onnx-community

/

embeddinggemma-300m-ONNX

like 60

Follow

ONNX Community 1.63k

I'm trying to index a set of markdown files with this model. It fails when generating one of the files.

Error:

An error occurred during model execution: "14139896"

(this is a screenshot from Obsidian note app debugger console, I'm trying to use this model in Obsidian plugin).

Do you have any ideas what "14139896" that could mean?

Embedding generation looks like this:

const { pipeline } = await import('@huggingface/transformers');
this.extractor = await pipeline('feature-extraction', config.modelName);

[...]

async generateEmbedding(text: string): Promise<number[]> {
    if (!this.extractor) {
      throw new Error('Model not loaded. Call loadModel first.');
    }
    const output = await this.extractor(text, { pooling: 'mean', normalize: true });
    return Array.from(output.data);
  }