-
due to the fact that we use a multidimentional tensor of embeddings made out of all the possible permutations of control points of all interpolation expressions to interpolate a prompt, using more tha…
-
### Problem Description
There is no built in method for converting a Qobj into a numpy array such that the `shape` of the numpy array reflects the `dims` of the Qobj. While fixing this is a oneliner …
-
## 🚀 Feature
Codegen tensor could support stride order, instead of assuming/requiring descending stride on domains.
## Motivation
This request arises from our channels last PR #1118. Where cu…
-
# New Operator
### Describe the operator
Swap two dynamically determined axis.
Current implementation of the `torch.transpose` op in ONNX requires known rank to create the permutation list.…
-
We are trying to using all reduce TP to slash the communication time. I noticed that you have implemented [Row split + all_reduce for MLP (not faster, disabled)](https://github.com/turboderp/exllamav2…
-
### What happened?
Compiling the ASR frontend with VMVX causes the following error:
```mlir
/tmp/iree/libri/compute_frontend.mlir:89:10: error:
'util.buffer.store' op operand #0 must be inde…
-
Hi, I'm learning your ProteinMPNN framework. When going through your script, protein_mpnn_utils.py, I am confused with the variable `order_mask_backward`, which is defined and used in line 1085 and li…
-
Let's discuss and decide on some terminology to use consistently **for dense tensors**:
1) A tensor which is addressed by *n* indices is:
- [ ] a: *n*-dimensional
- [ ] b: *n*-mode
-…
-
[https://people.csail.mit.edu/devadas/pubs/pldi24_fhelipe.pdf](https://people.csail.mit.edu/devadas/pubs/pldi24_fhelipe.pdf)
Paper builds further on top of HECO, for optimal data layout
Paper impr…
-
**Describe the Bug**
Hey,
I have been trying to install apex in colab but it has the bug as mentioned above in the subject.
It looks like it has something with the amp_C extension.
_Here is the …