Instructions to use ELBennett/Weevil-Recurrent-16M-Base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ELBennett/Weevil-Recurrent-16M-Base with Transformers:

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

pipe = pipeline("text-generation", model="ELBennett/Weevil-Recurrent-16M-Base", trust_remote_code=True)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("ELBennett/Weevil-Recurrent-16M-Base", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use ELBennett/Weevil-Recurrent-16M-Base with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ELBennett/Weevil-Recurrent-16M-Base"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ELBennett/Weevil-Recurrent-16M-Base",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/ELBennett/Weevil-Recurrent-16M-Base

SGLang

How to use ELBennett/Weevil-Recurrent-16M-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 "ELBennett/Weevil-Recurrent-16M-Base" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ELBennett/Weevil-Recurrent-16M-Base",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "ELBennett/Weevil-Recurrent-16M-Base" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ELBennett/Weevil-Recurrent-16M-Base",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use ELBennett/Weevil-Recurrent-16M-Base with Docker Model Runner:
```
docker model run hf.co/ELBennett/Weevil-Recurrent-16M-Base
```

Weevil-Recurrent-16M-Base

File size: 932 Bytes

f1b1d76

{
  "architectures": [
    "WeevilRecurrentForCausalLM"
  ],
  "auto_map": {
    "AutoConfig": "configuration_weevil_recurrent.WeevilRecurrentConfig",
    "AutoModel": "modeling_weevil_recurrent.WeevilRecurrentModel",
    "AutoModelForCausalLM": "modeling_weevil_recurrent.WeevilRecurrentForCausalLM"
  },
  "bos_token_id": 50281,
  "chunk_size": 32,
  "coda_layers": 2,
  "compute_dtype": "bfloat16",
  "conv_kernel": 4,
  "core_passes": 16,
  "dtype": "float32",
  "embedding_dim": 128,
  "eos_token_id": 50279,
  "head_dim": 16,
  "hidden_size": 256,
  "intermediate_size": 1024,
  "max_position_embeddings": 8192,
  "model_type": "weevil_recurrent",
  "norm_eps": 1e-06,
  "num_heads": 16,
  "pad_token_id": 50283,
  "pause_token_id": 50285,
  "prelude_layers": 2,
  "recurrent_layers": 4,
  "tie_word_embeddings": false,
  "transformers_version": "5.19.0",
  "unk_token_id": 50280,
  "use_cache": true,
  "vocab_size": 50368
}