# Modifications from the upstream model

`embeddinggemma-300m-memory-ft-v1` is not an unmodified copy of
`google/embeddinggemma-300m`. It is a Daecore-modified derivative and is not
endorsed by Google.

What changed:

- **Fine-tune.** The base weights were trained further for semantic retrieval
  over Daecore's governed memory corpus, using query and document role
  prefixes, graded positive supervision, hard negatives, Matryoshka
  dimensions, and attention LoRA (rank 16, alpha 32). The adapter was merged
  into the model weights. Relevance grades were frontier-model judgments under
  a frozen protocol, not human annotations.
- **Export.** The merged model was exported to ONNX as one graph taking
  `input_ids` and `attention_mask` and emitting 768-dimensional embeddings in
  FP32, with the parameters stored as external data.
- **Serving contract.** `serving.json` records the runtime and sequence
  geometry of the original export qualification: 128 query tokens, 1,024 passage
  tokens and a CUDA batch of 32. Production batching adapts to available memory
  within the separately qualified CPU, CUDA and DirectML limits.

Files modified or generated by Daecore:

- `model.onnx`: the generated ONNX graph of the fine-tuned model;
- `model.onnx.data`: the fine-tuned parameters used by that graph;
- `config.json`: the export configuration of the modified model;
- `serving.json`: the Daecore runtime and sequence-geometry contract.

Unchanged: `tokenizer.json`, `tokenizer_config.json`,
`special_tokens_map.json`, and `added_tokens.json` are the tokenizer files of
the selected model export and bind the exact text-processing contract.

This distribution is subject to the Gemma Terms of Use (`LICENSE`) and the
Gemma Prohibited Use Policy (`GEMMA_PROHIBITED_USE_POLICY.txt`).

- **Portable reshape correction.** Set `allowzero=0` on the 96 reshape operators
  whose target shape contains an inferred dimension. This avoids the DirectML
  interpretation failure without changing any valid reshape or model weights.
  CPU, CUDA and DirectML use the same graph.