mirror of
https://github.com/Mintplex-Labs/anything-llm.git
synced 2025-04-23 13:08:11 +00:00
persist token window for NIM and not only on model change
This commit is contained in:
parent
b2dd35fe15
commit
4b09a06590
2 changed files with 5 additions and 2 deletions
server/utils
|
@ -45,7 +45,7 @@ class NvidiaNimLLM {
|
||||||
}
|
}
|
||||||
|
|
||||||
/**
|
/**
|
||||||
* Set the model token limit `NVIDIA_NIM_LLM_TOKEN_LIMIT` for the given model ID
|
* Set the model token limit `NVIDIA_NIM_LLM_MODEL_TOKEN_LIMIT` for the given model ID
|
||||||
* @param {string} modelId
|
* @param {string} modelId
|
||||||
* @param {string} basePath
|
* @param {string} basePath
|
||||||
* @returns {Promise<void>}
|
* @returns {Promise<void>}
|
||||||
|
@ -69,7 +69,7 @@ class NvidiaNimLLM {
|
||||||
if (!model.length) return;
|
if (!model.length) return;
|
||||||
const modelInfo = model.find((model) => model.id === modelId);
|
const modelInfo = model.find((model) => model.id === modelId);
|
||||||
if (!modelInfo) return;
|
if (!modelInfo) return;
|
||||||
process.env.NVIDIA_NIM_LLM_TOKEN_LIMIT = Number(
|
process.env.NVIDIA_NIM_LLM_MODEL_TOKEN_LIMIT = Number(
|
||||||
modelInfo.max_model_len || 4096
|
modelInfo.max_model_len || 4096
|
||||||
);
|
);
|
||||||
}
|
}
|
||||||
|
|
|
@ -965,6 +965,9 @@ function dumpENV() {
|
||||||
"SIMPLE_SSO_ENABLED",
|
"SIMPLE_SSO_ENABLED",
|
||||||
// Community Hub
|
// Community Hub
|
||||||
"COMMUNITY_HUB_BUNDLE_DOWNLOADS_ENABLED",
|
"COMMUNITY_HUB_BUNDLE_DOWNLOADS_ENABLED",
|
||||||
|
|
||||||
|
// Nvidia NIM Keys that are automatically managed
|
||||||
|
"NVIDIA_NIM_LLM_MODEL_TOKEN_LIMIT",
|
||||||
];
|
];
|
||||||
|
|
||||||
// Simple sanitization of each value to prevent ENV injection via newline or quote escaping.
|
// Simple sanitization of each value to prevent ENV injection via newline or quote escaping.
|
||||||
|
|
Loading…
Add table
Reference in a new issue