Archives/gpt4all
2
0
Fork 0
mirror of https://github.com/nomic-ai/gpt4all synced 2024-11-02 09:40:42 +00:00
Code Issues Packages Projects Releases Wiki Activity

Erase the correct amount of logits when regenerating which is not the same

Browse Source 
as the number of tokens.
This commit is contained in:
Adam Treat 2023-04-15 09:19:06 -04:00
parent 12bf78bf24
commit 2f3a46c17f
2 changed files with 6 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
llm.cpp
View File

@ -20,6 +20,7 @@ LLMObject::LLMObject()
: QObject{nullptr}
, m_llmodel(new GPTJ)
, m_responseTokens(0)
, m_responseLogits(0)
{
moveToThread(&m_llmThread);
connect(&m_llmThread, &QThread::started, this, &LLMObject::loadModel);
@ -66,8 +67,9 @@ bool LLMObject::isModelLoaded() const
void LLMObject::resetResponse()
{
s_ctx.n_past -= m_responseTokens;
s_ctx.logits.erase(s_ctx.logits.end() -= m_responseTokens, s_ctx.logits.end());
s_ctx.logits.erase(s_ctx.logits.end() -= m_responseLogits, s_ctx.logits.end());
m_responseTokens = 0;
m_responseLogits = 0;
m_response = std::string();
emit responseChanged();
}
@ -110,7 +112,9 @@ bool LLMObject::prompt(const QString &prompt)
m_stopGenerating = false;
auto func = std::bind(&LLMObject::handleResponse, this, std::placeholders::_1);
emit responseStarted();
qint32 logitsBefore = s_ctx.logits.size();
m_llmodel->prompt(prompt.toStdString(), func, s_ctx, 4096 /*number of chars to predict*/);
m_responseLogits += s_ctx.logits.size() - logitsBefore;
emit responseStopped();
return true;
}

1
llm.h
View File

@ -42,6 +42,7 @@ private:
LLModel *m_llmodel;
std::string m_response;
quint32 m_responseTokens;
quint32 m_responseLogits;
QString m_modelName;
QThread m_llmThread;
std::atomic<bool> m_stopGenerating;