highperfocused/ollama
1
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2025-09-26 02:47:33 +02:00
Code Issues Packages Projects Releases Wiki Activity

model: add bos token if configured

Browse Source
This commit is contained in:
Michael Yang
2025-02-26 11:20:51 -08:00
parent e185c08ad9
commit 53d2990d9b
5 changed files with 32 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
model/models/llama/model.go
View File

@@ -37,7 +37,9 @@ func New(c ml.Config) (model.Model, error) {
Types: c.Uints("tokenizer.ggml.token_type"),
Merges: c.Strings("tokenizer.ggml.merges"),
BOS: int32(c.Uint("tokenizer.ggml.bos_token_id")),
AddBOS: c.Bool("tokenizer.ggml.add_bos_token", true),
EOS: int32(c.Uint("tokenizer.ggml.eos_token_id")),
AddEOS: c.Bool("tokenizer.ggml.add_eos_token", false),
},
),
Layers: make([]Layer, c.Uint("block_count")),

2
model/models/mllama/model.go
View File

@@ -33,7 +33,9 @@ func New(c ml.Config) (model.Model, error) {
Types: c.Uints("tokenizer.ggml.token_type"),
Merges: c.Strings("tokenizer.ggml.merges"),
BOS: int32(c.Uint("tokenizer.ggml.bos_token_id")),
AddBOS: c.Bool("tokenizer.ggml.add_bos_token", true),
EOS: int32(c.Uint("tokenizer.ggml.eos_token_id")),
AddEOS: c.Bool("tokenizer.ggml.add_eos_token", false),
},
),
ImageProcessor: newImageProcessor(c),

23
model/process_text.go
View File

@@ -30,7 +30,8 @@ type Vocabulary struct {
Scores []uint32
Merges []string
BOS, EOS int32
BOS, EOS int32
AddBOS, AddEOS bool
specialOnce sync.Once
special []string
@@ -281,6 +282,26 @@ func (bpe BytePairEncoding) Encode(s string) ([]int32, error) {
}
}
if len(ids) > 0 {
if bpe.vocab.AddBOS {
if ids[0] == bpe.vocab.BOS {
slog.Warn("adding bos token to prompt which already has it", "id", bpe.vocab.BOS)
}
slog.Debug("adding bos token to prompt", "id", bpe.vocab.BOS)
ids = append([]int32{bpe.vocab.BOS}, ids...)
}
if bpe.vocab.AddEOS {
if ids[len(ids)-1] == bpe.vocab.EOS {
slog.Warn("adding eos token to prompt which already has it", "id", bpe.vocab.EOS)
}
slog.Debug("adding eos token to prompt", "id", bpe.vocab.EOS)
ids = append(ids, bpe.vocab.EOS)
}
}
return ids, nil
}