End of training
Browse files- all_results.json +14 -0
- config.json +1 -1
- eval_results.json +9 -0
- pytorch_model.bin +2 -2
- train_results.json +8 -0
- trainer_state.json +1061 -0
all_results.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.0,
|
| 3 |
+
"eval_loss": 1.739702582359314,
|
| 4 |
+
"eval_runtime": 267.4125,
|
| 5 |
+
"eval_samples": 4216,
|
| 6 |
+
"eval_samples_per_second": 15.766,
|
| 7 |
+
"eval_steps_per_second": 15.766,
|
| 8 |
+
"perplexity": 5.695649184193736,
|
| 9 |
+
"train_loss": 1.8553011305630207,
|
| 10 |
+
"train_runtime": 15007.5766,
|
| 11 |
+
"train_samples": 32857,
|
| 12 |
+
"train_samples_per_second": 4.379,
|
| 13 |
+
"train_steps_per_second": 0.034
|
| 14 |
+
}
|
config.json
CHANGED
|
@@ -24,7 +24,7 @@
|
|
| 24 |
"skip_bias_add": true,
|
| 25 |
"skip_bias_add_qkv": false,
|
| 26 |
"slow_but_exact": false,
|
| 27 |
-
"torch_dtype": "
|
| 28 |
"transformers_version": "4.25.0.dev0",
|
| 29 |
"unk_token_id": 0,
|
| 30 |
"use_cache": true,
|
|
|
|
| 24 |
"skip_bias_add": true,
|
| 25 |
"skip_bias_add_qkv": false,
|
| 26 |
"slow_but_exact": false,
|
| 27 |
+
"torch_dtype": "bfloat16",
|
| 28 |
"transformers_version": "4.25.0.dev0",
|
| 29 |
"unk_token_id": 0,
|
| 30 |
"use_cache": true,
|
eval_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.0,
|
| 3 |
+
"eval_loss": 1.739702582359314,
|
| 4 |
+
"eval_runtime": 267.4125,
|
| 5 |
+
"eval_samples": 4216,
|
| 6 |
+
"eval_samples_per_second": 15.766,
|
| 7 |
+
"eval_steps_per_second": 15.766,
|
| 8 |
+
"perplexity": 5.695649184193736
|
| 9 |
+
}
|
pytorch_model.bin
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 sha256:140ba12f643cb906f6c72db1c86d084804488860f2d3e52dab9092012ad1ed11
|
| 3 |
+
size 2130113377
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.0,
|
| 3 |
+
"train_loss": 1.8553011305630207,
|
| 4 |
+
"train_runtime": 15007.5766,
|
| 5 |
+
"train_samples": 32857,
|
| 6 |
+
"train_samples_per_second": 4.379,
|
| 7 |
+
"train_steps_per_second": 0.034
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1061 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.99726094101893,
|
| 5 |
+
"global_step": 512,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.01,
|
| 12 |
+
"learning_rate": 1.3124999999999999e-05,
|
| 13 |
+
"loss": 3.3565,
|
| 14 |
+
"step": 3
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.02,
|
| 18 |
+
"learning_rate": 2.6249999999999998e-05,
|
| 19 |
+
"loss": 3.1293,
|
| 20 |
+
"step": 6
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.04,
|
| 24 |
+
"learning_rate": 3.9374999999999995e-05,
|
| 25 |
+
"loss": 2.987,
|
| 26 |
+
"step": 9
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.05,
|
| 30 |
+
"learning_rate": 5.2499999999999995e-05,
|
| 31 |
+
"loss": 2.9036,
|
| 32 |
+
"step": 12
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.06,
|
| 36 |
+
"learning_rate": 6.5625e-05,
|
| 37 |
+
"loss": 2.7414,
|
| 38 |
+
"step": 15
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.07,
|
| 42 |
+
"learning_rate": 6.99971917961865e-05,
|
| 43 |
+
"loss": 2.7513,
|
| 44 |
+
"step": 18
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.08,
|
| 48 |
+
"learning_rate": 6.998244995833964e-05,
|
| 49 |
+
"loss": 2.6567,
|
| 50 |
+
"step": 21
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.09,
|
| 54 |
+
"learning_rate": 6.995507775098683e-05,
|
| 55 |
+
"loss": 2.6567,
|
| 56 |
+
"step": 24
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.11,
|
| 60 |
+
"learning_rate": 6.991508505682909e-05,
|
| 61 |
+
"loss": 2.651,
|
| 62 |
+
"step": 27
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.12,
|
| 66 |
+
"learning_rate": 6.986248631517822e-05,
|
| 67 |
+
"loss": 2.5459,
|
| 68 |
+
"step": 30
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.13,
|
| 72 |
+
"learning_rate": 6.979730051674372e-05,
|
| 73 |
+
"loss": 2.5653,
|
| 74 |
+
"step": 33
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.14,
|
| 78 |
+
"learning_rate": 6.9719551196776e-05,
|
| 79 |
+
"loss": 2.5159,
|
| 80 |
+
"step": 36
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.15,
|
| 84 |
+
"learning_rate": 6.962926642656914e-05,
|
| 85 |
+
"loss": 2.4596,
|
| 86 |
+
"step": 39
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.16,
|
| 90 |
+
"learning_rate": 6.952647880332572e-05,
|
| 91 |
+
"loss": 2.5057,
|
| 92 |
+
"step": 42
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.18,
|
| 96 |
+
"learning_rate": 6.941122543838767e-05,
|
| 97 |
+
"loss": 2.4062,
|
| 98 |
+
"step": 45
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.19,
|
| 102 |
+
"learning_rate": 6.92835479438373e-05,
|
| 103 |
+
"loss": 2.44,
|
| 104 |
+
"step": 48
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.2,
|
| 108 |
+
"learning_rate": 6.914349241747322e-05,
|
| 109 |
+
"loss": 2.3659,
|
| 110 |
+
"step": 51
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.21,
|
| 114 |
+
"learning_rate": 6.899110942616686e-05,
|
| 115 |
+
"loss": 2.3672,
|
| 116 |
+
"step": 54
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.22,
|
| 120 |
+
"learning_rate": 6.882645398760536e-05,
|
| 121 |
+
"loss": 2.3759,
|
| 122 |
+
"step": 57
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.23,
|
| 126 |
+
"learning_rate": 6.864958555042743e-05,
|
| 127 |
+
"loss": 2.3567,
|
| 128 |
+
"step": 60
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.25,
|
| 132 |
+
"learning_rate": 6.846056797275964e-05,
|
| 133 |
+
"loss": 2.3606,
|
| 134 |
+
"step": 63
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.26,
|
| 138 |
+
"learning_rate": 6.825946949916035e-05,
|
| 139 |
+
"loss": 2.3166,
|
| 140 |
+
"step": 66
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.27,
|
| 144 |
+
"learning_rate": 6.804636273598024e-05,
|
| 145 |
+
"loss": 2.2707,
|
| 146 |
+
"step": 69
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.28,
|
| 150 |
+
"learning_rate": 6.782132462514781e-05,
|
| 151 |
+
"loss": 2.2617,
|
| 152 |
+
"step": 72
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.29,
|
| 156 |
+
"learning_rate": 6.758443641638958e-05,
|
| 157 |
+
"loss": 2.2521,
|
| 158 |
+
"step": 75
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.3,
|
| 162 |
+
"learning_rate": 6.733578363789503e-05,
|
| 163 |
+
"loss": 2.2746,
|
| 164 |
+
"step": 78
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.32,
|
| 168 |
+
"learning_rate": 6.707545606543672e-05,
|
| 169 |
+
"loss": 2.2112,
|
| 170 |
+
"step": 81
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.33,
|
| 174 |
+
"learning_rate": 6.680354768995687e-05,
|
| 175 |
+
"loss": 2.2237,
|
| 176 |
+
"step": 84
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.34,
|
| 180 |
+
"learning_rate": 6.652015668363205e-05,
|
| 181 |
+
"loss": 2.2225,
|
| 182 |
+
"step": 87
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.35,
|
| 186 |
+
"learning_rate": 6.622538536442822e-05,
|
| 187 |
+
"loss": 2.1471,
|
| 188 |
+
"step": 90
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.36,
|
| 192 |
+
"learning_rate": 6.5919340159159e-05,
|
| 193 |
+
"loss": 2.0738,
|
| 194 |
+
"step": 93
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.37,
|
| 198 |
+
"learning_rate": 6.560213156506037e-05,
|
| 199 |
+
"loss": 2.1797,
|
| 200 |
+
"step": 96
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.39,
|
| 204 |
+
"learning_rate": 6.527387410989579e-05,
|
| 205 |
+
"loss": 2.2545,
|
| 206 |
+
"step": 99
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.4,
|
| 210 |
+
"learning_rate": 6.493468631060607e-05,
|
| 211 |
+
"loss": 2.214,
|
| 212 |
+
"step": 102
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.41,
|
| 216 |
+
"learning_rate": 6.458469063051903e-05,
|
| 217 |
+
"loss": 2.1769,
|
| 218 |
+
"step": 105
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.42,
|
| 222 |
+
"learning_rate": 6.422401343513426e-05,
|
| 223 |
+
"loss": 2.1163,
|
| 224 |
+
"step": 108
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.43,
|
| 228 |
+
"learning_rate": 6.385278494649894e-05,
|
| 229 |
+
"loss": 2.1318,
|
| 230 |
+
"step": 111
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.44,
|
| 234 |
+
"learning_rate": 6.347113919619143e-05,
|
| 235 |
+
"loss": 2.1464,
|
| 236 |
+
"step": 114
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.46,
|
| 240 |
+
"learning_rate": 6.307921397692931e-05,
|
| 241 |
+
"loss": 2.1739,
|
| 242 |
+
"step": 117
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.47,
|
| 246 |
+
"learning_rate": 6.267715079281944e-05,
|
| 247 |
+
"loss": 2.1177,
|
| 248 |
+
"step": 120
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.48,
|
| 252 |
+
"learning_rate": 6.226509480826817e-05,
|
| 253 |
+
"loss": 2.1126,
|
| 254 |
+
"step": 123
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.49,
|
| 258 |
+
"learning_rate": 6.184319479556984e-05,
|
| 259 |
+
"loss": 2.1321,
|
| 260 |
+
"step": 126
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.5,
|
| 264 |
+
"learning_rate": 6.141160308119273e-05,
|
| 265 |
+
"loss": 2.0559,
|
| 266 |
+
"step": 129
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.51,
|
| 270 |
+
"learning_rate": 6.0970475490781874e-05,
|
| 271 |
+
"loss": 2.1131,
|
| 272 |
+
"step": 132
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.53,
|
| 276 |
+
"learning_rate": 6.0519971292898285e-05,
|
| 277 |
+
"loss": 2.1012,
|
| 278 |
+
"step": 135
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.54,
|
| 282 |
+
"learning_rate": 6.0060253141515295e-05,
|
| 283 |
+
"loss": 2.076,
|
| 284 |
+
"step": 138
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.55,
|
| 288 |
+
"learning_rate": 5.95914870172926e-05,
|
| 289 |
+
"loss": 2.0437,
|
| 290 |
+
"step": 141
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.56,
|
| 294 |
+
"learning_rate": 5.911384216764903e-05,
|
| 295 |
+
"loss": 2.0269,
|
| 296 |
+
"step": 144
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.57,
|
| 300 |
+
"learning_rate": 5.862749104565608e-05,
|
| 301 |
+
"loss": 2.0285,
|
| 302 |
+
"step": 147
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.58,
|
| 306 |
+
"learning_rate": 5.8132609247773926e-05,
|
| 307 |
+
"loss": 2.0499,
|
| 308 |
+
"step": 150
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.6,
|
| 312 |
+
"learning_rate": 5.762937545045251e-05,
|
| 313 |
+
"loss": 2.0388,
|
| 314 |
+
"step": 153
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.61,
|
| 318 |
+
"learning_rate": 5.711797134562063e-05,
|
| 319 |
+
"loss": 1.9678,
|
| 320 |
+
"step": 156
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.62,
|
| 324 |
+
"learning_rate": 5.6598581575086404e-05,
|
| 325 |
+
"loss": 2.0444,
|
| 326 |
+
"step": 159
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.63,
|
| 330 |
+
"learning_rate": 5.60713936638724e-05,
|
| 331 |
+
"loss": 2.0038,
|
| 332 |
+
"step": 162
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.64,
|
| 336 |
+
"learning_rate": 5.553659795251013e-05,
|
| 337 |
+
"loss": 1.9812,
|
| 338 |
+
"step": 165
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.65,
|
| 342 |
+
"learning_rate": 5.499438752831773e-05,
|
| 343 |
+
"loss": 1.9962,
|
| 344 |
+
"step": 168
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.67,
|
| 348 |
+
"learning_rate": 5.444495815568607e-05,
|
| 349 |
+
"loss": 1.9857,
|
| 350 |
+
"step": 171
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.68,
|
| 354 |
+
"learning_rate": 5.38885082053983e-05,
|
| 355 |
+
"loss": 1.9288,
|
| 356 |
+
"step": 174
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.69,
|
| 360 |
+
"learning_rate": 5.332523858300823e-05,
|
| 361 |
+
"loss": 1.9271,
|
| 362 |
+
"step": 177
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.7,
|
| 366 |
+
"learning_rate": 5.2755352656303755e-05,
|
| 367 |
+
"loss": 1.9128,
|
| 368 |
+
"step": 180
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.71,
|
| 372 |
+
"learning_rate": 5.217905618188108e-05,
|
| 373 |
+
"loss": 2.0431,
|
| 374 |
+
"step": 183
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.72,
|
| 378 |
+
"learning_rate": 5.1596557230856576e-05,
|
| 379 |
+
"loss": 1.8937,
|
| 380 |
+
"step": 186
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.74,
|
| 384 |
+
"learning_rate": 5.1008066113743024e-05,
|
| 385 |
+
"loss": 1.9296,
|
| 386 |
+
"step": 189
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.75,
|
| 390 |
+
"learning_rate": 5.04137953045172e-05,
|
| 391 |
+
"loss": 1.9901,
|
| 392 |
+
"step": 192
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.76,
|
| 396 |
+
"learning_rate": 4.981395936390644e-05,
|
| 397 |
+
"loss": 1.9698,
|
| 398 |
+
"step": 195
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.77,
|
| 402 |
+
"learning_rate": 4.920877486192174e-05,
|
| 403 |
+
"loss": 1.9076,
|
| 404 |
+
"step": 198
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.78,
|
| 408 |
+
"learning_rate": 4.85984602996655e-05,
|
| 409 |
+
"loss": 1.8708,
|
| 410 |
+
"step": 201
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.79,
|
| 414 |
+
"learning_rate": 4.798323603044187e-05,
|
| 415 |
+
"loss": 1.9115,
|
| 416 |
+
"step": 204
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.81,
|
| 420 |
+
"learning_rate": 4.736332418019853e-05,
|
| 421 |
+
"loss": 1.8812,
|
| 422 |
+
"step": 207
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.82,
|
| 426 |
+
"learning_rate": 4.67389485673284e-05,
|
| 427 |
+
"loss": 1.9152,
|
| 428 |
+
"step": 210
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.83,
|
| 432 |
+
"learning_rate": 4.6110334621860254e-05,
|
| 433 |
+
"loss": 2.0206,
|
| 434 |
+
"step": 213
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.84,
|
| 438 |
+
"learning_rate": 4.547770930406753e-05,
|
| 439 |
+
"loss": 1.9291,
|
| 440 |
+
"step": 216
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.85,
|
| 444 |
+
"learning_rate": 4.4841301022524574e-05,
|
| 445 |
+
"loss": 1.9792,
|
| 446 |
+
"step": 219
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.86,
|
| 450 |
+
"learning_rate": 4.420133955164006e-05,
|
| 451 |
+
"loss": 1.9464,
|
| 452 |
+
"step": 222
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.88,
|
| 456 |
+
"learning_rate": 4.3558055948697185e-05,
|
| 457 |
+
"loss": 1.835,
|
| 458 |
+
"step": 225
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.89,
|
| 462 |
+
"learning_rate": 4.291168247043072e-05,
|
| 463 |
+
"loss": 1.771,
|
| 464 |
+
"step": 228
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.9,
|
| 468 |
+
"learning_rate": 4.2262452489171054e-05,
|
| 469 |
+
"loss": 1.9584,
|
| 470 |
+
"step": 231
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.91,
|
| 474 |
+
"learning_rate": 4.1610600408585395e-05,
|
| 475 |
+
"loss": 2.0061,
|
| 476 |
+
"step": 234
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.92,
|
| 480 |
+
"learning_rate": 4.095636157904658e-05,
|
| 481 |
+
"loss": 1.8876,
|
| 482 |
+
"step": 237
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.93,
|
| 486 |
+
"learning_rate": 4.029997221266018e-05,
|
| 487 |
+
"loss": 1.8706,
|
| 488 |
+
"step": 240
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.95,
|
| 492 |
+
"learning_rate": 3.964166929798036e-05,
|
| 493 |
+
"loss": 1.884,
|
| 494 |
+
"step": 243
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.96,
|
| 498 |
+
"learning_rate": 3.898169051444552e-05,
|
| 499 |
+
"loss": 1.8348,
|
| 500 |
+
"step": 246
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.97,
|
| 504 |
+
"learning_rate": 3.8320274146564356e-05,
|
| 505 |
+
"loss": 1.8026,
|
| 506 |
+
"step": 249
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.98,
|
| 510 |
+
"learning_rate": 3.7657658997883615e-05,
|
| 511 |
+
"loss": 1.847,
|
| 512 |
+
"step": 252
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.99,
|
| 516 |
+
"learning_rate": 3.699408430476834e-05,
|
| 517 |
+
"loss": 1.8465,
|
| 518 |
+
"step": 255
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 1.0,
|
| 522 |
+
"eval_loss": 1.8656249046325684,
|
| 523 |
+
"eval_runtime": 309.3485,
|
| 524 |
+
"eval_samples_per_second": 13.629,
|
| 525 |
+
"eval_steps_per_second": 13.629,
|
| 526 |
+
"step": 256
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 1.01,
|
| 530 |
+
"learning_rate": 3.632978965002587e-05,
|
| 531 |
+
"loss": 2.1828,
|
| 532 |
+
"step": 258
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 1.02,
|
| 536 |
+
"learning_rate": 3.566501487640479e-05,
|
| 537 |
+
"loss": 1.6617,
|
| 538 |
+
"step": 261
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 1.03,
|
| 542 |
+
"learning_rate": 3.5e-05,
|
| 543 |
+
"loss": 1.6496,
|
| 544 |
+
"step": 264
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 1.04,
|
| 548 |
+
"learning_rate": 3.433498512359521e-05,
|
| 549 |
+
"loss": 1.5581,
|
| 550 |
+
"step": 267
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.05,
|
| 554 |
+
"learning_rate": 3.367021034997412e-05,
|
| 555 |
+
"loss": 1.6121,
|
| 556 |
+
"step": 270
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.07,
|
| 560 |
+
"learning_rate": 3.300591569523165e-05,
|
| 561 |
+
"loss": 1.6664,
|
| 562 |
+
"step": 273
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.08,
|
| 566 |
+
"learning_rate": 3.2342341002116385e-05,
|
| 567 |
+
"loss": 1.5229,
|
| 568 |
+
"step": 276
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 1.09,
|
| 572 |
+
"learning_rate": 3.1679725853435645e-05,
|
| 573 |
+
"loss": 1.5276,
|
| 574 |
+
"step": 279
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 1.1,
|
| 578 |
+
"learning_rate": 3.101830948555448e-05,
|
| 579 |
+
"loss": 1.6506,
|
| 580 |
+
"step": 282
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 1.11,
|
| 584 |
+
"learning_rate": 3.035833070201963e-05,
|
| 585 |
+
"loss": 1.5594,
|
| 586 |
+
"step": 285
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 1.12,
|
| 590 |
+
"learning_rate": 2.9700027787339826e-05,
|
| 591 |
+
"loss": 1.4913,
|
| 592 |
+
"step": 288
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.14,
|
| 596 |
+
"learning_rate": 2.904363842095341e-05,
|
| 597 |
+
"loss": 1.6122,
|
| 598 |
+
"step": 291
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.15,
|
| 602 |
+
"learning_rate": 2.838939959141461e-05,
|
| 603 |
+
"loss": 1.6086,
|
| 604 |
+
"step": 294
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.16,
|
| 608 |
+
"learning_rate": 2.7737547510828943e-05,
|
| 609 |
+
"loss": 1.5963,
|
| 610 |
+
"step": 297
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 1.17,
|
| 614 |
+
"learning_rate": 2.7088317529569277e-05,
|
| 615 |
+
"loss": 1.6068,
|
| 616 |
+
"step": 300
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 1.18,
|
| 620 |
+
"learning_rate": 2.6441944051302816e-05,
|
| 621 |
+
"loss": 1.528,
|
| 622 |
+
"step": 303
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 1.19,
|
| 626 |
+
"learning_rate": 2.5798660448359928e-05,
|
| 627 |
+
"loss": 1.629,
|
| 628 |
+
"step": 306
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 1.21,
|
| 632 |
+
"learning_rate": 2.5158698977475426e-05,
|
| 633 |
+
"loss": 1.6335,
|
| 634 |
+
"step": 309
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 1.22,
|
| 638 |
+
"learning_rate": 2.4522290695932468e-05,
|
| 639 |
+
"loss": 1.5641,
|
| 640 |
+
"step": 312
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.23,
|
| 644 |
+
"learning_rate": 2.3889665378139753e-05,
|
| 645 |
+
"loss": 1.5514,
|
| 646 |
+
"step": 315
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.24,
|
| 650 |
+
"learning_rate": 2.32610514326716e-05,
|
| 651 |
+
"loss": 1.6076,
|
| 652 |
+
"step": 318
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 1.25,
|
| 656 |
+
"learning_rate": 2.263667581980147e-05,
|
| 657 |
+
"loss": 1.5815,
|
| 658 |
+
"step": 321
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 1.26,
|
| 662 |
+
"learning_rate": 2.2016763969558128e-05,
|
| 663 |
+
"loss": 1.5281,
|
| 664 |
+
"step": 324
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 1.28,
|
| 668 |
+
"learning_rate": 2.140153970033449e-05,
|
| 669 |
+
"loss": 1.5254,
|
| 670 |
+
"step": 327
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 1.29,
|
| 674 |
+
"learning_rate": 2.0791225138078253e-05,
|
| 675 |
+
"loss": 1.5966,
|
| 676 |
+
"step": 330
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 1.3,
|
| 680 |
+
"learning_rate": 2.0186040636093567e-05,
|
| 681 |
+
"loss": 1.5427,
|
| 682 |
+
"step": 333
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.31,
|
| 686 |
+
"learning_rate": 1.9586204695482795e-05,
|
| 687 |
+
"loss": 1.6041,
|
| 688 |
+
"step": 336
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.32,
|
| 692 |
+
"learning_rate": 1.8991933886256963e-05,
|
| 693 |
+
"loss": 1.4729,
|
| 694 |
+
"step": 339
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 1.34,
|
| 698 |
+
"learning_rate": 1.840344276914342e-05,
|
| 699 |
+
"loss": 1.564,
|
| 700 |
+
"step": 342
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"epoch": 1.35,
|
| 704 |
+
"learning_rate": 1.7820943818118924e-05,
|
| 705 |
+
"loss": 1.5334,
|
| 706 |
+
"step": 345
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 1.36,
|
| 710 |
+
"learning_rate": 1.7244647343696252e-05,
|
| 711 |
+
"loss": 1.5026,
|
| 712 |
+
"step": 348
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 1.37,
|
| 716 |
+
"learning_rate": 1.6674761416991767e-05,
|
| 717 |
+
"loss": 1.586,
|
| 718 |
+
"step": 351
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 1.38,
|
| 722 |
+
"learning_rate": 1.611149179460171e-05,
|
| 723 |
+
"loss": 1.5368,
|
| 724 |
+
"step": 354
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.39,
|
| 728 |
+
"learning_rate": 1.555504184431393e-05,
|
| 729 |
+
"loss": 1.5588,
|
| 730 |
+
"step": 357
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.41,
|
| 734 |
+
"learning_rate": 1.500561247168226e-05,
|
| 735 |
+
"loss": 1.5276,
|
| 736 |
+
"step": 360
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 1.42,
|
| 740 |
+
"learning_rate": 1.446340204748987e-05,
|
| 741 |
+
"loss": 1.4957,
|
| 742 |
+
"step": 363
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 1.43,
|
| 746 |
+
"learning_rate": 1.3928606336127589e-05,
|
| 747 |
+
"loss": 1.5479,
|
| 748 |
+
"step": 366
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 1.44,
|
| 752 |
+
"learning_rate": 1.34014184249136e-05,
|
| 753 |
+
"loss": 1.4862,
|
| 754 |
+
"step": 369
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 1.45,
|
| 758 |
+
"learning_rate": 1.2882028654379362e-05,
|
| 759 |
+
"loss": 1.5849,
|
| 760 |
+
"step": 372
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 1.46,
|
| 764 |
+
"learning_rate": 1.2370624549547507e-05,
|
| 765 |
+
"loss": 1.5811,
|
| 766 |
+
"step": 375
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.48,
|
| 770 |
+
"learning_rate": 1.186739075222608e-05,
|
| 771 |
+
"loss": 1.4914,
|
| 772 |
+
"step": 378
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.49,
|
| 776 |
+
"learning_rate": 1.1372508954343916e-05,
|
| 777 |
+
"loss": 1.5442,
|
| 778 |
+
"step": 381
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 1.5,
|
| 782 |
+
"learning_rate": 1.0886157832350968e-05,
|
| 783 |
+
"loss": 1.528,
|
| 784 |
+
"step": 384
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 1.51,
|
| 788 |
+
"learning_rate": 1.0408512982707408e-05,
|
| 789 |
+
"loss": 1.51,
|
| 790 |
+
"step": 387
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 1.52,
|
| 794 |
+
"learning_rate": 9.939746858484699e-06,
|
| 795 |
+
"loss": 1.4226,
|
| 796 |
+
"step": 390
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 1.53,
|
| 800 |
+
"learning_rate": 9.480028707101716e-06,
|
| 801 |
+
"loss": 1.5474,
|
| 802 |
+
"step": 393
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 1.55,
|
| 806 |
+
"learning_rate": 9.02952450921813e-06,
|
| 807 |
+
"loss": 1.5436,
|
| 808 |
+
"step": 396
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.56,
|
| 812 |
+
"learning_rate": 8.588396918807265e-06,
|
| 813 |
+
"loss": 1.5102,
|
| 814 |
+
"step": 399
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.57,
|
| 818 |
+
"learning_rate": 8.156805204430163e-06,
|
| 819 |
+
"loss": 1.4774,
|
| 820 |
+
"step": 402
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 1.58,
|
| 824 |
+
"learning_rate": 7.734905191731818e-06,
|
| 825 |
+
"loss": 1.5731,
|
| 826 |
+
"step": 405
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 1.59,
|
| 830 |
+
"learning_rate": 7.322849207180554e-06,
|
| 831 |
+
"loss": 1.4854,
|
| 832 |
+
"step": 408
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 1.6,
|
| 836 |
+
"learning_rate": 6.92078602307069e-06,
|
| 837 |
+
"loss": 1.5024,
|
| 838 |
+
"step": 411
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 1.62,
|
| 842 |
+
"learning_rate": 6.528860803808572e-06,
|
| 843 |
+
"loss": 1.4762,
|
| 844 |
+
"step": 414
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 1.63,
|
| 848 |
+
"learning_rate": 6.147215053501067e-06,
|
| 849 |
+
"loss": 1.5015,
|
| 850 |
+
"step": 417
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.64,
|
| 854 |
+
"learning_rate": 5.775986564865746e-06,
|
| 855 |
+
"loss": 1.4655,
|
| 856 |
+
"step": 420
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.65,
|
| 860 |
+
"learning_rate": 5.4153093694809615e-06,
|
| 861 |
+
"loss": 1.3807,
|
| 862 |
+
"step": 423
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 1.66,
|
| 866 |
+
"learning_rate": 5.065313689393926e-06,
|
| 867 |
+
"loss": 1.5052,
|
| 868 |
+
"step": 426
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 1.67,
|
| 872 |
+
"learning_rate": 4.7261258901042164e-06,
|
| 873 |
+
"loss": 1.5502,
|
| 874 |
+
"step": 429
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 1.69,
|
| 878 |
+
"learning_rate": 4.397868434939627e-06,
|
| 879 |
+
"loss": 1.4234,
|
| 880 |
+
"step": 432
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 1.7,
|
| 884 |
+
"learning_rate": 4.080659840841001e-06,
|
| 885 |
+
"loss": 1.4875,
|
| 886 |
+
"step": 435
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 1.71,
|
| 890 |
+
"learning_rate": 3.774614635571785e-06,
|
| 891 |
+
"loss": 1.4858,
|
| 892 |
+
"step": 438
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.72,
|
| 896 |
+
"learning_rate": 3.4798433163679534e-06,
|
| 897 |
+
"loss": 1.4217,
|
| 898 |
+
"step": 441
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.73,
|
| 902 |
+
"learning_rate": 3.1964523100431278e-06,
|
| 903 |
+
"loss": 1.4572,
|
| 904 |
+
"step": 444
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 1.74,
|
| 908 |
+
"learning_rate": 2.924543934563277e-06,
|
| 909 |
+
"loss": 1.5039,
|
| 910 |
+
"step": 447
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 1.76,
|
| 914 |
+
"learning_rate": 2.664216362104964e-06,
|
| 915 |
+
"loss": 1.4897,
|
| 916 |
+
"step": 450
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 1.77,
|
| 920 |
+
"learning_rate": 2.4155635836104113e-06,
|
| 921 |
+
"loss": 1.5384,
|
| 922 |
+
"step": 453
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 1.78,
|
| 926 |
+
"learning_rate": 2.178675374852189e-06,
|
| 927 |
+
"loss": 1.505,
|
| 928 |
+
"step": 456
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 1.79,
|
| 932 |
+
"learning_rate": 1.953637264019748e-06,
|
| 933 |
+
"loss": 1.5019,
|
| 934 |
+
"step": 459
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.8,
|
| 938 |
+
"learning_rate": 1.7405305008396436e-06,
|
| 939 |
+
"loss": 1.5941,
|
| 940 |
+
"step": 462
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.81,
|
| 944 |
+
"learning_rate": 1.5394320272403605e-06,
|
| 945 |
+
"loss": 1.5265,
|
| 946 |
+
"step": 465
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 1.83,
|
| 950 |
+
"learning_rate": 1.3504144495725661e-06,
|
| 951 |
+
"loss": 1.5035,
|
| 952 |
+
"step": 468
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 1.84,
|
| 956 |
+
"learning_rate": 1.1735460123946455e-06,
|
| 957 |
+
"loss": 1.5002,
|
| 958 |
+
"step": 471
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 1.85,
|
| 962 |
+
"learning_rate": 1.0088905738331372e-06,
|
| 963 |
+
"loss": 1.4317,
|
| 964 |
+
"step": 474
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 1.86,
|
| 968 |
+
"learning_rate": 8.565075825267759e-07,
|
| 969 |
+
"loss": 1.4852,
|
| 970 |
+
"step": 477
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 1.87,
|
| 974 |
+
"learning_rate": 7.16452056162693e-07,
|
| 975 |
+
"loss": 1.5445,
|
| 976 |
+
"step": 480
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.88,
|
| 980 |
+
"learning_rate": 5.887745616123169e-07,
|
| 981 |
+
"loss": 1.4855,
|
| 982 |
+
"step": 483
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.9,
|
| 986 |
+
"learning_rate": 4.735211966742819e-07,
|
| 987 |
+
"loss": 1.4852,
|
| 988 |
+
"step": 486
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 1.91,
|
| 992 |
+
"learning_rate": 3.7073357343086086e-07,
|
| 993 |
+
"loss": 1.4545,
|
| 994 |
+
"step": 489
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 1.92,
|
| 998 |
+
"learning_rate": 2.804488032239943e-07,
|
| 999 |
+
"loss": 1.445,
|
| 1000 |
+
"step": 492
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 1.93,
|
| 1004 |
+
"learning_rate": 2.0269948325627227e-07,
|
| 1005 |
+
"loss": 1.4484,
|
| 1006 |
+
"step": 495
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 1.94,
|
| 1010 |
+
"learning_rate": 1.3751368482176727e-07,
|
| 1011 |
+
"loss": 1.4113,
|
| 1012 |
+
"step": 498
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 1.95,
|
| 1016 |
+
"learning_rate": 8.491494317091229e-08,
|
| 1017 |
+
"loss": 1.5394,
|
| 1018 |
+
"step": 501
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.97,
|
| 1022 |
+
"learning_rate": 4.492224901315344e-08,
|
| 1023 |
+
"loss": 1.4475,
|
| 1024 |
+
"step": 504
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.98,
|
| 1028 |
+
"learning_rate": 1.7550041660350232e-08,
|
| 1029 |
+
"loss": 1.4934,
|
| 1030 |
+
"step": 507
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 1.99,
|
| 1034 |
+
"learning_rate": 2.808203813499665e-09,
|
| 1035 |
+
"loss": 1.4903,
|
| 1036 |
+
"step": 510
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 2.0,
|
| 1040 |
+
"eval_loss": 1.7396166324615479,
|
| 1041 |
+
"eval_runtime": 309.3927,
|
| 1042 |
+
"eval_samples_per_second": 13.627,
|
| 1043 |
+
"eval_steps_per_second": 13.627,
|
| 1044 |
+
"step": 512
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 2.0,
|
| 1048 |
+
"step": 512,
|
| 1049 |
+
"total_flos": 2.7416092993388544e+17,
|
| 1050 |
+
"train_loss": 1.8553011305630207,
|
| 1051 |
+
"train_runtime": 15007.5766,
|
| 1052 |
+
"train_samples_per_second": 4.379,
|
| 1053 |
+
"train_steps_per_second": 0.034
|
| 1054 |
+
}
|
| 1055 |
+
],
|
| 1056 |
+
"max_steps": 512,
|
| 1057 |
+
"num_train_epochs": 2,
|
| 1058 |
+
"total_flos": 2.7416092993388544e+17,
|
| 1059 |
+
"trial_name": null,
|
| 1060 |
+
"trial_params": null
|
| 1061 |
+
}
|