File size: 800 Bytes
92d9d30
 
 
4b6addd
 
92d9d30
 
 
 
fd30bc1
92d9d30
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4b6addd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
---
tags:
- generated_from_trainer
- pythia
license: apache-2.0
---

## Model description

Sampling-based watermark distilled [Pythia 1.4B](https://huggingface.co/EleutherAI/pythia-1.4b) using the KGW \\(k=1, \gamma=0.25, \delta=2\\) watermarking strategy in the paper [On the Learnability of Watermarks for Language Models](https://arxiv.org/abs/2312.04469).

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 1e-05
- train_batch_size: 64
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 500
- num_epochs: 1.0

### Framework versions

- Transformers 4.29.2
- Pytorch 2.0.1+cu117
- Datasets 2.13.1
- Tokenizers 0.13.3