mirrors/ollama-for-amd
1
0
Fork 0
mirror of https://github.com/likelovewant/ollama-for-amd.git synced 2025-12-23 23:18:26 +00:00
Code Issues Packages Projects Releases Wiki Activity

model: ministral w/ llama4 scaling (#13292)

Browse Source 
This change:

* fixes rope scaling in the mistral converter
* updates ministral to include llama4 scaling
* includes a new ministral parser for parsing reasoning and tool calling

---------

Co-authored-by: jmorganca <jmorganca@gmail.com>
This commit is contained in:
Patrick Devine
2025-12-01 23:20:14 -08:00
committed by GitHub
parent 554172759c
commit d3e0a0dee4
9 changed files with 379 additions and 42 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
model/models/mistral3/model.go
View File

@@ -159,8 +159,9 @@ func (m *Model) PostTokenize(inputs []*input.Input) ([]*input.Input, error) {
func (m *Model) Forward(ctx ml.Context, batch input.Batch) (ml.Tensor, error) {
positions := ctx.Input().FromInts(batch.Positions, len(batch.Positions))
positionsScale := m.getScale(ctx, batch.Positions)
return m.TextModel.Forward(ctx, batch.Inputs, positions, batch.Outputs, batch, m.Cache), nil
return m.TextModel.Forward(ctx, batch.Inputs, positions, positionsScale, batch.Outputs, batch, m.Cache), nil
}
func init() {