File size: 1,800 Bytes
315dd06 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 | # Sources and attribution
## Shared model
This interface uses [convaiinnovations/laya](https://huggingface.co/convaiinnovations/laya/tree/55cf4c4ebb4ebe31b2550e8bdf3bd21b99753851) at revision `55cf4c4ebb4ebe31b2550e8bdf3bd21b99753851`. The runtime downloads and caches the base separately. Original Laya embeddings, encoder blocks and decision heads remain frozen.
## Software
Ariadne's runtime, prompts and benchmark preparation use code or recipes adapted from:
- [NandhaKishorM/laya](https://github.com/NandhaKishorM/laya/tree/4066d5d5fbf08b66c6757ddeedbd797bd7655bc0), commit `4066d5d5fbf08b66c6757ddeedbd797bd7655bc0`, Apache-2.0. Notice: [licenses/laya-LICENSE](licenses/laya-LICENSE).
- [AbdelStark/jev-benchmarks](https://github.com/AbdelStark/jev-benchmarks/tree/0d610cc53e79bcbec691312b0c4adb4a0e371642), commit `0d610cc53e79bcbec691312b0c4adb4a0e371642`, Apache-2.0. Notice: [licenses/jev-benchmarks-LICENSE](licenses/jev-benchmarks-LICENSE).
Ariadne's license is in [LICENSE](LICENSE). Other installed dependencies retain their own licenses. Independent adaptation; no affiliation with or endorsement by the source authors.
## Training data and evaluation
Dataset: [bitext/Bitext-customer-support-llm-chatbot-training-dataset](https://huggingface.co/datasets/bitext/Bitext-customer-support-llm-chatbot-training-dataset/tree/430d1a89bd93bd1fa23c16f29dd53e73f0087443), revision `430d1a89bd93bd1fa23c16f29dd53e73f0087443`. Dataset metadata is preserved in [source.json](source.json). The dataset retains its own license; training data is not included in this package.
See [README.md](README.md) for task scope and comparison limitations, and [metrics.json](metrics.json) for the public comparison model's revision and evaluation details. Public comparison weights are not included.
|