Update README.md
Browse files
README.md
CHANGED
|
@@ -12,35 +12,28 @@ tags:
|
|
| 12 |
|
| 13 |
|
| 14 |
# GOAT-70B-Storytelling model
|
| 15 |
-
|
| 16 |
GOAT-70B-Storytelling model trained by GOAT.AI lab as a core model for an autonomous story-writing agent.
|
| 17 |
|
| 18 |
# GOAT-Storytelling-Agent
|
| 19 |
-
|
| 20 |
This agent facilitates the generation of high-quality, cohesive, and captivating narratives, including stories and books. It achieves this by utilizing inputs such as plot outlines, character profiles, their interrelationships, and other relevant details. Examples are provided below.
|
| 21 |
|
| 22 |
# Model description
|
| 23 |
-
|
| 24 |
- **Base Architecture:** LLaMA 2 70B
|
| 25 |
- **License:** llama2
|
| 26 |
- **Context window length:** 4096 tokens
|
| 27 |
|
| 28 |
### Training details
|
| 29 |
-
|
| 30 |
Training was performed on a GPU cluster of 64xH100s and FSDP ZeRO-3 sharding is employed. We instruction finetune on a dataset of 18K examples for one epoch with batch size of 336, AdamW optimizer with learning rate 1e-5.
|
| 31 |
|
| 32 |
### Learn more
|
| 33 |
-
|
| 34 |
- **Blogpost:** [GOAT-Storytelling: Arbitrarily Long Story Writing Agent](https://www.blog.goat.ai/goat-st/)
|
| 35 |
- **GitHub:** [here](https://github.com/GOAT-AI-lab/GOAT-Storytelling-Agent)
|
| 36 |
- **Generated examples:** [here](https://huggingface.co/datasets/GOAT-AI/generated-novels/tree/main/generated-books)
|
| 37 |
|
| 38 |
## Uses
|
| 39 |
-
|
| 40 |
The main purpose of GOAT-70B-Storytelling is to generate books, novels, movie scripts and etc. as an agent in coping with our GOAT-Storytelling-Agent. It is specifically designed for storywriters.
|
| 41 |
|
| 42 |
## Usage
|
| 43 |
-
|
| 44 |
Usage can be either self-hosted via `transformers` or used with Spaces
|
| 45 |
|
| 46 |
```python
|
|
@@ -69,11 +62,9 @@ novel_scenes = goat.generate_story('treasure hunt in a jungle', form='novel')
|
|
| 69 |
```
|
| 70 |
|
| 71 |
## License
|
| 72 |
-
|
| 73 |
GOAT-70B-Storytelling model is based on [Meta's LLaMA-2-70b-hf](https://huggingface.co/meta-llama/Llama-2-70b-hf), and using own datasets.
|
| 74 |
|
| 75 |
GOAT-70B-Storytelling model weights are available under LLAMA-2 license.
|
| 76 |
|
| 77 |
### Risks and Biases
|
| 78 |
-
|
| 79 |
GOAT-70B-Storytelling model can produce factually incorrect output and should not be relied on to deliver factually accurate information. Therefore, the GOAT-70B-Storytelling model could possibly generate wrong, biased, or otherwise offensive outputs.
|
|
|
|
| 12 |
|
| 13 |
|
| 14 |
# GOAT-70B-Storytelling model
|
|
|
|
| 15 |
GOAT-70B-Storytelling model trained by GOAT.AI lab as a core model for an autonomous story-writing agent.
|
| 16 |
|
| 17 |
# GOAT-Storytelling-Agent
|
|
|
|
| 18 |
This agent facilitates the generation of high-quality, cohesive, and captivating narratives, including stories and books. It achieves this by utilizing inputs such as plot outlines, character profiles, their interrelationships, and other relevant details. Examples are provided below.
|
| 19 |
|
| 20 |
# Model description
|
|
|
|
| 21 |
- **Base Architecture:** LLaMA 2 70B
|
| 22 |
- **License:** llama2
|
| 23 |
- **Context window length:** 4096 tokens
|
| 24 |
|
| 25 |
### Training details
|
|
|
|
| 26 |
Training was performed on a GPU cluster of 64xH100s and FSDP ZeRO-3 sharding is employed. We instruction finetune on a dataset of 18K examples for one epoch with batch size of 336, AdamW optimizer with learning rate 1e-5.
|
| 27 |
|
| 28 |
### Learn more
|
|
|
|
| 29 |
- **Blogpost:** [GOAT-Storytelling: Arbitrarily Long Story Writing Agent](https://www.blog.goat.ai/goat-st/)
|
| 30 |
- **GitHub:** [here](https://github.com/GOAT-AI-lab/GOAT-Storytelling-Agent)
|
| 31 |
- **Generated examples:** [here](https://huggingface.co/datasets/GOAT-AI/generated-novels/tree/main/generated-books)
|
| 32 |
|
| 33 |
## Uses
|
|
|
|
| 34 |
The main purpose of GOAT-70B-Storytelling is to generate books, novels, movie scripts and etc. as an agent in coping with our GOAT-Storytelling-Agent. It is specifically designed for storywriters.
|
| 35 |
|
| 36 |
## Usage
|
|
|
|
| 37 |
Usage can be either self-hosted via `transformers` or used with Spaces
|
| 38 |
|
| 39 |
```python
|
|
|
|
| 62 |
```
|
| 63 |
|
| 64 |
## License
|
|
|
|
| 65 |
GOAT-70B-Storytelling model is based on [Meta's LLaMA-2-70b-hf](https://huggingface.co/meta-llama/Llama-2-70b-hf), and using own datasets.
|
| 66 |
|
| 67 |
GOAT-70B-Storytelling model weights are available under LLAMA-2 license.
|
| 68 |
|
| 69 |
### Risks and Biases
|
|
|
|
| 70 |
GOAT-70B-Storytelling model can produce factually incorrect output and should not be relied on to deliver factually accurate information. Therefore, the GOAT-70B-Storytelling model could possibly generate wrong, biased, or otherwise offensive outputs.
|