Instructions to use CohereLabs/tiny-aya-global with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use CohereLabs/tiny-aya-global with Transformers:

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

pipe = pipeline("text-generation", model="CohereLabs/tiny-aya-global")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("CohereLabs/tiny-aya-global")
model = AutoModelForCausalLM.from_pretrained("CohereLabs/tiny-aya-global")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Inference
HuggingChat
Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use CohereLabs/tiny-aya-global with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "CohereLabs/tiny-aya-global"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "CohereLabs/tiny-aya-global",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/CohereLabs/tiny-aya-global

SGLang

How to use CohereLabs/tiny-aya-global 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 "CohereLabs/tiny-aya-global" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "CohereLabs/tiny-aya-global",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "CohereLabs/tiny-aya-global" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "CohereLabs/tiny-aya-global",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use CohereLabs/tiny-aya-global with Docker Model Runner:
```
docker model run hf.co/CohereLabs/tiny-aya-global
```

Also providing MLX quants?

by bibproj - opened Feb 21

Discussion

bibproj

Feb 21

Hi CohereLabs

Congratulations to get 67 languages into such a 'tiny' model!

Thank you for already providing the GGUF quants of your new model. 👍 Would you perhaps consider to also provide MLX quants for people with Apple Macs and iPhones?

I would suggest BF16 for Apple Macs, and 4-bit versions for iPhones.

# Install MLX
pip install -U mlx-lm

# BF16 quants for Apple Macs
mlx_lm.convert --hf-path CohereLabs/tiny-aya-global --mlx-path tiny-aya-global-mlx-bf16 --dtype bfloat16
mlx_lm.convert --hf-path CohereLabs/tiny-aya-water --mlx-path tiny-aya-water-mlx-bf16 --dtype bfloat16
mlx_lm.convert --hf-path CohereLabs/tiny-aya-earth --mlx-path tiny-aya-earth-mlx-bf16 --dtype bfloat16
mlx_lm.convert --hf-path CohereLabs/tiny-aya-fire --mlx-path tiny-aya-fire-mlx-bf16 --dtype bfloat16

# 4-bit quants for Apple iPhones
mlx_lm.convert --hf-path CohereLabs/tiny-aya-global --mlx-path tiny-aya-global-mlx-4bit -q --q-bits 4
mlx_lm.convert --hf-path CohereLabs/tiny-aya-water --mlx-path tiny-aya-water-mlx-4bit -q --q-bits 4
mlx_lm.convert --hf-path CohereLabs/tiny-aya-earth --mlx-path tiny-aya-earth-mlx-4bit -q --q-bits 4
mlx_lm.convert --hf-path CohereLabs/tiny-aya-fire --mlx-path tiny-aya-fire-mlx-4bit -q --q-bits 4

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment