o iL@sTddlmZmZddlZddlmZddlmZddlm Z m Z ddl m Z ddl mZmZddlmZdd lmZmZdd lmZdd lmZd d lmZd dlmZmZmZmZm Z m!Z!m"Z"m#Z#m$Z$GdddeZ%Gddde!Z&GdddeZ'GdddeZ(Gddde"Z)Gddde Z*GdddeZ+GdddeZ,gdZ-dS) )CallableOptionalN)TransformersKwargs)Cache DynamicCache)layer_type_validation)create_causal_mask!create_sliding_window_causal_mask)BaseModelOutputWithPast)ROPE_INIT_FUNCTIONSrope_config_validation)ALL_ATTENTION_FUNCTIONS)Unpack) Olmo2Config) Olmo2AttentionOlmo2DecoderLayerOlmo2ForCausalLM Olmo2ModelOlmo2PreTrainedModel Olmo2RMSNormOlmo2RotaryEmbeddingapply_rotary_pos_embeager_attention_forwardcseZdZdZdZddddddddZdgd gfd d gd gfd gd gfd Z                    d fdd ZddZZ S)! Olmo3Configa This is the configuration class to store the configuration of a [`Olmo3Model`]. It is used to instantiate an OLMo3 model according to the specified arguments, defining the model architecture. Instantiating a configuration with the defaults will yield a similar configuration to that of the [allenai/OLMo-3-0725-1B](https://huggingface.co/allenai/OLMo-3-0725-1B). 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 50304): Vocabulary size of the Olmo3 model. Defines the number of different tokens that can be represented by the `inputs_ids` passed when calling [`Olmo3Model`] hidden_size (`int`, *optional*, defaults to 4096): Dimension of the hidden representations. intermediate_size (`int`, *optional*, defaults to 11008): Dimension of the MLP representations. num_hidden_layers (`int`, *optional*, defaults to 32): Number of hidden layers in the Transformer decoder. num_attention_heads (`int`, *optional*, defaults to 32): Number of attention heads for each attention layer in the Transformer decoder. num_key_value_heads (`int`, *optional*): 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 `num_attention_heads`. 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 2048): 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. 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`. pad_token_id (`int`, *optional*, defaults to 1): Padding token id. bos_token_id (`int`, *optional*): Beginning of stream token id. eos_token_id (`int`, *optional*, defaults to 50279): End of stream token id. tie_word_embeddings (`bool`, *optional*, defaults to `False`): Whether to tie weight embeddings rope_theta (`float`, *optional*, defaults to 10000.0): The base period of the RoPE embeddings. rope_scaling (`Dict`, *optional*): Dictionary containing the scaling configuration for the RoPE embeddings. NOTE: if you apply new rope type and you expect the model to work on longer `max_position_embeddings`, we recommend you to update this value accordingly. Expected contents: `rope_type` (`str`): The sub-variant of RoPE to use. Can be one of ['default', 'linear', 'dynamic', 'yarn', 'longrope', 'llama3'], with 'default' being the original RoPE implementation. `factor` (`float`, *optional*): Used with all rope types except 'default'. The scaling factor to apply to the RoPE embeddings. In most scaling types, a `factor` of x will enable the model to handle sequences of length x * original maximum pre-trained length. `original_max_position_embeddings` (`int`, *optional*): Used with 'dynamic', 'longrope' and 'llama3'. The original max position embeddings used during pretraining. `attention_factor` (`float`, *optional*): Used with 'yarn' and 'longrope'. The scaling factor to be applied on the attention computation. If unspecified, it defaults to value recommended by the implementation, using the `factor` field to infer the suggested value. `beta_fast` (`float`, *optional*): Only used with 'yarn'. Parameter to set the boundary for extrapolation (only) in the linear ramp function. If unspecified, it defaults to 32. `beta_slow` (`float`, *optional*): Only used with 'yarn'. Parameter to set the boundary for interpolation (only) in the linear ramp function. If unspecified, it defaults to 1. `short_factor` (`list[float]`, *optional*): Only used with 'longrope'. The scaling factor to be applied to short contexts (< `original_max_position_embeddings`). Must be a list of numbers with the same length as the hidden size divided by the number of attention heads divided by 2 `long_factor` (`list[float]`, *optional*): Only used with 'longrope'. The scaling factor to be applied to long contexts (< `original_max_position_embeddings`). Must be a list of numbers with the same length as the hidden size divided by the number of attention heads divided by 2 `low_freq_factor` (`float`, *optional*): Only used with 'llama3'. Scaling factor applied to low frequency components of the RoPE `high_freq_factor` (`float`, *optional*): Only used with 'llama3'. Scaling factor applied to high frequency components of the RoPE 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. attention_dropout (`float`, *optional*, defaults to 0.0): The dropout ratio for the attention probabilities. rms_norm_eps (`float`, *optional*, defaults to 1e-05): The epsilon used by the rms normalization layers. sliding_window (`int`, *optional*, defaults to 4096): Size of the sliding window for sliding window attention. layer_types (`list`, *optional*): Attention pattern for each layer. Defaults to sliding window attention for 3 out of 4 layers, and full attention for every 4th layer. ```python >>> from transformers import Olmo3Model, Olmo3Config >>> # Initializing a Olmo3 7B style configuration >>> configuration = Olmo3Config() >>> # Initializing a model from the Olmo3 7B style configuration >>> model = Olmo3Model(configuration) >>> # Accessing the model configuration >>> configuration = model.config ``` olmo3 colwise_rep rowwise_repcolwiserowwise)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_tokenslayersnorm+ Nsilu{Gz?TgF@h㈵>c stjdid|d|d|d|d|d|d|d|d | d | d | d | d | d|d|d|d|d|d||||_||_|jdurXddt|jD|_t|jdS)N vocab_size hidden_sizeintermediate_sizenum_hidden_layersnum_attention_headsnum_key_value_heads hidden_actmax_position_embeddingsinitializer_range use_cache pad_token_id bos_token_id eos_token_idtie_word_embeddings rope_theta rope_scalingattention_biasattention_dropout rms_norm_epscSs$g|]}|dddkrdndqS)r/rsliding_attentionfull_attention).0irJrJd/home/ubuntu/veenaModal/venv/lib/python3.10/site-packages/transformers/models/olmo3/modular_olmo3.py sz(Olmo3Config.__init__..rJ)super__init__sliding_window layer_typesranger7r)selfr4r5r6r7r8r9r:r;r<r=r>r?r@rArBrCrDrErFrQrRkwargs __class__rJrMrPs^       zOlmo3Config.__init__cCs t|dS)z< Validate the `rope_scaling` configuration. N)r )rTrJrJrM_rope_scaling_validations z$Olmo3Config._rope_scaling_validation)r(r)r*r+r+Nr,r-r.Tr/Nr0Fr1NFr2r3r)N) __name__ __module__ __qualname____doc__ model_typebase_model_tp_planbase_model_pp_planrPrX __classcell__rJrJrVrMr,sLo  8rc@ eZdZdS) Olmo3RMSNormNrYrZr[rJrJrJrMrbrbcseZdZdedeffdd Z  ddejdeejejfde ejd e e d e ej d e e d eeje ejffd dZZS)Olmo3Attentionconfig layer_idxcsJtj||d|jdusJ|j||_|jdkr |j|_dSd|_dS)N)rgrH)rOrPrRattention_typerQ)rTrfrgrVrJrMrPs  zOlmo3Attention.__init__Nr#position_embeddingsr$past_key_valuescache_positionrUreturncKs@|jdd}g|d|jR}|||} |||} ||} | |dd} | |dd} | |dd} |\} } t | | | | \} } |durc| | |d}| | | |j |\} } t }|j jdkrqt|j j}||| | | |f|js}dn|j|j|jd|\}}|jg|dR}||}||fS)Nr/r)sincosrkeagerr2)dropoutscalingrQ)shapehead_dimq_normq_projk_normk_projv_projview transposerupdatergrrf_attn_implementationrtrainingrErrrQreshape contiguouso_proj)rTr#rir$rjrkrU input_shape hidden_shape query_states key_states value_statesrorn cache_kwargsattention_interface attn_output attn_weightsrJrJrMforwards@       zOlmo3Attention.forwardNN)rYrZr[rintrPtorchTensortuplerr LongTensorrrrr`rJrJrVrMres& rec@ra)Olmo3DecoderLayerNrcrJrJrJrMr)rdrc@s$eZdZddedeefddZdS)Olmo3RotaryEmbeddingNrf rope_typecCstj||dur||_nt|dr&t|jtr&|jd|jd|_nd|_|jdus0J|j |_ |j |_ ||_ t |j|_||j |\}|_|jd|dd|j|_dS)NrCrtypedefaultinv_freqF) persistent)nnModulerPrhasattr isinstancerCdictgetr;max_seq_len_cachedoriginal_max_seq_lenrfr rope_init_fnattention_scalingregister_bufferroriginal_inv_freq)rTrfdevicerrrJrJrMrP0s   zOlmo3RotaryEmbedding.__init__r)rYrZr[rrstrrPrJrJrJrMr/src@ra)Olmo3PreTrainedModelNrcrJrJrJrMrFrdrcseZdZdeffdd Z       ddeejdeejdeejdee d eej d eejd ee d e e d efddZZS) Olmo3Modelrfcsfttjjd|_tfddtj D|_ t t ddt dd|_ |`dS)N)epscsg|]}t|qSrJ)r)rKrgrfrJrMrNRsz'Olmo3Model.__init__..r)rfrrrHrI)rOrPrbr5rFr'r ModuleListrSr7r& ModuleDictr rotary_embs rotary_emb)rTrfrVrrMrPNs  zOlmo3Model.__init__Nr!r$ position_idsrjr"rkr=rUrlc KsL|du|duAr td|dur||}|r!|dur!t|jd}|dur=|dur-|nd} tj| | |jd|jd}|durF| d}t |} t sf|j|||||d} t d i| t d i| d} |} |jd| ||jd | |d } |jd|jjD]}|| f| |jj|||| |jjd |} q|| } t| |d S)Nz:You must specify exactly one of input_ids or inputs_embedsrrr/)r)rf input_embedsr$rkrjr)rIrHrHrIr)r$rrjrkri)last_hidden_staterjrJ) ValueErrorr%rrfget_seq_lengthrarangersr unsqueezerrr r rr&r7 self_attnrhr'r )rTr!r$rrjr"rkr=rUpast_seen_tokenscausal_mask_mapping mask_kwargsr#position_embeddings_mapping decoder_layerrJrJrMr\sZ         zOlmo3Model.forward)NNNNNNN)rYrZr[rrPrrrrr FloatTensorboolrrr rr`rJrJrVrMrMs8  rc@ra)Olmo3ForCausalLMNrcrJrJrJrMrrdr)rrrr).typingrrrtorch.nnrtransformers.utils.genericr cache_utilsrrconfiguration_utilsr masking_utilsr r modeling_outputsr modeling_rope_utilsr r modeling_utilsrprocessing_utilsrolmo2.configuration_olmo2rolmo2.modeling_olmo2rrrrrrrrrrrbrerrrrr__all__rJrJrJrMs.       , @8U