Instructions to use StephanST/C-radiov4_quantized with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use StephanST/C-radiov4_quantized with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir C-radiov4_quantized StephanST/C-radiov4_quantized
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
Upload h/8bit-affine/README.md with huggingface_hub
Browse files- h/8bit-affine/README.md +1 -1
h/8bit-affine/README.md
CHANGED
|
@@ -56,7 +56,7 @@ Smoke-image 8-bit versus bf16 at `512x512`:
|
|
| 56 |
|
| 57 |
## Measured Speed
|
| 58 |
|
| 59 |
-
Packed
|
| 60 |
|
| 61 |
| p50 latency | Throughput |
|
| 62 |
| ---: | ---: |
|
|
|
|
| 56 |
|
| 57 |
## Measured Speed
|
| 58 |
|
| 59 |
+
Packed weight-only runtime, fast-kernel compiled-forward MLX at `512x512`, batch 1:
|
| 60 |
|
| 61 |
| p50 latency | Throughput |
|
| 62 |
| ---: | ---: |
|