Commit
•
315f3f6
1
Parent(s):
93e0bdb
Update README.md
Browse files
README.md
CHANGED
|
@@ -6,6 +6,8 @@ license: apache-2.0
|
|
| 6 |
|
| 7 |
# Model Card for Deita Complexity Scorer
|
| 8 |
|
|
|
|
|
|
|
| 9 |
Deita is an open-sourced project designed to facilitate **Automatic Data Selection** for instruction tuning in Large Language Models (LLMs).
|
| 10 |
|
| 11 |
Deita Complexity Scorer is a tool for automatically annotating the Instruction Complexity of SFT data.
|
|
@@ -24,8 +26,7 @@ Deita Complexity Scorer is a tool for automatically annotating the Instruction C
|
|
| 24 |
## Performance
|
| 25 |
|
| 26 |
|
| 27 |
-
|
| 28 |
-
<summary>See full evaluations</summary>
|
| 29 |
|
| 30 |
| Model | Align | Data Size | MT-Bench | AlpacaEval(%) | OpenLLM (Avg.) |
|
| 31 |
|------------------------------------------------|-----------|------------|----------|---------------|----------------|
|
|
@@ -60,7 +61,7 @@ Deita Complexity Scorer is a tool for automatically annotating the Instruction C
|
|
| 60 |
| DEITA-7B-v1.0 | SFT + DPO | 6K SFT + 10K DPO | 7.55 | 90.06 | 69.86 |
|
| 61 |
|
| 62 |
|
| 63 |
-
|
| 64 |
|
| 65 |
|
| 66 |
|
|
|
|
| 6 |
|
| 7 |
# Model Card for Deita Complexity Scorer
|
| 8 |
|
| 9 |
+
[GitHub](https://github.com/hkust-nlp/deita) | [Paper](https://arxiv.org/abs/2312.15685)
|
| 10 |
+
|
| 11 |
Deita is an open-sourced project designed to facilitate **Automatic Data Selection** for instruction tuning in Large Language Models (LLMs).
|
| 12 |
|
| 13 |
Deita Complexity Scorer is a tool for automatically annotating the Instruction Complexity of SFT data.
|
|
|
|
| 26 |
## Performance
|
| 27 |
|
| 28 |
|
| 29 |
+
|
|
|
|
| 30 |
|
| 31 |
| Model | Align | Data Size | MT-Bench | AlpacaEval(%) | OpenLLM (Avg.) |
|
| 32 |
|------------------------------------------------|-----------|------------|----------|---------------|----------------|
|
|
|
|
| 61 |
| DEITA-7B-v1.0 | SFT + DPO | 6K SFT + 10K DPO | 7.55 | 90.06 | 69.86 |
|
| 62 |
|
| 63 |
|
| 64 |
+
|
| 65 |
|
| 66 |
|
| 67 |
|