Commit
•
16cbde4
1
Parent(s):
9c4b54c
End of training
Browse files- README.md +20 -4
- all_results.json +22 -0
- eval_results.json +10 -0
- generated_predictions.txt +0 -0
- predict_results.json +9 -0
- train_results.json +8 -0
- trainer_state.json +614 -0
README.md
CHANGED
|
@@ -1,11 +1,27 @@
|
|
| 1 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
tags:
|
| 3 |
- generated_from_trainer
|
|
|
|
|
|
|
| 4 |
metrics:
|
| 5 |
- bleu
|
| 6 |
model-index:
|
| 7 |
- name: tst-gun-gub-pt
|
| 8 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
---
|
| 10 |
|
| 11 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -13,11 +29,11 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 13 |
|
| 14 |
# tst-gun-gub-pt
|
| 15 |
|
| 16 |
-
This model
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
- Loss: 1.8158
|
| 19 |
-
- Bleu: 8.
|
| 20 |
-
- Gen Len:
|
| 21 |
|
| 22 |
## Model description
|
| 23 |
|
|
|
|
| 1 |
---
|
| 2 |
+
language:
|
| 3 |
+
- multilingual
|
| 4 |
+
- pt
|
| 5 |
+
base_model: /content/opus-mt-en-mul
|
| 6 |
tags:
|
| 7 |
- generated_from_trainer
|
| 8 |
+
datasets:
|
| 9 |
+
- tiagoblima/translation-pt-indigenouns
|
| 10 |
metrics:
|
| 11 |
- bleu
|
| 12 |
model-index:
|
| 13 |
- name: tst-gun-gub-pt
|
| 14 |
+
results:
|
| 15 |
+
- task:
|
| 16 |
+
name: Translation
|
| 17 |
+
type: translation
|
| 18 |
+
dataset:
|
| 19 |
+
name: tiagoblima/translation-pt-indigenouns
|
| 20 |
+
type: tiagoblima/translation-pt-indigenouns
|
| 21 |
+
metrics:
|
| 22 |
+
- name: Bleu
|
| 23 |
+
type: bleu
|
| 24 |
+
value: 8.5368
|
| 25 |
---
|
| 26 |
|
| 27 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 29 |
|
| 30 |
# tst-gun-gub-pt
|
| 31 |
|
| 32 |
+
This model is a fine-tuned version of [/content/opus-mt-en-mul](https://huggingface.co//content/opus-mt-en-mul) on the tiagoblima/translation-pt-indigenouns dataset.
|
| 33 |
It achieves the following results on the evaluation set:
|
| 34 |
- Loss: 1.8158
|
| 35 |
+
- Bleu: 8.5368
|
| 36 |
+
- Gen Len: 59.24
|
| 37 |
|
| 38 |
## Model description
|
| 39 |
|
all_results.json
ADDED
|
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_bleu": 8.5368,
|
| 4 |
+
"eval_gen_len": 59.24,
|
| 5 |
+
"eval_loss": 1.8157883882522583,
|
| 6 |
+
"eval_runtime": 33.6814,
|
| 7 |
+
"eval_samples": 250,
|
| 8 |
+
"eval_samples_per_second": 7.422,
|
| 9 |
+
"eval_steps_per_second": 0.475,
|
| 10 |
+
"predict_bleu": 12.1638,
|
| 11 |
+
"predict_gen_len": 55.4231,
|
| 12 |
+
"predict_loss": 1.4852832555770874,
|
| 13 |
+
"predict_runtime": 468.6012,
|
| 14 |
+
"predict_samples": 3900,
|
| 15 |
+
"predict_samples_per_second": 8.323,
|
| 16 |
+
"predict_steps_per_second": 0.521,
|
| 17 |
+
"train_loss": 1.939443588256836,
|
| 18 |
+
"train_runtime": 15380.8623,
|
| 19 |
+
"train_samples": 217340,
|
| 20 |
+
"train_samples_per_second": 42.392,
|
| 21 |
+
"train_steps_per_second": 2.65
|
| 22 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_bleu": 8.5368,
|
| 4 |
+
"eval_gen_len": 59.24,
|
| 5 |
+
"eval_loss": 1.8157883882522583,
|
| 6 |
+
"eval_runtime": 33.6814,
|
| 7 |
+
"eval_samples": 250,
|
| 8 |
+
"eval_samples_per_second": 7.422,
|
| 9 |
+
"eval_steps_per_second": 0.475
|
| 10 |
+
}
|
generated_predictions.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
predict_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"predict_bleu": 12.1638,
|
| 3 |
+
"predict_gen_len": 55.4231,
|
| 4 |
+
"predict_loss": 1.4852832555770874,
|
| 5 |
+
"predict_runtime": 468.6012,
|
| 6 |
+
"predict_samples": 3900,
|
| 7 |
+
"predict_samples_per_second": 8.323,
|
| 8 |
+
"predict_steps_per_second": 0.521
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"train_loss": 1.939443588256836,
|
| 4 |
+
"train_runtime": 15380.8623,
|
| 5 |
+
"train_samples": 217340,
|
| 6 |
+
"train_samples_per_second": 42.392,
|
| 7 |
+
"train_steps_per_second": 2.65
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,614 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 1.8157883882522583,
|
| 3 |
+
"best_model_checkpoint": "/tmp/tst-gun-gub-pt/checkpoint-40000",
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 4000,
|
| 6 |
+
"global_step": 40752,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.04,
|
| 13 |
+
"learning_rate": 4.938653317628583e-05,
|
| 14 |
+
"loss": 3.3185,
|
| 15 |
+
"step": 500
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.07,
|
| 19 |
+
"learning_rate": 4.877306635257165e-05,
|
| 20 |
+
"loss": 2.967,
|
| 21 |
+
"step": 1000
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.11,
|
| 25 |
+
"learning_rate": 4.815959952885748e-05,
|
| 26 |
+
"loss": 2.803,
|
| 27 |
+
"step": 1500
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.15,
|
| 31 |
+
"learning_rate": 4.754613270514331e-05,
|
| 32 |
+
"loss": 2.6964,
|
| 33 |
+
"step": 2000
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.18,
|
| 37 |
+
"learning_rate": 4.6932665881429136e-05,
|
| 38 |
+
"loss": 2.6199,
|
| 39 |
+
"step": 2500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.22,
|
| 43 |
+
"learning_rate": 4.631919905771496e-05,
|
| 44 |
+
"loss": 2.5435,
|
| 45 |
+
"step": 3000
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.26,
|
| 49 |
+
"learning_rate": 4.5705732234000786e-05,
|
| 50 |
+
"loss": 2.4687,
|
| 51 |
+
"step": 3500
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.29,
|
| 55 |
+
"learning_rate": 4.509226541028662e-05,
|
| 56 |
+
"loss": 2.4421,
|
| 57 |
+
"step": 4000
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.29,
|
| 61 |
+
"eval_bleu": 3.3785,
|
| 62 |
+
"eval_gen_len": 68.48,
|
| 63 |
+
"eval_loss": 2.5284605026245117,
|
| 64 |
+
"eval_runtime": 39.0713,
|
| 65 |
+
"eval_samples_per_second": 6.399,
|
| 66 |
+
"eval_steps_per_second": 0.41,
|
| 67 |
+
"step": 4000
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 0.33,
|
| 71 |
+
"learning_rate": 4.4478798586572437e-05,
|
| 72 |
+
"loss": 2.3817,
|
| 73 |
+
"step": 4500
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.37,
|
| 77 |
+
"learning_rate": 4.386533176285827e-05,
|
| 78 |
+
"loss": 2.3424,
|
| 79 |
+
"step": 5000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.4,
|
| 83 |
+
"learning_rate": 4.3251864939144094e-05,
|
| 84 |
+
"loss": 2.3047,
|
| 85 |
+
"step": 5500
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 0.44,
|
| 89 |
+
"learning_rate": 4.263839811542992e-05,
|
| 90 |
+
"loss": 2.2823,
|
| 91 |
+
"step": 6000
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.48,
|
| 95 |
+
"learning_rate": 4.2024931291715744e-05,
|
| 96 |
+
"loss": 2.2436,
|
| 97 |
+
"step": 6500
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.52,
|
| 101 |
+
"learning_rate": 4.141146446800157e-05,
|
| 102 |
+
"loss": 2.2266,
|
| 103 |
+
"step": 7000
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"epoch": 0.55,
|
| 107 |
+
"learning_rate": 4.07979976442874e-05,
|
| 108 |
+
"loss": 2.1898,
|
| 109 |
+
"step": 7500
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 0.59,
|
| 113 |
+
"learning_rate": 4.018453082057323e-05,
|
| 114 |
+
"loss": 2.1667,
|
| 115 |
+
"step": 8000
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.59,
|
| 119 |
+
"eval_bleu": 4.5883,
|
| 120 |
+
"eval_gen_len": 58.6,
|
| 121 |
+
"eval_loss": 2.3017916679382324,
|
| 122 |
+
"eval_runtime": 34.6446,
|
| 123 |
+
"eval_samples_per_second": 7.216,
|
| 124 |
+
"eval_steps_per_second": 0.462,
|
| 125 |
+
"step": 8000
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 0.63,
|
| 129 |
+
"learning_rate": 3.957106399685905e-05,
|
| 130 |
+
"loss": 2.1403,
|
| 131 |
+
"step": 8500
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.66,
|
| 135 |
+
"learning_rate": 3.895759717314488e-05,
|
| 136 |
+
"loss": 2.1188,
|
| 137 |
+
"step": 9000
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.7,
|
| 141 |
+
"learning_rate": 3.834413034943071e-05,
|
| 142 |
+
"loss": 2.1002,
|
| 143 |
+
"step": 9500
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.74,
|
| 147 |
+
"learning_rate": 3.773066352571653e-05,
|
| 148 |
+
"loss": 2.0948,
|
| 149 |
+
"step": 10000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.77,
|
| 153 |
+
"learning_rate": 3.711719670200235e-05,
|
| 154 |
+
"loss": 2.0678,
|
| 155 |
+
"step": 10500
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.81,
|
| 159 |
+
"learning_rate": 3.6503729878288185e-05,
|
| 160 |
+
"loss": 2.0502,
|
| 161 |
+
"step": 11000
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.85,
|
| 165 |
+
"learning_rate": 3.589026305457401e-05,
|
| 166 |
+
"loss": 2.0416,
|
| 167 |
+
"step": 11500
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 0.88,
|
| 171 |
+
"learning_rate": 3.5276796230859836e-05,
|
| 172 |
+
"loss": 2.0255,
|
| 173 |
+
"step": 12000
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 0.88,
|
| 177 |
+
"eval_bleu": 5.1052,
|
| 178 |
+
"eval_gen_len": 67.3,
|
| 179 |
+
"eval_loss": 2.1290316581726074,
|
| 180 |
+
"eval_runtime": 40.7019,
|
| 181 |
+
"eval_samples_per_second": 6.142,
|
| 182 |
+
"eval_steps_per_second": 0.393,
|
| 183 |
+
"step": 12000
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.92,
|
| 187 |
+
"learning_rate": 3.466332940714566e-05,
|
| 188 |
+
"loss": 2.0057,
|
| 189 |
+
"step": 12500
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.96,
|
| 193 |
+
"learning_rate": 3.404986258343149e-05,
|
| 194 |
+
"loss": 1.9979,
|
| 195 |
+
"step": 13000
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.99,
|
| 199 |
+
"learning_rate": 3.343639575971731e-05,
|
| 200 |
+
"loss": 1.9726,
|
| 201 |
+
"step": 13500
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 1.03,
|
| 205 |
+
"learning_rate": 3.2822928936003144e-05,
|
| 206 |
+
"loss": 1.9261,
|
| 207 |
+
"step": 14000
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 1.07,
|
| 211 |
+
"learning_rate": 3.220946211228897e-05,
|
| 212 |
+
"loss": 1.9133,
|
| 213 |
+
"step": 14500
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 1.1,
|
| 217 |
+
"learning_rate": 3.15959952885748e-05,
|
| 218 |
+
"loss": 1.9109,
|
| 219 |
+
"step": 15000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.14,
|
| 223 |
+
"learning_rate": 3.098252846486062e-05,
|
| 224 |
+
"loss": 1.9171,
|
| 225 |
+
"step": 15500
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 1.18,
|
| 229 |
+
"learning_rate": 3.0369061641146445e-05,
|
| 230 |
+
"loss": 1.8995,
|
| 231 |
+
"step": 16000
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.18,
|
| 235 |
+
"eval_bleu": 7.8429,
|
| 236 |
+
"eval_gen_len": 55.48,
|
| 237 |
+
"eval_loss": 2.0535314083099365,
|
| 238 |
+
"eval_runtime": 32.581,
|
| 239 |
+
"eval_samples_per_second": 7.673,
|
| 240 |
+
"eval_steps_per_second": 0.491,
|
| 241 |
+
"step": 16000
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 1.21,
|
| 245 |
+
"learning_rate": 2.9755594817432277e-05,
|
| 246 |
+
"loss": 1.8771,
|
| 247 |
+
"step": 16500
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.25,
|
| 251 |
+
"learning_rate": 2.91421279937181e-05,
|
| 252 |
+
"loss": 1.8841,
|
| 253 |
+
"step": 17000
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 1.29,
|
| 257 |
+
"learning_rate": 2.8528661170003927e-05,
|
| 258 |
+
"loss": 1.8787,
|
| 259 |
+
"step": 17500
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 1.33,
|
| 263 |
+
"learning_rate": 2.7915194346289753e-05,
|
| 264 |
+
"loss": 1.8647,
|
| 265 |
+
"step": 18000
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 1.36,
|
| 269 |
+
"learning_rate": 2.730172752257558e-05,
|
| 270 |
+
"loss": 1.8488,
|
| 271 |
+
"step": 18500
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 1.4,
|
| 275 |
+
"learning_rate": 2.6688260698861407e-05,
|
| 276 |
+
"loss": 1.8342,
|
| 277 |
+
"step": 19000
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 1.44,
|
| 281 |
+
"learning_rate": 2.6074793875147235e-05,
|
| 282 |
+
"loss": 1.833,
|
| 283 |
+
"step": 19500
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 1.47,
|
| 287 |
+
"learning_rate": 2.546132705143306e-05,
|
| 288 |
+
"loss": 1.8322,
|
| 289 |
+
"step": 20000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.47,
|
| 293 |
+
"eval_bleu": 7.2663,
|
| 294 |
+
"eval_gen_len": 58.24,
|
| 295 |
+
"eval_loss": 1.9960261583328247,
|
| 296 |
+
"eval_runtime": 35.7526,
|
| 297 |
+
"eval_samples_per_second": 6.993,
|
| 298 |
+
"eval_steps_per_second": 0.448,
|
| 299 |
+
"step": 20000
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 1.51,
|
| 303 |
+
"learning_rate": 2.4847860227718886e-05,
|
| 304 |
+
"loss": 1.8187,
|
| 305 |
+
"step": 20500
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.55,
|
| 309 |
+
"learning_rate": 2.4234393404004714e-05,
|
| 310 |
+
"loss": 1.8118,
|
| 311 |
+
"step": 21000
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 1.58,
|
| 315 |
+
"learning_rate": 2.362092658029054e-05,
|
| 316 |
+
"loss": 1.8144,
|
| 317 |
+
"step": 21500
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.62,
|
| 321 |
+
"learning_rate": 2.3007459756576365e-05,
|
| 322 |
+
"loss": 1.7989,
|
| 323 |
+
"step": 22000
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 1.66,
|
| 327 |
+
"learning_rate": 2.2393992932862194e-05,
|
| 328 |
+
"loss": 1.8068,
|
| 329 |
+
"step": 22500
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 1.69,
|
| 333 |
+
"learning_rate": 2.178052610914802e-05,
|
| 334 |
+
"loss": 1.8043,
|
| 335 |
+
"step": 23000
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 1.73,
|
| 339 |
+
"learning_rate": 2.1167059285433844e-05,
|
| 340 |
+
"loss": 1.7863,
|
| 341 |
+
"step": 23500
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 1.77,
|
| 345 |
+
"learning_rate": 2.055359246171967e-05,
|
| 346 |
+
"loss": 1.7868,
|
| 347 |
+
"step": 24000
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.77,
|
| 351 |
+
"eval_bleu": 7.0981,
|
| 352 |
+
"eval_gen_len": 66.34,
|
| 353 |
+
"eval_loss": 1.922366976737976,
|
| 354 |
+
"eval_runtime": 40.7924,
|
| 355 |
+
"eval_samples_per_second": 6.129,
|
| 356 |
+
"eval_steps_per_second": 0.392,
|
| 357 |
+
"step": 24000
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 1.8,
|
| 361 |
+
"learning_rate": 1.9940125638005498e-05,
|
| 362 |
+
"loss": 1.7759,
|
| 363 |
+
"step": 24500
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 1.84,
|
| 367 |
+
"learning_rate": 1.9326658814291323e-05,
|
| 368 |
+
"loss": 1.7653,
|
| 369 |
+
"step": 25000
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 1.88,
|
| 373 |
+
"learning_rate": 1.871319199057715e-05,
|
| 374 |
+
"loss": 1.7689,
|
| 375 |
+
"step": 25500
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.91,
|
| 379 |
+
"learning_rate": 1.8099725166862977e-05,
|
| 380 |
+
"loss": 1.758,
|
| 381 |
+
"step": 26000
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.95,
|
| 385 |
+
"learning_rate": 1.7486258343148802e-05,
|
| 386 |
+
"loss": 1.754,
|
| 387 |
+
"step": 26500
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.99,
|
| 391 |
+
"learning_rate": 1.687279151943463e-05,
|
| 392 |
+
"loss": 1.7393,
|
| 393 |
+
"step": 27000
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 2.02,
|
| 397 |
+
"learning_rate": 1.6259324695720456e-05,
|
| 398 |
+
"loss": 1.7151,
|
| 399 |
+
"step": 27500
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 2.06,
|
| 403 |
+
"learning_rate": 1.5645857872006285e-05,
|
| 404 |
+
"loss": 1.7012,
|
| 405 |
+
"step": 28000
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 2.06,
|
| 409 |
+
"eval_bleu": 7.5657,
|
| 410 |
+
"eval_gen_len": 60.3,
|
| 411 |
+
"eval_loss": 1.8868523836135864,
|
| 412 |
+
"eval_runtime": 38.0123,
|
| 413 |
+
"eval_samples_per_second": 6.577,
|
| 414 |
+
"eval_steps_per_second": 0.421,
|
| 415 |
+
"step": 28000
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 2.1,
|
| 419 |
+
"learning_rate": 1.503239104829211e-05,
|
| 420 |
+
"loss": 1.6993,
|
| 421 |
+
"step": 28500
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 2.13,
|
| 425 |
+
"learning_rate": 1.4418924224577934e-05,
|
| 426 |
+
"loss": 1.6944,
|
| 427 |
+
"step": 29000
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 2.17,
|
| 431 |
+
"learning_rate": 1.380545740086376e-05,
|
| 432 |
+
"loss": 1.7064,
|
| 433 |
+
"step": 29500
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 2.21,
|
| 437 |
+
"learning_rate": 1.3191990577149588e-05,
|
| 438 |
+
"loss": 1.6993,
|
| 439 |
+
"step": 30000
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 2.25,
|
| 443 |
+
"learning_rate": 1.2578523753435415e-05,
|
| 444 |
+
"loss": 1.6893,
|
| 445 |
+
"step": 30500
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 2.28,
|
| 449 |
+
"learning_rate": 1.1965056929721242e-05,
|
| 450 |
+
"loss": 1.6874,
|
| 451 |
+
"step": 31000
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 2.32,
|
| 455 |
+
"learning_rate": 1.1351590106007069e-05,
|
| 456 |
+
"loss": 1.6793,
|
| 457 |
+
"step": 31500
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 2.36,
|
| 461 |
+
"learning_rate": 1.0738123282292894e-05,
|
| 462 |
+
"loss": 1.6773,
|
| 463 |
+
"step": 32000
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 2.36,
|
| 467 |
+
"eval_bleu": 7.9888,
|
| 468 |
+
"eval_gen_len": 61.18,
|
| 469 |
+
"eval_loss": 1.8613367080688477,
|
| 470 |
+
"eval_runtime": 38.7116,
|
| 471 |
+
"eval_samples_per_second": 6.458,
|
| 472 |
+
"eval_steps_per_second": 0.413,
|
| 473 |
+
"step": 32000
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 2.39,
|
| 477 |
+
"learning_rate": 1.012465645857872e-05,
|
| 478 |
+
"loss": 1.6798,
|
| 479 |
+
"step": 32500
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 2.43,
|
| 483 |
+
"learning_rate": 9.511189634864546e-06,
|
| 484 |
+
"loss": 1.6823,
|
| 485 |
+
"step": 33000
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.47,
|
| 489 |
+
"learning_rate": 8.897722811150373e-06,
|
| 490 |
+
"loss": 1.6711,
|
| 491 |
+
"step": 33500
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 2.5,
|
| 495 |
+
"learning_rate": 8.2842559874362e-06,
|
| 496 |
+
"loss": 1.679,
|
| 497 |
+
"step": 34000
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 2.54,
|
| 501 |
+
"learning_rate": 7.670789163722027e-06,
|
| 502 |
+
"loss": 1.6662,
|
| 503 |
+
"step": 34500
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 2.58,
|
| 507 |
+
"learning_rate": 7.057322340007853e-06,
|
| 508 |
+
"loss": 1.6716,
|
| 509 |
+
"step": 35000
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 2.61,
|
| 513 |
+
"learning_rate": 6.4438555162936784e-06,
|
| 514 |
+
"loss": 1.6569,
|
| 515 |
+
"step": 35500
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 2.65,
|
| 519 |
+
"learning_rate": 5.830388692579505e-06,
|
| 520 |
+
"loss": 1.6631,
|
| 521 |
+
"step": 36000
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 2.65,
|
| 525 |
+
"eval_bleu": 8.0862,
|
| 526 |
+
"eval_gen_len": 60.5,
|
| 527 |
+
"eval_loss": 1.8354450464248657,
|
| 528 |
+
"eval_runtime": 38.0466,
|
| 529 |
+
"eval_samples_per_second": 6.571,
|
| 530 |
+
"eval_steps_per_second": 0.421,
|
| 531 |
+
"step": 36000
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 2.69,
|
| 535 |
+
"learning_rate": 5.216921868865332e-06,
|
| 536 |
+
"loss": 1.6602,
|
| 537 |
+
"step": 36500
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 2.72,
|
| 541 |
+
"learning_rate": 4.6034550451511585e-06,
|
| 542 |
+
"loss": 1.6549,
|
| 543 |
+
"step": 37000
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 2.76,
|
| 547 |
+
"learning_rate": 3.989988221436985e-06,
|
| 548 |
+
"loss": 1.6521,
|
| 549 |
+
"step": 37500
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 2.8,
|
| 553 |
+
"learning_rate": 3.3765213977228115e-06,
|
| 554 |
+
"loss": 1.6643,
|
| 555 |
+
"step": 38000
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.83,
|
| 559 |
+
"learning_rate": 2.7630545740086376e-06,
|
| 560 |
+
"loss": 1.6533,
|
| 561 |
+
"step": 38500
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 2.87,
|
| 565 |
+
"learning_rate": 2.149587750294464e-06,
|
| 566 |
+
"loss": 1.6596,
|
| 567 |
+
"step": 39000
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 2.91,
|
| 571 |
+
"learning_rate": 1.5361209265802905e-06,
|
| 572 |
+
"loss": 1.6664,
|
| 573 |
+
"step": 39500
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 2.94,
|
| 577 |
+
"learning_rate": 9.22654102866117e-07,
|
| 578 |
+
"loss": 1.6379,
|
| 579 |
+
"step": 40000
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 2.94,
|
| 583 |
+
"eval_bleu": 8.4077,
|
| 584 |
+
"eval_gen_len": 60.18,
|
| 585 |
+
"eval_loss": 1.8157883882522583,
|
| 586 |
+
"eval_runtime": 35.7813,
|
| 587 |
+
"eval_samples_per_second": 6.987,
|
| 588 |
+
"eval_steps_per_second": 0.447,
|
| 589 |
+
"step": 40000
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 2.98,
|
| 593 |
+
"learning_rate": 3.091872791519435e-07,
|
| 594 |
+
"loss": 1.6493,
|
| 595 |
+
"step": 40500
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 3.0,
|
| 599 |
+
"step": 40752,
|
| 600 |
+
"total_flos": 3.4728164062396416e+16,
|
| 601 |
+
"train_loss": 1.939443588256836,
|
| 602 |
+
"train_runtime": 15380.8623,
|
| 603 |
+
"train_samples_per_second": 42.392,
|
| 604 |
+
"train_steps_per_second": 2.65
|
| 605 |
+
}
|
| 606 |
+
],
|
| 607 |
+
"logging_steps": 500,
|
| 608 |
+
"max_steps": 40752,
|
| 609 |
+
"num_train_epochs": 3,
|
| 610 |
+
"save_steps": 4000,
|
| 611 |
+
"total_flos": 3.4728164062396416e+16,
|
| 612 |
+
"trial_name": null,
|
| 613 |
+
"trial_params": null
|
| 614 |
+
}
|