AttributeError: 'Dataset' object has no attribute 'keys' about llama2lang

[ErikTromp]

What (parameters) dp you use to run the script?

[voulgeor]

this is how I run the script
python create_thread_prompts.py ./out/train "You are a generic chatbot that always answers in Greek ." out2

if I do a print a print(dataset.to_pandas().head())
I see data , so I suppose the datase is loaded.

[ErikTromp]

Looks like you are using the folder. You call the scripts like this:
python LLaMa2lang/translate_oasst.py --batch_size 10 --use_madlad --madlad_quant pl ./oasst_checkpoints/pl
python LLaMa2lang/combine_checkpoints.py ./oasst_checkpoints/pl ./oasst_checkpoints/pl_out
python LLaMa2lang/create_thread_prompts.py ./oasst_checkpoints/pl_out "test" ./oasst_checkpoints/pl_threads

You should not load the "train" subfolder but ./out instead if you followed this properly.