From afb99b5e82353b90a4a60e87989a44352b8e37cf Mon Sep 17 00:00:00 2001 From: Vincenzo Palazzo Date: Thu, 6 Aug 2026 20:54:03 +0200 Subject: [PATCH] docs: add on-disk sizes for GLM 5.2 model variants The DeepSeek Flash variants already document their sizes and RAM class in download_model.sh, but the GLM 5.2 entries only described the quant layout. Measured on disk after download: glm-antirez-iq2xxs ~211 GB (does not fully fit 128 GB RAM; use --ssd-streaming on single-node) glm-antirez-q2 ~262 GB (already documented) glm-antirez-q4 ~434 GB (already documented) glm-unsloth-q4 ~467 GB (11 shards) Also annotate the README download list with the same sizes so a quick glance is enough to pick a machine class. --- README.md | 8 ++++---- download_model.sh | 3 +++ 2 files changed, 7 insertions(+), 4 deletions(-) diff --git a/README.md b/README.md index d1a5b6b39..b9943747d 100644 --- a/README.md +++ b/README.md @@ -148,10 +148,10 @@ native tooling can be added later. GLM 5.2 support is limited to the GGUF files tested by this branch: ```sh -./download_model.sh glm-unsloth-q4 # Unsloth UD-Q4_K_XL, 11 shards -./download_model.sh glm-antirez-iq2xxs # antirez routed IQ2_XXS single-file GGUF -./download_model.sh glm-antirez-q2 # antirez routed Q2_K single-file GGUF -./download_model.sh glm-antirez-q4 # antirez routed Q4_K single-file GGUF +./download_model.sh glm-unsloth-q4 # Unsloth UD-Q4_K_XL, 11 shards (~467 GB) +./download_model.sh glm-antirez-iq2xxs # antirez routed IQ2_XXS single-file GGUF (~211 GB) +./download_model.sh glm-antirez-q2 # antirez routed Q2_K single-file GGUF (~262 GB) +./download_model.sh glm-antirez-q4 # antirez routed Q4_K single-file GGUF (~434 GB) ``` The supported GLM layout keeps dense/model-control tensors in the existing diff --git a/download_model.sh b/download_model.sh index 01f5b47e0..4e54495e1 100755 --- a/download_model.sh +++ b/download_model.sh @@ -95,6 +95,9 @@ Targets: glm-antirez-iq2xxs GLM 5.2 antirez routed IQ2_XXS GGUF from antirez/GLM-5.2-GGUF. Includes Q2_K block 78 and is intended for reduced-memory testing. + About 211 GB on disk. Does not fully fit in 128 GB RAM — use + --ssd-streaming (and --ssd-streaming-cache-experts) on single-node + 96/128 GB machines; two-node tensor parallelism also works. glm-antirez-q2 GLM 5.2 antirez routed Q2_K GGUF from antirez/GLM-5.2-GGUF.