Upload folder using huggingface_hub
Browse files- README.md +95 -3
- metadata.json +9 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +734 -0
- training_args.bin +3 -0
README.md
CHANGED
|
@@ -1,3 +1,95 @@
|
|
| 1 |
-
---
|
| 2 |
-
license:
|
| 3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
base_model: google/vit-base-patch16-224
|
| 4 |
+
tags:
|
| 5 |
+
- Image Regression
|
| 6 |
+
datasets:
|
| 7 |
+
- "LiamvL/foodplates"
|
| 8 |
+
metrics:
|
| 9 |
+
- accuracy
|
| 10 |
+
model-index:
|
| 11 |
+
- name: "fine-vlees-zorband25"
|
| 12 |
+
results: []
|
| 13 |
+
---
|
| 14 |
+
|
| 15 |
+
# fine-vlees-zorband25
|
| 16 |
+
## Image Regression Model
|
| 17 |
+
|
| 18 |
+
This model was trained with [Image Regression Model Trainer](https://github.com/TonyAssi/ImageRegression/tree/main). It takes an image as input and outputs a float value.
|
| 19 |
+
|
| 20 |
+
```python
|
| 21 |
+
from ImageRegression import predict
|
| 22 |
+
predict(repo_id='LiamvL/fine_vlees',image_path='image.jpg')
|
| 23 |
+
```
|
| 24 |
+
|
| 25 |
+
---
|
| 26 |
+
|
| 27 |
+
## Dataset
|
| 28 |
+
Dataset: LiamvL/foodplates\
|
| 29 |
+
Value Column: 'rest_weight_Vlees/Vis'\
|
| 30 |
+
Train Test Split: 0.2
|
| 31 |
+
|
| 32 |
+
---
|
| 33 |
+
|
| 34 |
+
## Training
|
| 35 |
+
Base Model: [google/vit-base-patch16-224](https://huggingface.co/google/vit-base-patch16-224)\
|
| 36 |
+
Epochs: 50\
|
| 37 |
+
Learning Rate: 0.0001
|
| 38 |
+
|
| 39 |
+
---
|
| 40 |
+
|
| 41 |
+
## Usage
|
| 42 |
+
|
| 43 |
+
### Download
|
| 44 |
+
```bash
|
| 45 |
+
git clone https://github.com/TonyAssi/ImageRegression.git
|
| 46 |
+
cd ImageRegression
|
| 47 |
+
```
|
| 48 |
+
|
| 49 |
+
### Installation
|
| 50 |
+
```bash
|
| 51 |
+
pip install -r requirements.txt
|
| 52 |
+
```
|
| 53 |
+
|
| 54 |
+
### Import
|
| 55 |
+
```python
|
| 56 |
+
from ImageRegression import train_model, upload_model, predict
|
| 57 |
+
```
|
| 58 |
+
|
| 59 |
+
### Inference (Prediction)
|
| 60 |
+
- **repo_id** 🤗 repo id of the model
|
| 61 |
+
- **image_path** path to image
|
| 62 |
+
```python
|
| 63 |
+
predict(repo_id='LiamvL/fine_vlees',
|
| 64 |
+
image_path='image.jpg')
|
| 65 |
+
```
|
| 66 |
+
The first time this function is called it'll download the safetensor model. Subsequent function calls will run faster.
|
| 67 |
+
|
| 68 |
+
### Train Model
|
| 69 |
+
- **dataset_id** 🤗 dataset id
|
| 70 |
+
- **value_column_name** column name of prediction values in dataset
|
| 71 |
+
- **test_split** test split of the train/test split
|
| 72 |
+
- **output_dir** the directory where the checkpoints will be saved
|
| 73 |
+
- **num_train_epochs** training epochs
|
| 74 |
+
- **learning_rate** learning rate
|
| 75 |
+
```python
|
| 76 |
+
train_model(dataset_id='LiamvL/foodplates',
|
| 77 |
+
value_column_name='rest_weight_Vlees/Vis',
|
| 78 |
+
test_split=0.2,
|
| 79 |
+
output_dir='./results',
|
| 80 |
+
num_train_epochs=50,
|
| 81 |
+
learning_rate=0.0001)
|
| 82 |
+
|
| 83 |
+
```
|
| 84 |
+
The trainer will save the checkpoints in the output_dir location. The model.safetensors are the trained weights you'll use for inference (predicton).
|
| 85 |
+
|
| 86 |
+
### Upload Model
|
| 87 |
+
This function will upload your model to the 🤗 Hub.
|
| 88 |
+
- **model_id** the name of the model id
|
| 89 |
+
- **token** go [here](https://huggingface.co/settings/tokens) to create a new 🤗 token
|
| 90 |
+
- **checkpoint_dir** checkpoint folder that will be uploaded
|
| 91 |
+
```python
|
| 92 |
+
upload_model(model_id='fine-vlees-zorband25',
|
| 93 |
+
token='YOUR_HF_TOKEN',
|
| 94 |
+
checkpoint_dir='./results/checkpoint-940')
|
| 95 |
+
```
|
metadata.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"dataset_id": "LiamvL/foodplates",
|
| 3 |
+
"value_column_name": "rest_weight_Vlees/Vis",
|
| 4 |
+
"test_split": 0.2,
|
| 5 |
+
"num_train_epochs": 50,
|
| 6 |
+
"learning_rate": 0.0001,
|
| 7 |
+
"max_value": 206,
|
| 8 |
+
"seed": 42
|
| 9 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ce49b3fafe7798261ea29cf5550f31ba45e740fc133a2703c9c133f7b4b50964
|
| 3 |
+
size 345583444
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:960723eb0298226fd8c830e4e96abc9adbc4328d9af6cc2972cad0b52500b3ed
|
| 3 |
+
size 691292427
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d1b2da445bdc8c10c1c9019867ae159eaa5e9b0a6d5d406fc41a87fe242c82a
|
| 3 |
+
size 14645
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9008af82ea2c682c9b64de9d5bd271649247b8bd3c3761983a4def45d64125f2
|
| 3 |
+
size 1465
|
trainer_state.json
ADDED
|
@@ -0,0 +1,734 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 50.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.5,
|
| 14 |
+
"grad_norm": 1.8985621929168701,
|
| 15 |
+
"learning_rate": 9.910000000000001e-05,
|
| 16 |
+
"loss": 0.2229,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 1.0,
|
| 21 |
+
"grad_norm": 2.6959447860717773,
|
| 22 |
+
"learning_rate": 9.81e-05,
|
| 23 |
+
"loss": 0.0395,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 1.5,
|
| 28 |
+
"grad_norm": 6.28496789932251,
|
| 29 |
+
"learning_rate": 9.71e-05,
|
| 30 |
+
"loss": 0.0261,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 2.0,
|
| 35 |
+
"grad_norm": 0.8198316097259521,
|
| 36 |
+
"learning_rate": 9.61e-05,
|
| 37 |
+
"loss": 0.0346,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 2.5,
|
| 42 |
+
"grad_norm": 3.930227518081665,
|
| 43 |
+
"learning_rate": 9.51e-05,
|
| 44 |
+
"loss": 0.0237,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 3.0,
|
| 49 |
+
"grad_norm": 0.24117116630077362,
|
| 50 |
+
"learning_rate": 9.41e-05,
|
| 51 |
+
"loss": 0.0144,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 3.5,
|
| 56 |
+
"grad_norm": 2.9592466354370117,
|
| 57 |
+
"learning_rate": 9.310000000000001e-05,
|
| 58 |
+
"loss": 0.0136,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 4.0,
|
| 63 |
+
"grad_norm": 6.1739702224731445,
|
| 64 |
+
"learning_rate": 9.21e-05,
|
| 65 |
+
"loss": 0.0143,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 4.5,
|
| 70 |
+
"grad_norm": 1.1602582931518555,
|
| 71 |
+
"learning_rate": 9.11e-05,
|
| 72 |
+
"loss": 0.0166,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 5.0,
|
| 77 |
+
"grad_norm": 3.7479774951934814,
|
| 78 |
+
"learning_rate": 9.010000000000001e-05,
|
| 79 |
+
"loss": 0.0174,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 5.5,
|
| 84 |
+
"grad_norm": 1.2943713665008545,
|
| 85 |
+
"learning_rate": 8.910000000000001e-05,
|
| 86 |
+
"loss": 0.0127,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 6.0,
|
| 91 |
+
"grad_norm": 1.9738529920578003,
|
| 92 |
+
"learning_rate": 8.81e-05,
|
| 93 |
+
"loss": 0.0056,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 6.5,
|
| 98 |
+
"grad_norm": 1.6200770139694214,
|
| 99 |
+
"learning_rate": 8.71e-05,
|
| 100 |
+
"loss": 0.0053,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 7.0,
|
| 105 |
+
"grad_norm": 2.116046667098999,
|
| 106 |
+
"learning_rate": 8.61e-05,
|
| 107 |
+
"loss": 0.0058,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 7.5,
|
| 112 |
+
"grad_norm": 1.7189213037490845,
|
| 113 |
+
"learning_rate": 8.510000000000001e-05,
|
| 114 |
+
"loss": 0.0035,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 8.0,
|
| 119 |
+
"grad_norm": 0.13006743788719177,
|
| 120 |
+
"learning_rate": 8.41e-05,
|
| 121 |
+
"loss": 0.0025,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 8.5,
|
| 126 |
+
"grad_norm": 0.35229212045669556,
|
| 127 |
+
"learning_rate": 8.31e-05,
|
| 128 |
+
"loss": 0.002,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 9.0,
|
| 133 |
+
"grad_norm": 0.5283310413360596,
|
| 134 |
+
"learning_rate": 8.21e-05,
|
| 135 |
+
"loss": 0.0026,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 9.5,
|
| 140 |
+
"grad_norm": 1.4066272974014282,
|
| 141 |
+
"learning_rate": 8.11e-05,
|
| 142 |
+
"loss": 0.0028,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 10.0,
|
| 147 |
+
"grad_norm": 0.6277774572372437,
|
| 148 |
+
"learning_rate": 8.010000000000001e-05,
|
| 149 |
+
"loss": 0.0018,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 10.5,
|
| 154 |
+
"grad_norm": 1.3672561645507812,
|
| 155 |
+
"learning_rate": 7.910000000000001e-05,
|
| 156 |
+
"loss": 0.0019,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 11.0,
|
| 161 |
+
"grad_norm": 0.849717915058136,
|
| 162 |
+
"learning_rate": 7.81e-05,
|
| 163 |
+
"loss": 0.0014,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 11.5,
|
| 168 |
+
"grad_norm": 1.2382935285568237,
|
| 169 |
+
"learning_rate": 7.71e-05,
|
| 170 |
+
"loss": 0.0017,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 12.0,
|
| 175 |
+
"grad_norm": 0.3568917512893677,
|
| 176 |
+
"learning_rate": 7.61e-05,
|
| 177 |
+
"loss": 0.0015,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 12.5,
|
| 182 |
+
"grad_norm": 0.02996239811182022,
|
| 183 |
+
"learning_rate": 7.510000000000001e-05,
|
| 184 |
+
"loss": 0.0009,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 13.0,
|
| 189 |
+
"grad_norm": 0.3278927206993103,
|
| 190 |
+
"learning_rate": 7.41e-05,
|
| 191 |
+
"loss": 0.0016,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 13.5,
|
| 196 |
+
"grad_norm": 0.78089839220047,
|
| 197 |
+
"learning_rate": 7.31e-05,
|
| 198 |
+
"loss": 0.0015,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 14.0,
|
| 203 |
+
"grad_norm": 1.049814224243164,
|
| 204 |
+
"learning_rate": 7.21e-05,
|
| 205 |
+
"loss": 0.0008,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 14.5,
|
| 210 |
+
"grad_norm": 0.05354577675461769,
|
| 211 |
+
"learning_rate": 7.11e-05,
|
| 212 |
+
"loss": 0.0012,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 15.0,
|
| 217 |
+
"grad_norm": 0.7721472978591919,
|
| 218 |
+
"learning_rate": 7.01e-05,
|
| 219 |
+
"loss": 0.0007,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 15.5,
|
| 224 |
+
"grad_norm": 0.5011315941810608,
|
| 225 |
+
"learning_rate": 6.91e-05,
|
| 226 |
+
"loss": 0.0006,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 16.0,
|
| 231 |
+
"grad_norm": 0.2097933441400528,
|
| 232 |
+
"learning_rate": 6.81e-05,
|
| 233 |
+
"loss": 0.001,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 16.5,
|
| 238 |
+
"grad_norm": 0.6485989689826965,
|
| 239 |
+
"learning_rate": 6.71e-05,
|
| 240 |
+
"loss": 0.0011,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 17.0,
|
| 245 |
+
"grad_norm": 0.46603426337242126,
|
| 246 |
+
"learning_rate": 6.610000000000001e-05,
|
| 247 |
+
"loss": 0.0008,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 17.5,
|
| 252 |
+
"grad_norm": 1.16153883934021,
|
| 253 |
+
"learning_rate": 6.510000000000001e-05,
|
| 254 |
+
"loss": 0.0012,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 18.0,
|
| 259 |
+
"grad_norm": 0.509899377822876,
|
| 260 |
+
"learning_rate": 6.41e-05,
|
| 261 |
+
"loss": 0.001,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 18.5,
|
| 266 |
+
"grad_norm": 0.22210121154785156,
|
| 267 |
+
"learning_rate": 6.31e-05,
|
| 268 |
+
"loss": 0.0006,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 19.0,
|
| 273 |
+
"grad_norm": 1.4337605237960815,
|
| 274 |
+
"learning_rate": 6.21e-05,
|
| 275 |
+
"loss": 0.0014,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 19.5,
|
| 280 |
+
"grad_norm": 0.6034935116767883,
|
| 281 |
+
"learning_rate": 6.110000000000001e-05,
|
| 282 |
+
"loss": 0.0006,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 20.0,
|
| 287 |
+
"grad_norm": 0.27057749032974243,
|
| 288 |
+
"learning_rate": 6.0100000000000004e-05,
|
| 289 |
+
"loss": 0.0009,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 20.5,
|
| 294 |
+
"grad_norm": 0.6695419549942017,
|
| 295 |
+
"learning_rate": 5.91e-05,
|
| 296 |
+
"loss": 0.0004,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 21.0,
|
| 301 |
+
"grad_norm": 1.8617427349090576,
|
| 302 |
+
"learning_rate": 5.8099999999999996e-05,
|
| 303 |
+
"loss": 0.002,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 21.5,
|
| 308 |
+
"grad_norm": 0.7135971784591675,
|
| 309 |
+
"learning_rate": 5.71e-05,
|
| 310 |
+
"loss": 0.0022,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 22.0,
|
| 315 |
+
"grad_norm": 1.104410171508789,
|
| 316 |
+
"learning_rate": 5.610000000000001e-05,
|
| 317 |
+
"loss": 0.0018,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 22.5,
|
| 322 |
+
"grad_norm": 0.5932751297950745,
|
| 323 |
+
"learning_rate": 5.5100000000000004e-05,
|
| 324 |
+
"loss": 0.0014,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 23.0,
|
| 329 |
+
"grad_norm": 0.4027913510799408,
|
| 330 |
+
"learning_rate": 5.410000000000001e-05,
|
| 331 |
+
"loss": 0.0008,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 23.5,
|
| 336 |
+
"grad_norm": 0.7953199148178101,
|
| 337 |
+
"learning_rate": 5.31e-05,
|
| 338 |
+
"loss": 0.0011,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 24.0,
|
| 343 |
+
"grad_norm": 0.25498589873313904,
|
| 344 |
+
"learning_rate": 5.2100000000000006e-05,
|
| 345 |
+
"loss": 0.0004,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 24.5,
|
| 350 |
+
"grad_norm": 0.8258208632469177,
|
| 351 |
+
"learning_rate": 5.11e-05,
|
| 352 |
+
"loss": 0.0007,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 25.0,
|
| 357 |
+
"grad_norm": 0.18510620296001434,
|
| 358 |
+
"learning_rate": 5.0100000000000005e-05,
|
| 359 |
+
"loss": 0.0007,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 25.5,
|
| 364 |
+
"grad_norm": 0.8307366371154785,
|
| 365 |
+
"learning_rate": 4.91e-05,
|
| 366 |
+
"loss": 0.0007,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 26.0,
|
| 371 |
+
"grad_norm": 0.7541914582252502,
|
| 372 |
+
"learning_rate": 4.8100000000000004e-05,
|
| 373 |
+
"loss": 0.0007,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 26.5,
|
| 378 |
+
"grad_norm": 0.3338281214237213,
|
| 379 |
+
"learning_rate": 4.71e-05,
|
| 380 |
+
"loss": 0.0004,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 27.0,
|
| 385 |
+
"grad_norm": 0.06798086315393448,
|
| 386 |
+
"learning_rate": 4.61e-05,
|
| 387 |
+
"loss": 0.0007,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 27.5,
|
| 392 |
+
"grad_norm": 0.15522146224975586,
|
| 393 |
+
"learning_rate": 4.5100000000000005e-05,
|
| 394 |
+
"loss": 0.0004,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 28.0,
|
| 399 |
+
"grad_norm": 0.4013696610927582,
|
| 400 |
+
"learning_rate": 4.41e-05,
|
| 401 |
+
"loss": 0.0006,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 28.5,
|
| 406 |
+
"grad_norm": 0.40657514333724976,
|
| 407 |
+
"learning_rate": 4.3100000000000004e-05,
|
| 408 |
+
"loss": 0.0007,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 29.0,
|
| 413 |
+
"grad_norm": 0.14566798508167267,
|
| 414 |
+
"learning_rate": 4.21e-05,
|
| 415 |
+
"loss": 0.0001,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 29.5,
|
| 420 |
+
"grad_norm": 0.4930070638656616,
|
| 421 |
+
"learning_rate": 4.11e-05,
|
| 422 |
+
"loss": 0.0003,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 30.0,
|
| 427 |
+
"grad_norm": 0.2327234447002411,
|
| 428 |
+
"learning_rate": 4.0100000000000006e-05,
|
| 429 |
+
"loss": 0.0005,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 30.5,
|
| 434 |
+
"grad_norm": 0.5080268383026123,
|
| 435 |
+
"learning_rate": 3.91e-05,
|
| 436 |
+
"loss": 0.0003,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 31.0,
|
| 441 |
+
"grad_norm": 0.5666539669036865,
|
| 442 |
+
"learning_rate": 3.8100000000000005e-05,
|
| 443 |
+
"loss": 0.0004,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 31.5,
|
| 448 |
+
"grad_norm": 0.03347349166870117,
|
| 449 |
+
"learning_rate": 3.71e-05,
|
| 450 |
+
"loss": 0.0006,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 32.0,
|
| 455 |
+
"grad_norm": 0.4377991557121277,
|
| 456 |
+
"learning_rate": 3.61e-05,
|
| 457 |
+
"loss": 0.0001,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 32.5,
|
| 462 |
+
"grad_norm": 0.19413357973098755,
|
| 463 |
+
"learning_rate": 3.51e-05,
|
| 464 |
+
"loss": 0.0005,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 33.0,
|
| 469 |
+
"grad_norm": 0.18701113760471344,
|
| 470 |
+
"learning_rate": 3.41e-05,
|
| 471 |
+
"loss": 0.0001,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 33.5,
|
| 476 |
+
"grad_norm": 0.31651440262794495,
|
| 477 |
+
"learning_rate": 3.3100000000000005e-05,
|
| 478 |
+
"loss": 0.0005,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 34.0,
|
| 483 |
+
"grad_norm": 0.02828579768538475,
|
| 484 |
+
"learning_rate": 3.21e-05,
|
| 485 |
+
"loss": 0.0,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 34.5,
|
| 490 |
+
"grad_norm": 0.34173232316970825,
|
| 491 |
+
"learning_rate": 3.1100000000000004e-05,
|
| 492 |
+
"loss": 0.0006,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 35.0,
|
| 497 |
+
"grad_norm": 0.09027812629938126,
|
| 498 |
+
"learning_rate": 3.01e-05,
|
| 499 |
+
"loss": 0.0001,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 35.5,
|
| 504 |
+
"grad_norm": 0.22914953529834747,
|
| 505 |
+
"learning_rate": 2.91e-05,
|
| 506 |
+
"loss": 0.0001,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 36.0,
|
| 511 |
+
"grad_norm": 0.2684924900531769,
|
| 512 |
+
"learning_rate": 2.8100000000000005e-05,
|
| 513 |
+
"loss": 0.0006,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 36.5,
|
| 518 |
+
"grad_norm": 0.29140010476112366,
|
| 519 |
+
"learning_rate": 2.7100000000000005e-05,
|
| 520 |
+
"loss": 0.0003,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 37.0,
|
| 525 |
+
"grad_norm": 0.39679646492004395,
|
| 526 |
+
"learning_rate": 2.61e-05,
|
| 527 |
+
"loss": 0.0003,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 37.5,
|
| 532 |
+
"grad_norm": 0.08030686527490616,
|
| 533 |
+
"learning_rate": 2.51e-05,
|
| 534 |
+
"loss": 0.0005,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 38.0,
|
| 539 |
+
"grad_norm": 0.06952106207609177,
|
| 540 |
+
"learning_rate": 2.41e-05,
|
| 541 |
+
"loss": 0.0,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 38.5,
|
| 546 |
+
"grad_norm": 0.03430015593767166,
|
| 547 |
+
"learning_rate": 2.3100000000000002e-05,
|
| 548 |
+
"loss": 0.0003,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 39.0,
|
| 553 |
+
"grad_norm": 0.31544044613838196,
|
| 554 |
+
"learning_rate": 2.2100000000000002e-05,
|
| 555 |
+
"loss": 0.0003,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 39.5,
|
| 560 |
+
"grad_norm": 0.20336127281188965,
|
| 561 |
+
"learning_rate": 2.11e-05,
|
| 562 |
+
"loss": 0.0,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 40.0,
|
| 567 |
+
"grad_norm": 0.028785208240151405,
|
| 568 |
+
"learning_rate": 2.01e-05,
|
| 569 |
+
"loss": 0.0005,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 40.5,
|
| 574 |
+
"grad_norm": 0.1161634773015976,
|
| 575 |
+
"learning_rate": 1.91e-05,
|
| 576 |
+
"loss": 0.0004,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 41.0,
|
| 581 |
+
"grad_norm": 0.11442368477582932,
|
| 582 |
+
"learning_rate": 1.81e-05,
|
| 583 |
+
"loss": 0.0,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 41.5,
|
| 588 |
+
"grad_norm": 0.010891513898968697,
|
| 589 |
+
"learning_rate": 1.7100000000000002e-05,
|
| 590 |
+
"loss": 0.0002,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 42.0,
|
| 595 |
+
"grad_norm": 0.2547859251499176,
|
| 596 |
+
"learning_rate": 1.6100000000000002e-05,
|
| 597 |
+
"loss": 0.0003,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 42.5,
|
| 602 |
+
"grad_norm": 0.006857059430330992,
|
| 603 |
+
"learning_rate": 1.51e-05,
|
| 604 |
+
"loss": 0.0002,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 43.0,
|
| 609 |
+
"grad_norm": 0.4951632618904114,
|
| 610 |
+
"learning_rate": 1.4099999999999999e-05,
|
| 611 |
+
"loss": 0.0003,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 43.5,
|
| 616 |
+
"grad_norm": 0.041084468364715576,
|
| 617 |
+
"learning_rate": 1.3100000000000002e-05,
|
| 618 |
+
"loss": 0.0004,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 44.0,
|
| 623 |
+
"grad_norm": 0.11843769252300262,
|
| 624 |
+
"learning_rate": 1.2100000000000001e-05,
|
| 625 |
+
"loss": 0.0,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 44.5,
|
| 630 |
+
"grad_norm": 0.03217615187168121,
|
| 631 |
+
"learning_rate": 1.11e-05,
|
| 632 |
+
"loss": 0.0005,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 45.0,
|
| 637 |
+
"grad_norm": 0.018256094306707382,
|
| 638 |
+
"learning_rate": 1.0100000000000002e-05,
|
| 639 |
+
"loss": 0.0,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 45.5,
|
| 644 |
+
"grad_norm": 0.042205873876810074,
|
| 645 |
+
"learning_rate": 9.100000000000001e-06,
|
| 646 |
+
"loss": 0.0,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 46.0,
|
| 651 |
+
"grad_norm": 0.05937511846423149,
|
| 652 |
+
"learning_rate": 8.1e-06,
|
| 653 |
+
"loss": 0.0004,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 46.5,
|
| 658 |
+
"grad_norm": 0.10673025250434875,
|
| 659 |
+
"learning_rate": 7.1e-06,
|
| 660 |
+
"loss": 0.0002,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 47.0,
|
| 665 |
+
"grad_norm": 0.1283574253320694,
|
| 666 |
+
"learning_rate": 6.1e-06,
|
| 667 |
+
"loss": 0.0002,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 47.5,
|
| 672 |
+
"grad_norm": 0.04566584527492523,
|
| 673 |
+
"learning_rate": 5.1e-06,
|
| 674 |
+
"loss": 0.0004,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 48.0,
|
| 679 |
+
"grad_norm": 0.08355249464511871,
|
| 680 |
+
"learning_rate": 4.1000000000000006e-06,
|
| 681 |
+
"loss": 0.0,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 48.5,
|
| 686 |
+
"grad_norm": 0.12251041829586029,
|
| 687 |
+
"learning_rate": 3.1e-06,
|
| 688 |
+
"loss": 0.0002,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 49.0,
|
| 693 |
+
"grad_norm": 0.16482886672019958,
|
| 694 |
+
"learning_rate": 2.1000000000000002e-06,
|
| 695 |
+
"loss": 0.0002,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 49.5,
|
| 700 |
+
"grad_norm": 0.008726044557988644,
|
| 701 |
+
"learning_rate": 1.1e-06,
|
| 702 |
+
"loss": 0.0002,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 50.0,
|
| 707 |
+
"grad_norm": 0.0353192463517189,
|
| 708 |
+
"learning_rate": 1.0000000000000001e-07,
|
| 709 |
+
"loss": 0.0002,
|
| 710 |
+
"step": 1000
|
| 711 |
+
}
|
| 712 |
+
],
|
| 713 |
+
"logging_steps": 10,
|
| 714 |
+
"max_steps": 1000,
|
| 715 |
+
"num_input_tokens_seen": 0,
|
| 716 |
+
"num_train_epochs": 50,
|
| 717 |
+
"save_steps": 10,
|
| 718 |
+
"stateful_callbacks": {
|
| 719 |
+
"TrainerControl": {
|
| 720 |
+
"args": {
|
| 721 |
+
"should_epoch_stop": false,
|
| 722 |
+
"should_evaluate": false,
|
| 723 |
+
"should_log": false,
|
| 724 |
+
"should_save": true,
|
| 725 |
+
"should_training_stop": true
|
| 726 |
+
},
|
| 727 |
+
"attributes": {}
|
| 728 |
+
}
|
| 729 |
+
},
|
| 730 |
+
"total_flos": 0.0,
|
| 731 |
+
"train_batch_size": 32,
|
| 732 |
+
"trial_name": null,
|
| 733 |
+
"trial_params": null
|
| 734 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:43657cd5a8f9e5df721e89daeddcaef29134d5f81701c9015e5e8a59cd6a2331
|
| 3 |
+
size 5777
|