mirrors/ollama
1
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2025-05-11 18:36:41 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

Move quantization to new backend (#10363)

Browse source 
* Move quantization logic to GGML via new backend

This moves the model aware logic to Go code and calls GGMLs quantization code for model creation.

* Remove "add model quantizations"

This is no longer needed now that quantization is implemented in Go+GGML code directly.
This commit is contained in:
Daniel Hiltgen 2025-05-06 11:20:48 -07:00 • committed by GitHub
parent 95e744beeb
commit 424810450f
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
39 changed files with 1854 additions and 440 deletions
Download patch file Download diff file Expand all files Collapse all files

2
llm/memory_test.go
View file

@ -25,7 +25,7 @@ func TestEstimateGPULayers(t *testing.T) {
defer f.Close()
inputLayerCount := 5
tensors := []ggml.Tensor{
tensors := []*ggml.Tensor{
{Name: "blk.0.attn.weight", Kind: uint32(0), Offset: uint64(0), Shape: []uint64{1, 1, 1, 1}, WriterTo: bytes.NewReader(make([]byte, 32))},
{Name: "blk.1.attn.weight", Kind: uint32(0), Offset: uint64(0), Shape: []uint64{1, 1, 1, 1}, WriterTo: bytes.NewReader(make([]byte, 32))},
{Name: "blk.2.attn.weight", Kind: uint32(0), Offset: uint64(0), Shape: []uint64{1, 1, 1, 1}, WriterTo: bytes.NewReader(make([]byte, 32))},