docs: add low-resource note in README; point to config guide; suggest OpenAI embeddings, SkyPilot remote build, and --no-recompute
This commit is contained in:
16
README.md
16
README.md
@@ -71,6 +71,11 @@ source .venv/bin/activate
|
||||
uv pip install leann
|
||||
```
|
||||
|
||||
> Low-resource setup: No local GPU? That's fine.
|
||||
> - Use OpenAI embeddings (`--embedding-mode openai --embedding-model text-embedding-3-small`) for fast indexing without local compute.
|
||||
> - Or run remote builds with SkyPilot using the template at `sky/leann-build.yaml` (see Configuration Guide for details).
|
||||
> - If build/search is slow, consider these two options or disable recomputation with `--no-recompute` (trade storage for speed).
|
||||
|
||||
<details>
|
||||
<summary>
|
||||
<strong>🔧 Build from Source (Recommended for development)</strong>
|
||||
@@ -545,17 +550,6 @@ Options:
|
||||
|
||||
**Backends:** HNSW (default) for most use cases, with optional DiskANN support for billion-scale datasets.
|
||||
|
||||
### Cloud Builds with SkyPilot (Optional)
|
||||
|
||||
If your local machine lacks a GPU or you want faster embedding/index builds, you can run LEANN builds on a cloud GPU VM using SkyPilot. A ready-to-use template is provided at `sky/leann-build.yaml`.
|
||||
|
||||
```bash
|
||||
sky launch -c leann-gpu sky/leann-build.yaml
|
||||
sky exec leann-gpu -- "leann build my-index --docs ~/leann-data --backend hnsw --complexity 64 --graph-degree 32"
|
||||
```
|
||||
|
||||
See the configuration guide section “Running Builds on SkyPilot (Optional)” for details.
|
||||
|
||||
## Benchmarks
|
||||
|
||||
|
||||
|
||||
@@ -261,7 +261,7 @@ The key is finding the right balance for your specific use case. Start small and
|
||||
|
||||
## Deep Dive: Critical Configuration Decisions
|
||||
|
||||
### When to Disable Recomputation
|
||||
### When to Disable Recomputation (Low-resource tip)
|
||||
|
||||
LEANN's recomputation feature provides exact distance calculations but can be disabled for extreme QPS requirements:
|
||||
|
||||
@@ -284,7 +284,7 @@ You can offload embedding generation and index building to a cloud GPU VM using
|
||||
|
||||
### Quick Start
|
||||
|
||||
1) Install SkyPilot by following their docs (`pip install skypilot`, then configure cloud credentials).
|
||||
1) Install SkyPilot by following their docs (`pip install skypilot`), then configure cloud credentials.
|
||||
|
||||
2) Use the provided SkyPilot template:
|
||||
|
||||
|
||||
Reference in New Issue
Block a user