o ei$@sPdZddlmZmZddlmZddlmZee Z GdddeZ dgZ dS)zQwen3 model configuration)PreTrainedConfiglayer_type_validation)RopeParameters)loggingc0s^eZdZdZdZdgZddddddddZdgdgfd d gd gfd gd gfd Z                    d3dedBdedBdedBdedBdedBdedBd edBd!e dBd"edBd#e dBd$e dBd%e dBd&e dBd'e e e e fBdBd(e dBd)e dBd*edBd+edBd,ee dBd-e dBd.edBd/edBd0edBf.fd1d2 ZZS)4 Qwen3Configa This is the configuration class to store the configuration of a [`Qwen3Model`]. It is used to instantiate a Qwen3 model according to the specified arguments, defining the model architecture. Instantiating a configuration with the defaults will yield a similar configuration to that of Qwen3-8B [Qwen/Qwen3-8B](https://huggingface.co/Qwen/Qwen3-8B). Configuration objects inherit from [`PreTrainedConfig`] and can be used to control the model outputs. Read the documentation from [`PreTrainedConfig`] for more information. Args: vocab_size (`int`, *optional*, defaults to 151936): Vocabulary size of the Qwen3 model. Defines the number of different tokens that can be represented by the `inputs_ids` passed when calling [`Qwen3Model`] hidden_size (`int`, *optional*, defaults to 4096): Dimension of the hidden representations. intermediate_size (`int`, *optional*, defaults to 22016): Dimension of the MLP representations. num_hidden_layers (`int`, *optional*, defaults to 32): Number of hidden layers in the Transformer encoder. num_attention_heads (`int`, *optional*, defaults to 32): Number of attention heads for each attention layer in the Transformer encoder. num_key_value_heads (`int`, *optional*, defaults to 32): This is the number of key_value heads that should be used to implement Grouped Query Attention. If `num_key_value_heads=num_attention_heads`, the model will use Multi Head Attention (MHA), if `num_key_value_heads=1` the model will use Multi Query Attention (MQA) otherwise GQA is used. When converting a multi-head checkpoint to a GQA checkpoint, each group key and value head should be constructed by meanpooling all the original heads within that group. For more details, check out [this paper](https://huggingface.co/papers/2305.13245). If it is not specified, will default to `32`. head_dim (`int`, *optional*, defaults to 128): The attention head dimension. hidden_act (`str` or `function`, *optional*, defaults to `"silu"`): The non-linear activation function (function or string) in the decoder. max_position_embeddings (`int`, *optional*, defaults to 32768): The maximum sequence length that this model might ever be used with. initializer_range (`float`, *optional*, defaults to 0.02): The standard deviation of the truncated_normal_initializer for initializing all weight matrices. rms_norm_eps (`float`, *optional*, defaults to 1e-06): The epsilon used by the rms normalization layers. use_cache (`bool`, *optional*, defaults to `True`): Whether or not the model should return the last key/values attentions (not used by all models). Only relevant if `config.is_decoder=True`. tie_word_embeddings (`bool`, *optional*, defaults to `False`): Whether the model's input and output word embeddings should be tied. rope_parameters (`RopeParameters`, *optional*): Dictionary containing the configuration parameters for the RoPE embeddings. The dictionary should contain a value for `rope_theta` and optionally parameters used for scaling in case you want to use RoPE with longer `max_position_embeddings`. attention_bias (`bool`, defaults to `False`, *optional*, defaults to `False`): Whether to use a bias in the query, key, value and output projection layers during self-attention. use_sliding_window (`bool`, *optional*, defaults to `False`): Whether to use sliding window attention. sliding_window (`int`, *optional*, defaults to 4096): Sliding window attention (SWA) window size. If not specified, will default to `4096`. max_window_layers (`int`, *optional*, defaults to 28): The number of layers using full attention. The first `max_window_layers` layers will use full attention, while any additional layer afterwards will use SWA (Sliding Window Attention). layer_types (`list`, *optional*): Attention pattern for each layer. attention_dropout (`float`, *optional*, defaults to 0.0): The dropout ratio for the attention probabilities. pad_token_id (`int`, *optional*): Padding token id. bos_token_id (`int`, *optional*): Beginning of stream token id. eos_token_id (`int`, *optional*): End of stream token id. ```python >>> from transformers import Qwen3Model, Qwen3Config >>> # Initializing a Qwen3 style configuration >>> configuration = Qwen3Config() >>> # Initializing a model from the Qwen3-8B style configuration >>> model = Qwen3Model(configuration) >>> # Accessing the model configuration >>> configuration = model.config ```qwen3past_key_valuescolwiserowwise)zlayers.*.self_attn.q_projzlayers.*.self_attn.k_projzlayers.*.self_attn.v_projzlayers.*.self_attn.o_projzlayers.*.mlp.gate_projzlayers.*.mlp.up_projzlayers.*.mlp.down_proj input_ids inputs_embeds hidden_statesattention_mask) embed_tokenslayersnormQV silu{Gz?ư>TFN vocab_size hidden_sizeintermediate_sizenum_hidden_layersnum_attention_headsnum_key_value_headshead_dim hidden_actmax_position_embeddingsinitializer_range rms_norm_eps use_cachetie_word_embeddingsrope_parametersattention_biasuse_sliding_windowsliding_windowmax_window_layers layer_typesattention_dropout pad_token_id bos_token_id eos_token_idc s|_| _|_|_|_|_|_jr|nd_|_|dur&|}|_ |_ |_ | _ | _ | _|_|_|_jdurSfddtjD_tjj|_|_|_| _|_tjdi|dS)Ncs(g|]}jdur|jkrdndqS)Nsliding_attentionfull_attention)r-r.).0iselfk/home/ubuntu/transcripts/venv/lib/python3.10/site-packages/transformers/models/qwen3/configuration_qwen3.py s z(Qwen3Config.__init__..r:)rr%rrr r!r,r-r.r"r#r$r&r'r(r+r0r/rangerr1r2r3r)r*super__init__)r9rrrr r!r"r#r$r%r&r'r(r)r*r+r,r-r.r/r0r1r2r3kwargs __class__r8r;r?}s>  zQwen3Config.__init__)rrrrrrrrrrrTFNFFrrNrNNN)__name__ __module__ __qualname____doc__ model_typekeys_to_ignore_at_inferencebase_model_tp_planbase_model_pp_planintstrfloatboolrdictlistr? __classcell__r:r:rAr;rsQ        rN) rFconfiguration_utilsrrmodeling_rope_utilsrutilsr get_loggerrCloggerr__all__r:r:r:r;s     +