ICONN 1
Introducing ICONN 1 Mini Beta, a cutting-edge open-source AI model with just 7 billion parameters — designed for natural, human-like language understanding and generation. Despite its compact size, it delivers powerful performance through efficient architecture and careful tuning. ICONN 1 Mini Beta represents the next step in accessible, conversational AI.
Developed entirely from scratch, ICONN-1-Mini-Beta is based on a new ICONN framework and comprises 7 billion parameters.
ICONN-1 is released in three distinct forms to serve different application needs:
- ICONN-1-Mini-Beta(This model) is a small 7B model trained for a lightweight alternative to ICONN 1.
- ICONN-1 is optimized for natural, emotionally resonant, and conversational interactions.
- ICONN-e1 is a specialized variant of the model fine-tuned for advanced reasoning, critical analysis, and complex problem-solving.
Together, these models represent a major leap forward in the evolution of AI systems—demonstrating not only deep reasoning but also a commitment to openness, accessibility, and human-aligned intelligence.
Usage
To run ICONN 1 Mini Beta, you need:
- Any hardware - CPU or GPU; Just make sure you have about 15GB storage space!
Run the code below to run ICONN 1 Mini Beta:
import os
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamer
from threading import Thread
model_id = "ICONNAI/ICONN-1-Mini-Beta"
try:
model = AutoModelForCausalLM.from_pretrained(
model_id, torch_dtype=torch.float16, device_map="auto", trust_remote_code=True
)
tokenizer = AutoTokenizer.from_pretrained(model_id)
except Exception as e:
exit(f"Exiting due to model loading error: {e}")
def generate_response(
message: str,
max_new_tokens: int = 2048,
temperature: float = 0.4,
top_p: float = 0.9,
top_k: int = 50,
repetition_penalty: float = 1.2,
) -> str:
conversation = [{"role": "user", "content": message}]
try:
input_ids = tokenizer.apply_chat_template(
conversation, return_tensors="pt", enable_thinking=True
)
except Exception as e:
return f"Error applying chat template: {e}"
input_ids = input_ids.to(model.device)
streamer = TextIteratorStreamer(tokenizer, timeout=20.0, skip_prompt=True, skip_special_tokens=True)
adjusted_top_k = int(max(1, top_k))
generate_kwargs = dict(
{"input_ids": input_ids},
streamer=streamer,
max_new_tokens=max_new_tokens,
do_sample=True,
top_p=top_p,
top_k=adjusted_top_k,
temperature=temperature,
num_beams=1,
repetition_penalty=repetition_penalty,
)
try:
t = Thread(target=model.generate, kwargs=generate_kwargs)
t.start()
except Exception as e:
return f"Error starting generation thread: {e}"
outputs = []
for text in streamer:
outputs.append(text)
return "".join(outputs)
if __name__ == "__main__":
question = "Can you explain briefly to me what is the Python programming language?"
print(f"User Question: {question}")
response = generate_response(question)
print(f"Bot Response: {response}")
Cite Us
If you use ICONN 1, please cite us as follows:
@misc{iconnai_2025,
author = { ICONNAI },
title = { ICONN-1-Mini-Beta (Revision e29b435) },
year = 2025,
url = { https://huggingface.co/ICONNAI/ICONN-1-Mini-Beta },
doi = { 10.57967/hf/5860 },
publisher = { Hugging Face }
}