DBRX (ml-explore#628)

* dbrx

* format

* format

* comments

* change scores slightly

* remove inadvertant import

Author: awni

llms/mlx_lm/models/dbrx.py

@@ -0,0 +1,255 @@
+from dataclasses import dataclass
+from typing import Dict, Optional, Tuple, Union
+
+import mlx.core as mx
+import mlx.nn as nn
+import numpy as np
+
+from .base import BaseModelArgs
+
+
+@dataclass
+class ModelArgs(BaseModelArgs):
+    model_type: str
+    vocab_size: int
+    d_model: int
+    ffn_config: dict
+    attn_config: dict
+    n_layers: int
+    n_heads: int
+
+
+class Attention(nn.Module):
+    def __init__(self, args: ModelArgs):
+        super().__init__()
+        self.num_heads = args.n_heads
+        self.d_model = args.d_model
+        self.head_dim = args.d_model // args.n_heads
+        self.num_key_value_heads = args.attn_config["kv_n_heads"]
+        self.clip_qkv = args.attn_config["clip_qkv"]
+        self.rope_theta = args.attn_config["rope_theta"]
+
+        self.scale = self.head_dim**-0.5
+
+        self.Wqkv = nn.Linear(
+            args.d_model,
+            (self.num_key_value_heads * 2 + self.num_heads) * self.head_dim,
+            bias=False,
+        )
+        self.out_proj = nn.Linear(args.d_model, args.d_model, bias=False)
+        self.rope = nn.RoPE(
+            self.head_dim,
+            traditional=False,
+            base=self.rope_theta,
+        )
+
+    def __call__(
+        self,
+        x: mx.array,
+        mask: Optional[mx.array] = None,
+        cache: Optional[Tuple[mx.array, mx.array]] = None,
+    ) -> mx.array:
+
+        qkv = self.Wqkv(x)
+        qkv = mx.clip(qkv, a_min=-self.clip_qkv, a_max=self.clip_qkv)
+        splits = [self.d_model, self.d_model + self.head_dim * self.num_key_value_heads]
+        queries, keys, values = mx.split(qkv, splits, axis=-1)
+
+        B, L, D = x.shape
+
+        # Prepare the queries, keys and values for the attention computation
+        queries = queries.reshape(B, L, self.num_heads, -1).transpose(0, 2, 1, 3)
+        keys = keys.reshape(B, L, self.num_key_value_heads, -1).transpose(0, 2, 1, 3)
+        values = values.reshape(B, L, self.num_key_value_heads, -1).transpose(
+            0, 2, 1, 3
+        )
+
+        if cache is not None:
+            key_cache, value_cache = cache
+            queries = self.rope(queries, offset=key_cache.shape[2])
+            keys = self.rope(keys, offset=key_cache.shape[2])
+            keys = mx.concatenate([key_cache, keys], axis=2)
+            values = mx.concatenate([value_cache, values], axis=2)
+        else:
+            queries = self.rope(queries)
+            keys = self.rope(keys)
+
+        output = mx.fast.scaled_dot_product_attention(
+            queries, keys, values, scale=self.scale, mask=mask
+        )
+        output = output.transpose(0, 2, 1, 3).reshape(B, L, -1)
+        return self.out_proj(output), (keys, values)
+
+
+class NormAttnNorm(nn.Module):
+    def __init__(self, args: ModelArgs):
+        super().__init__()
+        self.norm_1 = nn.LayerNorm(args.d_model, bias=False)
+        self.norm_2 = nn.LayerNorm(args.d_model, bias=False)
+        self.attn = Attention(args)
+
+    def __call__(
+        self,
+        x: mx.array,
+        mask: Optional[mx.array] = None,
+        cache: Optional[Tuple[mx.array, mx.array]] = None,
+    ) -> mx.array:
+        h, cache = self.attn(self.norm_1(x), mask=mask, cache=cache)
+        x = h + x
+        return x, self.norm_2(x), cache
+
+
+class MLP(nn.Module):
+    def __init__(self, d_model: int, ffn_dim: int):
+        super().__init__()
+        self.v1 = nn.Linear(d_model, ffn_dim, bias=False)
+        self.w1 = nn.Linear(d_model, ffn_dim, bias=False)
+        self.w2 = nn.Linear(ffn_dim, d_model, bias=False)
+        self.act_fn = nn.silu
+
+    def __call__(self, x: mx.array) -> mx.array:
+        current_hidden_states = self.act_fn(self.w1(x)) * self.v1(x)
+        current_hidden_states = self.w2(current_hidden_states)
+        return current_hidden_states
+
+
+class Router(nn.Module):
+    def __init__(self, d_model: int, num_experts: int):
+        super().__init__()
+        self.layer = nn.Linear(d_model, num_experts, bias=False)
+
+    def __call__(self, x: mx.array):
+        return self.layer(x)
+
+
+class SparseMoeBlock(nn.Module):
+    def __init__(self, args: ModelArgs):
+        super().__init__()
+        self.d_model = args.d_model
+        self.ffn_dim = args.ffn_config["ffn_hidden_size"]
+        self.num_experts = args.ffn_config["moe_num_experts"]
+        self.num_experts_per_tok = args.ffn_config["moe_top_k"]
+
+        self.router = Router(self.d_model, self.num_experts)
+        self.experts = [
+            MLP(self.d_model, self.ffn_dim) for _ in range(self.num_experts)
+        ]
+
+    def __call__(self, x: mx.array) -> mx.array:
+        ne = self.num_experts_per_tok
+        orig_shape = x.shape
+        x = x.reshape(-1, x.shape[-1])
+
+        gates = self.router(x)
+        gates = mx.softmax(gates.astype(mx.float32), axis=-1)
+
+        inds = mx.stop_gradient(mx.argpartition(-gates, kth=ne, axis=-1)[:, :ne])
+        scores = mx.take_along_axis(gates, inds, axis=-1)
+        scores = scores / mx.linalg.norm(scores, ord=1, axis=-1, keepdims=True)
+        scores = scores.astype(x.dtype)
+
+        if self.training:
+            inds = np.array(inds)
+            y = mx.zeros((x.shape[0], ne, x.shape[-1]), x.dtype)
+            for e, expert in enumerate(self.experts):
+                idx1, idx2 = map(mx.array, np.where(inds == e))
+                if idx1.size == 0:
+                    continue
+                y[idx1, idx2] = expert(x[idx1])
+
+            y = (y * scores[:, :, None]).sum(axis=1)
+        else:
+            y = []
+            for xt, st, it in zip(x, scores, inds.tolist()):
+                yt = mx.stack([self.experts[e](xt) for e in it], axis=-1)
+                yt = (yt * st).sum(axis=-1)
+                y.append(yt)
+            y = mx.stack(y, axis=0)
+
+        return y.reshape(orig_shape)
+
+
+class DecoderLayer(nn.Module):
+    def __init__(self, args: ModelArgs):
+        super().__init__()
+        self.ffn = SparseMoeBlock(args)
+        self.norm_attn_norm = NormAttnNorm(args)
+
+    def __call__(
+        self,
+        x: mx.array,
+        mask: Optional[mx.array] = None,
+        cache: Optional[Tuple[mx.array, mx.array]] = None,
+    ) -> mx.array:
+        r, h, cache = self.norm_attn_norm(x, mask, cache)
+        out = self.ffn(h) + r
+        return out, cache
+
+
+class DBRX(nn.Module):
+    def __init__(self, args: ModelArgs):
+        super().__init__()
+        self.vocab_size = args.vocab_size
+        self.wte = nn.Embedding(args.vocab_size, args.d_model)
+        self.blocks = [DecoderLayer(args=args) for _ in range(args.n_layers)]
+        self.norm_f = nn.LayerNorm(args.d_model, bias=False)
+
+    def __call__(
+        self,
+        inputs: mx.array,
+        cache=None,
+    ):
+        h = self.wte(inputs)
+
+        mask = None
+        T = h.shape[1]
+        if T > 1:
+            mask = nn.MultiHeadAttention.create_additive_causal_mask(T)
+            mask = mask.astype(h.dtype)
+
+        if cache is None:
+            cache = [None] * len(self.blocks)
+
+        for e, layer in enumerate(self.blocks):
+            h, cache[e] = layer(h, mask, cache[e])
+
+        return self.norm_f(h), cache
+
+
+class Model(nn.Module):
+    def __init__(self, args: ModelArgs):
+        super().__init__()
+        self.model_type = args.model_type
+        self.transformer = DBRX(args)
+        self.lm_head = nn.Linear(args.d_model, args.vocab_size, bias=False)
+        self.args = args
+
+    def __call__(
+        self,
+        inputs: mx.array,
+        cache=None,
+    ):
+        out, cache = self.transformer(inputs, cache)
+        return self.lm_head(out), cache
+
+    @property
+    def layers(self):
+        return self.transformer.blocks
+
+    def sanitize(self, weights):
+        # Split experts into sub matrices
+        num_experts = self.args.ffn_config["moe_num_experts"]
+        dim = self.args.ffn_config["ffn_hidden_size"]
+
+        pattern = "experts.mlp"
+        new_weights = {k: v for k, v in weights.items() if pattern not in k}
+        for k, v in weights.items():
+            if pattern in k:
+                experts = [
+                    (k.replace(".mlp", f".{e}") + ".weight", sv)
+                    for e, sv in enumerate(mx.split(v, num_experts, axis=0))
+                ]
+                if k.endswith("w2"):
+                    experts = [(s, sv.T) for s, sv in experts]
+                new_weights.update(experts)
+        return new_weights

llms/mlx_lm/models/mixtral.py

@@ -143,7 +143,6 @@ def __call__(self, x: mx.array) -> mx.array:
         ).astype(gates.dtype)
 
         if self.training:
-            mx.eval(inds)
             inds = np.array(inds)
             y = mx.zeros((x.shape[0], ne, x.shape[-1]), x.dtype)
             for e, expert in enumerate(self.experts):
@@ -156,7 +155,7 @@ def __call__(self, x: mx.array) -> mx.array:
         else:
             y = []
             for xt, st, it in zip(x, scores, inds.tolist()):
-                yt = mx.concatenate([self.experts[e](xt)[:, None] for e in it], axis=-1)
+                yt = mx.stack([self.experts[e](xt) for e in it], axis=-1)
                 yt = (yt * st).sum(axis=-1)
                 y.append(yt[None, :])
             y = mx.concatenate(y)

llms/mlx_lm/models/phixtral.py

@@ -125,7 +125,7 @@ def __call__(self, x: mx.array) -> mx.array:
         else:
             y = []
             for xt, st, it in zip(x, scores, inds.tolist()):
-                yt = mx.concatenate([self.mlp[e](xt)[:, None] for e in it], axis=-1)
+                yt = mx.stack([self.mlp[e](xt) for e in it], axis=-1)
                 yt = (yt * st).sum(axis=-1)
                 y.append(yt[None, :])
             y = mx.concatenate(y)

llms/mlx_lm/tuner/utils.py

@@ -60,6 +60,8 @@ def linear_to_lora_layers(
         keys = set(["att_proj"])
     elif model.model_type == "phi-msft":
         keys = set(["mixer.Wqkv", "moe.gate"])
+    elif model.model_type == "dbrx":
+        keys = set(["norm_attn_norm.attn.Wqkv", "ffn.router.layer"])
     else:
         raise ValueError(f"Lora does not support {model.model_type}")