mirror of
https://github.com/ml-explore/mlx-examples.git
synced 2025-06-24 17:31:18 +08:00
Merge pull request #45 from bbelescot/clarify-llama-readme-instructions
📝 Clarify python command for llama example
This commit is contained in:
commit
331690491f
@ -23,7 +23,7 @@ Alternatively, you can also download a select converted checkpoints from the [ml
|
||||
Convert the weights with:
|
||||
|
||||
```
|
||||
python convert.py <path_to_torch_weights> mlx_llama_weights.npz
|
||||
python convert.py <path_to_torch_weights> <path_to_mlx_llama_weights.npz>
|
||||
```
|
||||
|
||||
### Run
|
||||
@ -32,7 +32,7 @@ Once you've converted the weights to MLX format, you can interact with the
|
||||
LLaMA model:
|
||||
|
||||
```
|
||||
python llama.py mlx_llama_weights.npz <path_to_tokenizer.model> "hello"
|
||||
python llama.py <path_to_mlx_llama_weights.npz> <path_to_tokenizer.model> "hello"
|
||||
```
|
||||
|
||||
Run `python llama.py --help` for more details.
|
||||
|
Loading…
Reference in New Issue
Block a user