Hugging Face's logo

TheBloke
/
ALMA-13B-Pretrain-AWQ

Text Generation
Transformers
Safetensors
llama
text-generation-inference
4-bit precision
awq
Model card Files Community
1
TheBloke commited on
Commit
dc7ddbf
·
1 Parent(s): ea22966

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +3 -3
README.md CHANGED
@@ -5,7 +5,7 @@ license: mit
5
  model_creator: haoranxu
6
  model_name: ALMA 13B Pretrain
7
  model_type: llama
8
- prompt_template: 'Translate this from Chinese to English
9
 
10
  Chinese: {prompt}
11
 
@@ -61,7 +61,7 @@ It is also now supported by continuous batching server [vLLM](https://github.com
61
  ## Prompt template: ALMA
62
 
63
  ```
64
- Translate this from Chinese to English
65
  Chinese: {prompt}
66
  English:
67
 
@@ -160,7 +160,7 @@ model = AutoAWQForCausalLM.from_quantized(model_name_or_path, fuse_layers=True,
160
  tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, trust_remote_code=False)
161
 
162
  prompt = "Tell me about AI"
163
- prompt_template=f'''Translate this from Chinese to English
164
  Chinese: {prompt}
165
  English:
166
 
 
5
  model_creator: haoranxu
6
  model_name: ALMA 13B Pretrain
7
  model_type: llama
8
+ prompt_template: 'Translate this from Chinese to English:
9
 
10
  Chinese: {prompt}
11
 
 
61
  ## Prompt template: ALMA
62
 
63
  ```
64
+ Translate this from Chinese to English:
65
  Chinese: {prompt}
66
  English:
67
 
 
160
  tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, trust_remote_code=False)
161
 
162
  prompt = "Tell me about AI"
163
+ prompt_template=f'''Translate this from Chinese to English:
164
  Chinese: {prompt}
165
  English:
166