Datasets:

Modalities:
Text
Formats:
text
Libraries:
Datasets
License:
wjayesh commited on
Commit
f329e8e
·
verified ·
1 Parent(s): b2d4839

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +44 -1
README.md CHANGED
@@ -2,4 +2,47 @@
2
  license: mit
3
  ---
4
 
5
- This is how you use the llms.txt files: TODO
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2
  license: mit
3
  ---
4
 
5
+ # ZenML's llms.txt documentation
6
+
7
+ ## Available files
8
+
9
+ The following llms.txt files are available for ZenML.
10
+
11
+ ### basics.txt
12
+
13
+ Tokens: 120k
14
+
15
+ This file covers the User Guides and the Getting Started section of the ZenML documentation and can be used
16
+ for answering basic questions about ZenML. This file can also be used alongside other domain-specific files in
17
+ cases where you need better answers.
18
+
19
+ ### component-guide.txt
20
+
21
+ Tokens: 180k
22
+
23
+ This file covers all the stack components in ZenML and can be used when you want to find answers pertaining
24
+ to all of our integrations, how to configure/use them and more.
25
+
26
+ ### how-to-guides.txt
27
+
28
+ Tokens: 75k
29
+
30
+ This file contains all the doc pages in the how-to section of our documentation; each page summarized to contain
31
+ all useful information. For most cases, the how-to guides can answer all process questions.
32
+
33
+ ### llms-full.txt
34
+
35
+ Tokens: 600k
36
+
37
+ The whole ZenML documentation in its glory, un-summarized. Use this for the most accurate answers on ZenML.
38
+
39
+ ## Tips and recommendations
40
+
41
+ - Choose the file that pertains to the part of ZenML you want answers for.
42
+ - In every file, the text comes prefixed with the filename which means you can ask your LLM to return file
43
+ references when answering questions. This is particularly helpful when using the how-to guides which don't
44
+ the full text, but rather a summary of it.
45
+ - You can mix two files, as your context window allows to get more accurate results.
46
+ - While prompting, make sure you tell the LLM to not return an answer that it can't infer from the given
47
+ text file, to avoid getting hallucinated answers.
48
+