Skip to content

Updating test formatting in PyTorch FT #352

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
sreeram-11 wants to merge 2 commits into
base: main
Choose a base branch
from
+11 −10
Open

Updating test formatting in PyTorch FT #352

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
e9b069b
Updating test formatting in PyTorch FT
sreeram-11 Jun 6, 2026
f6df5d4
Merge branch 'main' into sreeram/py-ft-stx
sreeram-11 Jun 7, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
21 changes: 11 additions & 10 deletions playbooks/supplemental/pytorch-finetuning/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -32,8 +32,8 @@ This tutorial provides step-by-step examples for fine-tuning a large language mo
### 1. Setup

#### Create a Virtual Environment
<!-- @os:linux -->
<!-- @device:halo_box -->
<!-- @os:linux -->
<!-- @test:id=create-venv timeout=60 -->
```bash
sudo apt update
Expand All @@ -42,37 +42,32 @@ python3 -m venv finetune-venv --system-site-packages
source finetune-venv/bin/activate
```
<!-- @test:end -->
<!-- @device:end -->
<!-- @setup:id=activate-venv command="source finetune-venv/bin/activate" -->
<!-- @os:end -->

<!-- @os:windows -->
<!-- @device:halo_box -->
<!-- @test:id=create-venv timeout=60 -->
```powershell
python -m venv finetune-venv --system-site-packages
finetune-venv\Scripts\activate
```
<!-- @test:end -->
<!-- @device:end -->
<!-- @setup:id=activate-venv command="source finetune-venv\Scripts\activate" -->
<!-- @os:end -->
<!-- @device:end -->


<!-- @os:windows -->
<!-- @device:halo,stx,krk,rx7900xt,rx9070xt -->
<!-- @os:windows -->
<!-- @test:id=create-venv timeout=60 -->
```powershell
python -m venv finetune-venv
finetune-venv\Scripts\activate
```
<!-- @test:end -->
<!-- @device:end -->
<!-- @setup:id=activate-venv command="finetune-venv\Scripts\activate" -->
<!-- @os:end -->

<!-- @os:linux -->
<!-- @device:halo,stx,krk,rx7900xt,rx9070xt -->
<!-- @test:id=create-venv timeout=120 -->
```bash
sudo apt update
Expand All @@ -81,14 +76,13 @@ python3 -m venv finetune-venv
source finetune-venv/bin/activate
```
<!-- @test:end -->
<!-- @device:end -->
<!-- @setup:id=activate-venv command="source finetune-venv/bin/activate" -->
<!-- @os:end -->
<!-- @device:end -->

#### Installing Basic Dependencies
<!-- @require:pytorch -->


#### Additional Dependencies

<!-- @os:linux -->
Expand Down Expand Up @@ -197,6 +191,7 @@ r = subprocess.run([sys.executable, "train_full_finetuning.py"], timeout=600)
sys.exit(r.returncode)
```
<!-- @test:end -->
<!-- @device:end -->
---

## Understanding the Techniques
Expand Down Expand Up @@ -259,7 +254,13 @@ Below is a summary of the available training methods. Each method links to its s
| [`train_full_finetuning.py`](assets/train_full_finetuning.py) | **Full Fine-tuning** | Updates all model parameters. Maximum quality; highest memory and compute usage. | 40GB+ | Maximum quality; research; large VRAM |

<!-- @device:stx,krk -->
<!-- @os:linux -->
> **Note:** Full fine-tuning (`train_full_finetuning.py`) may require more than 64GB of system RAM and may not be feasible on this device. Consider using LoRA or QLoRA instead.
<!-- @os:end -->

<!-- @os:windows -->
> **Note:** Full fine-tuning (`train_full_finetuning.py`) may require more than 64GB of system RAM and may not be feasible on this device. Consider using LoRA instead.
<!-- @os:end -->
<!-- @device:end -->

Simply select your preferred `Training method`, download the corresponding script and execute it using the command keeping your virtual environment activated:
Expand Down
Loading