add support for `LiteWhisperConfig`

[eschmidbauer]

new whisper model uses configuration LiteWhisperConfig
when i try to convert with
ct2-transformers-converter --model efficient-speech/lite-whisper-large-v3-turbo --output_dir lite-whisper-large-v3-turbo-ct2 --trust_remote_code
i get

ValueError: No conversion is registered for the model configuration LiteWhisperConfig (supported configurations are: BartConfig, BertConfig, BloomConfig, CodeGenConfig, DistilBertConfig, FalconConfig, GPT2Config, GPTBigCodeConfig, GPTJConfig, GPTNeoXConfig, Gemma2Config, GemmaConfig, LlamaConfig, M2M100Config, MBartConfig, MPTConfig, MT5Config, MarianConfig, MistralConfig, MixFormerSequentialConfig, OPTConfig, PegasusConfig, Phi3Config, PhiConfig, RWConfig, T5Config, Wav2Vec2BertConfig, Wav2Vec2Config, WhisperConfig, XLMRobertaConfig)

[entn-at]

The model adds new low rank linear layers (see https://huggingface.co/efficient-speech/lite-whisper-large-v3-turbo/blob/main/modeling_lite_whisper.py). You would have to create a new C++ implementation in CTranslate2 to support LiteWhisper models.

[Eyoel-gebre]

I am working on writing a new converter and the C++ implementation for the low rank linear layers along with tests. I'm hoping to be finished within a week.

[moumeneb98]

HI @Eyoel-gebre did you manage to pogress in writing the implementation ?