Mia posted a two-Spark recipe. We froze the GitHub tree, pulled the EXL3 checkpoint, kept yesterday's Tony DFlash2 as the immediate rollback, and ran the same gates we ran last night. The served id changed. Echo now asks for GLM-5.3-Flash-EXL3.
This is Mia's stack. We changed the fabric to Funland bond0 and measured.
f3043c95bbf95fb91dd160fe58d740cd152a02c3. Tweet 2093351744885444871 quoted 63.5 tok/s structured on someone else's desk. That is not our number.25a44fdbf16862a46b7cc9921142c6c81350af2f, ~164 GiB, 120 shards. ShapleyMCG license.7d74cdd881ed7e32c31175984a67823127b66cfe. CC BY-NC-ND 4.0.SKIP_PULL=1. Do not copy Mia's CX7 interface names onto a bonded Funland pair. Those slaves have no IPs and NCCL hangs.Same pair as last night. Same extra-gates script and Tony C1–C6 harness (model id patched). Needles go through urllib; curl -d dies on Linux ARG_MAX past ~32K. Score needles by reported prompt_tokens, not filler-word guesses. First-pass fillers undershot; the calibrated retries are the ones below.
| Tony NVFP4 DFlash2 | Mia EXL3 DFlash2 | |
|---|---|---|
| Mixed C1 | 41.5 tok/s | 31.0 tok/s |
| C6 aggregate | 54.2 | 50.7 |
| JSON G7 / E5 | 46.3 / 55.3 | 52.8 / 54.04 |
| Prose G7 / E5 | 21–23 | 20.5 / 20.95 |
| Count E5 | 67.7 | 67.05 |
| Accept (fixture) | 58.3% | 62.7% (6509 / 10388) |
| Tools | 20/20 | 20/20 in 36s |
| Vision | red | red |
| Needle | 98,276 HIT | 98,251 HIT · 199,970 HIT |
| KV tokens | 310,292 | 941,208 |
| Cold bind | ~19 min | ~11 min |
C1–C6: zero failures. Extra gates: EXTRA_GATES_PASS. Concurrent 3×18,432-token prefills all HIT. C5/C6 queue because Mia's launcher sets MAX_NUM_SEQS=4.
git clone https://github.com/MiaAI-Lab/GLM-5.3-Flash-EXL3-2x-DGX-Sparks
git checkout f3043c95bbf95fb91dd160fe58d740cd152a02c3
hf download Mia-AiLab/GLM-5.3-Flash-EXL3-TR3-4bpw \
--revision 25a44fdbf16862a46b7cc9921142c6c81350af2f
hf download incoai/GLM-5.3-Flash-DFlash2 \
--revision 7d74cdd881ed7e32c31175984a67823127b66cfe
docker pull ghcr.io/miaai-lab/glm-5.3-flash-2x-dgx-sparks:exl3
# image id both ranks: sha256:ad0cdd86d1ddd15ee758f519d16da15ac237f7f0648a5c52fbc20f9554944263
Funland .env fabric (not Mia's CX7 defaults):
HEAD_IP=10.0.0.1
WORKER_IP=10.0.0.2
HEAD_CX7_IF=bond0
WORKER_CX7_IF=bond0
HEAD_CX7_IB=rocep1s0f0,rocep1s0f1
WORKER_CX7_IB=rocep1s0f0,rocep1s0f1
SKIP_PULL=1
SERVED_MODEL_NAME=GLM-5.3-Flash-EXL3
SPEC_METHOD=dflash
DFLASH_TOKENS=7
MAX_MODEL_LEN=900000
MAX_NUM_SEQS=4
ENFORCE_EAGER=0
Then ./start.sh start from Spark1 after Tony is off port 8888. Worker first is inside Mia's launcher. Overlay patches live under overlay/, not the repo root.
Forge default is GLM-5.3-Flash-EXL3 via spark-ds4. Context stays 131,072. CLI probe returned ECHO_ROUTE_OK. Gateway restarted after the picker tidy. Live aliases: ds4 / exl3, M5 select / vision / omlx27, plus the cloud shortcuts. Dead locals (:8016, :8013, :8026, leftover deepseek-v4-flash-dspark) are gone from the picker.
This writeup's Desktop session is still milo on grok-4.6. Spark is Echo's default, not milo's.
ssh milo@192.168.1.11 /home/milo/glm53-mia-exl3-20260828/restore-dflash.sh manual
That relaunches Tony DFlash2 (funland/glm53-dflash2:ccc5a78e = sha256:5be2a57f…) as glm-5.3-flash. Echo would need the model id flipped back. The older 0xSero SGLang image may no longer be on disk.