From 6e0baeb8e4b794bff083fdba855dc0783a8ce74d Mon Sep 17 00:00:00 2001 From: KevinMusgrave Date: Mon, 8 Jul 2024 14:24:56 -0400 Subject: [PATCH] update readme --- README.md | 1 + blog/act-mem-2/README.md | 5 +++++ blog/tp/README.md | 5 +++++ 3 files changed, 11 insertions(+) diff --git a/README.md b/README.md index 193fd94..7caa10a 100644 --- a/README.md +++ b/README.md @@ -16,6 +16,7 @@ This repository contains a variety of Determined examples that are not actively | [LLM Finetuning 2](blog/llm-finetuning-2) | Finetuning Mistral-7B on Text-to-SQL using LoRA and DeepSpeed. | | [LLM Finetuning 3](blog/llm-finetuning-3) | Finetuning Gemma-2B using DPO. | | [Python SDK demo](blog/python_sdk_demo) | Example usage of the Determined Python SDK to run and administer experiments. | +| [Tensor Parallelism](blog/tp) | Profiling tensor parallelism in PyTorch. | ## Computer Vision diff --git a/blog/act-mem-2/README.md b/blog/act-mem-2/README.md index 07f1547..9531f4c 100644 --- a/blog/act-mem-2/README.md +++ b/blog/act-mem-2/README.md @@ -9,3 +9,8 @@ memory. - `attn_script.py` shows the cost of activation memory in the attention layer. - Tests of the code are in `test.py`. - See `requirements.txt` for versions the code was built against. + + +## Contributors + +- [Garrett Goon](https://github.com/garrett361) \ No newline at end of file diff --git a/blog/tp/README.md b/blog/tp/README.md index 6ed42c1..79ea909 100644 --- a/blog/tp/README.md +++ b/blog/tp/README.md @@ -6,3 +6,8 @@ Code accompanying the deep-dive [blog post on Tensor Parallelism](https://determ - Matmul profiling code in `matmul_profiling.py` - MLP TP profiling code in `tp_profiling.py` - Tests of the rearranging tensor sums are in `test_dot_product_{local,distributed}.py` + + +## Contributors + +- [Garrett Goon](https://github.com/garrett361) \ No newline at end of file