Instructions to use transformers-community/contrastive-search with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use transformers-community/contrastive-search with Transformers:

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

pipe = pipeline("text-generation", model="transformers-community/contrastive-search")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("transformers-community/contrastive-search")
model = AutoModelForCausalLM.from_pretrained("transformers-community/contrastive-search")
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]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use transformers-community/contrastive-search with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/transformers-community/contrastive-search

SGLang

How to use transformers-community/contrastive-search 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 "transformers-community/contrastive-search" \
    --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": "transformers-community/contrastive-search",
		"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 "transformers-community/contrastive-search" \
        --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": "transformers-community/contrastive-search",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use transformers-community/contrastive-search with Docker Model Runner:
```
docker model run hf.co/transformers-community/contrastive-search
```

Description

Implementation of Contrastive Search, a decoding strategy that jointly optimizes model confidence and a degeneration penalty to produce fluent, coherent, and low-repetition text. At each step, the model considers the top-k candidate tokens and selects the one maximizing:

score(v) = (1 - alpha) * p(v | context) - alpha * max_cosine_similarity(h_v, H_context)

where alpha is the trade-off between confidence and the cosine-similarity-based penalty.

This strategy typically:

Reduces repetition compared to greedy/beam search
Preserves semantic coherence better than pure sampling

Base model

Qwen/Qwen2.5-0.5B-Instruct (example)

Model compatibility

Decoder and encoder-decoder transformer models for causal LM

Additional Arguments

top_k (int): Number of candidate tokens to consider each step (e.g., 4)
penalty_alpha (float): Weight of the degeneration penalty (e.g., 0.6)

Tips:

Larger top_k explores more candidates but increases compute
penalty_alpha in [0.3, 0.8] often works well; 0.0 reduces to greedy

Output Type changes

(none) — returns the same structure as standard transformers generation

Example usage

from transformers import AutoModelForCausalLM, AutoTokenizer, infer_device

device = infer_device()

model_id = "Qwen/Qwen2.5-0.5B-Instruct"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto").to(device)

inputs = tokenizer(["DeepMind Company is"], return_tensors="pt").to(device)

# Contrastive search
gen_out = model.generate(
    **inputs,
    custom_generate="contrastive_search",
    penalty_alpha=0.6,
    top_k=4,
    max_new_tokens=128,
    trust_remote_code=True,
)

print(tokenizer.batch_decode(gen_out, skip_special_tokens=True))

Downloads last month: 12

Safetensors

Model size

0.8B params

Tensor type

BF16

Collection including transformers-community/contrastive-search

Custom generation methods - Tutorials

Collection

Custom generation methods that are maintained by the `transformers` team -- tutorials and methods that were previously in the library • 5 items • Updated Sep 10, 2025 • 2