-
**Describe the bug**
I have an Image classification model that was trained using Microsoft CustomVision and exported as an ONNX model. I am able to run inferencing using this model with an average in…
-
I'm trying to install Sockeye with Horovod but in order to do so, I need access to the internet and access to CUDA/`nvcc`. The requirement seems to be mutually exclusive on Trixie. On the head node …
-
### System Info
test-embeddings-inference==v1.5.0
python==3.9
Run with CPU device
### Information
- [ ] Docker
- [X] The CLI directly
### Tasks
- [X] An officially supported command
…
-
### Bug description
### Steps/Code to reproduce bug
1. I have four partition in 4
train1/part_0.parquet
train1/part_1.parquet
train1/part_2.parquet
train1/part_3.parquet
and four gpus
I…
-
Using cntk 2.7, pyhton 3.5
I tried to export a model that I trained with Brainscript (ResNet50_ImageNet1K.cntk) to ONNX with this code:
import cntk as C
model_path = "ResNet_50_train_CNTK.mod…
-
One thing I found tedious is the runtime check the dimensionality match, for example, when writing a neural network, if you want to be safe, you need to check if the dimensionality of each layer match…
-
### 🚀 The feature, motivation and pitch
Due to how `FlattenParamsWrapper` is used in the current FSDP implementation, there doesn't seem to be a straightforward way to shard a parameter which doesn…
-
I want to simply make a RNN model like below. many2many (one output for each input)
```
def create_model(inp):
with C.layers.default_options(initial_state = 0.1):
m = Embedding(300)(…
-
Hi!
Wanted to be able to send longer inputs (of the range 8k input tokens) to a 7b mpt based model and hence switched from a single GPU instance (AWS sagemaker g5.2xlarge, 24GB GPU mem) to a multi-GP…
-
Before training, the function val evaluate the initial model on the NYUv2 test set, and the result is
abs_rel | sq_rel | rmse | rmse_log | a1 | a2 | a3 |
& 0.323 & 0.4…