Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
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
9 changes: 9 additions & 0 deletions deepmd/pretrained/registry.py
Original file line number Diff line number Diff line change
Expand Up @@ -6,6 +6,15 @@
)

MODEL_REGISTRY: dict[str, dict[str, Any]] = {
"DPA-3.3-1M": {
"urls": [
"https://huggingface.co/deepmodelingcommunity/DPA-3.3-1M/resolve/main/DPA-3.3-1M.pt?download=true",
"https://hf-mirror.com/deepmodelingcommunity/DPA-3.3-1M/resolve/main/DPA-3.3-1M.pt?download=true",
"https://modelscope.cn/models/DeepModelingCommunity/DPA-3.3-1M/resolve/master/DPA-3.3-1M.pt",
],
"filename": "DPA-3.3-1M.pt",
"sha256": "36fe440c111108d60cda54aa7d3fccac743794de25abef4d49564b9fb896a55b",
},
"DPA-3.2-5M": {
"urls": [
"https://huggingface.co/deepmodelingcommunity/DPA-3.2-5M/resolve/main/DPA-3.2-5M.pt?download=true",
Expand Down
1 change: 1 addition & 0 deletions doc/model/pretrained.md
Original file line number Diff line number Diff line change
Expand Up @@ -17,6 +17,7 @@ You can run `dp pretrained download -h` to see the currently supported model lis

Examples in this release include:

- `DPA-3.3-1M`
- `DPA-3.2-5M`
- `DPA-3.1-3M`
- `DPA3-Omol-Large`
Expand Down
7 changes: 4 additions & 3 deletions skills/deepmd-finetune-dpa3/SKILL.md
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
---
name: deepmd-finetune-dpa3
description: Fine-tune a DPA3 model in DeePMD-kit using the PyTorch backend. Use when the user wants to adapt a pre-trained DPA3 model to a new downstream dataset. Supports fine-tuning from a self-trained DPA3 model (.pt checkpoint), from a multi-task pre-trained model, or from a built-in pretrained model downloaded via `dp pretrained download` (e.g., DPA-3.1-3M, DPA-3.2-5M). Covers single-task and multi-task fine-tuning workflows.
description: Fine-tune a DPA3 model in DeePMD-kit using the PyTorch backend. Use when the user wants to adapt a pre-trained DPA3 model to a new downstream dataset. Supports fine-tuning from a self-trained DPA3 model (.pt checkpoint), from a multi-task pre-trained model, or from a built-in pretrained model downloaded via `dp pretrained download` (e.g., DPA-3.1-3M, DPA-3.2-5M, DPA-3.3-1M). Covers single-task and multi-task fine-tuning workflows.
compatibility: Requires deepmd-kit with PyTorch backend installed. GPU strongly recommended.
license: LGPL-3.0-or-later
metadata:
Expand All @@ -15,7 +15,7 @@ Fine-tune a pre-trained DPA3 model on a downstream dataset. This skill covers th

1. Fine-tuning from a self-trained single-task DPA3 model
1. Fine-tuning from a multi-task pre-trained DPA3 model
1. Fine-tuning from a built-in pretrained model (e.g., DPA-3.1-3M, DPA-3.2-5M) downloaded via `dp pretrained download`
1. Fine-tuning from a built-in pretrained model (e.g., DPA-3.1-3M, DPA-3.2-5M, DPA-3.3-1M) downloaded via `dp pretrained download`
Comment thread
anyangml marked this conversation as resolved.
Resolved

## Quick Start

Expand All @@ -32,7 +32,7 @@ dp --pt train input.json --finetune /path/to/DPA-3.2-5M.pt --use-pretrain-script

1. Determine the fine-tuning scenario:
- Does the user have a self-trained `.pt` model?
- Does the user want to use a built-in pretrained model (DPA-3.1-3M, DPA-3.2-5M, etc.)?
- Does the user want to use a built-in pretrained model (DPA-3.1-3M, DPA-3.2-5M, DPA-3.3-1M, etc.)?
- Is the pre-trained model single-task or multi-task?
1. If using a built-in pretrained model, download it first with `dp pretrained download`.
1. Collect the downstream training data paths and element types.
Expand Down Expand Up @@ -279,6 +279,7 @@ dp pretrained download -h

Currently available models include:

- `DPA-3.3-1M` — 1M parameter DPA3 pretrained model
- `DPA-3.2-5M` — latest large-scale pretrained model
- `DPA-3.1-3M` — 3M parameter DPA3 pretrained model
- `DPA3-Omol-Large` — large organic molecule model
Expand Down
3 changes: 2 additions & 1 deletion skills/deepmd-python-inference/SKILL.md
Original file line number Diff line number Diff line change
Expand Up @@ -58,7 +58,7 @@ dp = DeepPot("graph.pb")
dp = DeepPot("DPA-3.2-5M")
```

Built-in pretrained model names include `DPA-3.2-5M`, `DPA-3.1-3M`, `DPA3-Omol-Large`, etc. DeePMD-kit will automatically download and cache the model on first use.
Built-in pretrained model names include `DPA-3.3-1M`, `DPA-3.2-5M`, `DPA-3.1-3M`, `DPA3-Omol-Large`, etc. DeePMD-kit will automatically download and cache the model on first use.

### Predict Energy, Forces, and Virial

Expand Down Expand Up @@ -263,6 +263,7 @@ print(f"Forces:\n{f[0]}")
To download pretrained models explicitly:

```bash
dp pretrained download DPA-3.3-1M
dp pretrained download DPA-3.2-5M
dp pretrained download DPA-3.1-3M
dp pretrained download DPA-3.2-5M --cache-dir ./models
Expand Down
Loading