End of training
Browse files- README.md +20 -5
- all_results.json +16 -0
- eval_results.json +10 -0
- train_results.json +9 -0
- trainer_state.json +830 -0
README.md
CHANGED
|
@@ -1,13 +1,28 @@
|
|
| 1 |
---
|
|
|
|
|
|
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: google/mt5-small
|
| 4 |
tags:
|
| 5 |
- generated_from_trainer
|
|
|
|
|
|
|
| 6 |
metrics:
|
| 7 |
- bleu
|
| 8 |
model-index:
|
| 9 |
- name: ft-wmt14-5
|
| 10 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
---
|
| 12 |
|
| 13 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -15,11 +30,11 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# ft-wmt14-5
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [google/mt5-small](https://huggingface.co/google/mt5-small) on
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
-
- Loss: 2.
|
| 21 |
-
- Bleu: 20.
|
| 22 |
-
- Gen Len: 30.
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
|
|
|
| 1 |
---
|
| 2 |
+
language:
|
| 3 |
+
- de
|
| 4 |
+
- en
|
| 5 |
license: apache-2.0
|
| 6 |
base_model: google/mt5-small
|
| 7 |
tags:
|
| 8 |
- generated_from_trainer
|
| 9 |
+
datasets:
|
| 10 |
+
- lilferrit/wmt14-short
|
| 11 |
metrics:
|
| 12 |
- bleu
|
| 13 |
model-index:
|
| 14 |
- name: ft-wmt14-5
|
| 15 |
+
results:
|
| 16 |
+
- task:
|
| 17 |
+
name: Translation
|
| 18 |
+
type: translation
|
| 19 |
+
dataset:
|
| 20 |
+
name: lilferrit/wmt14-short
|
| 21 |
+
type: lilferrit/wmt14-short
|
| 22 |
+
metrics:
|
| 23 |
+
- name: Bleu
|
| 24 |
+
type: bleu
|
| 25 |
+
value: 20.7584
|
| 26 |
---
|
| 27 |
|
| 28 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 30 |
|
| 31 |
# ft-wmt14-5
|
| 32 |
|
| 33 |
+
This model is a fine-tuned version of [google/mt5-small](https://huggingface.co/google/mt5-small) on the lilferrit/wmt14-short dataset.
|
| 34 |
It achieves the following results on the evaluation set:
|
| 35 |
+
- Loss: 2.0604
|
| 36 |
+
- Bleu: 20.7584
|
| 37 |
+
- Gen Len: 30.499
|
| 38 |
|
| 39 |
## Model description
|
| 40 |
|
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.7777777777777777,
|
| 3 |
+
"eval_bleu": 20.7584,
|
| 4 |
+
"eval_gen_len": 30.499,
|
| 5 |
+
"eval_loss": 2.0603742599487305,
|
| 6 |
+
"eval_runtime": 371.1712,
|
| 7 |
+
"eval_samples": 3000,
|
| 8 |
+
"eval_samples_per_second": 8.083,
|
| 9 |
+
"eval_steps_per_second": 1.01,
|
| 10 |
+
"total_flos": 1.4240580791795712e+17,
|
| 11 |
+
"train_loss": 0.5475473999023438,
|
| 12 |
+
"train_runtime": 14821.2356,
|
| 13 |
+
"train_samples": 576000,
|
| 14 |
+
"train_samples_per_second": 107.953,
|
| 15 |
+
"train_steps_per_second": 6.747
|
| 16 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.7777777777777777,
|
| 3 |
+
"eval_bleu": 20.7584,
|
| 4 |
+
"eval_gen_len": 30.499,
|
| 5 |
+
"eval_loss": 2.0603742599487305,
|
| 6 |
+
"eval_runtime": 371.1712,
|
| 7 |
+
"eval_samples": 3000,
|
| 8 |
+
"eval_samples_per_second": 8.083,
|
| 9 |
+
"eval_steps_per_second": 1.01
|
| 10 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.7777777777777777,
|
| 3 |
+
"total_flos": 1.4240580791795712e+17,
|
| 4 |
+
"train_loss": 0.5475473999023438,
|
| 5 |
+
"train_runtime": 14821.2356,
|
| 6 |
+
"train_samples": 576000,
|
| 7 |
+
"train_samples_per_second": 107.953,
|
| 8 |
+
"train_steps_per_second": 6.747
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,830 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 20.7584,
|
| 3 |
+
"best_model_checkpoint": "/local1/hfs/gs_stuff/ft-wmt14-5/checkpoint-90000",
|
| 4 |
+
"epoch": 2.7777777777777777,
|
| 5 |
+
"eval_steps": 10000,
|
| 6 |
+
"global_step": 100000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.027777777777777776,
|
| 13 |
+
"grad_norm": 1.9314790964126587,
|
| 14 |
+
"learning_rate": 0.0005,
|
| 15 |
+
"loss": 3.3589,
|
| 16 |
+
"step": 1000
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.05555555555555555,
|
| 20 |
+
"grad_norm": 1.7348469495773315,
|
| 21 |
+
"learning_rate": 0.0005,
|
| 22 |
+
"loss": 2.5263,
|
| 23 |
+
"step": 2000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.08333333333333333,
|
| 27 |
+
"grad_norm": 1.9181748628616333,
|
| 28 |
+
"learning_rate": 0.0005,
|
| 29 |
+
"loss": 2.3365,
|
| 30 |
+
"step": 3000
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.1111111111111111,
|
| 34 |
+
"grad_norm": 1.6642646789550781,
|
| 35 |
+
"learning_rate": 0.0005,
|
| 36 |
+
"loss": 2.2207,
|
| 37 |
+
"step": 4000
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.1388888888888889,
|
| 41 |
+
"grad_norm": 1.1876742839813232,
|
| 42 |
+
"learning_rate": 0.0005,
|
| 43 |
+
"loss": 2.1363,
|
| 44 |
+
"step": 5000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.16666666666666666,
|
| 48 |
+
"grad_norm": 1.567658543586731,
|
| 49 |
+
"learning_rate": 0.0005,
|
| 50 |
+
"loss": 2.0733,
|
| 51 |
+
"step": 6000
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.19444444444444445,
|
| 55 |
+
"grad_norm": 1.2552471160888672,
|
| 56 |
+
"learning_rate": 0.0005,
|
| 57 |
+
"loss": 2.0262,
|
| 58 |
+
"step": 7000
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.2222222222222222,
|
| 62 |
+
"grad_norm": 1.049357533454895,
|
| 63 |
+
"learning_rate": 0.0005,
|
| 64 |
+
"loss": 1.9775,
|
| 65 |
+
"step": 8000
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.25,
|
| 69 |
+
"grad_norm": 1.303145170211792,
|
| 70 |
+
"learning_rate": 0.0005,
|
| 71 |
+
"loss": 1.9412,
|
| 72 |
+
"step": 9000
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.2777777777777778,
|
| 76 |
+
"grad_norm": 1.0213723182678223,
|
| 77 |
+
"learning_rate": 0.0005,
|
| 78 |
+
"loss": 1.9166,
|
| 79 |
+
"step": 10000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.2777777777777778,
|
| 83 |
+
"eval_bleu": 15.8119,
|
| 84 |
+
"eval_gen_len": 32.097,
|
| 85 |
+
"eval_loss": 2.31050968170166,
|
| 86 |
+
"eval_runtime": 410.6001,
|
| 87 |
+
"eval_samples_per_second": 7.306,
|
| 88 |
+
"eval_steps_per_second": 0.913,
|
| 89 |
+
"step": 10000
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.3055555555555556,
|
| 93 |
+
"grad_norm": 1.2851905822753906,
|
| 94 |
+
"learning_rate": 0.0005,
|
| 95 |
+
"loss": 1.8878,
|
| 96 |
+
"step": 11000
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 0.3333333333333333,
|
| 100 |
+
"grad_norm": 0.8447160720825195,
|
| 101 |
+
"learning_rate": 0.0005,
|
| 102 |
+
"loss": 1.8492,
|
| 103 |
+
"step": 12000
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"epoch": 0.3611111111111111,
|
| 107 |
+
"grad_norm": 1.1516064405441284,
|
| 108 |
+
"learning_rate": 0.0005,
|
| 109 |
+
"loss": 1.8309,
|
| 110 |
+
"step": 13000
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.3888888888888889,
|
| 114 |
+
"grad_norm": 1.0370670557022095,
|
| 115 |
+
"learning_rate": 0.0005,
|
| 116 |
+
"loss": 1.8057,
|
| 117 |
+
"step": 14000
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.4166666666666667,
|
| 121 |
+
"grad_norm": 1.1649495363235474,
|
| 122 |
+
"learning_rate": 0.0005,
|
| 123 |
+
"loss": 1.7867,
|
| 124 |
+
"step": 15000
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 0.4444444444444444,
|
| 128 |
+
"grad_norm": 1.2666045427322388,
|
| 129 |
+
"learning_rate": 0.0005,
|
| 130 |
+
"loss": 1.7679,
|
| 131 |
+
"step": 16000
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.4722222222222222,
|
| 135 |
+
"grad_norm": 1.0923264026641846,
|
| 136 |
+
"learning_rate": 0.0005,
|
| 137 |
+
"loss": 1.7563,
|
| 138 |
+
"step": 17000
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 0.5,
|
| 142 |
+
"grad_norm": 1.560994029045105,
|
| 143 |
+
"learning_rate": 0.0005,
|
| 144 |
+
"loss": 1.7342,
|
| 145 |
+
"step": 18000
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.5277777777777778,
|
| 149 |
+
"grad_norm": 0.9684827327728271,
|
| 150 |
+
"learning_rate": 0.0005,
|
| 151 |
+
"loss": 1.7228,
|
| 152 |
+
"step": 19000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.5555555555555556,
|
| 156 |
+
"grad_norm": 0.9182453751564026,
|
| 157 |
+
"learning_rate": 0.0005,
|
| 158 |
+
"loss": 1.7184,
|
| 159 |
+
"step": 20000
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.5555555555555556,
|
| 163 |
+
"eval_bleu": 17.5903,
|
| 164 |
+
"eval_gen_len": 31.1153,
|
| 165 |
+
"eval_loss": 2.19934344291687,
|
| 166 |
+
"eval_runtime": 393.3017,
|
| 167 |
+
"eval_samples_per_second": 7.628,
|
| 168 |
+
"eval_steps_per_second": 0.953,
|
| 169 |
+
"step": 20000
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.5833333333333334,
|
| 173 |
+
"grad_norm": 0.8953577280044556,
|
| 174 |
+
"learning_rate": 0.0005,
|
| 175 |
+
"loss": 1.7042,
|
| 176 |
+
"step": 21000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.6111111111111112,
|
| 180 |
+
"grad_norm": 0.9418250918388367,
|
| 181 |
+
"learning_rate": 0.0005,
|
| 182 |
+
"loss": 1.683,
|
| 183 |
+
"step": 22000
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.6388888888888888,
|
| 187 |
+
"grad_norm": 0.8577601909637451,
|
| 188 |
+
"learning_rate": 0.0005,
|
| 189 |
+
"loss": 1.6799,
|
| 190 |
+
"step": 23000
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.6666666666666666,
|
| 194 |
+
"grad_norm": 0.9786076545715332,
|
| 195 |
+
"learning_rate": 0.0005,
|
| 196 |
+
"loss": 1.6675,
|
| 197 |
+
"step": 24000
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.6944444444444444,
|
| 201 |
+
"grad_norm": 0.9262654781341553,
|
| 202 |
+
"learning_rate": 0.0005,
|
| 203 |
+
"loss": 1.6499,
|
| 204 |
+
"step": 25000
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.7222222222222222,
|
| 208 |
+
"grad_norm": 0.8759564757347107,
|
| 209 |
+
"learning_rate": 0.0005,
|
| 210 |
+
"loss": 1.6468,
|
| 211 |
+
"step": 26000
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 0.75,
|
| 215 |
+
"grad_norm": 1.0495752096176147,
|
| 216 |
+
"learning_rate": 0.0005,
|
| 217 |
+
"loss": 1.6285,
|
| 218 |
+
"step": 27000
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.7777777777777778,
|
| 222 |
+
"grad_norm": 1.092642068862915,
|
| 223 |
+
"learning_rate": 0.0005,
|
| 224 |
+
"loss": 1.6276,
|
| 225 |
+
"step": 28000
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.8055555555555556,
|
| 229 |
+
"grad_norm": 0.8775661587715149,
|
| 230 |
+
"learning_rate": 0.0005,
|
| 231 |
+
"loss": 1.6172,
|
| 232 |
+
"step": 29000
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 0.8333333333333334,
|
| 236 |
+
"grad_norm": 0.8970679044723511,
|
| 237 |
+
"learning_rate": 0.0005,
|
| 238 |
+
"loss": 1.6061,
|
| 239 |
+
"step": 30000
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.8333333333333334,
|
| 243 |
+
"eval_bleu": 18.9604,
|
| 244 |
+
"eval_gen_len": 30.327,
|
| 245 |
+
"eval_loss": 2.1379551887512207,
|
| 246 |
+
"eval_runtime": 380.095,
|
| 247 |
+
"eval_samples_per_second": 7.893,
|
| 248 |
+
"eval_steps_per_second": 0.987,
|
| 249 |
+
"step": 30000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.8611111111111112,
|
| 253 |
+
"grad_norm": 0.9657310247421265,
|
| 254 |
+
"learning_rate": 0.0005,
|
| 255 |
+
"loss": 1.5959,
|
| 256 |
+
"step": 31000
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.8888888888888888,
|
| 260 |
+
"grad_norm": 0.8748376369476318,
|
| 261 |
+
"learning_rate": 0.0005,
|
| 262 |
+
"loss": 1.5908,
|
| 263 |
+
"step": 32000
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 0.9166666666666666,
|
| 267 |
+
"grad_norm": 0.8462302088737488,
|
| 268 |
+
"learning_rate": 0.0005,
|
| 269 |
+
"loss": 1.5845,
|
| 270 |
+
"step": 33000
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 0.9444444444444444,
|
| 274 |
+
"grad_norm": 0.9005241394042969,
|
| 275 |
+
"learning_rate": 0.0005,
|
| 276 |
+
"loss": 1.5699,
|
| 277 |
+
"step": 34000
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 0.9722222222222222,
|
| 281 |
+
"grad_norm": 0.9596630930900574,
|
| 282 |
+
"learning_rate": 0.0005,
|
| 283 |
+
"loss": 1.5752,
|
| 284 |
+
"step": 35000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.0,
|
| 288 |
+
"grad_norm": 0.8307533860206604,
|
| 289 |
+
"learning_rate": 0.0005,
|
| 290 |
+
"loss": 1.5634,
|
| 291 |
+
"step": 36000
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 1.0277777777777777,
|
| 295 |
+
"grad_norm": 0.9918788075447083,
|
| 296 |
+
"learning_rate": 0.0005,
|
| 297 |
+
"loss": 1.5117,
|
| 298 |
+
"step": 37000
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 1.0555555555555556,
|
| 302 |
+
"grad_norm": 0.9118058085441589,
|
| 303 |
+
"learning_rate": 0.0005,
|
| 304 |
+
"loss": 1.5023,
|
| 305 |
+
"step": 38000
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.0833333333333333,
|
| 309 |
+
"grad_norm": 0.7213552594184875,
|
| 310 |
+
"learning_rate": 0.0005,
|
| 311 |
+
"loss": 1.5087,
|
| 312 |
+
"step": 39000
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 1.1111111111111112,
|
| 316 |
+
"grad_norm": 1.0255305767059326,
|
| 317 |
+
"learning_rate": 0.0005,
|
| 318 |
+
"loss": 1.516,
|
| 319 |
+
"step": 40000
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 1.1111111111111112,
|
| 323 |
+
"eval_bleu": 19.1444,
|
| 324 |
+
"eval_gen_len": 30.2727,
|
| 325 |
+
"eval_loss": 2.1365692615509033,
|
| 326 |
+
"eval_runtime": 377.1737,
|
| 327 |
+
"eval_samples_per_second": 7.954,
|
| 328 |
+
"eval_steps_per_second": 0.994,
|
| 329 |
+
"step": 40000
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 1.1388888888888888,
|
| 333 |
+
"grad_norm": 0.8766499161720276,
|
| 334 |
+
"learning_rate": 0.0005,
|
| 335 |
+
"loss": 1.5096,
|
| 336 |
+
"step": 41000
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 1.1666666666666667,
|
| 340 |
+
"grad_norm": 1.1786612272262573,
|
| 341 |
+
"learning_rate": 0.0005,
|
| 342 |
+
"loss": 1.4982,
|
| 343 |
+
"step": 42000
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 1.1944444444444444,
|
| 347 |
+
"grad_norm": 1.011268973350525,
|
| 348 |
+
"learning_rate": 0.0005,
|
| 349 |
+
"loss": 1.5013,
|
| 350 |
+
"step": 43000
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.2222222222222223,
|
| 354 |
+
"grad_norm": 1.0863969326019287,
|
| 355 |
+
"learning_rate": 0.0005,
|
| 356 |
+
"loss": 1.4878,
|
| 357 |
+
"step": 44000
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 1.25,
|
| 361 |
+
"grad_norm": 0.9729832410812378,
|
| 362 |
+
"learning_rate": 0.0005,
|
| 363 |
+
"loss": 1.4922,
|
| 364 |
+
"step": 45000
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 1.2777777777777777,
|
| 368 |
+
"grad_norm": 1.3476896286010742,
|
| 369 |
+
"learning_rate": 0.0005,
|
| 370 |
+
"loss": 1.4876,
|
| 371 |
+
"step": 46000
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 1.3055555555555556,
|
| 375 |
+
"grad_norm": 0.8493963479995728,
|
| 376 |
+
"learning_rate": 0.0005,
|
| 377 |
+
"loss": 1.4823,
|
| 378 |
+
"step": 47000
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 1.3333333333333333,
|
| 382 |
+
"grad_norm": 1.0311123132705688,
|
| 383 |
+
"learning_rate": 0.0005,
|
| 384 |
+
"loss": 1.4739,
|
| 385 |
+
"step": 48000
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 1.3611111111111112,
|
| 389 |
+
"grad_norm": 1.259581446647644,
|
| 390 |
+
"learning_rate": 0.0005,
|
| 391 |
+
"loss": 1.4747,
|
| 392 |
+
"step": 49000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.3888888888888888,
|
| 396 |
+
"grad_norm": 1.1934195756912231,
|
| 397 |
+
"learning_rate": 0.0005,
|
| 398 |
+
"loss": 1.4675,
|
| 399 |
+
"step": 50000
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.3888888888888888,
|
| 403 |
+
"eval_bleu": 19.7588,
|
| 404 |
+
"eval_gen_len": 30.1127,
|
| 405 |
+
"eval_loss": 2.120835781097412,
|
| 406 |
+
"eval_runtime": 372.4281,
|
| 407 |
+
"eval_samples_per_second": 8.055,
|
| 408 |
+
"eval_steps_per_second": 1.007,
|
| 409 |
+
"step": 50000
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.4166666666666667,
|
| 413 |
+
"grad_norm": 1.1824595928192139,
|
| 414 |
+
"learning_rate": 0.0005,
|
| 415 |
+
"loss": 1.4659,
|
| 416 |
+
"step": 51000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.4444444444444444,
|
| 420 |
+
"grad_norm": 1.1661032438278198,
|
| 421 |
+
"learning_rate": 0.0005,
|
| 422 |
+
"loss": 1.4737,
|
| 423 |
+
"step": 52000
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.4722222222222223,
|
| 427 |
+
"grad_norm": 0.7856634259223938,
|
| 428 |
+
"learning_rate": 0.0005,
|
| 429 |
+
"loss": 1.4595,
|
| 430 |
+
"step": 53000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.5,
|
| 434 |
+
"grad_norm": 0.9908609986305237,
|
| 435 |
+
"learning_rate": 0.0005,
|
| 436 |
+
"loss": 1.4656,
|
| 437 |
+
"step": 54000
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.5277777777777777,
|
| 441 |
+
"grad_norm": 0.9270644187927246,
|
| 442 |
+
"learning_rate": 0.0005,
|
| 443 |
+
"loss": 1.4524,
|
| 444 |
+
"step": 55000
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.5555555555555556,
|
| 448 |
+
"grad_norm": 0.9910904169082642,
|
| 449 |
+
"learning_rate": 0.0005,
|
| 450 |
+
"loss": 1.4453,
|
| 451 |
+
"step": 56000
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.5833333333333335,
|
| 455 |
+
"grad_norm": 1.0300639867782593,
|
| 456 |
+
"learning_rate": 0.0005,
|
| 457 |
+
"loss": 1.451,
|
| 458 |
+
"step": 57000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.6111111111111112,
|
| 462 |
+
"grad_norm": 0.809105396270752,
|
| 463 |
+
"learning_rate": 0.0005,
|
| 464 |
+
"loss": 1.444,
|
| 465 |
+
"step": 58000
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.6388888888888888,
|
| 469 |
+
"grad_norm": 0.7915866374969482,
|
| 470 |
+
"learning_rate": 0.0005,
|
| 471 |
+
"loss": 1.4421,
|
| 472 |
+
"step": 59000
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.6666666666666665,
|
| 476 |
+
"grad_norm": 0.9778928756713867,
|
| 477 |
+
"learning_rate": 0.0005,
|
| 478 |
+
"loss": 1.4416,
|
| 479 |
+
"step": 60000
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.6666666666666665,
|
| 483 |
+
"eval_bleu": 19.9263,
|
| 484 |
+
"eval_gen_len": 30.4463,
|
| 485 |
+
"eval_loss": 2.088862657546997,
|
| 486 |
+
"eval_runtime": 383.2772,
|
| 487 |
+
"eval_samples_per_second": 7.827,
|
| 488 |
+
"eval_steps_per_second": 0.978,
|
| 489 |
+
"step": 60000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 1.6944444444444444,
|
| 493 |
+
"grad_norm": 0.8484209775924683,
|
| 494 |
+
"learning_rate": 0.0005,
|
| 495 |
+
"loss": 1.4313,
|
| 496 |
+
"step": 61000
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 1.7222222222222223,
|
| 500 |
+
"grad_norm": 0.8703031539916992,
|
| 501 |
+
"learning_rate": 0.0005,
|
| 502 |
+
"loss": 1.4405,
|
| 503 |
+
"step": 62000
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 1.75,
|
| 507 |
+
"grad_norm": 1.4096006155014038,
|
| 508 |
+
"learning_rate": 0.0005,
|
| 509 |
+
"loss": 1.4375,
|
| 510 |
+
"step": 63000
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 1.7777777777777777,
|
| 514 |
+
"grad_norm": 0.9177774786949158,
|
| 515 |
+
"learning_rate": 0.0005,
|
| 516 |
+
"loss": 1.4262,
|
| 517 |
+
"step": 64000
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 1.8055555555555556,
|
| 521 |
+
"grad_norm": 1.2332441806793213,
|
| 522 |
+
"learning_rate": 0.0005,
|
| 523 |
+
"loss": 1.4233,
|
| 524 |
+
"step": 65000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 1.8333333333333335,
|
| 528 |
+
"grad_norm": 0.8750177621841431,
|
| 529 |
+
"learning_rate": 0.0005,
|
| 530 |
+
"loss": 1.4287,
|
| 531 |
+
"step": 66000
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 1.8611111111111112,
|
| 535 |
+
"grad_norm": 0.6736052632331848,
|
| 536 |
+
"learning_rate": 0.0005,
|
| 537 |
+
"loss": 1.4231,
|
| 538 |
+
"step": 67000
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 1.8888888888888888,
|
| 542 |
+
"grad_norm": 0.7802408933639526,
|
| 543 |
+
"learning_rate": 0.0005,
|
| 544 |
+
"loss": 1.4106,
|
| 545 |
+
"step": 68000
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 1.9166666666666665,
|
| 549 |
+
"grad_norm": 1.1860034465789795,
|
| 550 |
+
"learning_rate": 0.0005,
|
| 551 |
+
"loss": 1.4121,
|
| 552 |
+
"step": 69000
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 1.9444444444444444,
|
| 556 |
+
"grad_norm": 0.926054835319519,
|
| 557 |
+
"learning_rate": 0.0005,
|
| 558 |
+
"loss": 1.4111,
|
| 559 |
+
"step": 70000
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 1.9444444444444444,
|
| 563 |
+
"eval_bleu": 20.3323,
|
| 564 |
+
"eval_gen_len": 30.1207,
|
| 565 |
+
"eval_loss": 2.079472541809082,
|
| 566 |
+
"eval_runtime": 371.9755,
|
| 567 |
+
"eval_samples_per_second": 8.065,
|
| 568 |
+
"eval_steps_per_second": 1.008,
|
| 569 |
+
"step": 70000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.9722222222222223,
|
| 573 |
+
"grad_norm": 1.1691533327102661,
|
| 574 |
+
"learning_rate": 0.0005,
|
| 575 |
+
"loss": 1.407,
|
| 576 |
+
"step": 71000
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.0,
|
| 580 |
+
"grad_norm": 0.9077666997909546,
|
| 581 |
+
"learning_rate": 0.0005,
|
| 582 |
+
"loss": 1.4051,
|
| 583 |
+
"step": 72000
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.0277777777777777,
|
| 587 |
+
"grad_norm": 0.9149623513221741,
|
| 588 |
+
"learning_rate": 0.0005,
|
| 589 |
+
"loss": 1.3517,
|
| 590 |
+
"step": 73000
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.0555555555555554,
|
| 594 |
+
"grad_norm": 1.0772947072982788,
|
| 595 |
+
"learning_rate": 0.0005,
|
| 596 |
+
"loss": 1.3624,
|
| 597 |
+
"step": 74000
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.0833333333333335,
|
| 601 |
+
"grad_norm": 0.7283540964126587,
|
| 602 |
+
"learning_rate": 0.0005,
|
| 603 |
+
"loss": 1.355,
|
| 604 |
+
"step": 75000
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.111111111111111,
|
| 608 |
+
"grad_norm": 0.7279065847396851,
|
| 609 |
+
"learning_rate": 0.0005,
|
| 610 |
+
"loss": 1.3526,
|
| 611 |
+
"step": 76000
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.138888888888889,
|
| 615 |
+
"grad_norm": 1.2707905769348145,
|
| 616 |
+
"learning_rate": 0.0005,
|
| 617 |
+
"loss": 1.3535,
|
| 618 |
+
"step": 77000
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.1666666666666665,
|
| 622 |
+
"grad_norm": 0.9000493288040161,
|
| 623 |
+
"learning_rate": 0.0005,
|
| 624 |
+
"loss": 1.3519,
|
| 625 |
+
"step": 78000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.1944444444444446,
|
| 629 |
+
"grad_norm": 1.043967843055725,
|
| 630 |
+
"learning_rate": 0.0005,
|
| 631 |
+
"loss": 1.3567,
|
| 632 |
+
"step": 79000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.2222222222222223,
|
| 636 |
+
"grad_norm": 1.1248853206634521,
|
| 637 |
+
"learning_rate": 0.0005,
|
| 638 |
+
"loss": 1.3603,
|
| 639 |
+
"step": 80000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.2222222222222223,
|
| 643 |
+
"eval_bleu": 20.5373,
|
| 644 |
+
"eval_gen_len": 30.5943,
|
| 645 |
+
"eval_loss": 2.085047960281372,
|
| 646 |
+
"eval_runtime": 373.0705,
|
| 647 |
+
"eval_samples_per_second": 8.041,
|
| 648 |
+
"eval_steps_per_second": 1.005,
|
| 649 |
+
"step": 80000
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 2.25,
|
| 653 |
+
"grad_norm": 1.056221842765808,
|
| 654 |
+
"learning_rate": 0.0005,
|
| 655 |
+
"loss": 1.3657,
|
| 656 |
+
"step": 81000
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 2.2777777777777777,
|
| 660 |
+
"grad_norm": 0.9176587462425232,
|
| 661 |
+
"learning_rate": 0.0005,
|
| 662 |
+
"loss": 1.3572,
|
| 663 |
+
"step": 82000
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 2.3055555555555554,
|
| 667 |
+
"grad_norm": 1.0105085372924805,
|
| 668 |
+
"learning_rate": 0.0005,
|
| 669 |
+
"loss": 1.3498,
|
| 670 |
+
"step": 83000
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 2.3333333333333335,
|
| 674 |
+
"grad_norm": 1.1589380502700806,
|
| 675 |
+
"learning_rate": 0.0005,
|
| 676 |
+
"loss": 1.3567,
|
| 677 |
+
"step": 84000
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 2.361111111111111,
|
| 681 |
+
"grad_norm": 0.7733587622642517,
|
| 682 |
+
"learning_rate": 0.0005,
|
| 683 |
+
"loss": 1.3533,
|
| 684 |
+
"step": 85000
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 2.388888888888889,
|
| 688 |
+
"grad_norm": 1.036777138710022,
|
| 689 |
+
"learning_rate": 0.0005,
|
| 690 |
+
"loss": 1.3469,
|
| 691 |
+
"step": 86000
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 2.4166666666666665,
|
| 695 |
+
"grad_norm": 1.4935026168823242,
|
| 696 |
+
"learning_rate": 0.0005,
|
| 697 |
+
"loss": 1.3469,
|
| 698 |
+
"step": 87000
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 2.4444444444444446,
|
| 702 |
+
"grad_norm": 0.864630937576294,
|
| 703 |
+
"learning_rate": 0.0005,
|
| 704 |
+
"loss": 1.3506,
|
| 705 |
+
"step": 88000
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 2.4722222222222223,
|
| 709 |
+
"grad_norm": 0.8495751619338989,
|
| 710 |
+
"learning_rate": 0.0005,
|
| 711 |
+
"loss": 1.3408,
|
| 712 |
+
"step": 89000
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 2.5,
|
| 716 |
+
"grad_norm": 1.0840762853622437,
|
| 717 |
+
"learning_rate": 0.0005,
|
| 718 |
+
"loss": 1.3378,
|
| 719 |
+
"step": 90000
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 2.5,
|
| 723 |
+
"eval_bleu": 20.7584,
|
| 724 |
+
"eval_gen_len": 30.499,
|
| 725 |
+
"eval_loss": 2.0603742599487305,
|
| 726 |
+
"eval_runtime": 368.0992,
|
| 727 |
+
"eval_samples_per_second": 8.15,
|
| 728 |
+
"eval_steps_per_second": 1.019,
|
| 729 |
+
"step": 90000
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 2.5277777777777777,
|
| 733 |
+
"grad_norm": 0.7769622802734375,
|
| 734 |
+
"learning_rate": 0.0005,
|
| 735 |
+
"loss": 1.3409,
|
| 736 |
+
"step": 91000
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 2.5555555555555554,
|
| 740 |
+
"grad_norm": 1.049972414970398,
|
| 741 |
+
"learning_rate": 0.0005,
|
| 742 |
+
"loss": 1.3443,
|
| 743 |
+
"step": 92000
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 2.5833333333333335,
|
| 747 |
+
"grad_norm": 0.965621292591095,
|
| 748 |
+
"learning_rate": 0.0005,
|
| 749 |
+
"loss": 1.342,
|
| 750 |
+
"step": 93000
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 2.611111111111111,
|
| 754 |
+
"grad_norm": 0.8234182000160217,
|
| 755 |
+
"learning_rate": 0.0005,
|
| 756 |
+
"loss": 1.3297,
|
| 757 |
+
"step": 94000
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 2.638888888888889,
|
| 761 |
+
"grad_norm": 0.9464855790138245,
|
| 762 |
+
"learning_rate": 0.0005,
|
| 763 |
+
"loss": 1.3345,
|
| 764 |
+
"step": 95000
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 2.6666666666666665,
|
| 768 |
+
"grad_norm": 0.987382709980011,
|
| 769 |
+
"learning_rate": 0.0005,
|
| 770 |
+
"loss": 1.3284,
|
| 771 |
+
"step": 96000
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 2.6944444444444446,
|
| 775 |
+
"grad_norm": 0.6439863443374634,
|
| 776 |
+
"learning_rate": 0.0005,
|
| 777 |
+
"loss": 1.3285,
|
| 778 |
+
"step": 97000
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 2.7222222222222223,
|
| 782 |
+
"grad_norm": 0.8853390216827393,
|
| 783 |
+
"learning_rate": 0.0005,
|
| 784 |
+
"loss": 1.3339,
|
| 785 |
+
"step": 98000
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 2.75,
|
| 789 |
+
"grad_norm": 0.7582658529281616,
|
| 790 |
+
"learning_rate": 0.0005,
|
| 791 |
+
"loss": 1.3281,
|
| 792 |
+
"step": 99000
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 2.7777777777777777,
|
| 796 |
+
"grad_norm": 0.9061763882637024,
|
| 797 |
+
"learning_rate": 0.0005,
|
| 798 |
+
"loss": 1.3381,
|
| 799 |
+
"step": 100000
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 2.7777777777777777,
|
| 803 |
+
"eval_bleu": 20.6113,
|
| 804 |
+
"eval_gen_len": 30.701,
|
| 805 |
+
"eval_loss": 2.059664726257324,
|
| 806 |
+
"eval_runtime": 371.2241,
|
| 807 |
+
"eval_samples_per_second": 8.081,
|
| 808 |
+
"eval_steps_per_second": 1.01,
|
| 809 |
+
"step": 100000
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 2.7777777777777777,
|
| 813 |
+
"step": 100000,
|
| 814 |
+
"total_flos": 1.4240580791795712e+17,
|
| 815 |
+
"train_loss": 0.5475473999023438,
|
| 816 |
+
"train_runtime": 14821.2356,
|
| 817 |
+
"train_samples_per_second": 107.953,
|
| 818 |
+
"train_steps_per_second": 6.747
|
| 819 |
+
}
|
| 820 |
+
],
|
| 821 |
+
"logging_steps": 1000,
|
| 822 |
+
"max_steps": 100000,
|
| 823 |
+
"num_input_tokens_seen": 0,
|
| 824 |
+
"num_train_epochs": 3,
|
| 825 |
+
"save_steps": 10000,
|
| 826 |
+
"total_flos": 1.4240580791795712e+17,
|
| 827 |
+
"train_batch_size": 8,
|
| 828 |
+
"trial_name": null,
|
| 829 |
+
"trial_params": null
|
| 830 |
+
}
|