
+
"step": 7500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 3.15,
|
| 498 |
+
"learning_rate": 3.8194444444444444e-05,
|
| 499 |
+
"loss": 1.6997,
|
| 500 |
+
"step": 7600
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 3.2,
|
| 504 |
+
"learning_rate": 3.7962962962962964e-05,
|
| 505 |
+
"loss": 1.746,
|
| 506 |
+
"step": 7700
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 3.24,
|
| 510 |
+
"learning_rate": 3.7731481481481484e-05,
|
| 511 |
+
"loss": 1.6235,
|
| 512 |
+
"step": 7800
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 3.28,
|
| 516 |
+
"learning_rate": 3.7500000000000003e-05,
|
| 517 |
+
"loss": 1.564,
|
| 518 |
+
"step": 7900
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 3.32,
|
| 522 |
+
"learning_rate": 3.726851851851852e-05,
|
| 523 |
+
"loss": 1.5696,
|
| 524 |
+
"step": 8000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 3.32,
|
| 528 |
+
"eval_bleu": 53.8822,
|
| 529 |
+
"eval_em": 0.0,
|
| 530 |
+
"eval_gen_len": 48.7,
|
| 531 |
+
"eval_loss": 1.680190086364746,
|
| 532 |
+
"eval_rm": 1.0,
|
| 533 |
+
"eval_runtime": 52.3216,
|
| 534 |
+
"eval_samples_per_second": 0.956,
|
| 535 |
+
"eval_steps_per_second": 0.134,
|
| 536 |
+
"step": 8000
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 3.36,
|
| 540 |
+
"learning_rate": 3.7037037037037037e-05,
|
| 541 |
+
"loss": 1.5516,
|
| 542 |
+
"step": 8100
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 3.4,
|
| 546 |
+
"learning_rate": 3.6805555555555556e-05,
|
| 547 |
+
"loss": 1.5438,
|
| 548 |
+
"step": 8200
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 3.44,
|
| 552 |
+
"learning_rate": 3.6574074074074076e-05,
|
| 553 |
+
"loss": 1.5262,
|
| 554 |
+
"step": 8300
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 3.49,
|
| 558 |
+
"learning_rate": 3.6342592592592596e-05,
|
| 559 |
+
"loss": 1.5738,
|
| 560 |
+
"step": 8400
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 3.53,
|
| 564 |
+
"learning_rate": 3.611111111111111e-05,
|
| 565 |
+
"loss": 1.5799,
|
| 566 |
+
"step": 8500
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 3.57,
|
| 570 |
+
"learning_rate": 3.587962962962963e-05,
|
| 571 |
+
"loss": 1.5116,
|
| 572 |
+
"step": 8600
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 3.61,
|
| 576 |
+
"learning_rate": 3.564814814814815e-05,
|
| 577 |
+
"loss": 1.5699,
|
| 578 |
+
"step": 8700
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 3.65,
|
| 582 |
+
"learning_rate": 3.541666666666667e-05,
|
| 583 |
+
"loss": 1.4994,
|
| 584 |
+
"step": 8800
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 3.69,
|
| 588 |
+
"learning_rate": 3.518518518518519e-05,
|
| 589 |
+
"loss": 1.3772,
|
| 590 |
+
"step": 8900
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 3.73,
|
| 594 |
+
"learning_rate": 3.49537037037037e-05,
|
| 595 |
+
"loss": 1.4322,
|
| 596 |
+
"step": 9000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 3.78,
|
| 600 |
+
"learning_rate": 3.472222222222222e-05,
|
| 601 |
+
"loss": 1.4604,
|
| 602 |
+
"step": 9100
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 3.82,
|
| 606 |
+
"learning_rate": 3.449074074074074e-05,
|
| 607 |
+
"loss": 1.3924,
|
| 608 |
+
"step": 9200
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 3.86,
|
| 612 |
+
"learning_rate": 3.425925925925926e-05,
|
| 613 |
+
"loss": 1.348,
|
| 614 |
+
"step": 9300
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 3.9,
|
| 618 |
+
"learning_rate": 3.402777777777778e-05,
|
| 619 |
+
"loss": 1.3014,
|
| 620 |
+
"step": 9400
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 3.94,
|
| 624 |
+
"learning_rate": 3.3796296296296295e-05,
|
| 625 |
+
"loss": 1.3731,
|
| 626 |
+
"step": 9500
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 3.98,
|
| 630 |
+
"learning_rate": 3.3564814814814815e-05,
|
| 631 |
+
"loss": 1.39,
|
| 632 |
+
"step": 9600
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 4.02,
|
| 636 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 637 |
+
"loss": 1.2038,
|
| 638 |
+
"step": 9700
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 4.07,
|
| 642 |
+
"learning_rate": 3.3101851851851855e-05,
|
| 643 |
+
"loss": 1.1698,
|
| 644 |
+
"step": 9800
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 4.11,
|
| 648 |
+
"learning_rate": 3.2870370370370375e-05,
|
| 649 |
+
"loss": 1.1514,
|
| 650 |
+
"step": 9900
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 4.15,
|
| 654 |
+
"learning_rate": 3.263888888888889e-05,
|
| 655 |
+
"loss": 1.1359,
|
| 656 |
+
"step": 10000
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 4.15,
|
| 660 |
+
"eval_bleu": 64.5591,
|
| 661 |
+
"eval_em": 0.02,
|
| 662 |
+
"eval_gen_len": 45.2,
|
| 663 |
+
"eval_loss": 1.346365213394165,
|
| 664 |
+
"eval_rm": 0.973,
|
| 665 |
+
"eval_runtime": 47.5326,
|
| 666 |
+
"eval_samples_per_second": 1.052,
|
| 667 |
+
"eval_steps_per_second": 0.147,
|
| 668 |
+
"step": 10000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 4.19,
|
| 672 |
+
"learning_rate": 3.240740740740741e-05,
|
| 673 |
+
"loss": 1.1396,
|
| 674 |
+
"step": 10100
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 4.23,
|
| 678 |
+
"learning_rate": 3.217592592592593e-05,
|
| 679 |
+
"loss": 1.1749,
|
| 680 |
+
"step": 10200
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 4.27,
|
| 684 |
+
"learning_rate": 3.194444444444444e-05,
|
| 685 |
+
"loss": 1.1098,
|
| 686 |
+
"step": 10300
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 4.32,
|
| 690 |
+
"learning_rate": 3.171296296296297e-05,
|
| 691 |
+
"loss": 1.0741,
|
| 692 |
+
"step": 10400
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 4.36,
|
| 696 |
+
"learning_rate": 3.148148148148148e-05,
|
| 697 |
+
"loss": 1.1388,
|
| 698 |
+
"step": 10500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 4.4,
|
| 702 |
+
"learning_rate": 3.125e-05,
|
| 703 |
+
"loss": 1.0168,
|
| 704 |
+
"step": 10600
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 4.44,
|
| 708 |
+
"learning_rate": 3.101851851851852e-05,
|
| 709 |
+
"loss": 1.0862,
|
| 710 |
+
"step": 10700
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 4.48,
|
| 714 |
+
"learning_rate": 3.0787037037037034e-05,
|
| 715 |
+
"loss": 1.0689,
|
| 716 |
+
"step": 10800
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 4.52,
|
| 720 |
+
"learning_rate": 3.055555555555556e-05,
|
| 721 |
+
"loss": 1.0804,
|
| 722 |
+
"step": 10900
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 4.56,
|
| 726 |
+
"learning_rate": 3.0324074074074077e-05,
|
| 727 |
+
"loss": 1.0312,
|
| 728 |
+
"step": 11000
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 4.61,
|
| 732 |
+
"learning_rate": 3.0092592592592593e-05,
|
| 733 |
+
"loss": 1.0184,
|
| 734 |
+
"step": 11100
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 4.65,
|
| 738 |
+
"learning_rate": 2.9861111111111113e-05,
|
| 739 |
+
"loss": 0.992,
|
| 740 |
+
"step": 11200
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 4.69,
|
| 744 |
+
"learning_rate": 2.962962962962963e-05,
|
| 745 |
+
"loss": 1.0191,
|
| 746 |
+
"step": 11300
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 4.73,
|
| 750 |
+
"learning_rate": 2.9398148148148146e-05,
|
| 751 |
+
"loss": 1.0842,
|
| 752 |
+
"step": 11400
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 4.77,
|
| 756 |
+
"learning_rate": 2.916666666666667e-05,
|
| 757 |
+
"loss": 1.0508,
|
| 758 |
+
"step": 11500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 4.81,
|
| 762 |
+
"learning_rate": 2.8935185185185186e-05,
|
| 763 |
+
"loss": 0.9815,
|
| 764 |
+
"step": 11600
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 4.85,
|
| 768 |
+
"learning_rate": 2.8703703703703706e-05,
|
| 769 |
+
"loss": 0.9645,
|
| 770 |
+
"step": 11700
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 4.9,
|
| 774 |
+
"learning_rate": 2.8472222222222223e-05,
|
| 775 |
+
"loss": 0.9826,
|
| 776 |
+
"step": 11800
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 4.94,
|
| 780 |
+
"learning_rate": 2.824074074074074e-05,
|
| 781 |
+
"loss": 1.0036,
|
| 782 |
+
"step": 11900
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 4.98,
|
| 786 |
+
"learning_rate": 2.8009259259259263e-05,
|
| 787 |
+
"loss": 0.9994,
|
| 788 |
+
"step": 12000
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 4.98,
|
| 792 |
+
"eval_bleu": 68.0869,
|
| 793 |
+
"eval_em": 0.02,
|
| 794 |
+
"eval_gen_len": 47.76,
|
| 795 |
+
"eval_loss": 1.0576136112213135,
|
| 796 |
+
"eval_rm": 0.8889,
|
| 797 |
+
"eval_runtime": 48.2259,
|
| 798 |
+
"eval_samples_per_second": 1.037,
|
| 799 |
+
"eval_steps_per_second": 0.145,
|
| 800 |
+
"step": 12000
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 5.02,
|
| 804 |
+
"learning_rate": 2.777777777777778e-05,
|
| 805 |
+
"loss": 0.8445,
|
| 806 |
+
"step": 12100
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 5.06,
|
| 810 |
+
"learning_rate": 2.75462962962963e-05,
|
| 811 |
+
"loss": 0.833,
|
| 812 |
+
"step": 12200
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 5.1,
|
| 816 |
+
"learning_rate": 2.7314814814814816e-05,
|
| 817 |
+
"loss": 0.8034,
|
| 818 |
+
"step": 12300
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 5.15,
|
| 822 |
+
"learning_rate": 2.7083333333333332e-05,
|
| 823 |
+
"loss": 0.8154,
|
| 824 |
+
"step": 12400
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 5.19,
|
| 828 |
+
"learning_rate": 2.6851851851851855e-05,
|
| 829 |
+
"loss": 0.8026,
|
| 830 |
+
"step": 12500
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 5.23,
|
| 834 |
+
"learning_rate": 2.6620370370370372e-05,
|
| 835 |
+
"loss": 0.8147,
|
| 836 |
+
"step": 12600
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 5.27,
|
| 840 |
+
"learning_rate": 2.6388888888888892e-05,
|
| 841 |
+
"loss": 0.8466,
|
| 842 |
+
"step": 12700
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 5.31,
|
| 846 |
+
"learning_rate": 2.615740740740741e-05,
|
| 847 |
+
"loss": 0.7881,
|
| 848 |
+
"step": 12800
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 5.35,
|
| 852 |
+
"learning_rate": 2.5925925925925925e-05,
|
| 853 |
+
"loss": 0.7856,
|
| 854 |
+
"step": 12900
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 5.39,
|
| 858 |
+
"learning_rate": 2.5694444444444445e-05,
|
| 859 |
+
"loss": 0.8058,
|
| 860 |
+
"step": 13000
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 5.44,
|
| 864 |
+
"learning_rate": 2.5462962962962965e-05,
|
| 865 |
+
"loss": 0.7704,
|
| 866 |
+
"step": 13100
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 5.48,
|
| 870 |
+
"learning_rate": 2.5231481481481485e-05,
|
| 871 |
+
"loss": 0.7572,
|
| 872 |
+
"step": 13200
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 5.52,
|
| 876 |
+
"learning_rate": 2.5e-05,
|
| 877 |
+
"loss": 0.7757,
|
| 878 |
+
"step": 13300
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 5.56,
|
| 882 |
+
"learning_rate": 2.4768518518518518e-05,
|
| 883 |
+
"loss": 0.7598,
|
| 884 |
+
"step": 13400
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 5.6,
|
| 888 |
+
"learning_rate": 2.4537037037037038e-05,
|
| 889 |
+
"loss": 0.7268,
|
| 890 |
+
"step": 13500
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 5.64,
|
| 894 |
+
"learning_rate": 2.4305555555555558e-05,
|
| 895 |
+
"loss": 0.7455,
|
| 896 |
+
"step": 13600
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 5.68,
|
| 900 |
+
"learning_rate": 2.4074074074074074e-05,
|
| 901 |
+
"loss": 0.7525,
|
| 902 |
+
"step": 13700
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 5.73,
|
| 906 |
+
"learning_rate": 2.3842592592592594e-05,
|
| 907 |
+
"loss": 0.7205,
|
| 908 |
+
"step": 13800
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 5.77,
|
| 912 |
+
"learning_rate": 2.361111111111111e-05,
|
| 913 |
+
"loss": 0.7741,
|
| 914 |
+
"step": 13900
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 5.81,
|
| 918 |
+
"learning_rate": 2.337962962962963e-05,
|
| 919 |
+
"loss": 0.7275,
|
| 920 |
+
"step": 14000
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 5.81,
|
| 924 |
+
"eval_bleu": 74.1032,
|
| 925 |
+
"eval_em": 0.02,
|
| 926 |
+
"eval_gen_len": 46.52,
|
| 927 |
+
"eval_loss": 0.952226996421814,
|
| 928 |
+
"eval_rm": 0.9556,
|
| 929 |
+
"eval_runtime": 45.6891,
|
| 930 |
+
"eval_samples_per_second": 1.094,
|
| 931 |
+
"eval_steps_per_second": 0.153,
|
| 932 |
+
"step": 14000
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 5.85,
|
| 936 |
+
"learning_rate": 2.314814814814815e-05,
|
| 937 |
+
"loss": 0.7429,
|
| 938 |
+
"step": 14100
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 5.89,
|
| 942 |
+
"learning_rate": 2.2916666666666667e-05,
|
| 943 |
+
"loss": 0.7032,
|
| 944 |
+
"step": 14200
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 5.93,
|
| 948 |
+
"learning_rate": 2.2685185185185187e-05,
|
| 949 |
+
"loss": 0.749,
|
| 950 |
+
"step": 14300
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 5.98,
|
| 954 |
+
"learning_rate": 2.2453703703703703e-05,
|
| 955 |
+
"loss": 0.7122,
|
| 956 |
+
"step": 14400
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 6.02,
|
| 960 |
+
"learning_rate": 2.2222222222222223e-05,
|
| 961 |
+
"loss": 0.6559,
|
| 962 |
+
"step": 14500
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 6.06,
|
| 966 |
+
"learning_rate": 2.1990740740740743e-05,
|
| 967 |
+
"loss": 0.5757,
|
| 968 |
+
"step": 14600
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 6.1,
|
| 972 |
+
"learning_rate": 2.175925925925926e-05,
|
| 973 |
+
"loss": 0.5756,
|
| 974 |
+
"step": 14700
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 6.14,
|
| 978 |
+
"learning_rate": 2.152777777777778e-05,
|
| 979 |
+
"loss": 0.5777,
|
| 980 |
+
"step": 14800
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 6.18,
|
| 984 |
+
"learning_rate": 2.1296296296296296e-05,
|
| 985 |
+
"loss": 0.5904,
|
| 986 |
+
"step": 14900
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 6.22,
|
| 990 |
+
"learning_rate": 2.1064814814814816e-05,
|
| 991 |
+
"loss": 0.5798,
|
| 992 |
+
"step": 15000
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 6.27,
|
| 996 |
+
"learning_rate": 2.0833333333333336e-05,
|
| 997 |
+
"loss": 0.5939,
|
| 998 |
+
"step": 15100
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 6.31,
|
| 1002 |
+
"learning_rate": 2.0601851851851853e-05,
|
| 1003 |
+
"loss": 0.6095,
|
| 1004 |
+
"step": 15200
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 6.35,
|
| 1008 |
+
"learning_rate": 2.037037037037037e-05,
|
| 1009 |
+
"loss": 0.6163,
|
| 1010 |
+
"step": 15300
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 6.39,
|
| 1014 |
+
"learning_rate": 2.013888888888889e-05,
|
| 1015 |
+
"loss": 0.5634,
|
| 1016 |
+
"step": 15400
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 6.43,
|
| 1020 |
+
"learning_rate": 1.990740740740741e-05,
|
| 1021 |
+
"loss": 0.5909,
|
| 1022 |
+
"step": 15500
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 6.47,
|
| 1026 |
+
"learning_rate": 1.967592592592593e-05,
|
| 1027 |
+
"loss": 0.5408,
|
| 1028 |
+
"step": 15600
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 6.51,
|
| 1032 |
+
"learning_rate": 1.9444444444444445e-05,
|
| 1033 |
+
"loss": 0.5435,
|
| 1034 |
+
"step": 15700
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 6.56,
|
| 1038 |
+
"learning_rate": 1.9212962962962962e-05,
|
| 1039 |
+
"loss": 0.5753,
|
| 1040 |
+
"step": 15800
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 6.6,
|
| 1044 |
+
"learning_rate": 1.8981481481481482e-05,
|
| 1045 |
+
"loss": 0.5491,
|
| 1046 |
+
"step": 15900
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 6.64,
|
| 1050 |
+
"learning_rate": 1.8750000000000002e-05,
|
| 1051 |
+
"loss": 0.5868,
|
| 1052 |
+
"step": 16000
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 6.64,
|
| 1056 |
+
"eval_bleu": 71.6124,
|
| 1057 |
+
"eval_em": 0.02,
|
| 1058 |
+
"eval_gen_len": 47.52,
|
| 1059 |
+
"eval_loss": 0.9307076930999756,
|
| 1060 |
+
"eval_rm": 0.9556,
|
| 1061 |
+
"eval_runtime": 330.299,
|
| 1062 |
+
"eval_samples_per_second": 0.151,
|
| 1063 |
+
"eval_steps_per_second": 0.021,
|
| 1064 |
+
"step": 16000
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 6.68,
|
| 1068 |
+
"learning_rate": 1.8518518518518518e-05,
|
| 1069 |
+
"loss": 0.5088,
|
| 1070 |
+
"step": 16100
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 6.72,
|
| 1074 |
+
"learning_rate": 1.8287037037037038e-05,
|
| 1075 |
+
"loss": 0.6068,
|
| 1076 |
+
"step": 16200
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 6.76,
|
| 1080 |
+
"learning_rate": 1.8055555555555555e-05,
|
| 1081 |
+
"loss": 0.5327,
|
| 1082 |
+
"step": 16300
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 6.8,
|
| 1086 |
+
"learning_rate": 1.7824074074074075e-05,
|
| 1087 |
+
"loss": 0.5759,
|
| 1088 |
+
"step": 16400
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 6.85,
|
| 1092 |
+
"learning_rate": 1.7592592592592595e-05,
|
| 1093 |
+
"loss": 0.5849,
|
| 1094 |
+
"step": 16500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 6.89,
|
| 1098 |
+
"learning_rate": 1.736111111111111e-05,
|
| 1099 |
+
"loss": 0.5648,
|
| 1100 |
+
"step": 16600
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 6.93,
|
| 1104 |
+
"learning_rate": 1.712962962962963e-05,
|
| 1105 |
+
"loss": 0.5632,
|
| 1106 |
+
"step": 16700
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 6.97,
|
| 1110 |
+
"learning_rate": 1.6898148148148148e-05,
|
| 1111 |
+
"loss": 0.5093,
|
| 1112 |
+
"step": 16800
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 7.01,
|
| 1116 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 1117 |
+
"loss": 0.5295,
|
| 1118 |
+
"step": 16900
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 7.05,
|
| 1122 |
+
"learning_rate": 1.6435185185185187e-05,
|
| 1123 |
+
"loss": 0.433,
|
| 1124 |
+
"step": 17000
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 7.1,
|
| 1128 |
+
"learning_rate": 1.6203703703703704e-05,
|
| 1129 |
+
"loss": 0.4289,
|
| 1130 |
+
"step": 17100
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 7.14,
|
| 1134 |
+
"learning_rate": 1.597222222222222e-05,
|
| 1135 |
+
"loss": 0.4619,
|
| 1136 |
+
"step": 17200
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 7.18,
|
| 1140 |
+
"learning_rate": 1.574074074074074e-05,
|
| 1141 |
+
"loss": 0.458,
|
| 1142 |
+
"step": 17300
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 7.22,
|
| 1146 |
+
"learning_rate": 1.550925925925926e-05,
|
| 1147 |
+
"loss": 0.4155,
|
| 1148 |
+
"step": 17400
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 7.26,
|
| 1152 |
+
"learning_rate": 1.527777777777778e-05,
|
| 1153 |
+
"loss": 0.4391,
|
| 1154 |
+
"step": 17500
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 7.3,
|
| 1158 |
+
"learning_rate": 1.5046296296296297e-05,
|
| 1159 |
+
"loss": 0.4264,
|
| 1160 |
+
"step": 17600
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 7.34,
|
| 1164 |
+
"learning_rate": 1.4814814814814815e-05,
|
| 1165 |
+
"loss": 0.418,
|
| 1166 |
+
"step": 17700
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 7.39,
|
| 1170 |
+
"learning_rate": 1.4583333333333335e-05,
|
| 1171 |
+
"loss": 0.4285,
|
| 1172 |
+
"step": 17800
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 7.43,
|
| 1176 |
+
"learning_rate": 1.4351851851851853e-05,
|
| 1177 |
+
"loss": 0.4466,
|
| 1178 |
+
"step": 17900
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 7.47,
|
| 1182 |
+
"learning_rate": 1.412037037037037e-05,
|
| 1183 |
+
"loss": 0.4499,
|
| 1184 |
+
"step": 18000
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 7.47,
|
| 1188 |
+
"eval_bleu": 77.237,
|
| 1189 |
+
"eval_em": 0.06,
|
| 1190 |
+
"eval_gen_len": 46.0,
|
| 1191 |
+
"eval_loss": 0.8866045475006104,
|
| 1192 |
+
"eval_rm": 0.9574,
|
| 1193 |
+
"eval_runtime": 280.5418,
|
| 1194 |
+
"eval_samples_per_second": 0.178,
|
| 1195 |
+
"eval_steps_per_second": 0.025,
|
| 1196 |
+
"step": 18000
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 7.51,
|
| 1200 |
+
"learning_rate": 1.388888888888889e-05,
|
| 1201 |
+
"loss": 0.4415,
|
| 1202 |
+
"step": 18100
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 7.55,
|
| 1206 |
+
"learning_rate": 1.3657407407407408e-05,
|
| 1207 |
+
"loss": 0.4209,
|
| 1208 |
+
"step": 18200
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 7.59,
|
| 1212 |
+
"learning_rate": 1.3425925925925928e-05,
|
| 1213 |
+
"loss": 0.4357,
|
| 1214 |
+
"step": 18300
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 7.63,
|
| 1218 |
+
"learning_rate": 1.3194444444444446e-05,
|
| 1219 |
+
"loss": 0.437,
|
| 1220 |
+
"step": 18400
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 7.68,
|
| 1224 |
+
"learning_rate": 1.2962962962962962e-05,
|
| 1225 |
+
"loss": 0.4319,
|
| 1226 |
+
"step": 18500
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 7.72,
|
| 1230 |
+
"learning_rate": 1.2731481481481482e-05,
|
| 1231 |
+
"loss": 0.4578,
|
| 1232 |
+
"step": 18600
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 7.76,
|
| 1236 |
+
"learning_rate": 1.25e-05,
|
| 1237 |
+
"loss": 0.4244,
|
| 1238 |
+
"step": 18700
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 7.8,
|
| 1242 |
+
"learning_rate": 1.2268518518518519e-05,
|
| 1243 |
+
"loss": 0.4046,
|
| 1244 |
+
"step": 18800
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 7.84,
|
| 1248 |
+
"learning_rate": 1.2037037037037037e-05,
|
| 1249 |
+
"loss": 0.411,
|
| 1250 |
+
"step": 18900
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 7.88,
|
| 1254 |
+
"learning_rate": 1.1805555555555555e-05,
|
| 1255 |
+
"loss": 0.4219,
|
| 1256 |
+
"step": 19000
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 7.93,
|
| 1260 |
+
"learning_rate": 1.1574074074074075e-05,
|
| 1261 |
+
"loss": 0.3956,
|
| 1262 |
+
"step": 19100
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 7.97,
|
| 1266 |
+
"learning_rate": 1.1342592592592593e-05,
|
| 1267 |
+
"loss": 0.4333,
|
| 1268 |
+
"step": 19200
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 8.01,
|
| 1272 |
+
"learning_rate": 1.1111111111111112e-05,
|
| 1273 |
+
"loss": 0.4141,
|
| 1274 |
+
"step": 19300
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 8.05,
|
| 1278 |
+
"learning_rate": 1.087962962962963e-05,
|
| 1279 |
+
"loss": 0.3199,
|
| 1280 |
+
"step": 19400
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 8.09,
|
| 1284 |
+
"learning_rate": 1.0648148148148148e-05,
|
| 1285 |
+
"loss": 0.3236,
|
| 1286 |
+
"step": 19500
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 8.13,
|
| 1290 |
+
"learning_rate": 1.0416666666666668e-05,
|
| 1291 |
+
"loss": 0.3405,
|
| 1292 |
+
"step": 19600
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 8.17,
|
| 1296 |
+
"learning_rate": 1.0185185185185185e-05,
|
| 1297 |
+
"loss": 0.3411,
|
| 1298 |
+
"step": 19700
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 8.22,
|
| 1302 |
+
"learning_rate": 9.953703703703704e-06,
|
| 1303 |
+
"loss": 0.3166,
|
| 1304 |
+
"step": 19800
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 8.26,
|
| 1308 |
+
"learning_rate": 9.722222222222223e-06,
|
| 1309 |
+
"loss": 0.3463,
|
| 1310 |
+
"step": 19900
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 8.3,
|
| 1314 |
+
"learning_rate": 9.490740740740741e-06,
|
| 1315 |
+
"loss": 0.3515,
|
| 1316 |
+
"step": 20000
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 8.3,
|
| 1320 |
+
"eval_bleu": 77.5798,
|
| 1321 |
+
"eval_em": 0.08,
|
| 1322 |
+
"eval_gen_len": 47.5,
|
| 1323 |
+
"eval_loss": 0.9069581627845764,
|
| 1324 |
+
"eval_rm": 0.9574,
|
| 1325 |
+
"eval_runtime": 312.8421,
|
| 1326 |
+
"eval_samples_per_second": 0.16,
|
| 1327 |
+
"eval_steps_per_second": 0.022,
|
| 1328 |
+
"step": 20000
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 8.34,
|
| 1332 |
+
"learning_rate": 9.259259259259259e-06,
|
| 1333 |
+
"loss": 0.3363,
|
| 1334 |
+
"step": 20100
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 8.38,
|
| 1338 |
+
"learning_rate": 9.027777777777777e-06,
|
| 1339 |
+
"loss": 0.3477,
|
| 1340 |
+
"step": 20200
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 8.42,
|
| 1344 |
+
"learning_rate": 8.796296296296297e-06,
|
| 1345 |
+
"loss": 0.3454,
|
| 1346 |
+
"step": 20300
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 8.46,
|
| 1350 |
+
"learning_rate": 8.564814814814816e-06,
|
| 1351 |
+
"loss": 0.3267,
|
| 1352 |
+
"step": 20400
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 8.51,
|
| 1356 |
+
"learning_rate": 8.333333333333334e-06,
|
| 1357 |
+
"loss": 0.3406,
|
| 1358 |
+
"step": 20500
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 8.55,
|
| 1362 |
+
"learning_rate": 8.101851851851852e-06,
|
| 1363 |
+
"loss": 0.3189,
|
| 1364 |
+
"step": 20600
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 8.59,
|
| 1368 |
+
"learning_rate": 7.87037037037037e-06,
|
| 1369 |
+
"loss": 0.3355,
|
| 1370 |
+
"step": 20700
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 8.63,
|
| 1374 |
+
"learning_rate": 7.63888888888889e-06,
|
| 1375 |
+
"loss": 0.3315,
|
| 1376 |
+
"step": 20800
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 8.67,
|
| 1380 |
+
"learning_rate": 7.4074074074074075e-06,
|
| 1381 |
+
"loss": 0.345,
|
| 1382 |
+
"step": 20900
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 8.71,
|
| 1386 |
+
"learning_rate": 7.1759259259259266e-06,
|
| 1387 |
+
"loss": 0.3332,
|
| 1388 |
+
"step": 21000
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 8.76,
|
| 1392 |
+
"learning_rate": 6.944444444444445e-06,
|
| 1393 |
+
"loss": 0.3491,
|
| 1394 |
+
"step": 21100
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 8.8,
|
| 1398 |
+
"learning_rate": 6.712962962962964e-06,
|
| 1399 |
+
"loss": 0.3268,
|
| 1400 |
+
"step": 21200
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 8.84,
|
| 1404 |
+
"learning_rate": 6.481481481481481e-06,
|
| 1405 |
+
"loss": 0.3328,
|
| 1406 |
+
"step": 21300
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 8.88,
|
| 1410 |
+
"learning_rate": 6.25e-06,
|
| 1411 |
+
"loss": 0.3151,
|
| 1412 |
+
"step": 21400
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 8.92,
|
| 1416 |
+
"learning_rate": 6.0185185185185185e-06,
|
| 1417 |
+
"loss": 0.3529,
|
| 1418 |
+
"step": 21500
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 8.96,
|
| 1422 |
+
"learning_rate": 5.787037037037038e-06,
|
| 1423 |
+
"loss": 0.3087,
|
| 1424 |
+
"step": 21600
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 9.0,
|
| 1428 |
+
"learning_rate": 5.555555555555556e-06,
|
| 1429 |
+
"loss": 0.3228,
|
| 1430 |
+
"step": 21700
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 9.05,
|
| 1434 |
+
"learning_rate": 5.324074074074074e-06,
|
| 1435 |
+
"loss": 0.2613,
|
| 1436 |
+
"step": 21800
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 9.09,
|
| 1440 |
+
"learning_rate": 5.092592592592592e-06,
|
| 1441 |
+
"loss": 0.2721,
|
| 1442 |
+
"step": 21900
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 9.13,
|
| 1446 |
+
"learning_rate": 4.861111111111111e-06,
|
| 1447 |
+
"loss": 0.292,
|
| 1448 |
+
"step": 22000
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 9.13,
|
| 1452 |
+
"eval_bleu": 78.649,
|
| 1453 |
+
"eval_em": 0.06,
|
| 1454 |
+
"eval_gen_len": 47.96,
|
| 1455 |
+
"eval_loss": 0.8904868364334106,
|
| 1456 |
+
"eval_rm": 0.9574,
|
| 1457 |
+
"eval_runtime": 309.7604,
|
| 1458 |
+
"eval_samples_per_second": 0.161,
|
| 1459 |
+
"eval_steps_per_second": 0.023,
|
| 1460 |
+
"step": 22000
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 9.17,
|
| 1464 |
+
"learning_rate": 4.6296296296296296e-06,
|
| 1465 |
+
"loss": 0.2614,
|
| 1466 |
+
"step": 22100
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 9.21,
|
| 1470 |
+
"learning_rate": 4.398148148148149e-06,
|
| 1471 |
+
"loss": 0.2522,
|
| 1472 |
+
"step": 22200
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 9.25,
|
| 1476 |
+
"learning_rate": 4.166666666666667e-06,
|
| 1477 |
+
"loss": 0.289,
|
| 1478 |
+
"step": 22300
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 9.29,
|
| 1482 |
+
"learning_rate": 3.935185185185185e-06,
|
| 1483 |
+
"loss": 0.2875,
|
| 1484 |
+
"step": 22400
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 9.34,
|
| 1488 |
+
"learning_rate": 3.7037037037037037e-06,
|
| 1489 |
+
"loss": 0.2451,
|
| 1490 |
+
"step": 22500
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 9.38,
|
| 1494 |
+
"learning_rate": 3.4722222222222224e-06,
|
| 1495 |
+
"loss": 0.2855,
|
| 1496 |
+
"step": 22600
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 9.42,
|
| 1500 |
+
"learning_rate": 3.2407407407407406e-06,
|
| 1501 |
+
"loss": 0.2632,
|
| 1502 |
+
"step": 22700
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 9.46,
|
| 1506 |
+
"learning_rate": 3.0092592592592593e-06,
|
| 1507 |
+
"loss": 0.2726,
|
| 1508 |
+
"step": 22800
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 9.5,
|
| 1512 |
+
"learning_rate": 2.777777777777778e-06,
|
| 1513 |
+
"loss": 0.2768,
|
| 1514 |
+
"step": 22900
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 9.54,
|
| 1518 |
+
"learning_rate": 2.546296296296296e-06,
|
| 1519 |
+
"loss": 0.2729,
|
| 1520 |
+
"step": 23000
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 9.59,
|
| 1524 |
+
"learning_rate": 2.3148148148148148e-06,
|
| 1525 |
+
"loss": 0.2809,
|
| 1526 |
+
"step": 23100
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 9.63,
|
| 1530 |
+
"learning_rate": 2.0833333333333334e-06,
|
| 1531 |
+
"loss": 0.2714,
|
| 1532 |
+
"step": 23200
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 9.67,
|
| 1536 |
+
"learning_rate": 1.8518518518518519e-06,
|
| 1537 |
+
"loss": 0.2693,
|
| 1538 |
+
"step": 23300
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 9.71,
|
| 1542 |
+
"learning_rate": 1.6203703703703703e-06,
|
| 1543 |
+
"loss": 0.2639,
|
| 1544 |
+
"step": 23400
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 9.75,
|
| 1548 |
+
"learning_rate": 1.388888888888889e-06,
|
| 1549 |
+
"loss": 0.2667,
|
| 1550 |
+
"step": 23500
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 9.79,
|
| 1554 |
+
"learning_rate": 1.1574074074074074e-06,
|
| 1555 |
+
"loss": 0.2565,
|
| 1556 |
+
"step": 23600
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 9.83,
|
| 1560 |
+
"learning_rate": 9.259259259259259e-07,
|
| 1561 |
+
"loss": 0.2855,
|
| 1562 |
+
"step": 23700
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 9.88,
|
| 1566 |
+
"learning_rate": 6.944444444444445e-07,
|
| 1567 |
+
"loss": 0.2615,
|
| 1568 |
+
"step": 23800
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 9.92,
|
| 1572 |
+
"learning_rate": 4.6296296296296297e-07,
|
| 1573 |
+
"loss": 0.2709,
|
| 1574 |
+
"step": 23900
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 9.96,
|
| 1578 |
+
"learning_rate": 2.3148148148148148e-07,
|
| 1579 |
+
"loss": 0.2658,
|
| 1580 |
+
"step": 24000
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 9.96,
|
| 1584 |
+
"eval_bleu": 79.5906,
|
| 1585 |
+
"eval_em": 0.08,
|
| 1586 |
+
"eval_gen_len": 48.48,
|
| 1587 |
+
"eval_loss": 0.9002240896224976,
|
| 1588 |
+
"eval_rm": 0.9565,
|
| 1589 |
+
"eval_runtime": 351.0056,
|
| 1590 |
+
"eval_samples_per_second": 0.142,
|
| 1591 |
+
"eval_steps_per_second": 0.02,
|
| 1592 |
+
"step": 24000
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 10.0,
|
| 1596 |
+
"learning_rate": 0.0,
|
| 1597 |
+
"loss": 0.2602,
|
| 1598 |
+
"step": 24100
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 10.0,
|
| 1602 |
+
"step": 24100,
|
| 1603 |
+
"total_flos": 7011428612298048.0,
|
| 1604 |
+
"train_loss": 0.1476078914309933,
|
| 1605 |
+
"train_runtime": 5095.148,
|
| 1606 |
+
"train_samples_per_second": 37.832,
|
| 1607 |
+
"train_steps_per_second": 4.73
|
| 1608 |
+
}
|
| 1609 |
+
],
|
| 1610 |
+
"max_steps": 24100,
|
| 1611 |
+
"num_train_epochs": 10,
|
| 1612 |
+
"total_flos": 7011428612298048.0,
|
| 1613 |
+
"trial_name": null,
|
| 1614 |
+
"trial_params": null
|
| 1615 |
+
}
|