Hugging Face's logo

Social Post Explorers

community
Activity Feed

AI & ML interests

None defined yet.

Recent Activity

chansung  authored a paper about 2 months ago
KaSA: Knowledge-Aware Singular-Value Adaptation of Large Language Models
sayakpaul  authored a paper 2 months ago
A Noise is Worth Diffusion Guidance
dvilasuero  authored a paper 2 months ago
Global MMLU: Understanding and Addressing Cultural and Linguistic Biases in Multilingual Evaluation
View all activity

social-post-explorers's activity

grimjim 
posted an update 8 days ago
view post
Post
1772
A recent merge has provided another interesting result on the current Open LLM leaderboard.
open-llm-leaderboard/open_llm_leaderboard

Combining an o1 reasoning merge with VAGOsolutions's Llama-3.1 SauerkrautLM 8B Instruct model resulted in a lower IFEval, but a higher result in every other benchmark. This result is currently my best Llama 3.1 8B merge result to date.
grimjim/SauerHuatuoSkywork-o1-Llama-3.1-8B
The results suggest that defects in output format and/or output parsing may be limiting benchmark performance of various o1 models.
ariG23498 
posted an update 17 days ago
kadirnar 
posted an update 17 days ago
view post
Post
2709
I created my own AI image and video from scratch using the fal.ai platform 💫

Workflow: Flux Lora Training + Upscale + Kling AI(1.6)
·
ariG23498 
posted an update 20 days ago
thughost 
authored a paper 22 days ago
StephenGenusa 
posted an update 23 days ago
view post
Post
1181
I have a pro account and I am logged in. I have duplicated a space due to the error "You have exceeded your GPU quota", I am showing 0 GPU use, yet I am unable to use it "You have exceeded your GPU quota (60s requested vs. 44s left). Create a free account to get more daily usage quota." "Expert Support" is a pitch for consulting.
·
andrewrreed 
posted an update 29 days ago
view post
Post
2689
🚀 Supercharge your LLM apps with Langfuse on Hugging Face Spaces!

Langfuse brings end-to-end observability and tooling to accelerate your dev workflow from experiments through production

Now available as a Docker Space directly on the HF Hub! 🤗

🔍 Trace everything: monitor LLM calls, retrieval, and agent actions with popular frameworks
1⃣ One-click deployment: on Spaces with persistent storage and integrated OAuth
🛠 Simple Prompt Management: Version, edit, and update without redeployment
✅ Intuitive Evals: Collect user feedback, run model/prompt evaluations, and improve quality
📊 Dataset Creation: Build datasets directly from production data to enhance future performance

Kudos to the Langfuse team for this collab and the awesome, open-first product they’re building! 👏 @marcklingen @Clemo @MJannik

🔗 Space: langfuse/langfuse-template-space
🔗 Docs: https://huggingface.co/docs/hub/spaces-sdks-docker-langfuse
  • 1 reply
·
grimjim 
posted an update 29 days ago
view post
Post
1640
I've arrived at an interesting result on the current Open LLM leaderboard.
open-llm-leaderboard/open_llm_leaderboard
After I narrowed down the filter of models to be between 8-9B parameters, my recent merge of o1 reasoning models achieved the highest MATH eval result of any Llama 3.x 8B model currently on the board, hitting 33.99%, placing 973/2795.
grimjim/HuatuoSkywork-o1-Llama-3.1-8B

Unfortunately, I need more information to evaluate the parent models used in the merge.
The Skywork/Skywork-o1-Open-Llama-3.1-8B model scored 0% on the MATH eval, which I suspect was due to output formatting that was baked too hard into the model, and placed 2168/2795; the merge achieved a significant uplift in every benchmark across the board.
Unfortunately, FreedomIntelligence/HuatuoGPT-o1-8B was not currently benched as of this post, so I am unable to assess relative benchmarks. Nevertheless, it is intriguing that an ostensibly medical o1 model appears to have resulted in a sizable MATH boost.
natolambert 
authored 9 papers about 1 month ago
grimjim 
posted an update about 1 month ago
view post
Post
2768
I'm (finally) releasing a Python script that trims excess weights in Gemma2 full-weight models that bloated by ~1B parameters due to an early mergekit bug.
https://github.com/jim-plus/Gemma2-mergekit-remediation

I'd noticed something was off when merges of Gemma2 9B models ended up having ~10B parameters. The current mergekit package is fine, but there are still bloated models on HF that could stand to be fixed.

The script assumes that it will be run from the same directory as the model weights, and will trim the unnecessary lm_head.weight tensor and corresponding index entry.
  • 2 replies
·
nityan 
posted an update about 1 month ago
view post
Post
1642
#001 | A journey into open-source Hugging Face Models on Azure AI

December is the month for New Year resolutions - and this year I am determined to write more on Hugging Face. I kept putting this off thinking I wanted to have time to craft perfect long-form articles, but then I discovered we can do quick posts. So why wait till January?

I am a PhD, a Polyglot, a Parent, a Visual Storyteller, a Community Builder - and an AI Advocate at Microsoft. However, if I look back on my 25+ years in tech, what I love most is to help people learn by making complex concepts feel more accessible and actionable regardless of your background or expertise. And in 2025, I want to use a #NityaLearnsAI tagline as a way to share my learning journey, explore the vast space of AI tools and technologies, amplify our open-source community and put the fun back in fundamentals. I hope you find it useful and will join me!

My first post is on this Microsoft Ignite theater session delivered in Nov:
https://ignite.microsoft.com/en-US/sessions/THR502?source=sessions It was not recorded but can find the slides here: https://speakerdeck.com/nitya/thr502-journey-into-open-source-hugging-face-models-on-azure-ai - and the illustrated guide attached below summarizes the talk in one big picture.

At the core, this is about my growing interest in **Model Choice** and learning more about not just frontier models but the much larger ecosystem of open-source variants and the community creators who build them. See:

1. Oct / The Future of AI is model choice / https://techcommunity.microsoft.com/blog/aiplatformblog/the-future-of-ai-is-model-choice---from-structured-process-to-seamless-platform/4284091
2. Sep / HF Models Recap / https://techcommunity.microsoft.com/blog/aiplatformblog/new-hugging-face-models-on-azure-ai-phi-3-variants-from
3. Aug / HF Models Recap / https://techcommunity.microsoft.com/blog/aiplatformblog/new-hugging-face-models-on-azure-ai-multilingual-slm-and-biomed--july-2024-updat/4211881
  • 1 reply
·
ehristoforu 
posted an update about 1 month ago
view post
Post
3188
✒️ Ultraset - all-in-one dataset for SFT training in Alpaca format.
fluently-sets/ultraset

❓ Ultraset is a comprehensive dataset for training Large Language Models (LLMs) using the SFT (instruction-based Fine-Tuning) method. This dataset consists of over 785 thousand entries in eight languages, including English, Russian, French, Italian, Spanish, German, Chinese, and Korean.

🤯 Ultraset solves the problem faced by users when selecting an appropriate dataset for LLM training. It combines various types of data required to enhance the model's skills in areas such as text writing and editing, mathematics, coding, biology, medicine, finance, and multilingualism.

🤗 For effective use of the dataset, it is recommended to utilize only the "instruction," "input," and "output" columns and train the model for 1-3 epochs. The dataset does not include DPO or Instruct data, making it suitable for training various types of LLM models.

❇️ Ultraset is an excellent tool to improve your language model's skills in diverse knowledge areas.