baa-ai
/

Llama-3.1-70B-Instruct-SWAN-5bit-MLX

mixed-precision

4-bit precision

Model card Files Files and versions

Llama-3.1-70B-Instruct-SWAN-5bit-MLX / README.md

Commit History

Fix YAML metadata formatting

03e40fc
verified

tomkay commited on Apr 16

Fix: normalise line endings (CR removal) in README

eb52b3e
verified

tomkay commited on Apr 15

Upload README.md with huggingface_hub

805b502
verified

tomkay commited on Apr 15

Upload README.md with huggingface_hub

cbfbdff
verified

tomkay commited on Apr 15

Update model card: remove MINT/SWAN branding, optimised by baa.ai

e892cc1
verified

tomkay commited on Apr 5

Update model card: use generic language for allocation method

76932f9
verified

tomkay commited on Apr 3

Update model card: add MINT-UI link and custom quantization guide

fcf96d8
verified

tomkay commited on Apr 3

Update base_model format to include relation: quantized

51c1684
verified

tomkay commited on Mar 25

Add model card with metrics and usage

0c3be1b
verified

tomkay commited on Mar 20

Upload README.md with huggingface_hub

972fb37
verified

Trevor Kennedy commited on Mar 13

Upload README.md with huggingface_hub

cf0ae25
verified

Trevor Kennedy commited on Mar 13

SWAN adaptive mixed-precision quantization (5.51 avg bits, MAD bounds)

abf7841
verified

Trevor Kennedy commited on Mar 12