vikrahul77/

walton-hard-cosyn-90-10

Combined Walton (hard): 900 samples, CoSyn-400K: 100 samples

This dataset was processed using the data-preproc package for vision-language model training.

Base Model: Qwen/Qwen2.5-7B-Instruct Tokenizer: Qwen/Qwen2.5-7B-Instruct Sequence Length: 16384 Processing Type: Vision Language (VL)

input_ids: Tokenized input sequences attention_mask: Attention masks for the… See the full description on the dataset page: https://huggingface.co/datasets/vikrahul77/walton-hard-cosyn-90-10.