perf: Tev1 weights on fused LinearAttention ONNX-OPT graph

ATTRIBUTION.md

@@ -1,11 +1,25 @@
-# Attribution — browser WebGPU model slice
+# Attribution — Tev1 ONNX for edgextract
 
-## Tev1 ONNX (default)
+## Together Tev1-0.8B-experimental
 
-- Weights: [togethercomputer/Tev1-0.8B-experimental](https://huggingface.co/togethercomputer/Tev1-0.8B-experimental) — Together AI
-- Fine-tune release license: **being finalized** on the Hub card (exported with acknowledgment)
-- Topology: [onnx-community/Qwen3.5-0.8B-ONNX](https://huggingface.co/onnx-community/Qwen3.5-0.8B-ONNX) (Apache-2.0)
-- Base: [Qwen/Qwen3.5-0.8B](https://huggingface.co/Qwen/Qwen3.5-0.8B) (Apache-2.0)
-- Hub: [raphaelmansuy/tev1-0.8b-onnx-webgpu](https://huggingface.co/raphaelmansuy/tev1-0.8b-onnx-webgpu)
+- Hub: https://huggingface.co/togethercomputer/Tev1-0.8B-experimental
+- Publisher: Together AI
+- Code: https://github.com/togethercomputer/tev1
+- Blog: https://www.together.ai/blog/how-to-train-your-own-jev
 
-Full notes: [docs/THIRD_PARTY_NOTICES.md](../../../docs/THIRD_PARTY_NOTICES.md)
+**License:** the base Qwen3.5-0.8B is Apache-2.0. Together states that the
+release license for these **fine-tuned** weights is still being finalized.
+Do not redistribute this ONNX graph publicly until that license is clear, or
+until you accept that risk with ``--acknowledge-tev1-license-pending``.
+
+## Qwen3.5-0.8B (base)
+
+- Hub: https://huggingface.co/Qwen/Qwen3.5-0.8B
+- License: Apache-2.0
+
+## ONNX topology template
+
+- Hub: https://huggingface.co/onnx-community/Qwen3.5-0.8B-ONNX-OPT
+- License: Apache-2.0 (onnx-community / Qwen)
+
+See also `docs/THIRD_PARTY_NOTICES.md` in the edgextract repository.