-
Notifications
You must be signed in to change notification settings - Fork 1.9k
Open
Labels
P2Priority of the issue for triage purpose: Needs to be fixed at some point.Priority of the issue for triage purpose: Needs to be fixed at some point.enhancementNew feature or requestNew feature or request
Milestone
Description
Port CLIP tokenizer which leverages byte-level BPE. This tokenizer enables scenarios like StableDiffusion
May be dependent on #6992.
Reference:
https://huggingface.co/docs/transformers/main/en/model_doc/clip
https://github.com/huggingface/transformers/blob/0549000c5bf6c7249f411917f2a6f0b6d0f06da1/src/transformers/models/codegen/tokenization_codegen.py#L98
https://onnxruntime.ai/docs/tutorials/csharp/stable-diffusion-csharp.html#tokenization-with-onnx-runtime-extensions
Paper:
https://arxiv.org/abs/2103.00020
https://arxiv.org/pdf/2103.00020.pdf
arthurvb
Metadata
Metadata
Assignees
Labels
P2Priority of the issue for triage purpose: Needs to be fixed at some point.Priority of the issue for triage purpose: Needs to be fixed at some point.enhancementNew feature or requestNew feature or request