scriptoriumadmin/ReMe
1
0
Fork 0
mirror of https://github.com/agentscope-ai/ReMe.git synced 2026-08-28 05:25:04 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 11

fix typos

Browse source
This commit is contained in:
dengjiaji 2025-09-02 22:04:23 +08:00
parent cd6999bdad
commit 2ea8a950fc
2 changed files with 16 additions and 16 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

16
README.md
View file

@ -345,14 +345,14 @@ response = requests.post("http://localhost:8002/retrieve_task_memory", json={
## 🧪 Experiments
### 🌍 Appworld Experiment
### 🌍 [Appworld Experiment](./cookbook/appworld/quickstart.md)
We tested ReMe on Appworld using qwen3-8b:
| Method | pass@1 | pass@2 | pass@4 |
|--------------|-----------|-----------|-----------|
| without ReMe | 0.083 | 0.140 | 0.228 |
| with ReMe | **0.109** | **0.175** | **0.281** |
| Method | pass@1 | pass@2 | pass@4 |
|--------------|-------------------|-------------------|-------------------|
| without ReMe | 0.083 | 0.140 | 0.228 |
| with ReMe | 0.109 **(+2.6%)** | 0.175 **(+3.5%)** | 0.281 **(+5.3%)** |
Pass@K measures the probability that at least one of the K generated samples successfully completes the task (
score=1).
@ -360,7 +360,7 @@ The current experiment uses an internal AppWorld environment, which may have sli
You can find more details on reproducing the experiment in [quickstart.md](cookbook/appworld/quickstart.md).
### 🧊 Frozenlake Experiment
### 🧊 [Frozenlake Experiment](./cookbook/frozenlake/quickstart.md)
| without ReMe | with ReMe |
|:-------------------------------------------------------------------------------------------:|:-------------------------------------------------------------------------------------------:|
@ -371,11 +371,11 @@ We tested on 100 random frozenlake maps using qwen3-8b:
| Method | pass rate |
|--------------|------------------|
| without ReMe | 0.66 |
| with ReMe | 0.72 **(+9.1%)** |
| with ReMe | 0.72 **(+6.0%)** |
You can find more details on reproducing the experiment in [quickstart.md](cookbook/frozenlake/quickstart.md).
### 🔧 BFCL-V3 Experiment
### 🔧 [BFCL-V3 Experiment](./cookbook/bfcl/quickstart.md)
We tested ReMe on BFCL-V3 multi-turn-base (randomly split 50train/150val) using qwen3-8b:

16
README_ZH.md
View file

@ -327,14 +327,14 @@ response = requests.post("http://localhost:8002/retrieve_task_memory", json={
## 🧪 实验
### 🌍 Appworld 实验
### 🌍 [Appworld 实验](cookbook/appworld/quickstart.md)
我们在 Appworld 上使用 qwen3-8b 测试 ReMe：
| 方法 | pass@1 | pass@2 | pass@4 |
|--------------|-----------|-----------|-----------|
| without ReMe | 0.083 | 0.140 | 0.228 |
| with ReMe | **0.109** | **0.175** | **0.281** |
| 方法 | pass@1 | pass@2 | pass@4 |
|--------------|-------------------|-------------------|-------------------|
| without ReMe | 0.083 | 0.140 | 0.228 |
| with ReMe | 0.109 **(+2.6%)** | 0.175 **(+3.5%)** | 0.281 **(+5.3%)** |
Pass@K 衡量的是在生成的 K 个样本中，至少有一个成功完成任务（score=1）的概率。
当前实验使用的是一个内部的 AppWorld 环境，可能存在轻微差异。
@ -342,7 +342,7 @@ Pass@K 衡量的是在生成的 K 个样本中，至少有一个成功完成任
你可以在 [quickstart.md](cookbook/appworld/quickstart.md) 中找到复现实验的更多细节。
### 🧊 Frozenlake 实验
### 🧊 [Frozenlake 实验](cookbook/frozenlake/quickstart.md)
| 不使用ReMe | 使用ReMe |
|:-------------------------------------------------------------------------------------------:|:-------------------------------------------------------------------------------------------:|
@ -353,11 +353,11 @@ Pass@K 衡量的是在生成的 K 个样本中，至少有一个成功完成任
| 方法 | pass rate |
|--------------|------------------|
| without ReMe | 0.66 |
| with ReMe | 0.72 **(+9.1%)** |
| with ReMe | 0.72 **(+6.0%)** |
你可以在 [quickstart.md](cookbook/frozenlake/quickstart.md) 中找到复现实验的更多细节。
### 🔧 BFCL-V3 实验
### 🔧 [BFCL-V3 实验](cookbook/bfcl/quickstart.md)
我们在 BFCL-V3 multi-turn-base (随机划分50train/150val) 上使用 qwen3-8b 测试 ReMe：