Instructions to use ThingAI/Quark-270m-Base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ThingAI/Quark-270m-Base with Transformers:

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

pipe = pipeline("text-generation", model="ThingAI/Quark-270m-Base", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("ThingAI/Quark-270m-Base", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use ThingAI/Quark-270m-Base with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ThingAI/Quark-270m-Base"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ThingAI/Quark-270m-Base",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/ThingAI/Quark-270m-Base

SGLang

How to use ThingAI/Quark-270m-Base 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 "ThingAI/Quark-270m-Base" \
    --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": "ThingAI/Quark-270m-Base",
		"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 "ThingAI/Quark-270m-Base" \
        --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": "ThingAI/Quark-270m-Base",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use ThingAI/Quark-270m-Base with Docker Model Runner:
```
docker model run hf.co/ThingAI/Quark-270m-Base
```

ThingsAI commited on 6 days ago

Commit

c58e770

verified ·

1 Parent(s): 960cc50

Upload modeling_quark.py with huggingface_hub

Browse files

Files changed (1) hide show

modeling_quark.py +1 -1

modeling_quark.py CHANGED Viewed

@@ -39,7 +39,7 @@ class QuarkAttention(nn.Module):
         q=self.q_proj(x).view(B,T,self.nh,self.hd).transpose(1,2)
         k=self.k_proj(x).view(B,T,self.nkv,self.hd).transpose(1,2)
         v=self.v_proj(x).view(B,T,self.nkv,self.hd).transpose(1,2)
-        q,k=self.rope(q,k)
         if self.ng>1: k=k.repeat_interleave(self.ng,1); v=v.repeat_interleave(self.ng,1)
         return self.o_proj(F.scaled_dot_product_attention(q,k,v,is_causal=True).transpose(1,2).contiguous().view(B,T,-1))

         q=self.q_proj(x).view(B,T,self.nh,self.hd).transpose(1,2)
         k=self.k_proj(x).view(B,T,self.nkv,self.hd).transpose(1,2)
         v=self.v_proj(x).view(B,T,self.nkv,self.hd).transpose(1,2)
+        q,k=self.rope(q,k); q,k=q.to(v.dtype),k.to(v.dtype)
         if self.ng>1: k=k.repeat_interleave(self.ng,1); v=v.repeat_interleave(self.ng,1)
         return self.o_proj(F.scaled_dot_product_attention(q,k,v,is_causal=True).transpose(1,2).contiguous().view(B,T,-1))