RichardErkhov/zjunlp_-_OceanGPT-2B-v0.1-gguf

Quantization made by Richard Erkhov.

OceanGPT-2B-v0.1 - GGUF

Model creator: https://huggingface.co/zjunlp/
Original model: https://huggingface.co/zjunlp/OceanGPT-2B-v0.1/

Name	Quant method	Size
OceanGPT-2B-v0.1.Q2_K.gguf	Q2_K	1.12GB
OceanGPT-2B-v0.1.IQ3_XS.gguf	IQ3_XS	1.21GB
OceanGPT-2B-v0.1.IQ3_S.gguf	IQ3_S	1.26GB
OceanGPT-2B-v0.1.Q3_K_S.gguf	Q3_K_S	1.26GB
OceanGPT-2B-v0.1.IQ3_M.gguf	IQ3_M	1.32GB
OceanGPT-2B-v0.1.Q3_K.gguf	Q3_K	1.38GB
OceanGPT-2B-v0.1.Q3_K_M.gguf	Q3_K_M	1.38GB
OceanGPT-2B-v0.1.Q3_K_L.gguf	Q3_K_L	1.46GB
OceanGPT-2B-v0.1.IQ4_XS.gguf	IQ4_XS	1.45GB
OceanGPT-2B-v0.1.Q4_0.gguf	Q4_0	1.5GB
OceanGPT-2B-v0.1.IQ4_NL.gguf	IQ4_NL	1.51GB
OceanGPT-2B-v0.1.Q4_K_S.gguf	Q4_K_S	1.57GB
OceanGPT-2B-v0.1.Q4_K.gguf	Q4_K	1.68GB
OceanGPT-2B-v0.1.Q4_K_M.gguf	Q4_K_M	1.68GB
OceanGPT-2B-v0.1.Q4_1.gguf	Q4_1	1.64GB
OceanGPT-2B-v0.1.Q5_0.gguf	Q5_0	1.78GB
OceanGPT-2B-v0.1.Q5_K_S.gguf	Q5_K_S	1.81GB
OceanGPT-2B-v0.1.Q5_K.gguf	Q5_K	1.9GB
OceanGPT-2B-v0.1.Q5_K_M.gguf	Q5_K_M	1.9GB
OceanGPT-2B-v0.1.Q5_1.gguf	Q5_1	1.93GB
OceanGPT-2B-v0.1.Q6_K.gguf	Q6_K	2.2GB
OceanGPT-2B-v0.1.Q8_0.gguf	Q8_0	2.7GB

Original model description:

license: mit pipeline_tag: text-generation tags: - ocean - text-generation-inference - oceangpt language: - en - zh datasets: - zjunlp/OceanInstruct

OceanGPT(沧渊): A Large Language Model for Ocean Science Tasks

Project • Paper • Models • Web • Quickstart • Citation

OceanGPT-2B-v0.1 is based on MiniCPM-2B and has been trained on a bilingual dataset in the ocean domain, covering both Chinese and English.

❗Disclaimer: This project is purely an academic exploration rather than a product. Please be aware that due to the inherent limitations of large language models, there may be issues such as hallucinations.

⏩Quickstart

Download the model

Download the model: OceanGPT-2B-v0.1

git lfs install
git clone https://huggingface.co/zjunlp/OceanGPT-2B-v0.1

huggingface-cli download --resume-download zjunlp/OceanGPT-2B-v0.1 --local-dir OceanGPT-2B-v0.1 --local-dir-use-symlinks False

Inference

from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
device = "cuda" # the device to load the model onto
path = 'YOUR-MODEL-PATH'
model = AutoModelForCausalLM.from_pretrained(
    path,
    torch_dtype=torch.bfloat16,
    device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained(path)

prompt = "Which is the largest ocean in the world?"
messages = [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(device)

generated_ids = model.generate(
    model_inputs.input_ids,
    max_new_tokens=512
)
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]

📌Models

Model Name	HuggingFace	WiseModel	ModelScope
OceanGPT-14B-v0.1 (based on Qwen)	14B	14B	14B
OceanGPT-7B-v0.2 (based on Qwen)	7B	7B	7B
OceanGPT-2B-v0.1 (based on MiniCPM)	2B	2B	2B

🌻Acknowledgement

OceanGPT(沧渊) is trained based on the open-sourced large language models including Qwen, MiniCPM, LLaMA. Thanks for their great contributions!

Limitations

The model may have hallucination issues.
We did not optimize the identity and the model may generate identity information similar to that of Qwen/MiniCPM/LLaMA/GPT series models.
The model's output is influenced by prompt tokens, which may result in inconsistent results across multiple attempts.
The model requires the inclusion of specific simulator code instructions for training in order to possess simulated embodied intelligence capabilities (the simulator is subject to copyright restrictions and cannot be made available for now), and its current capabilities are quite limited.

🚩Citation

Please cite the following paper if you use OceanGPT in your work.

@article{bi2023oceangpt,
  title={OceanGPT: A Large Language Model for Ocean Science Tasks},
  author={Bi, Zhen and Zhang, Ningyu and Xue, Yida and Ou, Yixin and Ji, Daxiong and Zheng, Guozhou and Chen, Huajun},
  journal={arXiv preprint arXiv:2310.02031},
  year={2023}
}