Massimo Roberto Scamarcia's picture

Massimo Roberto Scamarcia PRO

mrs83

·

AI & ML interests

Natural Language Processing, Text Generation, Question Answering, Data Augmentation, Knowledge Transfer, Chain-of-Thought, ResearchOps, MLOps

Recent Activity

updated a model 5 days ago

ethicalabs/Echo-DSRN-114M-v0.1.2-Base

updated a model 5 days ago

ethicalabs/Echo-DSRN-114M-v0.1.2

reacted to qgallouedec's post with 🚀 6 days ago

TRL v1.3 ships day-one training support for Qwen 3.6 🚀 The new Qwen 3.6 family (`Qwen/Qwen3.6-27B`, `Qwen/Qwen3.6-35B-A3B`) reuses the Qwen3.5-MoE architecture but ships a slightly different chat template, so we updated the stack end-to-end: new training template with `{% generation %}` markers, tool-call response schema routing, tiny test models for the VLM matrix. SFT with assistant-only loss works out of the box: ```python from trl import SFTConfig, SFTTrainer trainer = SFTTrainer( model="Qwen/Qwen3.6-27B", args=SFTConfig(assistant_only_loss=True), train_dataset=dataset, ) trainer.train() ``` So does GRPO tool-calling — just hand `tools=[...]` to `GRPOTrainer`. v1.3 also brings a new experimental TPO trainer (Triple Preference Optimization), speculative decoding in `trl vllm-serve` (Qwen3 MTP / Eagle3 drafts), 12 more KTO ↔ DPO alignment PRs (KTO promotion to stable is now in reach), three more `{% generation %}` chat templates (Gemma/Gemma 2, Phi-3, GLM-4-MoE), and a chunky SFT entropy bug fix. Full release notes: https://github.com/huggingface/trl/releases/tag/v1.3.0

View all activity

Organizations

mrs83 's collections 3