Muthukumaran
commited on
Commit
·
bc95fec
1
Parent(s):
0d30fdd
Update README.md
Browse files
README.md
CHANGED
@@ -27,7 +27,7 @@ nasa-smd-ibm-v0.1 is a RoBERTa-based, Encoder-only transformer model, domain-ada
|
|
27 |
- PubMed abstracts
|
28 |
- PMC (commercial license subset)
|
29 |
|
30 |
-
|
31 |
|
32 |
## Training Procedure
|
33 |
- **Framework**: fairseq 0.12.1 with PyTorch 1.9.1
|
@@ -39,6 +39,7 @@ nasa-smd-ibm-v0.1 is a RoBERTa-based, Encoder-only transformer model, domain-ada
|
|
39 |
- Pruned SQuAD2.0 (SQ2) Benchmark
|
40 |
- NASA SMD Experts Benchmark (WIP)
|
41 |
|
|
|
42 |
![BLURB Benchmark Results](https://cdn-uploads.huggingface.co/production/uploads/61099e5d86580d4580767226/K0IpQnTQmrfQJ1JXxn1B6.png)
|
43 |
![SQ2 Benchmark Results](https://cdn-uploads.huggingface.co/production/uploads/61099e5d86580d4580767226/R4oMJquUz4puah3lvd5Ve.png)
|
44 |
|
@@ -89,4 +90,8 @@ NASA SMD
|
|
89 |
- Sanaz Vahidinia
|
90 |
- Ryan McGranaghan
|
91 |
- Armin Mehrabian
|
92 |
-
- Tsendgar Lee
|
|
|
|
|
|
|
|
|
|
27 |
- PubMed abstracts
|
28 |
- PMC (commercial license subset)
|
29 |
|
30 |
+
|
31 |
|
32 |
## Training Procedure
|
33 |
- **Framework**: fairseq 0.12.1 with PyTorch 1.9.1
|
|
|
39 |
- Pruned SQuAD2.0 (SQ2) Benchmark
|
40 |
- NASA SMD Experts Benchmark (WIP)
|
41 |
|
42 |
+
![Dataset Chart](https://cdn-uploads.huggingface.co/production/uploads/61099e5d86580d4580767226/h_Tf-mpBFzeUaRR2xeq7g.png)
|
43 |
![BLURB Benchmark Results](https://cdn-uploads.huggingface.co/production/uploads/61099e5d86580d4580767226/K0IpQnTQmrfQJ1JXxn1B6.png)
|
44 |
![SQ2 Benchmark Results](https://cdn-uploads.huggingface.co/production/uploads/61099e5d86580d4580767226/R4oMJquUz4puah3lvd5Ve.png)
|
45 |
|
|
|
90 |
- Sanaz Vahidinia
|
91 |
- Ryan McGranaghan
|
92 |
- Armin Mehrabian
|
93 |
+
- Tsendgar Lee
|
94 |
+
|
95 |
+
## Disclaimer
|
96 |
+
|
97 |
+
This Encoder-only model is experimental, and we are still exploring ways to improve the model.
|