Amazon SageMaker AI 推出训练会话 API 与训练 cookbook
Amazon SageMaker AI 推出 training session APIs,允许开发者在本地 Python 流程中编写训练循环,由服务托管 GPU、模型与 LoRA 加载、分布式执行、故障恢复和检查点。
推荐理由:文章拆解了训练会话 API、GRPO 循环和 Code RL 配方,并给出公开示例与 cookbook 入口,便于评估自定义训练循环的实现方式。
AK_QMUL at ArGuard Shared Tasks: A Single Fine-Tuned JAIS for Arabic Harmful-Prompt Detection
Ali Khairallah · Arkaitz Zubiaga
针对 ArGuard-2026 阿拉伯语有害提示检测共享任务,提交了二元安全/不安全任务(B1)和七类危害领域任务(B2)方案。两者均使用 JAIS-6.7B 作为骨干,采用 4-bit QLoRA 微调及基于最后 token 表示的线性分类头。B1 训练四元分类头以区分方言注册,并用 AraSafe 数据增强,通过 argmax 解码;B2 微调为七元分类器并平均五个随机种子的 softmax 输出。两项提交均获赛道第一,B1 宏 F1 为 0.984,B2 为 0.790。B1 在六种方言上准确率 93.6%–99.8%;B2 中六类 F1 介于 0.73 至 0.98,残余错误主要集中在与骚扰语义重叠的 Bully 类。
尚无解读。
arXiv:2610.36360v1 Announce Type: new Abstract: We present AK_QMUL's submissions to the ArGuard-2026 shared tasks on Arabic harmful-prompt detection: a binary safe/unsafe task (B1) and a seven-way harm-domain task (B2). Both submissions use one backbone -- JAIS-6.7B fine-tuned with 4-bit QLoRA and a linear classification head over its last-token representation. For B1 we train a four-way head that factors the label by dialect register {safe, unsafe} x {MSA, other}, we augment the training fold with MSA prompts from AraSafe, and decode with a plain argmax to score safe/unsafe. For B2 we fine-tune the same backbone as a seven-way classifier and average the softmax outputs of five random seeds. Both submissions ranked first on their track, with macro-F1 0.984 on B1 and 0.790 on B2. We report results by dialect and by class: on B1 the model is accurate across all six dialects (93.6--99.8%) and its predicted unsafe rate tracks the true rate closely; on B2 six of the seven classes reach F1 between 0.73 and 0.98, with the residual error concentrated in the Bully class, which overlaps semantically with harassment and harm-to-others.
Amazon SageMaker AI 推出 training session APIs,允许开发者在本地 Python 流程中编写训练循环,由服务托管 GPU、模型与 LoRA 加载、分布式执行、故障恢复和检查点。
推荐理由:文章拆解了训练会话 API、GRPO 循环和 Code RL 配方,并给出公开示例与 cookbook 入口,便于评估自定义训练循环的实现方式。