Text2Text Generation
Transformers
Safetensors
English
EtherAI commited on
Commit
d5c6f39
·
verified ·
1 Parent(s): 47610fe

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -2
README.md CHANGED
@@ -30,7 +30,7 @@ v1 is a language model designed for English text generation. The architecture em
30
 
31
  ### Direct Use Scenarios
32
 
33
- V1 can be used directly for various text generation tasks, subject to its current limitations:
34
 
35
  - **Content Creation:** Generating initial drafts of articles, blog posts, stories, poems, or scripts.
36
  - **Brainstorming:** Assisting with idea generation.
@@ -38,7 +38,7 @@ V1 can be used directly for various text generation tasks, subject to its curren
38
  - **Question Answering:** Answering questions based on its limited training data (factual accuracy should always be verified).
39
  - **Educational Exploration:** Assisting users in understanding language patterns.
40
 
41
- **Considerations for Direct Use:** When using the model directly, users should be mindful of the potential for overfitting or generating content reflecting the biases inherent in its small initial training set.
42
 
43
  ### Downstream Use and Fine-tuning
44
 
 
30
 
31
  ### Direct Use Scenarios
32
 
33
+ V1 can be used directly for various text generation tasks by downloaing via HuggingFace for ondevice training , subject to its current limitations:
34
 
35
  - **Content Creation:** Generating initial drafts of articles, blog posts, stories, poems, or scripts.
36
  - **Brainstorming:** Assisting with idea generation.
 
38
  - **Question Answering:** Answering questions based on its limited training data (factual accuracy should always be verified).
39
  - **Educational Exploration:** Assisting users in understanding language patterns.
40
 
41
+ **Considerations for Direct Use:** When using the model directly, users should be mindful of the potential for overfitting or generating content reflecting the biases inherent in its small initial training set but the model size is low as under 300MB.
42
 
43
  ### Downstream Use and Fine-tuning
44