Update README.md
Browse files
README.md
CHANGED
|
@@ -26,12 +26,15 @@ Mify-Coder-2.5B-v0.1 is a **2.5B-parameter code-focused language model**. It del
|
|
| 26 |
|
| 27 |
| **Category** | **Benchmark** | **# Shots** | **Metric** | **Scores** |
|
| 28 |
|----------------|----------------------|-------------|------------|-------------------|
|
| 29 |
-
| Code Gen | MBPP | 0 | pass@1 |
|
| 30 |
| Code Gen | MBPP+ | 0 | pass@1 | 88.89% |
|
| 31 |
| Code Gen | HumanEval | 0 | pass@1 | 53.05% |
|
| 32 |
| Code Gen | HumanEval+ | 0 | pass@1 | 46.95% |
|
| 33 |
| Code Gen | NumpyEval | 0 | pass@1 | 56.44% |
|
| 34 |
| Code Gen | PandasEval | 0 | pass@1 | 53.47% |
|
|
|
|
|
|
|
|
|
|
| 35 |
|
| 36 |
- Outperforms larger models on algorithmic reasoning tasks while maintaining competitive general coding and security-oriented capabilities.
|
| 37 |
|
|
|
|
| 26 |
|
| 27 |
| **Category** | **Benchmark** | **# Shots** | **Metric** | **Scores** |
|
| 28 |
|----------------|----------------------|-------------|------------|-------------------|
|
| 29 |
+
| Code Gen | MBPP | 0 | pass@1 | 89.23% |
|
| 30 |
| Code Gen | MBPP+ | 0 | pass@1 | 88.89% |
|
| 31 |
| Code Gen | HumanEval | 0 | pass@1 | 53.05% |
|
| 32 |
| Code Gen | HumanEval+ | 0 | pass@1 | 46.95% |
|
| 33 |
| Code Gen | NumpyEval | 0 | pass@1 | 56.44% |
|
| 34 |
| Code Gen | PandasEval | 0 | pass@1 | 53.47% |
|
| 35 |
+
| Tool Use | BFCL v1 | 0 | acc | 79.19% |
|
| 36 |
+
| Tool Use | BFCL v2 | 0 | acc | 55.26% |
|
| 37 |
+
|
| 38 |
|
| 39 |
- Outperforms larger models on algorithmic reasoning tasks while maintaining competitive general coding and security-oriented capabilities.
|
| 40 |
|