@ondeinference/cli - npm Package Compare versions

Comparing version 0.2.1 to 0.3.0

package.json

 {
     "name": "@ondeinference/cli",
-    "version": "0.2.1",
+    "version": "0.3.0",
     "keywords": [
@@ -35,9 +35,9 @@ "onde",
     "optionalDependencies": {
-        "@ondeinference/cli-darwin-arm64": "0.2.1",
-        "@ondeinference/cli-darwin-x64": "0.2.1",
-        "@ondeinference/cli-linux-arm64": "0.2.1",
-        "@ondeinference/cli-linux-x64": "0.2.1",
-        "@ondeinference/cli-windows-arm64": "0.2.1",
-        "@ondeinference/cli-windows-x64": "0.2.1"
+        "@ondeinference/cli-darwin-arm64": "0.3.0",
+        "@ondeinference/cli-darwin-x64": "0.3.0",
+        "@ondeinference/cli-linux-arm64": "0.3.0",
+        "@ondeinference/cli-linux-x64": "0.3.0",
+        "@ondeinference/cli-windows-arm64": "0.3.0",
+        "@ondeinference/cli-windows-x64": "0.3.0"
     }
 }

README.md

@@ -8,3 +8,3 @@ <p align="center">
 <p align="center">
-  Manage your Onde Inference account from the terminal.
+  A terminal app for your Onde Inference account, plus local model fine-tuning.
 </p>
@@ -27,4 +27,10 @@
 
-The right binary for your platform gets pulled in automatically. Works on macOS (Apple Silicon and Intel), Linux (x64 and arm64), and Windows (x64 and arm64).
+npm installs the right native binary for your platform automatically.
 
+It works on:
+
+- macOS (Apple Silicon and Intel)
+- Linux (x64 and arm64)
+- Windows (x64 and arm64)
+
 ### Other ways to install
@@ -34,10 +40,10 @@
 |---|---|
-| **Homebrew** | `brew install ondeinference/homebrew-tap/onde` |
-| **pip** | `pip install onde-cli` |
-| **uv** | `uv tool install onde-cli` |
-| **Cargo** | `cargo install onde-cli` |
+| Homebrew | `brew install ondeinference/homebrew-tap/onde` |
+| pip | `pip install onde-cli` |
+| uv | `uv tool install onde-cli` |
+| Cargo | `cargo install onde-cli` |
 
 ---
 
-## Usage
+## Run it
 
@@ -48,10 +54,22 @@ ```sh
 
-Opens a TUI. Sign up or sign in from there, no browser needed.
+That opens the terminal UI.
 
-| Key | What it does |
+From there you can:
+
+- sign up or sign in
+- create and manage apps
+- assign models
+- fine-tune supported local models
+- export merged models to GGUF
+
+No browser needed.
+
+## Basic keys
+
+| Key | Action |
 |---|---|
 | `Tab` | Move between fields |
-| `Enter` | Submit / sign out |
-| `Ctrl+L` | Sign in screen |
-| `Ctrl+N` | New account screen |
+| `Enter` | Submit or confirm |
+| `Ctrl+L` | Go to sign in |
+| `Ctrl+N` | Go to create account |
 | `Ctrl+C` | Quit |
@@ -61,6 +79,56 @@
 
-## What's Onde?
+## Fine-tuning
 
-[Onde Inference](https://ondeinference.com) runs LLMs on the user's device. No server round-trips, no data leaving the hardware. It ships as a native SDK for each platform:
+`onde` can fine-tune Qwen2, Qwen2.5, and Qwen3 safetensors models with LoRA.
 
+Training runs locally:
+- Metal on Apple Silicon
+- CPU on other platforms
+
+So yes, no cloud training setup and no Python environment to babysit.
+
+### Supported base models
+
+| Model | Size |
+|---|---|
+| `Qwen/Qwen3-0.6B` | ~1.2 GB |
+| `Qwen/Qwen2.5-1.5B-Instruct` | ~3.0 GB |
+| `Qwen/Qwen3-1.7B` | ~3.4 GB |
+| `Qwen/Qwen3-4B` | ~8.0 GB |
+
+Only safetensors models can be fine-tuned. GGUF models are already quantized, so their weights are not differentiable.
+
+### Training data
+
+Use one JSON object per line. Each object needs a `text` field containing the full conversation in Qwen's chat template.
+
+```jsonl
+{"text": "<|im_start|>user\nWhat is the boiling point of water?<|im_end|>\n<|im_start|>assistant\n100°C at sea level.<|im_end|>"}
+```
+
+### Running a fine-tune
+
+1. Open the Models tab.
+2. Pick a safetensors model with `↑` / `↓`.
+3. Press `f` to open the fine-tune config.
+4. Set your data path, LoRA rank (default `8`), epochs (default `3`), and learning rate (default `0.0001`).
+5. Start training.
+
+A rank-8 adapter for the 0.6B model is about 1.5 MB, so the output stays pretty small.
+
+### After training
+
+- Press `m` to merge the adapter into the base weights.
+- Press `g` to export the merged model to GGUF.
+
+The exported file loads directly in the [Onde SDK](https://ondeinference.com) for on-device inference.
+
+---
+
+## What is Onde?
+
+[Onde Inference](https://ondeinference.com) is for running LLMs on the user's device. No server round-trips, no sending prompts off to somebody else's machine.
+
+It ships as native SDKs for:
+
 <p align="center">
@@ -70,3 +138,3 @@ <a href="https://github.com/ondeinference/onde-swift">Swift</a>&ensp;·&ensp;<a href="https://pub.dev/packages/onde_inference">Flutter</a>&ensp;·&ensp;<a href="https://www.npmjs.com/package/@ondeinference/react-native">React Native</a>&ensp;·&ensp;<a href="https://crates.io/crates/onde">Rust</a>
 
-This CLI handles the account side of things. The SDKs handle inference.
+The CLI is for account management and local fine-tuning. The SDKs are what you ship in your app.
 
@@ -77,3 +145,3 @@ ---
 
-Logs go to `~/.cache/onde/debug.log`.
+Logs are written to `~/.cache/onde/debug.log`.
 
@@ -84,4 +152,4 @@ ## License
 
-<p align="center">
-  <sub>© 2026 <a href="https://ondeinference.com">Onde Inference</a> · <a href="https://apps.apple.com/se/developer/splitfire-ab/id1831430993">Splitfire AB</a></sub>
-</p>
+## Copyright
+
+© 2026 [Onde Inference](https://ondeinference.com) (Splitfire AB).

Fixed alerts

New author

Supply chain risk

A new npm collaborator published a version of the package for the first time. New collaborators are usually benign additions to a project, but do indicate a change to the security surface area of a package.

Improved metrics

Total package byte prevSize

6794

29.14%

Number of lines in readme file

147

86.08%

Number of medium supply chain risk alerts

1

-50%

No dependency changes