Uploading IQ4_K and IQ5_KSS

README.md

@@ -100,6 +100,123 @@ numactl -N 0 -m 0 \
 
 </details>
 
+## IQ5_KS 72.855 GiB (5.665 BPW)
+Final estimate: PPL = 4.5948 +/- 0.02815
+
+<details>
+
+<summary>👈 Secret Recipe</summary>
+
+```bash
+#!/usr/bin/env bash
+
+custom="
+# 47 Repeating Layers [0-46]
+# Note: All ffn_down.* layers are not divisible by 256 so have limited quantization options.
+
+# Attention
+blk\..*\.attn_q.*=iq5_ks
+blk\..*\.attn_k.*=q8_0
+blk\..*\.attn_v.*=q8_0
+blk\..*\.attn_output.*=iq5_ks
+
+# First 1 Dense Layers [0]
+blk\..*\.ffn_down\.weight=q6_0
+blk\..*\.ffn_(gate|up)\.weight=iq5_ks
+
+# Shared Expert Layers [1-46]
+blk\..*\.ffn_down_shexp\.weight=q6_0
+blk\..*\.ffn_(gate|up)_shexp\.weight=iq5_ks
+
+# Routed Experts Layers [1-46]
+blk\..*\.ffn_down_exps\.weight=q6_0
+blk\..*\.ffn_(gate|up)_exps\.weight=iq5_ks
+
+# NextN MTP Layer [46]
+blk\..*\.nextn\.embed_tokens\.weight=iq5_ks
+blk\..*\.nextn\.shared_head_head\.weight=iq5_ks
+blk\..*\.nextn\.eh_proj\.weight=q8_0
+
+# Non-Repeating Layers
+token_embd\.weight=iq4_k
+output\.weight=iq6_k
+"
+
+custom=$(
+  echo "$custom" | grep -v '^#' | \
+  sed -Ez 's:\n+:,:g;s:,$::;s:^,::'
+)
+
+numactl -N 0 -m 0 \
+./build/bin/llama-quantize \
+    --custom-q "$custom" \
+    --imatrix /mnt/raid/models/ubergarm/GLM-4.5-Air-GGUF/imatrix-GLM-4.5-Air-BF16.dat \
+    /mnt/raid/models/ubergarm/GLM-4.5-Air-GGUF/GLM-4.5-Air-128x9.4B-BF16-00001-of-00005.gguf \
+    /mnt/raid/models/ubergarm/GLM-4.5-Air-GGUF/GLM-4.5-Air-IQ5_KS.gguf \
+    IQ5_KS \
+    192
+```
+
+</details>
+
+## IQ4_K 62.910 GiB (4.892 BPW)
+Final estimate: PPL = 4.6273 +/- 0.02839
+
+<details>
+
+<summary>👈 Secret Recipe</summary>
+
+```bash
+#!/usr/bin/env bash
+
+custom="
+# 47 Repeating Layers [0-46]
+# Note: All ffn_down.* layers are not divisible by 256 so have limited quantization options.
+
+# Attention
+blk\..*\.attn_q.*=iq5_ks
+blk\..*\.attn_k.*=q8_0
+blk\..*\.attn_v.*=q8_0
+blk\..*\.attn_output.*=iq5_ks
+
+# First 1 Dense Layers [0]
+blk\..*\.ffn_down\.weight=q6_0
+blk\..*\.ffn_(gate|up)\.weight=iq5_ks
+
+# Shared Expert Layers [1-46]
+blk\..*\.ffn_down_shexp\.weight=q6_0
+blk\..*\.ffn_(gate|up)_shexp\.weight=iq5_ks
+
+# Routed Experts Layers [1-46]
+blk\..*\.ffn_down_exps\.weight=q5_0
+blk\..*\.ffn_(gate|up)_exps\.weight=iq4_k
+
+# NextN MTP Layer [46]
+blk\..*\.nextn\.embed_tokens\.weight=iq5_ks
+blk\..*\.nextn\.shared_head_head\.weight=iq5_ks
+blk\..*\.nextn\.eh_proj\.weight=q8_0
+
+# Non-Repeating Layers
+token_embd\.weight=iq4_k
+output\.weight=iq6_k
+"
+
+custom=$(
+  echo "$custom" | grep -v '^#' | \
+  sed -Ez 's:\n+:,:g;s:,$::;s:^,::'
+)
+
+numactl -N 1 -m 1 \
+./build/bin/llama-quantize \
+    --custom-q "$custom" \
+    --imatrix /mnt/raid/models/ubergarm/GLM-4.5-Air-GGUF/imatrix-GLM-4.5-Air-BF16.dat \
+    /mnt/raid/models/ubergarm/GLM-4.5-Air-GGUF/GLM-4.5-Air-128x9.4B-BF16-00001-of-00005.gguf \
+    /mnt/raid/models/ubergarm/GLM-4.5-Air-GGUF/GLM-4.5-Air-IQ4_K.gguf \
+    IQ4_K \
+    192
+```
+
+</details>
 
 ## IQ4_KSS 54.801 GiB (4.261 BPW)
 Final estimate: PPL = 4.7056 +/- 0.02909