Instructions to use uzlm/alloma-8B-Instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use uzlm/alloma-8B-Instruct with Transformers:

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

pipe = pipeline("text-generation", model="uzlm/alloma-8B-Instruct")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("uzlm/alloma-8B-Instruct")
model = AutoModelForCausalLM.from_pretrained("uzlm/alloma-8B-Instruct")
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]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use uzlm/alloma-8B-Instruct with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "uzlm/alloma-8B-Instruct"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "uzlm/alloma-8B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/uzlm/alloma-8B-Instruct

SGLang

How to use uzlm/alloma-8B-Instruct 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 "uzlm/alloma-8B-Instruct" \
    --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": "uzlm/alloma-8B-Instruct",
		"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 "uzlm/alloma-8B-Instruct" \
        --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": "uzlm/alloma-8B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use uzlm/alloma-8B-Instruct with Docker Model Runner:
```
docker model run hf.co/uzlm/alloma-8B-Instruct
```

alloma-8B-Instruct release

by UAzimov - opened Jan 7

Discussion

UAzimov

Jan 7

Hey, just wanted to say great job on alloma-8B-Instruct. The custom tokenizer getting down to 1.7 tokens per Uzbek word is really impressive, and those improvements on translation and sentiment benchmarks are solid. Nice to see more good Uzbek models coming out.
I noticed you used my uzbek-instruct-llm dataset for the SFT part (thanks for that, it shows up on my dataset page). Would you mind adding a quick mention in the model card? Something simple like "fine-tuned with UAzimov/uzbek-instruct-llm" or just putting it in the datasets YAML field would be helpful. No big deal if not. It would really help the dataset gaining popularity and expanding Uzbek community in HF.
Thanks again, looking forward to trying it out.

bxod

Uzbek LLM Lab org Jan 9

Hello, I hope you're doing great. I noticed your message just now.
First, thank you for putting effort on open-sourcing datasets for Uzbek NLP.

I surely added your dataset in the model card, that's why it poped up in your dashboard.
I also mentioned it in my Linkedin post previously. If you would like me to, I can also mention it inside the "Acknowledgements" part of our models.
Happy to hear from you, hope to stay connected!

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