Fix stale zaya1-godspeed GitHub links (repo renamed to zaya1-nvfp4-w4a4)

README.md

@@ -73,7 +73,7 @@ failing too) — neither currently offers a fix that preserves CUDA graphs for
 this failure mode. See [Pape, Evertz & Schönherr (arXiv:2605.19537)](https://arxiv.org/abs/2605.19537)
 for the general phenomenon of backend-dependent correctness drift in LLM
 serving. Full sweep, coherence re-verification, and citations in `RESEARCH.md`
-and `ROADMAP.md` on the [GitHub repo](https://github.com/t-timms/zaya1-godspeed).
+and `ROADMAP.md` on the [GitHub repo](https://github.com/t-timms/zaya1-nvfp4-w4a4).
 
 **Set `--enforce-eager` (or `enforce_eager=True`) when serving this checkpoint.**
 It is the only configuration confirmed to produce coherent output.
@@ -122,7 +122,7 @@ checkpoint, more so than the accuracy difference.
 > attempts to reproduce it on this project's own SM120 hardware hit the same
 > non-deterministic hang each time, ruling out the model and toolkit version
 > as causes; looks like a WSL2/driver-level issue. Not resolved as of
-> 2026-08-14 — see [GitHub repo](https://github.com/t-timms/zaya1-godspeed)
+> 2026-08-14 — see [GitHub repo](https://github.com/t-timms/zaya1-nvfp4-w4a4)
 > `RESEARCH.md` §5.16 for the full diagnostic log.
 >
 > **Update, same day:** the gap likely has a principled cause independent of
@@ -224,13 +224,13 @@ calibrate through the fake-quant `nn.Linear.forward` wrapper; and the NVFP4
 global-scale convention (`2688 / max_abs`, divisor form, block scales
 pre-multiplied) is undocumented — getting it wrong produces silent pad-token
 collapse, not an error. Full root-cause writeup in `RESEARCH.md` on the
-[GitHub repo](https://github.com/t-timms/zaya1-godspeed).
+[GitHub repo](https://github.com/t-timms/zaya1-nvfp4-w4a4).
 
 ## Usage
 
 Requires vLLM built from source with SM120 NVFP4 CUTLASS kernels (stock wheels
 don't include them) — see the
-[reproduce steps](https://github.com/t-timms/zaya1-godspeed#reproduce).
+[reproduce steps](https://github.com/t-timms/zaya1-nvfp4-w4a4#reproduce).
 Inference must run in `bfloat16` (not fp16/fp32).
 
 ```bash
@@ -247,7 +247,7 @@ See "Known issue" above before serving without it.
 decoding, validated **2.2× faster on coding-edit prompts** (no gain on
 free-form generation, since there's no prompt/output overlap to exploit).
 Detail: `RESEARCH.md` §5.18. The exact command above is
-[`scripts/serve.sh`](https://github.com/t-timms/zaya1-godspeed/blob/master/scripts/serve.sh)
+[`scripts/serve.sh`](https://github.com/t-timms/zaya1-nvfp4-w4a4/blob/master/scripts/serve.sh)
 in the repo.
 
 ## Evaluation
@@ -268,7 +268,7 @@ in the repo.
 > will serve you better.
 > **MMLU-Pro is 0-shot and is not comparable to Zyphra's 5-shot 74.2%** — that
 > gap is a protocol difference, not quantization damage. Full analysis:
-> [`RESEARCH.md` §5.22](https://github.com/t-timms/zaya1-godspeed/blob/master/RESEARCH.md).
+> [`RESEARCH.md` §5.22](https://github.com/t-timms/zaya1-nvfp4-w4a4/blob/master/RESEARCH.md).
 
 Budget-forced GPQA-Diamond using an s1-style harness that caps the reasoning
 trace and scores only the closed answer — stock lm-eval harnesses score near