Skip to content

Commit

Permalink
Explicit definition of audio_latency_block_size
Browse files Browse the repository at this point in the history
  • Loading branch information
saeeddhqan authored Nov 21, 2024
1 parent 058d99d commit 9f48b21
Showing 1 changed file with 4 additions and 0 deletions.
4 changes: 4 additions & 0 deletions ultravox/model/ultravox_config.py
Original file line number Diff line number Diff line change
Expand Up @@ -63,6 +63,8 @@ class UltravoxConfig(transformers.PretrainedConfig):
The LoRA configuration for finetuning the text model.
audio_model_lora_config (`LoraConfigSimplified`, *optional*):
The LoRA configuration for finetuning the audio model.
audio_latency_block_size (`int`, *optional*, defaults to `None`):
The latency block size for simulating audio streaming.
Example:
Expand Down Expand Up @@ -105,6 +107,7 @@ def __init__(
projector_act: str = "swiglu",
text_model_lora_config: Optional[LoraConfigSimplified] = None,
audio_model_lora_config: Optional[LoraConfigSimplified] = None,
audio_latency_block_size: Optional[int] = None,
**kwargs,
):
self.ignore_index = ignore_index
Expand Down Expand Up @@ -147,6 +150,7 @@ def __init__(
if isinstance(audio_model_lora_config, dict)
else dataclasses.asdict(audio_model_lora_config or LoraConfigSimplified())
)
self.audio_latency_block_size = audio_latency_block_size

self.vocab_size = self.text_config.vocab_size

Expand Down

0 comments on commit 9f48b21

Please sign in to comment.