Add LM Eval Harness Scores
Browse files
README.md
CHANGED
|
@@ -78,7 +78,18 @@ Yi-Ko series models are an auto-regressive language model that uses an optimized
|
|
| 78 |
|
| 79 |
## LM Eval Harness - Korean (polyglot branch)
|
| 80 |
|
| 81 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 82 |
|
| 83 |
## LICENSE
|
| 84 |
|
|
|
|
| 78 |
|
| 79 |
## LM Eval Harness - Korean (polyglot branch)
|
| 80 |
|
| 81 |
+
| beomi/Yi-Ko-6B | 0 | 5 | 10 | 50 |
|
| 82 |
+
|:---------------------------------|---------:|---------:|---------:|---------:|
|
| 83 |
+
| kobest_boolq (macro_f1) | 0.705806 | 0.79905 | 0.814299 | 0.81704 |
|
| 84 |
+
| kobest_copa (macro_f1) | 0.775604 | 0.808899 | 0.816866 | 0.842943 |
|
| 85 |
+
| kobest_hellaswag (macro_f1) | 0.500876 | 0.498673 | 0.493507 | 0.492183 |
|
| 86 |
+
| kobest_sentineg (macro_f1) | 0.404371 | 0.967254 | 0.982368 | 0.974811 |
|
| 87 |
+
| kohatespeech (macro_f1) | 0.353428 | 0.351804 | 0.402423 | 0.503764 |
|
| 88 |
+
| kohatespeech_apeach (macro_f1) | 0.337667 | 0.498679 | 0.471962 | 0.608401 |
|
| 89 |
+
| kohatespeech_gen_bias (macro_f1) | 0.124535 | 0.484745 | 0.474475 | 0.461714 |
|
| 90 |
+
| korunsmile (f1) | 0.382804 | 0.349344 | 0.391383 | 0.432875 |
|
| 91 |
+
| nsmc (acc) | 0.55064 | 0.8801 | 0.89866 | 0.9071 |
|
| 92 |
+
| pawsx_ko (acc) | 0.5145 | 0.54 | 0.538 | 0.5165 |
|
| 93 |
|
| 94 |
## LICENSE
|
| 95 |
|