You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In Issue #189 I have asked how to insert embeddings for a lot of plaintext articles (almost a million) using Ollama/llama 3.2 and nomic-embed-text. In the meantime, I create a small test project with 20 articles as follows
async function loadData() {
const result = await db.execute("SELECT fulltext FROM t LIMIT 20");
for (const row of result.rows) {
app.addLoader(new TextLoader({ text: row.fulltext }));
}
}
Then I query as follows
async function queryData(query) {
const ans = await app.query(query)
console.log(ans.content)
}
const query = 'Are there any species with maxillary and mandibular plates yellow?';
await queryData(query);
I get a reply that seems to be pulling in data from outside what I inserted. Is there any way to restrict the answers to just my corpus?
The text was updated successfully, but these errors were encountered:
In Issue #189 I have asked how to insert embeddings for a lot of plaintext articles (almost a million) using Ollama/llama 3.2 and
nomic-embed-text
. In the meantime, I create a small test project with 20 articles as followsThen I query as follows
I get a reply that seems to be pulling in data from outside what I inserted. Is there any way to restrict the answers to just my corpus?
The text was updated successfully, but these errors were encountered: