microsoft / DeepSpeed-MII

MII makes low-latency and high-throughput inference possible, powered by DeepSpeed.
Apache License 2.0
1.91k stars 175 forks source link

Configuration setting to pass parameters to tokenizer while encoding and decoding #522

Open krishnanpooja opened 2 months ago

krishnanpooja commented 2 months ago

We would like to have configuration options to pass parameters while encoding and decoding using Huggingface tokenizer. Ex:- in https://github.com/microsoft/DeepSpeed-MII/blob/2d5a2e77cb5d172c6652bbc7a6a1d6885d31d504/mii/batching/ragged_batching.py#L630, we need option to set truncation based on the max_length