# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained("EpistemeAI/Hercules1-8B-E2B-it")
model = AutoModelForCausalLM.from_pretrained("EpistemeAI/Hercules1-8B-E2B-it")
messages = [
    {"role": "user", "content": "Write me a Python function to calculate the nth fibonacci number.<think></think>"},
]
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=1024)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

lm eval result:

Tasks	Version	Filter	n-shot	Metric		Value
gpqa_diamond_zeroshot	1	none	0	acc/acc_norm	↑	0.3097

Uploaded finetuned model

Developed by: EpistemeAI
License: apache-2.0
Finetuned from model : unsloth/gemma-3n-e2b-it-unsloth-bnb-4bit

This gemma3n model was trained 2x faster with Unsloth and Huggingface's TRL library.