Commit
•
b3ccab2
1
Parent(s):
6d456f4
Update README.md (#1)
Browse files- Update README.md (a9ced8097f1e3aa55388964804e3492ccf19f3fb)
Co-authored-by: WeihaoZeng <AndrewZeng@users.noreply.huggingface.co>
README.md
CHANGED
|
@@ -25,24 +25,47 @@ Deita 7B V1.0 SFT is a fine-tuned version of Mistral-7B-v0.1 that was trained on
|
|
| 25 |
- **Model Family:** Other models and the dataset are found in the [Deita collection](https://huggingface.co/collections/hkust-nlp/deita-6569c198c174808d94cf5bd4).
|
| 26 |
|
| 27 |
## Performance
|
| 28 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 29 |
|------------------------------------------------|-----------|------------|----------|---------------|----------------|
|
| 30 |
| **Proprietary Models** | | | | | |
|
| 31 |
| GPT-4-Turbo | ? | -- | 9.32 | 97.70 | -- |
|
| 32 |
| GPT-4 | SFT + PPO | -- | 8.99 | 95.03 | -- |
|
| 33 |
| Claude-2 | SFT + PPO | -- | 8.06 | 91.36 | -- |
|
| 34 |
| GPT-3.5-turbo | SFT + PPO | -- | 7.94 | 89.37 | -- |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 35 |
| **Open-sourced Models based on Mistral-7B** | | | | | |
|
| 36 |
| Mistral-7B-Instruct-v0.1 | -- | -- | 6.84 | 69.65 | 60.45 |
|
| 37 |
| Zephyr-7B-sft | SFT | 200K SFT | 5.32 | 75.12 | 60.93 |
|
| 38 |
-
| Zephyr-7B-beta | SFT + DPO | 200K SFT + 60K DPO | 7.34 | 90.60 | 66.36 |
|
| 39 |
-
| OpenChat-3.5 | C-RLFT |
|
| 40 |
-
| Starling-7B | C-RLFT + APA |
|
| 41 |
| Random | SFT | 10K SFT | 5.89 | 56.90 | 61.72 |
|
| 42 |
-
| DEITA-7B-v1.0-sft | SFT | 6K SFT | 7.22 | 80.78 | 64.94 |
|
| 43 |
-
| DEITA-7B-v1.0-sft
|
| 44 |
| DEITA-7B-v1.0 | SFT + DPO | 6K SFT + 10K DPO | 7.55 | 90.06 | 69.86 |
|
| 45 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 46 |
## Input Format
|
| 47 |
|
| 48 |
The model is trained using the [vicuna_v1.1 template](https://github.com/lm-sys/FastChat/blob/main/fastchat/conversation.py)
|
|
|
|
| 25 |
- **Model Family:** Other models and the dataset are found in the [Deita collection](https://huggingface.co/collections/hkust-nlp/deita-6569c198c174808d94cf5bd4).
|
| 26 |
|
| 27 |
## Performance
|
| 28 |
+
|
| 29 |
+
|
| 30 |
+
<details>
|
| 31 |
+
<summary>See full evaluations</summary>
|
| 32 |
+
|
| 33 |
+
| Model | Align | Data Size | MT-Bench | AlpacaEval(%) | OpenLLM (Avg.) |
|
| 34 |
|------------------------------------------------|-----------|------------|----------|---------------|----------------|
|
| 35 |
| **Proprietary Models** | | | | | |
|
| 36 |
| GPT-4-Turbo | ? | -- | 9.32 | 97.70 | -- |
|
| 37 |
| GPT-4 | SFT + PPO | -- | 8.99 | 95.03 | -- |
|
| 38 |
| Claude-2 | SFT + PPO | -- | 8.06 | 91.36 | -- |
|
| 39 |
| GPT-3.5-turbo | SFT + PPO | -- | 7.94 | 89.37 | -- |
|
| 40 |
+
| **Open-sourced Models based on LLaMA-1-13B** | | | | | |
|
| 41 |
+
| LIMA | SFT | 1K SFT | 4.29 | 41.98 | 59.82 |
|
| 42 |
+
| WizardLM-13B | SFT | 70K SFT | 6.35 | 75.31 | 58.96 |
|
| 43 |
+
| Vicuna-13B-v1.3 | SFT | 125K SFT | 6.39 | 82.11 | 60.01 |
|
| 44 |
+
| Random | SFT | 10K SFT | 6.03 | 71.52 | 60.14 |
|
| 45 |
+
| DEITA-LLaMA1-13B-v1.0-sft | SFT | 10K SFT | 6.60 | 78.01 | 64.27 |
|
| 46 |
+
| **Open-sourced Models based on LLaMA-2-13B** | | | | | |
|
| 47 |
+
| Tulu-2-13B | SFT | 326K SFT | 6.70 | 78.90 | -- |
|
| 48 |
+
| Tulu-2-13B+DPO | SFT + DPO | 326K SFT + 60K DPO | 7.00 | 89.50 | -- |
|
| 49 |
+
| LLaMA2-13B-Chat | SFT + PPO | -- | 6.65 | 81.09 | -- |
|
| 50 |
+
| WizardLM-13B-v1.2 | SFT | >70K SFT | 7.09 | 89.17 | -- |
|
| 51 |
+
| Vicuna-13B-v1.5 | SFT | 125K SFT | 6.57 | 78.80 | 61.63 |
|
| 52 |
+
| Random | SFT | 10K SFT | 5.78 | 65.19 | 61.32 |
|
| 53 |
+
| DEITA-LLaMA2-13B-v1.0-sft | SFT | 10K SFT | 6.79 | 81.09 | 62.71 |
|
| 54 |
| **Open-sourced Models based on Mistral-7B** | | | | | |
|
| 55 |
| Mistral-7B-Instruct-v0.1 | -- | -- | 6.84 | 69.65 | 60.45 |
|
| 56 |
| Zephyr-7B-sft | SFT | 200K SFT | 5.32 | 75.12 | 60.93 |
|
| 57 |
+
| $\text{Zephyr-7B-}\beta$ | SFT + DPO | 200K SFT + 60K DPO | 7.34 | 90.60 | 66.36 |
|
| 58 |
+
| OpenChat-3.5 | C-RLFT | >> 70K C-RLFT | 7.81 | 88.51 | -- |
|
| 59 |
+
| Starling-7B | C-RLFT + APA | >>70K C-RLFT + 183K APA | 8.09 | 91.99 | -- |
|
| 60 |
| Random | SFT | 10K SFT | 5.89 | 56.90 | 61.72 |
|
| 61 |
+
| DEITA-7B-v1.0-sft (6K) | SFT | 6K SFT | 7.22 | 80.78 | 64.94 |
|
| 62 |
+
| DEITA-7B-v1.0-sft (10K) | SFT | 10K SFT | 7.32 | 81.67 | 64.00 |
|
| 63 |
| DEITA-7B-v1.0 | SFT + DPO | 6K SFT + 10K DPO | 7.55 | 90.06 | 69.86 |
|
| 64 |
|
| 65 |
+
|
| 66 |
+
</details>
|
| 67 |
+
|
| 68 |
+
|
| 69 |
## Input Format
|
| 70 |
|
| 71 |
The model is trained using the [vicuna_v1.1 template](https://github.com/lm-sys/FastChat/blob/main/fastchat/conversation.py)
|