Skip to content

Commit

Permalink
Read jinja templates as fallback
Browse files Browse the repository at this point in the history 
Signed-off-by: Ettore Di Giacinto <[email protected]>
  • Loading branch information
@mudler
mudler committed Dec 6, 2024
1 parent d4c1746 commit 4ed689a
Show file tree
Hide file tree
Showing 10 changed files with 395 additions and 207 deletions.
2 changes: 2 additions & 0 deletions core/config/backend_config.go
View file
Original file line number Diff line number Diff line change
Expand Up @@ -206,6 +206,8 @@ type TemplateConfig struct {
JoinChatMessagesByCharacter *string `yaml:"join_chat_messages_by_character"`

Multimodal string `yaml:"multimodal"`

JinjaTemplate bool `yaml:"jinja_template"`
}

func (c *BackendConfig) UnmarshalYAML(value *yaml.Node) error {
Expand Down
16 changes: 14 additions & 2 deletions core/config/guesser.go
View file
Original file line number Diff line number Diff line change
Expand Up @@ -26,14 +26,14 @@ const (
type settingsConfig struct {
StopWords []string
TemplateConfig TemplateConfig
RepeatPenalty float64
RepeatPenalty float64
}

// default settings to adopt with a given model family
var defaultsSettings map[familyType]settingsConfig = map[familyType]settingsConfig{
Gemma: {
RepeatPenalty: 1.0,
StopWords: []string{"<|im_end|>", "<end_of_turn>", "<start_of_turn>"},
StopWords: []string{"<|im_end|>", "<end_of_turn>", "<start_of_turn>"},
TemplateConfig: TemplateConfig{
Chat: "{{.Input }}\n<start_of_turn>model\n",
ChatMessage: "<start_of_turn>{{if eq .RoleName \"assistant\" }}model{{else}}{{ .RoleName }}{{end}}\n{{ if .Content -}}\n{{.Content -}}\n{{ end -}}<end_of_turn>",
Expand Down Expand Up @@ -200,6 +200,18 @@ func guessDefaultsFromFile(cfg *BackendConfig, modelPath string) {
} else {
log.Debug().Any("family", family).Msgf("guessDefaultsFromFile: no template found for family")
}

if cfg.HasTemplate() {
return
}

// identify from well known templates first, otherwise use the raw jinja template
chatTemplate, found := f.Header.MetadataKV.Get("tokenizer.chat_template")
if found {
// try to use the jinja template
cfg.TemplateConfig.JinjaTemplate = true
cfg.TemplateConfig.ChatMessage = chatTemplate.ValueString()
}
}

func identifyFamily(f *gguf.GGUFFile) familyType {
Expand Down
140 changes: 1 addition & 139 deletions core/http/endpoints/openai/chat.go
View file
Original file line number Diff line number Diff line change
Expand Up @@ -298,145 +298,7 @@ func ChatEndpoint(cl *config.BackendConfigLoader, ml *model.ModelLoader, startup
// If we are using the tokenizer template, we don't need to process the messages
// unless we are processing functions
if !config.TemplateConfig.UseTokenizerTemplate || shouldUseFn {
suppressConfigSystemPrompt := false
mess := []string{}
for messageIndex, i := range input.Messages {
var content string
role := i.Role

// if function call, we might want to customize the role so we can display better that the "assistant called a json action"
// if an "assistant_function_call" role is defined, we use it, otherwise we use the role that is passed by in the request
if (i.FunctionCall != nil || i.ToolCalls != nil) && i.Role == "assistant" {
roleFn := "assistant_function_call"
r := config.Roles[roleFn]
if r != "" {
role = roleFn
}
}
r := config.Roles[role]
contentExists := i.Content != nil && i.StringContent != ""

fcall := i.FunctionCall
if len(i.ToolCalls) > 0 {
fcall = i.ToolCalls
}

// First attempt to populate content via a chat message specific template
if config.TemplateConfig.ChatMessage != "" {
chatMessageData := model.ChatMessageTemplateData{
SystemPrompt: config.SystemPrompt,
Role: r,
RoleName: role,
Content: i.StringContent,
FunctionCall: fcall,
FunctionName: i.Name,
LastMessage: messageIndex == (len(input.Messages) - 1),
Function: config.Grammar != "" && (messageIndex == (len(input.Messages) - 1)),
MessageIndex: messageIndex,
}
templatedChatMessage, err := ml.EvaluateTemplateForChatMessage(config.TemplateConfig.ChatMessage, chatMessageData)
if err != nil {
log.Error().Err(err).Interface("message", chatMessageData).Str("template", config.TemplateConfig.ChatMessage).Msg("error processing message with template, skipping")
} else {
if templatedChatMessage == "" {
log.Warn().Msgf("template \"%s\" produced blank output for %+v. Skipping!", config.TemplateConfig.ChatMessage, chatMessageData)
continue // TODO: This continue is here intentionally to skip over the line `mess = append(mess, content)` below, and to prevent the sprintf
}
log.Debug().Msgf("templated message for chat: %s", templatedChatMessage)
content = templatedChatMessage
}
}

marshalAnyRole := func(f any) {
j, err := json.Marshal(f)
if err == nil {
if contentExists {
content += "\n" + fmt.Sprint(r, " ", string(j))
} else {
content = fmt.Sprint(r, " ", string(j))
}
}
}
marshalAny := func(f any) {
j, err := json.Marshal(f)
if err == nil {
if contentExists {
content += "\n" + string(j)
} else {
content = string(j)
}
}
}
// If this model doesn't have such a template, or if that template fails to return a value, template at the message level.
if content == "" {
if r != "" {
if contentExists {
content = fmt.Sprint(r, i.StringContent)
}

if i.FunctionCall != nil {
marshalAnyRole(i.FunctionCall)
}
if i.ToolCalls != nil {
marshalAnyRole(i.ToolCalls)
}
} else {
if contentExists {
content = fmt.Sprint(i.StringContent)
}
if i.FunctionCall != nil {
marshalAny(i.FunctionCall)
}
if i.ToolCalls != nil {
marshalAny(i.ToolCalls)
}
}
// Special Handling: System. We care if it was printed at all, not the r branch, so check seperately
if contentExists && role == "system" {
suppressConfigSystemPrompt = true
}
}

mess = append(mess, content)
}

joinCharacter := "\n"
if config.TemplateConfig.JoinChatMessagesByCharacter != nil {
joinCharacter = *config.TemplateConfig.JoinChatMessagesByCharacter
}

predInput = strings.Join(mess, joinCharacter)
log.Debug().Msgf("Prompt (before templating): %s", predInput)

templateFile := ""

// A model can have a "file.bin.tmpl" file associated with a prompt template prefix
if ml.ExistsInModelPath(fmt.Sprintf("%s.tmpl", config.Model)) {
templateFile = config.Model
}

if config.TemplateConfig.Chat != "" && !shouldUseFn {
templateFile = config.TemplateConfig.Chat
}

if config.TemplateConfig.Functions != "" && shouldUseFn {
templateFile = config.TemplateConfig.Functions
}

if templateFile != "" {
templatedInput, err := ml.EvaluateTemplateForPrompt(model.ChatPromptTemplate, templateFile, model.PromptTemplateData{
SystemPrompt: config.SystemPrompt,
SuppressSystemPrompt: suppressConfigSystemPrompt,
Input: predInput,
Functions: funcs,
})
if err == nil {
predInput = templatedInput
log.Debug().Msgf("Template found, input modified to: %s", predInput)
} else {
log.Debug().Msgf("Template failed loading: %s", err.Error())
}
}
predInput = ml.TemplateMessages(input.Messages, config, funcs, shouldUseFn)

log.Debug().Msgf("Prompt (after templating): %s", predInput)
if shouldUseFn && config.Grammar != "" {
Expand Down
44 changes: 14 additions & 30 deletions core/http/endpoints/openai/completion.go
View file
Original file line number Diff line number Diff line change
Expand Up @@ -94,33 +94,20 @@ func CompletionEndpoint(cl *config.BackendConfigLoader, ml *model.ModelLoader, a
c.Set("Transfer-Encoding", "chunked")
}

templateFile := ""

// A model can have a "file.bin.tmpl" file associated with a prompt template prefix
if ml.ExistsInModelPath(fmt.Sprintf("%s.tmpl", config.Model)) {
templateFile = config.Model
}

if config.TemplateConfig.Completion != "" {
templateFile = config.TemplateConfig.Completion
}

if input.Stream {
if len(config.PromptStrings) > 1 {
return errors.New("cannot handle more than 1 `PromptStrings` when Streaming")
}

predInput := config.PromptStrings[0]

if templateFile != "" {
templatedInput, err := ml.EvaluateTemplateForPrompt(model.CompletionPromptTemplate, templateFile, model.PromptTemplateData{
Input: predInput,
SystemPrompt: config.SystemPrompt,
})
if err == nil {
predInput = templatedInput
log.Debug().Msgf("Template found, input modified to: %s", predInput)
}
templatedInput, err := ml.EvaluateTemplateForPrompt(model.CompletionPromptTemplate, *config, model.PromptTemplateData{
Input: predInput,
SystemPrompt: config.SystemPrompt,
})
if err == nil {
predInput = templatedInput
log.Debug().Msgf("Template found, input modified to: %s", predInput)
}

responses := make(chan schema.OpenAIResponse)
Expand Down Expand Up @@ -165,16 +152,13 @@ func CompletionEndpoint(cl *config.BackendConfigLoader, ml *model.ModelLoader, a
totalTokenUsage := backend.TokenUsage{}

for k, i := range config.PromptStrings {
if templateFile != "" {
// A model can have a "file.bin.tmpl" file associated with a prompt template prefix
templatedInput, err := ml.EvaluateTemplateForPrompt(model.CompletionPromptTemplate, templateFile, model.PromptTemplateData{
SystemPrompt: config.SystemPrompt,
Input: i,
})
if err == nil {
i = templatedInput
log.Debug().Msgf("Template found, input modified to: %s", i)
}
templatedInput, err := ml.EvaluateTemplateForPrompt(model.CompletionPromptTemplate, *config, model.PromptTemplateData{
SystemPrompt: config.SystemPrompt,
Input: i,
})
if err == nil {
i = templatedInput
log.Debug().Msgf("Template found, input modified to: %s", i)
}

r, tokenUsage, err := ComputeChoices(
Expand Down
29 changes: 8 additions & 21 deletions core/http/endpoints/openai/edit.go
View file
Original file line number Diff line number Diff line change
Expand Up @@ -35,31 +35,18 @@ func EditEndpoint(cl *config.BackendConfigLoader, ml *model.ModelLoader, appConf

log.Debug().Msgf("Parameter Config: %+v", config)

templateFile := ""

// A model can have a "file.bin.tmpl" file associated with a prompt template prefix
if ml.ExistsInModelPath(fmt.Sprintf("%s.tmpl", config.Model)) {
templateFile = config.Model
}

if config.TemplateConfig.Edit != "" {
templateFile = config.TemplateConfig.Edit
}

var result []schema.Choice
totalTokenUsage := backend.TokenUsage{}

for _, i := range config.InputStrings {
if templateFile != "" {
templatedInput, err := ml.EvaluateTemplateForPrompt(model.EditPromptTemplate, templateFile, model.PromptTemplateData{
Input: i,
Instruction: input.Instruction,
SystemPrompt: config.SystemPrompt,
})
if err == nil {
i = templatedInput
log.Debug().Msgf("Template found, input modified to: %s", i)
}
templatedInput, err := ml.EvaluateTemplateForPrompt(model.EditPromptTemplate, *config, model.PromptTemplateData{
Input: i,
Instruction: input.Instruction,
SystemPrompt: config.SystemPrompt,
})
if err == nil {
i = templatedInput
log.Debug().Msgf("Template found, input modified to: %s", i)
}

r, tokenUsage, err := ComputeChoices(input, i, config, appConfig, ml, func(s string, c *[]schema.Choice) {
Expand Down
5 changes: 5 additions & 0 deletions go.mod
View file
Original file line number Diff line number Diff line change
Expand Up @@ -76,6 +76,7 @@ require (
cloud.google.com/go/auth v0.4.1 // indirect
cloud.google.com/go/auth/oauth2adapt v0.2.2 // indirect
cloud.google.com/go/compute/metadata v0.3.0 // indirect
github.com/dustin/go-humanize v1.0.1 // indirect
github.com/envoyproxy/protoc-gen-validate v1.0.4 // indirect
github.com/fasthttp/websocket v1.5.3 // indirect
github.com/felixge/httpsnoop v1.0.4 // indirect
Expand All @@ -84,8 +85,12 @@ require (
github.com/google/s2a-go v0.1.7 // indirect
github.com/googleapis/enterprise-certificate-proxy v0.3.2 // indirect
github.com/googleapis/gax-go/v2 v2.12.4 // indirect
github.com/json-iterator/go v1.1.12 // indirect
github.com/moby/docker-image-spec v1.3.1 // indirect
github.com/modern-go/concurrent v0.0.0-20180306012644-bacd9c7ef1dd // indirect
github.com/modern-go/reflect2 v1.0.2 // indirect
github.com/munnerz/goautoneg v0.0.0-20191010083416-a7dc8b61c822 // indirect
github.com/nikolalohinski/gonja/v2 v2.3.2 // indirect
github.com/pion/datachannel v1.5.8 // indirect
github.com/pion/dtls/v2 v2.2.12 // indirect
github.com/pion/ice/v2 v2.3.34 // indirect
Expand Down
12 changes: 12 additions & 0 deletions go.sum
View file
Original file line number Diff line number Diff line change
Expand Up @@ -140,6 +140,8 @@ github.com/dsnet/compress v0.0.2-0.20210315054119-f66993602bf5 h1:iFaUwBSo5Svw6L
github.com/dsnet/compress v0.0.2-0.20210315054119-f66993602bf5/go.mod h1:qssHWj60/X5sZFNxpG4HBPDHVqxNm4DfnCKgrbZOT+s=
github.com/dsnet/golib v0.0.0-20171103203638-1ea166775780/go.mod h1:Lj+Z9rebOhdfkVLjJ8T6VcRQv3SXugXy999NBtR9aFY=
github.com/dustin/go-humanize v1.0.0/go.mod h1:HtrtbFcZ19U5GC7JDqmcUSB87Iq5E25KnS6fMYU6eOk=
github.com/dustin/go-humanize v1.0.1 h1:GzkhY7T5VNhEkwH0PVJgjz+fX1rhBrR7pRT3mDkpeCY=
github.com/dustin/go-humanize v1.0.1/go.mod h1:Mu1zIs6XwVuF/gI1OepvI0qD18qycQx+mFykh5fBlto=
github.com/elastic/gosigar v0.12.0/go.mod h1:iXRIGg2tLnu7LBdpqzyQfGDEidKCfWcCMS0WKyPWoMs=
github.com/elastic/gosigar v0.14.3 h1:xwkKwPia+hSfg9GqrCUKYdId102m9qTJIIr7egmK/uo=
github.com/elastic/gosigar v0.14.3/go.mod h1:iXRIGg2tLnu7LBdpqzyQfGDEidKCfWcCMS0WKyPWoMs=
Expand Down Expand Up @@ -268,6 +270,7 @@ github.com/google/go-containerregistry v0.19.2 h1:TannFKE1QSajsP6hPWb5oJNgKe1IKj
github.com/google/go-containerregistry v0.19.2/go.mod h1:YCMFNQeeXeLF+dnhhWkqDItx/JSkH01j1Kis4PsjzFI=
github.com/google/go-github v17.0.0+incompatible/go.mod h1:zLgOLi98H3fifZn+44m+umXrS52loVEgC2AApnigrVQ=
github.com/google/go-querystring v1.0.0/go.mod h1:odCYkC5MyYFN7vkCjXpyrEuKhc/BUO6wN/zVPAxq5ck=
github.com/google/gofuzz v1.0.0/go.mod h1:dBl0BpW6vV/+mYPU4Po3pmUjxk6FQPldtuIdl/M65Eg=
github.com/google/gopacket v1.1.19 h1:ves8RnFZPGiFnTS0uPQStjwru6uO6h+nlr9j6fL7kF8=
github.com/google/gopacket v1.1.19/go.mod h1:iJ8V8n6KS+z2U1A8pUwu8bW5SyEMkXJB8Yo/Vo+TKTo=
github.com/google/martian v2.1.0+incompatible/go.mod h1:9I4somxYTbIHy5NJKHRl3wXiIaQGbYVAs8BPL6v8lEs=
Expand Down Expand Up @@ -353,6 +356,8 @@ github.com/joho/godotenv v1.5.1/go.mod h1:f4LDr5Voq0i2e/R5DDNOoa2zzDfwtkZa6DnEwA
github.com/josharian/intern v1.0.0 h1:vlS4z54oSdjm0bgjRigI+G1HpF+tI+9rE5LLzOg8HmY=
github.com/josharian/intern v1.0.0/go.mod h1:5DoeVV0s6jJacbCEi61lwdGj/aVlrQvzHFFd8Hwg//Y=
github.com/json-iterator/go v1.1.6/go.mod h1:+SdeFBvtyEkXs7REEP0seUULqWtbJapLOCVDaaPEHmU=
github.com/json-iterator/go v1.1.12 h1:PV8peI4a0ysnczrg+LtxykD8LfKY9ML6u2jnxaEnrnM=
github.com/json-iterator/go v1.1.12/go.mod h1:e30LSqwooZae/UwlEbR2852Gd8hjQvJoHmT4TnhNGBo=
github.com/jstemmer/go-junit-report v0.0.0-20190106144839-af01ea7f8024/go.mod h1:6v2b51hI/fHJwM22ozAgKL4VKDeJcHhJFhtBdhmNjmU=
github.com/jtolds/gls v4.20.0+incompatible h1:xdiiI2gbIgH/gLH7ADydsJ1uDOEzR8yvV7C0MuV77Wo=
github.com/jtolds/gls v4.20.0+incompatible/go.mod h1:QJZ7F/aHp+rZTRtaJ1ow/lLfFfVYBRgL+9YlvaHOwJU=
Expand Down Expand Up @@ -474,8 +479,12 @@ github.com/moby/sys/sequential v0.5.0 h1:OPvI35Lzn9K04PBbCLW0g4LcFAJgHsvXsRyewg5
github.com/moby/sys/sequential v0.5.0/go.mod h1:tH2cOOs5V9MlPiXcQzRC+eEyab644PWKGRYaaV5ZZlo=
github.com/moby/term v0.5.0 h1:xt8Q1nalod/v7BqbG21f8mQPqH+xAaC9C3N3wfWbVP0=
github.com/moby/term v0.5.0/go.mod h1:8FzsFHVUBGZdbDsJw/ot+X+d5HLUbvklYLJ9uGfcI3Y=
github.com/modern-go/concurrent v0.0.0-20180228061459-e0a39a4cb421/go.mod h1:6dJC0mAP4ikYIbvyc7fijjWJddQyLn8Ig3JB5CqoB9Q=
github.com/modern-go/concurrent v0.0.0-20180306012644-bacd9c7ef1dd h1:TRLaZ9cD/w8PVh93nsPXa1VrQ6jlwL5oN8l14QlcNfg=
github.com/modern-go/concurrent v0.0.0-20180306012644-bacd9c7ef1dd/go.mod h1:6dJC0mAP4ikYIbvyc7fijjWJddQyLn8Ig3JB5CqoB9Q=
github.com/modern-go/reflect2 v1.0.1/go.mod h1:bx2lNnkwVCuqBIxFjflWJWanXIb3RllmbCylyMrvgv0=
github.com/modern-go/reflect2 v1.0.2 h1:xBagoLtFs94CBntxluKeaWgTMpvLxC4ur3nMaC9Gz0M=
github.com/modern-go/reflect2 v1.0.2/go.mod h1:yWuevngMOJpCy52FWWMvUC8ws7m/LJsjYzDa0/r8luk=
github.com/mr-tron/base58 v1.1.2/go.mod h1:BinMc/sQntlIE1frQmRFPUoPA1Zkr8VRgBdjWI2mNwc=
github.com/mr-tron/base58 v1.2.0 h1:T/HDJBh4ZCPbU39/+c3rRvE0uKBQlU27+QI8LJ4t64o=
github.com/mr-tron/base58 v1.2.0/go.mod h1:BinMc/sQntlIE1frQmRFPUoPA1Zkr8VRgBdjWI2mNwc=
Expand Down Expand Up @@ -519,6 +528,9 @@ github.com/munnerz/goautoneg v0.0.0-20191010083416-a7dc8b61c822 h1:C3w9PqII01/Oq
github.com/munnerz/goautoneg v0.0.0-20191010083416-a7dc8b61c822/go.mod h1:+n7T8mK8HuQTcFwEeznm/DIxMOiR9yIdICNftLE1DvQ=
github.com/neelance/astrewrite v0.0.0-20160511093645-99348263ae86/go.mod h1:kHJEU3ofeGjhHklVoIGuVj85JJwZ6kWPaJwCIxgnFmo=
github.com/neelance/sourcemap v0.0.0-20151028013722-8c68805598ab/go.mod h1:Qr6/a/Q4r9LP1IltGz7tA7iOK1WonHEYhu1HRBA7ZiM=
github.com/nikolalohinski/gonja v1.5.3 h1:GsA+EEaZDZPGJ8JtpeGN78jidhOlxeJROpqMT9fTj9c=
github.com/nikolalohinski/gonja/v2 v2.3.2 h1:UgLFfqi7L9XfX0PEcE4eUpvGojVQL5KhBfJJaBp7ZxY=
github.com/nikolalohinski/gonja/v2 v2.3.2/go.mod h1:1Wcc/5huTu6y36e0sOFR1XQoFlylw3c3H3L5WOz0RDg=
github.com/nwaples/rardecode v1.1.0 h1:vSxaY8vQhOcVr4mm5e8XllHWTiM4JF507A0Katqw7MQ=
github.com/nwaples/rardecode v1.1.0/go.mod h1:5DzqNKiOdpKKBH87u8VlvAnPZMXcGRhxWkRpHbbfGS0=
github.com/nxadm/tail v1.4.11 h1:8feyoE3OzPrcshW5/MJ4sGESc5cqmGkGCWlco4l0bqY=
Expand Down
Loading

0 comments on commit 4ed689a

Please sign in to comment.