Jorge Portilla

65 posts

Jorge Portilla

Jorge Portilla

@jpswayne91

San Martin de Porres, Peru شامل ہوئے Ağustos 2019
219 فالونگ5 فالوورز
Oscar Salas Morales
Oscar Salas Morales@osalasmo·
@jpswayne91 @enzoelguera Solo habilitaron Apis para Perú, es decir desde una ip extranjera no se pueden descargar: te comparto por inbox el codigo en python para descargarlos
Español
3
0
1
449
Enzo Elguera Echegaray
Enzo Elguera Echegaray@enzoelguera·
Se realizó una proyección de resultados considerando el avance parcial del conteo de actas por región. Para ello, se tomó la votación actual de los cuatro primeros candidatos (Keiko, Aliaga, Nieto y Sánchez) en cada región y se estimó su resultado al 100%, manteniendo la misma proporción de votos registrada hasta el momento. Posteriormente, se consolidaron estas proyecciones a nivel nacional, lo que permitió simular un escenario con el total de actas procesadas. El ejercicio muestra que, bajo este supuesto, Sánchez subiría al tercer lugar en el resultado nacional; sin embargo, no alcanzaría a disputar el segundo puesto.
Enzo Elguera Echegaray tweet media
Español
70
49
246
45.2K
Oscar Salas Morales
Oscar Salas Morales@osalasmo·
Muy mal análisis, tienes que ver el avance de computo de cada provincia y la tendencia de cada candidato en esa provincia, lo mismo que en el extranjero y luego proyectas, a estas horas así cerrarían: RLA pierde en Perú pero gana en el global por el voto extranjero, solo le queda a RLA asegurar esos 81K del extrankerp. Es la foto de ahorita
Oscar Salas Morales tweet media
Español
11
17
44
6.7K
skatermemo
skatermemo@Mariosaico87·
@enzoelguera Descargue los datos distritales de la ONPE y genere una extrapolacion en base al porcentaje de actas que faltan contabilizar. El analisis da como que Roberto Sanchez pasa a segunda vuelta
skatermemo tweet media
Español
10
1
40
3.5K
Jorge Portilla ری ٹویٹ کیا
Josefina Miro Quesada Gayoso 🍉
Belmont no solo no tiene plan de gobierno, q a estas alturas a poca gente le importa. Pero además de estafador y racista, es un misógino q no tiene problemas en reconocerse como “pederasta”. Realmente, entre corruptos, mafiosos y abusadores, la valla es tan baja. Qué indignante.
Carlos Viguria@cviguria

WTF

Español
45
416
851
16.9K
Victoria Slocum
Victoria Slocum@victorialslocum·
Retrieval Augmented Generation (RAG) is one of the most popular techniques to leverage large language models for almost any type of use case. In a RAG workflow, the user asks a question, like “What is Weaviate?”, and that query is sent into a vector database to search for related documents or chunks to the sentence. Then, that context, along with the original query, are put together in a prompt for the LLM. The LLM will then output a conversational response to the original question using the provided context. There’s a lot of benefits to RAG systems. Unlike other techniques, like fine-tuning, RAG can be much easier to update with new data or delete old or outdated information without retraining a whole model. They can help reduce hallucinations and errors seen in LLM systems normally, and users can directly see the source information used by the model, allowing increased transparency and trust. Even though there are a lot of tutorials and resources for RAG now, it’s still a big technical undertaking to make a good, advanced RAG application. This is why we’ve build Verba, the golden RAGtriever - we wanted to provide advanced RAG techniques to anyone, even if they didn’t have a technical background, and make it super easy to use and customize. Check out the full explanation in our video: youtube.com/watch?v=swKKRd… GitHub: github.com/weaviate/verba Demo: verba.weaviate.io
YouTube video
YouTube
English
13
93
450
66.6K
nerdai
nerdai@_nerdai_·
Dropping a new PaperCard that summarizes the CRITIC paper recommended by @AndrewYNg here for learning about implementing Reflection with Agents. Going to try and implement this pattern with @LoganMarkewich for @llama_index in the coming weeks!! 😎 🦙 CRITIC: LLMs Can Self-Correct With Tool-Interactive Critiquing by Gou, Zhibin, et al. 2024 ICLR. Highlights: ======== 🌟 CRITIC is a framework / agent reasoning loop that uses external tools in order to verify and subsequently correct the outputs of a black-box LLM 🌟 The analyses conducted in this paper demonstrated the inadequacy of using LLMs for self-verification and self-correction i.e., external tools should be used. Links: ==== 📝 Paper lnkd.in/eEDWyiTv 🦙 LlamaIndex Agents: lnkd.in/eD2GyHVT
nerdai tweet media
Andrew Ng@AndrewYNg

Last week, I described four design patterns for AI agentic workflows that I believe will drive significant progress this year: Reflection, Tool use, Planning and Multi-agent collaboration. Instead of having an LLM generate its final output directly, an agentic workflow prompts the LLM multiple times, giving it opportunities to build step by step to higher-quality output. Here, I'd like to discuss Reflection. For a design pattern that’s relatively quick to implement, I've seen it lead to surprising performance gains. You may have had the experience of prompting ChatGPT/Claude/Gemini, receiving unsatisfactory output, delivering critical feedback to help the LLM improve its response, and then getting a better response. What if you automate the step of delivering critical feedback, so the model automatically criticizes its own output and improves its response? This is the crux of Reflection. Take the task of asking an LLM to write code. We can prompt it to generate the desired code directly to carry out some task X. After that, we can prompt it to reflect on its own output, perhaps as follows: Here’s code intended for task X: [previously generated code] Check the code carefully for correctness, style, and efficiency, and give constructive criticism for how to improve it. Sometimes this causes the LLM to spot problems and come up with constructive suggestions. Next, we can prompt the LLM with context including (i) the previously generated code and (ii) the constructive feedback, and ask it to use the feedback to rewrite the code. This can lead to a better response. Repeating the criticism/rewrite process might yield further improvements. This self-reflection process allows the LLM to spot gaps and improve its output on a variety of tasks including producing code, writing text, and answering questions. And we can go beyond self-reflection by giving the LLM tools that help evaluate its output; for example, running its code through a few unit tests to check whether it generates correct results on test cases or searching the web to double-check text output. Then it can reflect on any errors it found and come up with ideas for improvement. Further, we can implement Reflection using a multi-agent framework. I've found it convenient to create two different agents, one prompted to generate good outputs and the other prompted to give constructive criticism of the first agent's output. The resulting discussion between the two agents leads to improved responses. Reflection is a relatively basic type of agentic workflow, but I've been delighted by how much it improved my applications’ results in a few cases. I hope you will try it in your own work. If you’re interested in learning more about reflection, I recommend these papers: - Self-Refine: Iterative Refinement with Self-Feedback, by Madaan et al. (2023) - Reflexion: Language Agents with Verbal Reinforcement Learning, by Shinn et al. (2023) - CRITIC: Large Language Models Can Self-Correct with Tool-Interactive Critiquing, by Gou et al. (2024) I’ll discuss the other agentic design patterns as well in the future. [Original text: deeplearning.ai/the-batch/issu… ]

English
2
3
12
1.6K
Jerry Liu
Jerry Liu@jerryjliu0·
Inspired by @AndrewYNg’s threads on agents in the past few weeks, I’m excited to share my talk from the @weights_biases conference that outlines how to build a general context-augmented research assistant 🧑‍🔬 Naive RAG is mostly good for simple questions in a single-shot setting. But building a research agent enables the following: ✅ Multi-hop QA ✅ Query understanding and task planning ✅ Interfaces to interact with a broad range of external tools ✅ Reflection to iteratively correct behavior ✅ Memory for personalization We start with the agent ingredients and then compose full agents. Full guides to @llama_index resources linked below 👇 Check it out: docs.google.com/presentation/d…
Jerry Liu tweet media
English
13
153
826
86.1K
Akshay 🚀
Akshay 🚀@akshay_pachaar·
The architecture diagram presented below illustrates some of the key components & how they interact with each other! It will be followed by detailed descriptions & code for each component:
Akshay 🚀 tweet media
English
2
16
207
36.3K
Akshay 🚀
Akshay 🚀@akshay_pachaar·
Let's build a RAG app using MetaAI's Llama-3 (100% local):
English
57
356
2.9K
1.1M
Jorge Portilla ری ٹویٹ کیا
Santiago
Santiago@svpino·
@tunguz Can't install an NVIDIA GPU on this one, so I'm not surprised.
English
3
1
42
15.8K
Jorge Portilla
Jorge Portilla@jpswayne91·
@TatianaAstengoB @ocorpancho @adrianatudelag ¿Por qué sostienes que no tiene real interés en el país? Creo que es bueno que aprendas a dar sustentos, Adriana Tudela es de lo mejor de este congreso, que no te guste su postura política ya es otra cosa.
Español
1
0
15
0
Tatiana Astengo
Tatiana Astengo@TatianaAstengoB·
@ocorpancho @adrianatudelag Se tiene que respetar lo que la mayoría votó. Puede que existan vacíos, pero ya se votó por el NO. Además hay que dudar dependiendo de dónde viene el pedido… y esta señora no tiene un interés real en el país.
Español
35
14
189
0
Adriana Tudela Gutiérrez
Adriana Tudela Gutiérrez@adrianatudelag·
Estoy de acuerdo con elecciones generales adelantadas al 2024. Proceso debe venir acompañado de aprobación de reformas clave: -Bicameralidad -Mandato parlamentario de 2.5 años -Reelección de congresistas -Reforma del sist. electoral -Distritos uninominales
Español
6K
2.5K
13.1K
0
Jorge Portilla ری ٹویٹ کیا
Banco de Crédito BCP
Banco de Crédito BCP@BCPComunica·
Abrimos hilo 🧵 para contarte cómo es que estamos trabajando para mejorar cada día más en el BCP.
Banco de Crédito BCP tweet media
Español
29
3
14
0
Jorge Portilla ری ٹویٹ کیا
Martin Hidalgo
Martin Hidalgo@martinhidalgo·
Ojalá podamos ayudar a encontrar a Maruja. Los que puedan ayudar con un RT, sería de gran ayuda.
Martin Hidalgo tweet media
Español
2
559
320
0
C i r u
C i r u@NoCiruNo·
Mi día mejoró en un 100% con la existencia de Dj Salchicha
Español
444
5.5K
40.8K
0
Pedro Cateriano B
Pedro Cateriano B@PCaterianoB·
Felizmente todavía hay sensatez, en medio de la demagogia, improvisación e irresponsabilidad.
Español
44
207
738
0
Jorge Portilla ری ٹویٹ کیا
Gianfranco Ferrari
Gianfranco Ferrari@gferraridlC·
Ayer varios medios informaron sobre mi inclusión junto a más de 60 personas en la investigación sobre el proyecto Olmos. Comparto esta carta que envié a mi equipo del BCP explicando mi participación.
Gianfranco Ferrari tweet media
Español
15
35
230
0