Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Revert "fix: use /infill for llama.cpp code-completions (#513)" #533

Merged
merged 2 commits into from
May 8, 2024
Merged

Revert "fix: use /infill for llama.cpp code-completions (#513)" #533

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
57d225a
Revert "fix: use /infill for llama.cpp code-completions (#513)"
PhilKes May 7, 2024
34c0c51
Merge remote-tracking branch 'refs/remotes/origin/master' into revert…
PhilKes May 8, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion src/main/java/ee/carlrobert/codegpt/completions/CompletionRequestService.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -126,7 +126,7 @@ public EventSource getCodeCompletionAsync(
CodeCompletionRequestFactory.buildCustomRequest(requestDetails),
new OpenAITextCompletionEventSourceListener(eventListener));
case LLAMA_CPP -> CompletionClientProvider.getLlamaClient()
.getInfillAsync(
.getChatCompletionAsync(
CodeCompletionRequestFactory.buildLlamaRequest(requestDetails),
eventListener);
case OLLAMA -> CompletionClientProvider.getOllamaClient().getCompletionAsync(
Expand Down
17 changes: 8 additions & 9 deletions src/main/kotlin/ee/carlrobert/codegpt/codecompletions/CodeCompletionRequestFactory.kt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,6 @@ import ee.carlrobert.codegpt.settings.service.llama.LlamaSettingsState
import ee.carlrobert.codegpt.settings.service.ollama.OllamaSettings
import ee.carlrobert.codegpt.settings.service.openai.OpenAISettings
import ee.carlrobert.llm.client.llama.completion.LlamaCompletionRequest
import ee.carlrobert.llm.client.llama.completion.LlamaInfillRequest
import ee.carlrobert.llm.client.ollama.completion.request.OllamaCompletionRequest
import ee.carlrobert.llm.client.ollama.completion.request.OllamaParameters
import ee.carlrobert.llm.client.openai.completion.request.OpenAITextCompletionRequest
Expand Down Expand Up @@ -82,16 +81,16 @@ object CodeCompletionRequestFactory {
}

@JvmStatic
fun buildLlamaRequest(details: InfillRequestDetails): LlamaInfillRequest {
fun buildLlamaRequest(details: InfillRequestDetails): LlamaCompletionRequest {
val settings = LlamaSettings.getCurrentState()
val promptTemplate = getLlamaInfillPromptTemplate(settings)
return LlamaInfillRequest(
LlamaCompletionRequest.Builder(null)
.setN_predict(settings.codeCompletionMaxTokens)
.setStream(true)
.setTemperature(0.4)
.setStop(promptTemplate.stopTokens), details.prefix, details.suffix
)
val prompt = promptTemplate.buildPrompt(details.prefix, details.suffix)
return LlamaCompletionRequest.Builder(prompt)
.setN_predict(settings.codeCompletionMaxTokens)
.setStream(true)
.setTemperature(0.4)
.setStop(promptTemplate.stopTokens)
.build()
}

fun buildOllamaRequest(details: InfillRequestDetails): OllamaCompletionRequest {
Expand Down
6 changes: 3 additions & 3 deletions src/test/kotlin/ee/carlrobert/codegpt/codecompletions/CodeCompletionServiceTest.kt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -35,11 +35,11 @@ class CodeCompletionServiceTest : IntegrationTest() {
${"z".repeat(247)}
""".trimIndent() // 128 tokens
expectLlama(StreamHttpExchange { request: RequestEntity ->
assertThat(request.uri.path).isEqualTo("/infill")
assertThat(request.uri.path).isEqualTo("/completion")
assertThat(request.method).isEqualTo("POST")
assertThat(request.body)
.extracting("input_prefix", "input_suffix")
.containsExactly(prefix, suffix)
.extracting("prompt")
.isEqualTo(InfillPromptTemplate.CODE_LLAMA.buildPrompt(prefix, suffix))
listOf(jsonMapResponse(e("content", expectedCompletion), e("stop", true)))
})

Expand Down