andthattoo commited on
Commit
626e1a2
·
verified ·
1 Parent(s): 8241242

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +5 -1
README.md CHANGED
@@ -3,6 +3,7 @@ base_model: nisten/Biggie-SmoLlm-0.15B-Base
3
  license: mit
4
  datasets:
5
  - LDJnr/Capybara
 
6
  pipeline_tag: text-generation
7
  tags:
8
  - llama
@@ -11,7 +12,8 @@ tags:
11
 
12
  ### Fine-tuned [Biggie-SmoLlm-0.15B-Base](https://huggingface.co/nisten/Biggie-SmoLlm-0.15B-Base) for generating subqueries
13
 
14
- This dude is trained for boosting the performance of your IR app, or RAG
 
15
  My motivation was to tackle a core problem of IR with an extremely lightweight, but capable model.
16
 
17
  If queries are
@@ -19,6 +21,8 @@ If queries are
19
  - vague, ask follow up questions
20
  - multiple sub questions, generate multiple queries for each of them
21
 
 
 
22
 
23
  Heads up: [Ollama](https://ollama.com/andthattoo/subquery-smollm) version works 160 tps on 1 CPU core. No GPU? No worries. This little dude’s got you.
24
 
 
3
  license: mit
4
  datasets:
5
  - LDJnr/Capybara
6
+ - andthattoo/subqueries
7
  pipeline_tag: text-generation
8
  tags:
9
  - llama
 
12
 
13
  ### Fine-tuned [Biggie-SmoLlm-0.15B-Base](https://huggingface.co/nisten/Biggie-SmoLlm-0.15B-Base) for generating subqueries
14
 
15
+ This dude is trained for boosting the performance of your IR app, or RAG.
16
+
17
  My motivation was to tackle a core problem of IR with an extremely lightweight, but capable model.
18
 
19
  If queries are
 
21
  - vague, ask follow up questions
22
  - multiple sub questions, generate multiple queries for each of them
23
 
24
+ Training data was generated wit [Dria](dria.co). A decentralized p2p network for synthetic data. Join [discord](https://discord.gg/dria) to help decentralized data generation
25
+
26
 
27
  Heads up: [Ollama](https://ollama.com/andthattoo/subquery-smollm) version works 160 tps on 1 CPU core. No GPU? No worries. This little dude’s got you.
28