-
Refactor the Upload Task in GraphCore to make use of the newly created ClientFactory.
-
```
WARNING - Documentation file 'ai-testbed/cerebras/job-queuing-and-submission.md' contains a link to
'ai-testbed/cerebras/Steps-to-run-a-model-or-program.md' which is not found in t…
-
Hi, I follow the readme.md to run the whisper example program,
pip install optimum[graphcore] optuna
and my Python package is like the below:
transformers==4.25.1
optimum==1.6.1
optimum-…
-
We're generally constrained by the LLVM version that comes with Julia, but it'd be great to figure out what's preventing us from upgrading to newer versions of Poplar, for example 1.4 or 2.0 which sho…
-
Computing two hydrogen atoms using cpu backend works:
`python density_functional_theory.py -H -backend cpu -float32`
but with ipu backend
`python density_functional_theory.py -H -backend ipu -flo…
-
### Description & Motivation
The `Trainer`'s base precision class inherits from Fabric's base Precision: https://github.com/Lightning-AI/lightning/blob/master/src/lightning/pytorch/plugins/precision/…
-
Seems like app just crashes on Core mod loading because of lack of a localization for system region. Setting it to `English (USA)` in Win settings fixes the issue but idk it's not really a nice way.
…
-
bug in regression part of the hybridCElos, there should be a softmax applied first before inner product with `self.brackets` as reported by @callumm-graphcore
-
### Description
I did
" pip install jax==0.3.16+ipu jaxlib==0.3.15+ipu.sdk310 -f https://graphcore-research.github.io/jax-experimental/wheels.html "
as mentioned in the documentation.
When I o…
-
### 📖 Documentation
I've written two notebooks based on the tutorial that show how to use Molfeat + Graphcore IPUs to fine-tune a pre-trained Transformer model and to train a GINE model.
Idea…