Download script now defaults to new DS4.1 Flash checkpoint.

README.md

@@ -20,9 +20,7 @@ The repository also includes tools and data for GGUF, imatrix, quality, and spee
 
 ## Supported hardware
 
-* **Metal**, the primary target, on Macs with 96 GB or more. Smaller machines
-  can use SSD streaming. SSD streaming is also needed in order to run very
-  large models such as full GLM 5.x (not Flash) on 128GB systems.
+* **Metal**, the primary target, on Macs with 64 GB (for Qwen Flash next) and 96 GB or more (for DeepSeek and GLM). Smaller machines can use SSD streaming. SSD streaming is also needed in order to run very large models such as full GLM 5.x (not Flash) on 128GB systems.
 * **NVIDIA CUDA**, the DGX Spark is our main gaol. DwarfStar also supports multi-GPU systems that are not supported by other backends, for instance it can run DeepSeek v4 Flash on Ada Lovelace cards.
 * **ROCm** on Strix Halo systems such as the Framework Desktop.
 

download_model.sh

@@ -9,7 +9,7 @@ QWEN38_REPO="antirez/qwen3.8-flash-next-gguf"
 QWEN38_MMPROJ_REPO="ggml-org/Qwen3.8-Flash-Next-GGUF"
 REPO="antirez/deepseek-v4-gguf"
 DS41_REPO="antirez/deepseek-v4.1-flash-gguf"
-DS41_Q2_FILE="DeepSeek-V4.1-Flash-Q2.gguf"
+DS41_Q2_FILE="DeepSeek-V4.1-Flash-1017-IQ2_XXS-Q2_K-imatrix.gguf"
 DS41_Q4_FILE="DeepSeek-V4.1-Flash-Q4.gguf"
 DS41_VISION_FILE="DeepSeek-V4.1-Flash-Vision.gguf"
 DS4F_Q2_FILE="DeepSeek-V4-Flash-IQ2XXS-w2Q2K-AProjQ8-SExpQ8-OutQ8-chat-v2-imatrix-0731.gguf"