oli-mebberson/lattice-pulse

Lattice Pulse

Lattice Pulse is the conversational model in the Lattice Systems product line.

We fine-tuned Qwen/Qwen2.5-1.5B-Instruct on identity-mix instruction data (Lattice branding + Stanford Alpaca).

• Lattice Pulse — this model (fine-tuned Qwen2.5-1.5B-Instruct)
• Base weights: Qwen2.5 (Apache 2.0)

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "olimebberson/lattice-pulse"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id).to("cuda")

messages = [{"role": "user", "content": "Hello!"}]
text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
inputs = tokenizer(text, return_tensors="pt").to("cuda")
out = model.generate(**inputs, max_new_tokens=128, do_sample=True, temperature=0.7)
print(tokenizer.decode(out[0], skip_special_tokens=True))