Replace benchmark tables with link to Edge Inference Benchmarks space
Browse files
README.md
CHANGED
|
@@ -60,41 +60,11 @@ significantly reducing the memory footprint of model weights.
|
|
| 60 |
|
| 61 |
---
|
| 62 |
|
| 63 |
-
##
|
| 64 |
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
> [!WARNING]
|
| 68 |
-
> We have not been able to reproduce the baseline benchmarks reported by [nvidia/Cosmos-Reason2-2B](https://huggingface.co/nvidia/Cosmos-Reason2-2B)
|
| 69 |
-
> on the [Physical AI Bench Leaderboard](https://huggingface.co/spaces/shi-labs/physical-ai-bench-leaderboard),
|
| 70 |
-
> see related issue: https://github.com/nvidia-cosmos/cosmos-reason2/issues/52
|
| 71 |
-
|
| 72 |
-
### Overall + Category Scores
|
| 73 |
-
|
| 74 |
-
| Model | Overall | Embodied Reasoning | Common Sense |
|
| 75 |
-
|---------------------------------------------------------------------------------------------------|--------:|-------------------:|-------------:|
|
| 76 |
-
| [nvidia/Cosmos-Reason2-2B](https://huggingface.co/nvidia/Cosmos-Reason2-2B) | 50.60 | 53.93 | 47.19 |
|
| 77 |
-
| [embedl/Cosmos-Reason2-2B-NVFP4A16](https://huggingface.co/embedl/Cosmos-Reason2-2B-NVFP4A16) | 49.84 | 50.16 | 49.50 |
|
| 78 |
-
| [**embedl/Cosmos-Reason2-2B-W4A16**](https://huggingface.co/embedl/Cosmos-Reason2-2B-W4A16) | 48.68 | 50.49 | 46.85 |
|
| 79 |
-
| [embedl/Cosmos-Reason2-2B-W4A16-Edge2](https://huggingface.co/embedl/Cosmos-Reason2-2B-W4A16-Edge2) | 50.58 | 53.61 | 47.52 |
|
| 80 |
-
|
| 81 |
-
### Subcategory Scores
|
| 82 |
-
|
| 83 |
-
| Model | AV | Physical World | Time | Space | Agibot | HoloAssist | RoboFail | RoboVQA | BridgeData V2 |
|
| 84 |
-
|---------------------------------------------------------------------------------------------------|------:|---------------:|------:|------:|-------:|-----------:|---------:|--------:|--------------:|
|
| 85 |
-
| [nvidia/Cosmos-Reason2-2B](https://huggingface.co/nvidia/Cosmos-Reason2-2B) | 44.00 | 46.90 | 45.30 | 55.00 | 34.00 | 60.00 | 49.00 | 90.91 | 42.00 |
|
| 86 |
-
| [embedl/Cosmos-Reason2-2B-NVFP4A16](https://huggingface.co/embedl/Cosmos-Reason2-2B-NVFP4A16) | 44.00 | 45.13 | 52.01 | 52.50 | 28.00 | 58.00 | 51.00 | 84.55 | 32.00 |
|
| 87 |
-
| [**embedl/Cosmos-Reason2-2B-W4A16**](https://huggingface.co/embedl/Cosmos-Reason2-2B-W4A16) | 36.00 | 47.79 | 44.30 | 53.75 | 36.00 | 61.00 | 42.00 | 80.91 | 44.00 |
|
| 88 |
-
| [embedl/Cosmos-Reason2-2B-W4A16-Edge2](https://huggingface.co/embedl/Cosmos-Reason2-2B-W4A16-Edge2) | 45.00 | 44.25 | 48.66 | 52.50 | 32.00 | 59.00 | 54.00 | 85.45 | 43.00 |
|
| 89 |
-
|
| 90 |
-
---
|
| 91 |
-
|
| 92 |
-
## Performance
|
| 93 |
-
|
| 94 |
-
On-device performance benchmarks can be explored on [embedl/Edge-Inference-Benchmarks](https://huggingface.co/spaces/embedl/Edge-Inference-Benchmarks).
|
| 95 |
|
| 96 |
<img src="https://huggingface.co/datasets/embedl/documentation-images/resolve/main/Cosmos-Reason2-2B-W4A16/screenshot_edge_inference_benchmarks.png" alt="Screenshot Edge Inference Benchmarks" width="75%">
|
| 97 |
-
|
| 98 |
---
|
| 99 |
|
| 100 |
|
|
|
|
| 60 |
|
| 61 |
---
|
| 62 |
|
| 63 |
+
## Benchmarks
|
| 64 |
|
| 65 |
+
Accuracy and on-device latency benchmarks can be explored on [embedl/Edge-Inference-Benchmarks](https://huggingface.co/spaces/embedl/Edge-Inference-Benchmarks).
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 66 |
|
| 67 |
<img src="https://huggingface.co/datasets/embedl/documentation-images/resolve/main/Cosmos-Reason2-2B-W4A16/screenshot_edge_inference_benchmarks.png" alt="Screenshot Edge Inference Benchmarks" width="75%">
|
|
|
|
| 68 |
---
|
| 69 |
|
| 70 |
|