mirrors/ollama
1
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2025-05-11 10:26:53 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

llm: set done reason at server level (#9830)

Browse source 
No functional change. Many different done reasons can be set at the runner
level, so rather than obsuring them we should return them to the server
process and let it choose what to do with the done reason. This separates
the API concerns from the runner.
This commit is contained in:
Bruce MacDonald 2025-04-03 10:19:24 -07:00 • committed by GitHub
parent b51e0f397c
commit e53b3cbd0c
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
5 changed files with 54 additions and 42 deletions
Download patch file Download diff file Expand all files Collapse all files

20
server/routes.go
View file

@ -308,11 +308,10 @@ func (s *Server) GenerateHandler(c *gin.Context) {
Options: opts,
}, func(cr llm.CompletionResponse) {
res := api.GenerateResponse{
Model: req.Model,
CreatedAt: time.Now().UTC(),
Response: cr.Content,
Done: cr.Done,
DoneReason: cr.DoneReason,
Model: req.Model,
CreatedAt: time.Now().UTC(),
Response: cr.Content,
Done: cr.Done,
Metrics: api.Metrics{
PromptEvalCount: cr.PromptEvalCount,
PromptEvalDuration: cr.PromptEvalDuration,
@ -326,6 +325,7 @@ func (s *Server) GenerateHandler(c *gin.Context) {
}
if cr.Done {
res.DoneReason = cr.DoneReason.String()
res.TotalDuration = time.Since(checkpointStart)
res.LoadDuration = checkpointLoaded.Sub(checkpointStart)
@ -1533,11 +1533,10 @@ func (s *Server) ChatHandler(c *gin.Context) {
Options: opts,
}, func(r llm.CompletionResponse) {
res := api.ChatResponse{
Model: req.Model,
CreatedAt: time.Now().UTC(),
Message: api.Message{Role: "assistant", Content: r.Content},
Done: r.Done,
DoneReason: r.DoneReason,
Model: req.Model,
CreatedAt: time.Now().UTC(),
Message: api.Message{Role: "assistant", Content: r.Content},
Done: r.Done,
Metrics: api.Metrics{
PromptEvalCount: r.PromptEvalCount,
PromptEvalDuration: r.PromptEvalDuration,
@ -1547,6 +1546,7 @@ func (s *Server) ChatHandler(c *gin.Context) {
}
if r.Done {
res.DoneReason = r.DoneReason.String()
res.TotalDuration = time.Since(checkpointStart)
res.LoadDuration = checkpointLoaded.Sub(checkpointStart)
}