Skip to content
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

Add Phi-2 dropout note #890

Merged
merged 1 commit into from
Jan 18, 2024
Merged

Add Phi-2 dropout note #890

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
4 changes: 4 additions & 0 deletions tutorials/download_phi.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,6 +13,9 @@ python scripts/download.py --repo_id microsoft/phi-2 --from_safetensors True
python scripts/convert_hf_checkpoint.py --checkpoint_dir checkpoints/microsoft/phi-2
```

> [!WARNING]
> Phi-2 used [dropout](https://huggingface.co/microsoft/phi-2/blob/cb2f453/config.json#L26) during training which we don't model, so training will not be equal.

Inference the model in instruct mode:

```bash
Expand All @@ -22,6 +25,7 @@ python chat/base.py --checkpoint_dir checkpoints/microsoft/phi-2
>> Prompt: Write a detailed analogy between mathematics and a lighthouse.
>> Reply: Mathematics is like a lighthouse. Mathematics provides a method to guide us through the sometimes chaotic and confusing waters of life. It provides a structured approach to problems which can help us find our way and provide direction. Just as a lighthouse keeps watch over the sea, mathematics can provide us with the tools to try and make sense of the world. Furthermore, just as a lighthouse keeps a watchful eye on the horizon, mathematics can help us reach our goals by showing us the way.
```

> [!NOTE]
> In order to obtain appropriate answers, you may need to tweak the [input prompt](https://github.com/Lightning-AI/lit-gpt/blob/74b8df0c3f07fc31d9d1a49e870a1f7955329ad8/chat/base.py#L359). E.g. we found out that if using `"Instruct:{prompt}\nOutput:\n"` instead of `"Instruct:{prompt}\nOutput:"` the model generates longer answers in some cases.

Expand Down