Commit
•
8a2f935
1
Parent(s):
ded539e
End of training
Browse files- all_results.json +20 -20
- eval_pretrained_results.json +5 -5
- huggingface_training.py +2 -2
- pytorch_model.bin +1 -1
- test_finetuned_results.json +6 -6
- test_pretrained_results.json +5 -5
- train_results.json +5 -5
- trainer_state.json +764 -32
- training_args.bin +1 -1
all_results.json
CHANGED
|
@@ -1,32 +1,32 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
"eval_loss": 1.6191972494125366,
|
| 4 |
-
"eval_pretrained_loss": 1.
|
| 5 |
-
"eval_pretrained_runtime":
|
| 6 |
-
"eval_pretrained_samples_per_second":
|
| 7 |
-
"eval_pretrained_steps_per_second": 0.
|
| 8 |
-
"eval_pretrained_wer":
|
| 9 |
"eval_runtime": 56.3363,
|
| 10 |
"eval_samples_per_second": 0.071,
|
| 11 |
"eval_steps_per_second": 0.036,
|
| 12 |
"eval_wer": 153.2258064516129,
|
| 13 |
-
"test_finetuned_loss": 1.
|
| 14 |
-
"test_finetuned_runtime":
|
| 15 |
-
"test_finetuned_samples_per_second":
|
| 16 |
-
"test_finetuned_steps_per_second": 0.
|
| 17 |
-
"test_finetuned_wer":
|
| 18 |
"test_loss": 1.7568330764770508,
|
| 19 |
-
"test_pretrained_loss": 1.
|
| 20 |
-
"test_pretrained_runtime":
|
| 21 |
-
"test_pretrained_samples_per_second":
|
| 22 |
-
"test_pretrained_steps_per_second": 0.
|
| 23 |
-
"test_pretrained_wer":
|
| 24 |
"test_runtime": 37.8582,
|
| 25 |
"test_samples_per_second": 0.106,
|
| 26 |
"test_steps_per_second": 0.053,
|
| 27 |
"test_wer": 138.5964912280702,
|
| 28 |
-
"train_loss":
|
| 29 |
-
"train_runtime":
|
| 30 |
-
"train_samples_per_second":
|
| 31 |
-
"train_steps_per_second": 0.
|
| 32 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 2.32,
|
| 3 |
"eval_loss": 1.6191972494125366,
|
| 4 |
+
"eval_pretrained_loss": 1.710707187652588,
|
| 5 |
+
"eval_pretrained_runtime": 113.3653,
|
| 6 |
+
"eval_pretrained_samples_per_second": 2.911,
|
| 7 |
+
"eval_pretrained_steps_per_second": 0.097,
|
| 8 |
+
"eval_pretrained_wer": 258.2308797700783,
|
| 9 |
"eval_runtime": 56.3363,
|
| 10 |
"eval_samples_per_second": 0.071,
|
| 11 |
"eval_steps_per_second": 0.036,
|
| 12 |
"eval_wer": 153.2258064516129,
|
| 13 |
+
"test_finetuned_loss": 1.0526511669158936,
|
| 14 |
+
"test_finetuned_runtime": 167.4865,
|
| 15 |
+
"test_finetuned_samples_per_second": 4.532,
|
| 16 |
+
"test_finetuned_steps_per_second": 0.143,
|
| 17 |
+
"test_finetuned_wer": 176.51198300471353,
|
| 18 |
"test_loss": 1.7568330764770508,
|
| 19 |
+
"test_pretrained_loss": 1.723103642463684,
|
| 20 |
+
"test_pretrained_runtime": 171.9859,
|
| 21 |
+
"test_pretrained_samples_per_second": 4.413,
|
| 22 |
+
"test_pretrained_steps_per_second": 0.14,
|
| 23 |
+
"test_pretrained_wer": 261.9265750514506,
|
| 24 |
"test_runtime": 37.8582,
|
| 25 |
"test_samples_per_second": 0.106,
|
| 26 |
"test_steps_per_second": 0.053,
|
| 27 |
"test_wer": 138.5964912280702,
|
| 28 |
+
"train_loss": 0.9721650715385165,
|
| 29 |
+
"train_runtime": 1491.7987,
|
| 30 |
+
"train_samples_per_second": 4.805,
|
| 31 |
+
"train_steps_per_second": 0.075
|
| 32 |
}
|
eval_pretrained_results.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
-
"eval_pretrained_loss": 1.
|
| 3 |
-
"eval_pretrained_runtime":
|
| 4 |
-
"eval_pretrained_samples_per_second":
|
| 5 |
-
"eval_pretrained_steps_per_second": 0.
|
| 6 |
-
"eval_pretrained_wer":
|
| 7 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"eval_pretrained_loss": 1.710707187652588,
|
| 3 |
+
"eval_pretrained_runtime": 113.3653,
|
| 4 |
+
"eval_pretrained_samples_per_second": 2.911,
|
| 5 |
+
"eval_pretrained_steps_per_second": 0.097,
|
| 6 |
+
"eval_pretrained_wer": 258.2308797700783
|
| 7 |
}
|
huggingface_training.py
CHANGED
|
@@ -35,8 +35,8 @@ dataset = load_dataset(dataset_id, dataset_language_code, streaming=True)
|
|
| 35 |
|
| 36 |
"""The first time you run this code, make sure everything works fine using a small sample and low number of training steps. Just uncomment the next cell and run it. One note: since the dataset is loaded in streaming mode, the instruction will not be executed immediately. Instead, the dataset will be subsampled only when data will be needed during training."""
|
| 37 |
|
| 38 |
-
test_script = True
|
| 39 |
-
|
| 40 |
|
| 41 |
## Sample dataset for testing
|
| 42 |
if test_script is True:
|
|
|
|
| 35 |
|
| 36 |
"""The first time you run this code, make sure everything works fine using a small sample and low number of training steps. Just uncomment the next cell and run it. One note: since the dataset is loaded in streaming mode, the instruction will not be executed immediately. Instead, the dataset will be subsampled only when data will be needed during training."""
|
| 37 |
|
| 38 |
+
# test_script = True
|
| 39 |
+
test_script = False
|
| 40 |
|
| 41 |
## Sample dataset for testing
|
| 42 |
if test_script is True:
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 151098921
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c74db4ae6de9cc4fe7ef67b639a61333e374adb566ec56ff72f1129f43c25bf4
|
| 3 |
size 151098921
|
test_finetuned_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"test_finetuned_loss": 1.
|
| 4 |
-
"test_finetuned_runtime":
|
| 5 |
-
"test_finetuned_samples_per_second":
|
| 6 |
-
"test_finetuned_steps_per_second": 0.
|
| 7 |
-
"test_finetuned_wer":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 2.32,
|
| 3 |
+
"test_finetuned_loss": 1.0526511669158936,
|
| 4 |
+
"test_finetuned_runtime": 167.4865,
|
| 5 |
+
"test_finetuned_samples_per_second": 4.532,
|
| 6 |
+
"test_finetuned_steps_per_second": 0.143,
|
| 7 |
+
"test_finetuned_wer": 176.51198300471353
|
| 8 |
}
|
test_pretrained_results.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
-
"test_pretrained_loss": 1.
|
| 3 |
-
"test_pretrained_runtime":
|
| 4 |
-
"test_pretrained_samples_per_second":
|
| 5 |
-
"test_pretrained_steps_per_second": 0.
|
| 6 |
-
"test_pretrained_wer":
|
| 7 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"test_pretrained_loss": 1.723103642463684,
|
| 3 |
+
"test_pretrained_runtime": 171.9859,
|
| 4 |
+
"test_pretrained_samples_per_second": 4.413,
|
| 5 |
+
"test_pretrained_steps_per_second": 0.14,
|
| 6 |
+
"test_pretrained_wer": 261.9265750514506
|
| 7 |
}
|
train_results.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"train_loss":
|
| 4 |
-
"train_runtime":
|
| 5 |
-
"train_samples_per_second":
|
| 6 |
-
"train_steps_per_second": 0.
|
| 7 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 2.32,
|
| 3 |
+
"train_loss": 0.9721650715385165,
|
| 4 |
+
"train_runtime": 1491.7987,
|
| 5 |
+
"train_samples_per_second": 4.805,
|
| 6 |
+
"train_steps_per_second": 0.075
|
| 7 |
}
|
trainer_state.json
CHANGED
|
@@ -1,55 +1,787 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric":
|
| 3 |
-
"best_model_checkpoint": "./checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
-
"global_step":
|
| 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.
|
| 12 |
"learning_rate": 0.0,
|
| 13 |
-
"loss": 1.
|
| 14 |
"step": 1
|
| 15 |
},
|
| 16 |
{
|
| 17 |
-
"epoch": 0.
|
| 18 |
-
"
|
| 19 |
-
"
|
| 20 |
-
"
|
| 21 |
-
|
| 22 |
-
|
| 23 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 27 |
"learning_rate": 1e-05,
|
| 28 |
-
"loss": 1.
|
| 29 |
-
"step":
|
| 30 |
},
|
| 31 |
{
|
| 32 |
-
"epoch":
|
| 33 |
-
"
|
| 34 |
-
"
|
| 35 |
-
"
|
| 36 |
-
|
| 37 |
-
|
| 38 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 39 |
},
|
| 40 |
{
|
| 41 |
-
"epoch":
|
| 42 |
-
"step":
|
| 43 |
-
"total_flos":
|
| 44 |
-
"train_loss":
|
| 45 |
-
"train_runtime":
|
| 46 |
-
"train_samples_per_second":
|
| 47 |
-
"train_steps_per_second": 0.
|
| 48 |
}
|
| 49 |
],
|
| 50 |
-
"max_steps":
|
| 51 |
"num_train_epochs": 9223372036854775807,
|
| 52 |
-
"total_flos":
|
| 53 |
"trial_name": null,
|
| 54 |
"trial_params": null
|
| 55 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 158.1031454574485,
|
| 3 |
+
"best_model_checkpoint": "./checkpoint-88",
|
| 4 |
+
"epoch": 2.3214285714285716,
|
| 5 |
+
"global_step": 112,
|
| 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": 0.0,
|
| 13 |
+
"loss": 1.6569,
|
| 14 |
"step": 1
|
| 15 |
},
|
| 16 |
{
|
| 17 |
+
"epoch": 0.02,
|
| 18 |
+
"learning_rate": 4.347826086956522e-07,
|
| 19 |
+
"loss": 1.6564,
|
| 20 |
+
"step": 2
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.03,
|
| 24 |
+
"learning_rate": 8.695652173913044e-07,
|
| 25 |
+
"loss": 1.6518,
|
| 26 |
+
"step": 3
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.04,
|
| 30 |
+
"learning_rate": 1.3043478260869566e-06,
|
| 31 |
+
"loss": 1.6233,
|
| 32 |
+
"step": 4
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.04,
|
| 36 |
+
"learning_rate": 1.7391304347826088e-06,
|
| 37 |
+
"loss": 1.712,
|
| 38 |
+
"step": 5
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.05,
|
| 42 |
+
"learning_rate": 2.173913043478261e-06,
|
| 43 |
+
"loss": 1.6476,
|
| 44 |
+
"step": 6
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.06,
|
| 48 |
+
"learning_rate": 2.6086956521739132e-06,
|
| 49 |
+
"loss": 1.5716,
|
| 50 |
+
"step": 7
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.07,
|
| 54 |
+
"learning_rate": 3.043478260869566e-06,
|
| 55 |
+
"loss": 1.5909,
|
| 56 |
+
"step": 8
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.08,
|
| 60 |
+
"learning_rate": 3.4782608695652175e-06,
|
| 61 |
+
"loss": 1.5905,
|
| 62 |
+
"step": 9
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.09,
|
| 66 |
+
"learning_rate": 3.91304347826087e-06,
|
| 67 |
+
"loss": 1.5486,
|
| 68 |
+
"step": 10
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.1,
|
| 72 |
+
"learning_rate": 4.347826086956522e-06,
|
| 73 |
+
"loss": 1.5299,
|
| 74 |
+
"step": 11
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.1,
|
| 78 |
+
"eval_loss": 1.5621598958969116,
|
| 79 |
+
"eval_runtime": 105.0814,
|
| 80 |
+
"eval_samples_per_second": 3.14,
|
| 81 |
+
"eval_steps_per_second": 0.105,
|
| 82 |
+
"eval_wer": 219.67108414497844,
|
| 83 |
+
"step": 11
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 0.11,
|
| 87 |
+
"learning_rate": 4.782608695652174e-06,
|
| 88 |
+
"loss": 1.4537,
|
| 89 |
+
"step": 12
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.12,
|
| 93 |
+
"learning_rate": 5.2173913043478265e-06,
|
| 94 |
+
"loss": 1.3729,
|
| 95 |
+
"step": 13
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 0.12,
|
| 99 |
+
"learning_rate": 5.652173913043479e-06,
|
| 100 |
+
"loss": 1.4921,
|
| 101 |
+
"step": 14
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.13,
|
| 105 |
+
"learning_rate": 6.086956521739132e-06,
|
| 106 |
+
"loss": 1.5069,
|
| 107 |
+
"step": 15
|
| 108 |
},
|
| 109 |
{
|
| 110 |
+
"epoch": 0.14,
|
| 111 |
+
"learning_rate": 6.521739130434783e-06,
|
| 112 |
+
"loss": 1.388,
|
| 113 |
+
"step": 16
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 0.15,
|
| 117 |
+
"learning_rate": 6.956521739130435e-06,
|
| 118 |
+
"loss": 1.3857,
|
| 119 |
+
"step": 17
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 0.16,
|
| 123 |
+
"learning_rate": 7.391304347826087e-06,
|
| 124 |
+
"loss": 1.3389,
|
| 125 |
+
"step": 18
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 0.17,
|
| 129 |
+
"learning_rate": 7.82608695652174e-06,
|
| 130 |
+
"loss": 1.3089,
|
| 131 |
+
"step": 19
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.18,
|
| 135 |
+
"learning_rate": 8.260869565217392e-06,
|
| 136 |
+
"loss": 1.2118,
|
| 137 |
+
"step": 20
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.19,
|
| 141 |
+
"learning_rate": 8.695652173913044e-06,
|
| 142 |
+
"loss": 1.1634,
|
| 143 |
+
"step": 21
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.2,
|
| 147 |
+
"learning_rate": 9.130434782608697e-06,
|
| 148 |
+
"loss": 1.1908,
|
| 149 |
+
"step": 22
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.2,
|
| 153 |
+
"eval_loss": 1.3651723861694336,
|
| 154 |
+
"eval_runtime": 110.8115,
|
| 155 |
+
"eval_samples_per_second": 2.978,
|
| 156 |
+
"eval_steps_per_second": 0.099,
|
| 157 |
+
"eval_wer": 192.2401405077439,
|
| 158 |
+
"step": 22
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.21,
|
| 162 |
+
"learning_rate": 9.565217391304349e-06,
|
| 163 |
+
"loss": 1.1192,
|
| 164 |
+
"step": 23
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.21,
|
| 168 |
"learning_rate": 1e-05,
|
| 169 |
+
"loss": 1.2042,
|
| 170 |
+
"step": 24
|
| 171 |
},
|
| 172 |
{
|
| 173 |
+
"epoch": 0.22,
|
| 174 |
+
"learning_rate": 9.887640449438202e-06,
|
| 175 |
+
"loss": 1.1448,
|
| 176 |
+
"step": 25
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.23,
|
| 180 |
+
"learning_rate": 9.775280898876405e-06,
|
| 181 |
+
"loss": 1.1393,
|
| 182 |
+
"step": 26
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.24,
|
| 186 |
+
"learning_rate": 9.662921348314608e-06,
|
| 187 |
+
"loss": 1.1482,
|
| 188 |
+
"step": 27
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.25,
|
| 192 |
+
"learning_rate": 9.55056179775281e-06,
|
| 193 |
+
"loss": 1.179,
|
| 194 |
+
"step": 28
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.26,
|
| 198 |
+
"learning_rate": 9.438202247191012e-06,
|
| 199 |
+
"loss": 0.9847,
|
| 200 |
+
"step": 29
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.27,
|
| 204 |
+
"learning_rate": 9.325842696629213e-06,
|
| 205 |
+
"loss": 1.1149,
|
| 206 |
+
"step": 30
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.28,
|
| 210 |
+
"learning_rate": 9.213483146067417e-06,
|
| 211 |
+
"loss": 1.0657,
|
| 212 |
+
"step": 31
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.29,
|
| 216 |
+
"learning_rate": 9.101123595505619e-06,
|
| 217 |
+
"loss": 1.0028,
|
| 218 |
+
"step": 32
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.29,
|
| 222 |
+
"learning_rate": 8.988764044943822e-06,
|
| 223 |
+
"loss": 1.1161,
|
| 224 |
+
"step": 33
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.29,
|
| 228 |
+
"eval_loss": 1.1921106576919556,
|
| 229 |
+
"eval_runtime": 103.0628,
|
| 230 |
+
"eval_samples_per_second": 3.202,
|
| 231 |
+
"eval_steps_per_second": 0.107,
|
| 232 |
+
"eval_wer": 200.23950183618075,
|
| 233 |
+
"step": 33
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.3,
|
| 237 |
+
"learning_rate": 8.876404494382023e-06,
|
| 238 |
+
"loss": 1.0931,
|
| 239 |
+
"step": 34
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.31,
|
| 243 |
+
"learning_rate": 8.764044943820226e-06,
|
| 244 |
+
"loss": 1.0635,
|
| 245 |
+
"step": 35
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 0.32,
|
| 249 |
+
"learning_rate": 8.651685393258428e-06,
|
| 250 |
+
"loss": 1.0671,
|
| 251 |
+
"step": 36
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.33,
|
| 255 |
+
"learning_rate": 8.53932584269663e-06,
|
| 256 |
+
"loss": 1.0251,
|
| 257 |
+
"step": 37
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.34,
|
| 261 |
+
"learning_rate": 8.426966292134832e-06,
|
| 262 |
+
"loss": 1.0486,
|
| 263 |
+
"step": 38
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 1.01,
|
| 267 |
+
"learning_rate": 8.314606741573035e-06,
|
| 268 |
+
"loss": 0.9926,
|
| 269 |
+
"step": 39
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 1.02,
|
| 273 |
+
"learning_rate": 8.202247191011237e-06,
|
| 274 |
+
"loss": 1.0232,
|
| 275 |
+
"step": 40
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.03,
|
| 279 |
+
"learning_rate": 8.08988764044944e-06,
|
| 280 |
+
"loss": 0.9421,
|
| 281 |
+
"step": 41
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 1.04,
|
| 285 |
+
"learning_rate": 7.97752808988764e-06,
|
| 286 |
+
"loss": 0.9541,
|
| 287 |
+
"step": 42
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 1.04,
|
| 291 |
+
"learning_rate": 7.865168539325843e-06,
|
| 292 |
+
"loss": 0.9639,
|
| 293 |
+
"step": 43
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 1.05,
|
| 297 |
+
"learning_rate": 7.752808988764046e-06,
|
| 298 |
+
"loss": 0.9216,
|
| 299 |
+
"step": 44
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 1.05,
|
| 303 |
+
"eval_loss": 1.1263455152511597,
|
| 304 |
+
"eval_runtime": 93.6663,
|
| 305 |
+
"eval_samples_per_second": 3.523,
|
| 306 |
+
"eval_steps_per_second": 0.117,
|
| 307 |
+
"eval_wer": 186.52403001756346,
|
| 308 |
+
"step": 44
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.06,
|
| 312 |
+
"learning_rate": 7.640449438202247e-06,
|
| 313 |
+
"loss": 0.9097,
|
| 314 |
+
"step": 45
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 1.07,
|
| 318 |
+
"learning_rate": 7.5280898876404495e-06,
|
| 319 |
+
"loss": 0.8688,
|
| 320 |
+
"step": 46
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 1.08,
|
| 324 |
+
"learning_rate": 7.415730337078652e-06,
|
| 325 |
+
"loss": 0.9019,
|
| 326 |
+
"step": 47
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.09,
|
| 330 |
+
"learning_rate": 7.303370786516854e-06,
|
| 331 |
+
"loss": 0.9135,
|
| 332 |
+
"step": 48
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.1,
|
| 336 |
+
"learning_rate": 7.191011235955056e-06,
|
| 337 |
+
"loss": 0.9033,
|
| 338 |
+
"step": 49
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.11,
|
| 342 |
+
"learning_rate": 7.078651685393258e-06,
|
| 343 |
+
"loss": 0.8575,
|
| 344 |
+
"step": 50
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 1.12,
|
| 348 |
+
"learning_rate": 6.966292134831461e-06,
|
| 349 |
+
"loss": 0.8276,
|
| 350 |
+
"step": 51
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.12,
|
| 354 |
+
"learning_rate": 6.853932584269663e-06,
|
| 355 |
+
"loss": 0.9276,
|
| 356 |
+
"step": 52
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 1.13,
|
| 360 |
+
"learning_rate": 6.741573033707865e-06,
|
| 361 |
+
"loss": 0.9186,
|
| 362 |
+
"step": 53
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.14,
|
| 366 |
+
"learning_rate": 6.629213483146067e-06,
|
| 367 |
+
"loss": 0.8693,
|
| 368 |
+
"step": 54
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.15,
|
| 372 |
+
"learning_rate": 6.51685393258427e-06,
|
| 373 |
+
"loss": 0.8441,
|
| 374 |
+
"step": 55
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 1.15,
|
| 378 |
+
"eval_loss": 1.0945535898208618,
|
| 379 |
+
"eval_runtime": 102.0356,
|
| 380 |
+
"eval_samples_per_second": 3.234,
|
| 381 |
+
"eval_steps_per_second": 0.108,
|
| 382 |
+
"eval_wer": 179.32300814306242,
|
| 383 |
+
"step": 55
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 1.16,
|
| 387 |
+
"learning_rate": 6.404494382022472e-06,
|
| 388 |
+
"loss": 0.8189,
|
| 389 |
+
"step": 56
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.17,
|
| 393 |
+
"learning_rate": 6.292134831460674e-06,
|
| 394 |
+
"loss": 0.8047,
|
| 395 |
+
"step": 57
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.18,
|
| 399 |
+
"learning_rate": 6.179775280898876e-06,
|
| 400 |
+
"loss": 0.7858,
|
| 401 |
+
"step": 58
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.19,
|
| 405 |
+
"learning_rate": 6.06741573033708e-06,
|
| 406 |
+
"loss": 0.7312,
|
| 407 |
+
"step": 59
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 1.2,
|
| 411 |
+
"learning_rate": 5.955056179775281e-06,
|
| 412 |
+
"loss": 0.7638,
|
| 413 |
+
"step": 60
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 1.21,
|
| 417 |
+
"learning_rate": 5.842696629213483e-06,
|
| 418 |
+
"loss": 0.7374,
|
| 419 |
+
"step": 61
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 1.21,
|
| 423 |
+
"learning_rate": 5.730337078651685e-06,
|
| 424 |
+
"loss": 0.7986,
|
| 425 |
+
"step": 62
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 1.22,
|
| 429 |
+
"learning_rate": 5.617977528089889e-06,
|
| 430 |
+
"loss": 0.762,
|
| 431 |
+
"step": 63
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.23,
|
| 435 |
+
"learning_rate": 5.50561797752809e-06,
|
| 436 |
+
"loss": 0.785,
|
| 437 |
+
"step": 64
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.24,
|
| 441 |
+
"learning_rate": 5.393258426966292e-06,
|
| 442 |
+
"loss": 0.8349,
|
| 443 |
+
"step": 65
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.25,
|
| 447 |
+
"learning_rate": 5.280898876404494e-06,
|
| 448 |
+
"loss": 0.8505,
|
| 449 |
+
"step": 66
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 1.25,
|
| 453 |
+
"eval_loss": 1.0748353004455566,
|
| 454 |
+
"eval_runtime": 93.3893,
|
| 455 |
+
"eval_samples_per_second": 3.534,
|
| 456 |
+
"eval_steps_per_second": 0.118,
|
| 457 |
+
"eval_wer": 159.68385757624142,
|
| 458 |
+
"step": 66
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.26,
|
| 462 |
+
"learning_rate": 5.168539325842698e-06,
|
| 463 |
+
"loss": 0.7309,
|
| 464 |
+
"step": 67
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.27,
|
| 468 |
+
"learning_rate": 5.0561797752809e-06,
|
| 469 |
+
"loss": 0.8261,
|
| 470 |
+
"step": 68
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 1.28,
|
| 474 |
+
"learning_rate": 4.943820224719101e-06,
|
| 475 |
+
"loss": 0.8053,
|
| 476 |
+
"step": 69
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 1.29,
|
| 480 |
+
"learning_rate": 4.831460674157304e-06,
|
| 481 |
+
"loss": 0.7672,
|
| 482 |
+
"step": 70
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 1.29,
|
| 486 |
+
"learning_rate": 4.719101123595506e-06,
|
| 487 |
+
"loss": 0.8692,
|
| 488 |
+
"step": 71
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 1.3,
|
| 492 |
+
"learning_rate": 4.606741573033709e-06,
|
| 493 |
+
"loss": 0.8588,
|
| 494 |
+
"step": 72
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.31,
|
| 498 |
+
"learning_rate": 4.494382022471911e-06,
|
| 499 |
+
"loss": 0.8277,
|
| 500 |
+
"step": 73
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.32,
|
| 504 |
+
"learning_rate": 4.382022471910113e-06,
|
| 505 |
+
"loss": 0.8387,
|
| 506 |
+
"step": 74
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.33,
|
| 510 |
+
"learning_rate": 4.269662921348315e-06,
|
| 511 |
+
"loss": 0.801,
|
| 512 |
+
"step": 75
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 1.34,
|
| 516 |
+
"learning_rate": 4.157303370786518e-06,
|
| 517 |
+
"loss": 0.7801,
|
| 518 |
+
"step": 76
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 2.01,
|
| 522 |
+
"learning_rate": 4.04494382022472e-06,
|
| 523 |
+
"loss": 0.7844,
|
| 524 |
+
"step": 77
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 2.01,
|
| 528 |
+
"eval_loss": 1.0585265159606934,
|
| 529 |
+
"eval_runtime": 87.6428,
|
| 530 |
+
"eval_samples_per_second": 3.765,
|
| 531 |
+
"eval_steps_per_second": 0.126,
|
| 532 |
+
"eval_wer": 163.2923519080313,
|
| 533 |
+
"step": 77
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 2.02,
|
| 537 |
+
"learning_rate": 3.932584269662922e-06,
|
| 538 |
+
"loss": 0.8227,
|
| 539 |
+
"step": 78
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 2.03,
|
| 543 |
+
"learning_rate": 3.820224719101124e-06,
|
| 544 |
+
"loss": 0.757,
|
| 545 |
+
"step": 79
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 2.04,
|
| 549 |
+
"learning_rate": 3.707865168539326e-06,
|
| 550 |
+
"loss": 0.7713,
|
| 551 |
+
"step": 80
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 2.04,
|
| 555 |
+
"learning_rate": 3.595505617977528e-06,
|
| 556 |
+
"loss": 0.7782,
|
| 557 |
+
"step": 81
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 2.05,
|
| 561 |
+
"learning_rate": 3.4831460674157306e-06,
|
| 562 |
+
"loss": 0.7468,
|
| 563 |
+
"step": 82
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 2.06,
|
| 567 |
+
"learning_rate": 3.3707865168539327e-06,
|
| 568 |
+
"loss": 0.7457,
|
| 569 |
+
"step": 83
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.07,
|
| 573 |
+
"learning_rate": 3.258426966292135e-06,
|
| 574 |
+
"loss": 0.7126,
|
| 575 |
+
"step": 84
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 2.08,
|
| 579 |
+
"learning_rate": 3.146067415730337e-06,
|
| 580 |
+
"loss": 0.7476,
|
| 581 |
+
"step": 85
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 2.09,
|
| 585 |
+
"learning_rate": 3.03370786516854e-06,
|
| 586 |
+
"loss": 0.76,
|
| 587 |
+
"step": 86
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 2.1,
|
| 591 |
+
"learning_rate": 2.9213483146067416e-06,
|
| 592 |
+
"loss": 0.7673,
|
| 593 |
+
"step": 87
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 2.11,
|
| 597 |
+
"learning_rate": 2.8089887640449444e-06,
|
| 598 |
+
"loss": 0.7208,
|
| 599 |
+
"step": 88
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 2.11,
|
| 603 |
+
"eval_loss": 1.0490810871124268,
|
| 604 |
+
"eval_runtime": 106.8845,
|
| 605 |
+
"eval_samples_per_second": 3.087,
|
| 606 |
+
"eval_steps_per_second": 0.103,
|
| 607 |
+
"eval_wer": 158.1031454574485,
|
| 608 |
+
"step": 88
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 2.12,
|
| 612 |
+
"learning_rate": 2.696629213483146e-06,
|
| 613 |
+
"loss": 0.7045,
|
| 614 |
+
"step": 89
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 2.12,
|
| 618 |
+
"learning_rate": 2.584269662921349e-06,
|
| 619 |
+
"loss": 0.7887,
|
| 620 |
+
"step": 90
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 2.13,
|
| 624 |
+
"learning_rate": 2.4719101123595505e-06,
|
| 625 |
+
"loss": 0.7821,
|
| 626 |
+
"step": 91
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 2.14,
|
| 630 |
+
"learning_rate": 2.359550561797753e-06,
|
| 631 |
+
"loss": 0.7487,
|
| 632 |
+
"step": 92
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.15,
|
| 636 |
+
"learning_rate": 2.2471910112359554e-06,
|
| 637 |
+
"loss": 0.7281,
|
| 638 |
+
"step": 93
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 2.16,
|
| 642 |
+
"learning_rate": 2.1348314606741574e-06,
|
| 643 |
+
"loss": 0.7084,
|
| 644 |
+
"step": 94
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 2.17,
|
| 648 |
+
"learning_rate": 2.02247191011236e-06,
|
| 649 |
+
"loss": 0.6971,
|
| 650 |
+
"step": 95
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 2.18,
|
| 654 |
+
"learning_rate": 1.910112359550562e-06,
|
| 655 |
+
"loss": 0.6869,
|
| 656 |
+
"step": 96
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 2.19,
|
| 660 |
+
"learning_rate": 1.797752808988764e-06,
|
| 661 |
+
"loss": 0.6411,
|
| 662 |
+
"step": 97
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 2.2,
|
| 666 |
+
"learning_rate": 1.6853932584269663e-06,
|
| 667 |
+
"loss": 0.6658,
|
| 668 |
+
"step": 98
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 2.21,
|
| 672 |
+
"learning_rate": 1.5730337078651686e-06,
|
| 673 |
+
"loss": 0.6481,
|
| 674 |
+
"step": 99
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.21,
|
| 678 |
+
"eval_loss": 1.046801209449768,
|
| 679 |
+
"eval_runtime": 89.0792,
|
| 680 |
+
"eval_samples_per_second": 3.705,
|
| 681 |
+
"eval_steps_per_second": 0.123,
|
| 682 |
+
"eval_wer": 158.51828197349514,
|
| 683 |
+
"step": 99
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 2.21,
|
| 687 |
+
"learning_rate": 1.4606741573033708e-06,
|
| 688 |
+
"loss": 0.7064,
|
| 689 |
+
"step": 100
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 2.22,
|
| 693 |
+
"learning_rate": 1.348314606741573e-06,
|
| 694 |
+
"loss": 0.6699,
|
| 695 |
+
"step": 101
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 2.23,
|
| 699 |
+
"learning_rate": 1.2359550561797752e-06,
|
| 700 |
+
"loss": 0.6974,
|
| 701 |
+
"step": 102
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 2.24,
|
| 705 |
+
"learning_rate": 1.1235955056179777e-06,
|
| 706 |
+
"loss": 0.7523,
|
| 707 |
+
"step": 103
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 2.25,
|
| 711 |
+
"learning_rate": 1.01123595505618e-06,
|
| 712 |
+
"loss": 0.7621,
|
| 713 |
+
"step": 104
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 2.26,
|
| 717 |
+
"learning_rate": 8.98876404494382e-07,
|
| 718 |
+
"loss": 0.657,
|
| 719 |
+
"step": 105
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 2.27,
|
| 723 |
+
"learning_rate": 7.865168539325843e-07,
|
| 724 |
+
"loss": 0.7443,
|
| 725 |
+
"step": 106
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 2.28,
|
| 729 |
+
"learning_rate": 6.741573033707865e-07,
|
| 730 |
+
"loss": 0.729,
|
| 731 |
+
"step": 107
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 2.29,
|
| 735 |
+
"learning_rate": 5.617977528089888e-07,
|
| 736 |
+
"loss": 0.6983,
|
| 737 |
+
"step": 108
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 2.29,
|
| 741 |
+
"learning_rate": 4.49438202247191e-07,
|
| 742 |
+
"loss": 0.7963,
|
| 743 |
+
"step": 109
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 2.3,
|
| 747 |
+
"learning_rate": 3.3707865168539325e-07,
|
| 748 |
+
"loss": 0.7912,
|
| 749 |
+
"step": 110
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 2.3,
|
| 753 |
+
"eval_loss": 1.045613408088684,
|
| 754 |
+
"eval_runtime": 91.8976,
|
| 755 |
+
"eval_samples_per_second": 3.591,
|
| 756 |
+
"eval_steps_per_second": 0.12,
|
| 757 |
+
"eval_wer": 168.6092926712438,
|
| 758 |
+
"step": 110
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 2.31,
|
| 762 |
+
"learning_rate": 2.247191011235955e-07,
|
| 763 |
+
"loss": 0.7626,
|
| 764 |
+
"step": 111
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 2.32,
|
| 768 |
+
"learning_rate": 1.1235955056179776e-07,
|
| 769 |
+
"loss": 0.7754,
|
| 770 |
+
"step": 112
|
| 771 |
},
|
| 772 |
{
|
| 773 |
+
"epoch": 2.32,
|
| 774 |
+
"step": 112,
|
| 775 |
+
"total_flos": 1.7415399333888e+17,
|
| 776 |
+
"train_loss": 0.9721650715385165,
|
| 777 |
+
"train_runtime": 1491.7987,
|
| 778 |
+
"train_samples_per_second": 4.805,
|
| 779 |
+
"train_steps_per_second": 0.075
|
| 780 |
}
|
| 781 |
],
|
| 782 |
+
"max_steps": 112,
|
| 783 |
"num_train_epochs": 9223372036854775807,
|
| 784 |
+
"total_flos": 1.7415399333888e+17,
|
| 785 |
"trial_name": null,
|
| 786 |
"trial_params": null
|
| 787 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3579
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:23666f2916cd5411d616f8c83fed6cb219a0ba72c70f8943bef9ed5bb45800ed
|
| 3 |
size 3579
|