scriptoriumadmin/open-webui
1
0
Fork 0
mirror of https://github.com/open-webui/open-webui.git synced 2026-11-10 08:47:14 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 3

fix: read context usage tokens with the same key precedence as backend compaction

Browse source
This commit is contained in:
silentoplayz 2026-07-27 18:13:52 -04:00
parent 483adf7040
commit 91162a6af2
2 changed files with 36 additions and 12 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

24
src/lib/components/chat/Chat.svelte
View file

@ -281,13 +281,25 @@
for (let idx = activeMessages.length - 1; idx >= 0; idx -= 1) {
const usage = activeMessages[idx]?.usage ?? activeMessages[idx]?.info?.usage;
const inputTokens = usage?.input_tokens ?? usage?.prompt_tokens;
if (inputTokens) {
const usageTokens =
Number(
usage?.prompt_tokens ||
usage?.input_tokens ||
usage?.prompt_eval_count ||
usage?.prompt_n ||
0
) +
Number(
usage?.completion_tokens ||
usage?.output_tokens ||
usage?.eval_count ||
usage?.predicted_n ||
0
) +
Number(usage?.cache_n || 0);
if (usageTokens) {
hasUsageCheckpoint = true;
estimatedTokens =
Number(inputTokens || 0) +
Number(usage.output_tokens ?? usage.completion_tokens ?? 0) +
estimateMessagesTokens(activeMessages.slice(idx + 1));
estimatedTokens = usageTokens + estimateMessagesTokens(activeMessages.slice(idx + 1));
break;
}
}

24
src/lib/components/chat/MessageInput.svelte
View file

@ -444,13 +444,25 @@
for (let idx = activeMessages.length - 1; idx >= 0; idx -= 1) {
const usage = activeMessages[idx]?.usage ?? activeMessages[idx]?.info?.usage;
const inputTokens = usage?.input_tokens ?? usage?.prompt_tokens;
if (inputTokens) {
const usageTokens =
Number(
usage?.prompt_tokens ||
usage?.input_tokens ||
usage?.prompt_eval_count ||
usage?.prompt_n ||
0
) +
Number(
usage?.completion_tokens ||
usage?.output_tokens ||
usage?.eval_count ||
usage?.predicted_n ||
0
) +
Number(usage?.cache_n || 0);
if (usageTokens) {
hasUsageCheckpoint = true;
estimatedTokens =
Number(inputTokens || 0) +
Number(usage.output_tokens ?? usage.completion_tokens ?? 0) +
estimateMessagesTokens(activeMessages.slice(idx + 1));
estimatedTokens = usageTokens + estimateMessagesTokens(activeMessages.slice(idx + 1));
break;
}
}