Search

Code
94

janpaul123/valtownsemanticsearch/README.md

4 matches


It does this by comparing [embeddings from OpenAI](https://platform.openai.com/docs/guides/embeddings) generated for the code of all public vals, to an embedding of your search query.
I implemented three backends, which you can switch between in the search UI. Check out these vals for details on their implementation.
- **Neon:** storing and searching embeddings using the [pg_vector](https://neon.tech/docs/extensions/pgvector) extension in Neon's Postgres database.
  - Searching: [janpaul123/semanticSearchNeon](https://www.val.town/v/janpaul123/semanticSearchNeon)
  - Indexing: [janpaul123/indexValsNeon](https://www.val.town/v/janpaul123/indexValsNeon)
- **Blobs:** storing embeddings in Val Town's [standard blob storage](https://docs.val.town/std/blob/), and iterating through all of them to compute distance. Slow and terrible, but it works!
  - Searching: [janpaul123/semanticSearchBlobs](https://www.val.town/v/janpaul123/semanticSearchBlobs)

jacoblee93/conversationalRetrievalQAChainSummaryMemory/main.tsx

3 matches

  );
  const { OpenAIEmbeddings } = await import(
    "https://esm.sh/langchain/embeddings/openai"
  );
    [{ id: 2 }, { id: 1 }, { id: 3 }],
    new OpenAIEmbeddings({
      openAIApiKey: process.env.OPENAI_API_KEY,

maxm/actuallyGoodEmojiSearch/main.tsx

2 matches

import { searchEmojis } from "https://esm.town/v/maxm/emojiVectorEmbeddings";
import { extractValInfo } from "https://esm.town/v/pomdtr/extractValInfo";
      <br />
      Built on Val Town with sqlite vector search and openai embeddings.
      <br />

maxm/actuallyGoodEmojiSearch/README.md

1 matches


Uses vector embeddings to get "vibes" search on emojis

yawnxyz/ai/README.md

2 matches

```
async function calculateEmbeddings(text) {
    const url = `https://yawnxyz-ai.web.val.run/generate?embed=true&value=${encodeURIComponent(text)}`;
    } catch (error) {
        console.error('Error calculating embeddings:', error);
        return null;

jacoblee93/conversationalQAChainEx/main.tsx

3 matches

  );
  const { OpenAIEmbeddings } = await import(
    "https://esm.sh/langchain/embeddings/openai"
  );
    [{ id: 2 }, { id: 1 }, { id: 3 }, { id: 4 }, { id: 5 }],
    new OpenAIEmbeddings({
      openAIApiKey: process.env.OPENAI_API_KEY,

wilt/getOpenapiEmbedding/main.tsx

2 matches

/**
 * Call OpenAPI Embeddings api to vectorize a query string
 * Returns an array of 1536 numbers
}): Promise<number[]> =>
  fetchJSON("https://api.openai.com/v1/embeddings", {
    method: "POST",

jacoblee93/conversationalRetrievalQAChainStreamingExample/main.tsx

3 matches

  );
  const { OpenAIEmbeddings } = await import(
    "https://esm.sh/langchain/embeddings/openai"
  );
    [{ id: 2 }, { id: 1 }, { id: 3 }],
    new OpenAIEmbeddings({
      openAIApiKey: process.env.OPENAI_API_KEY,

AIWB/memorySampleSummary/main.tsx

1 matches

async function generateEmbedding(text: string): Promise<number[]> {
  const response = await openai.embeddings.create({
    model: "text-embedding-ada-002",

tmcw/surprisingEmbeddings/amicableCoralPorcupine

0 matches

tmcw/surprisingEmbeddings/index

11 matches

/** @jsxImportSource npm:hono@3/jsx */
import bots from "https://esm.town/v/tmcw/surprisingEmbeddings/bots"
import * as v from "jsr:@valibot/valibot"
    <p>
      Embeddings. They're one of the parts of the LLM/AI wave that I sort of like.
    </p>
    <p>
      Embeddings are pretty cool when they work, because they sort of capture the idea of{" "}
      <a href="https://blog.val.town/blog/val-vibes/">'vibes', which makes them useful for search applications</a>.
      This project is based on this{" "}
      <a href="https://www.linkedin.com/pulse/insanity-relying-vector-embeddings-why-rag-fails-michael-wood-4iexe/?trackingId=f0RxihkCgaZ4LppqqMP%2FUw%3D%3D">
        blog post I read last year
    const input = inputResult.output
    const embeddings = await client.embed({
      input,
        word,
        embedding: embeddings.data?.at(i)?.embedding,
      }
        <head>
          <title>Surprising embeddings</title>
          <link rel="stylesheet" href="https://unpkg.com/missing.css@1.1.3" />
              <h3>
                Surprising embeddings
                <sub-title>
              </script>
              <script type="module" src="https://esm.town/v/tmcw/surprisingEmbeddings/visualization"></script>
            </div>
      <head>
        <title>Surprising embeddings</title>
        <link rel="stylesheet" href="https://unpkg.com/missing.css@1.1.3" />
            <h3>
              Surprising embeddings
              <sub-title>

tmcw/surprisingEmbeddings/visualization

0 matches

tmcw/surprisingEmbeddings/embeddingGetter

2 matches


  const embeddings = await client.embed({
    input,
      word,
      embedding: embeddings.data?.at(i)?.embedding,
    }

yawnxyz/SidebarToArchive/main.tsx

4 matches

      try {
        // let embeddings = await modelProvider.gen({
        //   embed: true, 
        // });
        // console.log(`Adding link ${counter}:`, link.title, link.url, embeddings);
        console.log(`Adding link ${counter}:`, link.title, link.url);
          data: JSON.stringify(link),
          // embeddings: embeddings.embedding.join(','),
        });

janpaul123/semanticSearchBlobs/main.tsx

13 matches

export default async function semanticSearchPublicVals(query) {
  const allValsBlobEmbeddingsMeta = (await blob.getJSON(`allValsBlob${dimensions}EmbeddingsMeta`)) ?? {};
  const allBatchDataIndexes = _.uniq(Object.values(allValsBlobEmbeddingsMeta).map((item: any) => item.batchDataIndex));
  const embeddingsBatches = [];
  const allBatchDataIndexesPromises = [];
  for (const batchDataIndex of allBatchDataIndexes) {
    const embeddingsBatchBlobName = `allValsBlob${dimensions}EmbeddingsData_${batchDataIndex}`;
    const promise = blob.get(embeddingsBatchBlobName).then((response) => response.arrayBuffer());
    promise.then((data) => {
      embeddingsBatches[batchDataIndex as any] = data;
      console.log(`Loaded ${embeddingsBatchBlobName} (${data.byteLength} bytes)`);
    });
  const openai = new OpenAI();
  const queryEmbedding = (await openai.embeddings.create({
    model: "text-embedding-3-small",
  const res = [];
  for (const id in allValsBlobEmbeddingsMeta) {
    const meta = allValsBlobEmbeddingsMeta[id];
    const embedding = new Float32Array(
      embeddingsBatches[meta.batchDataIndex],
      dimensions * 4 * meta.valIndex,

Vals

tmcw

surprisingEmbeddings

Visualizing embedding distances

Public

yawnxyz

embeddingsSearchExample

Public

thomasatflexos

generateEmbeddings

Public

janpaul123

compareEmbeddings

Public

janpaul123

blogPostEmbeddingsDimensionalityReduction

Public

janpaul123

debugValEmbeddings

Public

maxm

emojiVectorEmbeddings

Public

Users

No users found

2 3 4 5

…

Search

Code94

Vals

Users

Code
94