Modelpublic

davidsyoung/GLM-5.3-EXL3-TR3-3.0bpw

sourceHugging Faceotherupdated 23d agoView on Hugging Face

3likes829downloads

Model Card

GLM-5.3 — EXL3/TR3 3.0 bpw (mixed K3/K4 trellis, data-free)

[!NOTE] Weights live and byte-verified (all 83 LFS files sha256-matched against the assembly manifest; structural gate at assembly verified all 58,368 promoted expert slots K3-shaped). KLD measurement to follow — the sibling releases' measured table below gives the ladder context; this checkpoint's row will be published here. Expectation revised 2026-08-30: brandonmusic's independent calibrated uniform-K3 draft scored 0.0375 (clean bf16 replay) on this lineage's exact confirmation windows; since this checkpoint is uniform-K3 without calibration and is scored through fp8-KV serving, expect ~0.04-0.05 here, not the naive ladder extrapolation (~0.030). Flat K3 concentrates damage in the worst-expert tail that the 3.25/3.42 siblings protect with K4 — that is the trade this artifact makes for context breadth. The reproduction kit in `kld/` scores it unmodified.

Trellis (EXL3) quantization of zai-org/GLM-5.3 (755B glmmoedsa MoE: 78 layers + MTP, 256 routed experts/layer), following the GLM-5.2 TR3 lineage layout. Fits TP4 on 4x 96 GB (RTX PRO 6000 Blackwell class) with FP8 KV cache.

From the upstream card: "GLM-5.3 uses the same base model as GLM-5.2 — every gain comes from post-training. Compared with GLM-5.2, it is much better at complex coding and long-horizon tasks." See the base model card for benchmarks, serving guides, and the technical report (arXiv:2602.15763).

What is quantized

component	treatment
routed experts (layers 3–78, incl. MTP-78)	EXL3 trellis, per layer 256 experts K3 (avg 3.0 bpw), mcg codebook
dense MLP (layers 0–2), all attention, norms, embeddings, lmhead, mlp.gate, ehproj	BF16, carried byte-exact
shared experts	BF16 in-checkpoint (online K6 at serve)

—Data-free encode: identity Hessian (H=I), rotations + trellis search only — no calibration capture anywhere. Deterministic: seeds derived from (layer, expert, projection, rank), seed_base 20260711.
—K4 selection: per layer, the experts with highest relative round-trip MSE under the K3 encode are re-encoded at K4 (downward remix: the 192 never-promoted K3 experts + the 64 banked donor-K3 payloads from the mix ledger — zero re-encode, byte-identical to a PROMOTE_K4=0 run).
—Per-expert tier map in tier_bitmap.json; encode provenance in config.json.hybrid_tr3_tail; file hashes in MANIFEST.sha256.

Serving

TR3 mixed-bit layout (carrier BF16 shards + per-layer trellis payloads) — serve with an exllamav3-b12x/sparkinfer-lineage stack, TP4 (+ DCP4/MTP3), FP8 KV. The mixed-K projection-tiers patch is REQUIRED; a stock loader that assumes a uniform K per layer will produce fluent garbage. Not loadable by vanilla exllamav3 model loading.

KLD vs BF16 teacher

Full-vocabulary, teacher-forced KL(teacher || student) against sealed BF16 GLM-5.3 logits — 4 held-out windows × 2,047 positions × 154,880 vocabulary, fp32 log-softmax both sides. Measured independently on two different 4× RTX PRO 6000 (96GB) machines:

weight quant	KV mode	this work	CN3 ([@dareposte](https://huggingface.co/dareposte))	Δ
3.42 bpw	fp8	0.024105	0.023966	−0.6%
3.25 bpw	fp8	0.026103	0.026776	+2.6%
3.25 bpw	nvfp4	0.035741	0.036661	+2.6%
3.42 bpw	nvfp4	0.037757	0.037060	−1.8%
3.42 bpw	nvfp4+rope8	0.039518	0.037695	−4.6%
3.25 bpw	nvfp4+rope8	—	0.039396	CN3 only

Readings: the 3.25↔3.42 weight step changes KLD by only ~0.002; the fp8→nvfp4 KV step costs ~7× more (~0.014) — cache format matters more than the extra 0.17 bpw. Under nvfp4 KV the weight-quant difference washes out entirely. The window SD (~0.02–0.03) is corpus heterogeneity — one citation-dense legal window is uniformly hardest; dialogue, explanatory prose, and reasoning-trace registers measure near-transparent.

<details> <summary>Per-window means (dialogue / legal / prose / reasoning-trace)</summary>

config	source	w0000	w0001	w0002	w0003
3.42 fp8	this work	0.0141	0.0537	0.0137	0.0148
3.42 fp8	CN3	0.0148	0.0542	0.0135	0.0134
3.25 fp8	this work	0.0188	0.0579	0.0138	0.0139
3.25 fp8	CN3	0.0183	0.0592	0.0142	0.0154
3.42 nvfp4	this work	0.0199	0.0816	0.0221	0.0274
3.42 nvfp4	CN3	0.0198	0.0775	0.0210	0.0300
3.25 nvfp4	this work	0.0256	0.0799	0.0184	0.0190
3.25 nvfp4	CN3	0.0255	0.0804	0.0189	0.0218
3.42 nvfp4+rope8	this work	0.0212	0.0812	0.0257	0.0300
3.42 nvfp4+rope8	CN3	0.0218	0.0761	0.0236	0.0292
3.25 nvfp4+rope8	CN3	0.0268	0.0883	0.0204	0.0221

</details>

Method (reproducible)

—Teacher: brandonmusic/GLM-5.3-BF16-full-logits, reference-full-panel confirmation lane (held out from every calibration fit), revision 427368f1.
—Student: this checkpoint, loaded by the digest-pinned r17 serving image (sha256:c5e96c5b…) — the real trellis kernels and online-K6 path, offline vllm.LLM, TP4, one teacher-forced prefill per window.
—Full runbook + runner: `kld/` in this repo (KLD-REPRODUCTION.md, prefill_kld_53.py, fetch-teacher.sh).
—Independent reproduction bundle (receipts, unedited logs, checksums, pinned revisions): `kld/cn3/`.

Capacity caveat (via the CN3 report): KV-pool token figures printed by KLD-profile boots (TP4/DCP1, 4,096-token envelope) are logical pool values for that profile only — they are not maximum context length or production serving capacity.

Credits

—[brandonmusic](https://huggingface.co/brandonmusic) — thank you for the GLM-5.3-BF16-full-logits teacher captures that make this measurement possible without a 1.5TB BF16 forward, and for the TR3 quantization references this release follows: the GLM-5.2-EXL3-TR3v4-3.5bpw-MTP78 method card, runbook, and the r10 reproducibility bundle whose encoder lineage (encode_tr3_v31.py) this checkpoint was produced with.
—[dareposte](https://huggingface.co/dareposte) — thank you for the independent CN3 reproduction: all six weight/KV configurations on separate hardware, within ±5% of our means, published here with receipts under `kld/cn3/`.
—willfalco/GLM-5.2-EXL3-TR3-3.25bpw — the 3.25 tier recipe and mixed-K checkpoint-format lineage.
—local-inference-lab — the qualified r17 serving stack these artifacts boot on.

Status

Context-maximal artifact of the release trio (3.42 quality / 3.25 balanced / 3.0 breadth): ~24 GiB lighter than 3.25 => roughly +530k KV tokens at fp8. Structurally gated at assembly (58,368 promoted slots shape-verified K3). KLD measurement to follow — expected ~0.030 fp8 by ladder extrapolation; the reproduction kit in kld/ scores it unmodified.

Quantized with encode_tr3_53.py (exllamav3 v0.0.43 vendored math, MIT) on 4x RTX PRO 6000 Blackwell. Encode + tooling notes ship in the repo.