End of training
Browse files
all_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss":
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 9.096017699115045,
|
| 3 |
+
"eval_accuracy": 0.7652173913043478,
|
| 4 |
+
"eval_loss": 0.6633358597755432,
|
| 5 |
+
"eval_runtime": 903.9019,
|
| 6 |
+
"eval_samples_per_second": 3.053,
|
| 7 |
+
"eval_steps_per_second": 0.096
|
| 8 |
}
|
runs/Feb13_14-20-30_DESKTOP-T04IOFP/events.out.tfevents.1739484113.DESKTOP-T04IOFP.18948.1
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7b2e7e3e087679270790296c73aa48b295a4bed2330633c7469e2f885d7654aa
|
| 3 |
+
size 734
|
test_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss":
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 9.096017699115045,
|
| 3 |
+
"eval_accuracy": 0.7652173913043478,
|
| 4 |
+
"eval_loss": 0.6633358597755432,
|
| 5 |
+
"eval_runtime": 903.9019,
|
| 6 |
+
"eval_samples_per_second": 3.053,
|
| 7 |
+
"eval_steps_per_second": 0.096
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,50 +1,1715 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "videomae-small-finetuned-kinetics-finetuned-2\\checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch":
|
| 13 |
-
"grad_norm": 2.
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss": 1.
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch":
|
| 20 |
-
"
|
| 21 |
-
"
|
| 22 |
-
"
|
| 23 |
-
"
|
| 24 |
-
"eval_steps_per_second": 0.217,
|
| 25 |
-
"step": 10
|
| 26 |
},
|
| 27 |
{
|
| 28 |
-
"epoch":
|
| 29 |
-
"
|
| 30 |
-
"
|
| 31 |
-
"
|
| 32 |
-
"
|
| 33 |
-
"train_samples_per_second": 0.104,
|
| 34 |
-
"train_steps_per_second": 0.009
|
| 35 |
},
|
| 36 |
{
|
| 37 |
-
"epoch":
|
| 38 |
-
"
|
| 39 |
-
"
|
| 40 |
-
"
|
| 41 |
-
"
|
| 42 |
-
|
| 43 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 44 |
}
|
| 45 |
],
|
| 46 |
"logging_steps": 10,
|
| 47 |
-
"max_steps":
|
| 48 |
"num_input_tokens_seen": 0,
|
| 49 |
"num_train_epochs": 9223372036854775807,
|
| 50 |
"save_steps": 500,
|
|
@@ -60,8 +1725,8 @@
|
|
| 60 |
"attributes": {}
|
| 61 |
}
|
| 62 |
},
|
| 63 |
-
"total_flos":
|
| 64 |
-
"train_batch_size":
|
| 65 |
"trial_name": null,
|
| 66 |
"trial_params": null
|
| 67 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.7665847665847666,
|
| 3 |
+
"best_model_checkpoint": "videomae-small-finetuned-kinetics-finetuned-2\\checkpoint-2260",
|
| 4 |
+
"epoch": 9.096017699115045,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 2260,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.004424778761061947,
|
| 13 |
+
"grad_norm": 2.1747984886169434,
|
| 14 |
+
"learning_rate": 2.2123893805309734e-06,
|
| 15 |
+
"loss": 1.0543,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.008849557522123894,
|
| 20 |
+
"grad_norm": 1.4438425302505493,
|
| 21 |
+
"learning_rate": 4.424778761061947e-06,
|
| 22 |
+
"loss": 1.0544,
|
| 23 |
+
"step": 20
|
|
|
|
|
|
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.01327433628318584,
|
| 27 |
+
"grad_norm": 1.5373220443725586,
|
| 28 |
+
"learning_rate": 6.6371681415929215e-06,
|
| 29 |
+
"loss": 1.0551,
|
| 30 |
+
"step": 30
|
|
|
|
|
|
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 0.017699115044247787,
|
| 34 |
+
"grad_norm": 1.477241039276123,
|
| 35 |
+
"learning_rate": 8.849557522123894e-06,
|
| 36 |
+
"loss": 1.0596,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.022123893805309734,
|
| 41 |
+
"grad_norm": 1.7869963645935059,
|
| 42 |
+
"learning_rate": 1.1061946902654869e-05,
|
| 43 |
+
"loss": 1.0579,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.02654867256637168,
|
| 48 |
+
"grad_norm": 1.406799554824829,
|
| 49 |
+
"learning_rate": 1.3274336283185843e-05,
|
| 50 |
+
"loss": 1.064,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.030973451327433628,
|
| 55 |
+
"grad_norm": 1.6087795495986938,
|
| 56 |
+
"learning_rate": 1.5486725663716813e-05,
|
| 57 |
+
"loss": 1.0334,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.035398230088495575,
|
| 62 |
+
"grad_norm": 1.539732575416565,
|
| 63 |
+
"learning_rate": 1.7699115044247787e-05,
|
| 64 |
+
"loss": 1.0502,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.03982300884955752,
|
| 69 |
+
"grad_norm": 1.7329670190811157,
|
| 70 |
+
"learning_rate": 1.991150442477876e-05,
|
| 71 |
+
"loss": 1.0408,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.04424778761061947,
|
| 76 |
+
"grad_norm": 1.849195957183838,
|
| 77 |
+
"learning_rate": 2.2123893805309738e-05,
|
| 78 |
+
"loss": 1.0403,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.048672566371681415,
|
| 83 |
+
"grad_norm": 1.7182987928390503,
|
| 84 |
+
"learning_rate": 2.433628318584071e-05,
|
| 85 |
+
"loss": 1.0406,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.05309734513274336,
|
| 90 |
+
"grad_norm": 1.8291207551956177,
|
| 91 |
+
"learning_rate": 2.6548672566371686e-05,
|
| 92 |
+
"loss": 1.0188,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.05752212389380531,
|
| 97 |
+
"grad_norm": 1.8635715246200562,
|
| 98 |
+
"learning_rate": 2.8761061946902656e-05,
|
| 99 |
+
"loss": 1.0132,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.061946902654867256,
|
| 104 |
+
"grad_norm": 2.1121060848236084,
|
| 105 |
+
"learning_rate": 3.097345132743363e-05,
|
| 106 |
+
"loss": 1.0194,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.06637168141592921,
|
| 111 |
+
"grad_norm": 1.605365514755249,
|
| 112 |
+
"learning_rate": 3.3185840707964604e-05,
|
| 113 |
+
"loss": 1.0276,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.07079646017699115,
|
| 118 |
+
"grad_norm": 1.4162675142288208,
|
| 119 |
+
"learning_rate": 3.5398230088495574e-05,
|
| 120 |
+
"loss": 1.0155,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.0752212389380531,
|
| 125 |
+
"grad_norm": 2.037018060684204,
|
| 126 |
+
"learning_rate": 3.7610619469026545e-05,
|
| 127 |
+
"loss": 0.9921,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.07964601769911504,
|
| 132 |
+
"grad_norm": 1.6612162590026855,
|
| 133 |
+
"learning_rate": 3.982300884955752e-05,
|
| 134 |
+
"loss": 0.9814,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.084070796460177,
|
| 139 |
+
"grad_norm": 1.4722100496292114,
|
| 140 |
+
"learning_rate": 4.20353982300885e-05,
|
| 141 |
+
"loss": 1.0092,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.08849557522123894,
|
| 146 |
+
"grad_norm": 1.9917899370193481,
|
| 147 |
+
"learning_rate": 4.4247787610619477e-05,
|
| 148 |
+
"loss": 0.9742,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.09292035398230089,
|
| 153 |
+
"grad_norm": 1.5193358659744263,
|
| 154 |
+
"learning_rate": 4.646017699115045e-05,
|
| 155 |
+
"loss": 0.9647,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.09734513274336283,
|
| 160 |
+
"grad_norm": 1.4704391956329346,
|
| 161 |
+
"learning_rate": 4.867256637168142e-05,
|
| 162 |
+
"loss": 0.9644,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.10044247787610619,
|
| 167 |
+
"eval_accuracy": 0.5777465777465778,
|
| 168 |
+
"eval_loss": 0.9732658863067627,
|
| 169 |
+
"eval_runtime": 1124.1661,
|
| 170 |
+
"eval_samples_per_second": 2.534,
|
| 171 |
+
"eval_steps_per_second": 0.08,
|
| 172 |
+
"step": 227
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 1.0013274336283187,
|
| 176 |
+
"grad_norm": 1.666140079498291,
|
| 177 |
+
"learning_rate": 4.990167158308752e-05,
|
| 178 |
+
"loss": 0.967,
|
| 179 |
+
"step": 230
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 1.0057522123893805,
|
| 183 |
+
"grad_norm": 1.6128636598587036,
|
| 184 |
+
"learning_rate": 4.9655850540806295e-05,
|
| 185 |
+
"loss": 0.9457,
|
| 186 |
+
"step": 240
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 1.0101769911504426,
|
| 190 |
+
"grad_norm": 1.2806826829910278,
|
| 191 |
+
"learning_rate": 4.941002949852507e-05,
|
| 192 |
+
"loss": 0.9482,
|
| 193 |
+
"step": 250
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 1.0146017699115044,
|
| 197 |
+
"grad_norm": 1.5180962085723877,
|
| 198 |
+
"learning_rate": 4.9164208456243856e-05,
|
| 199 |
+
"loss": 0.9656,
|
| 200 |
+
"step": 260
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.0190265486725665,
|
| 204 |
+
"grad_norm": 1.306519627571106,
|
| 205 |
+
"learning_rate": 4.891838741396263e-05,
|
| 206 |
+
"loss": 0.9615,
|
| 207 |
+
"step": 270
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 1.0234513274336283,
|
| 211 |
+
"grad_norm": 1.5663535594940186,
|
| 212 |
+
"learning_rate": 4.867256637168142e-05,
|
| 213 |
+
"loss": 0.9306,
|
| 214 |
+
"step": 280
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 1.0278761061946902,
|
| 218 |
+
"grad_norm": 1.3962496519088745,
|
| 219 |
+
"learning_rate": 4.8426745329400195e-05,
|
| 220 |
+
"loss": 0.9359,
|
| 221 |
+
"step": 290
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 1.0323008849557522,
|
| 225 |
+
"grad_norm": 1.3210822343826294,
|
| 226 |
+
"learning_rate": 4.818092428711898e-05,
|
| 227 |
+
"loss": 0.8929,
|
| 228 |
+
"step": 300
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 1.036725663716814,
|
| 232 |
+
"grad_norm": 1.5808290243148804,
|
| 233 |
+
"learning_rate": 4.7935103244837756e-05,
|
| 234 |
+
"loss": 0.8904,
|
| 235 |
+
"step": 310
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 1.0411504424778761,
|
| 239 |
+
"grad_norm": 1.4897513389587402,
|
| 240 |
+
"learning_rate": 4.768928220255654e-05,
|
| 241 |
+
"loss": 0.9128,
|
| 242 |
+
"step": 320
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.045575221238938,
|
| 246 |
+
"grad_norm": 1.4278603792190552,
|
| 247 |
+
"learning_rate": 4.7443461160275324e-05,
|
| 248 |
+
"loss": 0.9177,
|
| 249 |
+
"step": 330
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.05,
|
| 253 |
+
"grad_norm": 2.067490577697754,
|
| 254 |
+
"learning_rate": 4.71976401179941e-05,
|
| 255 |
+
"loss": 0.8882,
|
| 256 |
+
"step": 340
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 1.0544247787610619,
|
| 260 |
+
"grad_norm": 1.3919837474822998,
|
| 261 |
+
"learning_rate": 4.6951819075712886e-05,
|
| 262 |
+
"loss": 0.8824,
|
| 263 |
+
"step": 350
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 1.058849557522124,
|
| 267 |
+
"grad_norm": 1.398645281791687,
|
| 268 |
+
"learning_rate": 4.670599803343166e-05,
|
| 269 |
+
"loss": 0.905,
|
| 270 |
+
"step": 360
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 1.0632743362831858,
|
| 274 |
+
"grad_norm": 1.4657193422317505,
|
| 275 |
+
"learning_rate": 4.646017699115045e-05,
|
| 276 |
+
"loss": 0.8687,
|
| 277 |
+
"step": 370
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 1.0676991150442479,
|
| 281 |
+
"grad_norm": 1.2234872579574585,
|
| 282 |
+
"learning_rate": 4.6214355948869224e-05,
|
| 283 |
+
"loss": 0.8918,
|
| 284 |
+
"step": 380
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.0721238938053097,
|
| 288 |
+
"grad_norm": 1.321500539779663,
|
| 289 |
+
"learning_rate": 4.596853490658801e-05,
|
| 290 |
+
"loss": 0.8582,
|
| 291 |
+
"step": 390
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 1.0765486725663718,
|
| 295 |
+
"grad_norm": 1.3090804815292358,
|
| 296 |
+
"learning_rate": 4.5722713864306786e-05,
|
| 297 |
+
"loss": 0.868,
|
| 298 |
+
"step": 400
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 1.0809734513274336,
|
| 302 |
+
"grad_norm": 1.2257106304168701,
|
| 303 |
+
"learning_rate": 4.547689282202557e-05,
|
| 304 |
+
"loss": 0.8837,
|
| 305 |
+
"step": 410
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.0853982300884957,
|
| 309 |
+
"grad_norm": 1.2699074745178223,
|
| 310 |
+
"learning_rate": 4.523107177974435e-05,
|
| 311 |
+
"loss": 0.8863,
|
| 312 |
+
"step": 420
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 1.0898230088495575,
|
| 316 |
+
"grad_norm": 1.791901707649231,
|
| 317 |
+
"learning_rate": 4.498525073746313e-05,
|
| 318 |
+
"loss": 0.8883,
|
| 319 |
+
"step": 430
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 1.0942477876106196,
|
| 323 |
+
"grad_norm": 1.2082570791244507,
|
| 324 |
+
"learning_rate": 4.473942969518191e-05,
|
| 325 |
+
"loss": 0.8544,
|
| 326 |
+
"step": 440
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.0986725663716814,
|
| 330 |
+
"grad_norm": 1.3323192596435547,
|
| 331 |
+
"learning_rate": 4.449360865290069e-05,
|
| 332 |
+
"loss": 0.8823,
|
| 333 |
+
"step": 450
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.1004424778761062,
|
| 337 |
+
"eval_accuracy": 0.6676026676026676,
|
| 338 |
+
"eval_loss": 0.869060754776001,
|
| 339 |
+
"eval_runtime": 1028.2799,
|
| 340 |
+
"eval_samples_per_second": 2.771,
|
| 341 |
+
"eval_steps_per_second": 0.088,
|
| 342 |
+
"step": 454
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 2.0026548672566373,
|
| 346 |
+
"grad_norm": 1.2703664302825928,
|
| 347 |
+
"learning_rate": 4.4247787610619477e-05,
|
| 348 |
+
"loss": 0.846,
|
| 349 |
+
"step": 460
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 2.007079646017699,
|
| 353 |
+
"grad_norm": 1.2533842325210571,
|
| 354 |
+
"learning_rate": 4.4001966568338254e-05,
|
| 355 |
+
"loss": 0.8445,
|
| 356 |
+
"step": 470
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 2.011504424778761,
|
| 360 |
+
"grad_norm": 1.2655799388885498,
|
| 361 |
+
"learning_rate": 4.375614552605704e-05,
|
| 362 |
+
"loss": 0.8526,
|
| 363 |
+
"step": 480
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 2.015929203539823,
|
| 367 |
+
"grad_norm": 1.3053220510482788,
|
| 368 |
+
"learning_rate": 4.351032448377581e-05,
|
| 369 |
+
"loss": 0.8576,
|
| 370 |
+
"step": 490
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 2.020353982300885,
|
| 374 |
+
"grad_norm": 1.160129189491272,
|
| 375 |
+
"learning_rate": 4.326450344149459e-05,
|
| 376 |
+
"loss": 0.8595,
|
| 377 |
+
"step": 500
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 2.0247787610619468,
|
| 381 |
+
"grad_norm": 1.4154675006866455,
|
| 382 |
+
"learning_rate": 4.301868239921337e-05,
|
| 383 |
+
"loss": 0.8563,
|
| 384 |
+
"step": 510
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 2.029203539823009,
|
| 388 |
+
"grad_norm": 1.166509747505188,
|
| 389 |
+
"learning_rate": 4.2772861356932154e-05,
|
| 390 |
+
"loss": 0.8431,
|
| 391 |
+
"step": 520
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 2.033628318584071,
|
| 395 |
+
"grad_norm": 1.4635576009750366,
|
| 396 |
+
"learning_rate": 4.252704031465093e-05,
|
| 397 |
+
"loss": 0.8099,
|
| 398 |
+
"step": 530
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 2.038053097345133,
|
| 402 |
+
"grad_norm": 1.205285906791687,
|
| 403 |
+
"learning_rate": 4.2281219272369715e-05,
|
| 404 |
+
"loss": 0.8308,
|
| 405 |
+
"step": 540
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 2.0424778761061946,
|
| 409 |
+
"grad_norm": 1.521293044090271,
|
| 410 |
+
"learning_rate": 4.20353982300885e-05,
|
| 411 |
+
"loss": 0.8257,
|
| 412 |
+
"step": 550
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 2.0469026548672566,
|
| 416 |
+
"grad_norm": 1.212964415550232,
|
| 417 |
+
"learning_rate": 4.178957718780728e-05,
|
| 418 |
+
"loss": 0.8587,
|
| 419 |
+
"step": 560
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 2.0513274336283187,
|
| 423 |
+
"grad_norm": 1.156423807144165,
|
| 424 |
+
"learning_rate": 4.154375614552606e-05,
|
| 425 |
+
"loss": 0.812,
|
| 426 |
+
"step": 570
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 2.0557522123893803,
|
| 430 |
+
"grad_norm": 1.274183988571167,
|
| 431 |
+
"learning_rate": 4.129793510324484e-05,
|
| 432 |
+
"loss": 0.838,
|
| 433 |
+
"step": 580
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 2.0601769911504424,
|
| 437 |
+
"grad_norm": 1.8873164653778076,
|
| 438 |
+
"learning_rate": 4.105211406096362e-05,
|
| 439 |
+
"loss": 0.8067,
|
| 440 |
+
"step": 590
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 2.0646017699115045,
|
| 444 |
+
"grad_norm": 1.1822903156280518,
|
| 445 |
+
"learning_rate": 4.08062930186824e-05,
|
| 446 |
+
"loss": 0.7949,
|
| 447 |
+
"step": 600
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 2.0690265486725665,
|
| 451 |
+
"grad_norm": 1.2289406061172485,
|
| 452 |
+
"learning_rate": 4.0560471976401183e-05,
|
| 453 |
+
"loss": 0.8064,
|
| 454 |
+
"step": 610
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 2.073451327433628,
|
| 458 |
+
"grad_norm": 1.1396745443344116,
|
| 459 |
+
"learning_rate": 4.031465093411996e-05,
|
| 460 |
+
"loss": 0.8183,
|
| 461 |
+
"step": 620
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 2.07787610619469,
|
| 465 |
+
"grad_norm": 1.041188359260559,
|
| 466 |
+
"learning_rate": 4.0068829891838745e-05,
|
| 467 |
+
"loss": 0.8024,
|
| 468 |
+
"step": 630
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 2.0823008849557523,
|
| 472 |
+
"grad_norm": 1.2631102800369263,
|
| 473 |
+
"learning_rate": 3.982300884955752e-05,
|
| 474 |
+
"loss": 0.8168,
|
| 475 |
+
"step": 640
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 2.0867256637168143,
|
| 479 |
+
"grad_norm": 1.06856107711792,
|
| 480 |
+
"learning_rate": 3.9577187807276306e-05,
|
| 481 |
+
"loss": 0.7869,
|
| 482 |
+
"step": 650
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 2.091150442477876,
|
| 486 |
+
"grad_norm": 1.4375030994415283,
|
| 487 |
+
"learning_rate": 3.9331366764995083e-05,
|
| 488 |
+
"loss": 0.8004,
|
| 489 |
+
"step": 660
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 2.095575221238938,
|
| 493 |
+
"grad_norm": 1.1425267457962036,
|
| 494 |
+
"learning_rate": 3.908554572271387e-05,
|
| 495 |
+
"loss": 0.7626,
|
| 496 |
+
"step": 670
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 2.1,
|
| 500 |
+
"grad_norm": 1.9019882678985596,
|
| 501 |
+
"learning_rate": 3.883972468043265e-05,
|
| 502 |
+
"loss": 0.826,
|
| 503 |
+
"step": 680
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 2.1004424778761064,
|
| 507 |
+
"eval_accuracy": 0.7093717093717093,
|
| 508 |
+
"eval_loss": 0.8009993433952332,
|
| 509 |
+
"eval_runtime": 986.4291,
|
| 510 |
+
"eval_samples_per_second": 2.888,
|
| 511 |
+
"eval_steps_per_second": 0.091,
|
| 512 |
+
"step": 681
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 3.0039823008849558,
|
| 516 |
+
"grad_norm": 1.6283575296401978,
|
| 517 |
+
"learning_rate": 3.859390363815143e-05,
|
| 518 |
+
"loss": 0.7981,
|
| 519 |
+
"step": 690
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 3.008407079646018,
|
| 523 |
+
"grad_norm": 1.1492269039154053,
|
| 524 |
+
"learning_rate": 3.834808259587021e-05,
|
| 525 |
+
"loss": 0.8298,
|
| 526 |
+
"step": 700
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 3.0128318584070795,
|
| 530 |
+
"grad_norm": 1.1298381090164185,
|
| 531 |
+
"learning_rate": 3.810226155358899e-05,
|
| 532 |
+
"loss": 0.7982,
|
| 533 |
+
"step": 710
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 3.0172566371681415,
|
| 537 |
+
"grad_norm": 1.1305561065673828,
|
| 538 |
+
"learning_rate": 3.7856440511307774e-05,
|
| 539 |
+
"loss": 0.7666,
|
| 540 |
+
"step": 720
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 3.0216814159292036,
|
| 544 |
+
"grad_norm": 1.1142364740371704,
|
| 545 |
+
"learning_rate": 3.7610619469026545e-05,
|
| 546 |
+
"loss": 0.7775,
|
| 547 |
+
"step": 730
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 3.0261061946902656,
|
| 551 |
+
"grad_norm": 1.7376995086669922,
|
| 552 |
+
"learning_rate": 3.736479842674533e-05,
|
| 553 |
+
"loss": 0.7879,
|
| 554 |
+
"step": 740
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 3.0305309734513273,
|
| 558 |
+
"grad_norm": 1.1918601989746094,
|
| 559 |
+
"learning_rate": 3.711897738446411e-05,
|
| 560 |
+
"loss": 0.7587,
|
| 561 |
+
"step": 750
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 3.0349557522123893,
|
| 565 |
+
"grad_norm": 1.3769606351852417,
|
| 566 |
+
"learning_rate": 3.687315634218289e-05,
|
| 567 |
+
"loss": 0.7916,
|
| 568 |
+
"step": 760
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 3.0393805309734514,
|
| 572 |
+
"grad_norm": 1.269805669784546,
|
| 573 |
+
"learning_rate": 3.6627335299901674e-05,
|
| 574 |
+
"loss": 0.8069,
|
| 575 |
+
"step": 770
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 3.0438053097345135,
|
| 579 |
+
"grad_norm": 1.045404314994812,
|
| 580 |
+
"learning_rate": 3.638151425762045e-05,
|
| 581 |
+
"loss": 0.7894,
|
| 582 |
+
"step": 780
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 3.048230088495575,
|
| 586 |
+
"grad_norm": 1.1676949262619019,
|
| 587 |
+
"learning_rate": 3.6135693215339236e-05,
|
| 588 |
+
"loss": 0.8034,
|
| 589 |
+
"step": 790
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 3.052654867256637,
|
| 593 |
+
"grad_norm": 1.8563295602798462,
|
| 594 |
+
"learning_rate": 3.588987217305801e-05,
|
| 595 |
+
"loss": 0.7366,
|
| 596 |
+
"step": 800
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 3.057079646017699,
|
| 600 |
+
"grad_norm": 1.0208163261413574,
|
| 601 |
+
"learning_rate": 3.56440511307768e-05,
|
| 602 |
+
"loss": 0.7332,
|
| 603 |
+
"step": 810
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 3.0615044247787613,
|
| 607 |
+
"grad_norm": 1.3837121725082397,
|
| 608 |
+
"learning_rate": 3.5398230088495574e-05,
|
| 609 |
+
"loss": 0.7421,
|
| 610 |
+
"step": 820
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 3.065929203539823,
|
| 614 |
+
"grad_norm": 1.1950623989105225,
|
| 615 |
+
"learning_rate": 3.515240904621436e-05,
|
| 616 |
+
"loss": 0.7696,
|
| 617 |
+
"step": 830
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 3.070353982300885,
|
| 621 |
+
"grad_norm": 1.3049529790878296,
|
| 622 |
+
"learning_rate": 3.4906588003933136e-05,
|
| 623 |
+
"loss": 0.7897,
|
| 624 |
+
"step": 840
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 3.074778761061947,
|
| 628 |
+
"grad_norm": 1.3270063400268555,
|
| 629 |
+
"learning_rate": 3.466076696165192e-05,
|
| 630 |
+
"loss": 0.7884,
|
| 631 |
+
"step": 850
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 3.0792035398230087,
|
| 635 |
+
"grad_norm": 1.4187606573104858,
|
| 636 |
+
"learning_rate": 3.44149459193707e-05,
|
| 637 |
+
"loss": 0.7526,
|
| 638 |
+
"step": 860
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 3.0836283185840707,
|
| 642 |
+
"grad_norm": 1.0789133310317993,
|
| 643 |
+
"learning_rate": 3.416912487708948e-05,
|
| 644 |
+
"loss": 0.7535,
|
| 645 |
+
"step": 870
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 3.088053097345133,
|
| 649 |
+
"grad_norm": 1.1631813049316406,
|
| 650 |
+
"learning_rate": 3.3923303834808265e-05,
|
| 651 |
+
"loss": 0.7579,
|
| 652 |
+
"step": 880
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 3.092477876106195,
|
| 656 |
+
"grad_norm": 1.1757570505142212,
|
| 657 |
+
"learning_rate": 3.367748279252704e-05,
|
| 658 |
+
"loss": 0.79,
|
| 659 |
+
"step": 890
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 3.0969026548672565,
|
| 663 |
+
"grad_norm": 1.0601412057876587,
|
| 664 |
+
"learning_rate": 3.343166175024583e-05,
|
| 665 |
+
"loss": 0.7422,
|
| 666 |
+
"step": 900
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 3.1004424778761064,
|
| 670 |
+
"eval_accuracy": 0.7371007371007371,
|
| 671 |
+
"eval_loss": 0.7514493465423584,
|
| 672 |
+
"eval_runtime": 1046.1945,
|
| 673 |
+
"eval_samples_per_second": 2.723,
|
| 674 |
+
"eval_steps_per_second": 0.086,
|
| 675 |
+
"step": 908
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 4.000884955752213,
|
| 679 |
+
"grad_norm": 1.757256269454956,
|
| 680 |
+
"learning_rate": 3.3185840707964604e-05,
|
| 681 |
+
"loss": 0.7465,
|
| 682 |
+
"step": 910
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 4.005309734513275,
|
| 686 |
+
"grad_norm": 1.1318788528442383,
|
| 687 |
+
"learning_rate": 3.294001966568339e-05,
|
| 688 |
+
"loss": 0.7494,
|
| 689 |
+
"step": 920
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 4.009734513274337,
|
| 693 |
+
"grad_norm": 1.104137659072876,
|
| 694 |
+
"learning_rate": 3.2694198623402165e-05,
|
| 695 |
+
"loss": 0.7696,
|
| 696 |
+
"step": 930
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 4.014159292035398,
|
| 700 |
+
"grad_norm": 1.0731704235076904,
|
| 701 |
+
"learning_rate": 3.244837758112095e-05,
|
| 702 |
+
"loss": 0.748,
|
| 703 |
+
"step": 940
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 4.01858407079646,
|
| 707 |
+
"grad_norm": 1.1554673910140991,
|
| 708 |
+
"learning_rate": 3.220255653883973e-05,
|
| 709 |
+
"loss": 0.79,
|
| 710 |
+
"step": 950
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 4.023008849557522,
|
| 714 |
+
"grad_norm": 1.2223293781280518,
|
| 715 |
+
"learning_rate": 3.1956735496558504e-05,
|
| 716 |
+
"loss": 0.7281,
|
| 717 |
+
"step": 960
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 4.027433628318584,
|
| 721 |
+
"grad_norm": 1.1225980520248413,
|
| 722 |
+
"learning_rate": 3.171091445427729e-05,
|
| 723 |
+
"loss": 0.7219,
|
| 724 |
+
"step": 970
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 4.031858407079646,
|
| 728 |
+
"grad_norm": 1.0847655534744263,
|
| 729 |
+
"learning_rate": 3.1465093411996065e-05,
|
| 730 |
+
"loss": 0.7481,
|
| 731 |
+
"step": 980
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 4.036283185840708,
|
| 735 |
+
"grad_norm": 1.3014447689056396,
|
| 736 |
+
"learning_rate": 3.121927236971485e-05,
|
| 737 |
+
"loss": 0.7305,
|
| 738 |
+
"step": 990
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 4.04070796460177,
|
| 742 |
+
"grad_norm": 1.1104828119277954,
|
| 743 |
+
"learning_rate": 3.097345132743363e-05,
|
| 744 |
+
"loss": 0.7211,
|
| 745 |
+
"step": 1000
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 4.0451327433628315,
|
| 749 |
+
"grad_norm": 1.0301592350006104,
|
| 750 |
+
"learning_rate": 3.072763028515241e-05,
|
| 751 |
+
"loss": 0.7676,
|
| 752 |
+
"step": 1010
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 4.0495575221238935,
|
| 756 |
+
"grad_norm": 1.1068618297576904,
|
| 757 |
+
"learning_rate": 3.048180924287119e-05,
|
| 758 |
+
"loss": 0.7078,
|
| 759 |
+
"step": 1020
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 4.053982300884956,
|
| 763 |
+
"grad_norm": 1.4518685340881348,
|
| 764 |
+
"learning_rate": 3.0235988200589972e-05,
|
| 765 |
+
"loss": 0.7706,
|
| 766 |
+
"step": 1030
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 4.058407079646018,
|
| 770 |
+
"grad_norm": 0.9889740347862244,
|
| 771 |
+
"learning_rate": 2.9990167158308753e-05,
|
| 772 |
+
"loss": 0.7207,
|
| 773 |
+
"step": 1040
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 4.06283185840708,
|
| 777 |
+
"grad_norm": 1.1339548826217651,
|
| 778 |
+
"learning_rate": 2.9744346116027534e-05,
|
| 779 |
+
"loss": 0.7423,
|
| 780 |
+
"step": 1050
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 4.067256637168142,
|
| 784 |
+
"grad_norm": 1.3356149196624756,
|
| 785 |
+
"learning_rate": 2.9498525073746314e-05,
|
| 786 |
+
"loss": 0.7114,
|
| 787 |
+
"step": 1060
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 4.071681415929204,
|
| 791 |
+
"grad_norm": 1.0467292070388794,
|
| 792 |
+
"learning_rate": 2.9252704031465095e-05,
|
| 793 |
+
"loss": 0.6998,
|
| 794 |
+
"step": 1070
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 4.076106194690266,
|
| 798 |
+
"grad_norm": 1.0783181190490723,
|
| 799 |
+
"learning_rate": 2.9006882989183876e-05,
|
| 800 |
+
"loss": 0.7544,
|
| 801 |
+
"step": 1080
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 4.080530973451327,
|
| 805 |
+
"grad_norm": 1.0915840864181519,
|
| 806 |
+
"learning_rate": 2.8761061946902656e-05,
|
| 807 |
+
"loss": 0.7145,
|
| 808 |
+
"step": 1090
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 4.084955752212389,
|
| 812 |
+
"grad_norm": 1.646979570388794,
|
| 813 |
+
"learning_rate": 2.8515240904621437e-05,
|
| 814 |
+
"loss": 0.7145,
|
| 815 |
+
"step": 1100
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 4.089380530973451,
|
| 819 |
+
"grad_norm": 1.3354220390319824,
|
| 820 |
+
"learning_rate": 2.8269419862340218e-05,
|
| 821 |
+
"loss": 0.7329,
|
| 822 |
+
"step": 1110
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 4.093805309734513,
|
| 826 |
+
"grad_norm": 1.1131007671356201,
|
| 827 |
+
"learning_rate": 2.8023598820059e-05,
|
| 828 |
+
"loss": 0.6812,
|
| 829 |
+
"step": 1120
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 4.098230088495575,
|
| 833 |
+
"grad_norm": 1.260852575302124,
|
| 834 |
+
"learning_rate": 2.777777777777778e-05,
|
| 835 |
+
"loss": 0.7206,
|
| 836 |
+
"step": 1130
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 4.100442477876106,
|
| 840 |
+
"eval_accuracy": 0.7507897507897507,
|
| 841 |
+
"eval_loss": 0.7169636487960815,
|
| 842 |
+
"eval_runtime": 1087.6772,
|
| 843 |
+
"eval_samples_per_second": 2.619,
|
| 844 |
+
"eval_steps_per_second": 0.083,
|
| 845 |
+
"step": 1135
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 5.002212389380531,
|
| 849 |
+
"grad_norm": 1.1113187074661255,
|
| 850 |
+
"learning_rate": 2.753195673549656e-05,
|
| 851 |
+
"loss": 0.6946,
|
| 852 |
+
"step": 1140
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 5.006637168141593,
|
| 856 |
+
"grad_norm": 1.0982720851898193,
|
| 857 |
+
"learning_rate": 2.7286135693215344e-05,
|
| 858 |
+
"loss": 0.7624,
|
| 859 |
+
"step": 1150
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 5.011061946902655,
|
| 863 |
+
"grad_norm": 1.3066320419311523,
|
| 864 |
+
"learning_rate": 2.7040314650934125e-05,
|
| 865 |
+
"loss": 0.7023,
|
| 866 |
+
"step": 1160
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 5.015486725663717,
|
| 870 |
+
"grad_norm": 0.9856134057044983,
|
| 871 |
+
"learning_rate": 2.6794493608652905e-05,
|
| 872 |
+
"loss": 0.6863,
|
| 873 |
+
"step": 1170
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 5.019911504424779,
|
| 877 |
+
"grad_norm": 1.2251431941986084,
|
| 878 |
+
"learning_rate": 2.6548672566371686e-05,
|
| 879 |
+
"loss": 0.7327,
|
| 880 |
+
"step": 1180
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 5.024336283185841,
|
| 884 |
+
"grad_norm": 0.9257007837295532,
|
| 885 |
+
"learning_rate": 2.6302851524090467e-05,
|
| 886 |
+
"loss": 0.7002,
|
| 887 |
+
"step": 1190
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 5.028761061946903,
|
| 891 |
+
"grad_norm": 0.9643080830574036,
|
| 892 |
+
"learning_rate": 2.605703048180924e-05,
|
| 893 |
+
"loss": 0.6906,
|
| 894 |
+
"step": 1200
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 5.033185840707965,
|
| 898 |
+
"grad_norm": 1.1749197244644165,
|
| 899 |
+
"learning_rate": 2.581120943952802e-05,
|
| 900 |
+
"loss": 0.7327,
|
| 901 |
+
"step": 1210
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 5.037610619469026,
|
| 905 |
+
"grad_norm": 1.1557971239089966,
|
| 906 |
+
"learning_rate": 2.5565388397246802e-05,
|
| 907 |
+
"loss": 0.6845,
|
| 908 |
+
"step": 1220
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 5.042035398230088,
|
| 912 |
+
"grad_norm": 1.2948167324066162,
|
| 913 |
+
"learning_rate": 2.5319567354965586e-05,
|
| 914 |
+
"loss": 0.7152,
|
| 915 |
+
"step": 1230
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 5.04646017699115,
|
| 919 |
+
"grad_norm": 0.9071692228317261,
|
| 920 |
+
"learning_rate": 2.5073746312684367e-05,
|
| 921 |
+
"loss": 0.701,
|
| 922 |
+
"step": 1240
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 5.050884955752212,
|
| 926 |
+
"grad_norm": 1.0012332201004028,
|
| 927 |
+
"learning_rate": 2.4827925270403147e-05,
|
| 928 |
+
"loss": 0.7141,
|
| 929 |
+
"step": 1250
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 5.0553097345132745,
|
| 933 |
+
"grad_norm": 1.4386842250823975,
|
| 934 |
+
"learning_rate": 2.4582104228121928e-05,
|
| 935 |
+
"loss": 0.7405,
|
| 936 |
+
"step": 1260
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 5.0597345132743365,
|
| 940 |
+
"grad_norm": 1.1797550916671753,
|
| 941 |
+
"learning_rate": 2.433628318584071e-05,
|
| 942 |
+
"loss": 0.6717,
|
| 943 |
+
"step": 1270
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 5.064159292035399,
|
| 947 |
+
"grad_norm": 1.0993843078613281,
|
| 948 |
+
"learning_rate": 2.409046214355949e-05,
|
| 949 |
+
"loss": 0.6832,
|
| 950 |
+
"step": 1280
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 5.06858407079646,
|
| 954 |
+
"grad_norm": 0.9947831630706787,
|
| 955 |
+
"learning_rate": 2.384464110127827e-05,
|
| 956 |
+
"loss": 0.7164,
|
| 957 |
+
"step": 1290
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 5.073008849557522,
|
| 961 |
+
"grad_norm": 1.335589051246643,
|
| 962 |
+
"learning_rate": 2.359882005899705e-05,
|
| 963 |
+
"loss": 0.7541,
|
| 964 |
+
"step": 1300
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 5.077433628318584,
|
| 968 |
+
"grad_norm": 1.3028713464736938,
|
| 969 |
+
"learning_rate": 2.335299901671583e-05,
|
| 970 |
+
"loss": 0.7167,
|
| 971 |
+
"step": 1310
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 5.081858407079646,
|
| 975 |
+
"grad_norm": 1.1588274240493774,
|
| 976 |
+
"learning_rate": 2.3107177974434612e-05,
|
| 977 |
+
"loss": 0.7194,
|
| 978 |
+
"step": 1320
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 5.086283185840708,
|
| 982 |
+
"grad_norm": 0.9976550340652466,
|
| 983 |
+
"learning_rate": 2.2861356932153393e-05,
|
| 984 |
+
"loss": 0.7196,
|
| 985 |
+
"step": 1330
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 5.09070796460177,
|
| 989 |
+
"grad_norm": 0.8966324925422668,
|
| 990 |
+
"learning_rate": 2.2615535889872174e-05,
|
| 991 |
+
"loss": 0.7268,
|
| 992 |
+
"step": 1340
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 5.095132743362832,
|
| 996 |
+
"grad_norm": 1.0999760627746582,
|
| 997 |
+
"learning_rate": 2.2369714847590954e-05,
|
| 998 |
+
"loss": 0.7284,
|
| 999 |
+
"step": 1350
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 5.099557522123894,
|
| 1003 |
+
"grad_norm": 1.032329797744751,
|
| 1004 |
+
"learning_rate": 2.2123893805309738e-05,
|
| 1005 |
+
"loss": 0.6806,
|
| 1006 |
+
"step": 1360
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 5.100442477876106,
|
| 1010 |
+
"eval_accuracy": 0.7542997542997543,
|
| 1011 |
+
"eval_loss": 0.6924355626106262,
|
| 1012 |
+
"eval_runtime": 1080.3798,
|
| 1013 |
+
"eval_samples_per_second": 2.637,
|
| 1014 |
+
"eval_steps_per_second": 0.083,
|
| 1015 |
+
"step": 1362
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 6.0035398230088495,
|
| 1019 |
+
"grad_norm": 1.2827861309051514,
|
| 1020 |
+
"learning_rate": 2.187807276302852e-05,
|
| 1021 |
+
"loss": 0.768,
|
| 1022 |
+
"step": 1370
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 6.0079646017699115,
|
| 1026 |
+
"grad_norm": 1.1958467960357666,
|
| 1027 |
+
"learning_rate": 2.1632251720747296e-05,
|
| 1028 |
+
"loss": 0.7209,
|
| 1029 |
+
"step": 1380
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 6.012389380530974,
|
| 1033 |
+
"grad_norm": 1.0269005298614502,
|
| 1034 |
+
"learning_rate": 2.1386430678466077e-05,
|
| 1035 |
+
"loss": 0.7095,
|
| 1036 |
+
"step": 1390
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 6.016814159292036,
|
| 1040 |
+
"grad_norm": 0.9638125896453857,
|
| 1041 |
+
"learning_rate": 2.1140609636184858e-05,
|
| 1042 |
+
"loss": 0.6557,
|
| 1043 |
+
"step": 1400
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 6.021238938053098,
|
| 1047 |
+
"grad_norm": 1.3249033689498901,
|
| 1048 |
+
"learning_rate": 2.089478859390364e-05,
|
| 1049 |
+
"loss": 0.7129,
|
| 1050 |
+
"step": 1410
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 6.025663716814159,
|
| 1054 |
+
"grad_norm": 1.1330633163452148,
|
| 1055 |
+
"learning_rate": 2.064896755162242e-05,
|
| 1056 |
+
"loss": 0.6472,
|
| 1057 |
+
"step": 1420
|
| 1058 |
+
},
|
| 1059 |
+
{
|
| 1060 |
+
"epoch": 6.030088495575221,
|
| 1061 |
+
"grad_norm": 1.1240506172180176,
|
| 1062 |
+
"learning_rate": 2.04031465093412e-05,
|
| 1063 |
+
"loss": 0.6748,
|
| 1064 |
+
"step": 1430
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 6.034513274336283,
|
| 1068 |
+
"grad_norm": 1.091794729232788,
|
| 1069 |
+
"learning_rate": 2.015732546705998e-05,
|
| 1070 |
+
"loss": 0.7125,
|
| 1071 |
+
"step": 1440
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 6.038938053097345,
|
| 1075 |
+
"grad_norm": 0.892193078994751,
|
| 1076 |
+
"learning_rate": 1.991150442477876e-05,
|
| 1077 |
+
"loss": 0.7045,
|
| 1078 |
+
"step": 1450
|
| 1079 |
+
},
|
| 1080 |
+
{
|
| 1081 |
+
"epoch": 6.043362831858407,
|
| 1082 |
+
"grad_norm": 1.1631165742874146,
|
| 1083 |
+
"learning_rate": 1.9665683382497542e-05,
|
| 1084 |
+
"loss": 0.7432,
|
| 1085 |
+
"step": 1460
|
| 1086 |
+
},
|
| 1087 |
+
{
|
| 1088 |
+
"epoch": 6.047787610619469,
|
| 1089 |
+
"grad_norm": 0.9747660160064697,
|
| 1090 |
+
"learning_rate": 1.9419862340216326e-05,
|
| 1091 |
+
"loss": 0.6965,
|
| 1092 |
+
"step": 1470
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 6.052212389380531,
|
| 1096 |
+
"grad_norm": 0.9529617428779602,
|
| 1097 |
+
"learning_rate": 1.9174041297935107e-05,
|
| 1098 |
+
"loss": 0.6725,
|
| 1099 |
+
"step": 1480
|
| 1100 |
+
},
|
| 1101 |
+
{
|
| 1102 |
+
"epoch": 6.056637168141593,
|
| 1103 |
+
"grad_norm": 1.0859458446502686,
|
| 1104 |
+
"learning_rate": 1.8928220255653887e-05,
|
| 1105 |
+
"loss": 0.6744,
|
| 1106 |
+
"step": 1490
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 6.0610619469026545,
|
| 1110 |
+
"grad_norm": 1.4146475791931152,
|
| 1111 |
+
"learning_rate": 1.8682399213372664e-05,
|
| 1112 |
+
"loss": 0.6668,
|
| 1113 |
+
"step": 1500
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 6.065486725663717,
|
| 1117 |
+
"grad_norm": 1.5994105339050293,
|
| 1118 |
+
"learning_rate": 1.8436578171091445e-05,
|
| 1119 |
+
"loss": 0.6941,
|
| 1120 |
+
"step": 1510
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 6.069911504424779,
|
| 1124 |
+
"grad_norm": 1.22159743309021,
|
| 1125 |
+
"learning_rate": 1.8190757128810226e-05,
|
| 1126 |
+
"loss": 0.7201,
|
| 1127 |
+
"step": 1520
|
| 1128 |
+
},
|
| 1129 |
+
{
|
| 1130 |
+
"epoch": 6.074336283185841,
|
| 1131 |
+
"grad_norm": 1.196835994720459,
|
| 1132 |
+
"learning_rate": 1.7944936086529007e-05,
|
| 1133 |
+
"loss": 0.6506,
|
| 1134 |
+
"step": 1530
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 6.078761061946903,
|
| 1138 |
+
"grad_norm": 0.9212129712104797,
|
| 1139 |
+
"learning_rate": 1.7699115044247787e-05,
|
| 1140 |
+
"loss": 0.6931,
|
| 1141 |
+
"step": 1540
|
| 1142 |
+
},
|
| 1143 |
+
{
|
| 1144 |
+
"epoch": 6.083185840707965,
|
| 1145 |
+
"grad_norm": 1.0447068214416504,
|
| 1146 |
+
"learning_rate": 1.7453294001966568e-05,
|
| 1147 |
+
"loss": 0.7107,
|
| 1148 |
+
"step": 1550
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 6.087610619469027,
|
| 1152 |
+
"grad_norm": 1.2515738010406494,
|
| 1153 |
+
"learning_rate": 1.720747295968535e-05,
|
| 1154 |
+
"loss": 0.7178,
|
| 1155 |
+
"step": 1560
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 6.092035398230088,
|
| 1159 |
+
"grad_norm": 1.4285340309143066,
|
| 1160 |
+
"learning_rate": 1.6961651917404133e-05,
|
| 1161 |
+
"loss": 0.6929,
|
| 1162 |
+
"step": 1570
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"epoch": 6.09646017699115,
|
| 1166 |
+
"grad_norm": 1.3406482934951782,
|
| 1167 |
+
"learning_rate": 1.6715830875122913e-05,
|
| 1168 |
+
"loss": 0.6826,
|
| 1169 |
+
"step": 1580
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"epoch": 6.100442477876106,
|
| 1173 |
+
"eval_accuracy": 0.7585117585117586,
|
| 1174 |
+
"eval_loss": 0.675682008266449,
|
| 1175 |
+
"eval_runtime": 1089.8885,
|
| 1176 |
+
"eval_samples_per_second": 2.614,
|
| 1177 |
+
"eval_steps_per_second": 0.083,
|
| 1178 |
+
"step": 1589
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 7.000442477876106,
|
| 1182 |
+
"grad_norm": 1.920408844947815,
|
| 1183 |
+
"learning_rate": 1.6470009832841694e-05,
|
| 1184 |
+
"loss": 0.6845,
|
| 1185 |
+
"step": 1590
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 7.004867256637168,
|
| 1189 |
+
"grad_norm": 1.2930206060409546,
|
| 1190 |
+
"learning_rate": 1.6224188790560475e-05,
|
| 1191 |
+
"loss": 0.6906,
|
| 1192 |
+
"step": 1600
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 7.00929203539823,
|
| 1196 |
+
"grad_norm": 1.010541319847107,
|
| 1197 |
+
"learning_rate": 1.5978367748279252e-05,
|
| 1198 |
+
"loss": 0.7002,
|
| 1199 |
+
"step": 1610
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 7.013716814159292,
|
| 1203 |
+
"grad_norm": 1.3391730785369873,
|
| 1204 |
+
"learning_rate": 1.5732546705998033e-05,
|
| 1205 |
+
"loss": 0.6841,
|
| 1206 |
+
"step": 1620
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 7.018141592920354,
|
| 1210 |
+
"grad_norm": 1.4052796363830566,
|
| 1211 |
+
"learning_rate": 1.5486725663716813e-05,
|
| 1212 |
+
"loss": 0.6611,
|
| 1213 |
+
"step": 1630
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 7.022566371681416,
|
| 1217 |
+
"grad_norm": 1.2141647338867188,
|
| 1218 |
+
"learning_rate": 1.5240904621435596e-05,
|
| 1219 |
+
"loss": 0.6856,
|
| 1220 |
+
"step": 1640
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 7.026991150442478,
|
| 1224 |
+
"grad_norm": 1.3713358640670776,
|
| 1225 |
+
"learning_rate": 1.4995083579154376e-05,
|
| 1226 |
+
"loss": 0.6978,
|
| 1227 |
+
"step": 1650
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 7.03141592920354,
|
| 1231 |
+
"grad_norm": 0.9116381406784058,
|
| 1232 |
+
"learning_rate": 1.4749262536873157e-05,
|
| 1233 |
+
"loss": 0.696,
|
| 1234 |
+
"step": 1660
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 7.035840707964602,
|
| 1238 |
+
"grad_norm": 1.1704223155975342,
|
| 1239 |
+
"learning_rate": 1.4503441494591938e-05,
|
| 1240 |
+
"loss": 0.6695,
|
| 1241 |
+
"step": 1670
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 7.040265486725664,
|
| 1245 |
+
"grad_norm": 1.1459695100784302,
|
| 1246 |
+
"learning_rate": 1.4257620452310719e-05,
|
| 1247 |
+
"loss": 0.6575,
|
| 1248 |
+
"step": 1680
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 7.044690265486726,
|
| 1252 |
+
"grad_norm": 1.098761796951294,
|
| 1253 |
+
"learning_rate": 1.40117994100295e-05,
|
| 1254 |
+
"loss": 0.6905,
|
| 1255 |
+
"step": 1690
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 7.049115044247787,
|
| 1259 |
+
"grad_norm": 1.2097493410110474,
|
| 1260 |
+
"learning_rate": 1.376597836774828e-05,
|
| 1261 |
+
"loss": 0.7086,
|
| 1262 |
+
"step": 1700
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 7.053539823008849,
|
| 1266 |
+
"grad_norm": 1.3922789096832275,
|
| 1267 |
+
"learning_rate": 1.3520157325467062e-05,
|
| 1268 |
+
"loss": 0.6473,
|
| 1269 |
+
"step": 1710
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 7.057964601769911,
|
| 1273 |
+
"grad_norm": 1.064712405204773,
|
| 1274 |
+
"learning_rate": 1.3274336283185843e-05,
|
| 1275 |
+
"loss": 0.6987,
|
| 1276 |
+
"step": 1720
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 7.062389380530973,
|
| 1280 |
+
"grad_norm": 1.1013967990875244,
|
| 1281 |
+
"learning_rate": 1.302851524090462e-05,
|
| 1282 |
+
"loss": 0.6593,
|
| 1283 |
+
"step": 1730
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 7.0668141592920355,
|
| 1287 |
+
"grad_norm": 0.9073940515518188,
|
| 1288 |
+
"learning_rate": 1.2782694198623401e-05,
|
| 1289 |
+
"loss": 0.6873,
|
| 1290 |
+
"step": 1740
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 7.0712389380530976,
|
| 1294 |
+
"grad_norm": 1.1259492635726929,
|
| 1295 |
+
"learning_rate": 1.2536873156342183e-05,
|
| 1296 |
+
"loss": 0.6954,
|
| 1297 |
+
"step": 1750
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 7.07566371681416,
|
| 1301 |
+
"grad_norm": 0.8687026500701904,
|
| 1302 |
+
"learning_rate": 1.2291052114060964e-05,
|
| 1303 |
+
"loss": 0.6854,
|
| 1304 |
+
"step": 1760
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 7.080088495575222,
|
| 1308 |
+
"grad_norm": 1.687637448310852,
|
| 1309 |
+
"learning_rate": 1.2045231071779745e-05,
|
| 1310 |
+
"loss": 0.6781,
|
| 1311 |
+
"step": 1770
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 7.084513274336283,
|
| 1315 |
+
"grad_norm": 1.0679877996444702,
|
| 1316 |
+
"learning_rate": 1.1799410029498525e-05,
|
| 1317 |
+
"loss": 0.7,
|
| 1318 |
+
"step": 1780
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 7.088938053097345,
|
| 1322 |
+
"grad_norm": 1.1482867002487183,
|
| 1323 |
+
"learning_rate": 1.1553588987217306e-05,
|
| 1324 |
+
"loss": 0.6494,
|
| 1325 |
+
"step": 1790
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 7.093362831858407,
|
| 1329 |
+
"grad_norm": 1.042781949043274,
|
| 1330 |
+
"learning_rate": 1.1307767944936087e-05,
|
| 1331 |
+
"loss": 0.6776,
|
| 1332 |
+
"step": 1800
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 7.097787610619469,
|
| 1336 |
+
"grad_norm": 0.8880053162574768,
|
| 1337 |
+
"learning_rate": 1.1061946902654869e-05,
|
| 1338 |
+
"loss": 0.6756,
|
| 1339 |
+
"step": 1810
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 7.100442477876106,
|
| 1343 |
+
"eval_accuracy": 0.7630747630747631,
|
| 1344 |
+
"eval_loss": 0.6651633381843567,
|
| 1345 |
+
"eval_runtime": 1032.3877,
|
| 1346 |
+
"eval_samples_per_second": 2.76,
|
| 1347 |
+
"eval_steps_per_second": 0.087,
|
| 1348 |
+
"step": 1816
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 8.001769911504425,
|
| 1352 |
+
"grad_norm": 0.9909548163414001,
|
| 1353 |
+
"learning_rate": 1.0816125860373648e-05,
|
| 1354 |
+
"loss": 0.6837,
|
| 1355 |
+
"step": 1820
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 8.006194690265486,
|
| 1359 |
+
"grad_norm": 0.943515419960022,
|
| 1360 |
+
"learning_rate": 1.0570304818092429e-05,
|
| 1361 |
+
"loss": 0.665,
|
| 1362 |
+
"step": 1830
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 8.01061946902655,
|
| 1366 |
+
"grad_norm": 0.9931963682174683,
|
| 1367 |
+
"learning_rate": 1.032448377581121e-05,
|
| 1368 |
+
"loss": 0.6664,
|
| 1369 |
+
"step": 1840
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 8.01504424778761,
|
| 1373 |
+
"grad_norm": 0.8919005393981934,
|
| 1374 |
+
"learning_rate": 1.007866273352999e-05,
|
| 1375 |
+
"loss": 0.7089,
|
| 1376 |
+
"step": 1850
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 8.019469026548673,
|
| 1380 |
+
"grad_norm": 1.1344377994537354,
|
| 1381 |
+
"learning_rate": 9.832841691248771e-06,
|
| 1382 |
+
"loss": 0.6589,
|
| 1383 |
+
"step": 1860
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 8.023893805309735,
|
| 1387 |
+
"grad_norm": 1.4625028371810913,
|
| 1388 |
+
"learning_rate": 9.587020648967553e-06,
|
| 1389 |
+
"loss": 0.6771,
|
| 1390 |
+
"step": 1870
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 8.028318584070796,
|
| 1394 |
+
"grad_norm": 0.8750028610229492,
|
| 1395 |
+
"learning_rate": 9.341199606686332e-06,
|
| 1396 |
+
"loss": 0.6856,
|
| 1397 |
+
"step": 1880
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 8.032743362831859,
|
| 1401 |
+
"grad_norm": 1.1741547584533691,
|
| 1402 |
+
"learning_rate": 9.095378564405113e-06,
|
| 1403 |
+
"loss": 0.693,
|
| 1404 |
+
"step": 1890
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 8.03716814159292,
|
| 1408 |
+
"grad_norm": 1.0879665613174438,
|
| 1409 |
+
"learning_rate": 8.849557522123894e-06,
|
| 1410 |
+
"loss": 0.7053,
|
| 1411 |
+
"step": 1900
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 8.041592920353983,
|
| 1415 |
+
"grad_norm": 0.9571520686149597,
|
| 1416 |
+
"learning_rate": 8.603736479842674e-06,
|
| 1417 |
+
"loss": 0.6415,
|
| 1418 |
+
"step": 1910
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 8.046017699115044,
|
| 1422 |
+
"grad_norm": 1.060584545135498,
|
| 1423 |
+
"learning_rate": 8.357915437561457e-06,
|
| 1424 |
+
"loss": 0.6809,
|
| 1425 |
+
"step": 1920
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 8.050442477876107,
|
| 1429 |
+
"grad_norm": 1.0861510038375854,
|
| 1430 |
+
"learning_rate": 8.112094395280237e-06,
|
| 1431 |
+
"loss": 0.6771,
|
| 1432 |
+
"step": 1930
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 8.054867256637168,
|
| 1436 |
+
"grad_norm": 1.0085254907608032,
|
| 1437 |
+
"learning_rate": 7.866273352999016e-06,
|
| 1438 |
+
"loss": 0.6639,
|
| 1439 |
+
"step": 1940
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 8.05929203539823,
|
| 1443 |
+
"grad_norm": 1.0106779336929321,
|
| 1444 |
+
"learning_rate": 7.620452310717798e-06,
|
| 1445 |
+
"loss": 0.6665,
|
| 1446 |
+
"step": 1950
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 8.063716814159292,
|
| 1450 |
+
"grad_norm": 1.03801691532135,
|
| 1451 |
+
"learning_rate": 7.374631268436579e-06,
|
| 1452 |
+
"loss": 0.6768,
|
| 1453 |
+
"step": 1960
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 8.068141592920353,
|
| 1457 |
+
"grad_norm": 1.242561936378479,
|
| 1458 |
+
"learning_rate": 7.128810226155359e-06,
|
| 1459 |
+
"loss": 0.6154,
|
| 1460 |
+
"step": 1970
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 8.072566371681416,
|
| 1464 |
+
"grad_norm": 0.9356180429458618,
|
| 1465 |
+
"learning_rate": 6.88298918387414e-06,
|
| 1466 |
+
"loss": 0.6564,
|
| 1467 |
+
"step": 1980
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 8.076991150442478,
|
| 1471 |
+
"grad_norm": 1.3207311630249023,
|
| 1472 |
+
"learning_rate": 6.6371681415929215e-06,
|
| 1473 |
+
"loss": 0.7182,
|
| 1474 |
+
"step": 1990
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 8.08141592920354,
|
| 1478 |
+
"grad_norm": 1.6134599447250366,
|
| 1479 |
+
"learning_rate": 6.3913470993117005e-06,
|
| 1480 |
+
"loss": 0.6538,
|
| 1481 |
+
"step": 2000
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 8.085840707964602,
|
| 1485 |
+
"grad_norm": 1.317514419555664,
|
| 1486 |
+
"learning_rate": 6.145526057030482e-06,
|
| 1487 |
+
"loss": 0.6756,
|
| 1488 |
+
"step": 2010
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 8.090265486725663,
|
| 1492 |
+
"grad_norm": 1.5709384679794312,
|
| 1493 |
+
"learning_rate": 5.899705014749263e-06,
|
| 1494 |
+
"loss": 0.646,
|
| 1495 |
+
"step": 2020
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 8.094690265486726,
|
| 1499 |
+
"grad_norm": 0.9243668913841248,
|
| 1500 |
+
"learning_rate": 5.653883972468043e-06,
|
| 1501 |
+
"loss": 0.6908,
|
| 1502 |
+
"step": 2030
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 8.099115044247787,
|
| 1506 |
+
"grad_norm": 0.859581708908081,
|
| 1507 |
+
"learning_rate": 5.408062930186824e-06,
|
| 1508 |
+
"loss": 0.6964,
|
| 1509 |
+
"step": 2040
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 8.100442477876106,
|
| 1513 |
+
"eval_accuracy": 0.7655317655317655,
|
| 1514 |
+
"eval_loss": 0.6591010093688965,
|
| 1515 |
+
"eval_runtime": 997.7811,
|
| 1516 |
+
"eval_samples_per_second": 2.855,
|
| 1517 |
+
"eval_steps_per_second": 0.09,
|
| 1518 |
+
"step": 2043
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 9.003097345132744,
|
| 1522 |
+
"grad_norm": 1.7191717624664307,
|
| 1523 |
+
"learning_rate": 5.162241887905605e-06,
|
| 1524 |
+
"loss": 0.6788,
|
| 1525 |
+
"step": 2050
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 9.007522123893805,
|
| 1529 |
+
"grad_norm": 1.2770448923110962,
|
| 1530 |
+
"learning_rate": 4.9164208456243854e-06,
|
| 1531 |
+
"loss": 0.6437,
|
| 1532 |
+
"step": 2060
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 9.011946902654866,
|
| 1536 |
+
"grad_norm": 1.0872125625610352,
|
| 1537 |
+
"learning_rate": 4.670599803343166e-06,
|
| 1538 |
+
"loss": 0.6984,
|
| 1539 |
+
"step": 2070
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 9.01637168141593,
|
| 1543 |
+
"grad_norm": 0.8692576289176941,
|
| 1544 |
+
"learning_rate": 4.424778761061947e-06,
|
| 1545 |
+
"loss": 0.6377,
|
| 1546 |
+
"step": 2080
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 9.02079646017699,
|
| 1550 |
+
"grad_norm": 0.9820857048034668,
|
| 1551 |
+
"learning_rate": 4.178957718780728e-06,
|
| 1552 |
+
"loss": 0.6671,
|
| 1553 |
+
"step": 2090
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 9.025221238938054,
|
| 1557 |
+
"grad_norm": 1.6207181215286255,
|
| 1558 |
+
"learning_rate": 3.933136676499508e-06,
|
| 1559 |
+
"loss": 0.683,
|
| 1560 |
+
"step": 2100
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 9.029646017699115,
|
| 1564 |
+
"grad_norm": 1.0268634557724,
|
| 1565 |
+
"learning_rate": 3.6873156342182893e-06,
|
| 1566 |
+
"loss": 0.6381,
|
| 1567 |
+
"step": 2110
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 9.034070796460178,
|
| 1571 |
+
"grad_norm": 1.0367904901504517,
|
| 1572 |
+
"learning_rate": 3.44149459193707e-06,
|
| 1573 |
+
"loss": 0.6969,
|
| 1574 |
+
"step": 2120
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 9.038495575221239,
|
| 1578 |
+
"grad_norm": 1.4884945154190063,
|
| 1579 |
+
"learning_rate": 3.1956735496558502e-06,
|
| 1580 |
+
"loss": 0.6342,
|
| 1581 |
+
"step": 2130
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 9.042920353982302,
|
| 1585 |
+
"grad_norm": 1.086195707321167,
|
| 1586 |
+
"learning_rate": 2.9498525073746313e-06,
|
| 1587 |
+
"loss": 0.6649,
|
| 1588 |
+
"step": 2140
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 9.047345132743363,
|
| 1592 |
+
"grad_norm": 1.0393646955490112,
|
| 1593 |
+
"learning_rate": 2.704031465093412e-06,
|
| 1594 |
+
"loss": 0.6646,
|
| 1595 |
+
"step": 2150
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 9.051769911504424,
|
| 1599 |
+
"grad_norm": 1.188118815422058,
|
| 1600 |
+
"learning_rate": 2.4582104228121927e-06,
|
| 1601 |
+
"loss": 0.6681,
|
| 1602 |
+
"step": 2160
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 9.056194690265487,
|
| 1606 |
+
"grad_norm": 1.3695433139801025,
|
| 1607 |
+
"learning_rate": 2.2123893805309734e-06,
|
| 1608 |
+
"loss": 0.667,
|
| 1609 |
+
"step": 2170
|
| 1610 |
+
},
|
| 1611 |
+
{
|
| 1612 |
+
"epoch": 9.060619469026548,
|
| 1613 |
+
"grad_norm": 1.166283130645752,
|
| 1614 |
+
"learning_rate": 1.966568338249754e-06,
|
| 1615 |
+
"loss": 0.679,
|
| 1616 |
+
"step": 2180
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 9.065044247787611,
|
| 1620 |
+
"grad_norm": 0.956119179725647,
|
| 1621 |
+
"learning_rate": 1.720747295968535e-06,
|
| 1622 |
+
"loss": 0.6528,
|
| 1623 |
+
"step": 2190
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 9.069469026548672,
|
| 1627 |
+
"grad_norm": 1.2216787338256836,
|
| 1628 |
+
"learning_rate": 1.4749262536873157e-06,
|
| 1629 |
+
"loss": 0.6857,
|
| 1630 |
+
"step": 2200
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 9.073893805309735,
|
| 1634 |
+
"grad_norm": 1.2176066637039185,
|
| 1635 |
+
"learning_rate": 1.2291052114060964e-06,
|
| 1636 |
+
"loss": 0.6849,
|
| 1637 |
+
"step": 2210
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 9.078318584070797,
|
| 1641 |
+
"grad_norm": 0.9874552488327026,
|
| 1642 |
+
"learning_rate": 9.83284169124877e-07,
|
| 1643 |
+
"loss": 0.6788,
|
| 1644 |
+
"step": 2220
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 9.082743362831858,
|
| 1648 |
+
"grad_norm": 1.053183913230896,
|
| 1649 |
+
"learning_rate": 7.374631268436578e-07,
|
| 1650 |
+
"loss": 0.6791,
|
| 1651 |
+
"step": 2230
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 9.08716814159292,
|
| 1655 |
+
"grad_norm": 1.2449108362197876,
|
| 1656 |
+
"learning_rate": 4.916420845624385e-07,
|
| 1657 |
+
"loss": 0.663,
|
| 1658 |
+
"step": 2240
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 9.091592920353982,
|
| 1662 |
+
"grad_norm": 0.9063498377799988,
|
| 1663 |
+
"learning_rate": 2.4582104228121926e-07,
|
| 1664 |
+
"loss": 0.6794,
|
| 1665 |
+
"step": 2250
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 9.096017699115045,
|
| 1669 |
+
"grad_norm": 1.1721782684326172,
|
| 1670 |
+
"learning_rate": 0.0,
|
| 1671 |
+
"loss": 0.6943,
|
| 1672 |
+
"step": 2260
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 9.096017699115045,
|
| 1676 |
+
"eval_accuracy": 0.7665847665847666,
|
| 1677 |
+
"eval_loss": 0.6571447253227234,
|
| 1678 |
+
"eval_runtime": 1006.2992,
|
| 1679 |
+
"eval_samples_per_second": 2.831,
|
| 1680 |
+
"eval_steps_per_second": 0.089,
|
| 1681 |
+
"step": 2260
|
| 1682 |
+
},
|
| 1683 |
+
{
|
| 1684 |
+
"epoch": 9.096017699115045,
|
| 1685 |
+
"step": 2260,
|
| 1686 |
+
"total_flos": 2.277913504770854e+19,
|
| 1687 |
+
"train_loss": 0.7694057713567684,
|
| 1688 |
+
"train_runtime": 44318.5607,
|
| 1689 |
+
"train_samples_per_second": 1.632,
|
| 1690 |
+
"train_steps_per_second": 0.051
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 9.096017699115045,
|
| 1694 |
+
"eval_accuracy": 0.7652173913043478,
|
| 1695 |
+
"eval_loss": 0.663335919380188,
|
| 1696 |
+
"eval_runtime": 925.6485,
|
| 1697 |
+
"eval_samples_per_second": 2.982,
|
| 1698 |
+
"eval_steps_per_second": 0.094,
|
| 1699 |
+
"step": 2260
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"epoch": 9.096017699115045,
|
| 1703 |
+
"eval_accuracy": 0.7652173913043478,
|
| 1704 |
+
"eval_loss": 0.6633358597755432,
|
| 1705 |
+
"eval_runtime": 903.9019,
|
| 1706 |
+
"eval_samples_per_second": 3.053,
|
| 1707 |
+
"eval_steps_per_second": 0.096,
|
| 1708 |
+
"step": 2260
|
| 1709 |
}
|
| 1710 |
],
|
| 1711 |
"logging_steps": 10,
|
| 1712 |
+
"max_steps": 2260,
|
| 1713 |
"num_input_tokens_seen": 0,
|
| 1714 |
"num_train_epochs": 9223372036854775807,
|
| 1715 |
"save_steps": 500,
|
|
|
|
| 1725 |
"attributes": {}
|
| 1726 |
}
|
| 1727 |
},
|
| 1728 |
+
"total_flos": 2.277913504770854e+19,
|
| 1729 |
+
"train_batch_size": 32,
|
| 1730 |
"trial_name": null,
|
| 1731 |
"trial_params": null
|
| 1732 |
}
|