Fine-Tune Llama 3.1 and Deploy Using NVIDIA NIM Directly From Your Laptop Brev 2024-07-29 | NVIDIA Developer Program: developer.nvidia.com/nimGet Access To The NIM: build.nvidia.com/meta/llama-3_1-8b-instructFollow Along: https://brev.dev/llama3-1-nimIn this tutorial, we walk through getting access to the NVIDIA Developer program to gain access to the cutting edge Llama3.1 8B Instruct Model NIM. We fine-tune Llama 3.1 using a PEFT technique called LoRA using the NVIDIA Nemo Framework. And we deploy the fine-tuned model on an NVIDIA NIM. This video serves as an example for how you might develop an AI application with a fine-tuned LLAMA 3.1 model and deploy it for the fastest production level inference speeds on your own infrastructure.If you have any questions please leave them in the comment section below!
How do LLM Inference Optimizations Work? NVIDIA Coffee Chat Brev 2024-08-08 | How do LLM inference Optimizations Work? Nader (Director of Dev Tech at NVIDIA) had this question so he sat down with the legend, Kyle Kranen to learn more about them. NVIDIA is full of experts on anything AI related, is there something you want to learn about?
Fine-Tune Any LLM, Convert to GGUF, And Deploy Using Ollama Brev 2024-06-24 | Follow Along: https://console.brev.dev/launchable/deploy/now?userID=p2mzt91a8&orgID=ejmrvoj8m&name=Finetune+Llama3+and+Convert+to+Ollama+Modelfile&instance=A10G%40g5.2xlarge&diskStorage=500&cloudID=devplane-brev-1&python=3.10&cuda=12.0.1&file=https%3A%2F%2Fgithub.com%2Fbrevdev%2Fnotebooks%2Fblob%2Fmain%2Fllama3-to-ollama.ipynb&launchableID=env-2iBAoNimbkT4imNoAQXyIOXwrXdJoin Our Discord: discord.com/invite/NVDyv7TUgJRequest Access To the Model: huggingface.co/meta-llama/Meta-Llama-3-8B-InstructIn this guide, we walk through how to fine-tune Llama3-8B using ORPO fine-tuning and deploy it to Ollama to be run on any device (including your own laptop)! Please leave any future guides you would like made below!
How to Run Stable-Diffusion using TensorRT and ComfyUI Brev 2024-06-07 | Follow Along: https://console.brev.dev/launchable/deploy/now?userID=xswf1irzo&orgID=ejmrvoj8m&name=comfyUI-tensorRT-carter&instance=RTX+A6000%40NVIDIA-RTX+A6000-48Gi-PCIe-1x&diskStorage=500&cloudID=akash-brev-wg&baseImage=nvcr.io%2Fnvidia%2Ftensorrt%3A24.05-py3&ports=comfyui%3A8188&file=https%3A%2F%2Fgithub.com%2Fbrevdev%2Fnotebooks%2Fblob%2Fmain%2Ftensorrt-comfyui.ipynb&launchableID=env-2hU1R3hVyeFKQVnrmS8KPUiAUe2Join Our Discord: discord.com/invite/NVDyv7TUgJNVIDIA LinkedIn Post featuring Brev!: linkedin.com/posts/nvidia-ai_tensorrt-rtx-activity-7203849252198895617-diBaBlog Post Explaining TensorRT: https://brev.dev/blog/run-comfyui-with-tensorrtIn this guide, we demonstrate the power of TensorRT, developed by Nvidia, to dramatically increase the inference speeds of stable diffusion using comfyUI . TensorRT automatically detects the hardware (RTX GPU) and creates an engine to have the most efficient inference possible. We are using our platform, brev.dev to package this demo into a launchable so you can follow along!Please leave any future guides you would like made below!
How To FineTune Llama3 Brev 2024-05-27 | Follow Along: https://console.brev.dev/launchable/deploy/now?userID=p2mzt91a8&orgID=jnj0c501d&name=Fine-tune+Llama3+using+Direct+Preference+Optimization&instance=A100%40a100.1x&cloudID=crusoe-brev-wg&python=3.10&cuda=12.0.1&file=https%3A%2F%2Fgithub.com%2Fbrevdev%2Fnotebooks%2Fblob%2Fmain%2Fllama3dpo.ipynb&launchableID=env-2hRHIhATRIw8sCD6QwT11eYn6VfJoin Our Discord: discord.com/invite/NVDyv7TUgJRequest Access To the Model: huggingface.co/meta-llama/Meta-Llama-3-8B-InstructIn this guide, we fine tune the popular open sourced model, Llama3-8B using the powerful finetuning method DPO (Direct Preference Optimization). You can do this all from your laptop! Try it out and let us know what you thinkPlease leave any future guides you would like made below!
How To Fine-tune LLaVA Model (From Your Laptop!) Brev 2024-05-01 | Follow Along: https://console.brev.dev/launchable/deploy/now?userID=p2mzt91a8&orgID=jnj0c501d&launchableID=env-2hpxJ6HArVk5jzOYgJmFDJfvNmH&instance=A10G%40g5.12xlarge&diskStorage=300&cloudID=devplane-brev-1&python=3.10&cuda=12.2.2&file=https%3A%2F%2Fgithub.com%2Fbrevdev%2Fnotebooks%2Fblob%2Fmain%2Fllava-finetune.ipynb&name=Fine-tune+and+deploy+multimodal+LLaVA-1.5Join Our Discord: discord.com/invite/NVDyv7TUgJIn this guide, we fine tune the popular open sourced model, LLaVA (Large Language-and-Vision Assistant) on a dataset to be used in a visual classification application. You can perform the fine tuning yourself, regardless your level of experience, or the level of compute you have access to.Please leave any future guides you would like made below!
How to Fine-tune Mixtral 8x7B MoE on Your Own Dataset Brev 2024-02-01 | In this video, we show you how to fine-tune Mixtral, Mistral's 8x7B MoE (Mixture of Experts) model, on your own dataset.You'll be directed to another video where we fine-tune Mistral 7B (standard Mistral) on your own dataset, but you'll be using the notebook here:github.com/brevdev/notebooks/blob/main/mixtral-finetune-own-data.ipynbMy explanation on how QLoRA works: https://brev.dev/blog/how-qlora-worksFine-tune Mixtral MoE on a HuggingFace Dataset: youtu.be/zbKz4g100SQMore AI/ML notebooks: github.com/brevdev/notebooks Join the Discord: discord.gg/NVDyv7TUgJ Connect with me on π: https://x.com/HarperSCarroll
Chat with your PDFs with Open Source OCR (Optical Character Recognition) & Mistral Lite Brev 2024-01-27 | Hi there! Harper Carroll from Brev.dev here. In this tutorial, we go through a pre-made Jupyter Notebook to run OCR (Optical Character Recognition) on our uploaded PDFs to extract the text, and then we use Amazon's MistralLite to ask questions about those PDFs. Amazon's MistralLite is its fine-tuned version of Mistral 7B, which allows for context lengths of up to 32K tokens... this means we can fit more data into the model's prompt i.e. memory (with some lossiness as the context length is more utilized). Notebook: github.com/brevdev/notebooks/blob/main/ocr-pdf-analysis.ipynbMore AI/ML notebooks: github.com/brevdev/notebooks Join the Discord: discord.gg/NVDyv7TUgJ Connect with me on π: https://x.com/HarperSCarrollFind me on Reels: instagram.com/harpercarrollai
Generate AI Images with the Stable Diffusion Web UI AUTOMATIC1111 - No Code Tutorial Brev 2024-01-23 | Hi! Harper Carroll from Brev.dev here. In this video, we walk through how to run the most popular open source Stable Diffusion Web UI, AUTOMATIC1111. No coding of your own needed. Just plug & play.Notebook: github.com/brevdev/notebooks/blob/main/automatic1111-stable-diffusion-ui.ipynbOther AI/ML Notebooks: github.com/brevdev/notebooksFind me on π: https://x.com/HarperSCarroll/highlightsJoin our Discord: discord.com/invite/NVDyv7TUgJ
How to Fine-Tune Llama 2 (on your own data or on a HF dataset) - Video Walkthrough Brev 2024-01-19 | Fine-tune on your own dataset notebook: github.com/brevdev/notebooks/blob/main/llama2-finetune-own-data.ipynbVideo: youtube.com/watch?v=kmkcNVvEz-kFine-tune on a HuggingFace dataset notebook: github.com/brevdev/notebooks/blob/main/llama2-finetune.ipynbVideo: youtube.com/watch?v=zbKz4g100SQOther AI/ML Notebooks: github.com/brevdev/notebooksFind me on π: https://x.com/HarperSCarroll/highlightsJoin our Discord: discord.com/invite/NVDyv7TUgJ
How to Fine-tune Phi-2 - Video Walkthrough Brev 2024-01-10 | Fine-tune on your own dataset notebook: github.com/brevdev/notebooks/blob/main/phi2-finetune-own-data.ipynbVideo: youtube.com/watch?v=kmkcNVvEz-kFine-tune on a HuggingFace dataset notebook: github.com/brevdev/notebooks/blob/main/phi2-finetune.ipynbVideo: youtube.com/watch?v=zbKz4g100SQOther AI/ML Notebooks: github.com/brevdev/notebooksFind me on π: https://x.com/HarperSCarroll/highlightsJoin our Discord: discord.com/invite/NVDyv7TUgJ
Fine-Tune Mixtral 8x7B (Mistrals Mixture of Experts MoE) Model - Walkthrough Guide Brev 2023-12-21 | Hi! Harper Carroll from Brev.dev here. In this tutorial video, I walk you through how to fine-tune Mixtral, Mistralβs 8x7B Mixture of Experts (MoE) model, which outperforms Llama 2 70B on most tested benchmarks.Thanks everyone for 1K subscribers! So glad these are helpful. You rock. Notebook: github.com/brevdev/notebooks/blob/main/mixtral-finetune.ipynbMy explanation on how QLoRA works: https://brev.dev/blog/how-qlora-worksNotebook to fine-tune on own dataset: github.com/brevdev/notebooks/blob/main/mistral-finetune-own-data.ipynb More AI/ML notebooks: github.com/brevdev/notebooks Join the Discord: discord.gg/NVDyv7TUgJ Connect with me on π: https://x.com/HarperSCarroll
OpenAIs Q*?: Reinforcement Learning, Model-Based vs. Model-Free Methods, and Q-Learning Brev 2023-11-28 | In this Brev.dev Concepts video, Harper Carroll (Head of AI/ML) covers the basics of reinforcement learning, exploration and exploitation, model-based vs. model-free methods, Q-learning, Q*, and temporal difference learning. It is accessible to those of all backgrounds, and includes a little math for those interested. Find me on π: twitter.com/HarperSCarrollJoin our community on Discord: discord.gg/DndwhY6cjfAI/ML Tutorial Notebooks: github.com/brevdev/notebooksIntro: (0:00)Reinforcement Learning: (1:10)Exploration & Exploitation: (2:00)Model-Based Methods: (3:36)Model-Free Methods: (4:26)Temporal Difference Learning (estimating Q): (4:36)Q-Learning: (6:16)Q* at OpenAI?: (7:46)Conclusion: (8:24)
How to Fine-Tune Mistral 7B on Your Own Data Brev 2023-10-24 | Hi! Harper Carroll from Brev.dev here. In this tutorial video, I walk you through how to fine-tune Mistral 7B, which outperforms Llama 2 13B on all tested benchmarks, on your own data.... like how I do with my journal entries from over the years, teen angst and all.Notebook: github.com/brevdev/notebooks/blob/main/mistral-finetune-own-data.ipynbNotebook to fine-tune on a HF dataset: github.com/brevdev/notebooks/blob/main/mistral-finetune.ipynbMy explanation on how QLoRA works: https://brev.dev/blog/how-qlora-worksJoin the Discord: discord.com/invite/NVDyv7TUgJFind me on π: https://x.com/HarperSCarroll
How to easily use LoRAs on the Stable Diffusion AUTOMATIC1111 UI Brev 2023-07-25 | Want to create Stable Diffusion images that are as stunning as Midjourney's?Let's use LoRAs (low rank adaptation models) to create stunning Stable Diffusion models in our styles of choice!Link to the template & pre-set GPU: github.com/brevdev/AUTOMATIC1111-civitai-notebook(NVIDIA A10G - $1.21 per hour running; you can select a different machine in "Hardware" if you'd like to downgrade).Link to the LoRAs: civitai.comEasy remote GPU/CPU provisioner: https://brev.devπ€π€π€π€HAPPY CREATING π€π€π€π€
Setup and Run Dreambooth in 5 Minutes for free Brev 2022-11-03 | In this video we show you how to use Brev to run Dreambooth. Brev is a cloud dev environment tool that gives you 30 minutes of free GPU time!Link to get started: https://console.brev.dev/environment/new?setupRepo=https://github.com/brevdev/dreambooth&repo=https://github.com/brevdev/dreambooth&setupPath=.brev/setup.sh&instance=g5.2xlarge Text-guide: https://brev.dev/docs/guides/dreambooth0:00 Intro0:29 Creating an Environment1:48 Opening an Environment2:50 Uploading training data4:06 Running the training script4:48 Results
Build a fully deployed React app with Brev Brev 2021-04-26 | Build a a todo app with fully deployed backend and database.Follow along by cloning this repo: github.com/brevdev/todo and creating a Brev account here: https://app.brev.dev/alphasignup
