Cyrile commited on
Commit
ca18a8d
1 Parent(s): 812313c

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +14 -2
README.md CHANGED
@@ -15,9 +15,9 @@ widget:
15
  - text: </s>Hello, who are you?<s>
16
  ---
17
 
18
- Bloomz-560m-sft-chat
19
  --------------------
20
- We introduce the Bloomz-560m-sft-chat model, which is a fine-tuning of a Large Language Model (LLM) [bigscience/bloomz-560m](https://huggingface.co/bigscience/bloomz-560m). This model is notable for being pre-trained for a chatbot context and undergoing a transposition from float16 to bfloat16. Therefore, this model serves as a solid starting point for fine-tuning towards other more specific tasks.
21
 
22
  The model was trained equally on both French and English data, ensuring maximum efficiency for these two languages (and their interactions). Due to the transition from float16 to bfloat16, we do not guarantee the preservation of the original model's multilingual capabilities. However, fine-tuning can restore reasonable performance on other languages.
23
 
@@ -85,3 +85,15 @@ result = model("</s>Que veut dire NLP ?<s>", max_new_tokens=512)
85
  result
86
  [{'generated_text': "</s>Que veut dire NLP ?<s>NLP est un terme générique pour les techniques de traitement du langage naturel. Il englobe une variété de techniques, notamment la reconnaissance de la parole, la traduction automatique, la reconnaissance d'images et de vidéos, la reconnaissance vocale et la reconnaissance de la parole."}]
87
  ```
 
 
 
 
 
 
 
 
 
 
 
 
 
15
  - text: </s>Hello, who are you?<s>
16
  ---
17
 
18
+ bloomz-560m-sft-chat
19
  --------------------
20
+ We introduce the bloomz-560m-sft-chat model, which is a fine-tuning of a Large Language Model (LLM) [bigscience/bloomz-560m](https://huggingface.co/bigscience/bloomz-560m). This model is notable for being pre-trained for a chatbot context and undergoing a transposition from float16 to bfloat16. Therefore, this model serves as a solid starting point for fine-tuning towards other more specific tasks.
21
 
22
  The model was trained equally on both French and English data, ensuring maximum efficiency for these two languages (and their interactions). Due to the transition from float16 to bfloat16, we do not guarantee the preservation of the original model's multilingual capabilities. However, fine-tuning can restore reasonable performance on other languages.
23
 
 
85
  result
86
  [{'generated_text': "</s>Que veut dire NLP ?<s>NLP est un terme générique pour les techniques de traitement du langage naturel. Il englobe une variété de techniques, notamment la reconnaissance de la parole, la traduction automatique, la reconnaissance d'images et de vidéos, la reconnaissance vocale et la reconnaissance de la parole."}]
87
  ```
88
+
89
+ Citation
90
+ --------
91
+
92
+ ```bibtex
93
+ @online{delestre:hal-03674695,
94
+ AUTHOR = {Cyrile Delestre},
95
+ URL = {https://huggingface.co/cmarkea/bloomz-560m-sft-chat},
96
+ YEAR = {2023},
97
+ KEYWORDS = {NLP ; Transformers ; Bloomz},
98
+ }
99
+ ```