36 lines
965 B
Markdown
36 lines
965 B
Markdown
---
|
|
license: mit
|
|
datasets:
|
|
- wikipedia
|
|
language:
|
|
- ja
|
|
- en
|
|
---
|
|
|
|
# tiny-lm
|
|
|
|
This repository provides a tiny 16M parameters language model for debugging and testing purposes.
|
|
|
|
Trained on English and Japanese Wikipedia data.
|
|
|
|
## How to use
|
|
|
|
```python
|
|
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
|
|
|
|
model = AutoModelForCausalLM.from_pretrained("sbintuitions/tiny-lm", torch_dtype="auto")
|
|
tokenizer = AutoTokenizer.from_pretrained("sbintuitions/tiny-lm", use_fast=False)
|
|
generator = pipeline("text-generation", model=model, tokenizer=tokenizer)
|
|
print(generator("Hello", max_length=30, do_sample=True, top_k=100))
|
|
```
|
|
|
|
## Model architecture
|
|
A 4-layer, 512-hidden-size transformer-based language model.
|
|
|
|
## Training
|
|
The model was trained on English Wikipedia and Japanese Wikipedia to optimize a traditional language modelling objective for 25B tokens.
|
|
|
|
## License
|
|
[MIT License](https://huggingface.co/sbintuitions/tiny-lm/resolve/main/LICENSE)
|
|
|