Commit
·
635b256
verified
·
0
Parent(s):
Super-squash history to reclaim storage
Browse files- .gitattributes +81 -0
- README.md +240 -0
- RoboBrain2.0-7B-bf16.gguf +3 -0
- RoboBrain2.0-7B-bf16.mmproj +3 -0
- RoboBrain2.0-7B-bf16_q8_0.gguf +3 -0
- RoboBrain2.0-7B-f16.mmproj +3 -0
- RoboBrain2.0-7B-f16_q8_0.gguf +3 -0
- RoboBrain2.0-7B-f32.mmproj +3 -0
- RoboBrain2.0-7B-iq2_m.gguf +3 -0
- RoboBrain2.0-7B-iq2_s.gguf +3 -0
- RoboBrain2.0-7B-iq2_xs.gguf +3 -0
- RoboBrain2.0-7B-iq2_xxs.gguf +3 -0
- RoboBrain2.0-7B-iq3_m.gguf +3 -0
- RoboBrain2.0-7B-iq3_s.gguf +3 -0
- RoboBrain2.0-7B-iq3_xs.gguf +3 -0
- RoboBrain2.0-7B-iq3_xxs.gguf +3 -0
- RoboBrain2.0-7B-iq4_nl.gguf +3 -0
- RoboBrain2.0-7B-iq4_xs.gguf +3 -0
- RoboBrain2.0-7B-q2_k_m.gguf +3 -0
- RoboBrain2.0-7B-q2_k_s.gguf +3 -0
- RoboBrain2.0-7B-q3_k_m.gguf +3 -0
- RoboBrain2.0-7B-q3_k_s.gguf +3 -0
- RoboBrain2.0-7B-q4_0.gguf +3 -0
- RoboBrain2.0-7B-q4_1.gguf +3 -0
- RoboBrain2.0-7B-q4_k_m.gguf +3 -0
- RoboBrain2.0-7B-q4_k_s.gguf +3 -0
- RoboBrain2.0-7B-q5_0.gguf +3 -0
- RoboBrain2.0-7B-q5_1.gguf +3 -0
- RoboBrain2.0-7B-q5_k_m.gguf +3 -0
- RoboBrain2.0-7B-q5_k_s.gguf +3 -0
- RoboBrain2.0-7B-q6_k_m.gguf +3 -0
- RoboBrain2.0-7B-q8_0.gguf +3 -0
- RoboBrain2.0-7B-q8_0.mmproj +3 -0
- RoboBrain2.0-7B.imatrix +3 -0
.gitattributes
ADDED
|
@@ -0,0 +1,81 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
*.7z filter=lfs diff=lfs merge=lfs -text
|
| 2 |
+
*.arrow filter=lfs diff=lfs merge=lfs -text
|
| 3 |
+
*.bin filter=lfs diff=lfs merge=lfs -text
|
| 4 |
+
*.bz2 filter=lfs diff=lfs merge=lfs -text
|
| 5 |
+
*.ckpt filter=lfs diff=lfs merge=lfs -text
|
| 6 |
+
*.ftz filter=lfs diff=lfs merge=lfs -text
|
| 7 |
+
*.gz filter=lfs diff=lfs merge=lfs -text
|
| 8 |
+
*.h5 filter=lfs diff=lfs merge=lfs -text
|
| 9 |
+
*.joblib filter=lfs diff=lfs merge=lfs -text
|
| 10 |
+
*.lfs.* filter=lfs diff=lfs merge=lfs -text
|
| 11 |
+
*.mlmodel filter=lfs diff=lfs merge=lfs -text
|
| 12 |
+
*.model filter=lfs diff=lfs merge=lfs -text
|
| 13 |
+
*.msgpack filter=lfs diff=lfs merge=lfs -text
|
| 14 |
+
*.npy filter=lfs diff=lfs merge=lfs -text
|
| 15 |
+
*.npz filter=lfs diff=lfs merge=lfs -text
|
| 16 |
+
*.onnx filter=lfs diff=lfs merge=lfs -text
|
| 17 |
+
*.ot filter=lfs diff=lfs merge=lfs -text
|
| 18 |
+
*.parquet filter=lfs diff=lfs merge=lfs -text
|
| 19 |
+
*.pb filter=lfs diff=lfs merge=lfs -text
|
| 20 |
+
*.pickle filter=lfs diff=lfs merge=lfs -text
|
| 21 |
+
*.pkl filter=lfs diff=lfs merge=lfs -text
|
| 22 |
+
*.pt filter=lfs diff=lfs merge=lfs -text
|
| 23 |
+
*.pth filter=lfs diff=lfs merge=lfs -text
|
| 24 |
+
*.rar filter=lfs diff=lfs merge=lfs -text
|
| 25 |
+
*.safetensors filter=lfs diff=lfs merge=lfs -text
|
| 26 |
+
saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
| 27 |
+
*.tar.* filter=lfs diff=lfs merge=lfs -text
|
| 28 |
+
*.tar filter=lfs diff=lfs merge=lfs -text
|
| 29 |
+
*.tflite filter=lfs diff=lfs merge=lfs -text
|
| 30 |
+
*.tgz filter=lfs diff=lfs merge=lfs -text
|
| 31 |
+
*.wasm filter=lfs diff=lfs merge=lfs -text
|
| 32 |
+
*.xz filter=lfs diff=lfs merge=lfs -text
|
| 33 |
+
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
+
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
+
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
RoboBrain2.0-7B-f16.gguf filter=lfs diff=lfs merge=lfs -text
|
| 37 |
+
RoboBrain2.0-7B-f16_q8_0.gguf filter=lfs diff=lfs merge=lfs -text
|
| 38 |
+
RoboBrain2.0-7B-bf16_q8_0.gguf filter=lfs diff=lfs merge=lfs -text
|
| 39 |
+
RoboBrain2.0-7B-f16_q6_k.gguf filter=lfs diff=lfs merge=lfs -text
|
| 40 |
+
RoboBrain2.0-7B-bf16_q6_k.gguf filter=lfs diff=lfs merge=lfs -text
|
| 41 |
+
RoboBrain2.0-7B-f16_q4_k.gguf filter=lfs diff=lfs merge=lfs -text
|
| 42 |
+
RoboBrain2.0-7B-bf16_q4_k.gguf filter=lfs diff=lfs merge=lfs -text
|
| 43 |
+
RoboBrain2.0-7B-q2_k_l.gguf filter=lfs diff=lfs merge=lfs -text
|
| 44 |
+
RoboBrain2.0-7B-q3_k_l.gguf filter=lfs diff=lfs merge=lfs -text
|
| 45 |
+
RoboBrain2.0-7B-q4_k_l.gguf filter=lfs diff=lfs merge=lfs -text
|
| 46 |
+
RoboBrain2.0-7B-q5_k_l.gguf filter=lfs diff=lfs merge=lfs -text
|
| 47 |
+
RoboBrain2.0-7B-q6_k_l.gguf filter=lfs diff=lfs merge=lfs -text
|
| 48 |
+
RoboBrain2.0-7B-q2_k_m.gguf filter=lfs diff=lfs merge=lfs -text
|
| 49 |
+
RoboBrain2.0-7B-q2_k_s.gguf filter=lfs diff=lfs merge=lfs -text
|
| 50 |
+
RoboBrain2.0-7B-q3_k_m.gguf filter=lfs diff=lfs merge=lfs -text
|
| 51 |
+
RoboBrain2.0-7B-q3_k_s.gguf filter=lfs diff=lfs merge=lfs -text
|
| 52 |
+
RoboBrain2.0-7B-q4_k_m.gguf filter=lfs diff=lfs merge=lfs -text
|
| 53 |
+
RoboBrain2.0-7B-q4_k_s.gguf filter=lfs diff=lfs merge=lfs -text
|
| 54 |
+
RoboBrain2.0-7B-q5_k_m.gguf filter=lfs diff=lfs merge=lfs -text
|
| 55 |
+
RoboBrain2.0-7B-q5_k_s.gguf filter=lfs diff=lfs merge=lfs -text
|
| 56 |
+
RoboBrain2.0-7B-q6_k_m.gguf filter=lfs diff=lfs merge=lfs -text
|
| 57 |
+
RoboBrain2.0-7B-q8_0.gguf filter=lfs diff=lfs merge=lfs -text
|
| 58 |
+
RoboBrain2.0-7B-q4_0.gguf filter=lfs diff=lfs merge=lfs -text
|
| 59 |
+
RoboBrain2.0-7B-q4_1.gguf filter=lfs diff=lfs merge=lfs -text
|
| 60 |
+
RoboBrain2.0-7B-q4_0_l.gguf filter=lfs diff=lfs merge=lfs -text
|
| 61 |
+
RoboBrain2.0-7B-q4_1_l.gguf filter=lfs diff=lfs merge=lfs -text
|
| 62 |
+
RoboBrain2.0-7B-q5_0.gguf filter=lfs diff=lfs merge=lfs -text
|
| 63 |
+
RoboBrain2.0-7B-q5_1.gguf filter=lfs diff=lfs merge=lfs -text
|
| 64 |
+
RoboBrain2.0-7B-q5_0_l.gguf filter=lfs diff=lfs merge=lfs -text
|
| 65 |
+
RoboBrain2.0-7B-q5_1_l.gguf filter=lfs diff=lfs merge=lfs -text
|
| 66 |
+
RoboBrain2.0-7B-iq2_xs.gguf filter=lfs diff=lfs merge=lfs -text
|
| 67 |
+
RoboBrain2.0-7B-iq2_xxs.gguf filter=lfs diff=lfs merge=lfs -text
|
| 68 |
+
RoboBrain2.0-7B-iq2_s.gguf filter=lfs diff=lfs merge=lfs -text
|
| 69 |
+
RoboBrain2.0-7B-iq2_m.gguf filter=lfs diff=lfs merge=lfs -text
|
| 70 |
+
RoboBrain2.0-7B-iq3_xs.gguf filter=lfs diff=lfs merge=lfs -text
|
| 71 |
+
RoboBrain2.0-7B-iq3_xxs.gguf filter=lfs diff=lfs merge=lfs -text
|
| 72 |
+
RoboBrain2.0-7B-iq3_s.gguf filter=lfs diff=lfs merge=lfs -text
|
| 73 |
+
RoboBrain2.0-7B-iq3_m.gguf filter=lfs diff=lfs merge=lfs -text
|
| 74 |
+
RoboBrain2.0-7B-iq4_xs.gguf filter=lfs diff=lfs merge=lfs -text
|
| 75 |
+
RoboBrain2.0-7B-iq4_nl.gguf filter=lfs diff=lfs merge=lfs -text
|
| 76 |
+
RoboBrain2.0-7B.imatrix filter=lfs diff=lfs merge=lfs -text
|
| 77 |
+
RoboBrain2.0-7B-bf16.mmproj filter=lfs diff=lfs merge=lfs -text
|
| 78 |
+
RoboBrain2.0-7B-f32.mmproj filter=lfs diff=lfs merge=lfs -text
|
| 79 |
+
RoboBrain2.0-7B-q8_0.mmproj filter=lfs diff=lfs merge=lfs -text
|
| 80 |
+
RoboBrain2.0-7B-bf16.gguf filter=lfs diff=lfs merge=lfs -text
|
| 81 |
+
RoboBrain2.0-7B-f16.mmproj filter=lfs diff=lfs merge=lfs -text
|
README.md
ADDED
|
@@ -0,0 +1,240 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
library_name: transformers
|
| 4 |
+
pipeline_tag: robotics
|
| 5 |
+
---
|
| 6 |
+
|
| 7 |
+
# <span style="color: #7FFF7F;">RoboBrain2.0-7B GGUF Models</span>
|
| 8 |
+
|
| 9 |
+
|
| 10 |
+
## <span style="color: #7F7FFF;">Model Generation Details</span>
|
| 11 |
+
|
| 12 |
+
This model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`6adc3c3e`](https://github.com/ggerganov/llama.cpp/commit/6adc3c3ebc029af058ac950a8e2a825fdf18ecc6).
|
| 13 |
+
|
| 14 |
+
|
| 15 |
+
|
| 16 |
+
|
| 17 |
+
|
| 18 |
+
---
|
| 19 |
+
|
| 20 |
+
## <span style="color: #7FFF7F;">Quantization Beyond the IMatrix</span>
|
| 21 |
+
|
| 22 |
+
I've been experimenting with a new quantization approach that selectively elevates the precision of key layers beyond what the default IMatrix configuration provides.
|
| 23 |
+
|
| 24 |
+
In my testing, standard IMatrix quantization underperforms at lower bit depths, especially with Mixture of Experts (MoE) models. To address this, I'm using the `--tensor-type` option in `llama.cpp` to manually "bump" important layers to higher precision. You can see the implementation here:
|
| 25 |
+
👉 [Layer bumping with llama.cpp](https://github.com/Mungert69/GGUFModelBuilder/blob/main/model-converter/tensor_list_builder.py)
|
| 26 |
+
|
| 27 |
+
While this does increase model file size, it significantly improves precision for a given quantization level.
|
| 28 |
+
|
| 29 |
+
### **I'd love your feedback—have you tried this? How does it perform for you?**
|
| 30 |
+
|
| 31 |
+
|
| 32 |
+
|
| 33 |
+
|
| 34 |
+
---
|
| 35 |
+
|
| 36 |
+
<a href="https://readyforquantum.com/huggingface_gguf_selection_guide.html" style="color: #7FFF7F;">
|
| 37 |
+
Click here to get info on choosing the right GGUF model format
|
| 38 |
+
</a>
|
| 39 |
+
|
| 40 |
+
---
|
| 41 |
+
|
| 42 |
+
|
| 43 |
+
|
| 44 |
+
<!--Begin Original Model Card-->
|
| 45 |
+
|
| 46 |
+
|
| 47 |
+
<div align="center">
|
| 48 |
+
<img src="https://github.com/FlagOpen/RoboBrain2.0/raw/main/assets/logo2.png" width="500"/>
|
| 49 |
+
</div>
|
| 50 |
+
|
| 51 |
+
# RoboBrain 2.0: See Better. Think Harder. Do Smarter.
|
| 52 |
+
|
| 53 |
+
|
| 54 |
+
<p align="center">
|
| 55 |
+
</a>  ⭐️ <a href="https://superrobobrain.github.io/">Project</a></a>   |   ⭐️ <a href="https://github.com/FlagOpen/RoboBrain2.0">Github</a>   |   🤖 <a href="https://www.modelscope.cn/models/BAAI/RoboBrain2.0-7B/files/">ModelScope</a>   |   📑 <a href="">Technical Report (Coming soon)</a>   |   💬 <a href="./assets/wechat.png">WeChat</a>
|
| 56 |
+
</p>
|
| 57 |
+
<p align="center">
|
| 58 |
+
</a>  🎯 <a href="https://github.com/FlagOpen/RoboOS">RoboOS</a>: An Efficient Open-Source Multi-Robot Coordination System for RoboBrain.
|
| 59 |
+
</p>
|
| 60 |
+
<p align="center">
|
| 61 |
+
</a>  🌍 <a href="https://github.com/FlagOpen/RoboBrain">RoboBrain 1.0</a>: A Unified Brain Model for Robotic Manipulation from Abstract to Concrete.
|
| 62 |
+
</p>
|
| 63 |
+
|
| 64 |
+
## 🔥 Overview
|
| 65 |
+
We are excited to introduce **RoboBrain 2.0**, the most powerful open-source embodied brain model to date. Compared to its predecessor, RoboBrain1.0, our latest version significantly advances multi-agent task planning, spatial reasoning, and closed-loop execution. A detailed technical report will be released soon.
|
| 66 |
+
|
| 67 |
+
<div align="center">
|
| 68 |
+
<img src="https://github.com/FlagOpen/RoboBrain2.0/raw/main/assets/results.png" />
|
| 69 |
+
</div>
|
| 70 |
+
|
| 71 |
+
## 🗞️ News
|
| 72 |
+
- **`2025-06-07`**: 🎉 We highlight the training framework ([FlagScale](https://github.com/FlagOpen/FlagScale)) developed by *BAAI Framework R&D team*, and the evaluation framework ([FlagEvalMM](https://github.com/flageval-baai/FlagEvalMM)) by *BAAI FlagEval team*. Both are used for RoboBrain 2.0.
|
| 73 |
+
- **`2025-06-06`**: 🤗 [RoboBrain 2.0-7B](https://huggingface.co/BAAI/RoboBrain2.0-7B) model checkpoint has been released in Huggingface..
|
| 74 |
+
- **`2025-06-06`**: 🔥 We're excited to announce the release of our more powerful [RoboBrain 2.0](https://superrobobrain.github.io/).
|
| 75 |
+
- **`2025-04-11`**: 🎉 [RoboBrain 1.0](https://github.com/FlagOpen/RoboBrain/) was selected for CVPR 2025's official [Embodied AI Trends Commentary](https://cvpr.thecvf.com/Conferences/2025/News/AI_Enhanced_Robotics).
|
| 76 |
+
- **`2025-02-27`**: 🌍 [RoboBrain 1.0](http://arxiv.org/abs/2502.21257/) was accepted to CVPR2025.
|
| 77 |
+
|
| 78 |
+
|
| 79 |
+
## 📆 Todo
|
| 80 |
+
- [x] Release model checkpoint for RoboBrain 2.0-7B
|
| 81 |
+
- [x] Release quick inference example for RoboBrain 2.0
|
| 82 |
+
- [x] Release training codes for RoboBrain 2.0
|
| 83 |
+
- [ ] Release model checkpoint for RoboBrain 2.0-32B
|
| 84 |
+
|
| 85 |
+
## 🚀 Features
|
| 86 |
+
**RoboBrain 2.0** supports ***interactive reasoning*** with long-horizon planning and closed-loop feedback, ***spatial perception*** for precise point and bbox prediction from complex instructions, ***temporal perception*** for future trajectory estimation, and ***scene reasoning*** through real-time structured memory construction and update.
|
| 87 |
+
|
| 88 |
+
<div align="center">
|
| 89 |
+
<img src="https://github.com/FlagOpen/RoboBrain2.0/raw/main/assets/visualization.png" />
|
| 90 |
+
</div>
|
| 91 |
+
|
| 92 |
+
|
| 93 |
+
## ⭐️ Architecture
|
| 94 |
+
**RoboBrain 2.0** supports ***multi-image***, ***long video***, and ***high-resolution visual inputs***, along with complex task instructions and structured ***scene graphs*** on the language side. Visual inputs are processed via a Vision Encoder and MLP Projector, while textual inputs are tokenized into a unified token stream. All inputs are fed into a ***LLM Decoder*** that performs ***long-chain-of-thought reasoning*** and outputs structured plans, spatial relations, and both ***relative*** and ***absolute coordinates***.
|
| 95 |
+
|
| 96 |
+
<div align="center">
|
| 97 |
+
<img src="https://github.com/FlagOpen/RoboBrain2.0/raw/main/assets/arch.png" />
|
| 98 |
+
</div>
|
| 99 |
+
|
| 100 |
+
|
| 101 |
+
## 🤗 Model Zoo
|
| 102 |
+
|
| 103 |
+
|
| 104 |
+
| Models | Checkpoint | Description |
|
| 105 |
+
|----------------------|----------------------------------------------------------------|------------------------------------------------------------|
|
| 106 |
+
| RoboBrain 2.0 7B | [🤗 BAAI/RoboBrain2.0-7B](https://huggingface.co/BAAI/RoboBrain2.0-7B) | 7B parameter version of the RoboBrain2.0 |
|
| 107 |
+
| RoboBrain 2.0 32B | [🤗 BAAI/RoboBrain2.0-32B](https://huggingface.co/BAAI/RoboBrain2.0-32B) | 32B parameter version of the RoboBrain2.0 *(Coming soon)*
|
| 108 |
+
|
| 109 |
+
|
| 110 |
+
## 🛠️ Setup
|
| 111 |
+
|
| 112 |
+
```bash
|
| 113 |
+
# clone repo.
|
| 114 |
+
git clone https://github.com/FlagOpen/RoboBrain2.0.git
|
| 115 |
+
cd RoboBrain
|
| 116 |
+
|
| 117 |
+
# build conda env.
|
| 118 |
+
conda create -n robobrain2 python=3.10
|
| 119 |
+
conda activate robobrain2
|
| 120 |
+
pip install -r requirements.txt
|
| 121 |
+
```
|
| 122 |
+
|
| 123 |
+
|
| 124 |
+
## 🤖 Simple Inference
|
| 125 |
+
**Note: Please refer to [RoboBrain 2.0 Github](https://github.com/FlagOpen/RoboBrain2.0) for the usage of RoboBrain 2.0**
|
| 126 |
+
|
| 127 |
+
|
| 128 |
+
## 😊 More Results
|
| 129 |
+
|
| 130 |
+
**Benchmark comparison across spatial reasoning and task planning.** ***RoboBrain2.0-32B*** achieves state-of-the-art performance on four key embodied intelligence benchmarks: ***BLINK-Spatial***, ***CV-Bench***, ***EmbSpatial***, and ***RefSpatial***. It not only outperforms leading open-source models such as o4-mini and Qwen2.5-VL, but also surpasses closed-source models like Gemini 2.5 Pro and Claude Sonnet 4 — especially in the challenging ***RefSpatial*** benchmark, where ***RoboBrain2.0*** shows a >50% absolute improvement.
|
| 131 |
+
|
| 132 |
+
<div align="center">
|
| 133 |
+
<img src="https://github.com/FlagOpen/RoboBrain2.0/raw/main/assets/results_table.png" />
|
| 134 |
+
</div>
|
| 135 |
+
|
| 136 |
+
|
| 137 |
+
## 📑 Citation
|
| 138 |
+
If you find this project useful, welcome to cite us.
|
| 139 |
+
```bib
|
| 140 |
+
@article{RoboBrain 2.0 Technical Report,
|
| 141 |
+
title={RoboBrain 2.0 Technical Report},
|
| 142 |
+
author={BAAI RoboBrain Team},
|
| 143 |
+
journal={arXiv preprint arXiv:TODO},
|
| 144 |
+
year={2025}
|
| 145 |
+
}
|
| 146 |
+
|
| 147 |
+
@article{RoboBrain 1.0,
|
| 148 |
+
title={Robobrain: A unified brain model for robotic manipulation from abstract to concrete},
|
| 149 |
+
author={Ji, Yuheng and Tan, Huajie and Shi, Jiayu and Hao, Xiaoshuai and Zhang, Yuan and Zhang, Hengyuan and Wang, Pengwei and Zhao, Mengdi and Mu, Yao and An, Pengju and others},
|
| 150 |
+
journal={arXiv preprint arXiv:2502.21257},
|
| 151 |
+
year={2025}
|
| 152 |
+
}
|
| 153 |
+
|
| 154 |
+
@article{RoboOS,
|
| 155 |
+
title={RoboOS: A Hierarchical Embodied Framework for Cross-Embodiment and Multi-Agent Collaboration},
|
| 156 |
+
author={Tan, Huajie and Hao, Xiaoshuai and Lin, Minglan and Wang, Pengwei and Lyu, Yaoxu and Cao, Mingyu and Wang, Zhongyuan and Zhang, Shanghang},
|
| 157 |
+
journal={arXiv preprint arXiv:2505.03673},
|
| 158 |
+
year={2025}
|
| 159 |
+
}
|
| 160 |
+
|
| 161 |
+
@article{zhou2025roborefer,
|
| 162 |
+
title={RoboRefer: Towards Spatial Referring with Reasoning in Vision-Language Models for Robotics},
|
| 163 |
+
author={Zhou, Enshen and An, Jingkun and Chi, Cheng and Han, Yi and Rong, Shanyu and Zhang, Chi and Wang, Pengwei and Wang, Zhongyuan and Huang, Tiejun and Sheng, Lu and others},
|
| 164 |
+
journal={arXiv preprint arXiv:2506.04308},
|
| 165 |
+
year={2025}
|
| 166 |
+
}
|
| 167 |
+
|
| 168 |
+
@article{Reason-RFT,
|
| 169 |
+
title={Reason-rft: Reinforcement fine-tuning for visual reasoning},
|
| 170 |
+
author={Tan, Huajie and Ji, Yuheng and Hao, Xiaoshuai and Lin, Minglan and Wang, Pengwei and Wang, Zhongyuan and Zhang, Shanghang},
|
| 171 |
+
journal={arXiv preprint arXiv:2503.20752},
|
| 172 |
+
year={2025}
|
| 173 |
+
}
|
| 174 |
+
|
| 175 |
+
@article{Code-as-Monitor,
|
| 176 |
+
title={Code-as-Monitor: Constraint-aware Visual Programming for Reactive and Proactive Robotic Failure Detection},
|
| 177 |
+
author={Zhou, Enshen and Su, Qi and Chi, Cheng and Zhang, Zhizheng and Wang, Zhongyuan and Huang, Tiejun and Sheng, Lu and Wang, He},
|
| 178 |
+
journal={arXiv preprint arXiv:2412.04455},
|
| 179 |
+
year={2024}
|
| 180 |
+
}
|
| 181 |
+
```
|
| 182 |
+
|
| 183 |
+
<!--End Original Model Card-->
|
| 184 |
+
|
| 185 |
+
---
|
| 186 |
+
|
| 187 |
+
# <span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>
|
| 188 |
+
|
| 189 |
+
Help me test my **AI-Powered Quantum Network Monitor Assistant** with **quantum-ready security checks**:
|
| 190 |
+
|
| 191 |
+
👉 [Quantum Network Monitor](https://readyforquantum.com/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme)
|
| 192 |
+
|
| 193 |
+
|
| 194 |
+
The full Open Source Code for the Quantum Network Monitor Service available at my github repos ( repos with NetworkMonitor in the name) : [Source Code Quantum Network Monitor](https://github.com/Mungert69). You will also find the code I use to quantize the models if you want to do it yourself [GGUFModelBuilder](https://github.com/Mungert69/GGUFModelBuilder)
|
| 195 |
+
|
| 196 |
+
💬 **How to test**:
|
| 197 |
+
Choose an **AI assistant type**:
|
| 198 |
+
- `TurboLLM` (GPT-4.1-mini)
|
| 199 |
+
- `HugLLM` (Hugginface Open-source models)
|
| 200 |
+
- `TestLLM` (Experimental CPU-only)
|
| 201 |
+
|
| 202 |
+
### **What I’m Testing**
|
| 203 |
+
I’m pushing the limits of **small open-source models for AI network monitoring**, specifically:
|
| 204 |
+
- **Function calling** against live network services
|
| 205 |
+
- **How small can a model go** while still handling:
|
| 206 |
+
- Automated **Nmap security scans**
|
| 207 |
+
- **Quantum-readiness checks**
|
| 208 |
+
- **Network Monitoring tasks**
|
| 209 |
+
|
| 210 |
+
🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads on huggingface docker space):
|
| 211 |
+
- ✅ **Zero-configuration setup**
|
| 212 |
+
- ⏳ 30s load time (slow inference but **no API costs**) . No token limited as the cost is low.
|
| 213 |
+
- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate!
|
| 214 |
+
|
| 215 |
+
### **Other Assistants**
|
| 216 |
+
🟢 **TurboLLM** – Uses **gpt-4.1-mini** :
|
| 217 |
+
- **It performs very well but unfortunatly OpenAI charges per token. For this reason tokens usage is limited.
|
| 218 |
+
- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**
|
| 219 |
+
- **Real-time network diagnostics and monitoring**
|
| 220 |
+
- **Security Audits**
|
| 221 |
+
- **Penetration testing** (Nmap/Metasploit)
|
| 222 |
+
|
| 223 |
+
🔵 **HugLLM** – Latest Open-source models:
|
| 224 |
+
- 🌐 Runs on Hugging Face Inference API. Performs pretty well using the lastest models hosted on Novita.
|
| 225 |
+
|
| 226 |
+
### 💡 **Example commands you could test**:
|
| 227 |
+
1. `"Give me info on my websites SSL certificate"`
|
| 228 |
+
2. `"Check if my server is using quantum safe encyption for communication"`
|
| 229 |
+
3. `"Run a comprehensive security audit on my server"`
|
| 230 |
+
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a [Quantum Network Monitor Agent](https://readyforquantum.com/Download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) to run the .net code on. This is a very flexible and powerful feature. Use with caution!
|
| 231 |
+
|
| 232 |
+
### Final Word
|
| 233 |
+
|
| 234 |
+
I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.
|
| 235 |
+
|
| 236 |
+
If you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.
|
| 237 |
+
|
| 238 |
+
I'm also open to job opportunities or sponsorship.
|
| 239 |
+
|
| 240 |
+
Thank you! 😊
|
RoboBrain2.0-7B-bf16.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:655490fbc5dee94943c07c71645ddb975b7ab880e159f6fb9d6ab7e757a3ae41
|
| 3 |
+
size 15237851872
|
RoboBrain2.0-7B-bf16.mmproj
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0249d890288fa57a04b6813750fb905f76ebf3180c95381cf771c465985a1335
|
| 3 |
+
size 1354162912
|
RoboBrain2.0-7B-bf16_q8_0.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7fc56a68b23deffd53d7cdd58c80239a81490afadfcdf37c3587ab4b1a440367
|
| 3 |
+
size 11287997152
|
RoboBrain2.0-7B-f16.mmproj
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:db0de65afd17803b5e50eb7571d65d92f5f4fc36e9d49cf84ec83fefd07f158d
|
| 3 |
+
size 1354162912
|
RoboBrain2.0-7B-f16_q8_0.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:db0053bcd984d2397f0e4b14ba4666015935b36a299e749b5001282fad00195c
|
| 3 |
+
size 11287997152
|
RoboBrain2.0-7B-f32.mmproj
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:df6fcfd7c65ff2e7071bdf29ec0fcec7387601bcb07084c35ff01a5fd092b796
|
| 3 |
+
size 2703221472
|
RoboBrain2.0-7B-iq2_m.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:69274248612a28c7471bbf7ef85146c2ec8a08d3c29215e3d0b0ebfae5b9629e
|
| 3 |
+
size 3039120384
|
RoboBrain2.0-7B-iq2_s.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9967dcc724182a64e375fb580ebd99ff1e2ed5c5cbf03fc02bf6ae4555cef4fb
|
| 3 |
+
size 2912963584
|
RoboBrain2.0-7B-iq2_xs.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:72bc59edf765d5d3d99ff041bbba39138efb20a9d10ece78c8596a99d8bcc122
|
| 3 |
+
size 2839333888
|
RoboBrain2.0-7B-iq2_xxs.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2d1b5368670ff265b135d76cdf850246afec28c7fa31601956400d7565424a6d
|
| 3 |
+
size 2650901504
|
RoboBrain2.0-7B-iq3_m.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e79bc586ea029a15795fc3163e85219f4c119b7d1990b65a5dfccf82b69e8a5f
|
| 3 |
+
size 3603385344
|
RoboBrain2.0-7B-iq3_s.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bdfacd977be38fe4ca3151b5cf7a98feed65036ec0d8e01aa0020d017f252c02
|
| 3 |
+
size 3565853696
|
RoboBrain2.0-7B-iq3_xs.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b487c8b52b01168b0a881a54f072a283a7751cd75c9e85d4da0b38c456576a50
|
| 3 |
+
size 3412917248
|
RoboBrain2.0-7B-iq3_xxs.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2715b3b51f8e3a3180938f54c5071f7f5dcabda58d7000e33c41bff710a98d47
|
| 3 |
+
size 3272653824
|
RoboBrain2.0-7B-iq4_nl.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c1906d83863b65ec45f8428c0408497a40f8c88c33b4af330b9333aa86a4bb06
|
| 3 |
+
size 4437812224
|
RoboBrain2.0-7B-iq4_xs.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7ca18579af31ed713a9908ef83c37e9a4b0dac1b29ff01b6352f30a95767fd6e
|
| 3 |
+
size 4218471424
|
RoboBrain2.0-7B-q2_k_m.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8aaa1fd0ad8497267385a84bc08abaa25b94ec3fb7f71283c79aee5086f5d906
|
| 3 |
+
size 3259493376
|
RoboBrain2.0-7B-q2_k_s.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b176eb7e3a72df5e73c641ba4027630ea88d6e196baff5dce01370c1456ca34
|
| 3 |
+
size 2934324224
|
RoboBrain2.0-7B-q3_k_m.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:634763598ed6ef2c0ef6c42ed472ac12900932cdf21d04ccbe44e34722de3dc6
|
| 3 |
+
size 3996851200
|
RoboBrain2.0-7B-q3_k_s.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:33eac9546adb8f64f37116b9de7da86cd86ec2f48f41d18eb47e012ffce337e1
|
| 3 |
+
size 3610811392
|
RoboBrain2.0-7B-q4_0.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:32356fa3017cd77ee49da6a47df9f759e18ccb8529de62ba6429b0767e537136
|
| 3 |
+
size 4290882560
|
RoboBrain2.0-7B-q4_1.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2781eb5f6f5fa8b4a3ec63a3626b826a6ae426a8ac3ea92d25e07e26712926b8
|
| 3 |
+
size 4766837760
|
RoboBrain2.0-7B-q4_k_m.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f0a68d073f50762b2a70c894adb648f64503ef9a9a311ae4aa6a577103d6f1c0
|
| 3 |
+
size 4777647104
|
RoboBrain2.0-7B-q4_k_s.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:15847f2218fd6347a4a86636d2ad5de5b6744c3be09dceaebadfe708aad126dc
|
| 3 |
+
size 4634057728
|
RoboBrain2.0-7B-q5_0.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bf9ceb025a1906ca97d645f7a9c6b952b517deeaabafef0d9b54928ee55b529a
|
| 3 |
+
size 5242792960
|
RoboBrain2.0-7B-q5_1.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:345743fd2de825f6d6207e03046120f3b3bd3efd16017e42251b6d400d534fd8
|
| 3 |
+
size 5718748160
|
RoboBrain2.0-7B-q5_k_m.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:da08184937fdd15c575952619c0082f41a84545e3dccf9a2fcf0fcc9a5c0a6cd
|
| 3 |
+
size 5527448576
|
RoboBrain2.0-7B-q5_k_s.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6b6eaccad846e4133675d96186ef7136c184d6bc87ebeb65deed4079545525a8
|
| 3 |
+
size 5453360128
|
RoboBrain2.0-7B-q6_k_m.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0dba596ccec08f5c32e70ef59a87c438dee44ca729310d28c6030171ce89d37d
|
| 3 |
+
size 6254197760
|
RoboBrain2.0-7B-q8_0.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bbf5dd95f5e33f17b9ce6a4efc6e7d47a2ebf807a95ffa1f1a6a18a48f3132da
|
| 3 |
+
size 8098523872
|
RoboBrain2.0-7B-q8_0.mmproj
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4d9606b484a1ae69f5f4672dfbeddcc39f25a3e192b58e404d49aeadb360e3b0
|
| 3 |
+
size 853119712
|
RoboBrain2.0-7B.imatrix
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e7b97ca196f8517153c3de04b18f021d31e90a6910fea85807e4f08ca2e430f5
|
| 3 |
+
size 4536678
|