Commit
•
c601b4e
1
Parent(s):
cb4c9bf
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,12 +1,36 @@
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
base_model: tokyotech-llm/Swallow-MX-8x7b-NVE-v0.1
|
| 4 |
tags:
|
| 5 |
- generated_from_trainer
|
| 6 |
-
model-index:
|
| 7 |
-
- name: outputs/basemodel-swallowmx-8x22b
|
| 8 |
-
results: []
|
| 9 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
|
| 11 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 12 |
should probably proofread and complete it, then remove this comment. -->
|
|
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
| 3 |
+
language:
|
| 4 |
+
- ja
|
| 5 |
+
- en
|
| 6 |
+
datasets:
|
| 7 |
+
- augmxnt/ultra-orca-boros-en-ja-v1
|
| 8 |
base_model: tokyotech-llm/Swallow-MX-8x7b-NVE-v0.1
|
| 9 |
tags:
|
| 10 |
- generated_from_trainer
|
|
|
|
|
|
|
|
|
|
| 11 |
---
|
| 12 |
+
shisa-v2 Base Model ablation
|
| 13 |
+
|
| 14 |
+
Using a [fork](https://github.com/shisa-ai/shaberi) of [Lightblue's Shaberi benchmark framework](https://github.com/lightblue-tech/japanese_llm_eval):
|
| 15 |
+
|
| 16 |
+
| Model | Average | ELYZA-tasks-100 | MT-Bench | Rakuda | Tengu-Bench |
|
| 17 |
+
|----------------------------------------|---------|-----------------|----------|--------|-------------|
|
| 18 |
+
| gpt-4-turbo-2024-04-09 | 8.75 | 8.78 | 8.74 | 9.18 | 8.31 |
|
| 19 |
+
| CohereForAI/c4ai-command-r-plus | 7.69 | 7.50 | 7.43 | 9.05 | 6.79 |
|
| 20 |
+
| karakuri-ai/karakuri-lm-70b-chat-v0.1 | 6.84 | 6.86 | 6.43 | 7.85 | 6.23 |
|
| 21 |
+
| lightblue/ao-karasu-72B | 6.81 | 7.19 | 6.54 | 7.25 | 6.27 |
|
| 22 |
+
| shisa-ai/shisa-llama3-8b-v1^ | 6.29 | 6.62 | 6.41 | 7.05 | 5.07 |
|
| 23 |
+
| **shisa-ai/shisa-swallowmx-13a47b-v1^**| **6.17**| **6.48** | **6.07** | **7.11**| **5.03** |
|
| 24 |
+
| Rakuten/RakutenAI-7B-chat | 5.58 | 5.92 | 4.60 | 6.58 | 5.24 |
|
| 25 |
+
| shisa-ai/shisa-gemma-7b-v1 | 5.64 | 6.50 | 5.42 | 5.10 | 5.55 |
|
| 26 |
+
| augmxnt/shisa-gamma-7b-v1 | 5.56 | 5.84 | 4.00 | 6.73 | 5.68 |
|
| 27 |
+
| lightblue/qarasu-14B-chat-plus-unleashed | 5.20 | 5.58 | 4.74 | 5.46 | 5.01 |
|
| 28 |
+
| cyberagent/calm2-7b-chat | 4.76 | 4.90 | 3.58 | 5.75 | 4.81 |
|
| 29 |
+
| mistralai/Mistral-7B-Instruct-v0.2 | 4.69 | 5.78 | 4.65 | 3.80 | 4.53 |
|
| 30 |
+
| shisa-ai/shisa-yi1.5-9b-v1 | 4.63 | 5.98 | 4.28 | 3.26 | 5.00 |
|
| 31 |
+
|
| 32 |
+
^ Sampler settings: temperature 0.2, min_p 0.1, frequency_penalty 0.5
|
| 33 |
+
|
| 34 |
|
| 35 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 36 |
should probably proofread and complete it, then remove this comment. -->
|