Commit
•
02f405f
1
Parent(s):
102b01e
Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language:
|
| 3 |
+
- en
|
| 4 |
+
tags:
|
| 5 |
+
- text generation
|
| 6 |
+
- pytorch
|
| 7 |
+
- causal-lm
|
| 8 |
+
- gpt_neox
|
| 9 |
+
license: mit
|
| 10 |
+
datasets:
|
| 11 |
+
- hoskinson-center/proof-pile
|
| 12 |
+
---
|
| 13 |
+
|
| 14 |
+
# ProofGPT-v0.1
|
| 15 |
+
|
| 16 |
+
# Model Description
|
| 17 |
+
ProofGPT-v0.1 is a 6.7B parameter language model based on the GPT-NeoX architecture and trained on the [proof-pile](https://huggingface.co/datasets/hoskinson-center/proof-pile) (v1.1).
|
| 18 |
+
We initiailized training with pythia-6.7b weights, a precursor to the [pythia-6.9b](https://huggingface.co/EleutherAI/pythia-6.9b) model that has roughly equivalent performance.
|
| 19 |
+
|
| 20 |
+
Detailed evaluations coming soon :)
|