Update README.md
Browse files
README.md
CHANGED
|
@@ -24,6 +24,8 @@ of assistant conversations collected through the
|
|
| 24 |
[https://open-assistant.io/](https://open-assistant.io/) human feedback web
|
| 25 |
app before April 12, 2023.
|
| 26 |
|
|
|
|
|
|
|
| 27 |
## Model Details
|
| 28 |
|
| 29 |
- **Developed by:** [Open-Assistant Contributors](https://open-assistant.io/) and [iKala](https://ikala.ai/)
|
|
@@ -49,8 +51,8 @@ start generating the assistant reply.
|
|
| 49 |
|
| 50 |
| model | MMLU | BBH | Humaneval @10 |
|
| 51 |
|---|---|---|---|
|
| 52 |
-
| ikala/redpajama-3b-chat | 24.6 | 29.3 | 4.
|
| 53 |
-
| ikala/bloom-zh-chat-3b | 31.4 | 30.
|
| 54 |
| llama-7b (reference) | 30.9 | 27.6 | 10.3 |
|
| 55 |
|
| 56 |
## Dev Details
|
|
|
|
| 24 |
[https://open-assistant.io/](https://open-assistant.io/) human feedback web
|
| 25 |
app before April 12, 2023.
|
| 26 |
|
| 27 |
+
supervised finetune on sequence of 5120 conversation dataset
|
| 28 |
+
|
| 29 |
## Model Details
|
| 30 |
|
| 31 |
- **Developed by:** [Open-Assistant Contributors](https://open-assistant.io/) and [iKala](https://ikala.ai/)
|
|
|
|
| 51 |
|
| 52 |
| model | MMLU | BBH | Humaneval @10 |
|
| 53 |
|---|---|---|---|
|
| 54 |
+
| ikala/redpajama-3b-chat | 24.6 | 29.3 | 4.8 |
|
| 55 |
+
| ikala/bloom-zh-chat-3b | 31.4 | 30.2 | 0.0 |
|
| 56 |
| llama-7b (reference) | 30.9 | 27.6 | 10.3 |
|
| 57 |
|
| 58 |
## Dev Details
|