Modelpublic

Chimanwakis/qwen_manim_animation_q4_k_m_v5

sourceHugging Faceupdated 1mo agoView on Hugging Face

0likes107downloads

Model Card

Chimanwakis/qwenmanimanimationq4kmv5

GGUF export of a Manim-specialized SFT model. The LoRA adapter was merged into unsloth/Qwen2.5-Coder-3B-Instruct-bnb-4bit before conversion with Unsloth.

Training dataset: primary_maths_manim_qwen_messages_final_6023.jsonl.

Dataset SHA-256: 1a784d85752316042438700de8c83ab5d27d246cb990b09789ac88dd34cc8fbf.

Files

File	Size
`qwen2.5-coder-3b-instruct.Q4_K_M.gguf`	1.80 GiB

Quantization requested: Q4_K_M. File hashes are available in SHA256SUMS.

llama.cpp

bash

llama-cli -m qwen2.5-coder-3b-instruct.Q4_K_M.gguf -cnv -c 4096

Increase -c only when the computer has enough memory.