Qwen3-0.6B / 1.7B SFT-distilled from Qwen3-32B on Divij/qwen3-32b-mas-traces (planner/executor/verifier). 4 epochs, bf16.
-
STEVENZHANG904/Qwen3-0.6B-planner-sft
Text Generation • 0.6B • Updated • 73 -
STEVENZHANG904/Qwen3-0.6B-executor-sft
Text Generation • 0.6B • Updated • 25 -
STEVENZHANG904/Qwen3-1.7B-executor-sft
Text Generation • 2B • Updated • 24 -
STEVENZHANG904/Qwen3-0.6B-verifier-sft
Text Generation • 0.6B • Updated • 30