ceos/unsloth
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Updated Home (markdown)

Daniel Han 2025-03-07 07:15:38 -08:00
commit 231c3eaaf7
1 changed files with 1 additions and 36 deletions
Download patch file Download diff file Expand all files Collapse all files

37
Home.md

@ -997,39 +997,4 @@ services:
build:
context: .
dockerfile: Dockerfile
```
### -We are hiring
***
[Updated 10th November 2024] Want to work on cool Triton kernels, optimizations and maths algorithms to make LLMs and AI more accessible? Come join us! We currently have over [2.5 million](https://huggingface.co/unsloth) monthly Hugging Face model downloads and collaborate with Meta, Google, Hugging Face on open models. We fixed dozens of bugs in [Gemma](https://x.com/danielhanchen/status/1765446273661075609), [Llama & Phi](https://unsloth.ai/blog/phi3), [Mistral](https://unsloth.ai/blog/mistral-nemo), helped fix a [gradient accumulation bug](https://x.com/danielhanchen/status/1846235913443262891), showed how [gradient checkpointing](https://unsloth.ai/blog/long-context) can be improved to reduce VRAM and more!
We value engineers who are proactive, independent and who ship features and ideas quickly - if stuff breaks, that's fine with us! Internships are 3 monthly renewable roles ($100K - $120K USD pa) SF focused or remote. Full time roles ($150K - $250K USD pa with equity) SF only. We're [Y Combinator S24](https://www.ycombinator.com/companies/unsloth-ai) alumni & backed by Github!
**!! 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 accepted PR on 1 item below. 📚=Software Eng 🔢=Kernels 🛠️=Infra:
* 📚Unoptimized or optimized (extra points) 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
* 🔢Add float8 + QLoRA finetuning support via Torch AO into Unsloth.
* 🔢Bitsandbytes 4bit QLoRA dequant Triton kernel - must be faster than CUDA version.
* 📚Add TPU (maybe JAX?) & AMD support into Unsloth. TPU - Colab & Kaggle notebooks. AMD - Runpod equivalent.
* 🔢Use torch AO and add MXFP4 support in preparation for Blackwell. Show experiments on loss curves matching.
* 🔢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)
* 🛠️Modal but using spot instances
* 🔢Use binary tensor cores for fast Hamming distances. Show this works in approximate nearest neighbors
* 🔢Make `torch.compile` work on gradient checkpointing with compiled autograd and removing `torch._dynamo.disable`
* 🔢Port Flex Attention to Unsloth for all models (Llama, Mistral, Gemma etc). Must be torch compilable and faster than naive SDPA.
**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)
2. Low Level Technicals of LLMs. [Youtube link](https://www.youtube.com/watch?v=pRM_P6UfdIc)
3. Fixing bugs in Llama, Mistral, Gemma. [Youtube link](https://www.youtube.com/watch?v=TKmfBnW0mQA)
4. PyTorch Conference Mini Talk. [Youtube link](https://www.youtube.com/watch?v=PdtKkc5jB4g)
5. PyTorch Engineers Meeting Talk. [Youtube link](https://www.youtube.com/watch?v=MQwryfkydc0)
6. Hugging Face Collab Blog. [Blog link](https://huggingface.co/blog/unsloth-trl)
```