chore(lora): support mixtral in lora example (#343)

lora/convert.py

@@ -20,7 +20,13 @@ def quantize(weights, config, args):
     model.load_weights(list(weights.items()))
 
     # Quantize the model:
-    nn.QuantizedLinear.quantize_module(model, args.q_group_size, args.q_bits)
+    nn.QuantizedLinear.quantize_module(
+        model,
+        args.q_group_size,
+        args.q_bits,
+        linear_class_predicate=lambda m: isinstance(m, nn.Linear)
+        and m.weight.shape[0] != 8,
+    )
 
     # Update the config:
     quantized_config["quantization"] = {

lora/fuse.py

@@ -56,6 +56,8 @@ if __name__ == "__main__":
     for l in model.model.layers[-lora_layers:]:
         l.self_attn.q_proj = LoRALinear.from_linear(l.self_attn.q_proj)
         l.self_attn.v_proj = LoRALinear.from_linear(l.self_attn.v_proj)
+        if hasattr(l, "block_sparse_moe"):
+            l.block_sparse_moe.gate = LoRALinear.from_linear(l.block_sparse_moe.gate)
 
     model.update(tree_unflatten(adapters))
     fused_linears = [

lora/lora.py

@@ -315,6 +315,8 @@ if __name__ == "__main__":
     for l in model.model.layers[len(model.model.layers) - args.lora_layers :]:
         l.self_attn.q_proj = LoRALinear.from_linear(l.self_attn.q_proj)
         l.self_attn.v_proj = LoRALinear.from_linear(l.self_attn.v_proj)
+        if hasattr(l, "block_sparse_moe"):
+            l.block_sparse_moe.gate = LoRALinear.from_linear(l.block_sparse_moe.gate)
 
     p = sum(v.size for _, v in tree_flatten(model.parameters())) / 10**6
     print(f"Total parameters {p:.3f}M")
@@ -349,7 +351,7 @@ if __name__ == "__main__":
 
     if args.test:
         print("Testing")
-
+        model.eval()
         test_loss = evaluate(
             model,
             test_set,

lora/models.py

@@ -328,7 +328,12 @@ def load(path_or_hf_repo: str):
 
     model = Model(model_args)
     if quantization is not None:
-        nn.QuantizedLinear.quantize_module(model, **quantization)
+        nn.QuantizedLinear.quantize_module(
+            model,
+            **quantization,
+            linear_class_predicate=lambda m: isinstance(m, nn.Linear)
+            and m.weight.shape[0] != 8,
+        )
 
     model.load_weights(list(weights.items()))
 

lora/models/mixtral.py

@@ -0,0 +1,253 @@
+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):
+    vocab_size: int = 32000
+    max_position_embeddings: int = 4096 * 32
+    hidden_size: int = 4096
+    intermediate_size: int = 14336
+    num_hidden_layers: int = 32
+    num_attention_heads: int = 32
+    num_experts_per_tok: int = 2
+    num_key_value_heads: int = 8
+    num_local_experts: int = 8
+    rms_norm_eps: float = 1e-5
+    vocab_size: int
+    rope_theta: float = 1e6
+    rope_traditional: bool = False
+    model_type: str = None
+    rope_scaling: Optional[Dict[str, Union[float, str]]] = None
+
+    def __post_init__(self):
+        if self.num_key_value_heads is None:
+            self.num_key_value_heads = self.num_attention_heads
+
+
+class RMSNorm(nn.Module):
+    def __init__(self, dims: int, eps: float = 1e-5):
+        super().__init__()
+        self.weight = mx.ones((dims,))
+        self.eps = eps
+
+    def _norm(self, x):
+        return x * mx.rsqrt(x.square().mean(-1, keepdims=True) + self.eps)
+
+    def __call__(self, x):
+        output = self._norm(x.astype(mx.float32)).astype(x.dtype)
+        return self.weight * output
+
+
+class MixtralAttention(nn.Module):
+    def __init__(self, args: ModelArgs):
+        super().__init__()
+        self.hidden_size = args.hidden_size
+        self.num_heads = args.num_attention_heads
+        self.head_dim = self.hidden_size // self.num_heads
+        self.num_key_value_heads = args.num_key_value_heads
+        self.max_position_embeddings = args.max_position_embeddings
+        self.rope_theta = args.rope_theta
+
+        self.repeats = self.num_heads // self.num_key_value_heads
+
+        self.scale = self.head_dim**-0.5
+
+        self.q_proj = nn.Linear(
+            self.hidden_size, self.num_heads * self.head_dim, bias=False
+        )
+        self.k_proj = nn.Linear(
+            self.hidden_size, self.num_key_value_heads * self.head_dim, bias=False
+        )
+        self.v_proj = nn.Linear(
+            self.hidden_size, self.num_key_value_heads * self.head_dim, bias=False
+        )
+        self.o_proj = nn.Linear(
+            self.num_heads * self.head_dim, self.hidden_size, bias=False
+        )
+
+        self.rope = nn.RoPE(
+            self.head_dim,
+            traditional=args.rope_traditional,
+            base=args.rope_theta,
+        )
+
+    def __call__(
+        self,
+        x: mx.array,
+        mask: Optional[mx.array] = None,
+        cache: Optional[Tuple[mx.array, mx.array]] = 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.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
+        )
+
+        def repeat(a):
+            a = mx.concatenate([mx.expand_dims(a, 2)] * self.repeats, axis=2)
+            return a.reshape([B, self.num_heads, L, -1])
+
+        if self.repeats > 1:
+            keys, values = map(repeat, (keys, values))
+
+        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)
+
+        scores = (queries * self.scale) @ keys.transpose(0, 1, 3, 2)
+        if mask is not None:
+            scores += mask
+        scores = mx.softmax(scores.astype(mx.float32), axis=-1).astype(scores.dtype)
+        output = (scores @ values).transpose(0, 2, 1, 3).reshape(B, L, -1)
+        return self.o_proj(output), (keys, values)
+
+
+class MixtralBLockSparseTop2MLP(nn.Module):
+    def __init__(self, args: ModelArgs):
+        super().__init__()
+        self.ffn_dim = args.intermediate_size
+        self.hidden_dim = args.hidden_size
+
+        self.w1 = nn.Linear(self.hidden_dim, self.ffn_dim, bias=False)
+        self.w2 = nn.Linear(self.ffn_dim, self.hidden_dim, bias=False)
+        self.w3 = nn.Linear(self.hidden_dim, self.ffn_dim, 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.w3(x)
+        current_hidden_states = self.w2(current_hidden_states)
+        return current_hidden_states
+
+
+class MixtralSparseMoeBlock(nn.Module):
+    def __init__(self, args: ModelArgs):
+        super().__init__()
+        self.hidden_dim = args.hidden_size
+        self.ffn_dim = args.intermediate_size
+        self.num_experts = args.num_local_experts
+        self.num_experts_per_tok = args.num_experts_per_tok
+
+        # gating
+        self.gate = nn.Linear(self.hidden_dim, self.num_experts, bias=False)
+
+        self.experts = [
+            MixtralBLockSparseTop2MLP(args=args) 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.gate(x)
+        inds = mx.stop_gradient(mx.argpartition(-gates, kth=ne, axis=-1)[:, :ne])
+
+        scores = mx.softmax(
+            mx.take_along_axis(gates, inds, axis=-1).astype(mx.float32),
+            axis=-1,
+        ).astype(gates.dtype)
+
+        mx.eval(inds)
+        inds = np.array(inds)
+        y = mx.zeros((x.shape[0], ne, x.shape[-1]))
+        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)
+
+        return y.reshape(orig_shape)
+
+
+class MixtralDecoderLayer(nn.Module):
+    def __init__(self, args: ModelArgs):
+        super().__init__()
+        self.hidden_size = args.hidden_size
+
+        self.self_attn = MixtralAttention(args)
+
+        self.block_sparse_moe = MixtralSparseMoeBlock(args)
+        self.input_layernorm = RMSNorm(args.hidden_size, eps=args.rms_norm_eps)
+        self.post_attention_layernorm = RMSNorm(args.hidden_size, eps=args.rms_norm_eps)
+
+    def __call__(
+        self,
+        x: mx.array,
+        mask: Optional[mx.array] = None,
+        cache: Optional[Tuple[mx.array, mx.array]] = None,
+    ) -> mx.array:
+        r, cache = self.self_attn(self.input_layernorm(x), mask, cache)
+        h = x + r
+        r = self.block_sparse_moe(self.post_attention_layernorm(h))
+        out = h + r
+        return out, cache
+
+
+class MixtralModel(nn.Module):
+    def __init__(self, args: ModelArgs):
+        super().__init__()
+        self.vocab_size = args.vocab_size
+        self.num_hidden_layers = args.num_hidden_layers
+
+        self.embed_tokens = nn.Embedding(args.vocab_size, args.hidden_size)
+        self.layers = [
+            MixtralDecoderLayer(args=args) for _ in range(args.num_hidden_layers)
+        ]
+        self.norm = RMSNorm(args.hidden_size, eps=args.rms_norm_eps)
+
+    def __call__(
+        self,
+        inputs: mx.array,
+        cache=None,
+    ):
+        h = self.embed_tokens(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.layers)
+
+        for e, layer in enumerate(self.layers):
+            h, cache[e] = layer(h, mask, cache[e])
+
+        return self.norm(h), cache
+
+
+class Model(nn.Module):
+    def __init__(self, args: ModelArgs):
+        super().__init__()
+        self.model = MixtralModel(args)
+        self.lm_head = nn.Linear(args.hidden_size, args.vocab_size, bias=False)
+
+    def __call__(
+        self,
+        inputs: mx.array,
+        cache=None,
+    ):
+        out, cache = self.model(inputs, cache)
+        return self.lm_head(out), cache

lora/utils.py

@@ -9,6 +9,7 @@ from typing import Generator
 import mlx.core as mx
 import mlx.nn as nn
 import models.llama as llama
+import models.mixtral as mixtral
 import models.phi2 as phi2
 import transformers
 from huggingface_hub import snapshot_download
@@ -18,6 +19,7 @@ MODEL_MAPPING = {
     "llama": llama,
     "mistral": llama,  # mistral is compatible with llama
     "phi": phi2,
+    "mixtral": mixtral,
 }
 
 
@@ -150,7 +152,12 @@ def load(path_or_hf_repo: str):
     model_args = model_args_class.from_dict(config)
     model = model_class(model_args)
     if quantization is not None:
-        nn.QuantizedLinear.quantize_module(model, **quantization)
+        nn.QuantizedLinear.quantize_module(
+            model,
+            **quantization,
+            linear_class_predicate=lambda m: isinstance(m, nn.Linear)
+            and m.weight.shape[0] != 8,
+        )
 
     model.load_weights(list(weights.items()))