LLM/moltbot
1
0
Fork 0
mirror of https://github.com/moltbot/moltbot.git synced 2026-03-08 06:54:24 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix: correctly identify Groq TPM limits as rate limits instead of context overflow (#16176)

Browse Source 
Co-authored-by: Howard <dddabtc@users.noreply.github.com>
This commit is contained in:
Alice Losasso
2026-02-23 11:32:53 -04:00
committed by GitHub
parent c1b75ab8e2
commit 652099cd5c
1 changed files with 15 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
src/agents/pi-embedded-helpers/errors.ts
View File

@@ -39,6 +39,12 @@ export function isContextOverflowError(errorMessage?: string): boolean {
return false;
}
const lower = errorMessage.toLowerCase();
// Groq uses 413 for TPM (tokens per minute) limits, which is a rate limit, not context overflow.
if (lower.includes("tpm") || lower.includes("tokens per minute")) {
return false;
}
const hasRequestSizeExceeds = lower.includes("request size exceeds");
const hasContextWindow =
lower.includes("context window") ||
@@ -72,6 +78,13 @@ export function isLikelyContextOverflowError(errorMessage?: string): boolean {
if (!errorMessage) {
return false;
}
// Groq uses 413 for TPM (tokens per minute) limits, which is a rate limit, not context overflow.
const lower = errorMessage.toLowerCase();
if (lower.includes("tpm") || lower.includes("tokens per minute")) {
return false;
}
if (CONTEXT_WINDOW_TOO_SMALL_RE.test(errorMessage)) {
return false;
}
@@ -571,6 +584,8 @@ const ERROR_PATTERNS = {
"quota exceeded",
"resource_exhausted",
"usage limit",
"tpm",
"tokens per minute",
],
overloaded: [
/overloaded_error|"type"\s*:\s*"overloaded_error"/i,