Commit ·
53ac067
1
Parent(s): c11ef27
Update README.md
Browse files
README.md
CHANGED
|
@@ -16,6 +16,7 @@ Trained from scratch using Tensorflow Keras
|
|
| 16 |
- Type: decoder-only
|
| 17 |
- Tokenizer: BPE
|
| 18 |
- Vocabulary size: 32000
|
|
|
|
| 19 |
- Hidden size: 768
|
| 20 |
- FFN size: 3072
|
| 21 |
- Attention heads: 24
|
|
|
|
| 16 |
- Type: decoder-only
|
| 17 |
- Tokenizer: BPE
|
| 18 |
- Vocabulary size: 32000
|
| 19 |
+
- Max sequence length: 120
|
| 20 |
- Hidden size: 768
|
| 21 |
- FFN size: 3072
|
| 22 |
- Attention heads: 24
|