thevox commited on
Commit
fc9367c
1 Parent(s): c9ce2f8

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +20 -2
README.md CHANGED
@@ -1,21 +1,39 @@
1
  ---
2
  language:
3
  - en
 
 
4
  library_name: transformers
5
  tags:
6
  - gpt
7
  - llm
8
  - large language model
9
  - h2o-llmstudio
10
- inference: false
11
- thumbnail: https://h2o.ai/etc.clientlibs/h2o/clientlibs/clientlib-site/resources/images/favicon.ico
 
 
 
 
 
 
 
 
12
  ---
13
  # Model Card
14
  ## Summary
15
 
 
 
 
 
 
 
 
16
  This model was trained using [H2O LLM Studio](https://github.com/h2oai/h2o-llmstudio).
17
  - Base model: [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1)
18
 
 
19
 
20
  ## Usage
21
 
 
1
  ---
2
  language:
3
  - en
4
+ - 'no'
5
+ - nb
6
  library_name: transformers
7
  tags:
8
  - gpt
9
  - llm
10
  - large language model
11
  - h2o-llmstudio
12
+ - translation
13
+ inference: true
14
+ thumbnail: >-
15
+ https://h2o.ai/etc.clientlibs/h2o/clientlibs/clientlib-site/resources/images/favicon.ico
16
+ datasets:
17
+ - thevox/en-nb-15k
18
+ license: mpl-2.0
19
+ metrics:
20
+ - perplexity
21
+ pipeline_tag: text-generation
22
  ---
23
  # Model Card
24
  ## Summary
25
 
26
+ English to Norwegian translation model, to rival DeepL natural translations.
27
+
28
+ Context length is 1024 for input and output (2048).
29
+
30
+ Outputs context, translation and improved translation.
31
+
32
+ ### Training
33
  This model was trained using [H2O LLM Studio](https://github.com/h2oai/h2o-llmstudio).
34
  - Base model: [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1)
35
 
36
+ Trained with 4-Bit Lora (R=64) for 2 epochs. Perplexity on validation: 1.245
37
 
38
  ## Usage
39