exo/lm-lib
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix: logging for prompt length warnings in ChatCompletionClient

Browse Source
This commit is contained in:
extio-de
2026-07-12 17:08:56 +02:00
parent 7adc5d8d30
commit 4a85d9e65f
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
src/main/java/de/extio/lmlib/client/oai/completion/chat/ChatCompletionClient.java
View File

@@ -338,7 +338,7 @@ public class ChatCompletionClient extends AbstractCompletionClient {
final int maxLength = modelProfile.maxContextLength() - modelProfile.maxTokens() - 15; // 15 is a margin for EOT and other special tokens final int maxLength = modelProfile.maxContextLength() - modelProfile.maxTokens() - 15; // 15 is a margin for EOT and other special tokens
if (tokens.addAndGet(textTokens) > maxLength) { if (tokens.addAndGet(textTokens) > maxLength) {
var tokenized = this.tokenizerResolver.tokenize(prompt, modelProfile); var tokenized = this.tokenizerResolver.tokenize(prompt, modelProfile);
LOGGER.warn("Prompt too long: {} tokens / {} max. Prompt will be cut!", tokenized.size(), maxLength); LOGGER.warn("Prompt too long: {} tokens / {} max. Prompt will be cut!", tokens.get(), maxLength);
tokenized = tokenized.subList(0, tokenized.size() - (tokens.get() - maxLength)); tokenized = tokenized.subList(0, tokenized.size() - (tokens.get() - maxLength));
prompt = this.tokenizerResolver.detokenize(tokenized, modelProfile); prompt = this.tokenizerResolver.detokenize(tokenized, modelProfile);
trimmed = true; trimmed = true;