46 lines
1.3 KiB
Markdown
46 lines
1.3 KiB
Markdown
---
|
|
license: apache-2.0
|
|
base_model: laion/r2egym-nl2bash-stack-bugsseq-fixthink-again
|
|
tags:
|
|
- reinforcement-learning
|
|
- code
|
|
- r2egym
|
|
- rl
|
|
- rloo-n
|
|
- terminus-structured
|
|
language:
|
|
- en
|
|
pipeline_tag: text-generation
|
|
library_name: transformers
|
|
---
|
|
|
|
# rl_r2egym-full_terminus-structured
|
|
|
|
RL-trained Qwen3-8B with structured tool calls. Continued from mixed step 37 with full r2egym dataset (1785 tasks) for 18 more steps.
|
|
|
|
SWEBench-100: 42% pass@3. Training pass@8 peaked at 90.6%.
|
|
|
|
## Training Details
|
|
|
|
- **Base model**: [laion/r2egym-nl2bash-stack-bugsseq-fixthink-again](https://huggingface.co/laion/r2egym-nl2bash-stack-bugsseq-fixthink-again)
|
|
- **Training method**: rloo-n with terminus-structured agent (structured tool calls: bash, view, edit, create, search)
|
|
- **Framework**: BenSkyRL + Harbor
|
|
- **Context**: 32k (24k input + 8k output)
|
|
- **Learning rate**: 1e-5
|
|
|
|
## SWEBench-Verified Results (100 tasks, pass@3)
|
|
|
|
| Model | SWEBench pass@3 |
|
|
|---|---|
|
|
| Base SFT (terminus-2) | 37% |
|
|
| This model (terminus-structured) | See eval results |
|
|
|
|
## Usage
|
|
|
|
```python
|
|
from transformers import AutoModelForCausalLM, AutoTokenizer
|
|
|
|
model = AutoModelForCausalLM.from_pretrained("laion/rl_r2egym-full_terminus-structured")
|
|
tokenizer = AutoTokenizer.from_pretrained("laion/rl_r2egym-full_terminus-structured")
|
|
```
|