Universal Assisted Generation: Faster Decoding with Any Assistant Model
•
51
Lewis Tunstall PRO
lewtun
AI & ML interests
LLMs, LLMs, LLMs
Recent Activity
liked
a Space
4 days ago
data-agents/jupyter-agent
liked
a model
5 days ago
answerdotai/ModernBERT-large
Articles
Organizations
lewtun's activity
posted
an
update
9 days ago
Post
6430
We outperform Llama 70B with Llama 3B on hard math by scaling test-time compute 🔥
How? By combining step-wise reward models with tree search algorithms :)
We show that smol models can match or exceed the performance of their much larger siblings when given enough "time to think"
We're open sourcing the full recipe and sharing a detailed blog post.
In our blog post we cover:
📈 Compute-optimal scaling: How we implemented DeepMind's recipe to boost the mathematical capabilities of open models at test-time.
🎄 Diverse Verifier Tree Search (DVTS): An unpublished extension we developed to the verifier-guided tree search technique. This simple yet effective method improves diversity and delivers better performance, particularly at large test-time compute budgets.
🧭 Search and Learn: A lightweight toolkit for implementing search strategies with LLMs and built for speed with vLLM
Here's the links:
- Blog post: HuggingFaceH4/blogpost-scaling-test-time-compute
- Code: https://github.com/huggingface/search-and-learn
Enjoy!
How? By combining step-wise reward models with tree search algorithms :)
We show that smol models can match or exceed the performance of their much larger siblings when given enough "time to think"
We're open sourcing the full recipe and sharing a detailed blog post.
In our blog post we cover:
📈 Compute-optimal scaling: How we implemented DeepMind's recipe to boost the mathematical capabilities of open models at test-time.
🎄 Diverse Verifier Tree Search (DVTS): An unpublished extension we developed to the verifier-guided tree search technique. This simple yet effective method improves diversity and delivers better performance, particularly at large test-time compute budgets.
🧭 Search and Learn: A lightweight toolkit for implementing search strategies with LLMs and built for speed with vLLM
Here's the links:
- Blog post: HuggingFaceH4/blogpost-scaling-test-time-compute
- Code: https://github.com/huggingface/search-and-learn
Enjoy!
- 2 replies
reacted to
julien-c's
post with 🤗❤️🔥
15 days ago
Post
7601
After some heated discussion 🔥, we clarify our intent re. storage limits on the Hub
TL;DR:
- public storage is free, and (unless blatant abuse) unlimited. We do ask that you consider upgrading to PRO and/or Enterprise Hub if possible
- private storage is paid above a significant free tier (1TB if you have a paid account, 100GB otherwise)
docs: https://huggingface.co./docs/hub/storage-limits
We optimize our infrastructure continuously to scale our storage for the coming years of growth in Machine learning, to the benefit of the community 🔥
cc: @reach-vb @pierric @victor and the HF team
TL;DR:
- public storage is free, and (unless blatant abuse) unlimited. We do ask that you consider upgrading to PRO and/or Enterprise Hub if possible
- private storage is paid above a significant free tier (1TB if you have a paid account, 100GB otherwise)
docs: https://huggingface.co./docs/hub/storage-limits
We optimize our infrastructure continuously to scale our storage for the coming years of growth in Machine learning, to the benefit of the community 🔥
cc: @reach-vb @pierric @victor and the HF team
- 28 replies
replied to
dvilasuero's
post
6 months ago
Welcome to the team @dvilasuero and Argilla! It’s been really nice collaborating with you on various projects around LLM alignment and I’m excited to see what we’ll build next together!
reacted to
dvilasuero's
post with 🤝❤️🤗🚀🔥
6 months ago
Post
8073
Today is a huge day in Argilla’s history. We couldn’t be more excited to share this with the community: we’re joining Hugging Face!
We’re embracing a larger mission, becoming part of a brilliant and kind team and a shared vision about the future of AI.
Over the past year, we’ve been collaborating with Hugging Face on countless projects: launching partner of Docker Spaces, empowering the community to clean Alpaca translations into Spanish and other languages, launching argilla/notus-7b-v1 building on Zephyr’s learnings, the Data is Better Together initiative with hundreds of community contributors, or releasing argilla/OpenHermesPreferences, one of the largest open preference tuning datasets
After more than 2,000 Slack messages and over 60 people collaborating for over a year, it already felt like we were part of the same team, pushing in the same direction. After a week of the smoothest transition you can imagine, we’re now the same team.
To those of you who’ve been following us, this won’t be a huge surprise, but it will be a big deal in the coming months. This acquisition means we’ll double down on empowering the community to build and collaborate on high quality datasets, we’ll bring full support for multimodal datasets, and we’ll be in a better place to collaborate with the Open Source AI community. For enterprises, this means that the Enterprise Hub will unlock highly requested features like single sign-on and integration with Inference Endpoints.
As a founder, I am proud of the Argilla team. We're now part of something bigger and a larger team but with the same values, culture, and goals. Grateful to have shared this journey with my beloved co-founders Paco and Amélie.
Finally, huge thanks to the Chief Llama Officer @osanseviero for sparking this and being such a great partner during the acquisition process.
Would love to answer any questions you have so feel free to add them below!
We’re embracing a larger mission, becoming part of a brilliant and kind team and a shared vision about the future of AI.
Over the past year, we’ve been collaborating with Hugging Face on countless projects: launching partner of Docker Spaces, empowering the community to clean Alpaca translations into Spanish and other languages, launching argilla/notus-7b-v1 building on Zephyr’s learnings, the Data is Better Together initiative with hundreds of community contributors, or releasing argilla/OpenHermesPreferences, one of the largest open preference tuning datasets
After more than 2,000 Slack messages and over 60 people collaborating for over a year, it already felt like we were part of the same team, pushing in the same direction. After a week of the smoothest transition you can imagine, we’re now the same team.
To those of you who’ve been following us, this won’t be a huge surprise, but it will be a big deal in the coming months. This acquisition means we’ll double down on empowering the community to build and collaborate on high quality datasets, we’ll bring full support for multimodal datasets, and we’ll be in a better place to collaborate with the Open Source AI community. For enterprises, this means that the Enterprise Hub will unlock highly requested features like single sign-on and integration with Inference Endpoints.
As a founder, I am proud of the Argilla team. We're now part of something bigger and a larger team but with the same values, culture, and goals. Grateful to have shared this journey with my beloved co-founders Paco and Amélie.
Finally, huge thanks to the Chief Llama Officer @osanseviero for sparking this and being such a great partner during the acquisition process.
Would love to answer any questions you have so feel free to add them below!
- 28 replies
replied to
BramVanroy's
post
7 months ago
I am not aware of any public ablations which validate this, but I suspect it has become less important for chat models where one is more interested in the performance via human evaluation instead of academic benchmarks like MMLU (which are OK for selecting base models, but less so for chat/instruct ones)
reacted to
JustinLin610's
post with 🚀🔥
8 months ago
Post
3005
Finally, Qwen1.5-110B is out! With weights and demo!
Blog: https://qwenlm.github.io/blog/qwen1.5-110b/
Demo: Qwen/Qwen1.5-110B-Chat-demo
Base: Qwen/Qwen1.5-110B
Chat: Qwen/Qwen1.5-110B-Chat
This model has some specific features:
* GQA
* 32K token context length
* Multilingual support
We feel good about its performance on benchmarks, including those for base models and chat models, but we still need more of your testing and feedback to help us know its capabilities and limitations!
Additionally, the base model has not learned chatml tokens. Yeah if you use chatml format, you need to be careful about it!
Enjoy and stay tuned for Qwen2!
Blog: https://qwenlm.github.io/blog/qwen1.5-110b/
Demo: Qwen/Qwen1.5-110B-Chat-demo
Base: Qwen/Qwen1.5-110B
Chat: Qwen/Qwen1.5-110B-Chat
This model has some specific features:
* GQA
* 32K token context length
* Multilingual support
We feel good about its performance on benchmarks, including those for base models and chat models, but we still need more of your testing and feedback to help us know its capabilities and limitations!
Additionally, the base model has not learned chatml tokens. Yeah if you use chatml format, you need to be careful about it!
Enjoy and stay tuned for Qwen2!
reacted to
Sentdex's
post with 👍
8 months ago
Post
5764
Benchmarks!
I have lately been diving deep into the main benchmarks we all use to evaluate and compare models.
If you've never actually looked under the hood for how benchmarks work, check out the LM eval harness from EleutherAI: https://github.com/EleutherAI/lm-evaluation-harness
+ check out the benchmark datasets, you can find the ones for the LLM leaderboard on the about tab here: https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard, then click the dataset and actually peak at the data that comprises these benchmarks.
It feels to me like benchmarks only represent a tiny portion of what we actually use and want LLMs for, and I doubt I'm alone in that sentiment.
Beyond this, the actual evaluations of responses from models are extremely strict and often use even rudimentary NLP techniques when, at this point, we have LLMs themselves that are more than capable at evaluating and scoring responses.
It feels like we've made great strides in the quality of LLMs themselves, but almost no change in the quality of how we benchmark.
If you have any ideas for how benchmarks could be a better assessment of an LLM, or know of good research papers that tackle this challenge, please share!
I have lately been diving deep into the main benchmarks we all use to evaluate and compare models.
If you've never actually looked under the hood for how benchmarks work, check out the LM eval harness from EleutherAI: https://github.com/EleutherAI/lm-evaluation-harness
+ check out the benchmark datasets, you can find the ones for the LLM leaderboard on the about tab here: https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard, then click the dataset and actually peak at the data that comprises these benchmarks.
It feels to me like benchmarks only represent a tiny portion of what we actually use and want LLMs for, and I doubt I'm alone in that sentiment.
Beyond this, the actual evaluations of responses from models are extremely strict and often use even rudimentary NLP techniques when, at this point, we have LLMs themselves that are more than capable at evaluating and scoring responses.
It feels like we've made great strides in the quality of LLMs themselves, but almost no change in the quality of how we benchmark.
If you have any ideas for how benchmarks could be a better assessment of an LLM, or know of good research papers that tackle this challenge, please share!
- 3 replies
reacted to
VictorSanh's
post with ❤️🚀🔥
8 months ago
Post
2785
Glad to see Idefics2 making its way into the awesome OpenVLM Leaderboard which ranks VLMs. 🏆
2nd in its category (<10B parameters and open weights)!
While InternLM-XComposer2 uses proprietary data, Idefics2 is built solely using openly available data.
Leaderboard: opencompass/open_vlm_leaderboard
Model: HuggingFaceM4/idefics2-8b
2nd in its category (<10B parameters and open weights)!
While InternLM-XComposer2 uses proprietary data, Idefics2 is built solely using openly available data.
Leaderboard: opencompass/open_vlm_leaderboard
Model: HuggingFaceM4/idefics2-8b
-
- 9 replies
posted
an
update
9 months ago
Post
5028
Introducing Zephyr 141B-A35B 🪁:
HuggingFaceH4/zephyr-orpo-141b-A35b-v0.1
Yesterday, Mistral released their latest base model (via magnet link of course 😅) and the community quickly converted it to transformers format and pushed it to the Hub: mistral-community/Mixtral-8x22B-v0.1
Early evals of this model looked extremely strong, so we teamed up with Argilla and KAIST AI to cook up a Zephyr recipe with a few new alignment techniques that came out recently:
🧑🍳 Align the base model with Odds Ratio Preference Optimisation (ORPO). This novel algorithm developed by @JW17 and @nlee-208 and @j6mes and does not require an SFT step to achieve high performance and is thus much more computationally efficient than methods like DPO and PPO.
🦫 Use a brand new dataset of 7k high-quality, multi-turn preferences that has been developed by our friends at Argilla. To create this dataset, they took the excellent Capybara SFT dataset from @LDJnr LDJnr/Capybara and converted it into a preference dataset by augmenting the final turn with responses from new LLMs that were then ranked by GPT-4.
What we find especially neat about this approach is that training on 7k samples only takes ~1.3h on 4 H100 nodes, yet produces a model that is very strong on chat benchmarks like IFEval and BBH.
Kudos to @alvarobartt @JW17 and @nlee-208 for this very nice and fast-paced collab!
For more details on the paper and dataset, checkout our collection: HuggingFaceH4/zephyr-orpo-6617eba2c5c0e2cc3c151524
HuggingFaceH4/zephyr-orpo-141b-A35b-v0.1
Yesterday, Mistral released their latest base model (via magnet link of course 😅) and the community quickly converted it to transformers format and pushed it to the Hub: mistral-community/Mixtral-8x22B-v0.1
Early evals of this model looked extremely strong, so we teamed up with Argilla and KAIST AI to cook up a Zephyr recipe with a few new alignment techniques that came out recently:
🧑🍳 Align the base model with Odds Ratio Preference Optimisation (ORPO). This novel algorithm developed by @JW17 and @nlee-208 and @j6mes and does not require an SFT step to achieve high performance and is thus much more computationally efficient than methods like DPO and PPO.
🦫 Use a brand new dataset of 7k high-quality, multi-turn preferences that has been developed by our friends at Argilla. To create this dataset, they took the excellent Capybara SFT dataset from @LDJnr LDJnr/Capybara and converted it into a preference dataset by augmenting the final turn with responses from new LLMs that were then ranked by GPT-4.
What we find especially neat about this approach is that training on 7k samples only takes ~1.3h on 4 H100 nodes, yet produces a model that is very strong on chat benchmarks like IFEval and BBH.
Kudos to @alvarobartt @JW17 and @nlee-208 for this very nice and fast-paced collab!
For more details on the paper and dataset, checkout our collection: HuggingFaceH4/zephyr-orpo-6617eba2c5c0e2cc3c151524
reacted to
trisfromgoogle's
post with ❤️🔥
9 months ago
Post
1839
Very excited to share the first two official Gemma variants from Google! Today at Google Cloud Next, we announced cutting-edge models for code and research!
First, google/codegemma-release-66152ac7b683e2667abdee11 - a new set of code-focused Gemma models at 2B and 7B, in both pretrained and instruction-tuned variants. These exhibit outstanding performance on academic benchmarks and (in my experience) real-life usage. Read more in the excellent HuggingFace blog: https://huggingface.co./blog/codegemma
Second, ( google/recurrentgemma-release-66152cbdd2d6619cb1665b7a), which is based on the outstanding Google DeepMind research in Griffin: https://arxiv.org/abs/2402.19427. RecurrentGemma is a research variant that enables higher throughput and vastly improved memory usage. We are excited about new architectures, especially in the lightweight Gemma sizes, where innovations like RecurrentGemma can scale modern AI to many more use cases.
For details on the launches of these models, check out our launch blog -- and please do not hesitate to send us feedback. We are excited to see what you build with CodeGemma and RecurrentGemma!
Huge thanks to the Hugging Face team for helping ensure that these models work flawlessly in the Hugging Face ecosystem at launch!
First, google/codegemma-release-66152ac7b683e2667abdee11 - a new set of code-focused Gemma models at 2B and 7B, in both pretrained and instruction-tuned variants. These exhibit outstanding performance on academic benchmarks and (in my experience) real-life usage. Read more in the excellent HuggingFace blog: https://huggingface.co./blog/codegemma
Second, ( google/recurrentgemma-release-66152cbdd2d6619cb1665b7a), which is based on the outstanding Google DeepMind research in Griffin: https://arxiv.org/abs/2402.19427. RecurrentGemma is a research variant that enables higher throughput and vastly improved memory usage. We are excited about new architectures, especially in the lightweight Gemma sizes, where innovations like RecurrentGemma can scale modern AI to many more use cases.
For details on the launches of these models, check out our launch blog -- and please do not hesitate to send us feedback. We are excited to see what you build with CodeGemma and RecurrentGemma!
Huge thanks to the Hugging Face team for helping ensure that these models work flawlessly in the Hugging Face ecosystem at launch!
-
- 3 replies