Appearance
diffuse-coordinator model pull
Fetch a model: a name from the catalogue, or a repository on a hub.
Synopsis
diffuse-coordinator model pull <REFERENCE> [OPTIONS]Arguments
| argument | required | description |
|---|---|---|
REFERENCE | yes | e.g. HuggingFaceTB/SmolLM2-135M-Instruct |
Options
| flag | value | default | description |
|---|---|---|---|
--config | <CONFIG> | $DIFFUSE_COORDINATOR_CONFIG | Configuration file. Its [admin] section says where to connect |
--endpoint | <ENDPOINT> | $DIFFUSE_COORDINATOR_ENDPOINT | Coordinator endpoint, e.g. https://coordinator.internal:7443 |
--ca-cert | <CA_CERT> | $DIFFUSE_CA_CERT | The deployment CA certificate (PEM) |
--cert | <CERT> | $DIFFUSE_CERT | This process's certificate chain (PEM) |
--key | <KEY> | $DIFFUSE_KEY | This process's private key (PEM) |
--revision | <REVISION> | - | Branch, tag or commit. Resolved to a commit and recorded as one |
--as | <MODEL_KEY> | - | Slug to file it under. Derived from the reference by default |
--quantization | <QUANTIZATION> | - | Which quantisation to take, when a repository publishes several |
--no-precheck | flag | - | Fetch even when no machine here has the memory to serve the result |
Notes
Fetches from the publisher and installs into this deployment's model store. The download happens once, on the coordinator; nodes take only the shards of the layers they are given.
Examples
bash
$ diffuse-coordinator model pull Qwen/Qwen2.5-3B-Instruct --as qwen2.5-3bresolving Qwen/Qwen2.5-3B-Instruct@main
README.md
config.json
model.safetensors
source https://huggingface.co/Qwen/Qwen2.5-3B-Instruct
revision 7ae557604adf67be50417f59c2c2f167def9a775
licence apache-2.0