Check the init parameters you can specify for NvidiaTextEmbeder.
YAML Init Parameters
These are the parameters you can specify in pipeline YAML:
Parameter | Type | Possible values | Description |
---|---|---|---|
api_key | Secret | Default: Secret.from_env_var("NVIDIA_API_KEY") | The Nvidia API key. Required. |
model | String | Default: NV-Embed-QA , | The name of the model to use. Required. |
api_url | String | Default: "https://ai.api.nvidia.com/v1/retrieval/nvidia" | Custom API URL for NVIDIA NIM. The URL follows the format: https://host:port .Required. |
prefix | String | Default: "" (empty string) | A string to add at the beginning of each text. Can be used to prepend the text with an instruction, as required by some embedding models, such as E5 and bge. Required. |
suffix | String | Default: "" (empty string) | A text string to add at the end of each text. Required. |
batch_size | Integer | Default: 32 | The count of documents to embed at once. Must be a value below 50. Required. |
progress_bar | Boolean | True False Default: True | If True , shows a progress bar during embedding.Required. |
meta_fields_to_embed | List of strings | Default: None | A list of metadata fields you want to embed along with the document text. Optional. |
embedding_separator | String | Default: "\n" | Separator to concatenate metadata fields into the document text. Required. |
truncate | Union of EmbeddingTruncateMode and string | START END Default: None | Specifies how inputs longer than the maximum token length should be truncated. Possible values: - START : The input is truncated from the start.- END : The input is truncated from the end.- None : The behavior depends on the model. See model documentation for information.Optional. |