HuggingFaceTransformersEmbeddings supports local model path

[spring1915]

Checked

• I searched existing ideas and did not find a similar one
• I added a very descriptive title
• I've clearly described the feature request and motivation for it

Feature request

This langchainjs doc only shows how the script downloads the embedding model. I want to load the model that has been manually downloaded to a local path due to security concerns.

Is there a way to do that?

Motivation

• We need to have a model downloaded by hand earlier as our network prevents direct retrieval from HuggingFace.
• We want to develop our internal embedding model.

Proposal (If applicable)

No response