Lm-format-enforcer

- [83](https://github.com/noamgat/lm-format-enforcer/pull/83) - Supporting added model tokens
- Improved support for out-of-token-vocabulary characters

- [80](https://github.com/noamgat/lm-format-enforcer/issues/80) - Fixed bug where comma could start json list
- [34](https://github.com/noamgat/lm-format-enforcer/issues/34) - Fixed llama-cpp-python low max_tokens default in sample

- Fixed build errors in certain edge cases

- [68](https://github.com/noamgat/lm-format-enforcer/pull/68) Added NVIDIA TensorRT-LLM Support, NVIDIA's contribution by [Ahmet Erdem](https://github.com/aerdem4). Thanks!
- Much faster TokenizerData initialization, new JSON freetext token caching algorithm.
- More robust error reporting.

- [67](https://github.com/noamgat/lm-format-enforcer/issues/67) Updating vLLM integration to support v0.3.0
- [63](https://github.com/noamgat/lm-format-enforcer/issues/63) JSONSchemaParser: Empty list cannot be closed after a newline

Several `JsonSchemaParser` improvements:
- [32](https://github.com/noamgat/lm-format-enforcer/issues/32) Added limited support for regex-constrained string in JSON using the `pattern` field. See `test_phone_number_in_string()`.
- [54](https://github.com/noamgat/lm-format-enforcer/issues/54) Fixed regression bug caused by limited-length JSON string caching.
- [53](https://github.com/noamgat/lm-format-enforcer/issues/53) Fixed problems with arrays of union types.