|
--- |
|
license: apache-2.0 |
|
datasets: |
|
- wenbopan/RefGPT-Fact-v2-8x |
|
- wenbopan/anti-haystack |
|
- wenbopan/OpenHermes-2.5-zh |
|
language: |
|
- zh |
|
- en |
|
--- |
|
|
|
 |
|
|
|
# Fi-9B |
|
|
|
Fi-9B is an improved [Yi-9B-200K](https://huggingface.co/01-ai/Yi-9B-200K) with extensive instruction tuning on [Fusang-V1](https://huggingface.co/datasets/wenbopan/Fusang-v1). Compared to Yi-9B-200K, Fi-9B has gained greater capability in various downstream tasks and long-context modeling thanks to the large-scale synthetic data in Fusang-V1. |
|
|
|
## Performance |
|
|
|
Fi-9B enhances its ability compared to Yi-9B-200K in most dimensions, especially in long-range modeling and bilingual (English, Chinese) understanding. Fi is competitive among all open-sourced models at around 9B parameters. Fi-9B is good at both factual tasks and preferred by LLM-judges. |
|
|
|
### Fact-based Evaluation (Open LLM Leaderboard) |
|
|
|
| **Metric** | **MMLU** | GSM8K | **HellaSwag** | **TruthfulQA** | **Arc** | **Winogrande** | |
|
| -------------- | --------- | --------- | ------------- | -------------- | ----------- | -------------- | |
|
| **Yi-9B-200K** | 65.73 | 50.49 | 56.72 | 33.80 | 69.25 | 71.67 | |
|
| **Fi-9B-200K** | **68.80** | **63.08** | **57.28** | **40.86** | **72.58** | 71.11 | |
|
|
|
### Long-context Modeling (LongBench) |
|
|
|
| **Name** | **Average_zh** | **Average_en** | **Code Completion** | |
|
|----------------|----------------|----------------|---------------------| |
|
| **Yi-9B-200K** | 30.288 | 36.7071 | 72.2 | |
|
| **Fi-9B-200K** | **41.092** | **40.9536** | 46.0 | |
|
|
|
<details> |
|
<summary>Score breakdown</summary> |
|
|
|
| **Name** | **Few-shot Learning_en** | **Synthetic Tasks_en** | **Single-Doc QA_en** | **Multi-Doc QA_en** | **Summarization_en** | **Few-shot Learning_zh** | **Synthetic Tasks_zh** | **Single-Doc QA_zh** | **Multi-Doc QA_zh** | **Summarization_zh** | |
|
|----------------|--------------------------|------------------------|----------------------|---------------------|----------------------|--------------------------|------------------------|----------------------|---------------------|----------------------| |
|
| **Yi-9B-200K** | 60.6 | 22.8 | 30.9 | 38.9 | 25.8 | 46.5 | 28.0 | 49.6 | 17.7 | 9.7 | |
|
| **Fi-9B-200K** | **63.8** | **40.2** | **36.2** | 38.0 | **26.3** | 30.0 | **75.1** | **55.6** | **30.7** | **14.1** | |
|
|
|
</details> |
|
|
|
<!--### Performance on Preference TODO--> |
|
|
|
### Bilingual Ability (CMMLU & MMLU) |
|
|
|
| **Name** | MMLU | **CMMLU** | |
|
| -------------- | --------- | --------- | |
|
| **Yi-9B-200K** | 65.73 | 71.97 | |
|
| **Fi-9B-200K** | **68.80** | **73.28** | |
|
|
|
|
|
## Current Limitations |
|
|
|
- This version of Fi-9B may not be able to stop generation in some scenarios. I will fix that soon. |
|
- Compared to the original Yi-9B-200K, Fi-9B has degraded ability for code completion. This may be due to the lack of raw code data during instruction tuning. |
