Instructions to use goldfish-models/kaa_latn_full with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use goldfish-models/kaa_latn_full with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="goldfish-models/kaa_latn_full")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("goldfish-models/kaa_latn_full")
model = AutoModelForCausalLM.from_pretrained("goldfish-models/kaa_latn_full")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use goldfish-models/kaa_latn_full with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "goldfish-models/kaa_latn_full"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "goldfish-models/kaa_latn_full",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/goldfish-models/kaa_latn_full

SGLang

How to use goldfish-models/kaa_latn_full with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "goldfish-models/kaa_latn_full" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "goldfish-models/kaa_latn_full",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "goldfish-models/kaa_latn_full" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "goldfish-models/kaa_latn_full",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use goldfish-models/kaa_latn_full with Docker Model Runner:
```
docker model run hf.co/goldfish-models/kaa_latn_full
```

kaa_latn_full

Goldfish is a suite of monolingual language models trained for 350 languages. This model is the Kara-Kalpak (Latin script) model trained on 165MB of data (all our data in the language), after accounting for an estimated byte premium of 1.23; content-matched text in Kara-Kalpak takes on average 1.23x as many UTF-8 bytes to encode as English. The Goldfish models are trained primarily for comparability across languages and for low-resource languages; Goldfish performance for high-resource languages is not designed to be comparable with modern large language models (LLMs).

Note: This language is available in Goldfish with other scripts (writing systems). See: kaa_cyrl.

Note: kaa_latn is an individual language code. It is not contained in any macrolanguage codes contained in Goldfish (for script latn).

All training and hyperparameter details are in our paper, Goldfish: Monolingual Language Models for 350 Languages (Chang et al., 2024).

Training code and sample usage: https://github.com/tylerachang/goldfish

Sample usage also in this Google Colab: link

Model details:

To access all Goldfish model details programmatically, see https://github.com/tylerachang/goldfish/blob/main/model_details.json. All models are trained with a [CLS] (same as [BOS]) token prepended, and a [SEP] (same as [EOS]) token separating sequences. For best results, make sure that [CLS] is prepended to your input sequence (see sample usage linked above)! Details for this model specifically:

Architecture: gpt2
Parameters: 124770816
Maximum sequence length: 512 tokens
Training text data (raw): 202.78MB
Training text data (byte premium scaled): 165.225MB
Training tokens: 38767104 (x10 epochs)
Vocabulary size: 50000
Compute cost: 1.97850849214464e+17 FLOPs or ~18.7 NVIDIA A6000 GPU hours

Training datasets (percentages prior to deduplication):

97.52670%: MADLAD-400 (CommonCrawl)
1.65677%: Wikipedia 2023/08
0.81649%: Glot500, including Tatoeba, Wikipedia Hugging Face
0.00004%: Tatoeba

Citation

If you use this model, please cite:

@article{chang-etal-2024-goldfish,
  title={Goldfish: Monolingual Language Models for 350 Languages},
  author={Chang, Tyler A. and Arnett, Catherine and Tu, Zhuowen and Bergen, Benjamin K.},
  journal={Preprint},
  year={2024},
  url={https://www.arxiv.org/abs/2408.10441},
}

Downloads last month: 7

Datasets used to train goldfish-models/kaa_latn_full

Space using goldfish-models/kaa_latn_full 1

Paper for goldfish-models/kaa_latn_full

Goldfish: Monolingual Language Models for 350 Languages

Paper • 2408.10441 • Published Aug 19, 2024