Instructions to use Devil-Assassin/iVision with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Keras
How to use Devil-Assassin/iVision with Keras:
# !pip install -U keras tensorflow huggingface_hub # Keras needs TensorFlow installed to read "hf://" paths, so the tensorflow backend is selected here; # "jax" and "torch" also work for computation once TensorFlow is installed. import os os.environ["KERAS_BACKEND"] = "tensorflow" import keras model = keras.saving.load_model("hf://Devil-Assassin/iVision") - Notebooks
- Google Colab
- Kaggle
Object Detection: Uploaded Legacy Files
Browse filesThe YOLO models that we previously used but now we don't, either due to
issues with the export process used or due to them becoming obsolete in
our app.
Some of the issues were:
- Old YOLO export was incompatible with the new LiteRT API, so
Ultralytics updated their export to use litert-torch instead, and
they fixed inherent incompatibilities with the YOLO model itself, to
be able to export it with full GPU delegation.
- The new export they introduced caused the models to be exported with
NCHW inputs followed by an unnecessary transpose operation, causing
NCHW inputs to turn back to NHWC anyways. At the time, I couldn't
find a solution around it, but later on I did and now we have NHWC
models with no useless transpose operation.
- The new export process removed the nms argument, so the default for
LiteRT exports was that the output had to be postprocessed with NMS
later on. At the time, I forced nms=False anyways but this stopped
full GPU delegation and I wasn't concerned about it at the time, so
I've exported new models with the pre-NMS outputs and applied NMS
post-processing instead.
- COCO: Original YOLO models pretrained on COCO, exported to LiteRT.
- Precision: float32
- Resolution: 320x320 (letterboxed)
- Input: 1x320x320x3 (NHWC)
- Output: 1x300x6 (post-NMS)
- OD v1: YOLO models trained on OD v1 dataset, exported to LiteRT.
- Precision: float32 and w8a32
- Resolution: 320x320 (letterboxed)
- Input: 1x3x320x320 (NCHW)
- Output: 1x300x6 (post-NMS)
- object_detection/_legacy/coco/yolo26n-seg.pt +3 -0
- object_detection/_legacy/coco/yolo26n-seg_f32.tflite +3 -0
- object_detection/_legacy/coco/yolo26n.pt +3 -0
- object_detection/_legacy/coco/yolo26n_f32.tflite +3 -0
- object_detection/_legacy/od_v1/yolo26n-v1-f32.tflite +3 -0
- object_detection/_legacy/od_v1/yolo26n-v1-w8a32.tflite +3 -0
- object_detection/_legacy/od_v1/yolo26n-v1.pt +3 -0
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:361fbfabab285c3237700b6bb91d7ecfa602cd945fffda8dbe1242829b71e73f
|
| 3 |
+
size 6719965
|
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fc0f68168298e396b98105b8eb7db742b358be009c2332f4e2dd289d69e5e9aa
|
| 3 |
+
size 11616703
|
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9b09cc8bf347f0fc8a5f7657480587f25db09b34bf33b0652110fb03a8ad4fef
|
| 3 |
+
size 5544453
|
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a285a655eaa6dd9c8b2cd51146378976295e7b1ed02bf63ff29b3dd418dc3a29
|
| 3 |
+
size 10193240
|
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:37955fa932fef92dde5f562fe079e3e7cbeb6ac6e73014639f8a24fc569ab0d6
|
| 3 |
+
size 10110170
|
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:85e0fdaa3e5e1caad3d507e8e122c5f5c0b23902e652994217213c02a958745a
|
| 3 |
+
size 3032502
|
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8806059c02c42226c42a01bb020a2c2718e3b0fffca9074bb054788fff3f7d70
|
| 3 |
+
size 5448901
|