Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,20 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: apache-2.0
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
datasets:
|
| 4 |
+
- HuggingFaceTB/smollm-corpus
|
| 5 |
+
language:
|
| 6 |
+
- en
|
| 7 |
+
pipeline_tag: text2text-generation
|
| 8 |
+
library_name: transformers
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
|
| 12 |
+
# tFINE-900m-e16-d32-1024ctx
|
| 13 |
+
|
| 14 |
+
|
| 15 |
+
Pretrained T5 model with nanoT5:
|
| 16 |
+
|
| 17 |
+
- ~900m parameters, 16 layers in encoder, 32 layers in decoder
|
| 18 |
+
- sentencepiece tokenizer with 48k vocab & byte-pair fallback
|
| 19 |
+
- handles whitespaces etc correctly (unlike standard T5 tokenizer)
|
| 20 |
+
- 1024 ctx during pretrain
|