evan/nix
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: add opencode package and update models

Browse Source
This commit is contained in:
Evan Reichard
2025-12-20 20:19:19 -05:00
parent c3aa7802ad
commit a33790cfc7
6 changed files with 326 additions and 31 deletions
Download Patch File Download Diff File Expand all files Collapse all files

56
systems/x86_64-linux/lin-va-desktop/default.nix
View File

@@ -91,41 +91,41 @@ in
package = pkgs.reichard.llama-swap;
settings = {
models = {
# https://huggingface.co/mradermacher/gpt-oss-20b-heretic-v2-i1-GGUF/tree/main
# --chat-template-kwargs '{\"reasoning_effort\":\"low\"}'
"gpt-oss-20b-thinking" = {
name = "GPT OSS (20B) - Thinking";
cmd = "${pkgs.reichard.llama-cpp}/bin/llama-server --port \${PORT} -m /mnt/ssd/Models/gpt-oss-20b-heretic-v2.i1-MXFP4_MOE.gguf -c 131072 --temp 1.0 --top-p 1.0 --top-k 40 -dev CUDA0";
};
# https://huggingface.co/unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF/tree/main
"qwen3-coder-30b-instruct" = {
name = "Qwen3 Coder (30B) - Instruct";
cmd = "${pkgs.reichard.llama-cpp}/bin/llama-server --port \${PORT} -m /mnt/ssd/Models/Qwen3-Coder-30B-A3B-Instruct-Q4_K_M.gguf -c 262144 --temp 0.7 --min-p 0.0 --top-p 0.8 --top-k 20 --repeat-penalty 1.05 --cache-type-k q8_0 --cache-type-v q8_0 -ts 70,30";
};
# https://huggingface.co/unsloth/Qwen3-30B-A3B-Instruct-2507-GGUF/tree/main
"qwen3-30b-2507-instruct" = {
name = "Qwen3 2507 (30B) - Instruct";
cmd = "${pkgs.reichard.llama-cpp}/bin/llama-server --port \${PORT} -m /mnt/ssd/Models/Qwen3-30B-A3B-Instruct-2507-Q4_K_M.gguf -c 262144 --temp 0.7 --min-p 0.0 --top-p 0.8 --top-k 20 --repeat-penalty 1.05 --cache-type-k q8_0 --cache-type-v q8_0 -ts 70,30";
};
# https://huggingface.co/unsloth/Qwen3-30B-A3B-Thinking-2507-GGUF/tree/main
"qwen3-30b-2507-thinking" = {
name = "Qwen3 2507 (30B) - Thinking";
cmd = "${pkgs.reichard.llama-cpp}/bin/llama-server --port \${PORT} -m /mnt/ssd/Models/Qwen3-30B-A3B-Thinking-2507-UD-Q4_K_XL.gguf -c 262144 --temp 0.7 --min-p 0.0 --top-p 0.8 --top-k 20 --repeat-penalty 1.05 --cache-type-k q8_0 --cache-type-v q8_0 -ts 70,30";
# https://huggingface.co/unsloth/Devstral-Small-2-24B-Instruct-2512-GGUF/tree/main
"devstral-small-2-instruct" = {
name = "Devstral Small 2 (24B) - Instruct";
cmd = "${pkgs.reichard.llama-cpp}/bin/llama-server --port \${PORT} -m /mnt/ssd/Models/Devstral-Small-2-24B-Instruct-2512-UD-Q4_K_XL.gguf -c 98304 -ctk q8_0 -ctv q8_0 -fit off -dev CUDA0";
};
# https://huggingface.co/unsloth/Qwen3-Next-80B-A3B-Instruct-GGUF/tree/main
"qwen3-next-80b-instruct" = {
name = "Qwen3 Next (80B) - Instruct";
cmd = "${pkgs.reichard.llama-cpp}/bin/llama-server --port \${PORT} -m /mnt/ssd/Models/Qwen3-Next-80B-A3B-Instruct-UD-Q4_K_XL.gguf --ctx-size 262144 --temp 0.7 --min-p 0.0 --top-p 0.8 --top-k 20 --repeat-penalty 1.05 --cache-type-k q8_0 --cache-type-v q8_0";
cmd = "${pkgs.reichard.llama-cpp}/bin/llama-server --port \${PORT} -m /mnt/ssd/Models/Qwen3-Next-80B-A3B-Instruct-UD-Q4_K_XL.gguf -c 131072 --temp 0.7 --min-p 0.0 --top-p 0.8 --top-k 20 --repeat-penalty 1.05 -ctk q8_0 -ctv q8_0 -fit off -ncmoe 15 -ts 77,23";
};
# https://huggingface.co/unsloth/Devstral-Small-2-24B-Instruct-2512-GGUF/tree/main
"devstral-small-2-instruct" = {
name = "Devstral Small 2 (24B) - Instruct";
cmd = "${pkgs.reichard.llama-cpp}/bin/llama-server --port \${PORT} -m /mnt/ssd/Models/Devstral-Small-2-24B-Instruct-2512-UD-Q4_K_XL.gguf -c 98304 -ctk q8_0 -ctv q8_0 -dev CUDA0";
# https://huggingface.co/unsloth/Qwen3-30B-A3B-Instruct-2507-GGUF/tree/main
"qwen3-30b-2507-instruct" = {
name = "Qwen3 2507 (30B) - Instruct";
cmd = "${pkgs.reichard.llama-cpp}/bin/llama-server --port \${PORT} -m /mnt/ssd/Models/Qwen3-30B-A3B-Instruct-2507-Q4_K_M.gguf -c 262144 --temp 0.7 --min-p 0.0 --top-p 0.8 --top-k 20 --repeat-penalty 1.05 -ctk q8_0 -ctv q8_0 -ts 70,30";
};
# https://huggingface.co/unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF/tree/main
"qwen3-coder-30b-instruct" = {
name = "Qwen3 Coder (30B) - Instruct";
cmd = "${pkgs.reichard.llama-cpp}/bin/llama-server --port \${PORT} -m /mnt/ssd/Models/Qwen3-Coder-30B-A3B-Instruct-Q4_K_M.gguf -c 262144 --temp 0.7 --min-p 0.0 --top-p 0.8 --top-k 20 --repeat-penalty 1.05 -ctk q8_0 -ctv q8_0 -ts 70,30";
};
# https://huggingface.co/unsloth/Qwen3-30B-A3B-Thinking-2507-GGUF/tree/main
"qwen3-30b-2507-thinking" = {
name = "Qwen3 2507 (30B) - Thinking";
cmd = "${pkgs.reichard.llama-cpp}/bin/llama-server --port \${PORT} -m /mnt/ssd/Models/Qwen3-30B-A3B-Thinking-2507-UD-Q4_K_XL.gguf -c 262144 --temp 0.7 --min-p 0.0 --top-p 0.8 --top-k 20 --repeat-penalty 1.05 -ctk q8_0 -ctv q8_0 -ts 70,30";
};
# https://huggingface.co/mradermacher/gpt-oss-20b-heretic-v2-i1-GGUF/tree/main
# --chat-template-kwargs '{\"reasoning_effort\":\"low\"}'
"gpt-oss-20b-thinking" = {
name = "GPT OSS (20B) - Thinking";
cmd = "${pkgs.reichard.llama-cpp}/bin/llama-server --port \${PORT} -m /mnt/ssd/Models/gpt-oss-20b-heretic-v2.i1-MXFP4_MOE.gguf -c 131072 --temp 1.0 --top-p 1.0 --top-k 40 -dev CUDA0";
};
# https://huggingface.co/unsloth/Qwen3-VL-8B-Instruct-GGUF/tree/main