ReSAIL: Mitigating Collapse in Iterative Agent Self-Distillation. Qwen3 4B/8B checkpoints for ALFWorld and TextCraft.
-
HuggingJin/resail-qwen3-4b-alfworld-oel-c3
Text Generation • 4B • Updated • 110 -
HuggingJin/resail-qwen3-4b-alfworld-sdpo-c3
Text Generation • 4B • Updated • 114 -
HuggingJin/resail-qwen3-4b-textcraft-oel-c3
Text Generation • 4B • Updated • 113 -
HuggingJin/resail-qwen3-4b-textcraft-sdpo-c3
Text Generation • 4B • Updated • 122