Model save
Browse files- README.md +71 -0
- all_results.json +13 -0
- eval_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +2172 -0
README.md
ADDED
|
@@ -0,0 +1,71 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
base_model: HuggingFaceH4/zephyr-7b-beta
|
| 4 |
+
tags:
|
| 5 |
+
- generated_from_trainer
|
| 6 |
+
model-index:
|
| 7 |
+
- name: zephyr-convsearch-7b
|
| 8 |
+
results: []
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 12 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 13 |
+
|
| 14 |
+
# zephyr-convsearch-7b
|
| 15 |
+
|
| 16 |
+
This model is a fine-tuned version of [HuggingFaceH4/zephyr-7b-beta](https://huggingface.co/HuggingFaceH4/zephyr-7b-beta) on an unknown dataset.
|
| 17 |
+
It achieves the following results on the evaluation set:
|
| 18 |
+
- Loss: 0.4475
|
| 19 |
+
|
| 20 |
+
## Model description
|
| 21 |
+
|
| 22 |
+
More information needed
|
| 23 |
+
|
| 24 |
+
## Intended uses & limitations
|
| 25 |
+
|
| 26 |
+
More information needed
|
| 27 |
+
|
| 28 |
+
## Training and evaluation data
|
| 29 |
+
|
| 30 |
+
More information needed
|
| 31 |
+
|
| 32 |
+
## Training procedure
|
| 33 |
+
|
| 34 |
+
### Training hyperparameters
|
| 35 |
+
|
| 36 |
+
The following hyperparameters were used during training:
|
| 37 |
+
- learning_rate: 2e-05
|
| 38 |
+
- train_batch_size: 8
|
| 39 |
+
- eval_batch_size: 8
|
| 40 |
+
- seed: 42
|
| 41 |
+
- distributed_type: multi-GPU
|
| 42 |
+
- num_devices: 2
|
| 43 |
+
- gradient_accumulation_steps: 32
|
| 44 |
+
- total_train_batch_size: 512
|
| 45 |
+
- total_eval_batch_size: 16
|
| 46 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 47 |
+
- lr_scheduler_type: cosine
|
| 48 |
+
- num_epochs: 10
|
| 49 |
+
|
| 50 |
+
### Training results
|
| 51 |
+
|
| 52 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 53 |
+
|:-------------:|:-----:|:----:|:---------------:|
|
| 54 |
+
| 0.5796 | 0.62 | 34 | 0.5780 |
|
| 55 |
+
| 0.5119 | 1.62 | 68 | 0.5122 |
|
| 56 |
+
| 0.4799 | 2.63 | 103 | 0.4831 |
|
| 57 |
+
| 0.4712 | 3.62 | 137 | 0.4706 |
|
| 58 |
+
| 0.4615 | 4.63 | 172 | 0.4631 |
|
| 59 |
+
| 0.4584 | 5.62 | 206 | 0.4574 |
|
| 60 |
+
| 0.4524 | 6.63 | 241 | 0.4529 |
|
| 61 |
+
| 0.4507 | 7.62 | 275 | 0.4502 |
|
| 62 |
+
| 0.4478 | 8.63 | 310 | 0.4480 |
|
| 63 |
+
| 0.4467 | 9.62 | 344 | 0.4475 |
|
| 64 |
+
|
| 65 |
+
|
| 66 |
+
### Framework versions
|
| 67 |
+
|
| 68 |
+
- Transformers 4.35.0
|
| 69 |
+
- Pytorch 2.1.1+cu118
|
| 70 |
+
- Datasets 2.14.6
|
| 71 |
+
- Tokenizers 0.14.1
|
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.62,
|
| 3 |
+
"eval_loss": 0.44749271869659424,
|
| 4 |
+
"eval_runtime": 23.1279,
|
| 5 |
+
"eval_samples": 200,
|
| 6 |
+
"eval_samples_per_second": 8.648,
|
| 7 |
+
"eval_steps_per_second": 0.562,
|
| 8 |
+
"train_loss": 0.4900048969443454,
|
| 9 |
+
"train_runtime": 108147.3514,
|
| 10 |
+
"train_samples": 27870,
|
| 11 |
+
"train_samples_per_second": 2.577,
|
| 12 |
+
"train_steps_per_second": 0.005
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.62,
|
| 3 |
+
"eval_loss": 0.44749271869659424,
|
| 4 |
+
"eval_runtime": 23.1279,
|
| 5 |
+
"eval_samples": 200,
|
| 6 |
+
"eval_samples_per_second": 8.648,
|
| 7 |
+
"eval_steps_per_second": 0.562
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.62,
|
| 3 |
+
"train_loss": 0.4900048969443454,
|
| 4 |
+
"train_runtime": 108147.3514,
|
| 5 |
+
"train_samples": 27870,
|
| 6 |
+
"train_samples_per_second": 2.577,
|
| 7 |
+
"train_steps_per_second": 0.005
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2172 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.620551090700344,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 344,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.02,
|
| 13 |
+
"learning_rate": 1.9999830768577445e-05,
|
| 14 |
+
"loss": 0.7381,
|
| 15 |
+
"step": 1
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.04,
|
| 19 |
+
"learning_rate": 1.9999323080037623e-05,
|
| 20 |
+
"loss": 0.7407,
|
| 21 |
+
"step": 2
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.06,
|
| 25 |
+
"learning_rate": 1.9998476951563914e-05,
|
| 26 |
+
"loss": 0.7359,
|
| 27 |
+
"step": 3
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.07,
|
| 31 |
+
"learning_rate": 1.999729241179462e-05,
|
| 32 |
+
"loss": 0.7298,
|
| 33 |
+
"step": 4
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.09,
|
| 37 |
+
"learning_rate": 1.9995769500822007e-05,
|
| 38 |
+
"loss": 0.7305,
|
| 39 |
+
"step": 5
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.11,
|
| 43 |
+
"learning_rate": 1.999390827019096e-05,
|
| 44 |
+
"loss": 0.7124,
|
| 45 |
+
"step": 6
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.13,
|
| 49 |
+
"learning_rate": 1.9991708782897214e-05,
|
| 50 |
+
"loss": 0.7076,
|
| 51 |
+
"step": 7
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.15,
|
| 55 |
+
"learning_rate": 1.998917111338525e-05,
|
| 56 |
+
"loss": 0.6927,
|
| 57 |
+
"step": 8
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.17,
|
| 61 |
+
"learning_rate": 1.9986295347545738e-05,
|
| 62 |
+
"loss": 0.6858,
|
| 63 |
+
"step": 9
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.18,
|
| 67 |
+
"learning_rate": 1.9983081582712684e-05,
|
| 68 |
+
"loss": 0.6812,
|
| 69 |
+
"step": 10
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.2,
|
| 73 |
+
"learning_rate": 1.9979529927660076e-05,
|
| 74 |
+
"loss": 0.6788,
|
| 75 |
+
"step": 11
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.22,
|
| 79 |
+
"learning_rate": 1.9975640502598243e-05,
|
| 80 |
+
"loss": 0.6633,
|
| 81 |
+
"step": 12
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.24,
|
| 85 |
+
"learning_rate": 1.9971413439169777e-05,
|
| 86 |
+
"loss": 0.6589,
|
| 87 |
+
"step": 13
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.26,
|
| 91 |
+
"learning_rate": 1.996684888044506e-05,
|
| 92 |
+
"loss": 0.6471,
|
| 93 |
+
"step": 14
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.28,
|
| 97 |
+
"learning_rate": 1.9961946980917457e-05,
|
| 98 |
+
"loss": 0.6474,
|
| 99 |
+
"step": 15
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.29,
|
| 103 |
+
"learning_rate": 1.9956707906498046e-05,
|
| 104 |
+
"loss": 0.6419,
|
| 105 |
+
"step": 16
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.31,
|
| 109 |
+
"learning_rate": 1.9951131834510034e-05,
|
| 110 |
+
"loss": 0.6345,
|
| 111 |
+
"step": 17
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.33,
|
| 115 |
+
"learning_rate": 1.9945218953682736e-05,
|
| 116 |
+
"loss": 0.635,
|
| 117 |
+
"step": 18
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.35,
|
| 121 |
+
"learning_rate": 1.99389694641452e-05,
|
| 122 |
+
"loss": 0.6309,
|
| 123 |
+
"step": 19
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.37,
|
| 127 |
+
"learning_rate": 1.9932383577419432e-05,
|
| 128 |
+
"loss": 0.619,
|
| 129 |
+
"step": 20
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.39,
|
| 133 |
+
"learning_rate": 1.9925461516413224e-05,
|
| 134 |
+
"loss": 0.6184,
|
| 135 |
+
"step": 21
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.4,
|
| 139 |
+
"learning_rate": 1.9918203515412616e-05,
|
| 140 |
+
"loss": 0.6127,
|
| 141 |
+
"step": 22
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.42,
|
| 145 |
+
"learning_rate": 1.9910609820073986e-05,
|
| 146 |
+
"loss": 0.6108,
|
| 147 |
+
"step": 23
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.44,
|
| 151 |
+
"learning_rate": 1.9902680687415704e-05,
|
| 152 |
+
"loss": 0.6138,
|
| 153 |
+
"step": 24
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.46,
|
| 157 |
+
"learning_rate": 1.9894416385809444e-05,
|
| 158 |
+
"loss": 0.6039,
|
| 159 |
+
"step": 25
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.48,
|
| 163 |
+
"learning_rate": 1.9885817194971116e-05,
|
| 164 |
+
"loss": 0.6123,
|
| 165 |
+
"step": 26
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.5,
|
| 169 |
+
"learning_rate": 1.9876883405951378e-05,
|
| 170 |
+
"loss": 0.5973,
|
| 171 |
+
"step": 27
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.51,
|
| 175 |
+
"learning_rate": 1.9867615321125796e-05,
|
| 176 |
+
"loss": 0.6002,
|
| 177 |
+
"step": 28
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.53,
|
| 181 |
+
"learning_rate": 1.9858013254184597e-05,
|
| 182 |
+
"loss": 0.5908,
|
| 183 |
+
"step": 29
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.55,
|
| 187 |
+
"learning_rate": 1.9848077530122083e-05,
|
| 188 |
+
"loss": 0.596,
|
| 189 |
+
"step": 30
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.57,
|
| 193 |
+
"learning_rate": 1.983780848522559e-05,
|
| 194 |
+
"loss": 0.5886,
|
| 195 |
+
"step": 31
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.59,
|
| 199 |
+
"learning_rate": 1.9827206467064133e-05,
|
| 200 |
+
"loss": 0.5908,
|
| 201 |
+
"step": 32
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.61,
|
| 205 |
+
"learning_rate": 1.9816271834476642e-05,
|
| 206 |
+
"loss": 0.5821,
|
| 207 |
+
"step": 33
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.62,
|
| 211 |
+
"learning_rate": 1.9805004957559795e-05,
|
| 212 |
+
"loss": 0.5796,
|
| 213 |
+
"step": 34
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.62,
|
| 217 |
+
"eval_loss": 0.5779749751091003,
|
| 218 |
+
"eval_runtime": 24.4017,
|
| 219 |
+
"eval_samples_per_second": 8.196,
|
| 220 |
+
"eval_steps_per_second": 0.533,
|
| 221 |
+
"step": 34
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 1.01,
|
| 225 |
+
"learning_rate": 1.9793406217655516e-05,
|
| 226 |
+
"loss": 0.5719,
|
| 227 |
+
"step": 35
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 1.03,
|
| 231 |
+
"learning_rate": 1.9781476007338058e-05,
|
| 232 |
+
"loss": 0.5771,
|
| 233 |
+
"step": 36
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.05,
|
| 237 |
+
"learning_rate": 1.976921473040071e-05,
|
| 238 |
+
"loss": 0.5709,
|
| 239 |
+
"step": 37
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 1.06,
|
| 243 |
+
"learning_rate": 1.9756622801842144e-05,
|
| 244 |
+
"loss": 0.567,
|
| 245 |
+
"step": 38
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 1.08,
|
| 249 |
+
"learning_rate": 1.9743700647852356e-05,
|
| 250 |
+
"loss": 0.5746,
|
| 251 |
+
"step": 39
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 1.1,
|
| 255 |
+
"learning_rate": 1.973044870579824e-05,
|
| 256 |
+
"loss": 0.5651,
|
| 257 |
+
"step": 40
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 1.12,
|
| 261 |
+
"learning_rate": 1.9716867424208805e-05,
|
| 262 |
+
"loss": 0.5621,
|
| 263 |
+
"step": 41
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 1.14,
|
| 267 |
+
"learning_rate": 1.9702957262759964e-05,
|
| 268 |
+
"loss": 0.5581,
|
| 269 |
+
"step": 42
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 1.16,
|
| 273 |
+
"learning_rate": 1.9688718692259007e-05,
|
| 274 |
+
"loss": 0.5555,
|
| 275 |
+
"step": 43
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.18,
|
| 279 |
+
"learning_rate": 1.967415219462864e-05,
|
| 280 |
+
"loss": 0.5512,
|
| 281 |
+
"step": 44
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 1.19,
|
| 285 |
+
"learning_rate": 1.9659258262890683e-05,
|
| 286 |
+
"loss": 0.554,
|
| 287 |
+
"step": 45
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 1.21,
|
| 291 |
+
"learning_rate": 1.964403740114939e-05,
|
| 292 |
+
"loss": 0.5507,
|
| 293 |
+
"step": 46
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 1.23,
|
| 297 |
+
"learning_rate": 1.962849012457438e-05,
|
| 298 |
+
"loss": 0.5482,
|
| 299 |
+
"step": 47
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 1.25,
|
| 303 |
+
"learning_rate": 1.961261695938319e-05,
|
| 304 |
+
"loss": 0.5439,
|
| 305 |
+
"step": 48
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.27,
|
| 309 |
+
"learning_rate": 1.9596418442823495e-05,
|
| 310 |
+
"loss": 0.5387,
|
| 311 |
+
"step": 49
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 1.29,
|
| 315 |
+
"learning_rate": 1.957989512315489e-05,
|
| 316 |
+
"loss": 0.5434,
|
| 317 |
+
"step": 50
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.3,
|
| 321 |
+
"learning_rate": 1.9563047559630356e-05,
|
| 322 |
+
"loss": 0.5388,
|
| 323 |
+
"step": 51
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 1.32,
|
| 327 |
+
"learning_rate": 1.954587632247732e-05,
|
| 328 |
+
"loss": 0.5385,
|
| 329 |
+
"step": 52
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 1.34,
|
| 333 |
+
"learning_rate": 1.9528381992878362e-05,
|
| 334 |
+
"loss": 0.5364,
|
| 335 |
+
"step": 53
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 1.36,
|
| 339 |
+
"learning_rate": 1.9510565162951538e-05,
|
| 340 |
+
"loss": 0.535,
|
| 341 |
+
"step": 54
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 1.38,
|
| 345 |
+
"learning_rate": 1.949242643573034e-05,
|
| 346 |
+
"loss": 0.5253,
|
| 347 |
+
"step": 55
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.4,
|
| 351 |
+
"learning_rate": 1.9473966425143292e-05,
|
| 352 |
+
"loss": 0.5278,
|
| 353 |
+
"step": 56
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 1.41,
|
| 357 |
+
"learning_rate": 1.945518575599317e-05,
|
| 358 |
+
"loss": 0.5316,
|
| 359 |
+
"step": 57
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.43,
|
| 363 |
+
"learning_rate": 1.9436085063935837e-05,
|
| 364 |
+
"loss": 0.5267,
|
| 365 |
+
"step": 58
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 1.45,
|
| 369 |
+
"learning_rate": 1.9416664995458756e-05,
|
| 370 |
+
"loss": 0.5265,
|
| 371 |
+
"step": 59
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 1.47,
|
| 375 |
+
"learning_rate": 1.9396926207859085e-05,
|
| 376 |
+
"loss": 0.5235,
|
| 377 |
+
"step": 60
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 1.49,
|
| 381 |
+
"learning_rate": 1.937686936922145e-05,
|
| 382 |
+
"loss": 0.5217,
|
| 383 |
+
"step": 61
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 1.51,
|
| 387 |
+
"learning_rate": 1.9356495158395317e-05,
|
| 388 |
+
"loss": 0.5243,
|
| 389 |
+
"step": 62
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.52,
|
| 393 |
+
"learning_rate": 1.9335804264972018e-05,
|
| 394 |
+
"loss": 0.5189,
|
| 395 |
+
"step": 63
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.54,
|
| 399 |
+
"learning_rate": 1.9314797389261426e-05,
|
| 400 |
+
"loss": 0.5193,
|
| 401 |
+
"step": 64
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.56,
|
| 405 |
+
"learning_rate": 1.9293475242268224e-05,
|
| 406 |
+
"loss": 0.5182,
|
| 407 |
+
"step": 65
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 1.58,
|
| 411 |
+
"learning_rate": 1.9271838545667876e-05,
|
| 412 |
+
"loss": 0.5222,
|
| 413 |
+
"step": 66
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 1.6,
|
| 417 |
+
"learning_rate": 1.924988803178216e-05,
|
| 418 |
+
"loss": 0.516,
|
| 419 |
+
"step": 67
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 1.62,
|
| 423 |
+
"learning_rate": 1.9227624443554425e-05,
|
| 424 |
+
"loss": 0.5119,
|
| 425 |
+
"step": 68
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 1.62,
|
| 429 |
+
"eval_loss": 0.5121633410453796,
|
| 430 |
+
"eval_runtime": 23.8105,
|
| 431 |
+
"eval_samples_per_second": 8.4,
|
| 432 |
+
"eval_steps_per_second": 0.546,
|
| 433 |
+
"step": 68
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 2.0,
|
| 437 |
+
"learning_rate": 1.9205048534524405e-05,
|
| 438 |
+
"loss": 0.5095,
|
| 439 |
+
"step": 69
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 2.02,
|
| 443 |
+
"learning_rate": 1.9182161068802742e-05,
|
| 444 |
+
"loss": 0.5091,
|
| 445 |
+
"step": 70
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 2.04,
|
| 449 |
+
"learning_rate": 1.9158962821045113e-05,
|
| 450 |
+
"loss": 0.5086,
|
| 451 |
+
"step": 71
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 2.06,
|
| 455 |
+
"learning_rate": 1.913545457642601e-05,
|
| 456 |
+
"loss": 0.507,
|
| 457 |
+
"step": 72
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 2.07,
|
| 461 |
+
"learning_rate": 1.9111637130612172e-05,
|
| 462 |
+
"loss": 0.5096,
|
| 463 |
+
"step": 73
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 2.09,
|
| 467 |
+
"learning_rate": 1.9087511289735646e-05,
|
| 468 |
+
"loss": 0.5124,
|
| 469 |
+
"step": 74
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 2.11,
|
| 473 |
+
"learning_rate": 1.9063077870366504e-05,
|
| 474 |
+
"loss": 0.5013,
|
| 475 |
+
"step": 75
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 2.13,
|
| 479 |
+
"learning_rate": 1.9038337699485207e-05,
|
| 480 |
+
"loss": 0.5048,
|
| 481 |
+
"step": 76
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 2.15,
|
| 485 |
+
"learning_rate": 1.9013291614454622e-05,
|
| 486 |
+
"loss": 0.4988,
|
| 487 |
+
"step": 77
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 2.17,
|
| 491 |
+
"learning_rate": 1.8987940462991673e-05,
|
| 492 |
+
"loss": 0.502,
|
| 493 |
+
"step": 78
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 2.18,
|
| 497 |
+
"learning_rate": 1.8962285103138637e-05,
|
| 498 |
+
"loss": 0.4969,
|
| 499 |
+
"step": 79
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.2,
|
| 503 |
+
"learning_rate": 1.8936326403234125e-05,
|
| 504 |
+
"loss": 0.503,
|
| 505 |
+
"step": 80
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 2.22,
|
| 509 |
+
"learning_rate": 1.891006524188368e-05,
|
| 510 |
+
"loss": 0.4978,
|
| 511 |
+
"step": 81
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 2.24,
|
| 515 |
+
"learning_rate": 1.8883502507930044e-05,
|
| 516 |
+
"loss": 0.4954,
|
| 517 |
+
"step": 82
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 2.26,
|
| 521 |
+
"learning_rate": 1.8856639100423062e-05,
|
| 522 |
+
"loss": 0.4939,
|
| 523 |
+
"step": 83
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 2.28,
|
| 527 |
+
"learning_rate": 1.8829475928589272e-05,
|
| 528 |
+
"loss": 0.4942,
|
| 529 |
+
"step": 84
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 2.3,
|
| 533 |
+
"learning_rate": 1.880201391180111e-05,
|
| 534 |
+
"loss": 0.4933,
|
| 535 |
+
"step": 85
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 2.31,
|
| 539 |
+
"learning_rate": 1.877425397954582e-05,
|
| 540 |
+
"loss": 0.4946,
|
| 541 |
+
"step": 86
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.33,
|
| 545 |
+
"learning_rate": 1.874619707139396e-05,
|
| 546 |
+
"loss": 0.4939,
|
| 547 |
+
"step": 87
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 2.35,
|
| 551 |
+
"learning_rate": 1.8717844136967626e-05,
|
| 552 |
+
"loss": 0.4922,
|
| 553 |
+
"step": 88
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 2.37,
|
| 557 |
+
"learning_rate": 1.8689196135908303e-05,
|
| 558 |
+
"loss": 0.4925,
|
| 559 |
+
"step": 89
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 2.39,
|
| 563 |
+
"learning_rate": 1.866025403784439e-05,
|
| 564 |
+
"loss": 0.4902,
|
| 565 |
+
"step": 90
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 2.41,
|
| 569 |
+
"learning_rate": 1.8631018822358363e-05,
|
| 570 |
+
"loss": 0.4866,
|
| 571 |
+
"step": 91
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 2.42,
|
| 575 |
+
"learning_rate": 1.860149147895366e-05,
|
| 576 |
+
"loss": 0.4937,
|
| 577 |
+
"step": 92
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 2.44,
|
| 581 |
+
"learning_rate": 1.8571673007021124e-05,
|
| 582 |
+
"loss": 0.4897,
|
| 583 |
+
"step": 93
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.46,
|
| 587 |
+
"learning_rate": 1.854156441580526e-05,
|
| 588 |
+
"loss": 0.4879,
|
| 589 |
+
"step": 94
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 2.48,
|
| 593 |
+
"learning_rate": 1.8511166724369997e-05,
|
| 594 |
+
"loss": 0.4888,
|
| 595 |
+
"step": 95
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 2.5,
|
| 599 |
+
"learning_rate": 1.848048096156426e-05,
|
| 600 |
+
"loss": 0.4908,
|
| 601 |
+
"step": 96
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 2.52,
|
| 605 |
+
"learning_rate": 1.8449508165987106e-05,
|
| 606 |
+
"loss": 0.4856,
|
| 607 |
+
"step": 97
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 2.53,
|
| 611 |
+
"learning_rate": 1.8418249385952575e-05,
|
| 612 |
+
"loss": 0.4842,
|
| 613 |
+
"step": 98
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 2.55,
|
| 617 |
+
"learning_rate": 1.8386705679454243e-05,
|
| 618 |
+
"loss": 0.4877,
|
| 619 |
+
"step": 99
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 2.57,
|
| 623 |
+
"learning_rate": 1.8354878114129368e-05,
|
| 624 |
+
"loss": 0.491,
|
| 625 |
+
"step": 100
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.59,
|
| 629 |
+
"learning_rate": 1.832276776722278e-05,
|
| 630 |
+
"loss": 0.4877,
|
| 631 |
+
"step": 101
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 2.61,
|
| 635 |
+
"learning_rate": 1.8290375725550417e-05,
|
| 636 |
+
"loss": 0.4863,
|
| 637 |
+
"step": 102
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 2.63,
|
| 641 |
+
"learning_rate": 1.8257703085462542e-05,
|
| 642 |
+
"loss": 0.4799,
|
| 643 |
+
"step": 103
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 2.63,
|
| 647 |
+
"eval_loss": 0.4830611050128937,
|
| 648 |
+
"eval_runtime": 23.8066,
|
| 649 |
+
"eval_samples_per_second": 8.401,
|
| 650 |
+
"eval_steps_per_second": 0.546,
|
| 651 |
+
"step": 103
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 3.01,
|
| 655 |
+
"learning_rate": 1.8224750952806626e-05,
|
| 656 |
+
"loss": 0.4797,
|
| 657 |
+
"step": 104
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 3.03,
|
| 661 |
+
"learning_rate": 1.819152044288992e-05,
|
| 662 |
+
"loss": 0.4833,
|
| 663 |
+
"step": 105
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 3.05,
|
| 667 |
+
"learning_rate": 1.8158012680441723e-05,
|
| 668 |
+
"loss": 0.4833,
|
| 669 |
+
"step": 106
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 3.07,
|
| 673 |
+
"learning_rate": 1.8124228799575295e-05,
|
| 674 |
+
"loss": 0.4818,
|
| 675 |
+
"step": 107
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 3.08,
|
| 679 |
+
"learning_rate": 1.8090169943749477e-05,
|
| 680 |
+
"loss": 0.4885,
|
| 681 |
+
"step": 108
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 3.1,
|
| 685 |
+
"learning_rate": 1.8055837265729996e-05,
|
| 686 |
+
"loss": 0.4785,
|
| 687 |
+
"step": 109
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 3.12,
|
| 691 |
+
"learning_rate": 1.802123192755044e-05,
|
| 692 |
+
"loss": 0.4789,
|
| 693 |
+
"step": 110
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 3.14,
|
| 697 |
+
"learning_rate": 1.798635510047293e-05,
|
| 698 |
+
"loss": 0.4795,
|
| 699 |
+
"step": 111
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 3.16,
|
| 703 |
+
"learning_rate": 1.795120796494848e-05,
|
| 704 |
+
"loss": 0.478,
|
| 705 |
+
"step": 112
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 3.18,
|
| 709 |
+
"learning_rate": 1.7915791710577035e-05,
|
| 710 |
+
"loss": 0.477,
|
| 711 |
+
"step": 113
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 3.19,
|
| 715 |
+
"learning_rate": 1.788010753606722e-05,
|
| 716 |
+
"loss": 0.4746,
|
| 717 |
+
"step": 114
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 3.21,
|
| 721 |
+
"learning_rate": 1.784415664919576e-05,
|
| 722 |
+
"loss": 0.4811,
|
| 723 |
+
"step": 115
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.23,
|
| 727 |
+
"learning_rate": 1.7807940266766595e-05,
|
| 728 |
+
"loss": 0.477,
|
| 729 |
+
"step": 116
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 3.25,
|
| 733 |
+
"learning_rate": 1.777145961456971e-05,
|
| 734 |
+
"loss": 0.4744,
|
| 735 |
+
"step": 117
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 3.27,
|
| 739 |
+
"learning_rate": 1.7734715927339642e-05,
|
| 740 |
+
"loss": 0.4753,
|
| 741 |
+
"step": 118
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 3.29,
|
| 745 |
+
"learning_rate": 1.769771044871368e-05,
|
| 746 |
+
"loss": 0.4754,
|
| 747 |
+
"step": 119
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 3.3,
|
| 751 |
+
"learning_rate": 1.766044443118978e-05,
|
| 752 |
+
"loss": 0.4773,
|
| 753 |
+
"step": 120
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 3.32,
|
| 757 |
+
"learning_rate": 1.7622919136084183e-05,
|
| 758 |
+
"loss": 0.4743,
|
| 759 |
+
"step": 121
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 3.34,
|
| 763 |
+
"learning_rate": 1.7585135833488692e-05,
|
| 764 |
+
"loss": 0.479,
|
| 765 |
+
"step": 122
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.36,
|
| 769 |
+
"learning_rate": 1.7547095802227723e-05,
|
| 770 |
+
"loss": 0.4747,
|
| 771 |
+
"step": 123
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 3.38,
|
| 775 |
+
"learning_rate": 1.7508800329814993e-05,
|
| 776 |
+
"loss": 0.4763,
|
| 777 |
+
"step": 124
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 3.4,
|
| 781 |
+
"learning_rate": 1.7470250712409963e-05,
|
| 782 |
+
"loss": 0.47,
|
| 783 |
+
"step": 125
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 3.42,
|
| 787 |
+
"learning_rate": 1.7431448254773943e-05,
|
| 788 |
+
"loss": 0.4741,
|
| 789 |
+
"step": 126
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 3.43,
|
| 793 |
+
"learning_rate": 1.739239427022596e-05,
|
| 794 |
+
"loss": 0.4753,
|
| 795 |
+
"step": 127
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 3.45,
|
| 799 |
+
"learning_rate": 1.735309008059829e-05,
|
| 800 |
+
"loss": 0.4738,
|
| 801 |
+
"step": 128
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 3.47,
|
| 805 |
+
"learning_rate": 1.7313537016191706e-05,
|
| 806 |
+
"loss": 0.4769,
|
| 807 |
+
"step": 129
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.49,
|
| 811 |
+
"learning_rate": 1.7273736415730488e-05,
|
| 812 |
+
"loss": 0.4749,
|
| 813 |
+
"step": 130
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 3.51,
|
| 817 |
+
"learning_rate": 1.723368962631708e-05,
|
| 818 |
+
"loss": 0.4703,
|
| 819 |
+
"step": 131
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 3.53,
|
| 823 |
+
"learning_rate": 1.7193398003386514e-05,
|
| 824 |
+
"loss": 0.4722,
|
| 825 |
+
"step": 132
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 3.54,
|
| 829 |
+
"learning_rate": 1.7152862910660516e-05,
|
| 830 |
+
"loss": 0.474,
|
| 831 |
+
"step": 133
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 3.56,
|
| 835 |
+
"learning_rate": 1.711208572010137e-05,
|
| 836 |
+
"loss": 0.4721,
|
| 837 |
+
"step": 134
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 3.58,
|
| 841 |
+
"learning_rate": 1.7071067811865477e-05,
|
| 842 |
+
"loss": 0.4755,
|
| 843 |
+
"step": 135
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 3.6,
|
| 847 |
+
"learning_rate": 1.702981057425662e-05,
|
| 848 |
+
"loss": 0.4717,
|
| 849 |
+
"step": 136
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.62,
|
| 853 |
+
"learning_rate": 1.6988315403679e-05,
|
| 854 |
+
"loss": 0.4712,
|
| 855 |
+
"step": 137
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 3.62,
|
| 859 |
+
"eval_loss": 0.470627099275589,
|
| 860 |
+
"eval_runtime": 23.8372,
|
| 861 |
+
"eval_samples_per_second": 8.39,
|
| 862 |
+
"eval_steps_per_second": 0.545,
|
| 863 |
+
"step": 137
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 4.0,
|
| 867 |
+
"learning_rate": 1.6946583704589973e-05,
|
| 868 |
+
"loss": 0.4679,
|
| 869 |
+
"step": 138
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 4.02,
|
| 873 |
+
"learning_rate": 1.6904616889452497e-05,
|
| 874 |
+
"loss": 0.4707,
|
| 875 |
+
"step": 139
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 4.04,
|
| 879 |
+
"learning_rate": 1.686241637868734e-05,
|
| 880 |
+
"loss": 0.4683,
|
| 881 |
+
"step": 140
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 4.06,
|
| 885 |
+
"learning_rate": 1.6819983600624986e-05,
|
| 886 |
+
"loss": 0.4715,
|
| 887 |
+
"step": 141
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 4.08,
|
| 891 |
+
"learning_rate": 1.6777319991457325e-05,
|
| 892 |
+
"loss": 0.4721,
|
| 893 |
+
"step": 142
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 4.09,
|
| 897 |
+
"learning_rate": 1.6734426995189003e-05,
|
| 898 |
+
"loss": 0.4719,
|
| 899 |
+
"step": 143
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 4.11,
|
| 903 |
+
"learning_rate": 1.6691306063588583e-05,
|
| 904 |
+
"loss": 0.4662,
|
| 905 |
+
"step": 144
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 4.13,
|
| 909 |
+
"learning_rate": 1.6647958656139377e-05,
|
| 910 |
+
"loss": 0.4684,
|
| 911 |
+
"step": 145
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 4.15,
|
| 915 |
+
"learning_rate": 1.6604386239990077e-05,
|
| 916 |
+
"loss": 0.4671,
|
| 917 |
+
"step": 146
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 4.17,
|
| 921 |
+
"learning_rate": 1.6560590289905074e-05,
|
| 922 |
+
"loss": 0.4671,
|
| 923 |
+
"step": 147
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 4.19,
|
| 927 |
+
"learning_rate": 1.6516572288214555e-05,
|
| 928 |
+
"loss": 0.4642,
|
| 929 |
+
"step": 148
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 4.2,
|
| 933 |
+
"learning_rate": 1.6472333724764326e-05,
|
| 934 |
+
"loss": 0.468,
|
| 935 |
+
"step": 149
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 4.22,
|
| 939 |
+
"learning_rate": 1.6427876096865394e-05,
|
| 940 |
+
"loss": 0.4699,
|
| 941 |
+
"step": 150
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 4.24,
|
| 945 |
+
"learning_rate": 1.6383200909243285e-05,
|
| 946 |
+
"loss": 0.4646,
|
| 947 |
+
"step": 151
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 4.26,
|
| 951 |
+
"learning_rate": 1.63383096739871e-05,
|
| 952 |
+
"loss": 0.4626,
|
| 953 |
+
"step": 152
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 4.28,
|
| 957 |
+
"learning_rate": 1.6293203910498375e-05,
|
| 958 |
+
"loss": 0.4665,
|
| 959 |
+
"step": 153
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 4.3,
|
| 963 |
+
"learning_rate": 1.6247885145439602e-05,
|
| 964 |
+
"loss": 0.4654,
|
| 965 |
+
"step": 154
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 4.31,
|
| 969 |
+
"learning_rate": 1.6202354912682602e-05,
|
| 970 |
+
"loss": 0.4695,
|
| 971 |
+
"step": 155
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 4.33,
|
| 975 |
+
"learning_rate": 1.6156614753256583e-05,
|
| 976 |
+
"loss": 0.4646,
|
| 977 |
+
"step": 156
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 4.35,
|
| 981 |
+
"learning_rate": 1.6110666215296e-05,
|
| 982 |
+
"loss": 0.4686,
|
| 983 |
+
"step": 157
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 4.37,
|
| 987 |
+
"learning_rate": 1.6064510853988137e-05,
|
| 988 |
+
"loss": 0.4635,
|
| 989 |
+
"step": 158
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 4.39,
|
| 993 |
+
"learning_rate": 1.6018150231520486e-05,
|
| 994 |
+
"loss": 0.4638,
|
| 995 |
+
"step": 159
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 4.41,
|
| 999 |
+
"learning_rate": 1.5971585917027864e-05,
|
| 1000 |
+
"loss": 0.4666,
|
| 1001 |
+
"step": 160
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 4.42,
|
| 1005 |
+
"learning_rate": 1.592481948653931e-05,
|
| 1006 |
+
"loss": 0.4668,
|
| 1007 |
+
"step": 161
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 4.44,
|
| 1011 |
+
"learning_rate": 1.5877852522924733e-05,
|
| 1012 |
+
"loss": 0.4656,
|
| 1013 |
+
"step": 162
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 4.46,
|
| 1017 |
+
"learning_rate": 1.5830686615841348e-05,
|
| 1018 |
+
"loss": 0.4652,
|
| 1019 |
+
"step": 163
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 4.48,
|
| 1023 |
+
"learning_rate": 1.5783323361679865e-05,
|
| 1024 |
+
"loss": 0.4651,
|
| 1025 |
+
"step": 164
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 4.5,
|
| 1029 |
+
"learning_rate": 1.573576436351046e-05,
|
| 1030 |
+
"loss": 0.4657,
|
| 1031 |
+
"step": 165
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 4.52,
|
| 1035 |
+
"learning_rate": 1.568801123102852e-05,
|
| 1036 |
+
"loss": 0.4654,
|
| 1037 |
+
"step": 166
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 4.54,
|
| 1041 |
+
"learning_rate": 1.5640065580500146e-05,
|
| 1042 |
+
"loss": 0.4633,
|
| 1043 |
+
"step": 167
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 4.55,
|
| 1047 |
+
"learning_rate": 1.5591929034707468e-05,
|
| 1048 |
+
"loss": 0.4628,
|
| 1049 |
+
"step": 168
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 4.57,
|
| 1053 |
+
"learning_rate": 1.5543603222893718e-05,
|
| 1054 |
+
"loss": 0.4691,
|
| 1055 |
+
"step": 169
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 4.59,
|
| 1059 |
+
"learning_rate": 1.5495089780708062e-05,
|
| 1060 |
+
"loss": 0.4644,
|
| 1061 |
+
"step": 170
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 4.61,
|
| 1065 |
+
"learning_rate": 1.5446390350150272e-05,
|
| 1066 |
+
"loss": 0.4633,
|
| 1067 |
+
"step": 171
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 4.63,
|
| 1071 |
+
"learning_rate": 1.539750657951513e-05,
|
| 1072 |
+
"loss": 0.4615,
|
| 1073 |
+
"step": 172
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 4.63,
|
| 1077 |
+
"eval_loss": 0.4630681276321411,
|
| 1078 |
+
"eval_runtime": 23.8164,
|
| 1079 |
+
"eval_samples_per_second": 8.398,
|
| 1080 |
+
"eval_steps_per_second": 0.546,
|
| 1081 |
+
"step": 172
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 5.01,
|
| 1085 |
+
"learning_rate": 1.5348440123336647e-05,
|
| 1086 |
+
"loss": 0.4604,
|
| 1087 |
+
"step": 173
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 5.03,
|
| 1091 |
+
"learning_rate": 1.529919264233205e-05,
|
| 1092 |
+
"loss": 0.4624,
|
| 1093 |
+
"step": 174
|
| 1094 |
+
},
|
| 1095 |
+
{
|
| 1096 |
+
"epoch": 5.05,
|
| 1097 |
+
"learning_rate": 1.5249765803345602e-05,
|
| 1098 |
+
"loss": 0.4629,
|
| 1099 |
+
"step": 175
|
| 1100 |
+
},
|
| 1101 |
+
{
|
| 1102 |
+
"epoch": 5.07,
|
| 1103 |
+
"learning_rate": 1.5200161279292154e-05,
|
| 1104 |
+
"loss": 0.4616,
|
| 1105 |
+
"step": 176
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 5.09,
|
| 1109 |
+
"learning_rate": 1.5150380749100545e-05,
|
| 1110 |
+
"loss": 0.4681,
|
| 1111 |
+
"step": 177
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 5.1,
|
| 1115 |
+
"learning_rate": 1.5100425897656754e-05,
|
| 1116 |
+
"loss": 0.4653,
|
| 1117 |
+
"step": 178
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 5.12,
|
| 1121 |
+
"learning_rate": 1.5050298415746903e-05,
|
| 1122 |
+
"loss": 0.459,
|
| 1123 |
+
"step": 179
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 5.14,
|
| 1127 |
+
"learning_rate": 1.5000000000000002e-05,
|
| 1128 |
+
"loss": 0.4589,
|
| 1129 |
+
"step": 180
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 5.16,
|
| 1133 |
+
"learning_rate": 1.4949532352830543e-05,
|
| 1134 |
+
"loss": 0.4627,
|
| 1135 |
+
"step": 181
|
| 1136 |
+
},
|
| 1137 |
+
{
|
| 1138 |
+
"epoch": 5.18,
|
| 1139 |
+
"learning_rate": 1.4898897182380872e-05,
|
| 1140 |
+
"loss": 0.4579,
|
| 1141 |
+
"step": 182
|
| 1142 |
+
},
|
| 1143 |
+
{
|
| 1144 |
+
"epoch": 5.2,
|
| 1145 |
+
"learning_rate": 1.4848096202463373e-05,
|
| 1146 |
+
"loss": 0.4612,
|
| 1147 |
+
"step": 183
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 5.21,
|
| 1151 |
+
"learning_rate": 1.4797131132502464e-05,
|
| 1152 |
+
"loss": 0.4598,
|
| 1153 |
+
"step": 184
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 5.23,
|
| 1157 |
+
"learning_rate": 1.4746003697476406e-05,
|
| 1158 |
+
"loss": 0.4615,
|
| 1159 |
+
"step": 185
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 5.25,
|
| 1163 |
+
"learning_rate": 1.469471562785891e-05,
|
| 1164 |
+
"loss": 0.4571,
|
| 1165 |
+
"step": 186
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 5.27,
|
| 1169 |
+
"learning_rate": 1.4643268659560571e-05,
|
| 1170 |
+
"loss": 0.4589,
|
| 1171 |
+
"step": 187
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 5.29,
|
| 1175 |
+
"learning_rate": 1.4591664533870118e-05,
|
| 1176 |
+
"loss": 0.4608,
|
| 1177 |
+
"step": 188
|
| 1178 |
+
},
|
| 1179 |
+
{
|
| 1180 |
+
"epoch": 5.31,
|
| 1181 |
+
"learning_rate": 1.4539904997395468e-05,
|
| 1182 |
+
"loss": 0.4606,
|
| 1183 |
+
"step": 189
|
| 1184 |
+
},
|
| 1185 |
+
{
|
| 1186 |
+
"epoch": 5.32,
|
| 1187 |
+
"learning_rate": 1.4487991802004625e-05,
|
| 1188 |
+
"loss": 0.4607,
|
| 1189 |
+
"step": 190
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 5.34,
|
| 1193 |
+
"learning_rate": 1.4435926704766364e-05,
|
| 1194 |
+
"loss": 0.4605,
|
| 1195 |
+
"step": 191
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 5.36,
|
| 1199 |
+
"learning_rate": 1.4383711467890776e-05,
|
| 1200 |
+
"loss": 0.4598,
|
| 1201 |
+
"step": 192
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 5.38,
|
| 1205 |
+
"learning_rate": 1.4331347858669631e-05,
|
| 1206 |
+
"loss": 0.4588,
|
| 1207 |
+
"step": 193
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 5.4,
|
| 1211 |
+
"learning_rate": 1.4278837649416543e-05,
|
| 1212 |
+
"loss": 0.4575,
|
| 1213 |
+
"step": 194
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 5.42,
|
| 1217 |
+
"learning_rate": 1.4226182617406996e-05,
|
| 1218 |
+
"loss": 0.46,
|
| 1219 |
+
"step": 195
|
| 1220 |
+
},
|
| 1221 |
+
{
|
| 1222 |
+
"epoch": 5.43,
|
| 1223 |
+
"learning_rate": 1.417338454481818e-05,
|
| 1224 |
+
"loss": 0.4605,
|
| 1225 |
+
"step": 196
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 5.45,
|
| 1229 |
+
"learning_rate": 1.4120445218668687e-05,
|
| 1230 |
+
"loss": 0.4606,
|
| 1231 |
+
"step": 197
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 5.47,
|
| 1235 |
+
"learning_rate": 1.4067366430758004e-05,
|
| 1236 |
+
"loss": 0.4584,
|
| 1237 |
+
"step": 198
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 5.49,
|
| 1241 |
+
"learning_rate": 1.4014149977605893e-05,
|
| 1242 |
+
"loss": 0.4623,
|
| 1243 |
+
"step": 199
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 5.51,
|
| 1247 |
+
"learning_rate": 1.396079766039157e-05,
|
| 1248 |
+
"loss": 0.4586,
|
| 1249 |
+
"step": 200
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 5.53,
|
| 1253 |
+
"learning_rate": 1.3907311284892737e-05,
|
| 1254 |
+
"loss": 0.457,
|
| 1255 |
+
"step": 201
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 5.54,
|
| 1259 |
+
"learning_rate": 1.3853692661424485e-05,
|
| 1260 |
+
"loss": 0.4577,
|
| 1261 |
+
"step": 202
|
| 1262 |
+
},
|
| 1263 |
+
{
|
| 1264 |
+
"epoch": 5.56,
|
| 1265 |
+
"learning_rate": 1.3799943604777993e-05,
|
| 1266 |
+
"loss": 0.4604,
|
| 1267 |
+
"step": 203
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"epoch": 5.58,
|
| 1271 |
+
"learning_rate": 1.3746065934159123e-05,
|
| 1272 |
+
"loss": 0.4603,
|
| 1273 |
+
"step": 204
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 5.6,
|
| 1277 |
+
"learning_rate": 1.3692061473126845e-05,
|
| 1278 |
+
"loss": 0.46,
|
| 1279 |
+
"step": 205
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"epoch": 5.62,
|
| 1283 |
+
"learning_rate": 1.3637932049531517e-05,
|
| 1284 |
+
"loss": 0.4584,
|
| 1285 |
+
"step": 206
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 5.62,
|
| 1289 |
+
"eval_loss": 0.4574292004108429,
|
| 1290 |
+
"eval_runtime": 23.8087,
|
| 1291 |
+
"eval_samples_per_second": 8.4,
|
| 1292 |
+
"eval_steps_per_second": 0.546,
|
| 1293 |
+
"step": 206
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 6.0,
|
| 1297 |
+
"learning_rate": 1.3583679495453e-05,
|
| 1298 |
+
"loss": 0.4533,
|
| 1299 |
+
"step": 207
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 6.02,
|
| 1303 |
+
"learning_rate": 1.3529305647138689e-05,
|
| 1304 |
+
"loss": 0.456,
|
| 1305 |
+
"step": 208
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 6.04,
|
| 1309 |
+
"learning_rate": 1.3474812344941315e-05,
|
| 1310 |
+
"loss": 0.4592,
|
| 1311 |
+
"step": 209
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 6.06,
|
| 1315 |
+
"learning_rate": 1.342020143325669e-05,
|
| 1316 |
+
"loss": 0.4559,
|
| 1317 |
+
"step": 210
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 6.08,
|
| 1321 |
+
"learning_rate": 1.3365474760461265e-05,
|
| 1322 |
+
"loss": 0.4606,
|
| 1323 |
+
"step": 211
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 6.1,
|
| 1327 |
+
"learning_rate": 1.3310634178849583e-05,
|
| 1328 |
+
"loss": 0.4607,
|
| 1329 |
+
"step": 212
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 6.11,
|
| 1333 |
+
"learning_rate": 1.3255681544571568e-05,
|
| 1334 |
+
"loss": 0.4557,
|
| 1335 |
+
"step": 213
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 6.13,
|
| 1339 |
+
"learning_rate": 1.3200618717569716e-05,
|
| 1340 |
+
"loss": 0.4549,
|
| 1341 |
+
"step": 214
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 6.15,
|
| 1345 |
+
"learning_rate": 1.3145447561516138e-05,
|
| 1346 |
+
"loss": 0.4555,
|
| 1347 |
+
"step": 215
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 6.17,
|
| 1351 |
+
"learning_rate": 1.3090169943749475e-05,
|
| 1352 |
+
"loss": 0.4557,
|
| 1353 |
+
"step": 216
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 6.19,
|
| 1357 |
+
"learning_rate": 1.3034787735211708e-05,
|
| 1358 |
+
"loss": 0.4536,
|
| 1359 |
+
"step": 217
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 6.21,
|
| 1363 |
+
"learning_rate": 1.297930281038482e-05,
|
| 1364 |
+
"loss": 0.459,
|
| 1365 |
+
"step": 218
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 6.22,
|
| 1369 |
+
"learning_rate": 1.2923717047227368e-05,
|
| 1370 |
+
"loss": 0.4543,
|
| 1371 |
+
"step": 219
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 6.24,
|
| 1375 |
+
"learning_rate": 1.2868032327110904e-05,
|
| 1376 |
+
"loss": 0.4528,
|
| 1377 |
+
"step": 220
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 6.26,
|
| 1381 |
+
"learning_rate": 1.2812250534756307e-05,
|
| 1382 |
+
"loss": 0.4574,
|
| 1383 |
+
"step": 221
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 6.28,
|
| 1387 |
+
"learning_rate": 1.2756373558169992e-05,
|
| 1388 |
+
"loss": 0.4533,
|
| 1389 |
+
"step": 222
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 6.3,
|
| 1393 |
+
"learning_rate": 1.270040328858001e-05,
|
| 1394 |
+
"loss": 0.4558,
|
| 1395 |
+
"step": 223
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 6.32,
|
| 1399 |
+
"learning_rate": 1.2644341620372025e-05,
|
| 1400 |
+
"loss": 0.4584,
|
| 1401 |
+
"step": 224
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 6.33,
|
| 1405 |
+
"learning_rate": 1.2588190451025209e-05,
|
| 1406 |
+
"loss": 0.4547,
|
| 1407 |
+
"step": 225
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 6.35,
|
| 1411 |
+
"learning_rate": 1.253195168104802e-05,
|
| 1412 |
+
"loss": 0.4568,
|
| 1413 |
+
"step": 226
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 6.37,
|
| 1417 |
+
"learning_rate": 1.2475627213913861e-05,
|
| 1418 |
+
"loss": 0.4547,
|
| 1419 |
+
"step": 227
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 6.39,
|
| 1423 |
+
"learning_rate": 1.2419218955996677e-05,
|
| 1424 |
+
"loss": 0.4523,
|
| 1425 |
+
"step": 228
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 6.41,
|
| 1429 |
+
"learning_rate": 1.2362728816506418e-05,
|
| 1430 |
+
"loss": 0.4556,
|
| 1431 |
+
"step": 229
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 6.43,
|
| 1435 |
+
"learning_rate": 1.2306158707424402e-05,
|
| 1436 |
+
"loss": 0.4561,
|
| 1437 |
+
"step": 230
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 6.44,
|
| 1441 |
+
"learning_rate": 1.2249510543438652e-05,
|
| 1442 |
+
"loss": 0.4548,
|
| 1443 |
+
"step": 231
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 6.46,
|
| 1447 |
+
"learning_rate": 1.2192786241879033e-05,
|
| 1448 |
+
"loss": 0.4569,
|
| 1449 |
+
"step": 232
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 6.48,
|
| 1453 |
+
"learning_rate": 1.2135987722652403e-05,
|
| 1454 |
+
"loss": 0.4561,
|
| 1455 |
+
"step": 233
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 6.5,
|
| 1459 |
+
"learning_rate": 1.2079116908177592e-05,
|
| 1460 |
+
"loss": 0.4535,
|
| 1461 |
+
"step": 234
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 6.52,
|
| 1465 |
+
"learning_rate": 1.2022175723320382e-05,
|
| 1466 |
+
"loss": 0.4538,
|
| 1467 |
+
"step": 235
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 6.54,
|
| 1471 |
+
"learning_rate": 1.1965166095328302e-05,
|
| 1472 |
+
"loss": 0.454,
|
| 1473 |
+
"step": 236
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 6.55,
|
| 1477 |
+
"learning_rate": 1.190808995376545e-05,
|
| 1478 |
+
"loss": 0.4555,
|
| 1479 |
+
"step": 237
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 6.57,
|
| 1483 |
+
"learning_rate": 1.1850949230447146e-05,
|
| 1484 |
+
"loss": 0.4567,
|
| 1485 |
+
"step": 238
|
| 1486 |
+
},
|
| 1487 |
+
{
|
| 1488 |
+
"epoch": 6.59,
|
| 1489 |
+
"learning_rate": 1.1793745859374575e-05,
|
| 1490 |
+
"loss": 0.4571,
|
| 1491 |
+
"step": 239
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 6.61,
|
| 1495 |
+
"learning_rate": 1.1736481776669307e-05,
|
| 1496 |
+
"loss": 0.4535,
|
| 1497 |
+
"step": 240
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 6.63,
|
| 1501 |
+
"learning_rate": 1.1679158920507773e-05,
|
| 1502 |
+
"loss": 0.4524,
|
| 1503 |
+
"step": 241
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 6.63,
|
| 1507 |
+
"eval_loss": 0.45291438698768616,
|
| 1508 |
+
"eval_runtime": 23.8049,
|
| 1509 |
+
"eval_samples_per_second": 8.402,
|
| 1510 |
+
"eval_steps_per_second": 0.546,
|
| 1511 |
+
"step": 241
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 7.01,
|
| 1515 |
+
"learning_rate": 1.1621779231055677e-05,
|
| 1516 |
+
"loss": 0.4501,
|
| 1517 |
+
"step": 242
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 7.03,
|
| 1521 |
+
"learning_rate": 1.156434465040231e-05,
|
| 1522 |
+
"loss": 0.4529,
|
| 1523 |
+
"step": 243
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 7.05,
|
| 1527 |
+
"learning_rate": 1.1506857122494832e-05,
|
| 1528 |
+
"loss": 0.4559,
|
| 1529 |
+
"step": 244
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 7.07,
|
| 1533 |
+
"learning_rate": 1.1449318593072468e-05,
|
| 1534 |
+
"loss": 0.4523,
|
| 1535 |
+
"step": 245
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 7.09,
|
| 1539 |
+
"learning_rate": 1.1391731009600655e-05,
|
| 1540 |
+
"loss": 0.4572,
|
| 1541 |
+
"step": 246
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 7.11,
|
| 1545 |
+
"learning_rate": 1.1334096321205129e-05,
|
| 1546 |
+
"loss": 0.4543,
|
| 1547 |
+
"step": 247
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 7.12,
|
| 1551 |
+
"learning_rate": 1.127641647860595e-05,
|
| 1552 |
+
"loss": 0.452,
|
| 1553 |
+
"step": 248
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 7.14,
|
| 1557 |
+
"learning_rate": 1.1218693434051475e-05,
|
| 1558 |
+
"loss": 0.4529,
|
| 1559 |
+
"step": 249
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 7.16,
|
| 1563 |
+
"learning_rate": 1.1160929141252303e-05,
|
| 1564 |
+
"loss": 0.4528,
|
| 1565 |
+
"step": 250
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"epoch": 7.18,
|
| 1569 |
+
"learning_rate": 1.110312555531512e-05,
|
| 1570 |
+
"loss": 0.4505,
|
| 1571 |
+
"step": 251
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 7.2,
|
| 1575 |
+
"learning_rate": 1.1045284632676535e-05,
|
| 1576 |
+
"loss": 0.4491,
|
| 1577 |
+
"step": 252
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 7.22,
|
| 1581 |
+
"learning_rate": 1.0987408331036879e-05,
|
| 1582 |
+
"loss": 0.4544,
|
| 1583 |
+
"step": 253
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 7.23,
|
| 1587 |
+
"learning_rate": 1.0929498609293925e-05,
|
| 1588 |
+
"loss": 0.4509,
|
| 1589 |
+
"step": 254
|
| 1590 |
+
},
|
| 1591 |
+
{
|
| 1592 |
+
"epoch": 7.25,
|
| 1593 |
+
"learning_rate": 1.0871557427476585e-05,
|
| 1594 |
+
"loss": 0.4517,
|
| 1595 |
+
"step": 255
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 7.27,
|
| 1599 |
+
"learning_rate": 1.0813586746678584e-05,
|
| 1600 |
+
"loss": 0.4505,
|
| 1601 |
+
"step": 256
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 7.29,
|
| 1605 |
+
"learning_rate": 1.0755588528992082e-05,
|
| 1606 |
+
"loss": 0.4508,
|
| 1607 |
+
"step": 257
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 7.31,
|
| 1611 |
+
"learning_rate": 1.0697564737441254e-05,
|
| 1612 |
+
"loss": 0.4537,
|
| 1613 |
+
"step": 258
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 7.33,
|
| 1617 |
+
"learning_rate": 1.0639517335915857e-05,
|
| 1618 |
+
"loss": 0.4543,
|
| 1619 |
+
"step": 259
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 7.34,
|
| 1623 |
+
"learning_rate": 1.0581448289104759e-05,
|
| 1624 |
+
"loss": 0.4532,
|
| 1625 |
+
"step": 260
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 7.36,
|
| 1629 |
+
"learning_rate": 1.0523359562429441e-05,
|
| 1630 |
+
"loss": 0.4532,
|
| 1631 |
+
"step": 261
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 7.38,
|
| 1635 |
+
"learning_rate": 1.046525312197747e-05,
|
| 1636 |
+
"loss": 0.4507,
|
| 1637 |
+
"step": 262
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 7.4,
|
| 1641 |
+
"learning_rate": 1.040713093443596e-05,
|
| 1642 |
+
"loss": 0.4494,
|
| 1643 |
+
"step": 263
|
| 1644 |
+
},
|
| 1645 |
+
{
|
| 1646 |
+
"epoch": 7.42,
|
| 1647 |
+
"learning_rate": 1.0348994967025012e-05,
|
| 1648 |
+
"loss": 0.4531,
|
| 1649 |
+
"step": 264
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 7.44,
|
| 1653 |
+
"learning_rate": 1.0290847187431115e-05,
|
| 1654 |
+
"loss": 0.4525,
|
| 1655 |
+
"step": 265
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 7.45,
|
| 1659 |
+
"learning_rate": 1.0232689563740563e-05,
|
| 1660 |
+
"loss": 0.452,
|
| 1661 |
+
"step": 266
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 7.47,
|
| 1665 |
+
"learning_rate": 1.0174524064372837e-05,
|
| 1666 |
+
"loss": 0.4527,
|
| 1667 |
+
"step": 267
|
| 1668 |
+
},
|
| 1669 |
+
{
|
| 1670 |
+
"epoch": 7.49,
|
| 1671 |
+
"learning_rate": 1.0116352658013973e-05,
|
| 1672 |
+
"loss": 0.4532,
|
| 1673 |
+
"step": 268
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"epoch": 7.51,
|
| 1677 |
+
"learning_rate": 1.005817731354994e-05,
|
| 1678 |
+
"loss": 0.4507,
|
| 1679 |
+
"step": 269
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 7.53,
|
| 1683 |
+
"learning_rate": 1e-05,
|
| 1684 |
+
"loss": 0.4504,
|
| 1685 |
+
"step": 270
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 7.55,
|
| 1689 |
+
"learning_rate": 9.941822686450061e-06,
|
| 1690 |
+
"loss": 0.4536,
|
| 1691 |
+
"step": 271
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 7.56,
|
| 1695 |
+
"learning_rate": 9.883647341986032e-06,
|
| 1696 |
+
"loss": 0.4515,
|
| 1697 |
+
"step": 272
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 7.58,
|
| 1701 |
+
"learning_rate": 9.825475935627165e-06,
|
| 1702 |
+
"loss": 0.4537,
|
| 1703 |
+
"step": 273
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 7.6,
|
| 1707 |
+
"learning_rate": 9.767310436259438e-06,
|
| 1708 |
+
"loss": 0.4527,
|
| 1709 |
+
"step": 274
|
| 1710 |
+
},
|
| 1711 |
+
{
|
| 1712 |
+
"epoch": 7.62,
|
| 1713 |
+
"learning_rate": 9.709152812568886e-06,
|
| 1714 |
+
"loss": 0.4507,
|
| 1715 |
+
"step": 275
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 7.62,
|
| 1719 |
+
"eval_loss": 0.4502379596233368,
|
| 1720 |
+
"eval_runtime": 23.8128,
|
| 1721 |
+
"eval_samples_per_second": 8.399,
|
| 1722 |
+
"eval_steps_per_second": 0.546,
|
| 1723 |
+
"step": 275
|
| 1724 |
+
},
|
| 1725 |
+
{
|
| 1726 |
+
"epoch": 8.0,
|
| 1727 |
+
"learning_rate": 9.651005032974994e-06,
|
| 1728 |
+
"loss": 0.4474,
|
| 1729 |
+
"step": 276
|
| 1730 |
+
},
|
| 1731 |
+
{
|
| 1732 |
+
"epoch": 8.02,
|
| 1733 |
+
"learning_rate": 9.592869065564043e-06,
|
| 1734 |
+
"loss": 0.4489,
|
| 1735 |
+
"step": 277
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 8.04,
|
| 1739 |
+
"learning_rate": 9.534746878022533e-06,
|
| 1740 |
+
"loss": 0.4527,
|
| 1741 |
+
"step": 278
|
| 1742 |
+
},
|
| 1743 |
+
{
|
| 1744 |
+
"epoch": 8.06,
|
| 1745 |
+
"learning_rate": 9.476640437570562e-06,
|
| 1746 |
+
"loss": 0.4495,
|
| 1747 |
+
"step": 279
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 8.08,
|
| 1751 |
+
"learning_rate": 9.418551710895243e-06,
|
| 1752 |
+
"loss": 0.4567,
|
| 1753 |
+
"step": 280
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 8.1,
|
| 1757 |
+
"learning_rate": 9.360482664084144e-06,
|
| 1758 |
+
"loss": 0.4523,
|
| 1759 |
+
"step": 281
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 8.11,
|
| 1763 |
+
"learning_rate": 9.302435262558748e-06,
|
| 1764 |
+
"loss": 0.4485,
|
| 1765 |
+
"step": 282
|
| 1766 |
+
},
|
| 1767 |
+
{
|
| 1768 |
+
"epoch": 8.13,
|
| 1769 |
+
"learning_rate": 9.244411471007923e-06,
|
| 1770 |
+
"loss": 0.4492,
|
| 1771 |
+
"step": 283
|
| 1772 |
+
},
|
| 1773 |
+
{
|
| 1774 |
+
"epoch": 8.15,
|
| 1775 |
+
"learning_rate": 9.18641325332142e-06,
|
| 1776 |
+
"loss": 0.4508,
|
| 1777 |
+
"step": 284
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"epoch": 8.17,
|
| 1781 |
+
"learning_rate": 9.128442572523418e-06,
|
| 1782 |
+
"loss": 0.4486,
|
| 1783 |
+
"step": 285
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"epoch": 8.19,
|
| 1787 |
+
"learning_rate": 9.07050139070608e-06,
|
| 1788 |
+
"loss": 0.4495,
|
| 1789 |
+
"step": 286
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"epoch": 8.21,
|
| 1793 |
+
"learning_rate": 9.012591668963123e-06,
|
| 1794 |
+
"loss": 0.4521,
|
| 1795 |
+
"step": 287
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 8.23,
|
| 1799 |
+
"learning_rate": 8.954715367323468e-06,
|
| 1800 |
+
"loss": 0.4472,
|
| 1801 |
+
"step": 288
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 8.24,
|
| 1805 |
+
"learning_rate": 8.896874444684882e-06,
|
| 1806 |
+
"loss": 0.4494,
|
| 1807 |
+
"step": 289
|
| 1808 |
+
},
|
| 1809 |
+
{
|
| 1810 |
+
"epoch": 8.26,
|
| 1811 |
+
"learning_rate": 8.839070858747697e-06,
|
| 1812 |
+
"loss": 0.4482,
|
| 1813 |
+
"step": 290
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"epoch": 8.28,
|
| 1817 |
+
"learning_rate": 8.781306565948528e-06,
|
| 1818 |
+
"loss": 0.4488,
|
| 1819 |
+
"step": 291
|
| 1820 |
+
},
|
| 1821 |
+
{
|
| 1822 |
+
"epoch": 8.3,
|
| 1823 |
+
"learning_rate": 8.723583521394054e-06,
|
| 1824 |
+
"loss": 0.4527,
|
| 1825 |
+
"step": 292
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 8.32,
|
| 1829 |
+
"learning_rate": 8.665903678794873e-06,
|
| 1830 |
+
"loss": 0.4499,
|
| 1831 |
+
"step": 293
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 8.34,
|
| 1835 |
+
"learning_rate": 8.60826899039935e-06,
|
| 1836 |
+
"loss": 0.4511,
|
| 1837 |
+
"step": 294
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 8.35,
|
| 1841 |
+
"learning_rate": 8.550681406927534e-06,
|
| 1842 |
+
"loss": 0.4528,
|
| 1843 |
+
"step": 295
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 8.37,
|
| 1847 |
+
"learning_rate": 8.49314287750517e-06,
|
| 1848 |
+
"loss": 0.4465,
|
| 1849 |
+
"step": 296
|
| 1850 |
+
},
|
| 1851 |
+
{
|
| 1852 |
+
"epoch": 8.39,
|
| 1853 |
+
"learning_rate": 8.43565534959769e-06,
|
| 1854 |
+
"loss": 0.4467,
|
| 1855 |
+
"step": 297
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"epoch": 8.41,
|
| 1859 |
+
"learning_rate": 8.378220768944328e-06,
|
| 1860 |
+
"loss": 0.4519,
|
| 1861 |
+
"step": 298
|
| 1862 |
+
},
|
| 1863 |
+
{
|
| 1864 |
+
"epoch": 8.43,
|
| 1865 |
+
"learning_rate": 8.32084107949223e-06,
|
| 1866 |
+
"loss": 0.4517,
|
| 1867 |
+
"step": 299
|
| 1868 |
+
},
|
| 1869 |
+
{
|
| 1870 |
+
"epoch": 8.45,
|
| 1871 |
+
"learning_rate": 8.263518223330698e-06,
|
| 1872 |
+
"loss": 0.4487,
|
| 1873 |
+
"step": 300
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"epoch": 8.46,
|
| 1877 |
+
"learning_rate": 8.206254140625425e-06,
|
| 1878 |
+
"loss": 0.4514,
|
| 1879 |
+
"step": 301
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 8.48,
|
| 1883 |
+
"learning_rate": 8.149050769552856e-06,
|
| 1884 |
+
"loss": 0.4524,
|
| 1885 |
+
"step": 302
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 8.5,
|
| 1889 |
+
"learning_rate": 8.091910046234552e-06,
|
| 1890 |
+
"loss": 0.4493,
|
| 1891 |
+
"step": 303
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"epoch": 8.52,
|
| 1895 |
+
"learning_rate": 8.034833904671698e-06,
|
| 1896 |
+
"loss": 0.4495,
|
| 1897 |
+
"step": 304
|
| 1898 |
+
},
|
| 1899 |
+
{
|
| 1900 |
+
"epoch": 8.54,
|
| 1901 |
+
"learning_rate": 7.977824276679623e-06,
|
| 1902 |
+
"loss": 0.4501,
|
| 1903 |
+
"step": 305
|
| 1904 |
+
},
|
| 1905 |
+
{
|
| 1906 |
+
"epoch": 8.56,
|
| 1907 |
+
"learning_rate": 7.92088309182241e-06,
|
| 1908 |
+
"loss": 0.4482,
|
| 1909 |
+
"step": 306
|
| 1910 |
+
},
|
| 1911 |
+
{
|
| 1912 |
+
"epoch": 8.57,
|
| 1913 |
+
"learning_rate": 7.864012277347602e-06,
|
| 1914 |
+
"loss": 0.4539,
|
| 1915 |
+
"step": 307
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"epoch": 8.59,
|
| 1919 |
+
"learning_rate": 7.807213758120965e-06,
|
| 1920 |
+
"loss": 0.449,
|
| 1921 |
+
"step": 308
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 8.61,
|
| 1925 |
+
"learning_rate": 7.750489456561351e-06,
|
| 1926 |
+
"loss": 0.4486,
|
| 1927 |
+
"step": 309
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 8.63,
|
| 1931 |
+
"learning_rate": 7.6938412925756e-06,
|
| 1932 |
+
"loss": 0.4478,
|
| 1933 |
+
"step": 310
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"epoch": 8.63,
|
| 1937 |
+
"eval_loss": 0.4479629397392273,
|
| 1938 |
+
"eval_runtime": 23.8304,
|
| 1939 |
+
"eval_samples_per_second": 8.393,
|
| 1940 |
+
"eval_steps_per_second": 0.546,
|
| 1941 |
+
"step": 310
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 9.01,
|
| 1945 |
+
"learning_rate": 7.637271183493587e-06,
|
| 1946 |
+
"loss": 0.4482,
|
| 1947 |
+
"step": 311
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 9.03,
|
| 1951 |
+
"learning_rate": 7.580781044003324e-06,
|
| 1952 |
+
"loss": 0.4462,
|
| 1953 |
+
"step": 312
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 9.05,
|
| 1957 |
+
"learning_rate": 7.524372786086143e-06,
|
| 1958 |
+
"loss": 0.4488,
|
| 1959 |
+
"step": 313
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 9.07,
|
| 1963 |
+
"learning_rate": 7.468048318951983e-06,
|
| 1964 |
+
"loss": 0.4521,
|
| 1965 |
+
"step": 314
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 9.09,
|
| 1969 |
+
"learning_rate": 7.411809548974792e-06,
|
| 1970 |
+
"loss": 0.4528,
|
| 1971 |
+
"step": 315
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 9.11,
|
| 1975 |
+
"learning_rate": 7.355658379627981e-06,
|
| 1976 |
+
"loss": 0.4497,
|
| 1977 |
+
"step": 316
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 9.12,
|
| 1981 |
+
"learning_rate": 7.299596711419994e-06,
|
| 1982 |
+
"loss": 0.447,
|
| 1983 |
+
"step": 317
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 9.14,
|
| 1987 |
+
"learning_rate": 7.243626441830009e-06,
|
| 1988 |
+
"loss": 0.4488,
|
| 1989 |
+
"step": 318
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 9.16,
|
| 1993 |
+
"learning_rate": 7.187749465243694e-06,
|
| 1994 |
+
"loss": 0.4473,
|
| 1995 |
+
"step": 319
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 9.18,
|
| 1999 |
+
"learning_rate": 7.131967672889101e-06,
|
| 2000 |
+
"loss": 0.4494,
|
| 2001 |
+
"step": 320
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 9.2,
|
| 2005 |
+
"learning_rate": 7.076282952772634e-06,
|
| 2006 |
+
"loss": 0.4461,
|
| 2007 |
+
"step": 321
|
| 2008 |
+
},
|
| 2009 |
+
{
|
| 2010 |
+
"epoch": 9.22,
|
| 2011 |
+
"learning_rate": 7.02069718961518e-06,
|
| 2012 |
+
"loss": 0.4505,
|
| 2013 |
+
"step": 322
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 9.23,
|
| 2017 |
+
"learning_rate": 6.9652122647882966e-06,
|
| 2018 |
+
"loss": 0.4476,
|
| 2019 |
+
"step": 323
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 9.25,
|
| 2023 |
+
"learning_rate": 6.909830056250527e-06,
|
| 2024 |
+
"loss": 0.445,
|
| 2025 |
+
"step": 324
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 9.27,
|
| 2029 |
+
"learning_rate": 6.854552438483866e-06,
|
| 2030 |
+
"loss": 0.4494,
|
| 2031 |
+
"step": 325
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 9.29,
|
| 2035 |
+
"learning_rate": 6.799381282430284e-06,
|
| 2036 |
+
"loss": 0.4464,
|
| 2037 |
+
"step": 326
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 9.31,
|
| 2041 |
+
"learning_rate": 6.744318455428436e-06,
|
| 2042 |
+
"loss": 0.4503,
|
| 2043 |
+
"step": 327
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 9.33,
|
| 2047 |
+
"learning_rate": 6.689365821150421e-06,
|
| 2048 |
+
"loss": 0.4503,
|
| 2049 |
+
"step": 328
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 9.35,
|
| 2053 |
+
"learning_rate": 6.634525239538736e-06,
|
| 2054 |
+
"loss": 0.4485,
|
| 2055 |
+
"step": 329
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 9.36,
|
| 2059 |
+
"learning_rate": 6.579798566743314e-06,
|
| 2060 |
+
"loss": 0.4492,
|
| 2061 |
+
"step": 330
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 9.38,
|
| 2065 |
+
"learning_rate": 6.525187655058687e-06,
|
| 2066 |
+
"loss": 0.4462,
|
| 2067 |
+
"step": 331
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 9.4,
|
| 2071 |
+
"learning_rate": 6.4706943528613135e-06,
|
| 2072 |
+
"loss": 0.4499,
|
| 2073 |
+
"step": 332
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 9.42,
|
| 2077 |
+
"learning_rate": 6.4163205045469975e-06,
|
| 2078 |
+
"loss": 0.4478,
|
| 2079 |
+
"step": 333
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 9.44,
|
| 2083 |
+
"learning_rate": 6.362067950468489e-06,
|
| 2084 |
+
"loss": 0.4491,
|
| 2085 |
+
"step": 334
|
| 2086 |
+
},
|
| 2087 |
+
{
|
| 2088 |
+
"epoch": 9.46,
|
| 2089 |
+
"learning_rate": 6.3079385268731575e-06,
|
| 2090 |
+
"loss": 0.4481,
|
| 2091 |
+
"step": 335
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 9.47,
|
| 2095 |
+
"learning_rate": 6.25393406584088e-06,
|
| 2096 |
+
"loss": 0.4509,
|
| 2097 |
+
"step": 336
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 9.49,
|
| 2101 |
+
"learning_rate": 6.200056395222012e-06,
|
| 2102 |
+
"loss": 0.4489,
|
| 2103 |
+
"step": 337
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 9.51,
|
| 2107 |
+
"learning_rate": 6.146307338575519e-06,
|
| 2108 |
+
"loss": 0.4469,
|
| 2109 |
+
"step": 338
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 9.53,
|
| 2113 |
+
"learning_rate": 6.092688715107265e-06,
|
| 2114 |
+
"loss": 0.4472,
|
| 2115 |
+
"step": 339
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 9.55,
|
| 2119 |
+
"learning_rate": 6.039202339608432e-06,
|
| 2120 |
+
"loss": 0.4499,
|
| 2121 |
+
"step": 340
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 9.57,
|
| 2125 |
+
"learning_rate": 5.9858500223941066e-06,
|
| 2126 |
+
"loss": 0.4483,
|
| 2127 |
+
"step": 341
|
| 2128 |
+
},
|
| 2129 |
+
{
|
| 2130 |
+
"epoch": 9.58,
|
| 2131 |
+
"learning_rate": 5.932633569242e-06,
|
| 2132 |
+
"loss": 0.4494,
|
| 2133 |
+
"step": 342
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 9.6,
|
| 2137 |
+
"learning_rate": 5.879554781331317e-06,
|
| 2138 |
+
"loss": 0.4528,
|
| 2139 |
+
"step": 343
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 9.62,
|
| 2143 |
+
"learning_rate": 5.8266154551818225e-06,
|
| 2144 |
+
"loss": 0.4467,
|
| 2145 |
+
"step": 344
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 9.62,
|
| 2149 |
+
"eval_loss": 0.4474964439868927,
|
| 2150 |
+
"eval_runtime": 23.8091,
|
| 2151 |
+
"eval_samples_per_second": 8.4,
|
| 2152 |
+
"eval_steps_per_second": 0.546,
|
| 2153 |
+
"step": 344
|
| 2154 |
+
},
|
| 2155 |
+
{
|
| 2156 |
+
"epoch": 9.62,
|
| 2157 |
+
"step": 344,
|
| 2158 |
+
"total_flos": 6.414434999835034e+16,
|
| 2159 |
+
"train_loss": 0.4900048969443454,
|
| 2160 |
+
"train_runtime": 108147.3514,
|
| 2161 |
+
"train_samples_per_second": 2.577,
|
| 2162 |
+
"train_steps_per_second": 0.005
|
| 2163 |
+
}
|
| 2164 |
+
],
|
| 2165 |
+
"logging_steps": 1,
|
| 2166 |
+
"max_steps": 540,
|
| 2167 |
+
"num_train_epochs": 10,
|
| 2168 |
+
"save_steps": 500,
|
| 2169 |
+
"total_flos": 6.414434999835034e+16,
|
| 2170 |
+
"trial_name": null,
|
| 2171 |
+
"trial_params": null
|
| 2172 |
+
}
|