LFM2.5-1.2B-Thinking

Reasoning-focused 1.2B parameter model with enhanced math and logical problem-solving

← Back to Text Models

LFM2.5-1.2B-Thinking is optimized for reasoning tasks, delivering strong performance on math, logic, and multi-step problem-solving. Built on the LFM2.5 architecture with specialized training for chain-of-thought reasoning.

HFGGUFMLXONNX

Specifications

Property	Value
Parameters	1.2B
Context Length	32K tokens
Architecture	LFM2.5 (Dense)

Math & Logic

Strong arithmetic and logical reasoning

Chain-of-Thought

Step-by-step problem decomposition

Fine-tunable

TRL compatible (SFT, DPO, GRPO)

Quick Start

Transformers

Install:

pip install "transformers>=5.0.0" torch accelerate

Download & Run:

from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "LiquidAI/LFM2.5-1.2B-Thinking"
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    device_map="auto",
    dtype="bfloat16",
)
tokenizer = AutoTokenizer.from_pretrained(model_id)

input_ids = tokenizer.apply_chat_template(
    [{"role": "user", "content": "What is machine learning?"}],
    add_generation_prompt=True,
    return_tensors="pt",
    tokenize=True,
).to(model.device)

output = model.generate(input_ids, max_new_tokens=512)
response = tokenizer.decode(output[0][len(input_ids[0]):], skip_special_tokens=True)
print(response)

llama.cpp

Install:

brew install llama.cpp

Run:

llama-cli -hf LiquidAI/LFM2.5-1.2B-Thinking-GGUF -c 4096 --color -i

The -hf flag downloads the model directly from Hugging Face. For other installation methods and advanced usage, see the llama.cpp guide.

vLLM

Install:

pip install vllm==0.14

Run:

from vllm import LLM, SamplingParams

llm = LLM(model="LiquidAI/LFM2.5-1.2B-Thinking")

sampling_params = SamplingParams(
    temperature=0.3,
    min_p=0.15,
    repetition_penalty=1.05,
    max_tokens=512,
)

output = llm.chat("What is machine learning?", sampling_params)
print(output[0].outputs[0].text)