bitext-innovations commited on
Commit
00f1393
1 Parent(s): 4baa121

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +10 -8
README.md CHANGED
@@ -7,11 +7,13 @@ sdk: static
7
  pinned: false
8
  ---
9
 
10
- Bitext provides NLP/NLG services to 3 of the top 5 companies on NASDAQ. Bitext Automates Text Data Services for Multilingual GenAI. We cover:
11
-
12
- - Generation of Synthetic Text, based on proprietary reliable NLG technology (not generative technology)
13
- - Automation of Data Labelling and Annotation (DAL), combining GenAI models and NLP tools with a human-in-the-loop approach
14
- - Verticalization of General-Purpose models (GPT, Mistral) in 20 domains (Customer Support, Banking, Travel)
15
- - Training and Evaluation of General-Purpose models (GPT, Mistral…) for Conversational AI
16
-
17
- We have developed NLP/NLG technology for 77 languages (including Arabic, Japanese, Chinese, Hindi, Urdu…) and 25 regional variants (like Egyptian Arabic, Canadian French, Indian English…)
 
 
 
7
  pinned: false
8
  ---
9
 
10
+ Bitext provides NLP/NLG services to 3 of the top 5 companies on NASDAQ. Bitext automates Text Data Services for Multilingual GenAI, covering:
11
+
12
+ - Generation of Synthetic Text based on proprietary NLG technology (not generative)
13
+ - Automation of Data Labelling and Annotation (DAL) using GenAI models and NLP tools with a human-in-the-loop approach
14
+ - Verticalization of General-Purpose models (GPT, Mistral, OpenELM) in 20 domains (Customer Support, Banking, Travel)
15
+ - Training and Evaluation of General-Purpose models for Conversational AI
16
+
17
+ We offer hybrid synthetic datasets to fine-tune LLMs like GPT, Mistral, and OpenELM, showcasing domain adaptation in sectors like Retail Banking. Our two-step approach allows clients to create customized LLMs by first using our dataset and then fine-tuning with their own data.
18
+
19
+ Our technology supports 77 languages (including Arabic, Japanese, Chinese, Hindi, Urdu) and 25 regional variants (like Egyptian Arabic, Canadian French, Indian English). More details can be found [From General-Purpose LLMs to Verticalized Enterprise Models](https://www.bitext.com/blog/general-purpose-models-verticalized-enterprise-genai/).