Instructions to use NoesisLab/Asterisk-135M with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use NoesisLab/Asterisk-135M with Transformers:

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

pipe = pipeline("text-generation", model="NoesisLab/Asterisk-135M", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("NoesisLab/Asterisk-135M", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use NoesisLab/Asterisk-135M with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "NoesisLab/Asterisk-135M"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "NoesisLab/Asterisk-135M",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/NoesisLab/Asterisk-135M

SGLang

How to use NoesisLab/Asterisk-135M 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 "NoesisLab/Asterisk-135M" \
    --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": "NoesisLab/Asterisk-135M",
		"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 "NoesisLab/Asterisk-135M" \
        --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": "NoesisLab/Asterisk-135M",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use NoesisLab/Asterisk-135M with Docker Model Runner:
```
docker model run hf.co/NoesisLab/Asterisk-135M
```

OzTianlu commited on Jan 19

Commit

5bd36fb

verified ·

1 Parent(s): 7e01414

Update README.md

Browse files

Files changed (1) hide show

README.md +1 -2

README.md CHANGED Viewed

@@ -59,7 +59,6 @@ output = gate * ASPP(x) + (1-gate) * Attention(x)
 ```python
 class ASPPOperator:
     """
-    Simplified ASPP without neighbor gathering to reduce overfitting
     Forward pass:
     1. Optional dimensionality reduction: h_t = down_proj(hidden_states)
@@ -246,7 +245,7 @@ Exponential convergence to fixed points with rate c=0.76 under Lipschitz continu
 ### Turing Completeness
 Proven via cyclic tag system simulation - ASPP can compute any Turing-computable function given sufficient depth.
-**Implementation Note**: This implementation simplifies theoretical ASPP to point-wise evolution (no neighbor gathering) to reduce overfitting while maintaining iterative refinement benefits.
 ## Files in Checkpoint

 ```python
 class ASPPOperator:
     """
     Forward pass:
     1. Optional dimensionality reduction: h_t = down_proj(hidden_states)
 ### Turing Completeness
 Proven via cyclic tag system simulation - ASPP can compute any Turing-computable function given sufficient depth.
+**Implementation Note**: This implementation simplifies theoretical ASPP to point-wise evolution to reduce overfitting while maintaining iterative refinement benefits.
 ## Files in Checkpoint