Instructions to use robertmyers/sybil-0.1-base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use robertmyers/sybil-0.1-base with Transformers:

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

pipe = pipeline("text-generation", model="robertmyers/sybil-0.1-base")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("robertmyers/sybil-0.1-base")
model = AutoModelForCausalLM.from_pretrained("robertmyers/sybil-0.1-base")
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 Settings

vLLM

How to use robertmyers/sybil-0.1-base with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "robertmyers/sybil-0.1-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": "robertmyers/sybil-0.1-base",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/robertmyers/sybil-0.1-base

SGLang

How to use robertmyers/sybil-0.1-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 "robertmyers/sybil-0.1-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": "robertmyers/sybil-0.1-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 "robertmyers/sybil-0.1-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": "robertmyers/sybil-0.1-base",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use robertmyers/sybil-0.1-base with Docker Model Runner:
```
docker model run hf.co/robertmyers/sybil-0.1-base
```

Sybil 0.1

Sybil 0.1 is an experimental base language model derived from GLM-5.2. It is intended for researchers and builders who want a GLM-5.2-lineage foundation checkpoint for further evaluation, adaptation, and text-generation experiments.

This is a base model, not an instruction-tuned assistant. Prompts should be designed accordingly, and downstream behavior depends heavily on sampling settings, prompt format, and any additional fine-tuning or alignment applied by the user.

Model Lineage

Model name: Sybil 0.1
Base lineage: GLM-5.2
Model type: Base text-generation model
Languages: English and Chinese
License: MIT
Library: Transformers-compatible GLM architecture

Sybil 0.1 preserves the GLM-5.2 model family as its foundation while being packaged as a separate base checkpoint for experimentation.

Intended Use

Sybil 0.1 is suitable for:

Research on GLM-5.2-derived base models
Continued pretraining or supervised adaptation
Prompting and sampling experiments
Evaluation of base-model behavior before downstream tuning

It is not presented as a production assistant, safety-filtered chatbot, or drop-in replacement for an instruction-tuned model.

Usage Notes

Use an inference stack that supports the GLM-5.2 architecture and the model's checkpoint format. Because this is a base model, start with conservative generation settings and evaluate outputs carefully for your use case.

Example areas to validate before deployment include instruction following, factuality, multilingual behavior, long-context behavior, refusal behavior, domain-specific accuracy, and safety characteristics.

Evaluation

No Sybil 0.1-specific benchmark results are claimed in this README. Users should run their own evaluations on the tasks, prompts, and inference settings relevant to their intended use.

Attribution

Sybil 0.1 is derived from GLM-5.2. For details on the original GLM-5 model family, see the GLM-5 technical report and upstream project materials.

Downloads last month: -

Safetensors

Model size

753B params

Tensor type

F32

BF16

F8_E4M3