Edit model card

LongReward-glm4-9b-SFT

🤗 [LongReward Dataset] • 💻 [Github Repo] • 📃 [LongReward Paper]

LongReward-glm4-9b-SFT is supervisedly fined-tuned from glm-4-9b using the sft split of LongReward-10k dataset, and supports a maximum context window of up to 64K tokens.

Environment: Same environment requirement as glm-4-9b-chat (transforemrs>=4.43.0).

A simple demo for deployment of the model:

import torch
from transformers import AutoTokenizer, AutoModelForCausalLM

model_path = "NeoZ123/LongReward-glm4-9b-SFT"
tokenizer = AutoTokenizer.from_pretrained(model_path, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(model_path, torch_dtype=torch.bfloat16, trust_remote_code=True, device_map='auto')
context = '''
W. Russell Todd, 94, United States Army general (b. 1928). February 13. Tim Aymar, 59, heavy metal singer (Pharaoh) (b. 1963). Marshall \"Eddie\" Conway, 76, Black Panther Party leader (b. 1946). Roger Bonk, 78, football player (North Dakota Fighting Sioux, Winnipeg Blue Bombers) (b. 1944). Conrad Dobler, 72, football player (St. Louis Cardinals, New Orleans Saints, Buffalo Bills) (b. 1950). Brian DuBois, 55, baseball player (Detroit Tigers) (b. 1967). Robert Geddes, 99, architect, dean of the Princeton University School of Architecture (1965–1982) (b. 1923). Tom Luddy, 79, film producer (Barfly, The Secret Garden), co-founder of the Telluride Film Festival (b. 1943). David Singmaster, 84, mathematician (b. 1938).
'''
query = "What was Robert Geddes' profession?"
prompt = context + '\n\n' + query
response, _ = model.chat(tokenizer, prompt, temprature=1, max_new_tokens=1024)
print(response)

You can also deploy the model with vllm for faster inference:

import torch
from vllm import LLM, SamplingParams

model_path = "NeoZ123/LongReward-glm4-9b-SFT"
model = LLM(
    model= model_path,
    dtype=torch.bfloat16,
    trust_remote_code=True,
    tensor_parallel_size=1,
    max_model_len=65536,
    gpu_memory_utilization=1,
)
tokenizer = model.get_tokenizer()
context = '''
W. Russell Todd, 94, United States Army general (b. 1928). February 13. Tim Aymar, 59, heavy metal singer (Pharaoh) (b. 1963). Marshall \"Eddie\" Conway, 76, Black Panther Party leader (b. 1946). Roger Bonk, 78, football player (North Dakota Fighting Sioux, Winnipeg Blue Bombers) (b. 1944). Conrad Dobler, 72, football player (St. Louis Cardinals, New Orleans Saints, Buffalo Bills) (b. 1950). Brian DuBois, 55, baseball player (Detroit Tigers) (b. 1967). Robert Geddes, 99, architect, dean of the Princeton University School of Architecture (1965–1982) (b. 1923). Tom Luddy, 79, film producer (Barfly, The Secret Garden), co-founder of the Telluride Film Festival (b. 1943). David Singmaster, 84, mathematician (b. 1938).
'''
query = "What was Robert Geddes' profession?"
prompt = context + '\n\n' + query
inputs = tokenizer.build_chat_input(prompt, history=[], role='user')
eos_token_id = [tokenizer.eos_token_id, tokenizer.get_command("<|user|>"), tokenizer.get_command("<|observation|>")]
generation_params = SamplingParams(
    temperature=0.95,
    top_p=0.7,
    max_tokens=1024,
    stop_token_ids=eos_token_id,
)
input_ids = inputs.input_ids[0].tolist()
outputs = model.generate(sampling_params=generation_params, prompt_token_ids=[input_ids])
response = tokenizer.decode(outputs[0].outputs[0].token_ids[:-1])
print(response)

License

glm-4-9b License

Citation

If you find our work useful, please consider citing LongReward:

@article{zhang2024longreward,
  title = {LongReward: Improving Long-context Large Language Models
with AI Feedback} 
  author={Jiajie Zhang and Zhongni Hou and Xin Lv and Shulin Cao and Zhenyu Hou and Yilin Niu and Lei Hou and Yuxiao Dong and Ling Feng and Juanzi Li},
  journal={arXiv preprint arXiv:2410.21252},
  year={2024}
}
Downloads last month
15
Safetensors
Model size
9.4B params
Tensor type
BF16
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Dataset used to train NeoZ123/LongReward-glm4-9b-SFT