Instructions to use simplescaling/step-conditional-control-old with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use simplescaling/step-conditional-control-old with Transformers:

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

pipe = pipeline("text-generation", model="simplescaling/step-conditional-control-old")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("simplescaling/step-conditional-control-old")
model = AutoModelForCausalLM.from_pretrained("simplescaling/step-conditional-control-old")
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 simplescaling/step-conditional-control-old with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/simplescaling/step-conditional-control-old

SGLang

How to use simplescaling/step-conditional-control-old 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 "simplescaling/step-conditional-control-old" \
    --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": "simplescaling/step-conditional-control-old",
		"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 "simplescaling/step-conditional-control-old" \
        --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": "simplescaling/step-conditional-control-old",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use simplescaling/step-conditional-control-old with Docker Model Runner:
```
docker model run hf.co/simplescaling/step-conditional-control-old
```

Muennighoff commited on Feb 1, 2025

Commit

0221ed8

verified ·

1 Parent(s): 4532948

Update README.md

Browse files

Files changed (1) hide show

README.md +11 -34

README.md CHANGED Viewed

@@ -1,58 +1,35 @@
 ---
 base_model: Qwen/Qwen2.5-32B-Instruct
 library_name: transformers
-model_name: Qwen2.5-32B-Instruct-20250104_095632
 tags:
 - generated_from_trainer
 - trl
 - sft
-licence: license
 ---
-# Model Card for Qwen2.5-32B-Instruct-20250104_095632
-This model is a fine-tuned version of [Qwen/Qwen2.5-32B-Instruct](https://huggingface.co/Qwen/Qwen2.5-32B-Instruct).
-It has been trained using [TRL](https://github.com/huggingface/trl).
-## Quick start
-```python
-from transformers import pipeline
-question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
-generator = pipeline("text-generation", model="qfq/Qwen2.5-32B-Instruct-20250104_095632", device="cuda")
-output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
-print(output["generated_text"])
-```
-## Training procedure
 [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/hashimoto-group/o1/runs/rkflww8e)
-This model was trained with SFT.
-### Framework versions
 - TRL: 0.13.0
-- Transformers: 4.47.1
 - Pytorch: 2.3.1
 - Datasets: 3.0.1
 - Tokenizers: 0.21.0
-## Citations
-Cite TRL as:
 ```bibtex
-@misc{vonwerra2022trl,
-	title        = {{TRL: Transformer Reinforcement Learning}},
-	author       = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallouédec},
-	year         = 2020,
-	journal      = {GitHub repository},
-	publisher    = {GitHub},
-	howpublished = {\url{https://github.com/huggingface/trl}}
-}
 ```

 ---
 base_model: Qwen/Qwen2.5-32B-Instruct
 library_name: transformers
+model_name: step-conditional-control
 tags:
 - generated_from_trainer
 - trl
 - sft
+license: apache-2.0
 ---
+# Model Summary
+- **Repository:** [simplescaling/s1](https://github.com/simplescaling/s1)
+- **Paper:** TODO
+# Use
+This is an older step-conditional control model for our paper used only for the Discussion setion. You can evaluate using the information [here](https://github.com/simplescaling/s1?tab=readme-ov-file#evaluation).
+# Training information
 [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/hashimoto-group/o1/runs/rkflww8e)
 - TRL: 0.13.0
+- Transformers: 4.48.0
 - Pytorch: 2.3.1
 - Datasets: 3.0.1
 - Tokenizers: 0.21.0
+# Citation
 ```bibtex
+TODO
 ```