reach-vb HF staff commited on
Commit
d09db19
1 Parent(s): 71bfc8d

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +37 -0
README.md ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ pipeline_tag: text-generation
3
+ library_name: mlx
4
+ inference: false
5
+ tags:
6
+ - facebook
7
+ - meta
8
+ - llama
9
+ - llama-2
10
+ - mlx
11
+ license: llama2
12
+ ---
13
+
14
+ # **Llama 2**
15
+
16
+ Llama 2 is a collection of pretrained and fine-tuned generative text models ranging in scale from 7 billion to 70 billion parameters. This is the repository for the 7B fine-tuned model, in `npz` format suitable for use in Apple's MLX framework.
17
+
18
+ Weights have been converted to `float16` from the original `bfloat16` type, because `numpy` is not compatible with `bfloat16` out of the box.
19
+
20
+ How to use with [MLX](https://github.com/ml-explore/mlx).
21
+
22
+ ```bash
23
+ # Install mlx, mlx-examples, huggingface-cli
24
+ pip install mlx
25
+ pip install huggingface_hub hf_transfer
26
+ git clone https://github.com/ml-explore/mlx-examples.git
27
+
28
+ # Download model
29
+ export HF_HUB_ENABLE_HF_TRANSFER=1
30
+ huggingface-cli download --local-dir Llama-2-7b-chat-mlx mlx-llama/Llama-2-7b-chat-mlx
31
+
32
+ # Run example
33
+ python mlx-examples/llama/llama.py --prompt "My name is " Llama-2-7b-chat-mlx/ Llama-2-7b-chat-mlx/tokenizer.model
34
+ ```
35
+
36
+ Please, refer to the [original model card](https://huggingface.co/meta-llama/Llama-2-7b-chat) for details on Llama 2.
37
+