Skip to content

Qwen3.8-Flash-Next on any consumer hardware: one-click install for Windows / Linux. Strata inference engine, OpenAI/Anthropic API on localhost, optional image input.

C++

setup: DeepSeek family line names IST-DAS for the official 2-bit GGUFs and Unsloth for UD-Q8_K_XL; says who made the files of the chosen version

Niko1221b6cc7e681b3e
diff --git a/setup.py b/setup.py
index 2f11e46a0..ca24e8aa6 100644
--- a/setup.py
+++ b/setup.py
@@ -704,7 +704,7 @@ DSV41_ALL_FILES = {**DSV41_FILES, **DSV41_ISTA_FILES}
DSV41_PROFILES = ROOT / "data" / "dsv41" # the shipped hot-expert profiles (profile-2.00.bin, profile-2.25.bin, + .sha256)
DSV41_CONTEXT = 8192 # the context the engine was checked at (4096 and 8192; 64K is not tried yet)
MULTI_FILE_FAMILIES = {
- "deepseek41": {"title": "DeepSeek-V4.1-Flash", "by": "DeepSeek; GGUFs by Unsloth", "arch": "deepseek41",
+ "deepseek41": {"title": "DeepSeek-V4.1-Flash", "by": "DeepSeek; the 2-bit GGUFs by IST-DAS, UD-Q8_K_XL by Unsloth", "arch": "deepseek41",
"about": "a big model that runs on a PC with a 12 GB card and 64 GB of RAM (the two official 2-bit versions: 257 or "
"274 GB of disk, about 5-7 words a second) or, in Unsloth's 508 GB version, on several graphics cards",
"hf": hf(DSV41_REPO) + "UD-Q8_K_XL/", "tag": "deepseek-v4.1-flash-", "name": "deepseek-v4.1-flash",
@@ -7408,6 +7408,8 @@ def install_dsv41(a, family, gpu, hip, amd, cuda_tk, multi, port, ram, adopted,
model = names[int(ask("Which version?", [str(i) for i in range(1, len(names) + 1)], "1", a.yes)) - 1]
ok(f"size: {model}")
md = fam["models"][model]
+ say(" Its GGUF files: " + ("IST-DAS (huggingface.co/helcig/tmp-oct26-2), with the DSpark drafter and a small metadata file from Unsloth"
+ if md.get("recipe") == "ista" else "Unsloth (huggingface.co/unsloth/DeepSeek-V4.1-Flash-1017-GGUF)"))
ista = md.get("recipe") == "ista"
tag = fam["tag"] + model
cards = dsv41_cards(gpu, all_cards, bool(a.gpu is not None or a.gpus))