jiwasawa commited on
Commit
7a203a4
1 Parent(s): 2e8eaf7

Update README

Browse files
Files changed (1) hide show
  1. README.md +48 -3
README.md CHANGED
@@ -1,3 +1,48 @@
1
- ---
2
- license: llama3
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - en
4
+ - ja
5
+ library_name: transformers
6
+ pipeline_tag: text-generation
7
+ license: llama3
8
+ model_type: llama
9
+ ---
10
+
11
+ # Llama3-Preferred-MedSwallow-70B
12
+
13
+ ## Model Description
14
+
15
+ Llama3-Preferred-MedSwallow-70B is a finetuned model based on [tokyotech-llm/Llama-3-Swallow-70B-v0.1](https://huggingface.co/tokyotech-llm/Llama-3-Swallow-70B-v0.1), which has undergone continued pretraining on an original corpus of medical-related text.
16
+ For more details, please refer to our blog post at https://tech.preferred.jp/ja/blog/llama3-preferred-medswallow-70b/.
17
+ The model is released under the [META LLAMA 3 COMMUNITY LICENSE](https://llama.meta.com/llama3/license/).
18
+
19
+ ## Model Performance
20
+
21
+ The table below shows the performance on the Japanese national medical licensing examinations from 2018 to 2022 ([IgakuQA](https://github.com/jungokasai/IgakuQA)).
22
+ | Model ID | Average | 2018 | 2019 | 2020 | 2021 | 2022 |
23
+ |:--------------------------------------------------------------------------------------------------------------------|-------------------:|-------:|-------:|-------:|-------:|-------:|
24
+ | **Llama3-Preferred-MedSwallow-70B** | **395.2** | **407** | **390** | **391** | 393 | **395** |
25
+ GPT-4 | 388.8 | 382 | 385 | 387 | **398** | 392 |
26
+ | [Llama-3-Swallow-70B-v0.1](https://huggingface.co/tokyotech-llm/Llama-3-Swallow-70B-v0.1) | 348.6 | 353 | 347 | 353 | 345 | 345 |
27
+ | [Meta-Llama-3-70B](https://huggingface.co/meta-llama/Meta-Llama-3-70B) | 334.6 | 353 | 340 | 348 | 314 | 318 |
28
+ | [Qwen2-72B](https://huggingface.co/Qwen/Qwen2-72B) | 331.2 | 320 | 325 | 325 | 326 | 360 |
29
+ | [gemma-2-27b](https://huggingface.co/google/gemma-2-27b) | 316 | 337 | 298 | 327 | 296 | 322 |
30
+ | [Swallow-70b-NVE-hf](https://huggingface.co/tokyotech-llm/Swallow-70b-NVE-hf) | 291.6 | 283 | 280 | 300 | 295 | 300 |
31
+ | [Swallow-MX-8x7b-NVE-v0.1](https://huggingface.co/tokyotech-llm/Swallow-MX-8x7b-NVE-v0.1) | 280.8 | 262 | 273 | 291 | 284 | 294 |
32
+ ChatGPT | 273.2 | 266 | 250 | 266 | 297 | 287 |
33
+
34
+ ## Limitations
35
+
36
+ The model was developed for research purposes and is not intended for clinical diagnosis.
37
+ It is the users responsibility to ensure compliance with applicable rules and regulations.
38
+
39
+ ## Contributors
40
+
41
+ Preferred Networks, Inc.
42
+ - Junichiro Iwasawa
43
+ - Keita Suzuki
44
+ - Wataru Kawakami
45
+
46
+ ## License
47
+
48
+ [META LLAMA 3 COMMUNITY LICENSE](https://llama.meta.com/llama3/license/)