jpacifico commited on
Commit
613013c
·
verified ·
1 Parent(s): 3d51135

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -0
README.md CHANGED
@@ -18,6 +18,8 @@ This is a merge of pre-trained language models created using [mergekit](https://
18
 
19
  **Interpretation:** Significant gains on language understanding & pragmatic reasoning (ARC-C/E, Wino, BoolQ, HellaSwag, TriviaQA) with stability on other skills. Math/code are not the optimization target; GSM8K stays essentially stable relative to the BitNet 1.58-bit baseline.
20
 
 
 
21
  | Benchmark (metric) | microsoft/bitnet-b1.58-2B-4T-bf16 | bitnet-dpo-merged-modelstock7 |
22
  |------------------------------------|-----------------------------------|--------------------------------|
23
  | arc_challenge 0 shot | 47.95 | **51.62** |
 
18
 
19
  **Interpretation:** Significant gains on language understanding & pragmatic reasoning (ARC-C/E, Wino, BoolQ, HellaSwag, TriviaQA) with stability on other skills. Math/code are not the optimization target; GSM8K stays essentially stable relative to the BitNet 1.58-bit baseline.
20
 
21
+ **ARC-Challenge:** 51.62 (First-ever ≥50 score for a model in the 2B category, i.e., >1.5B and <2.5B params)
22
+
23
  | Benchmark (metric) | microsoft/bitnet-b1.58-2B-4T-bf16 | bitnet-dpo-merged-modelstock7 |
24
  |------------------------------------|-----------------------------------|--------------------------------|
25
  | arc_challenge 0 shot | 47.95 | **51.62** |