Joakimpalm-Zen
/

Qwen3-4B-Runner-ToolUse-Q4_K_M

function-calling

reproducible-training

xyntetik-runner

Model card Files Files and versions

Qwen3-4B-Runner-ToolUse-Q4_K_M

4.55 GB

Ctrl+K

Ctrl+K

1 contributor

History: 30 commits

Joakimpalm-Zen's picture

Card: the off-distribution result beside the tool-use result, limit first (R8.13.1)

b2a6a07 verified about 1 hour ago

assets
demo gifs re-recorded with neutral paths (hygiene) 3 months ago
data
Qwen3-4B tool-use LoRA trained directly through the Q4_K_M serving artifact; provenance, evals, and the three-precision adaptation study 3 months ago
evals
Upload evals/logprob-study/logprob-study.json with huggingface_hub 3 months ago
merged
merge study: --merge-lora artifacts — Q4 merge erases the adapter (measured), Q8_0/F16 preserve it; standalone merged-Q8_0 + provenance 3 months ago
study
Qwen3-4B tool-use LoRA trained directly through the Q4_K_M serving artifact; provenance, evals, and the three-precision adaptation study 3 months ago
.gitattributes

1.87 kB
merge study: --merge-lora artifacts — Q4 merge erases the adapter (measured), Q8_0/F16 preserve it; standalone merged-Q8_0 + provenance 3 months ago
Qwen3-4B-ToolUse-LoRA.gguf

66.1 MB
xet

Qwen3-4B tool-use LoRA trained directly through the Q4_K_M serving artifact; provenance, evals, and the three-precision adaptation study 3 months ago
Qwen3-4B-ToolUse-LoRA.gguf.envelope.json

745 Bytes
adapter envelope: measured capability + neutral-corpus drift, dated 3 months ago
Qwen3-4B-ToolUse-LoRA.gguf.train.json

658 Bytes
Qwen3-4B tool-use LoRA trained directly through the Q4_K_M serving artifact; provenance, evals, and the three-precision adaptation study 3 months ago
README.md

23.4 kB
Card: the off-distribution result beside the tool-use result, limit first (R8.13.1) about 1 hour ago