cudabenchmarktest/personaplex-7b-turbo2bit at main

personaplex-7b-turbo2bit

5.01 GB

Ctrl+K

Ctrl+K

1 contributor

History: 17 commits

cudabenchmarktest's picture

cudabenchmarktest

fix: Update open-agents link to npm (npmjs.com/package/open-agents-ai)

1441cf6 verified 2 months ago

voices
Add OverBarn voice 2 months ago
.gitattributes

1.52 kB
initial commit 2 months ago
README.md

5.39 kB
fix: Update open-agents link to npm (npmjs.com/package/open-agents-ai) 2 months ago
clone-voice.py

15.1 kB
Add clone-voice.py 2 months ago
config.json

692 Bytes
Add config.json — inference harness 2 months ago
dequant-loader.py

5.79 kB
Add dequant-loader.py — inference harness 2 months ago
dequantize.py

5.74 kB
Add dequantize.py — inference harness 2 months ago
linear2bit.py

18.9 kB
Add linear2bit.py — inference harness 2 months ago
model-2bit-calibrated.safetensors

2.4 GB
xet

Add calibrated 2-bit weights (GPTQ-style, still experimental) 2 months ago
model-turbo2bit.safetensors

2.22 GB
xet

PersonaPlex 7B TurboQuant 2-bit (NF2+WHT) — 7.5x compression, 2.07 GB 2 months ago
quantize-weights.py

6.5 kB
Add quantize-weights.py 2 months ago
tokenizer-e351c8d8-checkpoint125.safetensors

385 MB
xet

Add tokenizer-e351c8d8-checkpoint125.safetensors (no HF_TOKEN needed for full setup) 2 months ago
tokenizer_spm_32k_3.model

553 kB
xet

Add tokenizer_spm_32k_3.model (no HF_TOKEN needed for full setup) 2 months ago