foxAIAI๊ฐ€ ์›€์ง์ด๊ณ  ๋ฏธ๋ž˜๊ฐ€ ์—…๋ฐ์ดํŠธ๋ฉ๋‹ˆ๋‹ค.
ํ•œ๊ตญ์–ด
ไธญๆ–‡Englishๆ—ฅๆœฌ่ชžํ•œ๊ตญ์–ด

Qwopus3.5-4B-Coder-MTP-GGUF

๐Ÿ’ก What is Qwopus3.5-4B-Coder? ๐Ÿช Qwopus3.5-4B-Coder is a compact coding and agent model built on the Qwen3.5 4B family. It is optimized for local execution, code debugging, structured tool-use behavior, and reasoning-heavy developer workflows. The training recipe follows the Qwopus Coder line: Trace Inversion for learnable reasoning traces, high-quality agent trajectories for tool behavior, and curriculum SFT to preserve formatting stability under longer contexts. ๐Ÿงฉ Structured Debugging Targets bug localization, minimal patch reasoning, and environment-verified code repair behavior. ๐Ÿชถ Agent Trace Alignment Learns from tool-call trajectories that include real feedback loops, not only isolated prompt-response pairs.

๋ฒˆ์—ญ ์ค€๋น„ ์ค‘์ž…๋‹ˆ๋‹ค. ์›๋ฌธ์„ ํ‘œ์‹œํ•ฉ๋‹ˆ๋‹ค.

๐Ÿ’ก What is Qwopus3.5-4B-Coder? ๐Ÿช Qwopus3.5-4B-Coder is a compact coding and agent model built on the Qwen3.5 4B family. It is optimized for local execution, code debugging, structured tool-use behavior, and reasoning-heavy developer workflows. The training recipe follows the Qwopus Coder line: Trace Inversion for learnable reasoning traces, high-quality agent trajectories for tool behavior, and curriculum SFT to preserve formatting stability under longer contexts. ๐Ÿงฉ Structured Debugging Targets bug localization, minimal patch reasoning, and environment-verified code repair behavior. ๐Ÿชถ Agent Trace Alignment Learns from tool-call trajectories that include real feedback loops, not only isolated prompt-response pairs.

์ถœ์ฒ˜

huggingface.co โ†—