Batched inference final version

v0.11.0

Hotfix for attention mask

Fix Batched inference

v0.10.0

Fix batching

Handle loading word embeddings directly

v0.8

Handle loading state dict tensor

Use tokenizer path from argument

v0.7

Add github actions