Selene Mini Resources
Use these links to get started with our open source eval model.
Use Selene Mini through Hugging Face Transformers.
Get API Access
Copied to clipboard
from transformers import AutoModelForCausalLM, AutoTokenizer
device = "cuda" # the device to load the model onto
model_id = "AtlaAI/Selene-1-Mini-Llama-3.1-8B"
model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto")
tokenizer = AutoTokenizer.from_pretrained(model_id)
prompt = "I heard you can evaluate my responses?" # replace with your eval prompt
messages = [{"role": "user", "content": prompt}]
text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
model_inputs = tokenizer([text], return_tensors="pt").to(device)
generated_ids = model.generate(model_inputs.input_ids, max_new_tokens=512, do_sample=True)
generated_ids = [output_ids[len(input_ids):] for input_ids, output_ids in
zip(model_inputs.input_ids, generated_ids)]
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]Selene models
Explore the right size and implementation methods
for your evaluation needs.
for your evaluation needs.
Light & Open Source
Selene 1 Mini
The best evaluation model of its size (8B). Excels particularly
in scoring tasks.
Powerfully accurate, built for production
Selene 1
The best model for evaluation on the market. Capable of accurately judging a wide variety of evaluation tasks, as well as adapting to custom evaluation criteria. Available via API.
Cost
Intelligence
Evaluate with Selene 1
Use our API to integrate Selene 1, our latest and most powerful evaluation model, into your production workflows.
