From 0bec3bb3f0dd4ed75cc1a1228fe09382196d5a8e Mon Sep 17 00:00:00 2001 From: Joey Grasty Date: Sun, 30 Aug 2026 17:40:21 -0500 Subject: [PATCH] Add ep3 Drive link, keep ep2 available for comparison ep3 uploaded to the team Drive folder and linked as the current production checkpoint. ep2's link is kept in place per request, labeled as the previous production checkpoint rather than removed. --- README.md | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/README.md b/README.md index da88c82..c01930c 100644 --- a/README.md +++ b/README.md @@ -25,12 +25,17 @@ The 208 training/holdout packages themselves are already embedded inside ### 1. Get the model file Download `Qwen3-30B-A3B-VoxDay-Nuttall-SFT-ep3-Q8_0.gguf` (~31GB) from -Google Drive: +Google Drive — **this is the current production checkpoint**, switched from +ep2 on 2026-08-30 based on separability and length-calibration evidence +(see `JOINT_TRAINING_CLOSEOUT_REPORT.md`): + +https://drive.google.com/open?id=1QZT488Uu1b9Pxoie9Hs2lnxtOhMD4Pwr + +The previous production checkpoint, `Qwen3-30B-A3B-VoxDay-Nuttall-SFT-ep2-Q8_0.gguf`, +remains available for comparison: https://drive.google.com/open?id=1q1w_XMp6oRmfx8xqXDV-HvDx_krVxQtD -**Note (2026-08-30): the link above still points to the ep2 file.** Production was just switched to ep3 (separability + length-calibration evidence, see `JOINT_TRAINING_CLOSEOUT_REPORT.md`); the ep3 GGUF upload and link swap are pending. - Memory needed: ~31GB for the weights plus ~3GB of KV cache at the full 32768-token context, so ~35GB total. Confirmed working on a DGX Spark; should comfortably fit a 48GB L40 as well. If a machine has less than that,