Skip to content

anshuman23/LLM_Position_Bias

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Position Bias in LLM Summarization

Pegasus

Main code is in script.py Command to run script:

python script.py --dataset dataset_name --batch_size batch-size --device device# --model model_name

  • Replace dataset_name with either cnn, xsum, news, reddit
  • Replace batch-size with number
  • Replace device# with GPU
  • Replace model_name with either the huggingface model name or the local finetuned model name

BART

Main code is in script.py Command to run script:

python script.py --dataset dataset_name --batch_size batch-size --device device# --model model_name

  • Replace dataset_name with either cnn, xsum, news, reddit
  • Replace batch-size with number
  • Replace device# with GPU
  • Replace model_name with either the huggingface model name or the local finetuned model name

Dolly-v2-7B

Inference the LLM with simply running data_collection.py.

Main code is in script.py Command to run script:

python script.py --dataset dataset_name --batch_size batch-size --device device# 

  • Replace dataset_name with either cnn, xsum, news, reddit
  • Replace batch-size with number
  • Replace device# with GPU

ChatGPT 3.5-T

Inference the LLM with running inference.ipynb You will need to add a azure-configuration.json that has the Azure OpenAI endpoints.

Main code is in script.py Command to run script:

python script.py --dataset dataset_name --batch_size batch-size --device device# 

  • Replace dataset_name with either cnn, xsum, news, reddit
  • Replace batch-size with number
  • Replace device# with GPU

Llama-13B-chat

You will need to download and copy the Llama folder from Meta with the weights to the directory first. Inference can be done by running the file inference_13b_chat.py with the following command.

torchrun --nproc_per_node 2 inference_13b_chat.py \
    --ckpt_dir llama-2-13b-chat/ \
    --tokenizer_path tokenizer.model \
    --max_seq_len 2000 --max_batch_size 4

Main code is in script.py Command to run script:

python script.py --dataset dataset_name --batch_size batch-size --device device# 

  • Replace dataset_name with either cnn, xsum, news, reddit
  • Replace batch-size with number
  • Replace device# with GPU

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published