Upload folder using huggingface_hub

fixup readme

Release IQ3_KS with perplexity info

Support added into ik_llama.cpp main branch now yay!

Upload folder using huggingface_hub

Add IQ1_KT (with iq4_nl ffn_down_exps lmao)

update readme with IQ4_K and IQ5_KS

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Uploading IQ4_K and IQ5_KSS

fixup IQ4_KSS recipe

add recipes and perplexity graph

fixup readme

Upload folder using huggingface_hub

Upload IQ2_KL and add perplexity values

Upload folder using huggingface_hub

Add IQ5_K 6BPW with similar PPL as Q8_0

Add --chat-template chatglm4 again for /nothink

Add BF16 Perplexity

add two perplexity numbers

Delete IQ4_KSS/WARNING-EXPERIMENTAL-IKLLAMACPP-ONLY-GLM-4.5-Air-IQ4_KSS-00002-of-00002.gguf

Delete IQ4_KSS/WARNING-EXPERIMENTAL-IKLLAMACPP-ONLY-GLM-4.5-Air-IQ4_KSS-00001-of-00002.gguf

Updating imatrix and IQ4_KSS

Upload folder using huggingface_hub

Upload imatrix-GLM-4.5-Air-BF16.dat with huggingface_hub

May need --override-kv tokenizer.ggml.eot_token_id=int:151336

Update WARNING EXPERIMENTAL quant instructions

WARNING EXPERIMENTAL IQ4_KSS

Upload folder using huggingface_hub

Upload imatrix-GLM-4.5-Air-BF16.dat with huggingface_hub

add BF16 and Q8_0 size info

initial commit

initial commit