Update README.md
Browse files
README.md
CHANGED
|
@@ -43,6 +43,17 @@ and is comparable with Mistral-7B-Instruct-v0.1 on MMLU and MT-Bench in English.
|
|
| 43 |
|
| 44 |
\* Few-shot learning cannot effectively guide the model to generate the proper answer.
|
| 45 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 46 |
## Inference Performance
|
| 47 |
|
| 48 |
| Models | Speed (char/sec) |Max Input Length (TC Char)|
|
|
|
|
| 43 |
|
| 44 |
\* Few-shot learning cannot effectively guide the model to generate the proper answer.
|
| 45 |
|
| 46 |
+
| Models on 5 shot TMMLU+ | STEM | Social Science | Humanities | Other |
|
| 47 |
+
|-----------------------------------------------------|--------------|----------------|------------|------------|
|
| 48 |
+
| MediaTek-Research/Breeze-7B-Base-v0.1 | 35.74 | 46.08 | 40.29 | 40.35 |
|
| 49 |
+
| mistralai/Mistral-7B-v0.1 | | | | |
|
| 50 |
+
| yentinglin/Taiwan-LLM-7B-v2.1-base | | | | |
|
| 51 |
+
| yentinglin/Taiwan-LLM-13B-v2.0-base | | | | |
|
| 52 |
+
| 01-ai/Yi-6B | | | | |
|
| 53 |
+
| 01-ai/Yi-34B | | | | |
|
| 54 |
+
| Qwen/Qwen-7B | | | | |
|
| 55 |
+
| Qwen/Qwen-14B | | | | |
|
| 56 |
+
|
| 57 |
## Inference Performance
|
| 58 |
|
| 59 |
| Models | Speed (char/sec) |Max Input Length (TC Char)|
|