bridge.models.qwen_omni.modeling_qwen25_omni.transformer_config#
Module Contents#
Classes#
Configuration for Qwen2.5 Omni transformer with vision, audio, and language components. |
API#
- class bridge.models.qwen_omni.modeling_qwen25_omni.transformer_config.Qwen25OmniTransformerConfig#
Bases:
megatron.core.transformer.transformer_config.TransformerConfigConfiguration for Qwen2.5 Omni transformer with vision, audio, and language components.
- vocab_size: int#
152064
- language_max_sequence_length: int#
4096
- patch_size: int#
14
- temporal_patch_size: int#
2
- in_channels: int#
3
- spatial_merge_size: int#
2
- apply_rotary_pos_emb_in_fp32: bool#
False
- fp16_lm_cross_entropy: bool#
False
False
- rotary_percent: float#
1.0
- rotary_base: float#
10000
- mrope_section: list[int]#
‘field(…)’
- apply_rope_fusion: bool#
False
- image_token_id: int#
151655
- video_token_id: int#
151656
- audio_token_id: int#
151646
- vision_start_token_id: int#
151652
- audio_start_token_id: int#
151647
- position_id_per_seconds: int#
25
- seconds_per_chunk: int#
2
- qk_layernorm: bool#
False