jeans/ollama
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

mistral3 arch

Browse Source
This commit is contained in:
Bruce MacDonald 2025-03-20 12:44:02 -07:00 committed by jmorganca
parent 9a12fd1067
commit 3b4ad00a4b
4 changed files with 86 additions and 43 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
convert/convert.go
View File

@ -185,7 +185,7 @@ func ConvertModel(fsys fs.FS, ws io.WriteSeeker) error {
case "LlamaForCausalLM": case "LlamaForCausalLM":
conv = &llamaModel{} conv = &llamaModel{}
case "Mistral3ForConditionalGeneration": case "Mistral3ForConditionalGeneration":
conv = &mistralModel{} conv = &mistral3Model{}
case "MixtralForCausalLM": case "MixtralForCausalLM":
conv = &mixtralModel{} conv = &mixtralModel{}
case "GemmaForCausalLM": case "GemmaForCausalLM":

85
convert/convert_mistral.go
View File

@ -11,8 +11,11 @@ import (
"github.com/ollama/ollama/fs/ggml" "github.com/ollama/ollama/fs/ggml"
) )
type mistralModel struct { type mistral3Model struct {
ModelParameters ModelParameters
// ImageTokenIndex uint32 `json:"image_token_index"`
// SpatialMergeSize uint32 `json:"spatial_merge_size"`
// VisionFeatureLayer int32 `json:"vision_feature_layer"`
TextModel struct { TextModel struct {
NumHiddenLayers uint32 `json:"num_hidden_layers"` NumHiddenLayers uint32 `json:"num_hidden_layers"`
MaxPositionEmbeddings uint32 `json:"max_position_embeddings"` MaxPositionEmbeddings uint32 `json:"max_position_embeddings"`
@ -23,30 +26,62 @@ type mistralModel struct {
RopeTheta float32 `json:"rope_theta"` RopeTheta float32 `json:"rope_theta"`
RMSNormEPS float32 `json:"rms_norm_eps"` RMSNormEPS float32 `json:"rms_norm_eps"`
HeadDim uint32 `json:"head_dim"` HeadDim uint32 `json:"head_dim"`
SlidingWindow *uint32 `json:"sliding_window"`
HiddenAct string `json:"hidden_act"`
VocabSize uint32 `json:"vocab_size"`
} `json:"text_config"` } `json:"text_config"`
// VisionModel struct {
// NumAttentionHeads uint32 `json:"num_attention_heads"`
// NumHiddenLayers uint32 `json:"num_hidden_layers"`
// HiddenSize uint32 `json:"hidden_size"`
// IntermediateSize uint32 `json:"intermediate_size"`
// ImageSize uint32 `json:"image_size"`
// NumChannels uint32 `json:"num_channels"`
// PatchSize uint32 `json:"patch_size"`
// HeadDim uint32 `json:"head_dim"`
// HiddenAct string `json:"hidden_act"`
// RopeTheta float32 `json:"rope_theta"`
// } `json:"vision_config"`
// MultiModalProjectorBias bool `json:"multimodal_projector_bias"`
// ProjectorHiddenAct string `json:"projector_hidden_act"`
} }
func (p *mistralModel) KV(t *Tokenizer) ggml.KV { func (p *mistral3Model) KV(t *Tokenizer) ggml.KV {
kv := p.ModelParameters.KV(t) kv := p.ModelParameters.KV(t)
kv["general.architecture"] = "mistral" kv["general.architecture"] = "mistral3"
kv["mistral.vocab_size"] = p.VocabSize kv["mistral3.vocab_size"] = p.TextModel.VocabSize
kv["mistral.block_count"] = p.TextModel.NumHiddenLayers // Text configuration
kv["mistral.context_length"] = p.TextModel.MaxPositionEmbeddings kv["mistral3.block_count"] = p.TextModel.NumHiddenLayers
kv["mistral.embedding_length"] = p.TextModel.HiddenSize kv["mistral3.context_length"] = p.TextModel.MaxPositionEmbeddings
kv["mistral.feed_forward_length"] = p.TextModel.IntermediateSize kv["mistral3.embedding_length"] = p.TextModel.HiddenSize
kv["mistral.attention.head_count"] = p.TextModel.NumAttentionHeads kv["mistral3.feed_forward_length"] = p.TextModel.IntermediateSize
kv["mistral.rope.dimension_count"] = p.TextModel.HiddenSize / p.TextModel.NumHiddenLayers kv["mistral3.attention.head_count"] = p.TextModel.NumAttentionHeads
kv["mistral.rope.freq_base"] = p.TextModel.RopeTheta kv["mistral3.attention.head_count_kv"] = p.TextModel.NumKeyValueHeads
kv["mistral.attention.head_count_kv"] = p.TextModel.NumKeyValueHeads kv["mistral3.attention.layer_norm_rms_epsilon"] = p.TextModel.RMSNormEPS
kv["mistral.attention.layer_norm_rms_epsilon"] = p.TextModel.RMSNormEPS kv["mistral3.attention.key_length"] = p.TextModel.HeadDim
kv["mistral.attention.key_length"] = p.TextModel.HeadDim kv["mistral3.attention.value_length"] = p.TextModel.HeadDim
kv["mistral.attention.value_length"] = p.TextModel.HeadDim kv["mistral3.rope.dimension_count"] = p.TextModel.HiddenSize / p.TextModel.NumHiddenLayers
kv["mistral3.rope.freq_base"] = p.TextModel.RopeTheta
// Multimodal configuration
// kv["mistral3.image_token_index"] = p.ImageTokenIndex
// kv["mistral3.spatial_merge_size"] = p.SpatialMergeSize
// if p.VisionFeatureLayer != 0 {
// kv["mistral3.vision_feature_layer"] = p.VisionFeatureLayer
// }
// kv["mistral3.mm.projector_bias"] = p.MultiModalProjectorBias
// if p.ProjectorHiddenAct != "" {
// kv["mistral3.mm.projector_hidden_act"] = p.ProjectorHiddenAct
// }
return kv return kv
} }
func (p *mistralModel) Tensors(ts []Tensor) []ggml.Tensor { func (p *mistral3Model) Tensors(ts []Tensor) []ggml.Tensor {
var out []ggml.Tensor var out []ggml.Tensor
for _, t := range ts { for _, t := range ts {
@ -55,10 +90,8 @@ func (p *mistralModel) Tensors(ts []Tensor) []ggml.Tensor {
t.SetRepacker(p.repack) t.SetRepacker(p.repack)
} }
if strings.HasPrefix(t.Name(), "patch_merger.") || // Skip certain vision model tensors that might need special handling
strings.HasPrefix(t.Name(), "pre_mm_projector_output_norm.") || if strings.HasPrefix(t.Name(), "patch_merger.") || strings.HasPrefix(t.Name(), "pre_mm_projector_output_norm.") {
strings.HasPrefix(t.Name(), "vision_encoder.") ||
strings.HasPrefix(t.Name(), "vision_language_adapter.") {
continue continue
} }
@ -73,8 +106,9 @@ func (p *mistralModel) Tensors(ts []Tensor) []ggml.Tensor {
return out return out
} }
func (p *mistralModel) Replacements() []string { func (p *mistral3Model) Replacements() []string {
return []string{ return []string{
// Text model replacements
"model.layers", "blk", "model.layers", "blk",
"input_layernorm", "attn_norm", "input_layernorm", "attn_norm",
"post_attention_layernorm", "ffn_norm", "post_attention_layernorm", "ffn_norm",
@ -121,14 +155,21 @@ func (p *mistralModel) Replacements() []string {
"vision_tower.transformer.layers.*.ffn_norm", "v.ffn_norm", "vision_tower.transformer.layers.*.ffn_norm", "v.ffn_norm",
"vision_tower.ln_pre", "v.encoder_norm", "vision_tower.ln_pre", "v.encoder_norm",
"vision_tower.patch_conv", "v.patch_conv", "vision_tower.patch_conv", "v.patch_conv",
"vision_tower.embeddings", "v.embeddings",
// Alternative vision model paths
"vision_model.vision_model.embeddings", "v.embeddings",
"vision_model.vision_model", "v",
"vision_model.layers", "v.blk",
// Multimodal projector components // Multimodal projector components
"multi_modal_projector.patch_merger", "mm.patch_merger", "multi_modal_projector.patch_merger", "mm.patch_merger",
"multi_modal_projector.norm", "mm.norm", "multi_modal_projector.norm", "mm.norm",
"multi_modal_projector.linear", "mm.projection",
} }
} }
func (p *mistralModel) repack(name string, data []float32, shape []uint64) ([]float32, error) { func (p *mistral3Model) repack(name string, data []float32, shape []uint64) ([]float32, error) {
var dims []int var dims []int
for _, dim := range shape { for _, dim := range shape {
dims = append(dims, int(dim)) dims = append(dims, int(dim))

22
model/models/gemma3/model_text.go
View File

@ -10,7 +10,7 @@ import (
"github.com/ollama/ollama/model/input" "github.com/ollama/ollama/model/input"
) )
type TextOptions struct { type TextConfig struct {
hiddenSize, numHeads, numKVHeads int hiddenSize, numHeads, numKVHeads int
attnKeyLen, attnValLen int attnKeyLen, attnValLen int
eps, ropeScale float32 eps, ropeScale float32
@ -27,7 +27,7 @@ type TextModel struct {
OutputNorm *nn.RMSNorm `gguf:"output_norm"` OutputNorm *nn.RMSNorm `gguf:"output_norm"`
Output *nn.Linear `gguf:"output,alt:token_embd"` Output *nn.Linear `gguf:"output,alt:token_embd"`
*TextOptions *TextConfig
} }
const ( const (
@ -55,7 +55,7 @@ func newTextModel(c ml.Config) *TextModel {
}, },
), ),
Layers: make([]TextLayer, numBlocks), Layers: make([]TextLayer, numBlocks),
TextOptions: &TextOptions{ TextConfig: &TextConfig{
hiddenSize: int(c.Uint("embedding_length")), hiddenSize: int(c.Uint("embedding_length")),
numHeads: int(c.Uint("attention.head_count")), numHeads: int(c.Uint("attention.head_count")),
numKVHeads: int(c.Uint("attention.head_count_kv")), numKVHeads: int(c.Uint("attention.head_count_kv")),
@ -84,7 +84,7 @@ type TextSelfAttention struct {
Output *nn.Linear `gguf:"attn_output"` Output *nn.Linear `gguf:"attn_output"`
} }
func (sa *TextSelfAttention) Forward(ctx ml.Context, layer int, hiddenState, positionIDs ml.Tensor, cache kvcache.Cache, opts *TextOptions) ml.Tensor { func (sa *TextSelfAttention) Forward(ctx ml.Context, layer int, hiddenState, positionIDs ml.Tensor, cache kvcache.Cache, opts *TextConfig) ml.Tensor {
batchSize := hiddenState.Dim(1) batchSize := hiddenState.Dim(1)
ropeType := uint32(2) ropeType := uint32(2)
@ -120,12 +120,12 @@ func (sa *TextSelfAttention) Forward(ctx ml.Context, layer int, hiddenState, pos
} }
func (m *TextModel) Shift(ctx ml.Context, layer int, key, shift ml.Tensor) (ml.Tensor, error) { func (m *TextModel) Shift(ctx ml.Context, layer int, key, shift ml.Tensor) (ml.Tensor, error) {
ropeBase := m.TextOptions.ropeLocalBase ropeBase := m.TextConfig.ropeLocalBase
if (layer+1)%gemmaGlobalCacheCount == 0 { if (layer+1)%gemmaGlobalCacheCount == 0 {
ropeBase = m.TextOptions.ropeGlobalBase ropeBase = m.TextConfig.ropeGlobalBase
} }
return key.RoPE(ctx, shift, nil, uint32(m.TextOptions.attnKeyLen), uint32(2), ropeBase, m.TextOptions.ropeScale), nil return key.RoPE(ctx, shift, nil, uint32(m.TextConfig.attnKeyLen), uint32(2), ropeBase, m.TextConfig.ropeScale), nil
} }
type TextMLP struct { type TextMLP struct {
@ -134,7 +134,7 @@ type TextMLP struct {
Gate *nn.Linear `gguf:"ffn_gate"` Gate *nn.Linear `gguf:"ffn_gate"`
} }
func (mlp *TextMLP) Forward(ctx ml.Context, hiddenState ml.Tensor, opts *TextOptions) ml.Tensor { func (mlp *TextMLP) Forward(ctx ml.Context, hiddenState ml.Tensor, opts *TextConfig) ml.Tensor {
hiddenState = mlp.Gate.Forward(ctx, hiddenState).GELU(ctx).Mul(ctx, mlp.Up.Forward(ctx, hiddenState)) hiddenState = mlp.Gate.Forward(ctx, hiddenState).GELU(ctx).Mul(ctx, mlp.Up.Forward(ctx, hiddenState))
return mlp.Down.Forward(ctx, hiddenState) return mlp.Down.Forward(ctx, hiddenState)
} }
@ -148,7 +148,7 @@ type TextLayer struct {
PostMLPNorm *nn.RMSNorm `gguf:"post_ffw_norm"` PostMLPNorm *nn.RMSNorm `gguf:"post_ffw_norm"`
} }
func (l *TextLayer) Forward(ctx ml.Context, layer int, hiddenState, positionIDs, outputs ml.Tensor, cache kvcache.Cache, opts *TextOptions) ml.Tensor { func (l *TextLayer) Forward(ctx ml.Context, layer int, hiddenState, positionIDs, outputs ml.Tensor, cache kvcache.Cache, opts *TextConfig) ml.Tensor {
residual := hiddenState residual := hiddenState
hiddenState = l.AttentionNorm.Forward(ctx, hiddenState, opts.eps) hiddenState = l.AttentionNorm.Forward(ctx, hiddenState, opts.eps)
@ -173,7 +173,7 @@ func (l *TextLayer) Forward(ctx ml.Context, layer int, hiddenState, positionIDs,
func (m *TextModel) Forward(ctx ml.Context, inputs, positions, outputs ml.Tensor, batch input.Batch, cache kvcache.Cache) ml.Tensor { func (m *TextModel) Forward(ctx ml.Context, inputs, positions, outputs ml.Tensor, batch input.Batch, cache kvcache.Cache) ml.Tensor {
hiddenState := m.TokenEmbedding.Forward(ctx, inputs) hiddenState := m.TokenEmbedding.Forward(ctx, inputs)
hiddenState = hiddenState.Scale(ctx, math.Sqrt(float64(m.TextOptions.hiddenSize))) hiddenState = hiddenState.Scale(ctx, math.Sqrt(float64(m.TextConfig.hiddenSize)))
// set image embeddings // set image embeddings
var except []int var except []int
@ -206,7 +206,7 @@ func (m *TextModel) Forward(ctx ml.Context, inputs, positions, outputs ml.Tensor
lastLayerOutputs = outputs lastLayerOutputs = outputs
} }
hiddenState = layer.Forward(ctx, i, hiddenState, positions, lastLayerOutputs, cache, m.TextOptions) hiddenState = layer.Forward(ctx, i, hiddenState, positions, lastLayerOutputs, cache, m.TextConfig)
} }
hiddenState = m.OutputNorm.Forward(ctx, hiddenState, m.eps) hiddenState = m.OutputNorm.Forward(ctx, hiddenState, m.eps)

20
model/models/mistral/model.go
View File

@ -1,4 +1,4 @@
package llama package mistral3
import ( import (
"fmt" "fmt"
@ -12,7 +12,7 @@ import (
"github.com/ollama/ollama/model/input" "github.com/ollama/ollama/model/input"
) )
type Options struct { type TextOptions struct {
hiddenSize, numHeads, numKVHeads, headDim int hiddenSize, numHeads, numKVHeads, headDim int
eps, ropeBase, ropeScale float32 eps, ropeBase, ropeScale float32
ropeDim uint32 ropeDim uint32
@ -27,7 +27,7 @@ type Model struct {
OutputNorm *nn.RMSNorm `gguf:"output_norm"` OutputNorm *nn.RMSNorm `gguf:"output_norm"`
Output *nn.Linear `gguf:"output,alt:token_embd"` Output *nn.Linear `gguf:"output,alt:token_embd"`
*Options *TextOptions
} }
func New(c ml.Config) (model.Model, error) { func New(c ml.Config) (model.Model, error) {
@ -49,7 +49,7 @@ func New(c ml.Config) (model.Model, error) {
}, },
), ),
Layers: make([]Layer, c.Uint("block_count")), Layers: make([]Layer, c.Uint("block_count")),
Options: &Options{ TextOptions: &TextOptions{
hiddenSize: int(c.Uint("embedding_length")), hiddenSize: int(c.Uint("embedding_length")),
numHeads: int(c.Uint("attention.head_count")), numHeads: int(c.Uint("attention.head_count")),
numKVHeads: int(c.Uint("attention.head_count_kv")), numKVHeads: int(c.Uint("attention.head_count_kv")),
@ -74,7 +74,7 @@ type SelfAttention struct {
RopeFactors ml.Tensor `gguf:"rope_freqs.weight"` RopeFactors ml.Tensor `gguf:"rope_freqs.weight"`
} }
func (sa *SelfAttention) Forward(ctx ml.Context, hiddenState, positionIDs ml.Tensor, cache kvcache.Cache, opts *Options) ml.Tensor { func (sa *SelfAttention) Forward(ctx ml.Context, hiddenState, positionIDs ml.Tensor, cache kvcache.Cache, opts *TextOptions) ml.Tensor {
batchSize := hiddenState.Dim(1) batchSize := hiddenState.Dim(1)
ropeType := uint32(0) ropeType := uint32(0)
// Get head dimension - use explicit value if available, otherwise calculate // Get head dimension - use explicit value if available, otherwise calculate
@ -119,7 +119,7 @@ type MLP struct {
Gate *nn.Linear `gguf:"ffn_gate"` Gate *nn.Linear `gguf:"ffn_gate"`
} }
func (mlp *MLP) Forward(ctx ml.Context, hiddenState ml.Tensor, opts *Options) ml.Tensor { func (mlp *MLP) Forward(ctx ml.Context, hiddenState ml.Tensor, opts *TextOptions) ml.Tensor {
hiddenState = mlp.Gate.Forward(ctx, hiddenState).SILU(ctx).Mul(ctx, mlp.Up.Forward(ctx, hiddenState)) hiddenState = mlp.Gate.Forward(ctx, hiddenState).SILU(ctx).Mul(ctx, mlp.Up.Forward(ctx, hiddenState))
return mlp.Down.Forward(ctx, hiddenState) return mlp.Down.Forward(ctx, hiddenState)
} }
@ -131,7 +131,7 @@ type Layer struct {
MLP *MLP MLP *MLP
} }
func (l *Layer) Forward(ctx ml.Context, hiddenState, positionIDs, outputs ml.Tensor, cache kvcache.Cache, opts *Options) ml.Tensor { func (l *Layer) Forward(ctx ml.Context, hiddenState, positionIDs, outputs ml.Tensor, cache kvcache.Cache, opts *TextOptions) ml.Tensor {
residual := hiddenState residual := hiddenState
hiddenState = l.AttentionNorm.Forward(ctx, hiddenState, opts.eps) hiddenState = l.AttentionNorm.Forward(ctx, hiddenState, opts.eps)
@ -168,8 +168,10 @@ func (m *Model) Forward(ctx ml.Context, opts input.Options) (ml.Tensor, error) {
return nil, err return nil, err
} }
// Process text inputs
hiddenState := m.TokenEmbedding.Forward(ctx, inputs) hiddenState := m.TokenEmbedding.Forward(ctx, inputs)
// Process through text transformer layers
for i, layer := range m.Layers { for i, layer := range m.Layers {
m.Cache.SetLayer(i) m.Cache.SetLayer(i)
@ -178,7 +180,7 @@ func (m *Model) Forward(ctx ml.Context, opts input.Options) (ml.Tensor, error) {
lastLayerOutputs = outputs lastLayerOutputs = outputs
} }
hiddenState = layer.Forward(ctx, hiddenState, positions, lastLayerOutputs, m.Cache, m.Options) hiddenState = layer.Forward(ctx, hiddenState, positions, lastLayerOutputs, m.Cache, m.TextOptions)
} }
hiddenState = m.OutputNorm.Forward(ctx, hiddenState, m.eps) hiddenState = m.OutputNorm.Forward(ctx, hiddenState, m.eps)
@ -186,5 +188,5 @@ func (m *Model) Forward(ctx ml.Context, opts input.Options) (ml.Tensor, error) {
} }
func init() { func init() {
model.Register("mistral", New) model.Register("mistral3", New)
} }