from dataclasses import dataclass from typing import Any, Dict, List, Optional, Union import mlx.core as mx import mlx.nn as nn from mlx_lm.models.rope_utils import initialize_rope from ..base import ( LanguageModelOutput, create_attention_mask, scaled_dot_product_attention, ) from ..cache import KVCache, RotatingKVCache from ..pixtral.language import Mistral from .config import TextConfig def _get_llama_4_attn_scale( start: int, stop: int, beta: float, max_position_embeddings: int ): scaling = 1 + beta * mx.log( 1 + mx.floor(mx.arange(start, stop) / max_position_embeddings) ) return scaling[:, None] class Attention(nn.Module): def __init__(self, config: TextConfig): super().__init__() dim = config.hidden_size self.n_heads = n_heads = config.num_attention_heads self.n_kv_heads = n_kv_heads = config.num_key_value_heads self.head_dim = head_dim = config.head_dim or config.hidden_size // n_heads self.scale = head_dim**-0.5 self.q_proj = nn.Linear(dim, n_heads * head_dim, bias=False) self.k_proj = nn.Linear(dim, n_kv_heads * head_dim, bias=False) self.v_proj = nn.Linear(dim, n_kv_heads * head_dim, bias=False) self.o_proj = nn.Linear(n_heads * head_dim, dim, bias=False) self.rope = initialize_rope( self.head_dim, config.rope_parameters["rope_theta"], config.rope_traditional, config.rope_parameters, config.max_position_embeddings, ) def __call__( self, x: mx.array, attn_scale: mx.array, mask: Optional[mx.array] = None, cache: Optional[Any] = None, ) -> mx.array: B, L, D = x.shape queries, keys, values = self.q_proj(x), self.k_proj(x), self.v_proj(x) # Prepare the queries, keys and values for the attention computation queries = queries.reshape(B, L, self.n_heads, -1).transpose(0, 2, 1, 3) keys = keys.reshape(B, L, self.n_kv_heads, -1).transpose(0, 2, 1, 3) values = values.reshape(B, L, self.n_kv_heads, -1).transpose(0, 2, 1, 3) offset = 0 if cache is not None: offset = cache.offset queries = self.rope(queries, offset=offset) keys = self.rope(keys, offset=offset) keys, values = cache.update_and_fetch(keys, values) else: queries = self.rope(queries) keys = self.rope(keys) queries = queries * attn_scale output = scaled_dot_product_attention( queries, keys, values, cache=cache, scale=self.scale, mask=mask ) output = output.transpose(0, 2, 1, 3).reshape(B, L, -1) return self.o_proj(output) class MLP(nn.Module): def __init__(self, config: TextConfig): super().__init__() dim = config.hidden_size hidden_dim = config.intermediate_size self.gate_proj = nn.Linear(dim, hidden_dim, bias=False) self.down_proj = nn.Linear(hidden_dim, dim, bias=False) self.up_proj = nn.Linear(dim, hidden_dim, bias=False) def __call__(self, x) -> mx.array: return self.down_proj(nn.silu(self.gate_proj(x)) * self.up_proj(x)) class TransformerBlock(nn.Module): def __init__(self, config: TextConfig, use_sliding: bool = False): super().__init__() self.num_attention_heads = config.num_attention_heads self.hidden_size = config.hidden_size self.use_sliding = use_sliding self.self_attn = Attention(config) self.mlp = MLP(config) self.input_layernorm = nn.RMSNorm(config.hidden_size, eps=config.rms_norm_eps) self.post_attention_layernorm = nn.RMSNorm( config.hidden_size, eps=config.rms_norm_eps ) self.config = config def __call__( self, x: mx.array, attn_scale: mx.array, mask: Optional[mx.array] = None, cache: Optional[Any] = None, ) -> mx.array: r = self.self_attn(self.input_layernorm(x), attn_scale, mask, cache) h = x + r r = self.mlp(self.post_attention_layernorm(h)) out = h + r return out class Ministral3(nn.Module): def __init__(self, config: TextConfig): super().__init__() self.config = config self.vocab_size = config.vocab_size self.num_hidden_layers = config.num_hidden_layers self.layer_types = config.layer_types self.sliding_window = config.sliding_window self.embed_tokens = nn.Embedding(config.vocab_size, config.hidden_size) self.layers = [ TransformerBlock( config=config, use_sliding=layer_type == "sliding_attention" ) for layer_type in self.layer_types ] self.norm = nn.RMSNorm(config.hidden_size, eps=config.rms_norm_eps) self.fa_idx = self.layer_types.index("full_attention") self.swa_idx = None for e, l in enumerate(self.layers): if l.use_sliding: self.swa_idx = e break def __call__( self, inputs: mx.array, cache=None, inputs_embeds: Optional[mx.array] = None, ): if inputs_embeds is not None: h = inputs_embeds else: h = self.embed_tokens(inputs) if cache is None: cache = [None] * len(self.layers) offset = 0 else: offset = cache[0].offset fa_mask = create_attention_mask(h, cache[self.fa_idx]) if self.swa_idx is not None: swa_mask = create_attention_mask( h, cache[self.swa_idx], window_size=self.sliding_window ) attn_scale = _get_llama_4_attn_scale( offset, offset + inputs.shape[1], self.config.rope_parameters["llama_4_scaling_beta"], self.config.rope_parameters["original_max_position_embeddings"], ).astype(h.dtype) for layer, cache in zip(self.layers, cache): mask = swa_mask if layer.use_sliding else fa_mask h = layer(h, attn_scale, mask, cache=cache) return self.norm(h) class LanguageModel(nn.Module): def __init__(self, config: TextConfig): super().__init__() self.config = config self.model_type = config.model_type if self.model_type == "ministral3": self.model = Ministral3(config) else: self.model = Mistral(config) if not config.tie_word_embeddings: self.lm_head = nn.Linear(config.hidden_size, config.vocab_size, bias=False) def __call__( self, inputs: mx.array, cache=None, inputs_embeds: Optional[mx.array] = None, ): out = self.model(inputs=inputs, cache=cache, inputs_embeds=inputs_embeds) if self.config.tie_word_embeddings: out = self.model.embed_tokens.as_linear(out) else: out = self.lm_head(out) return LanguageModelOutput(logits=out) def sanitize(self, weights): # Remove unused precomputed rotary freqs weights = { k: v for k, v in weights.items() if "self_attn.rotary_emb.inv_freq" not in k } if self.config.tie_word_embeddings: weights.pop("lm_head.weight", None) new_weights = {} for k, v in weights.items(): if "weight_scale_inv" in k: scale_inv = v wk = k.replace("_scale_inv", "") weight = weights[wk] new_weights[wk] = weight * scale_inv elif "activation_scale" in k: continue elif k not in new_weights: new_weights[k] = v weights = new_weights return weights @property def layers(self): return self.model.layers def make_cache(self): return [ ( RotatingKVCache(max_size=self.model.sliding_window) if getattr(layer, "use_sliding", False) else KVCache() ) for layer in self.layers ]