Hugging Face's logo

GerbilLab
/
GerbilBlender-A-15m

Text Generation
Transformers
PyTorch
gptj
Inference Endpoints
Model card Files Community
1
crumb commited on
Commit
9e0e9ba
·
1 Parent(s): 7261716

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -2
README.md CHANGED
@@ -1,8 +1,8 @@
1
  ---
2
  license: apache-2.0
3
  ---
4
- | Model Name | Parameters | Class | Ratio | Tokens | Batch Size (Tokens) | Training Loss ↓ | hellaswag acc ↑ | lambada ppl ↓ | lambada acc ↑ | winogrande acc ↑
5
- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
6
  | [GerbilLab/GerbilBlender-A-15m](https://hf.co/GerbilLab/GerbilBlender-A-15m) | 15m | A-Class | 20 | 280M | 131k | 4.9642 |
7
 
8
  "Blender" models, inspired by UL2 pretraining, are trained equally in fill-in-the-middle, causal modelling, and masked language modelling tasks. Special tokens for these models include:
 
1
  ---
2
  license: apache-2.0
3
  ---
4
+ | Model Name | Parameters | Class | Ratio | Tokens | Batch Size (Tokens) | Training Loss ↓ |
5
+ | --- | --- | --- | --- | --- | --- | --- |
6
  | [GerbilLab/GerbilBlender-A-15m](https://hf.co/GerbilLab/GerbilBlender-A-15m) | 15m | A-Class | 20 | 280M | 131k | 4.9642 |
7
 
8
  "Blender" models, inspired by UL2 pretraining, are trained equally in fill-in-the-middle, causal modelling, and masked language modelling tasks. Special tokens for these models include: