Feature Extraction
Transformers
PyTorch
Safetensors
Hebrew
bert
custom_code
text-embeddings-inference
Instructions to use dicta-il/dictabert-lex with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use dicta-il/dictabert-lex with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("feature-extraction", model="dicta-il/dictabert-lex", trust_remote_code=True)# pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModel tokenizer = AutoTokenizer.from_pretrained("dicta-il/dictabert-lex", trust_remote_code=True) model = AutoModel.from_pretrained("dicta-il/dictabert-lex", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
Document lexicon mode (beta) and its CC BY-NC licensing
#2
by ChuckLind - opened
The use_lexicon=True option has been present in the synced modeling code since 2026-09-30 but undocumented. This adds:
- A beta section describing what the option does, its requirements and cost, a worked example, and measured limitations (verified against wordlist v1.1.5).
- Licensing: the default lemmatizer stays CC BY 4.0. Lexicon mode depends on the CC BY-NC 4.0
dicta-il/wordlistdataset, so that mode falls under non-commercial terms. Includes a contact address for a free small-business commercial licence.
Note: the repo's license frontmatter field still reads cc-by-4.0; the split is expressed in prose only, since the field holds a single value.
ChuckLind changed pull request status to merged
ChuckLind deleted the
refs/pr/2 ref