From f48c554eb924bbec4ae4ccebb71a31e08a0c3d09 Mon Sep 17 00:00:00 2001 From: Daniel Han Date: Thu, 31 Oct 2024 02:25:49 -0700 Subject: [PATCH] Updated Home (markdown) --- Home.md | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/Home.md b/Home.md index f0f6f3e..fac2ba4 100644 --- a/Home.md +++ b/Home.md @@ -1008,7 +1008,7 @@ We value engineers who are proactive, independent and who ship features and idea **!! On our criteria for being considered for an internship / full time role !!** 1. Debug, solve urgent issues / bugs and make 3 merged PRs for interns / 6 merged PRs for a full time engineer -2. **OR** Create a high quality PR on 1 item below (only 1 needed): +2. **OR** Create a high quality accepted PR on 1 item below (only 1 needed): * Somewhat optimized QLoRA for FLUX / stable diffusion models. Diffusers just added 4bit QLoRA support - make this faster. Provide a Colab notebook tutorial on how to use it. Must fit in 1 x 16GB Tesla T4. * Unoptimized Apple Silicon / Metal support LoRA - MLX, Core ML Tools etc support. * Utilities to export Unsloth finetunes to vLLM, SGLang & Ollama - LoRA adapters only. Provide a standalone serving interface to vLLM, SGLang @@ -1019,6 +1019,8 @@ We value engineers who are proactive, independent and who ship features and idea * Add fully optimized Deepseek finetuning support - investigate [Scatter MoE](https://github.com/shawntan/scattermoe) - confirm loss curves match. * Add full finetuning / pretraining support in Unsloth - Triton kernels for all. * FSDPv2 + QLoRA (maybe via Torch AO) + `torch.compile`. Investigate PyTorch native Pipeline, Sequence & Tensor parallelism. + * Using spot instances to train models with a checkpoint recovery mechanism (like SkyPilot) + * Use binary tensor cores for fast Hamming distances for approximate nearest neighbors **For for details, email me, or ask me questions on Discord! For more information about us, see our:** 1. CUDA / GPU Mode lecture Talk. [Youtube link](https://www.youtube.com/watch?v=hfb_AIhDYnA)