jzhang533 ckl117 commited on
Commit
dc1dd18
·
verified ·
1 Parent(s): e0d3ac5

Update config.json (#15)

Browse files

- Update config.json (dd15cfb90079d7cd428538ef75e01a8230c7ab75)
- minor update to readme (262fa420ec92a220874a928b6e557a72db72b4df)


Co-authored-by: ckl <[email protected]>

Files changed (2) hide show
  1. README.md +2 -3
  2. config.json +1 -0
README.md CHANGED
@@ -72,8 +72,7 @@ ERNIE-4.5-21B-A3B is a text MoE Post-trained model, with 21B total parameters an
72
 
73
  ### Using `transformers` library
74
 
75
- **Note**: Before using the model, please ensure you have the `transformers` library installed
76
- (upcoming version 4.54.0 or [the latest version](https://github.com/huggingface/transformers?tab=readme-ov-file#installation))
77
 
78
  The following contains a code snippet illustrating how to use the model generate content based on given inputs.
79
 
@@ -120,7 +119,7 @@ print("generate_text:", generate_text)
120
  [vllm](https://github.com/vllm-project/vllm/tree/main) github library. Python-only [build](https://docs.vllm.ai/en/latest/getting_started/installation/gpu.html#set-up-using-python-only-build-without-compilation).
121
 
122
  ```bash
123
- vllm serve baidu/ERNIE-4.5-21B-A3B-PT --trust-remote-code
124
  ```
125
 
126
  ## License
 
72
 
73
  ### Using `transformers` library
74
 
75
+ **Note**: You'll need the `transformers` library (version 4.54.0 or newer) installed to use this model.
 
76
 
77
  The following contains a code snippet illustrating how to use the model generate content based on given inputs.
78
 
 
119
  [vllm](https://github.com/vllm-project/vllm/tree/main) github library. Python-only [build](https://docs.vllm.ai/en/latest/getting_started/installation/gpu.html#set-up-using-python-only-build-without-compilation).
120
 
121
  ```bash
122
+ vllm serve baidu/ERNIE-4.5-21B-A3B-PT
123
  ```
124
 
125
  ## License
config.json CHANGED
@@ -28,6 +28,7 @@
28
  "rope_scaling": null,
29
  "rope_theta": 500000.0,
30
  "router_aux_loss_coef": 0.001,
 
31
  "torch_dtype": "bfloat16",
32
  "transformers_version": "4.54.0.dev0",
33
  "use_bias": false,
 
28
  "rope_scaling": null,
29
  "rope_theta": 500000.0,
30
  "router_aux_loss_coef": 0.001,
31
+ "tie_word_embeddings": true,
32
  "torch_dtype": "bfloat16",
33
  "transformers_version": "4.54.0.dev0",
34
  "use_bias": false,