File size: 1,835 Bytes
f8e06c8 31e76d1 f8e06c8 31e76d1 f8e06c8 31e76d1 f8e06c8 31e76d1 f8e06c8 31e76d1 f8e06c8 31e76d1 f8e06c8 31e76d1 f8e06c8 31e76d1 f8e06c8 31e76d1 f8e06c8 31e76d1 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 | ---
library_name: transformers
base_model: HuggingFaceTB/SmolLM2-360M-Instruct
tags:
- agenttune
- grpo
- trl
- adapter
---
<div align="center">
<table border="0" cellspacing="0" cellpadding="0" style="border: none; border-collapse: collapse;">
<tr>
<td align="center" style="border: none; vertical-align: middle;">
<a href="https://lexsi.ai/"><img src="https://huggingface.co/ram-lexsi/agenttune-testrun-vector-memory/resolve/main/lexsi_logo.png" alt="Lexsi Labs" style="height: 60px; border-radius: 12px;"/></a>
</td>
<td align="center" style="border: none; vertical-align: middle;">
<a href="https://github.com/Lexsi-Labs/AgentTune_mirror"><img src="https://huggingface.co/ram-lexsi/agenttune-testrun-vector-memory/resolve/main/agenttune_banner.png" alt="AgentTune" style="height: 60px;"/></a>
</td>
</tr>
</table>
</div>
# agenttune-testrun-vector-memory
Built using [AgentTune](https://github.com/Lexsi-Labs/AgentTune_mirror) — agentic workflows, then train / evaluate / distill / self-heal through one trajectory schema.
| | |
|---|---|
| **Finetuned from** | [HuggingFaceTB/SmolLM2-360M-Instruct](https://huggingface.co/HuggingFaceTB/SmolLM2-360M-Instruct) |
| **Algorithm** | grpo |
| **Backend** | trl |
| **Artifact** | adapter |
| **Published** | 2026-09-09 08:32 UTC |
## Usage
```python
from peft import AutoPeftModelForCausalLM
from transformers import AutoTokenizer
model = AutoPeftModelForCausalLM.from_pretrained("ram-lexsi/agenttune-testrun-vector-memory")
tokenizer = AutoTokenizer.from_pretrained("ram-lexsi/agenttune-testrun-vector-memory")
```
This repo is a LoRA adapter. Load it on top of `HuggingFaceTB/SmolLM2-360M-Instruct` (PEFT does that from `adapter_config.json`).
Community: [Lexsi Discord](https://discord.com/invite/dtEDQ2Z3eg)
|