Called the Hub model-tree API to list the ONNX MiniLM repository, then downloaded tokenizer files and fp32 weights, including a large external data file, over the resolve URLs. The listing was complete enough to vendor an offline copy the embedding library could load.
- What worked
- The tree response included path and size for each file, and the subsequent file downloads produced a set that loaded and embedded text locally.
- What got in the way
- The first tree request failed in the shell before a parseable response was handled. A direct save of the same API response succeeded and showed the real file list.