mirrors/ollama-for-amd
1
0
Fork 0
mirror of https://github.com/likelovewant/ollama-for-amd.git synced 2025-12-21 22:33:56 +00:00
Code Issues Packages Projects Releases Wiki Activity

feat: add trace log level (#10650)

Browse Source 
reduce prompt log to trace level
This commit is contained in:
Michael Yang
2025-05-12 11:43:00 -07:00
committed by GitHub
parent 82a9e9462a
commit f95a1f2bef
13 changed files with 141 additions and 114 deletions
Download Patch File Download Diff File Expand all files Collapse all files

59
llm/server.go
View File

@@ -17,6 +17,7 @@ import (
"os/exec"
"path/filepath"
"runtime"
"slices"
"strconv"
"strings"
"sync"
@@ -30,9 +31,37 @@ import (
"github.com/ollama/ollama/format"
"github.com/ollama/ollama/fs/ggml"
"github.com/ollama/ollama/llama"
"github.com/ollama/ollama/logutil"
"github.com/ollama/ollama/model"
)
type filteredEnv []string
func (e filteredEnv) LogValue() slog.Value {
var attrs []slog.Attr
for _, env := range e {
if key, value, ok := strings.Cut(env, "="); ok {
switch {
case strings.HasPrefix(key, "OLLAMA_"),
strings.HasPrefix(key, "CUDA_"),
strings.HasPrefix(key, "ROCR_"),
strings.HasPrefix(key, "ROCM_"),
strings.HasPrefix(key, "HIP_"),
strings.HasPrefix(key, "GPU_"),
strings.HasPrefix(key, "HSA_"),
strings.HasPrefix(key, "GGML_"),
slices.Contains([]string{
"PATH",
"LD_LIBRARY_PATH",
"DYLD_LIBRARY_PATH",
}, key):
attrs = append(attrs, slog.String(key, value))
}
}
}
return slog.GroupValue(attrs...)
}
type LlamaServer interface {
Ping(ctx context.Context) error
WaitUntilRunning(ctx context.Context) error
@@ -148,10 +177,6 @@ func NewLlamaServer(gpus discover.GpuInfoList, modelPath string, f *ggml.GGML, a
params = append(params, "--n-gpu-layers", strconv.Itoa(opts.NumGPU))
}
if envconfig.Debug() {
params = append(params, "--verbose")
}
if opts.MainGPU > 0 {
params = append(params, "--main-gpu", strconv.Itoa(opts.MainGPU))
}
@@ -404,26 +429,7 @@ func NewLlamaServer(gpus discover.GpuInfoList, modelPath string, f *ggml.GGML, a
}
slog.Info("starting llama server", "cmd", s.cmd)
if envconfig.Debug() {
filteredEnv := []string{}
for _, ev := range s.cmd.Env {
if strings.HasPrefix(ev, "OLLAMA_") ||
strings.HasPrefix(ev, "CUDA_") ||
strings.HasPrefix(ev, "ROCR_") ||
strings.HasPrefix(ev, "ROCM_") ||
strings.HasPrefix(ev, "HIP_") ||
strings.HasPrefix(ev, "GPU_") ||
strings.HasPrefix(ev, "HSA_") ||
strings.HasPrefix(ev, "GGML_") ||
strings.HasPrefix(ev, "PATH=") ||
strings.HasPrefix(ev, "LD_LIBRARY_PATH=") ||
strings.HasPrefix(ev, "DYLD_LIBRARY_PATH=") {
filteredEnv = append(filteredEnv, ev)
}
}
// Log at debug as the environment is inherited and might contain sensitive information
slog.Debug("subprocess", "environment", filteredEnv)
}
slog.Debug("subprocess", "", filteredEnv(s.cmd.Env))
if err = s.cmd.Start(); err != nil {
var msg string
@@ -721,6 +727,9 @@ type CompletionResponse struct {
}
func (s *llmServer) Completion(ctx context.Context, req CompletionRequest, fn func(CompletionResponse)) error {
slog.Debug("completion request", "images", len(req.Images), "prompt", len(req.Prompt), "format", string(req.Format))
slog.Log(ctx, logutil.LevelTrace, "completion request", "prompt", req.Prompt)
if len(req.Format) > 0 {
switch string(req.Format) {
case `null`, `""`:
@@ -884,6 +893,8 @@ type EmbeddingResponse struct {
}
func (s *llmServer) Embedding(ctx context.Context, input string) ([]float32, error) {
slog.Log(ctx, logutil.LevelTrace, "embedding request", "input", input)
if err := s.sem.Acquire(ctx, 1); err != nil {
if errors.Is(err, context.Canceled) {
slog.Info("aborting embedding request due to client closing the connection")