1 61 22

Krinal Joshi

krinal

kjdeveloper8

AI & ML interests

NLP, Speech

Recent Activity

upvoted an article 3 days ago

From Chunks to Blocks: Accelerating Uploads and Downloads on the Hub

reacted to AdinaY's post with 👍 3 days ago

InspireMusic 🎵🔥 an open music generation framework by Alibaba FunAudio Lab Model: https://huggingface.co/FunAudioLLM/InspireMusic-1.5B-Long Demo: https://huggingface.co/spaces/FunAudioLLM/InspireMusic ✨ Music, songs, audio - ALL IN ONE ✨ High quality audio: 24kHz & 48kHz sampling rates ✨ Long-Form Generation: enables extended audio creation ✨ Efficient Fine-Tuning: precision (BF16, FP16, FP32) with user-friendly scripts

reacted to lewtun's post with 👍 5 days ago

Introducing OpenR1-Math-220k! https://huggingface.co/datasets/open-r1/OpenR1-Math-220k The community has been busy distilling DeepSeek-R1 from inference providers, but we decided to have a go at doing it ourselves from scratch 💪 What’s new compared to existing reasoning datasets? ♾ Based on https://huggingface.co/datasets/AI-MO/NuminaMath-1.5: we focus on math reasoning traces and generate answers for problems in NuminaMath 1.5, an improved version of the popular NuminaMath-CoT dataset. 🐳 800k R1 reasoning traces: We generate two answers for 400k problems using DeepSeek R1. The filtered dataset contains 220k problems with correct reasoning traces. 📀 512 H100s running locally: Instead of relying on an API, we leverage vLLM and SGLang to run generations locally on our science cluster, generating 180k reasoning traces per day. ⏳ Automated filtering: We apply Math Verify to only retain problems with at least one correct answer. We also leverage Llama3.3-70B-Instruct as a judge to retrieve more correct examples (e.g for cases with malformed answers that can’t be verified with a rules-based parser) 📊 We match the performance of DeepSeek-Distill-Qwen-7B by finetuning Qwen-7B-Math-Instruct on our dataset. 🔎 Read our blog post for all the nitty gritty details: https://huggingface.co/blog/open-r1/update-2

View all activity

Organizations

krinal's activity

upvoted an article 3 days ago

Article

From Chunks to Blocks: Accelerating Uploads and Downloads on the Hub

5 days ago

• 42

upvoted an article 9 days ago

Article

Smol but Mighty: Can Small Models Reason well? 🤔

•

12 days ago

• 7

upvoted 2 articles 11 days ago

Article

Open-source DeepResearch – Freeing our search agents

13 days ago

• 984

Article

G2P Shrinks Speech Models

•

11 days ago

• 23

upvoted an article 12 days ago

Article

Powerful ASR + diarization + speculative decoding with Hugging Face Inference Endpoints

May 1, 2024

• 72

upvoted 4 articles 13 days ago

Article

Visualize and understand GPU memory in PyTorch

Dec 24, 2024

• 186

Article

Deploying Speech-to-Speech on Hugging Face

Oct 22, 2024

• 37

Article

Open-R1: a fully open reproduction of DeepSeek-R1

20 days ago

• 748

Article

How to deploy and fine-tune DeepSeek models on AWS

18 days ago

• 45

upvoted an article 17 days ago

Article

Welcome to Inference Providers on the Hub 🔥

20 days ago

• 361

upvoted an article about 1 month ago

Article

Train 400x faster Static Embedding Models with Sentence Transformers

Jan 15

• 144

upvoted a paper about 1 month ago

Tensor Product Attention Is All You Need

Paper • 2501.06425 • Published Jan 11 • 83

upvoted 2 articles about 1 month ago

Article

Mastering Tensor Dimensions in Transformers

•

Jan 12

• 44

Article

Use Models from the Hugging Face Hub in LM Studio

•

Nov 28, 2024

• 139

upvoted a collection 7 months ago

Llama 3.1

Collection

This collection hosts the transformers and original repos of the Llama 3.1, Llama Guard 3 and Prompt Guard models • 11 items • Updated Dec 6, 2024 • 648

upvoted a paper 10 months ago

FlashSpeech: Efficient Zero-Shot Speech Synthesis

Paper • 2404.14700 • Published Apr 23, 2024 • 31

upvoted an article 10 months ago

Article

Welcome Llama 3 - Meta's new open LLM

Apr 18, 2024

• 283

upvoted 3 papers about 1 year ago