mirror of
https://github.com/ollama/ollama.git
synced 2026-10-21 19:11:17 +00:00
2e036e7cdf
The MLX runner is the only Go inference runner left and is no longer experimental, so its packages leave x/. The bindings become a top-level mlx package beside the carried patches in mlx/compat, mirroring how llama/ holds the llama.cpp integration, and the runner becomes mlxrunner with the architectures nested under the package they implement. Subpackages move with their parent unless listed. x/mlxrunner/mlx mlx x/internal/mlxthread mlx/mlxthread x/internal/mlxthreadtest mlx/mlxthread/mlxthreadtest x/internal/mlxtest mlx/mlxtest x/quant mlx/quant mlx/compat/*.patch mlx/compat/mlx-c (MLX patches go in mlx/compat/mlx) x/mlxrunner mlxrunner x/models/nn mlxrunner/nn x/models/<arch> mlxrunner/model/<arch> x/mlxrunner/imports.go mlxrunner/model/architectures (new package) x/create create x/safetensors fs/safetensors x/tokenizer mlxrunner/tokenizer Every package keeps its name, so the Go changes are the import path rewrites the moves force, and the CMake, Dockerfile, CI cache keys, drift check and Darwin payload script follow the new paths. Four edits are not paths: the runner's blank architecture imports become the package mlxrunner/model/architectures, so the list to extend for a new model sits beside the architecture directories; a depguard rule keeps the two test harnesses out of non-test code, as the x/internal placement used to; the CI change filter's two entries for the long-deleted x/imagegen/mlx now name the bindings' CMake project and the carried patches, so a change to either builds the payload; and the tokenizer parity test reads its fixtures from its own testdata instead of walking out of x/. x/server and x/imagegen/manifest stay for the next two commits.
93 lines
2.2 KiB
Go
93 lines
2.2 KiB
Go
package model
|
|
|
|
import (
|
|
"github.com/ollama/ollama/mlx"
|
|
"github.com/ollama/ollama/mlxrunner/nn"
|
|
)
|
|
|
|
// LinearFactory builds linear layers using shared tensor maps and quant defaults.
|
|
type LinearFactory struct {
|
|
tensors map[string]*mlx.Array
|
|
defaultGroupSize int
|
|
defaultBits int
|
|
defaultMode string
|
|
tensorQuant map[string]*TensorQuantInfo
|
|
}
|
|
|
|
// NewLinearFactory creates a reusable constructor for model linear layers.
|
|
func NewLinearFactory(
|
|
tensors map[string]*mlx.Array,
|
|
defaultGroupSize, defaultBits int,
|
|
defaultMode string,
|
|
tensorQuant map[string]*TensorQuantInfo,
|
|
) LinearFactory {
|
|
return LinearFactory{
|
|
tensors: tensors,
|
|
defaultGroupSize: defaultGroupSize,
|
|
defaultBits: defaultBits,
|
|
defaultMode: defaultMode,
|
|
tensorQuant: tensorQuant,
|
|
}
|
|
}
|
|
|
|
// Make constructs a linear layer at path.
|
|
func (f LinearFactory) Make(path string) nn.LinearLayer {
|
|
return MakeLinearLayer(
|
|
f.tensors,
|
|
path,
|
|
f.defaultGroupSize,
|
|
f.defaultBits,
|
|
f.defaultMode,
|
|
f.tensorQuant,
|
|
)
|
|
}
|
|
|
|
// MakeLinearLayer constructs a linear layer from a tensor map.
|
|
//
|
|
// For quantized tensors (path.weight + path.weight_scale), it resolves per-tensor
|
|
// quant params via TensorQuant metadata (with shape-based affine fallback).
|
|
// For non-quantized tensors, it returns a standard nn.Linear.
|
|
func MakeLinearLayer(
|
|
tensors map[string]*mlx.Array,
|
|
path string,
|
|
defaultGroupSize, defaultBits int,
|
|
defaultMode string,
|
|
tensorQuant map[string]*TensorQuantInfo,
|
|
) nn.LinearLayer {
|
|
w := tensors[path+".weight"]
|
|
if w == nil {
|
|
return nil
|
|
}
|
|
|
|
scales := tensors[path+".weight_scale"]
|
|
if scales != nil {
|
|
qbiases := tensors[path+".weight_qbias"]
|
|
bias := tensors[path+".bias"]
|
|
|
|
groupSize, bits, mode := ResolveLinearQuantParams(
|
|
defaultGroupSize,
|
|
defaultBits,
|
|
defaultMode,
|
|
tensorQuant,
|
|
path+".weight",
|
|
w,
|
|
scales,
|
|
)
|
|
|
|
globalScale, _ := ReadGlobalScale(tensors, path+".weight")
|
|
|
|
return &nn.QuantizedLinear{
|
|
Weight: w,
|
|
Scales: scales,
|
|
QBiases: qbiases,
|
|
Bias: bias,
|
|
GlobalScale: globalScale,
|
|
GroupSize: groupSize,
|
|
Bits: bits,
|
|
Mode: mode,
|
|
}
|
|
}
|
|
|
|
bias := tensors[path+".bias"]
|
|
return nn.NewLinear(w, bias)
|
|
}
|