Instructions to use silx-ai/Quasar-Preview with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use silx-ai/Quasar-Preview with Transformers:

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

pipe = pipeline("text-generation", model="silx-ai/Quasar-Preview", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("silx-ai/Quasar-Preview", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use silx-ai/Quasar-Preview with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/silx-ai/Quasar-Preview

SGLang

How to use silx-ai/Quasar-Preview 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 "silx-ai/Quasar-Preview" \
    --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": "silx-ai/Quasar-Preview",
		"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 "silx-ai/Quasar-Preview" \
        --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": "silx-ai/Quasar-Preview",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use silx-ai/Quasar-Preview with Docker Model Runner:
```
docker model run hf.co/silx-ai/Quasar-Preview
```

Missing Folder or Package

by Akicou - opened 15 days ago

Discussion

Akicou

15 days ago

This comment has been hidden (marked as Resolved)

Akicou changed discussion status to closed 14 days ago

Akicou changed discussion status to open 14 days ago

Akicou

14 days ago

Hi, I’m working on llama.cpp (my fork specifically) support for Quasar-Preview.

The model config enables hybrid layers with Quasar/Raven/GLA branches, and modeling_quasar_long.py imports:

   from raven.layers.raven import RavenAttention

However, the HF repo does not seem to include a raven/ folder or raven.layers.raven implementation. The
code also checks for a local raven/ directory and raises if missing.

Could you please clarify where the RavenAttention source is available? Is it supposed to be bundled in this
repo, installed from another package, or disabled for inference?

Without the Raven implementation, exact faithful inference/llama.cpp support is blocked. Thanks!

mira1eval

13 days ago

•

edited 13 days ago

Also why add the raven folder then delete it? could we please get an explanation. we just want to understand

eyad-silx

SILX org 11 days ago

Hi, Raven attention is already supported in a public repo. You can check it here:

https://github.com/goombalab/raven

We were uploading Raven to Hugging Face, but that created some import issues

We are creating a specific repo for evaluation, but the link above is the main Raven repo from the original authors.

mira1eval

11 days ago

Hi, Raven attention is already supported in a public repo. You can check it here:

https://github.com/goombalab/raven

We were uploading Raven to Hugging Face, but that created some import issues

We are creating a specific repo for evaluation, but the link above is the main Raven repo from the original authors.

Thank you for clarifying and for sharing the Raven repository. We had seen the upstream repo, but were unsure whether there were any Quasar Preview specific changes, so we wanted to be cautious before relying on it.

That makes sense now. Thanks again for the quick response.

Akicou changed discussion status to closed 11 days ago

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