Awni Hannun comments

Results 1014 comments of


                                            Awni Hannun

Support quantization for non-multiples of 32.

Nice. I think (at least for the non-group axis) fixing this is mostly a matter of bounds checking in the matmul.. but @angeloskath can say more about if / when...

Support quantization for non-multiples of 32.

Aslo https://github.com/ml-explore/mlx-examples/issues/279

Support quantization for non-multiples of 32.

@mzbac not sure you saw, in 0.0.10 this should be fixed. We can add yayi2 to mlx-lm now 😃

Support quantization for non-multiples of 32.

Oh that would be amazing.

`repr` and `str` for bool arrays should print `True` and `False` instead of `true` and `false`

Possible we could allow for some basic number formatting in the printing of arrays. The python code can then set the appropriate format for `True` and `False`.

`repr` and `str` for bool arrays should print `True` and `False` instead of `true` and `false`

The array printing code is [here](https://github.com/ml-explore/mlx/blob/main/mlx/utils.cpp#L124). It looks like NumPy uses some global state (see [set_format_options](https://numpy.org/doc/stable/reference/generated/numpy.set_printoptions.html#numpy.set_printoptions) to manage the printing style. That may be a bit heavy handed for a...

Awni Hannun

Support quantization for non-multiples of 32.

Support quantization for non-multiples of 32.

Support quantization for non-multiples of 32.

Support quantization for non-multiples of 32.

`repr` and `str` for bool arrays should print `True` and `False` instead of `true` and `false`

`repr` and `str` for bool arrays should print `True` and `False` instead of `true` and `false`

Training seg faults after a few iterations

Training seg faults after a few iterations

Training seg faults after a few iterations

Training seg faults after a few iterations