tcsenpai/ollama
1
0
Fork 0
mirror of https://github.com/tcsenpai/ollama.git synced 2025-06-07 11:45:21 +00:00
Code Issues Packages Projects Releases Wiki Activity

remove dependency on llm

Browse Source
This commit is contained in:
jmorganca 2024-05-26 23:23:09 -07:00
parent c0b94376b2
commit ce15ed6d69
3 changed files with 3 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
llama/llama.go
View File

@ -35,8 +35,6 @@ import (
"runtime" "runtime"
"strings" "strings"
"unsafe" "unsafe"
"github.com/ollama/ollama/llm"
) )
func BackendInit() { func BackendInit() {
@ -227,7 +225,7 @@ func (m *Model) Tokenize(text string, maxTokens int, addSpecial bool, parseSpeci
return tokens, nil return tokens, nil
} }
func Quantize(infile, outfile string, ftype llm.FileType) error { func Quantize(infile, outfile string, ftype uint32) error {
cinfile := C.CString(infile) cinfile := C.CString(infile)
defer C.free(unsafe.Pointer(cinfile)) defer C.free(unsafe.Pointer(cinfile))
@ -236,7 +234,7 @@ func Quantize(infile, outfile string, ftype llm.FileType) error {
params := C.llama_model_quantize_default_params() params := C.llama_model_quantize_default_params()
params.nthread = -1 params.nthread = -1
params.ftype = ftype.Value() params.ftype = ftype
if rc := C.llama_model_quantize(cinfile, coutfile, &params); rc != 0 { if rc := C.llama_model_quantize(cinfile, coutfile, &params); rc != 0 {
return fmt.Errorf("llama_model_quantize: %d", rc) return fmt.Errorf("llama_model_quantize: %d", rc)

1
llama/runner/runner.go
View File

@ -209,7 +209,6 @@ func (s *Server) handler(w http.ResponseWriter, r *http.Request) {
for i, sq := range s.seqs { for i, sq := range s.seqs {
if sq == nil { if sq == nil {
s.seqs[i] = seq s.seqs[i] = seq
fmt.Println("signal")
s.cond.Signal() s.cond.Signal()
break break
} }

2
llama/sampling_ext.cpp vendored
View File

@ -17,7 +17,7 @@ struct llama_sampling_context* llama_sampling_cinit(struct llama_sampling_cparam
sparams.mirostat_eta = params->mirostat_eta; sparams.mirostat_eta = params->mirostat_eta;
sparams.penalize_nl = params->penalize_nl; sparams.penalize_nl = params->penalize_nl;
sparams.seed = params->seed; sparams.seed = params->seed;
sparams.grammar = std::string(params->grammar); sparams.grammar = params->grammar;
return llama_sampling_init(sparams); return llama_sampling_init(sparams);
} }