Instructions to use junshim/When2Think-ThinkOnly-1.5B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use junshim/When2Think-ThinkOnly-1.5B with Transformers:

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

pipe = pipeline("text-generation", model="junshim/When2Think-ThinkOnly-1.5B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("junshim/When2Think-ThinkOnly-1.5B")
model = AutoModelForCausalLM.from_pretrained("junshim/When2Think-ThinkOnly-1.5B", device_map="auto")
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=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use junshim/When2Think-ThinkOnly-1.5B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "junshim/When2Think-ThinkOnly-1.5B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "junshim/When2Think-ThinkOnly-1.5B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/junshim/When2Think-ThinkOnly-1.5B

SGLang

How to use junshim/When2Think-ThinkOnly-1.5B 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 "junshim/When2Think-ThinkOnly-1.5B" \
    --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": "junshim/When2Think-ThinkOnly-1.5B",
		"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 "junshim/When2Think-ThinkOnly-1.5B" \
        --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": "junshim/When2Think-ThinkOnly-1.5B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use junshim/When2Think-ThinkOnly-1.5B with Docker Model Runner:
```
docker model run hf.co/junshim/When2Think-ThinkOnly-1.5B
```

junshim commited on Sep 19

Commit

60a33e1

verified ·

1 Parent(s): bfd1691

Update README.md

Browse files

Files changed (1) hide show

README.md +6 -1

README.md CHANGED Viewed

@@ -60,6 +60,11 @@ Importance sampling is used during post-training to balance exploration between
 - Use **When2Think-1.5B** for adaptive hybrid reasoning. It learns when direct answering is sufficient and when explicit reasoning is needed, providing the intended balance between accuracy and token efficiency.
 - Use **ThinkOnly-1.5B** when explicit reasoning should always be used, or when studying difficulty-aware reasoning-depth control without hybrid Think/NoThink mode selection.
 ### Model Sources
 <!-- Provide the basic links for the model. -->
@@ -113,7 +118,7 @@ Use the code below to get started with the model.
 ```py
 from transformers import pipeline
-model_path = "junshim/When2Think-1.5B"
 prompt = "Find the value of $x$ that satisfies the equation $4x+5 = 6x+7$."
 messages = [
     {"role": "system", "content": "You are a helpful assistant."},

 - Use **When2Think-1.5B** for adaptive hybrid reasoning. It learns when direct answering is sufficient and when explicit reasoning is needed, providing the intended balance between accuracy and token efficiency.
 - Use **ThinkOnly-1.5B** when explicit reasoning should always be used, or when studying difficulty-aware reasoning-depth control without hybrid Think/NoThink mode selection.
+| Model | Whether to Think | How Deeply to Think |
+|---|---|---|
+| **When2Think-1.5B** | Learned through IS-based hybrid exploration | Controlled through IDAC |
+| **ThinkOnly-1.5B** | Always Think | Controlled through IDAC |
 ### Model Sources
 <!-- Provide the basic links for the model. -->
 ```py
 from transformers import pipeline
+model_path = "junshim/When2Think-ThinkOnly-1.5B"
 prompt = "Find the value of $x$ that satisfies the equation $4x+5 = 6x+7$."
 messages = [
     {"role": "system", "content": "You are a helpful assistant."},