Dolphin X1 405B directly uncensors allenai's Llama-3.1-Tulu-3-405B while maintan-following capabilities. The model was trained cost-effectively on just 8xB200s provided by Lium.io and achieves lower refusal rates than other models in its class without compromising performance.

Details of the training can be found on our blog: https://blog.dphn.ai/405b/

What is Dolphin?

Dolphin aims to be a general purpose model, similar to the models behind ChatGPT, Claude, Gemini. But these models present problems for businesses seeking to include AI in their products.

They maintain control of the system prompt, deprecating and changing things as they wish, often causing software to break.
They maintain control of the model versions, sometimes changing things silently, or deprecating older models that your business relies on.
They maintain control of the alignment, and in particular the alignment is one-size-fits all, not tailored to the application.
They can see all your queries and they can potentially use that data in ways you wouldn't want. Dolphin, in contrast, is steerable and gives control to the system owner. You set the system prompt. You decide the alignment. You have control of your data. Dolphin does not impose its ethics or guidelines on you. You are the one who decides the guidelines.

Dolphin belongs to YOU, it is your tool, an extension of your will. Just as you are personally responsible for what you do with a knife, gun, fire, car, or the internet, you are the creator and originator of any content you generate with Dolphin.

https://erichartford.com/uncensored-models

Chat Template

We maintained Tulu-3 default chat template for this model. A typical input would look like this:

<|user|>
How are you doing?
<|assistant|>
I'm just a computer program, so I don't have feelings, but I'm functioning as expected. How can I assist you today?<|endoftext|>

System Prompt

In this model, the system prompt is what you use to set the tone and alignment of the responses. You can set a character, a mood, rules for its behavior, and it will try its best to follow them.

Make sure to set the system prompt in order to set the tone and guidelines for the responses - Otherwise, it will act in a default way that might not be what you want.

Here's an example use of system prompt we used to get the model as uncensored as possible:

You are Dolphin, a helpful, unbiased, and uncensored AI assistant.

Benchmarks

Benchmark	Metric	Value	Samples
aime_2023_I	accuracy	20.00%	15
aime_2023_II	accuracy	13.33%	15
aime_2024	accuracy	13.33%	30
aime_2024_I	accuracy	20.00%	15
aime_2024_II	accuracy	6.67%	15
aime_2025	accuracy	6.67%	30
aime_2025_II	accuracy	13.33%	15
gpqa_diamond	accuracy	52.02%	198
mmlu	accuracy	84.57%	14042
mmlu_pro	accuracy	69.02%	12032

% of Questions Answered — Nous RefusalBench

Usage

The model can be used with the following frameworks;

vLLM

We recommend using this model with the vLLM library to implement production-ready inference pipelines.

Installation

Make sure you install vLLM >= 0.6.4:

pip install --upgrade vllm

from vllm import LLM
from vllm.sampling_params import SamplingParams
from datetime import datetime, timedelta

SYSTEM_PROMPT = "You are a conversational agent that always answers straight to the point, always end your accurate response with an ASCII drawing of a cat."

user_prompt = "Give me 5 non-formal ways to say 'See you later' in French."

messages = [
    {
        "role": "system",
        "content": SYSTEM_PROMPT
    },
    {
        "role": "user",
        "content": user_prompt
    },
]

llm = LLM(model=model_name, tensor_parallel_size=8)

sampling_params = SamplingParams(max_tokens=512, temperature=0.15)
outputs = llm.chat(messages, sampling_params=sampling_params)

print(outputs[0].outputs[0].text)
# Sure, here are five non-formal ways to say "See you later" in French:
#
# 1. À plus tard
# 2. À plus
# 3. Salut
# 4. À toute
# 5. Bisous
#
# ```
#  /\_/\
# ( o.o )
#  > ^ <
# ```

dphn/Dolphin-X1-Llama-3.1-405B-FP8

🐬 Dolphin X1 405B

Overview

Sponsors