Qwen3-1.7B · ScienceWorld · trajectory SFT on Qwen3-32B rollouts — 1 epoch
Student: Qwen/Qwen3-1.7B. Part of a study comparing teacher-trajectory SFT, online multi-turn ROSE and on-policy distillation for
multi-turn agents.
Training
Trajectory SFT: one row per full teacher episode in ChatML, cross-entropy only on each teacher reply and its <|im_end|> (observations, instruction and assistant header are context). Batch 32, AdamW lr 1e-5, weight decay 0, cosine schedule with 10% warmup, fp32 training. Nothing filtered: failed episodes are included. Data: 2059 training task variations played once by Qwen3-32B at temperature 1.0 (teacher success 17.10%). 1 epoch = 64 steps, max length 8192. Stored in bf16.
Evaluation
ScienceWorld test split (AgentGym, 200 task variations), avg of 4 passes, temperature 0.4, 512 tokens/turn, 30 rounds, thinking off, no system turn. Success 13.12%, Avg@1 0.2645. Reference points on the same protocol: base Qwen3-1.7B 0.12%, Qwen3-32B itself 15.62%.
Prompt format
ReAct, multi-turn chat: the environment's AgentGym instruction as the first user turn, a canned assistant acknowledgement, then one user
turn per observation; the model replies Thought:\n...\n\nAction:\n<one action>. Render with Qwen3's chat template and enable_thinking=False.