Israel Olalla (@iolalla) header
Israel Olalla (@iolalla) profile picture

Usually will post about cloud, ML, AI and computing, and yes my opinions are mine and can be wrong, but my data ...

Spain
Joined February 2010
Photo by Israel Olalla (@iolalla)Photo by Israel Olalla (@iolalla)Photo by Israel Olalla (@iolalla)Photo by Israel Olalla (@iolalla)Photo by Israel Olalla (@iolalla)Photo by Israel Olalla (@iolalla)Photo by Israel Olalla (@iolalla)Photo by Israel Olalla (@iolalla)Photo by Israel Olalla (@iolalla)Photo by Israel Olalla (@iolalla)
Israel Olalla retweeted
Check out NotebookLM! Create a notebook, upload one or more sources (e.g. PDFs of research papers, your favorite PhD thesis, a newspaper article, etc) then click on 'Generate' to create a podcast of two voices talking about the content you've uploaded. blog.google/technology/ai/no…
90
363
122
1,977
391,605
Muchas gracias @RiuHoteles y especialmente al hotel de Ocho Rios Jamaica y a su director por su amabilidad y por cuidar de los pequeños detalles. ¡¡¡Seguro que repetiremos!!!
2
1
1
170
the book
9
11
1
142
14,532
Pues ya tenemos disponible "Ayúdame a Escribir" en @googledocs y basado en #gemini ¿Te apuntas a probarlo?
We’re excited to announce that 'Help me write' in @googledocs and @gmail is available in Spanish and Portuguese. Sign up to try Gemini for #GoogleWorkspace today. → goo.gle/4aWkMrT
4
113
Google Cloud and Oracle announced an exciting new strategic partnership that aims to provide our customers with a variety of best-in-class solutions to migrate, modernize, and manage their Oracle-based applications in the cloud. Read the announcement: google.smh.re/45lF
1
1
85
Israel Olalla retweeted
Google Cloud’s new partnership with Oracle provides a path for the world’s largest enterprises to use next generation cloud and AI by combining Oracle Cloud database and Google Cloud services. cloud.google.com/blog/produc…
40
157
15
740
158,598
Israel Olalla retweeted
This is really a 'WOW' paper. 🤯 Claims that MatMul operations can be completely eliminated from LLMs while maintaining strong performance at billion-parameter scales and by utilizing an optimized kernel during inference, their model’s memory consumption can be reduced by more than 10× compared to unoptimized models. 🤯 'Scalable MatMul-free Language Modeling' Concludes that it is possible to create the first scalable MatMul-free LLM that achieves performance on par with state-of-the-art Transformers at billion-parameter scales. 📌 The proposed MatMul-free LLM replaces MatMul operations in dense layers with ternary accumulations using weights constrained to {-1, 0, +1}. This reduces computational cost and memory utilization while preserving network expressiveness. 📌 To remove MatMul from self-attention, the Gated Recurrent Unit (GRU) is optimized to rely solely on element-wise products, creating the MatMul-free Linear GRU (MLGRU) token mixer. The MLGRU simplifies the GRU by removing hidden-state related weights, enabling parallel computation, and replacing remaining weights with ternary matrices. 📌 For MatMul-free channel mixing, the Gated Linear Unit (GLU) is adapted to use BitLinear layers with ternary weights, eliminating expensive MatMuls while maintaining effectiveness in mixing information across channels. 📌 The paper introduces a hardware-efficient fused BitLinear layer that optimizes RMSNorm and BitLinear operations. By fusing these operations and utilizing shared memory, training speed improves by 25.6% and memory consumption reduces by 61% over an unoptimized baseline. 📌 Experimental results show that the MatMul-free LLM achieves competitive performance compared to Transformer++ baselines on downstream tasks, with the performance gap narrowing as model size increases. The scaling law projections suggest MatMul-free LLM can outperform Transformer++ in efficiency and potentially in loss when scaled up. 📌 A custom FPGA accelerator is built to exploit the lightweight operations of the MatMul-free LLM. The accelerator processes billion-parameter scale models at 13W beyond human-readable throughput, demonstrating the potential for brain-like efficiency in future lightweight LLMs.
104
818
270
4,708
2,452,796
Israel Olalla retweeted
Day 1 of the International Conference on Education and #Innovation in #Museums (#ICEIM) in Riyadh! Excited to share our insights on creativity-tech tools for enhancing visitor experiences and learning. Incorporating global expertise from institutions and colleagues! 🌍✨
3
22
5
42
9,925
Tengo 2 entradas en pista para AC/DC de un amigo que no puede ir por problemas médicos. ¿Algún interesado en la sala? Y si me ayudáis a difundirlo, os agradecería.
1
24
1
9
2,632
For Reinforcement Learning to be successful on top of LLMs, it is critical to have a very powerful and accurate reward model. Reward in most language tasks isn't as clearly defined as, say, in chess, where the winning condition is a simple computation. Generative reward models are the most powerful class as they are themselves general. It's great to see how the best LLMs are starting to top this important benchmark! huggingface.co/spaces/allena…
4
32
1
195
31,296
Everything is ready for our largest ever Google event in Spain, Google Cloud Summit Madrid. Looking forward to hosting thousands of customers and partners Today #AI #GoogleCloud
1
3
18
500
WoW, hacer un AV con un LLM suena un exceso, pero...
2
126
Israel Olalla retweeted
The ScaNN vector search library was open-sourced in 2020 to highlight innovations in vector search algorithms, critical for many #ML applications. Today, learn how SOAR introduces redundancy to ScaNN’s vector index to improve vector search efficiency → goo.gle/3vUKkao
11
65
5
291
52,260
Israel Olalla retweeted
Google Maps just got a MASSIVE upgrade. Now, you can use AI in your Google Maps. Here are 6 features you can't miss out:↓
Readers added context they thought people might want to know
These features were released last year. blog.google/products/maps/…
205
1,925
139
11,314
3,419,434
Israel Olalla retweeted
I am really excited to reveal what @GoogleDeepMind's Open Endedness Team has been up to 🚀. We introduce Genie 🧞, a foundation world model trained exclusively from Internet videos that can generate an endless variety of action-controllable 2D worlds given image prompts.
129
531
265
2,388
833,724
Es para mí un honor anunciar que por 14avo año estaré gestionando el programa de becas de verano de Google, GSoC, con un elenco de mentoras y mentores de las comunidades Google. Si alguien tiene interés en nuestro proyecto @_liquidgalaxy la info está en liquidgalaxy.eu/2023/09/goog…
1
10
1
23
1,059
Miquel Moll Riudavets, Caterina Caules Ferrer i @Alex_ao8 expliquen la seva vida d'estudiants STEM #STEMTalksMenorca
3
6
217
Google just killed all search engines using AI! You can now use AI in Google. Here are the 3 new AI features at Google:
57
31
5
178
74,363