From doctest to runnable Markdown
•
3
AI & ML interests
The AI community building the future.
Recent Activity
View all activity
Papers
Qualixar OS: A Universal Operating System for AI Agent Orchestration
FineVision: Open Data Is All You Need
Articles
nielsr
updated a
Space about 6 hours ago
lightx2v/Wan2.1-I2V-14B-720P-StepDistill-CfgDistill-Lightx2v
#5192 opened 6 months ago
by 
Grtfr134567
alvarobartt
updated a
dataset about 21 hours ago
sayakpaul
updated a
dataset 2 days ago
Upload feetech-reset-offset.png
1
#604 opened 2 days ago
by 
maximellerbach
[FEEDBACK] Daily Papers
🔥❤️ 21
193
#32 opened almost 2 years ago
by 
kramp
Unordered questions and remarks
4
#1 opened 2 months ago
by 
qgallouedec
sergiopaniego
posted an update 3 days ago
Post
183
Great experience yesterday at PyTorch Conf Europe in Paris 🇫🇷
We (w/ @kashif ) talked about training LLMs through interaction, using trajectories across games, browsers, or simulators
Room was packed, a clear sign of interest in where RL post-training is heading.
sharing the slides! 🤓
https://drive.google.com/file/d/16k7YRnf5EJEo0XjXGlRJ_hVeLoFWKyNP/view?usp=sharing
We (w/ @kashif ) talked about training LLMs through interaction, using trajectories across games, browsers, or simulators
Room was packed, a clear sign of interest in where RL post-training is heading.
sharing the slides! 🤓
https://drive.google.com/file/d/16k7YRnf5EJEo0XjXGlRJ_hVeLoFWKyNP/view?usp=sharing
Post
252
🌐 I've just published Sentence Transformers v5.4 to make the project fully multimodal for embeddings and reranking. The release also includes a modular CrossEncoder, and automatic Flash Attention 2 input flattening. Details:
You can now use SentenceTransformer and CrossEncoder with text, images, audio, and video, with the same familiar API. That means you can compute embeddings for an image and a text query using model.encode(), compare them with model.similarity(), and it just works. Models like Qwen3-VL-Embedding-2B and jinaai/jina-reranker-m0 are supported out of the box.
Beyond multimodal, I also fully modularized the CrossEncoder class. It's now a torch.nn.Sequential of composable modules, just like SentenceTransformer has been. This unlocked support for generative rerankers (CausalLM-based models like mxbai-rerank-v2 and the Qwen3 rerankers) via a new LogitScore module, which wasn't possible before without custom code.
Also, Flash Attention 2 now automatically skips padding for text-only inputs. If your batch has a mix of short and long texts, this gives you a nice speedup and lower VRAM usage for free.
I wrote a blog post walking through the multimodal features with practical examples. Check it out if you want to get started, or just point your Agent to the URL: https://huggingface.co/blog/multimodal-sentence-transformers
This release has set up the groundwork for more easily introducing late-interaction models (both text-only and multimodal) into Sentence Transformers in the next major release. I'm looking forward to it!
You can now use SentenceTransformer and CrossEncoder with text, images, audio, and video, with the same familiar API. That means you can compute embeddings for an image and a text query using model.encode(), compare them with model.similarity(), and it just works. Models like Qwen3-VL-Embedding-2B and jinaai/jina-reranker-m0 are supported out of the box.
Beyond multimodal, I also fully modularized the CrossEncoder class. It's now a torch.nn.Sequential of composable modules, just like SentenceTransformer has been. This unlocked support for generative rerankers (CausalLM-based models like mxbai-rerank-v2 and the Qwen3 rerankers) via a new LogitScore module, which wasn't possible before without custom code.
Also, Flash Attention 2 now automatically skips padding for text-only inputs. If your batch has a mix of short and long texts, this gives you a nice speedup and lower VRAM usage for free.
I wrote a blog post walking through the multimodal features with practical examples. Check it out if you want to get started, or just point your Agent to the URL: https://huggingface.co/blog/multimodal-sentence-transformers
This release has set up the groundwork for more easily introducing late-interaction models (both text-only and multimodal) into Sentence Transformers in the next major release. I'm looking forward to it!
akhaliq
submitted a
paper to Daily Papers 9 days ago
sergiopaniego
posted an update 10 days ago
Post
2699
Gemma 4 💎 is here and it’s strong!
to celebrate, we’re rolling out in TRL:
> support for multimodal tool responses for environments (OpenEnv)
> an example to train it in CARLA for autonomous driving with image-based tool calls
go check it out 🏎️🏎️
blog: https://huggingface.co/blog/gemma4
script: https://github.com/huggingface/trl/blob/main/examples/scripts/openenv/carla_vlm_gemma.py
to celebrate, we’re rolling out in TRL:
> support for multimodal tool responses for environments (OpenEnv)
> an example to train it in CARLA for autonomous driving with image-based tool calls
go check it out 🏎️🏎️
blog: https://huggingface.co/blog/gemma4
script: https://github.com/huggingface/trl/blob/main/examples/scripts/openenv/carla_vlm_gemma.py
sergiopaniego
posted an update 12 days ago
Post
1954
TRL is officially an adult 🥳
excited to announce TRL v1.0❗️
head to the blog to see how we got here and what’s next for this post-training library, designed to keep pace with the field
https://huggingface.co/blog/trl-v1
excited to announce TRL v1.0❗️
head to the blog to see how we got here and what’s next for this post-training library, designed to keep pace with the field
https://huggingface.co/blog/trl-v1
- 2 replies
AdinaY
submitted a
paper to Daily Papers 12 days ago
akhaliq
submitted a
paper to Daily Papers 16 days ago
akhaliq
submitted a
paper to Daily Papers 25 days ago
akhaliq
submitted a
paper to Daily Papers 28 days ago
sergiopaniego
posted an update about 1 month ago
Post
742
ICYMI, great blog by @kashif and @stas on Ulysses Sequence Parallelism: train with million-token contexts
on 4×H100s: 12x longer sequences, 3.7x throughput
learn how to integrate it with Accelerate, Transformers, and TRL ⤵️
https://huggingface.co/blog/ulysses-sp
on 4×H100s: 12x longer sequences, 3.7x throughput
learn how to integrate it with Accelerate, Transformers, and TRL ⤵️
https://huggingface.co/blog/ulysses-sp
AdinaY
submitted a
paper to Daily Papers about 1 month ago
sergiopaniego
posted an update about 1 month ago
Post
426
We just released a big blog surveying 16 OSS frameworks for async RL training of LLMs!
We're building a new async GRPO trainer for TRL and as first step, we needed to understand how the ecosystem solves this problem today.
The problem: in synchronous RL training, generation dominates wall-clock time. 32K-token rollouts on a 32B model take hours while training GPUs sit completely idle. With reasoning models and agentic RL making rollouts longer and more variable, this only gets worse.
The ecosystem converged on the same fix: separate inference + training onto different GPU pools, rollout buffer, and async weight sync.
We compared 16 frameworks across 7 axes: orchestration, buffer design, weight sync, staleness management, partial rollouts, LoRA, and MoE support.
This survey is step one. The async GRPO trainer for TRL is next!
https://huggingface.co/blog/async-rl-training-landscape
We're building a new async GRPO trainer for TRL and as first step, we needed to understand how the ecosystem solves this problem today.
The problem: in synchronous RL training, generation dominates wall-clock time. 32K-token rollouts on a 32B model take hours while training GPUs sit completely idle. With reasoning models and agentic RL making rollouts longer and more variable, this only gets worse.
The ecosystem converged on the same fix: separate inference + training onto different GPU pools, rollout buffer, and async weight sync.
We compared 16 frameworks across 7 axes: orchestration, buffer design, weight sync, staleness management, partial rollouts, LoRA, and MoE support.
This survey is step one. The async GRPO trainer for TRL is next!
https://huggingface.co/blog/async-rl-training-landscape