Filter
Exclude
Time range
-
Near
Deploy and manage models flexibly and consistently on #RHEL or #OpenShift. Get supported, fit-for-purpose Granite models and tune them easily with InstructLab. Bring #AI to your enterprise with #RedHat AI. 🔑 red.ht/3Z1RF2D
1
6
484
Deploy and manage models flexibly and consistently on #RHEL or #OpenShift. Get supported, fit-for-purpose Granite models and tune them easily with InstructLab. Bring #AI to your enterprise with #RedHat AI. 🔑 red.ht/3Z1RF2D
1
3
223
Deploy and manage models flexibly and consistently on #RHEL or #OpenShift. Get supported, fit-for-purpose Granite models and tune them easily with InstructLab. Bring #AI to your enterprise with #RedHat AI. 🔑 red.ht/3Z1RF2D
2
6
263
Deploy and manage models flexibly and consistently on #RHEL or #OpenShift. Get supported, fit-for-purpose Granite models and tune them easily with InstructLab. Bring #AI to your enterprise with #RedHat AI. 🔑 red.ht/3Z1RF2D
2
5
263
Deploy and manage models flexibly and consistently on #RHEL or #OpenShift. Get supported, fit-for-purpose Granite models and tune them easily with InstructLab. Bring #AI to your enterprise with #RedHat AI. 🔑 red.ht/47tRxhm
3
385
InstructLab @InstructLab とは redhat.com/ja/topics/ai/what… InstructLab は、プライベートデータを使用してLLMをカスタマイズします。 * OpenAI の GPT * Anthropic の Claude * Meta の Llama * Mistral * IBM の Granite などもあります。
1
1
2
90
#instructlab | DeepWiki deepwiki.com/instructlab/ins… >IBMとRed Hatが共同開発 >OSSのLLMを改良・再学習するためのフレームワーク >人間が自然言語で書いたルールや知識を「構造化されたデータ(taxonomy)」としてモデルに教えられます。 → Human-in-the-Loopで使えるとのこと🤗
来年は、真面目に金融AI(SLM):ヘッジファンドAIを作る。なのでGPUボードがほしい。1から作るの大変だから、まずは出来合いのモデルをfine-tuningから。 fine-tuningのデータセットの作成やトレーニングは"Instructlab"を使う。 github.com/instructlab
1
2
3
1,273
来年は、真面目に金融AI(SLM):ヘッジファンドAIを作る。なのでGPUボードがほしい。1から作るの大変だから、まずは出来合いのモデルをfine-tuningから。 fine-tuningのデータセットの作成やトレーニングは"Instructlab"を使う。 github.com/instructlab
最終的には金融専用のSLM開発(自社のノウハウなどの学習)が差別化要因になると思う。エージェンティックAIもSLMをベースとしたモデルになると思う。これはある意味、特定のAIサービスベンダー脱却で、脱OpenAIであり、脱Geminiでもあり得る。
1
50
10,868
GoogleがGemini API向けにRAGシステム「File Search Tool」を発表! ついにRAGが本命感でてきましたね🔥 読書会『AI時代の価値のつくりかた』では、InstructLabについても学びましたが、企業独自のカスタマイズにはRAGが手っ取り早いのでしょう。 次のグレッグさんのワークショップでもRAG実習があります。 理解を深めるチャンス。 codezine.jp/news/detail/2252… #uxforai
2
145
Deploy and manage models flexibly and consistently on #RHEL or #OpenShift. Get supported, fit-for-purpose Granite models and tune them easily with InstructLab. Bring #AI to your enterprise with #RedHat AI. 🔑 red.ht/47tRxhm
1
8
586
1) buy a Mac Studio or newer full spec Mac Mini 2) install open-notebook, vLLM, InstructLab, and OpenCode 3) draw the rest of the owl (meaning build out a pipeline to structure your data, train the small model, frontend stuff, harden the box, etc) 4) test and iterate
1
4
263
Honestly, if you have never done any sort of tuning, I would start with InstructLab. I’m a bit biased here, but it really is a good way to work with taxonomic data and get it into a domain-focused model. Look at IBM’s Docling-Granite for unstructured stuff.
6
422
📢 Hi all! The #InstructLab Community is evolving to better align with technical changes, with the project components being separated and relocated. Here is the full announcement: blog.instructlab.ai/2025/09/… Thank you for all your participation and contributions in the past year! 🐶

3
4
351
Deploy and manage models flexibly and consistently on #RHEL or #OpenShift. Get supported, fit-for-purpose Granite models and tune them easily with InstructLab. Bring #AI to your enterprise with #RedHat AI. 🔑 red.ht/4m22U4s
3
471
Last week, @IBM hosted me at @USOpen Mixed Doubles Finals match in New York. Just before the match started, I spoke to Dr. @MaryamAshoori, VP Product at @IBM about state of applications in AI era and the role of developers going forward. I will post the full video of our talk next week. Let’s celebrate developer labor this #LaborDay! You all have a great #LongWeekend! @furrier @dvellante @RemyBlaireNews @matteastwood @Craw @rneelmani @nyike @IsForAt @rwang0 @RealStrech #USOpen #USOpen2025 #IBM #HybridCloud #MCP #A2A #GraniteModels #InstructLab @IBMDeveloper @IBMResearch @usta @zkerravala @ArvindKrishna
4
8
23
2,452
28 Jun 2025
Replying to @oscrhong @JeffDean
IBM Research Chief Scientist Ruchir Puri proposes a path to jump over the data wall for chip design Inspired by their work on refactoring large legacy codebases (COBOL) into modern languages Open source models are currently pull-only (no way to contribute into the base model). @InstructLab looks to fill this gap Aiming for OpenEDA LLM by next year
1
143
Last week during @Databricks #DataAISummit, I spoke to Stephanie Valarezo, Director, Product Management, @IBM Data & AI Products. We talked about IBM announcements during the event and I looked at two demos: 1. Data Integration for Agentic Applications 2. IBM Data Product Hub We covered AI based data assistant, data integration from multiple sources, LowCode/NoCode tooling, Access Control, Data Privacy, exposing data to developer, vector databases, data linage, data pipelines, RAG, metadata tagging for enhancing semantic richness, data products, IBM Data Product Hub. All thee services sit under IBM Watsonx AI and IBM Watsonx Orchestrate buckets. For more on IBM announcements during last week at Data AI Summit, please visit: ibm.com/events/databricks-da… #DataAISummit #AgenticApplications #GraniteModels #IBMPartner #InstructLab #DataPipelines #RAG @IBMwatsonx @IBMData @IBMcloud @ibmconsulting @dvellante @furrier @RealStrech @ShellyKramer @SanjMo @ggilbert41 @AkwyZ @Craw @matteastwood @robhof @bjbaumann2014 @IsForAt
1
1
3
384
It’s day 2️⃣ at @DevConf_CZ where we’ve been talking all things open source, now at the conference evening social (with beautiful weather) to recap a big day! Today’s keynote was on Bootable Containers with @Podman_io, plus our @InstructLab workshop, and so much more 🥳
4
11
846