metadata

language:
  - en
  - fr
  - es
  - pt
tags:
  - falcon3
base_model: tiiuae/Falcon3-7B-Base
license: other
license_name: falcon-llm-license
license_link: https://falconllm.tii.ae/falcon-terms-and-conditions.html

Falcon3-7B-Instruct

Falcon3 family of Open Foundation Models is a set of pretrained and instruct LLMs ranging from 1B to 10B.

This repository contains the Falcon3-7B-Instruct. It achieves state of art results (at release's time) on reasoning, language understanding, instruction following, code and mathematics tasks. Falcon3-7B-Instruct supports 4 languages (english, french, spanish, portuguese) and a context length up to 32K.

Model Details

Architecture
- Transformer based causal decoder only architecture
- 28 decoder blocks
- Grouped query attention (GQA) for faster inference: 12 query heads and 4 KV heads
- Wider head dimension: 256
- High RoPE value to support long context understanding: 1000042
- Uses SwiGLU and RMSNorm
- 32k context length
- 131k vocab size
Pretrained on 14 Gigatokens of datasets comprising of web, code, STEM, high quality and mutlilingual data using 2048 H100 GPU chips
Postrained on 1.2 million samples of STEM, conversations, code, safety and function call data
Supports EN, FR, ES, PT
Developed by Technology Innovation Institute
License: TII Falcon-LLM License 2.0
Model Release Date: December 2024

Getting started

Click to expand

from transformers import AutoTokenizer, AutoModelForCausalLM


from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "tiiuae/Falcon3-7B-Instruct"

model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype="auto",
    device_map="auto"]
)
tokenizer = AutoTokenizer.from_pretrained(model_name)

prompt = "How many hours in one day?"
messages = [
    {"role": "system", "content": "You are a helpful friendly assistant Falcon3 from TII, try to follow instructions as much as possible."},
    {"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)

generated_ids = model.generate(
    **model_inputs,
    max_new_tokens=1024
)
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]
print(response)

Benchmarks

We report in the following table our internal pipeline benchmarks:

Category	Benchmark	Llama-3.1-8B-Instruct	Qwen2.5-7B-Instruct	Falcon3-7B-Instruct
General	MMLU (5-shot)	55.9	72.4	68
	MMLU-PRO (5-shot)	21.8	35.8	40.7
	IFEval	78.8	74.7	76.5
Math	GSM8K (5-shot)	19.2	33.7	78.8
	GSM8k (8-shot, COT)	79.8	72.7	80.9
	MATH(4-shot)	10.4	26	33.1
Reasoning	Arc Challenge (25-shot)	46.6	55.7	65.9
	GPQA (0-shot)	33.6	31.9	32
	GPQA (0-shot, COT)	9.6	13.8	22.3
	MUSR (0-shot)	38.6	40.7	46.4
	BBH (3-shot)	43.7	53.9	52.4
	BBH (3-shot, COT)	6.7	21.2	69.3
CommonSense Understanding	PIQA (0-shot)	78.9	73.7	78.8
	SciQ (0-shot)	80.2	50.9	94.7
	Winogrande (0-shot)	TODO	TODO	70.4
	OpenbookQA (0-shot)	46.2	42.4	45.8
Instructions following	MT-Bench (avg)	7.86	8.54	8.36
Instructions following	Alapaca (WC)	26.57	31.5	26.13

Citation

If Falcon3 family were helpful to your work, feel free to give us a cite.

@misc{Falcon3,
    title = {The Falcon 3 family of Open Models},
    author = {TII Team},
    month = {December},
    year = {2024}
}