← back to catalog · registered 2026-08-22 13:56

ENOSYS/Kimi-Linear-48B-A3B-Instruct-abliterated-1000-v1-GGUF

ENOSYS Kimi 48B GGUF MoE second-order 1.0M ctx
Your rig guess connected
? Why do I need an app?
Reading your rig…

This is a rough estimate. Install the free app - we'll show exact numbers.

Reading real hardware from your app right now. Numbers below are exact.

Below is the per-quantization compatibility for this model.

curl -H "Authorization: Bearer $ABL_KEY" \
     "https://abliteration.org/api/v1/models/ENOSYS%2FKimi-Linear-48B-A3B-Instruct-abliterated-1000-v1-GGUF"
Response includes
  • classification m8
  • files 43
  • hub_downloads_all_time 2,790
  • author_summary 2 models
  • readme_text full
10 credits · hourly refresh · ~4 KB payload Get an API key →
Abliteration classifier · v1.0.0
M8
Primary method

Repackaging (quantization)

Applied on top of direct removal inherited from the base model.
Confidence
MEDIUM
Inherited from base model
Why this label 3 signals
Method inferred from partial signals - repository name, related files, or tag patterns. Producer identity not confirmed; label may sharpen or shift as we gather more evidence.
  • 'abliterated' in name/tags
  • is_gguf=1
  • assume M1 (base ablation) + M8 (GGUF quant) - default when producer unknown
Refusal direction extraction

No specific extraction method could be identified for this model. The producer either did not document it or used a proprietary pipeline.

What is a refusal direction? →
Downloads · lifetime
3K
308 last 30d - stable
Likes
1
Model age
5mo ago
created 2026-05-12

Training datasets

Corpora the author lists in the model card. Datasets tracked in our /datasets catalog carry a category badge linking to the workflow stage. Others open on Hugging Face.

Downloads over time
Now2.9K→from616↑369%
5031.4K2.2K3.1K616 on May 132.9K on Oct 11MayJunJulAugSepOct
May 13 → Oct 11 · 61 snapshots · spans 151 days

Genealogy 0 direct forks

Full fork graph →

This model's place in the market. Above: what it was derived from. Below: the tree of everything derived from it.

Metadata

License
mit
Languages
ru en multilingual
Tags
transformers gguf uncensored abliterated text-generation ru en multilingual dataset:eaddario/imatrix-calibration license:mit endpoints_compatible region:us

Related

Total size
206 GB
Files
43
Quantizations
1
Registered
2026-08-22 13:56
Last updated on HF
2026-05-12 16:49

Files by quantization

Auxiliary files 43 files 206 GB
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW5.00.gguf 28.6 GB 13623f70 download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW4.75.gguf 27.2 GB 20e9def7 download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW4.50.gguf 25.7 GB 9c0a086a download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW4.25.gguf 24.3 GB aed0ddcf download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW4.00.gguf 22.9 GB 171f23ee download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW3.75.gguf 21.5 GB c8583ab1 download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW3.50.gguf 20.0 GB a13839e0 download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW3.25.gguf 18.6 GB fa070a24 download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW3.00.gguf 17.2 GB beee7801 download
imatrix.gguf 146 MB ddd3b0be download
dataset.txt 2.34 MB 7d3c7041 download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW4.00-quantize.log 249 KB 53e16e40 download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW3.75-quantize.log 249 KB fc56f567 download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW4.25-quantize.log 249 KB ccf9dbeb download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW4.75-quantize.log 249 KB 13ea6b96 download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW5.00-quantize.log 249 KB 8ba48503 download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW3.00-quantize.log 249 KB 9c252d35 download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW3.25-quantize.log 249 KB 350392fd download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW3.50-quantize.log 249 KB cb21d66d download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW4.50-quantize.log 249 KB b2da633c download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW4.75-kl-divergence.log 45.3 KB f6adf741 download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW4.25-kl-divergence.log 45.3 KB 95ae73b9 download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW4.50-kl-divergence.log 45.2 KB 4f5d0b53 download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW5.00-kl-divergence.log 45.2 KB d02042fa download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW3.75-kl-divergence.log 44.4 KB 27a43e84 download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW4.00-kl-divergence.log 43.7 KB ef9ea079 download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW3.50-kl-divergence.log 34.0 KB 3a475df9 download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW3.00-kl-divergence.log 33.9 KB b59183cc download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW3.25-kl-divergence.log 33.9 KB 14d38050 download
kl-divergence-base.log 27.4 KB f0edf15d download
imatrix.log 26.7 KB f803086c download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW3.00-quants.txt 19.4 KB edecb5c9 download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW4.25-quants.txt 19.3 KB a797c6ac download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW4.00-quants.txt 19.3 KB 2b49f82c download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW4.50-quants.txt 19.3 KB 0d5132db download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW3.25-quants.txt 19.3 KB 16b37c8e download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW3.75-quants.txt 19.3 KB 48046f23 download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW3.50-quants.txt 19.2 KB 9afc1028 download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW4.75-quants.txt 19.0 KB 0e342bda download
Kimi-Linear-48B-A3B-Instruct-abliterated-BPW5.00-quants.txt 18.9 KB eefc2efd download
README.md 2.57 KB 30cdf911 download
.gitattributes 1.53 KB f15b49c2 download
statistics.md 1.51 KB 353c31eb download

README current version from Hugging Face


license: mit
pipeline_tag: text-generation
base_model:

  • huihui-ai/Huihui-Kimi-Linear-48B-A3B-Instruct-abliterated
    library_name: transformers
    tags:
  • uncensored
  • abliterated
    datasets:
  • eaddario/imatrix-calibration
    language:
  • ru
  • en
  • multilingual

Experimental global target bits‑per‑weight quantization of huihui-ai/Huihui-Kimi-Linear-48B-A3B-Instruct-abliterated

  • Using non-standard (forked) LLaMA C++ branch for quantization.
  • Using a CLI tool to build KLD evaluation and imatrix calibration datasets for GGUF models, sourced from eaddario/imatrix-calibration.
  • Using dataset sources: tools, math, code, text_en, text_ru.
  • Using dataset chunks: 1000.
  • Tensors quantinization F16 instead of BF16, Nvidia Pascal architecture friendly like P100.
  • Small set of patches added.

Many thanks to Ed Addario for an impressive job.

Quantization comparison

BPW/TGS SIZE (MiB) PPL correlation PPL mean ratio ΔPPL Mean KLD Median KLD Maximum KLD 99.9% KLD Mean Δp RMS Δp
3.000 17574 97.54% 1.099746 ± 0.001198 0.703413 ± 0.008959 0.122940 ± 0.000426 0.066714 9.294135 2.637652 -2.015 ± 0.019 % 10.272 ± 0.039 %
3.250 19038 98.25% 1.064443 ± 0.000969 0.454454 ± 0.006978 0.089130 ± 0.000308 0.050624 10.323897 1.992209 -1.732 ± 0.016 % 8.814 ± 0.035 %
3.500 20502 98.69% 1.042289 ± 0.000821 0.298224 ± 0.005835 0.064519 ± 0.000265 0.032550 8.311751 1.765223 -1.217 ± 0.014 % 7.481 ± 0.033 %
3.750 21966 99.04% 1.030411 ± 0.000697 0.214459 ± 0.004989 0.047054 ± 0.000201 0.022846 8.346386 1.258301 -0.654 ± 0.012 % 6.340 ± 0.031 %
4.000 23430 99.15% 1.028018 ± 0.000653 0.197583 ± 0.004668 0.042084 ± 0.000178 0.020843 7.031458 1.095171 -0.639 ± 0.011 % 5.992 ± 0.029 %
4.250 24894 99.25% 1.021574 ± 0.000610 0.152141 ± 0.004327 0.036486 ± 0.000158 0.018297 6.763760 1.009924 -0.568 ± 0.011 % 5.619 ± 0.028 %
4.500 26358 99.50% 1.012304 ± 0.000495 0.086766 ± 0.003506 0.024419 ± 0.000105 0.012078 5.742187 0.643191 -0.204 ± 0.009 % 4.565 ± 0.023 %
4.750 27822 99.61% 1.007470 ± 0.000434 0.052677 ± 0.003063 0.019040 ± 0.000086 0.009403 4.860535 0.504090 -0.200 ± 0.008 % 4.035 ± 0.020 %
5.000 29286 99.67% 1.006065 ± 0.000397 0.042769 ± 0.002799 0.015935 ± 0.000073 0.008021 3.658825 0.425957 -0.201 ± 0.007 % 3.731 ± 0.020 %

README history 3 versions

The author's README evolved over time. Click a version to see its content at that point.

  1. 2026-05-12Upload files57aaccb2.6 KB
    Loading...
  2. 2026-05-12Update README.md10273621.1 KB
    Loading...
  3. 2026-05-12initial commit118234828 B
    Loading...
Catalog is the map. Apps are the tools.

Run models on your own machine, not in the cloud.

Every model page has an "Open in Abliteration" button that hands the model directly to the first-party desktop client, at the quantization your rig can actually run. No API keys, no subscription, no prompt leakage.

Open in Abliteration