Update README.md
Browse files
README.md
CHANGED
|
@@ -189,16 +189,18 @@ evalplus.evaluate \
|
|
| 189 |
|
| 190 |
| Metric | ibm-granite/granite-3.1-8b-instruct | neuralmagic-ent/granite-3.1-8b-instruct-quantized.w4a16 |
|
| 191 |
|-----------------------------------------|:---------------------------------:|:-------------------------------------------:|
|
| 192 |
-
| ARC-Challenge (Acc-Norm, 25-shot) | |
|
| 193 |
-
| GSM8K (Strict-Match, 5-shot) | |
|
| 194 |
-
| HellaSwag (Acc-Norm, 10-shot) | |
|
| 195 |
-
| MMLU (Acc, 5-shot) | |
|
| 196 |
-
| TruthfulQA (MC2, 0-shot) | | |
|
| 197 |
-
| Winogrande (Acc, 5-shot) | | |
|
| 198 |
-
| **Average Score** |
|
| 199 |
-
| **Recovery** | **100.00**
|
| 200 |
|
| 201 |
#### HumanEval pass@1 scores
|
| 202 |
-
|
|
|
|
|
|
|
| 203 |
|
| 204 |
|
|
|
|
| 189 |
|
| 190 |
| Metric | ibm-granite/granite-3.1-8b-instruct | neuralmagic-ent/granite-3.1-8b-instruct-quantized.w4a16 |
|
| 191 |
|-----------------------------------------|:---------------------------------:|:-------------------------------------------:|
|
| 192 |
+
| ARC-Challenge (Acc-Norm, 25-shot) | 66.81 | 66.98 |
|
| 193 |
+
| GSM8K (Strict-Match, 5-shot) | 64.52 | 68.08 |
|
| 194 |
+
| HellaSwag (Acc-Norm, 10-shot) | 84.18 | 83.30 |
|
| 195 |
+
| MMLU (Acc, 5-shot) | 65.52 | 63.96 |
|
| 196 |
+
| TruthfulQA (MC2, 0-shot) | 60.57 | 60.62 |
|
| 197 |
+
| Winogrande (Acc, 5-shot) | 80.19 | 78.61 |
|
| 198 |
+
| **Average Score** | **70.30** | **70.26** |
|
| 199 |
+
| **Recovery** | **100.00** | **99.94** |
|
| 200 |
|
| 201 |
#### HumanEval pass@1 scores
|
| 202 |
+
| Metric | ibm-granite/granite-3.1-8b-instruct | neuralmagic-ent/granite-3.1-8b-instruct-quantized.w4a16 |
|
| 203 |
+
|-----------------------------------------|:---------------------------------:|:-------------------------------------------:|
|
| 204 |
+
| HumanEval Pass@1 | 71.00 | 70.90 |
|
| 205 |
|
| 206 |
|