nguyenbh commited on
Commit
ae6cb90
1 Parent(s): a09cede

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +3 -1
README.md CHANGED
@@ -288,6 +288,8 @@ The prompt is the same as the [CLIcK paper](https://arxiv.org/abs/2403.06412) pr
288
  - GPT-4-turbo: 2024-04-09 version
289
  - GPT-3.5-turbo: 2023-06-13 version
290
 
 
 
291
  | Benchmarks | Phi-3.5-MoE-Instruct | Phi-3.0-Mini-128k-Instruct (June2024) | Llama-3.1-8B-Instruct | GPT-4o | GPT-4o-mini | GPT-4-turbo | GPT-3.5-turbo |
292
  |:-------------------------|-----------------------:|--------------------------------:|------------------------:|---------:|--------------:|--------------:|----------------:|
293
  | CLIcK | 56.44 | 29.12 | 47.82 | 80.46 | 68.5 | 72.82 | 50.98 |
@@ -296,7 +298,7 @@ The prompt is the same as the [CLIcK paper](https://arxiv.org/abs/2403.06412) pr
296
  | KMMLU (5-shot) | 47.92 | 29.98 | 20.21 | 64.28 | 51.62 | 59.29 | 42.28 |
297
  | KMMLU-HARD (0-shot, CoT) | 25.34 | 25.68 | 24.03 | 39.62 | 24.56 | 30.56 | 20.97 |
298
  | KMMLU-HARD (5-shot) | 25.66 | 25.73 | 15.81 | 40.94 | 24.63 | 31.12 | 21.19 |
299
- | Average | 45.82 | 29.99 | 29.29 | 62.54 | 50.08 | 56.74 | 39.61 |
300
 
301
  #### CLIcK (Cultural and Linguistic Intelligence in Korean)
302
 
 
288
  - GPT-4-turbo: 2024-04-09 version
289
  - GPT-3.5-turbo: 2023-06-13 version
290
 
291
+ Overall, the Phi-3.5 MoE model with just 6.6B active params outperforms GPT-3.5-Turbo.
292
+
293
  | Benchmarks | Phi-3.5-MoE-Instruct | Phi-3.0-Mini-128k-Instruct (June2024) | Llama-3.1-8B-Instruct | GPT-4o | GPT-4o-mini | GPT-4-turbo | GPT-3.5-turbo |
294
  |:-------------------------|-----------------------:|--------------------------------:|------------------------:|---------:|--------------:|--------------:|----------------:|
295
  | CLIcK | 56.44 | 29.12 | 47.82 | 80.46 | 68.5 | 72.82 | 50.98 |
 
298
  | KMMLU (5-shot) | 47.92 | 29.98 | 20.21 | 64.28 | 51.62 | 59.29 | 42.28 |
299
  | KMMLU-HARD (0-shot, CoT) | 25.34 | 25.68 | 24.03 | 39.62 | 24.56 | 30.56 | 20.97 |
300
  | KMMLU-HARD (5-shot) | 25.66 | 25.73 | 15.81 | 40.94 | 24.63 | 31.12 | 21.19 |
301
+ | **Average** | **45.82** | **29.99** | **29.29** | **62.54** | **50.08** | **56.74** | **39.61** |
302
 
303
  #### CLIcK (Cultural and Linguistic Intelligence in Korean)
304