Ram20307/qwen3-0.6b-gsm8k-arm-a-seed42
0.6B • Updated • 9
slm-reasoning-experiments
Qwen3-0.6B reasoning-format SFT/DPO experiments on GSM8K: arms A/B/C across seeds, plus raw-DPO and OpenMath SFT+DPO variants.
-
-
Ram20307/qwen3-0.6b-gsm8k-arm-a-seed123
0.6B • Updated • 14 -
Ram20307/qwen3-0.6b-gsm8k-arm-b-seed42
0.6B • Updated • 10 -
Ram20307/qwen3-0.6b-gsm8k-arm-b-seed123
0.6B • Updated • 9 -
Ram20307/qwen3-0.6b-gsm8k-arm-c-seed42
0.6B • Updated • 9 -
Ram20307/qwen3-0.6b-gsm8k-raw-dpo-seed42
0.6B • Updated • 6 -
Ram20307/qwen3-0.6b-gsm8k-openmath-sft-dpo-seed42
0.6B • Updated • 8