Model save
Browse files- README.md +67 -0
- adapter_model.safetensors +1 -1
- all_results.json +13 -0
- eval_results.json +8 -0
- runs/Feb13_16-50-16_aida1/events.out.tfevents.1707839513.aida1.2927848.0 +2 -2
- runs/Feb13_16-50-16_aida1/events.out.tfevents.1707897760.aida1.2927848.1 +3 -0
- train_results.json +8 -0
- trainer_state.json +1742 -0
README.md
ADDED
|
@@ -0,0 +1,67 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
library_name: peft
|
| 4 |
+
tags:
|
| 5 |
+
- trl
|
| 6 |
+
- sft
|
| 7 |
+
- generated_from_trainer
|
| 8 |
+
datasets:
|
| 9 |
+
- generator
|
| 10 |
+
base_model: HuggingFaceH4/zephyr-7b-beta
|
| 11 |
+
model-index:
|
| 12 |
+
- name: conversational-PCRA-and-CCRA-zephyr-7b-beta-sft-qlora
|
| 13 |
+
results: []
|
| 14 |
+
---
|
| 15 |
+
|
| 16 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 17 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 18 |
+
|
| 19 |
+
# conversational-PCRA-and-CCRA-zephyr-7b-beta-sft-qlora
|
| 20 |
+
|
| 21 |
+
This model is a fine-tuned version of [HuggingFaceH4/zephyr-7b-beta](https://huggingface.co/HuggingFaceH4/zephyr-7b-beta) on the generator dataset.
|
| 22 |
+
It achieves the following results on the evaluation set:
|
| 23 |
+
- Loss: 1.3551
|
| 24 |
+
|
| 25 |
+
## Model description
|
| 26 |
+
|
| 27 |
+
More information needed
|
| 28 |
+
|
| 29 |
+
## Intended uses & limitations
|
| 30 |
+
|
| 31 |
+
More information needed
|
| 32 |
+
|
| 33 |
+
## Training and evaluation data
|
| 34 |
+
|
| 35 |
+
More information needed
|
| 36 |
+
|
| 37 |
+
## Training procedure
|
| 38 |
+
|
| 39 |
+
### Training hyperparameters
|
| 40 |
+
|
| 41 |
+
The following hyperparameters were used during training:
|
| 42 |
+
- learning_rate: 0.0002
|
| 43 |
+
- train_batch_size: 4
|
| 44 |
+
- eval_batch_size: 8
|
| 45 |
+
- seed: 42
|
| 46 |
+
- distributed_type: multi-GPU
|
| 47 |
+
- gradient_accumulation_steps: 2
|
| 48 |
+
- total_train_batch_size: 8
|
| 49 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 50 |
+
- lr_scheduler_type: cosine
|
| 51 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 52 |
+
- num_epochs: 1
|
| 53 |
+
|
| 54 |
+
### Training results
|
| 55 |
+
|
| 56 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 57 |
+
|:-------------:|:-----:|:----:|:---------------:|
|
| 58 |
+
| 1.3504 | 1.0 | 1419 | 1.3551 |
|
| 59 |
+
|
| 60 |
+
|
| 61 |
+
### Framework versions
|
| 62 |
+
|
| 63 |
+
- PEFT 0.7.1
|
| 64 |
+
- Transformers 4.36.2
|
| 65 |
+
- Pytorch 2.1.2
|
| 66 |
+
- Datasets 2.14.6
|
| 67 |
+
- Tokenizers 0.15.2
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 83946192
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:65676469f296a6e9f7a2679efffc3621d21780d634bdf17f7b770a5cb00b1e8a
|
| 3 |
size 83946192
|
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"eval_loss": 1.355061411857605,
|
| 4 |
+
"eval_runtime": 1843.1454,
|
| 5 |
+
"eval_samples": 3466,
|
| 6 |
+
"eval_samples_per_second": 0.715,
|
| 7 |
+
"eval_steps_per_second": 0.09,
|
| 8 |
+
"train_loss": 1.4210990094399267,
|
| 9 |
+
"train_runtime": 56404.0081,
|
| 10 |
+
"train_samples": 30392,
|
| 11 |
+
"train_samples_per_second": 0.201,
|
| 12 |
+
"train_steps_per_second": 0.025
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"eval_loss": 1.355061411857605,
|
| 4 |
+
"eval_runtime": 1843.1454,
|
| 5 |
+
"eval_samples": 3466,
|
| 6 |
+
"eval_samples_per_second": 0.715,
|
| 7 |
+
"eval_steps_per_second": 0.09
|
| 8 |
+
}
|
runs/Feb13_16-50-16_aida1/events.out.tfevents.1707839513.aida1.2927848.0
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fdc2f5e9353c876b639ba7a8ebc254993f8ae9e39e0199b3975b00de03e1361d
|
| 3 |
+
size 50107
|
runs/Feb13_16-50-16_aida1/events.out.tfevents.1707897760.aida1.2927848.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c5adc816e3f65409f4300af0ed689e51f5ab3bab4b37b06f035cbda8310dff03
|
| 3 |
+
size 359
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"train_loss": 1.4210990094399267,
|
| 4 |
+
"train_runtime": 56404.0081,
|
| 5 |
+
"train_samples": 30392,
|
| 6 |
+
"train_samples_per_second": 0.201,
|
| 7 |
+
"train_steps_per_second": 0.025
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1742 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9996477632969355,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1419,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0,
|
| 13 |
+
"learning_rate": 1.4084507042253521e-06,
|
| 14 |
+
"loss": 1.9425,
|
| 15 |
+
"step": 1
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.0,
|
| 19 |
+
"learning_rate": 7.042253521126762e-06,
|
| 20 |
+
"loss": 2.0234,
|
| 21 |
+
"step": 5
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.01,
|
| 25 |
+
"learning_rate": 1.4084507042253523e-05,
|
| 26 |
+
"loss": 1.9715,
|
| 27 |
+
"step": 10
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.01,
|
| 31 |
+
"learning_rate": 2.112676056338028e-05,
|
| 32 |
+
"loss": 1.9477,
|
| 33 |
+
"step": 15
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.01,
|
| 37 |
+
"learning_rate": 2.8169014084507046e-05,
|
| 38 |
+
"loss": 1.8815,
|
| 39 |
+
"step": 20
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.02,
|
| 43 |
+
"learning_rate": 3.5211267605633805e-05,
|
| 44 |
+
"loss": 1.8108,
|
| 45 |
+
"step": 25
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.02,
|
| 49 |
+
"learning_rate": 4.225352112676056e-05,
|
| 50 |
+
"loss": 1.7765,
|
| 51 |
+
"step": 30
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.02,
|
| 55 |
+
"learning_rate": 4.929577464788733e-05,
|
| 56 |
+
"loss": 1.7328,
|
| 57 |
+
"step": 35
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.03,
|
| 61 |
+
"learning_rate": 5.633802816901409e-05,
|
| 62 |
+
"loss": 1.7477,
|
| 63 |
+
"step": 40
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.03,
|
| 67 |
+
"learning_rate": 6.338028169014085e-05,
|
| 68 |
+
"loss": 1.6816,
|
| 69 |
+
"step": 45
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.04,
|
| 73 |
+
"learning_rate": 7.042253521126761e-05,
|
| 74 |
+
"loss": 1.6714,
|
| 75 |
+
"step": 50
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.04,
|
| 79 |
+
"learning_rate": 7.746478873239437e-05,
|
| 80 |
+
"loss": 1.642,
|
| 81 |
+
"step": 55
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.04,
|
| 85 |
+
"learning_rate": 8.450704225352113e-05,
|
| 86 |
+
"loss": 1.612,
|
| 87 |
+
"step": 60
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.05,
|
| 91 |
+
"learning_rate": 9.15492957746479e-05,
|
| 92 |
+
"loss": 1.592,
|
| 93 |
+
"step": 65
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.05,
|
| 97 |
+
"learning_rate": 9.859154929577466e-05,
|
| 98 |
+
"loss": 1.6068,
|
| 99 |
+
"step": 70
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.05,
|
| 103 |
+
"learning_rate": 0.0001056338028169014,
|
| 104 |
+
"loss": 1.5911,
|
| 105 |
+
"step": 75
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.06,
|
| 109 |
+
"learning_rate": 0.00011267605633802819,
|
| 110 |
+
"loss": 1.5727,
|
| 111 |
+
"step": 80
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.06,
|
| 115 |
+
"learning_rate": 0.00011971830985915493,
|
| 116 |
+
"loss": 1.5929,
|
| 117 |
+
"step": 85
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.06,
|
| 121 |
+
"learning_rate": 0.0001267605633802817,
|
| 122 |
+
"loss": 1.5441,
|
| 123 |
+
"step": 90
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.07,
|
| 127 |
+
"learning_rate": 0.00013380281690140845,
|
| 128 |
+
"loss": 1.5659,
|
| 129 |
+
"step": 95
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.07,
|
| 133 |
+
"learning_rate": 0.00014084507042253522,
|
| 134 |
+
"loss": 1.5422,
|
| 135 |
+
"step": 100
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.07,
|
| 139 |
+
"learning_rate": 0.000147887323943662,
|
| 140 |
+
"loss": 1.5782,
|
| 141 |
+
"step": 105
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.08,
|
| 145 |
+
"learning_rate": 0.00015492957746478874,
|
| 146 |
+
"loss": 1.5553,
|
| 147 |
+
"step": 110
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.08,
|
| 151 |
+
"learning_rate": 0.0001619718309859155,
|
| 152 |
+
"loss": 1.5324,
|
| 153 |
+
"step": 115
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.08,
|
| 157 |
+
"learning_rate": 0.00016901408450704225,
|
| 158 |
+
"loss": 1.4999,
|
| 159 |
+
"step": 120
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.09,
|
| 163 |
+
"learning_rate": 0.00017605633802816902,
|
| 164 |
+
"loss": 1.5137,
|
| 165 |
+
"step": 125
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.09,
|
| 169 |
+
"learning_rate": 0.0001830985915492958,
|
| 170 |
+
"loss": 1.5549,
|
| 171 |
+
"step": 130
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.1,
|
| 175 |
+
"learning_rate": 0.00019014084507042254,
|
| 176 |
+
"loss": 1.5369,
|
| 177 |
+
"step": 135
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.1,
|
| 181 |
+
"learning_rate": 0.0001971830985915493,
|
| 182 |
+
"loss": 1.5427,
|
| 183 |
+
"step": 140
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.1,
|
| 187 |
+
"learning_rate": 0.00019999727649301603,
|
| 188 |
+
"loss": 1.5195,
|
| 189 |
+
"step": 145
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.11,
|
| 193 |
+
"learning_rate": 0.00019998063337645585,
|
| 194 |
+
"loss": 1.5184,
|
| 195 |
+
"step": 150
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.11,
|
| 199 |
+
"learning_rate": 0.0001999488627179007,
|
| 200 |
+
"loss": 1.5382,
|
| 201 |
+
"step": 155
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.11,
|
| 205 |
+
"learning_rate": 0.00019990196932440123,
|
| 206 |
+
"loss": 1.5211,
|
| 207 |
+
"step": 160
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.12,
|
| 211 |
+
"learning_rate": 0.00019983996029114938,
|
| 212 |
+
"loss": 1.5237,
|
| 213 |
+
"step": 165
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.12,
|
| 217 |
+
"learning_rate": 0.00019976284500040467,
|
| 218 |
+
"loss": 1.5225,
|
| 219 |
+
"step": 170
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.12,
|
| 223 |
+
"learning_rate": 0.00019967063512007482,
|
| 224 |
+
"loss": 1.5393,
|
| 225 |
+
"step": 175
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.13,
|
| 229 |
+
"learning_rate": 0.00019956334460195015,
|
| 230 |
+
"loss": 1.4915,
|
| 231 |
+
"step": 180
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.13,
|
| 235 |
+
"learning_rate": 0.00019944098967959278,
|
| 236 |
+
"loss": 1.492,
|
| 237 |
+
"step": 185
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.13,
|
| 241 |
+
"learning_rate": 0.00019930358886588025,
|
| 242 |
+
"loss": 1.5237,
|
| 243 |
+
"step": 190
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.14,
|
| 247 |
+
"learning_rate": 0.00019915116295020458,
|
| 248 |
+
"loss": 1.5198,
|
| 249 |
+
"step": 195
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.14,
|
| 253 |
+
"learning_rate": 0.00019898373499532666,
|
| 254 |
+
"loss": 1.5046,
|
| 255 |
+
"step": 200
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.14,
|
| 259 |
+
"learning_rate": 0.0001988013303338867,
|
| 260 |
+
"loss": 1.4883,
|
| 261 |
+
"step": 205
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.15,
|
| 265 |
+
"learning_rate": 0.00019860397656457147,
|
| 266 |
+
"loss": 1.4738,
|
| 267 |
+
"step": 210
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.15,
|
| 271 |
+
"learning_rate": 0.00019839170354793827,
|
| 272 |
+
"loss": 1.5288,
|
| 273 |
+
"step": 215
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.15,
|
| 277 |
+
"learning_rate": 0.00019816454340189693,
|
| 278 |
+
"loss": 1.5271,
|
| 279 |
+
"step": 220
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.16,
|
| 283 |
+
"learning_rate": 0.0001979225304968504,
|
| 284 |
+
"loss": 1.4791,
|
| 285 |
+
"step": 225
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.16,
|
| 289 |
+
"learning_rate": 0.00019766570145049407,
|
| 290 |
+
"loss": 1.4587,
|
| 291 |
+
"step": 230
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 0.17,
|
| 295 |
+
"learning_rate": 0.0001973940951222756,
|
| 296 |
+
"loss": 1.4508,
|
| 297 |
+
"step": 235
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.17,
|
| 301 |
+
"learning_rate": 0.0001971077526075151,
|
| 302 |
+
"loss": 1.5056,
|
| 303 |
+
"step": 240
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.17,
|
| 307 |
+
"learning_rate": 0.00019680671723118734,
|
| 308 |
+
"loss": 1.4492,
|
| 309 |
+
"step": 245
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.18,
|
| 313 |
+
"learning_rate": 0.0001964910345413664,
|
| 314 |
+
"loss": 1.5118,
|
| 315 |
+
"step": 250
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.18,
|
| 319 |
+
"learning_rate": 0.00019616075230233408,
|
| 320 |
+
"loss": 1.4772,
|
| 321 |
+
"step": 255
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.18,
|
| 325 |
+
"learning_rate": 0.00019581592048735296,
|
| 326 |
+
"loss": 1.4862,
|
| 327 |
+
"step": 260
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.19,
|
| 331 |
+
"learning_rate": 0.00019545659127110507,
|
| 332 |
+
"loss": 1.4979,
|
| 333 |
+
"step": 265
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.19,
|
| 337 |
+
"learning_rate": 0.00019508281902179782,
|
| 338 |
+
"loss": 1.4564,
|
| 339 |
+
"step": 270
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.19,
|
| 343 |
+
"learning_rate": 0.00019469466029293776,
|
| 344 |
+
"loss": 1.4446,
|
| 345 |
+
"step": 275
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.2,
|
| 349 |
+
"learning_rate": 0.00019429217381477357,
|
| 350 |
+
"loss": 1.4828,
|
| 351 |
+
"step": 280
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.2,
|
| 355 |
+
"learning_rate": 0.00019387542048541023,
|
| 356 |
+
"loss": 1.4632,
|
| 357 |
+
"step": 285
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 0.2,
|
| 361 |
+
"learning_rate": 0.0001934444633615946,
|
| 362 |
+
"loss": 1.4543,
|
| 363 |
+
"step": 290
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.21,
|
| 367 |
+
"learning_rate": 0.0001929993676491747,
|
| 368 |
+
"loss": 1.4742,
|
| 369 |
+
"step": 295
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.21,
|
| 373 |
+
"learning_rate": 0.00019254020069323386,
|
| 374 |
+
"loss": 1.4811,
|
| 375 |
+
"step": 300
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.21,
|
| 379 |
+
"learning_rate": 0.00019206703196790096,
|
| 380 |
+
"loss": 1.4505,
|
| 381 |
+
"step": 305
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.22,
|
| 385 |
+
"learning_rate": 0.00019157993306583867,
|
| 386 |
+
"loss": 1.4237,
|
| 387 |
+
"step": 310
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.22,
|
| 391 |
+
"learning_rate": 0.00019107897768741127,
|
| 392 |
+
"loss": 1.4285,
|
| 393 |
+
"step": 315
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.23,
|
| 397 |
+
"learning_rate": 0.0001905642416295333,
|
| 398 |
+
"loss": 1.4454,
|
| 399 |
+
"step": 320
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.23,
|
| 403 |
+
"learning_rate": 0.0001900358027742012,
|
| 404 |
+
"loss": 1.42,
|
| 405 |
+
"step": 325
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.23,
|
| 409 |
+
"learning_rate": 0.00018949374107670935,
|
| 410 |
+
"loss": 1.4507,
|
| 411 |
+
"step": 330
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 0.24,
|
| 415 |
+
"learning_rate": 0.0001889381385535525,
|
| 416 |
+
"loss": 1.4222,
|
| 417 |
+
"step": 335
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.24,
|
| 421 |
+
"learning_rate": 0.00018836907927001628,
|
| 422 |
+
"loss": 1.4502,
|
| 423 |
+
"step": 340
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.24,
|
| 427 |
+
"learning_rate": 0.00018778664932745772,
|
| 428 |
+
"loss": 1.4255,
|
| 429 |
+
"step": 345
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.25,
|
| 433 |
+
"learning_rate": 0.0001871909368502777,
|
| 434 |
+
"loss": 1.4597,
|
| 435 |
+
"step": 350
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 0.25,
|
| 439 |
+
"learning_rate": 0.00018658203197258728,
|
| 440 |
+
"loss": 1.432,
|
| 441 |
+
"step": 355
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.25,
|
| 445 |
+
"learning_rate": 0.0001859600268245701,
|
| 446 |
+
"loss": 1.4676,
|
| 447 |
+
"step": 360
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 0.26,
|
| 451 |
+
"learning_rate": 0.00018532501551854242,
|
| 452 |
+
"loss": 1.4267,
|
| 453 |
+
"step": 365
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 0.26,
|
| 457 |
+
"learning_rate": 0.00018467709413471378,
|
| 458 |
+
"loss": 1.4392,
|
| 459 |
+
"step": 370
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.26,
|
| 463 |
+
"learning_rate": 0.00018401636070664945,
|
| 464 |
+
"loss": 1.4067,
|
| 465 |
+
"step": 375
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.27,
|
| 469 |
+
"learning_rate": 0.0001833429152064374,
|
| 470 |
+
"loss": 1.4427,
|
| 471 |
+
"step": 380
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.27,
|
| 475 |
+
"learning_rate": 0.00018265685952956222,
|
| 476 |
+
"loss": 1.4519,
|
| 477 |
+
"step": 385
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.27,
|
| 481 |
+
"learning_rate": 0.00018195829747948773,
|
| 482 |
+
"loss": 1.4449,
|
| 483 |
+
"step": 390
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 0.28,
|
| 487 |
+
"learning_rate": 0.00018124733475195117,
|
| 488 |
+
"loss": 1.424,
|
| 489 |
+
"step": 395
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.28,
|
| 493 |
+
"learning_rate": 0.00018052407891897075,
|
| 494 |
+
"loss": 1.447,
|
| 495 |
+
"step": 400
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.29,
|
| 499 |
+
"learning_rate": 0.0001797886394125696,
|
| 500 |
+
"loss": 1.4363,
|
| 501 |
+
"step": 405
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.29,
|
| 505 |
+
"learning_rate": 0.00017904112750821824,
|
| 506 |
+
"loss": 1.4152,
|
| 507 |
+
"step": 410
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.29,
|
| 511 |
+
"learning_rate": 0.00017828165630799796,
|
| 512 |
+
"loss": 1.41,
|
| 513 |
+
"step": 415
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.3,
|
| 517 |
+
"learning_rate": 0.000177510340723488,
|
| 518 |
+
"loss": 1.446,
|
| 519 |
+
"step": 420
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.3,
|
| 523 |
+
"learning_rate": 0.0001767272974583789,
|
| 524 |
+
"loss": 1.4449,
|
| 525 |
+
"step": 425
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 0.3,
|
| 529 |
+
"learning_rate": 0.00017593264499081465,
|
| 530 |
+
"loss": 1.4519,
|
| 531 |
+
"step": 430
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 0.31,
|
| 535 |
+
"learning_rate": 0.00017512650355546634,
|
| 536 |
+
"loss": 1.4323,
|
| 537 |
+
"step": 435
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.31,
|
| 541 |
+
"learning_rate": 0.00017430899512534026,
|
| 542 |
+
"loss": 1.424,
|
| 543 |
+
"step": 440
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 0.31,
|
| 547 |
+
"learning_rate": 0.00017348024339332259,
|
| 548 |
+
"loss": 1.3989,
|
| 549 |
+
"step": 445
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.32,
|
| 553 |
+
"learning_rate": 0.0001726403737534642,
|
| 554 |
+
"loss": 1.4005,
|
| 555 |
+
"step": 450
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.32,
|
| 559 |
+
"learning_rate": 0.00017178951328200798,
|
| 560 |
+
"loss": 1.4457,
|
| 561 |
+
"step": 455
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.32,
|
| 565 |
+
"learning_rate": 0.0001709277907181615,
|
| 566 |
+
"loss": 1.4335,
|
| 567 |
+
"step": 460
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 0.33,
|
| 571 |
+
"learning_rate": 0.0001700553364446182,
|
| 572 |
+
"loss": 1.4705,
|
| 573 |
+
"step": 465
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 0.33,
|
| 577 |
+
"learning_rate": 0.00016917228246782987,
|
| 578 |
+
"loss": 1.4012,
|
| 579 |
+
"step": 470
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.33,
|
| 583 |
+
"learning_rate": 0.00016827876239803352,
|
| 584 |
+
"loss": 1.4038,
|
| 585 |
+
"step": 475
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.34,
|
| 589 |
+
"learning_rate": 0.0001673749114290354,
|
| 590 |
+
"loss": 1.4353,
|
| 591 |
+
"step": 480
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.34,
|
| 595 |
+
"learning_rate": 0.00016646086631775563,
|
| 596 |
+
"loss": 1.4034,
|
| 597 |
+
"step": 485
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.35,
|
| 601 |
+
"learning_rate": 0.00016553676536353612,
|
| 602 |
+
"loss": 1.4265,
|
| 603 |
+
"step": 490
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 0.35,
|
| 607 |
+
"learning_rate": 0.00016460274838721545,
|
| 608 |
+
"loss": 1.4354,
|
| 609 |
+
"step": 495
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 0.35,
|
| 613 |
+
"learning_rate": 0.00016365895670997305,
|
| 614 |
+
"loss": 1.4099,
|
| 615 |
+
"step": 500
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 0.36,
|
| 619 |
+
"learning_rate": 0.00016270553313194684,
|
| 620 |
+
"loss": 1.4237,
|
| 621 |
+
"step": 505
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.36,
|
| 625 |
+
"learning_rate": 0.00016174262191062674,
|
| 626 |
+
"loss": 1.4616,
|
| 627 |
+
"step": 510
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 0.36,
|
| 631 |
+
"learning_rate": 0.00016077036873902797,
|
| 632 |
+
"loss": 1.4041,
|
| 633 |
+
"step": 515
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.37,
|
| 637 |
+
"learning_rate": 0.00015978892072364694,
|
| 638 |
+
"loss": 1.4019,
|
| 639 |
+
"step": 520
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.37,
|
| 643 |
+
"learning_rate": 0.00015879842636220328,
|
| 644 |
+
"loss": 1.3983,
|
| 645 |
+
"step": 525
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 0.37,
|
| 649 |
+
"learning_rate": 0.00015779903552117153,
|
| 650 |
+
"loss": 1.4043,
|
| 651 |
+
"step": 530
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 0.38,
|
| 655 |
+
"learning_rate": 0.00015679089941310575,
|
| 656 |
+
"loss": 1.373,
|
| 657 |
+
"step": 535
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 0.38,
|
| 661 |
+
"learning_rate": 0.00015577417057376,
|
| 662 |
+
"loss": 1.4002,
|
| 663 |
+
"step": 540
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 0.38,
|
| 667 |
+
"learning_rate": 0.00015474900283900923,
|
| 668 |
+
"loss": 1.3802,
|
| 669 |
+
"step": 545
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.39,
|
| 673 |
+
"learning_rate": 0.00015371555132157318,
|
| 674 |
+
"loss": 1.3933,
|
| 675 |
+
"step": 550
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.39,
|
| 679 |
+
"learning_rate": 0.00015267397238754693,
|
| 680 |
+
"loss": 1.4463,
|
| 681 |
+
"step": 555
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.39,
|
| 685 |
+
"learning_rate": 0.00015162442363274214,
|
| 686 |
+
"loss": 1.3943,
|
| 687 |
+
"step": 560
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 0.4,
|
| 691 |
+
"learning_rate": 0.00015056706385884196,
|
| 692 |
+
"loss": 1.4336,
|
| 693 |
+
"step": 565
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 0.4,
|
| 697 |
+
"learning_rate": 0.00014950205304937368,
|
| 698 |
+
"loss": 1.3881,
|
| 699 |
+
"step": 570
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.41,
|
| 703 |
+
"learning_rate": 0.00014842955234550231,
|
| 704 |
+
"loss": 1.3875,
|
| 705 |
+
"step": 575
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.41,
|
| 709 |
+
"learning_rate": 0.00014734972402164932,
|
| 710 |
+
"loss": 1.4053,
|
| 711 |
+
"step": 580
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.41,
|
| 715 |
+
"learning_rate": 0.00014626273146093967,
|
| 716 |
+
"loss": 1.404,
|
| 717 |
+
"step": 585
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.42,
|
| 721 |
+
"learning_rate": 0.00014516873913048117,
|
| 722 |
+
"loss": 1.4101,
|
| 723 |
+
"step": 590
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.42,
|
| 727 |
+
"learning_rate": 0.0001440679125564799,
|
| 728 |
+
"loss": 1.3958,
|
| 729 |
+
"step": 595
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.42,
|
| 733 |
+
"learning_rate": 0.00014296041829919522,
|
| 734 |
+
"loss": 1.3972,
|
| 735 |
+
"step": 600
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 0.43,
|
| 739 |
+
"learning_rate": 0.00014184642392773845,
|
| 740 |
+
"loss": 1.4385,
|
| 741 |
+
"step": 605
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 0.43,
|
| 745 |
+
"learning_rate": 0.00014072609799471896,
|
| 746 |
+
"loss": 1.4154,
|
| 747 |
+
"step": 610
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 0.43,
|
| 751 |
+
"learning_rate": 0.00013959961001074113,
|
| 752 |
+
"loss": 1.392,
|
| 753 |
+
"step": 615
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.44,
|
| 757 |
+
"learning_rate": 0.00013846713041875693,
|
| 758 |
+
"loss": 1.3893,
|
| 759 |
+
"step": 620
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.44,
|
| 763 |
+
"learning_rate": 0.00013732883056827684,
|
| 764 |
+
"loss": 1.3978,
|
| 765 |
+
"step": 625
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.44,
|
| 769 |
+
"learning_rate": 0.0001361848826894439,
|
| 770 |
+
"loss": 1.414,
|
| 771 |
+
"step": 630
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 0.45,
|
| 775 |
+
"learning_rate": 0.00013503545986697456,
|
| 776 |
+
"loss": 1.3606,
|
| 777 |
+
"step": 635
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 0.45,
|
| 781 |
+
"learning_rate": 0.00013388073601397008,
|
| 782 |
+
"loss": 1.4059,
|
| 783 |
+
"step": 640
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 0.45,
|
| 787 |
+
"learning_rate": 0.0001327208858456027,
|
| 788 |
+
"loss": 1.3912,
|
| 789 |
+
"step": 645
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.46,
|
| 793 |
+
"learning_rate": 0.00013155608485268031,
|
| 794 |
+
"loss": 1.4211,
|
| 795 |
+
"step": 650
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 0.46,
|
| 799 |
+
"learning_rate": 0.00013038650927509404,
|
| 800 |
+
"loss": 1.384,
|
| 801 |
+
"step": 655
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.46,
|
| 805 |
+
"learning_rate": 0.00012921233607515216,
|
| 806 |
+
"loss": 1.3631,
|
| 807 |
+
"step": 660
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.47,
|
| 811 |
+
"learning_rate": 0.0001280337429108049,
|
| 812 |
+
"loss": 1.3803,
|
| 813 |
+
"step": 665
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 0.47,
|
| 817 |
+
"learning_rate": 0.00012685090810876407,
|
| 818 |
+
"loss": 1.3631,
|
| 819 |
+
"step": 670
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.48,
|
| 823 |
+
"learning_rate": 0.0001256640106375212,
|
| 824 |
+
"loss": 1.4074,
|
| 825 |
+
"step": 675
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 0.48,
|
| 829 |
+
"learning_rate": 0.0001244732300802689,
|
| 830 |
+
"loss": 1.386,
|
| 831 |
+
"step": 680
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.48,
|
| 835 |
+
"learning_rate": 0.00012327874660772898,
|
| 836 |
+
"loss": 1.4198,
|
| 837 |
+
"step": 685
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.49,
|
| 841 |
+
"learning_rate": 0.00012208074095089192,
|
| 842 |
+
"loss": 1.3607,
|
| 843 |
+
"step": 690
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.49,
|
| 847 |
+
"learning_rate": 0.00012087939437367126,
|
| 848 |
+
"loss": 1.3762,
|
| 849 |
+
"step": 695
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.49,
|
| 853 |
+
"learning_rate": 0.0001196748886454775,
|
| 854 |
+
"loss": 1.389,
|
| 855 |
+
"step": 700
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 0.5,
|
| 859 |
+
"learning_rate": 0.00011846740601371576,
|
| 860 |
+
"loss": 1.4058,
|
| 861 |
+
"step": 705
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 0.5,
|
| 865 |
+
"learning_rate": 0.00011725712917621059,
|
| 866 |
+
"loss": 1.3868,
|
| 867 |
+
"step": 710
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 0.5,
|
| 871 |
+
"learning_rate": 0.00011604424125356312,
|
| 872 |
+
"loss": 1.3644,
|
| 873 |
+
"step": 715
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.51,
|
| 877 |
+
"learning_rate": 0.00011482892576144405,
|
| 878 |
+
"loss": 1.3894,
|
| 879 |
+
"step": 720
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.51,
|
| 883 |
+
"learning_rate": 0.00011361136658282663,
|
| 884 |
+
"loss": 1.363,
|
| 885 |
+
"step": 725
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.51,
|
| 889 |
+
"learning_rate": 0.00011239174794016469,
|
| 890 |
+
"loss": 1.3882,
|
| 891 |
+
"step": 730
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.52,
|
| 895 |
+
"learning_rate": 0.00011117025436751855,
|
| 896 |
+
"loss": 1.409,
|
| 897 |
+
"step": 735
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.52,
|
| 901 |
+
"learning_rate": 0.00010994707068263434,
|
| 902 |
+
"loss": 1.3615,
|
| 903 |
+
"step": 740
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.52,
|
| 907 |
+
"learning_rate": 0.00010872238195898019,
|
| 908 |
+
"loss": 1.3623,
|
| 909 |
+
"step": 745
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 0.53,
|
| 913 |
+
"learning_rate": 0.00010749637349774357,
|
| 914 |
+
"loss": 1.3658,
|
| 915 |
+
"step": 750
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 0.53,
|
| 919 |
+
"learning_rate": 0.00010626923079979465,
|
| 920 |
+
"loss": 1.3687,
|
| 921 |
+
"step": 755
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 0.54,
|
| 925 |
+
"learning_rate": 0.00010504113953761885,
|
| 926 |
+
"loss": 1.3753,
|
| 927 |
+
"step": 760
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.54,
|
| 931 |
+
"learning_rate": 0.00010381228552722392,
|
| 932 |
+
"loss": 1.3861,
|
| 933 |
+
"step": 765
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.54,
|
| 937 |
+
"learning_rate": 0.00010258285470002494,
|
| 938 |
+
"loss": 1.3645,
|
| 939 |
+
"step": 770
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 0.55,
|
| 943 |
+
"learning_rate": 0.00010135303307471213,
|
| 944 |
+
"loss": 1.3761,
|
| 945 |
+
"step": 775
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 0.55,
|
| 949 |
+
"learning_rate": 0.00010012300672910535,
|
| 950 |
+
"loss": 1.3553,
|
| 951 |
+
"step": 780
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 0.55,
|
| 955 |
+
"learning_rate": 9.889296177199954e-05,
|
| 956 |
+
"loss": 1.3739,
|
| 957 |
+
"step": 785
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 0.56,
|
| 961 |
+
"learning_rate": 9.766308431500566e-05,
|
| 962 |
+
"loss": 1.3452,
|
| 963 |
+
"step": 790
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 0.56,
|
| 967 |
+
"learning_rate": 9.64335604443912e-05,
|
| 968 |
+
"loss": 1.3559,
|
| 969 |
+
"step": 795
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.56,
|
| 973 |
+
"learning_rate": 9.520457619292423e-05,
|
| 974 |
+
"loss": 1.3802,
|
| 975 |
+
"step": 800
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.57,
|
| 979 |
+
"learning_rate": 9.397631751172601e-05,
|
| 980 |
+
"loss": 1.3469,
|
| 981 |
+
"step": 805
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 0.57,
|
| 985 |
+
"learning_rate": 9.274897024213544e-05,
|
| 986 |
+
"loss": 1.3818,
|
| 987 |
+
"step": 810
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 0.57,
|
| 991 |
+
"learning_rate": 9.152272008759076e-05,
|
| 992 |
+
"loss": 1.3351,
|
| 993 |
+
"step": 815
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 0.58,
|
| 997 |
+
"learning_rate": 9.029775258553128e-05,
|
| 998 |
+
"loss": 1.3855,
|
| 999 |
+
"step": 820
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.58,
|
| 1003 |
+
"learning_rate": 8.907425307932514e-05,
|
| 1004 |
+
"loss": 1.3522,
|
| 1005 |
+
"step": 825
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 0.58,
|
| 1009 |
+
"learning_rate": 8.785240669022568e-05,
|
| 1010 |
+
"loss": 1.344,
|
| 1011 |
+
"step": 830
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.59,
|
| 1015 |
+
"learning_rate": 8.663239828936174e-05,
|
| 1016 |
+
"loss": 1.3704,
|
| 1017 |
+
"step": 835
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.59,
|
| 1021 |
+
"learning_rate": 8.541441246976607e-05,
|
| 1022 |
+
"loss": 1.3571,
|
| 1023 |
+
"step": 840
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 0.6,
|
| 1027 |
+
"learning_rate": 8.419863351844508e-05,
|
| 1028 |
+
"loss": 1.3621,
|
| 1029 |
+
"step": 845
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 0.6,
|
| 1033 |
+
"learning_rate": 8.298524538849576e-05,
|
| 1034 |
+
"loss": 1.3889,
|
| 1035 |
+
"step": 850
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 0.6,
|
| 1039 |
+
"learning_rate": 8.177443167127244e-05,
|
| 1040 |
+
"loss": 1.3597,
|
| 1041 |
+
"step": 855
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 0.61,
|
| 1045 |
+
"learning_rate": 8.056637556860872e-05,
|
| 1046 |
+
"loss": 1.3719,
|
| 1047 |
+
"step": 860
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 0.61,
|
| 1051 |
+
"learning_rate": 7.936125986509803e-05,
|
| 1052 |
+
"loss": 1.3565,
|
| 1053 |
+
"step": 865
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.61,
|
| 1057 |
+
"learning_rate": 7.815926690043756e-05,
|
| 1058 |
+
"loss": 1.3909,
|
| 1059 |
+
"step": 870
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.62,
|
| 1063 |
+
"learning_rate": 7.696057854183935e-05,
|
| 1064 |
+
"loss": 1.3711,
|
| 1065 |
+
"step": 875
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 0.62,
|
| 1069 |
+
"learning_rate": 7.576537615651295e-05,
|
| 1070 |
+
"loss": 1.3712,
|
| 1071 |
+
"step": 880
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 0.62,
|
| 1075 |
+
"learning_rate": 7.457384058422368e-05,
|
| 1076 |
+
"loss": 1.3663,
|
| 1077 |
+
"step": 885
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 0.63,
|
| 1081 |
+
"learning_rate": 7.338615210993074e-05,
|
| 1082 |
+
"loss": 1.3771,
|
| 1083 |
+
"step": 890
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 0.63,
|
| 1087 |
+
"learning_rate": 7.220249043650918e-05,
|
| 1088 |
+
"loss": 1.3454,
|
| 1089 |
+
"step": 895
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 0.63,
|
| 1093 |
+
"learning_rate": 7.102303465756019e-05,
|
| 1094 |
+
"loss": 1.3879,
|
| 1095 |
+
"step": 900
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.64,
|
| 1099 |
+
"learning_rate": 6.98479632303131e-05,
|
| 1100 |
+
"loss": 1.3803,
|
| 1101 |
+
"step": 905
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.64,
|
| 1105 |
+
"learning_rate": 6.867745394862422e-05,
|
| 1106 |
+
"loss": 1.3628,
|
| 1107 |
+
"step": 910
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 0.64,
|
| 1111 |
+
"learning_rate": 6.751168391607576e-05,
|
| 1112 |
+
"loss": 1.4021,
|
| 1113 |
+
"step": 915
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 0.65,
|
| 1117 |
+
"learning_rate": 6.635082951917897e-05,
|
| 1118 |
+
"loss": 1.3821,
|
| 1119 |
+
"step": 920
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 0.65,
|
| 1123 |
+
"learning_rate": 6.519506640068638e-05,
|
| 1124 |
+
"loss": 1.3592,
|
| 1125 |
+
"step": 925
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 0.66,
|
| 1129 |
+
"learning_rate": 6.404456943301592e-05,
|
| 1130 |
+
"loss": 1.3659,
|
| 1131 |
+
"step": 930
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 0.66,
|
| 1135 |
+
"learning_rate": 6.289951269179215e-05,
|
| 1136 |
+
"loss": 1.3887,
|
| 1137 |
+
"step": 935
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.66,
|
| 1141 |
+
"learning_rate": 6.17600694295076e-05,
|
| 1142 |
+
"loss": 1.3603,
|
| 1143 |
+
"step": 940
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.67,
|
| 1147 |
+
"learning_rate": 6.0626412049308965e-05,
|
| 1148 |
+
"loss": 1.3654,
|
| 1149 |
+
"step": 945
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 0.67,
|
| 1153 |
+
"learning_rate": 5.949871207891162e-05,
|
| 1154 |
+
"loss": 1.3842,
|
| 1155 |
+
"step": 950
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 0.67,
|
| 1159 |
+
"learning_rate": 5.837714014464677e-05,
|
| 1160 |
+
"loss": 1.3333,
|
| 1161 |
+
"step": 955
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.68,
|
| 1165 |
+
"learning_rate": 5.7261865945644664e-05,
|
| 1166 |
+
"loss": 1.3426,
|
| 1167 |
+
"step": 960
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 0.68,
|
| 1171 |
+
"learning_rate": 5.6153058228158686e-05,
|
| 1172 |
+
"loss": 1.3652,
|
| 1173 |
+
"step": 965
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 0.68,
|
| 1177 |
+
"learning_rate": 5.505088476003284e-05,
|
| 1178 |
+
"loss": 1.3833,
|
| 1179 |
+
"step": 970
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.69,
|
| 1183 |
+
"learning_rate": 5.395551230531797e-05,
|
| 1184 |
+
"loss": 1.387,
|
| 1185 |
+
"step": 975
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.69,
|
| 1189 |
+
"learning_rate": 5.286710659903937e-05,
|
| 1190 |
+
"loss": 1.3584,
|
| 1191 |
+
"step": 980
|
| 1192 |
+
},
|
| 1193 |
+
{
|
| 1194 |
+
"epoch": 0.69,
|
| 1195 |
+
"learning_rate": 5.1785832322120574e-05,
|
| 1196 |
+
"loss": 1.3588,
|
| 1197 |
+
"step": 985
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 0.7,
|
| 1201 |
+
"learning_rate": 5.0711853076466045e-05,
|
| 1202 |
+
"loss": 1.403,
|
| 1203 |
+
"step": 990
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 0.7,
|
| 1207 |
+
"learning_rate": 4.9645331360207505e-05,
|
| 1208 |
+
"loss": 1.3297,
|
| 1209 |
+
"step": 995
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 0.7,
|
| 1213 |
+
"learning_rate": 4.8586428543117434e-05,
|
| 1214 |
+
"loss": 1.3314,
|
| 1215 |
+
"step": 1000
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 0.71,
|
| 1219 |
+
"learning_rate": 4.753530484219279e-05,
|
| 1220 |
+
"loss": 1.3619,
|
| 1221 |
+
"step": 1005
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.71,
|
| 1225 |
+
"learning_rate": 4.6492119297413475e-05,
|
| 1226 |
+
"loss": 1.3421,
|
| 1227 |
+
"step": 1010
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.72,
|
| 1231 |
+
"learning_rate": 4.545702974767891e-05,
|
| 1232 |
+
"loss": 1.356,
|
| 1233 |
+
"step": 1015
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 0.72,
|
| 1237 |
+
"learning_rate": 4.443019280692638e-05,
|
| 1238 |
+
"loss": 1.3057,
|
| 1239 |
+
"step": 1020
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 0.72,
|
| 1243 |
+
"learning_rate": 4.341176384043416e-05,
|
| 1244 |
+
"loss": 1.3334,
|
| 1245 |
+
"step": 1025
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 0.73,
|
| 1249 |
+
"learning_rate": 4.240189694131458e-05,
|
| 1250 |
+
"loss": 1.3574,
|
| 1251 |
+
"step": 1030
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 0.73,
|
| 1255 |
+
"learning_rate": 4.140074490719861e-05,
|
| 1256 |
+
"loss": 1.3476,
|
| 1257 |
+
"step": 1035
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 0.73,
|
| 1261 |
+
"learning_rate": 4.040845921711702e-05,
|
| 1262 |
+
"loss": 1.36,
|
| 1263 |
+
"step": 1040
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 0.74,
|
| 1267 |
+
"learning_rate": 3.9425190008580826e-05,
|
| 1268 |
+
"loss": 1.3517,
|
| 1269 |
+
"step": 1045
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.74,
|
| 1273 |
+
"learning_rate": 3.8451086054864896e-05,
|
| 1274 |
+
"loss": 1.3709,
|
| 1275 |
+
"step": 1050
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 0.74,
|
| 1279 |
+
"learning_rate": 3.748629474249766e-05,
|
| 1280 |
+
"loss": 1.3551,
|
| 1281 |
+
"step": 1055
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 0.75,
|
| 1285 |
+
"learning_rate": 3.6530962048960973e-05,
|
| 1286 |
+
"loss": 1.3444,
|
| 1287 |
+
"step": 1060
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 0.75,
|
| 1291 |
+
"learning_rate": 3.558523252060295e-05,
|
| 1292 |
+
"loss": 1.3683,
|
| 1293 |
+
"step": 1065
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 0.75,
|
| 1297 |
+
"learning_rate": 3.464924925076758e-05,
|
| 1298 |
+
"loss": 1.3673,
|
| 1299 |
+
"step": 1070
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 0.76,
|
| 1303 |
+
"learning_rate": 3.3723153858143784e-05,
|
| 1304 |
+
"loss": 1.3692,
|
| 1305 |
+
"step": 1075
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 0.76,
|
| 1309 |
+
"learning_rate": 3.280708646533799e-05,
|
| 1310 |
+
"loss": 1.3456,
|
| 1311 |
+
"step": 1080
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.76,
|
| 1315 |
+
"learning_rate": 3.1901185677672976e-05,
|
| 1316 |
+
"loss": 1.3424,
|
| 1317 |
+
"step": 1085
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 0.77,
|
| 1321 |
+
"learning_rate": 3.100558856221606e-05,
|
| 1322 |
+
"loss": 1.3521,
|
| 1323 |
+
"step": 1090
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 0.77,
|
| 1327 |
+
"learning_rate": 3.012043062704034e-05,
|
| 1328 |
+
"loss": 1.3721,
|
| 1329 |
+
"step": 1095
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 0.77,
|
| 1333 |
+
"learning_rate": 2.9245845800721574e-05,
|
| 1334 |
+
"loss": 1.3416,
|
| 1335 |
+
"step": 1100
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 0.78,
|
| 1339 |
+
"learning_rate": 2.838196641207438e-05,
|
| 1340 |
+
"loss": 1.3671,
|
| 1341 |
+
"step": 1105
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 0.78,
|
| 1345 |
+
"learning_rate": 2.752892317012985e-05,
|
| 1346 |
+
"loss": 1.3855,
|
| 1347 |
+
"step": 1110
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 0.79,
|
| 1351 |
+
"learning_rate": 2.668684514435914e-05,
|
| 1352 |
+
"loss": 1.388,
|
| 1353 |
+
"step": 1115
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.79,
|
| 1357 |
+
"learning_rate": 2.5855859745144284e-05,
|
| 1358 |
+
"loss": 1.3535,
|
| 1359 |
+
"step": 1120
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 0.79,
|
| 1363 |
+
"learning_rate": 2.503609270450058e-05,
|
| 1364 |
+
"loss": 1.3519,
|
| 1365 |
+
"step": 1125
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 0.8,
|
| 1369 |
+
"learning_rate": 2.4227668057052634e-05,
|
| 1370 |
+
"loss": 1.3591,
|
| 1371 |
+
"step": 1130
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 0.8,
|
| 1375 |
+
"learning_rate": 2.343070812126752e-05,
|
| 1376 |
+
"loss": 1.3823,
|
| 1377 |
+
"step": 1135
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 0.8,
|
| 1381 |
+
"learning_rate": 2.2645333480947173e-05,
|
| 1382 |
+
"loss": 1.3482,
|
| 1383 |
+
"step": 1140
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 0.81,
|
| 1387 |
+
"learning_rate": 2.1871662966983663e-05,
|
| 1388 |
+
"loss": 1.3656,
|
| 1389 |
+
"step": 1145
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 0.81,
|
| 1393 |
+
"learning_rate": 2.1109813639379427e-05,
|
| 1394 |
+
"loss": 1.3696,
|
| 1395 |
+
"step": 1150
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.81,
|
| 1399 |
+
"learning_rate": 2.035990076953569e-05,
|
| 1400 |
+
"loss": 1.3406,
|
| 1401 |
+
"step": 1155
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 0.82,
|
| 1405 |
+
"learning_rate": 1.9622037822811123e-05,
|
| 1406 |
+
"loss": 1.3501,
|
| 1407 |
+
"step": 1160
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 0.82,
|
| 1411 |
+
"learning_rate": 1.8896336441354167e-05,
|
| 1412 |
+
"loss": 1.3447,
|
| 1413 |
+
"step": 1165
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.82,
|
| 1417 |
+
"learning_rate": 1.8182906427211043e-05,
|
| 1418 |
+
"loss": 1.3734,
|
| 1419 |
+
"step": 1170
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 0.83,
|
| 1423 |
+
"learning_rate": 1.7481855725712093e-05,
|
| 1424 |
+
"loss": 1.3903,
|
| 1425 |
+
"step": 1175
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 0.83,
|
| 1429 |
+
"learning_rate": 1.6793290409139207e-05,
|
| 1430 |
+
"loss": 1.3574,
|
| 1431 |
+
"step": 1180
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 0.83,
|
| 1435 |
+
"learning_rate": 1.611731466067655e-05,
|
| 1436 |
+
"loss": 1.3495,
|
| 1437 |
+
"step": 1185
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.84,
|
| 1441 |
+
"learning_rate": 1.54540307586473e-05,
|
| 1442 |
+
"loss": 1.342,
|
| 1443 |
+
"step": 1190
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 0.84,
|
| 1447 |
+
"learning_rate": 1.4803539061038119e-05,
|
| 1448 |
+
"loss": 1.3346,
|
| 1449 |
+
"step": 1195
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 0.85,
|
| 1453 |
+
"learning_rate": 1.4165937990314915e-05,
|
| 1454 |
+
"loss": 1.3269,
|
| 1455 |
+
"step": 1200
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 0.85,
|
| 1459 |
+
"learning_rate": 1.3541324018530788e-05,
|
| 1460 |
+
"loss": 1.37,
|
| 1461 |
+
"step": 1205
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 0.85,
|
| 1465 |
+
"learning_rate": 1.2929791652729439e-05,
|
| 1466 |
+
"loss": 1.3472,
|
| 1467 |
+
"step": 1210
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 0.86,
|
| 1471 |
+
"learning_rate": 1.233143342064581e-05,
|
| 1472 |
+
"loss": 1.3459,
|
| 1473 |
+
"step": 1215
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 0.86,
|
| 1477 |
+
"learning_rate": 1.1746339856706234e-05,
|
| 1478 |
+
"loss": 1.3591,
|
| 1479 |
+
"step": 1220
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.86,
|
| 1483 |
+
"learning_rate": 1.1174599488330051e-05,
|
| 1484 |
+
"loss": 1.361,
|
| 1485 |
+
"step": 1225
|
| 1486 |
+
},
|
| 1487 |
+
{
|
| 1488 |
+
"epoch": 0.87,
|
| 1489 |
+
"learning_rate": 1.0616298822535087e-05,
|
| 1490 |
+
"loss": 1.3585,
|
| 1491 |
+
"step": 1230
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 0.87,
|
| 1495 |
+
"learning_rate": 1.007152233284866e-05,
|
| 1496 |
+
"loss": 1.3768,
|
| 1497 |
+
"step": 1235
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 0.87,
|
| 1501 |
+
"learning_rate": 9.540352446526479e-06,
|
| 1502 |
+
"loss": 1.3655,
|
| 1503 |
+
"step": 1240
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 0.88,
|
| 1507 |
+
"learning_rate": 9.022869532080735e-06,
|
| 1508 |
+
"loss": 1.3734,
|
| 1509 |
+
"step": 1245
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 0.88,
|
| 1513 |
+
"learning_rate": 8.51915188712027e-06,
|
| 1514 |
+
"loss": 1.3404,
|
| 1515 |
+
"step": 1250
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 0.88,
|
| 1519 |
+
"learning_rate": 8.029275726503626e-06,
|
| 1520 |
+
"loss": 1.3352,
|
| 1521 |
+
"step": 1255
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 0.89,
|
| 1525 |
+
"learning_rate": 7.553315170807373e-06,
|
| 1526 |
+
"loss": 1.3394,
|
| 1527 |
+
"step": 1260
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 0.89,
|
| 1531 |
+
"learning_rate": 7.09134223511132e-06,
|
| 1532 |
+
"loss": 1.3762,
|
| 1533 |
+
"step": 1265
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 0.89,
|
| 1537 |
+
"learning_rate": 6.643426818102339e-06,
|
| 1538 |
+
"loss": 1.373,
|
| 1539 |
+
"step": 1270
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 0.9,
|
| 1543 |
+
"learning_rate": 6.209636691498366e-06,
|
| 1544 |
+
"loss": 1.3304,
|
| 1545 |
+
"step": 1275
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 0.9,
|
| 1549 |
+
"learning_rate": 5.790037489794109e-06,
|
| 1550 |
+
"loss": 1.3483,
|
| 1551 |
+
"step": 1280
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 0.91,
|
| 1555 |
+
"learning_rate": 5.384692700330407e-06,
|
| 1556 |
+
"loss": 1.3342,
|
| 1557 |
+
"step": 1285
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 0.91,
|
| 1561 |
+
"learning_rate": 4.993663653688152e-06,
|
| 1562 |
+
"loss": 1.3695,
|
| 1563 |
+
"step": 1290
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 0.91,
|
| 1567 |
+
"learning_rate": 4.617009514408788e-06,
|
| 1568 |
+
"loss": 1.3326,
|
| 1569 |
+
"step": 1295
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 0.92,
|
| 1573 |
+
"learning_rate": 4.254787272042315e-06,
|
| 1574 |
+
"loss": 1.352,
|
| 1575 |
+
"step": 1300
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 0.92,
|
| 1579 |
+
"learning_rate": 3.907051732524669e-06,
|
| 1580 |
+
"loss": 1.3532,
|
| 1581 |
+
"step": 1305
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 0.92,
|
| 1585 |
+
"learning_rate": 3.573855509885171e-06,
|
| 1586 |
+
"loss": 1.3466,
|
| 1587 |
+
"step": 1310
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 0.93,
|
| 1591 |
+
"learning_rate": 3.2552490182858443e-06,
|
| 1592 |
+
"loss": 1.3799,
|
| 1593 |
+
"step": 1315
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 0.93,
|
| 1597 |
+
"learning_rate": 2.95128046439348e-06,
|
| 1598 |
+
"loss": 1.3706,
|
| 1599 |
+
"step": 1320
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 0.93,
|
| 1603 |
+
"learning_rate": 2.6619958400858513e-06,
|
| 1604 |
+
"loss": 1.3794,
|
| 1605 |
+
"step": 1325
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.94,
|
| 1609 |
+
"learning_rate": 2.3874389154927233e-06,
|
| 1610 |
+
"loss": 1.3892,
|
| 1611 |
+
"step": 1330
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 0.94,
|
| 1615 |
+
"learning_rate": 2.127651232373429e-06,
|
| 1616 |
+
"loss": 1.3396,
|
| 1617 |
+
"step": 1335
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 0.94,
|
| 1621 |
+
"learning_rate": 1.8826720978313062e-06,
|
| 1622 |
+
"loss": 1.352,
|
| 1623 |
+
"step": 1340
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 0.95,
|
| 1627 |
+
"learning_rate": 1.6525385783663028e-06,
|
| 1628 |
+
"loss": 1.3655,
|
| 1629 |
+
"step": 1345
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 0.95,
|
| 1633 |
+
"learning_rate": 1.4372854942667602e-06,
|
| 1634 |
+
"loss": 1.3788,
|
| 1635 |
+
"step": 1350
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 0.95,
|
| 1639 |
+
"learning_rate": 1.2369454143408289e-06,
|
| 1640 |
+
"loss": 1.339,
|
| 1641 |
+
"step": 1355
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 0.96,
|
| 1645 |
+
"learning_rate": 1.051548650988754e-06,
|
| 1646 |
+
"loss": 1.3178,
|
| 1647 |
+
"step": 1360
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.96,
|
| 1651 |
+
"learning_rate": 8.811232556163451e-07,
|
| 1652 |
+
"loss": 1.3495,
|
| 1653 |
+
"step": 1365
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 0.97,
|
| 1657 |
+
"learning_rate": 7.256950143907815e-07,
|
| 1658 |
+
"loss": 1.3702,
|
| 1659 |
+
"step": 1370
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 0.97,
|
| 1663 |
+
"learning_rate": 5.852874443389888e-07,
|
| 1664 |
+
"loss": 1.3459,
|
| 1665 |
+
"step": 1375
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 0.97,
|
| 1669 |
+
"learning_rate": 4.5992178978938593e-07,
|
| 1670 |
+
"loss": 1.3747,
|
| 1671 |
+
"step": 1380
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 0.98,
|
| 1675 |
+
"learning_rate": 3.496170191575554e-07,
|
| 1676 |
+
"loss": 1.3402,
|
| 1677 |
+
"step": 1385
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 0.98,
|
| 1681 |
+
"learning_rate": 2.5438982207622866e-07,
|
| 1682 |
+
"loss": 1.3601,
|
| 1683 |
+
"step": 1390
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 0.98,
|
| 1687 |
+
"learning_rate": 1.742546068700168e-07,
|
| 1688 |
+
"loss": 1.3837,
|
| 1689 |
+
"step": 1395
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.99,
|
| 1693 |
+
"learning_rate": 1.0922349837545476e-07,
|
| 1694 |
+
"loss": 1.4011,
|
| 1695 |
+
"step": 1400
|
| 1696 |
+
},
|
| 1697 |
+
{
|
| 1698 |
+
"epoch": 0.99,
|
| 1699 |
+
"learning_rate": 5.9306336106346174e-08,
|
| 1700 |
+
"loss": 1.3838,
|
| 1701 |
+
"step": 1405
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 0.99,
|
| 1705 |
+
"learning_rate": 2.4510672765087894e-08,
|
| 1706 |
+
"loss": 1.3608,
|
| 1707 |
+
"step": 1410
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 1.0,
|
| 1711 |
+
"learning_rate": 4.8417730998728814e-09,
|
| 1712 |
+
"loss": 1.3504,
|
| 1713 |
+
"step": 1415
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 1.0,
|
| 1717 |
+
"eval_loss": 1.355061411857605,
|
| 1718 |
+
"eval_runtime": 1858.7192,
|
| 1719 |
+
"eval_samples_per_second": 0.709,
|
| 1720 |
+
"eval_steps_per_second": 0.089,
|
| 1721 |
+
"step": 1419
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 1.0,
|
| 1725 |
+
"step": 1419,
|
| 1726 |
+
"total_flos": 9.977407964782264e+17,
|
| 1727 |
+
"train_loss": 1.4210990094399267,
|
| 1728 |
+
"train_runtime": 56404.0081,
|
| 1729 |
+
"train_samples_per_second": 0.201,
|
| 1730 |
+
"train_steps_per_second": 0.025
|
| 1731 |
+
}
|
| 1732 |
+
],
|
| 1733 |
+
"logging_steps": 5,
|
| 1734 |
+
"max_steps": 1419,
|
| 1735 |
+
"num_input_tokens_seen": 0,
|
| 1736 |
+
"num_train_epochs": 1,
|
| 1737 |
+
"save_steps": 100,
|
| 1738 |
+
"total_flos": 9.977407964782264e+17,
|
| 1739 |
+
"train_batch_size": 4,
|
| 1740 |
+
"trial_name": null,
|
| 1741 |
+
"trial_params": null
|
| 1742 |
+
}
|