Logan Adams

Results 294 comments of Logan Adams

MoQ problem ：'str' object has no attribute 'size'

AMD support updated as well, please try with recent DeepSpeed and re-open if the issue persists.

remove force-multi and fix None val check in base tuner in autotuning

These changes were already applied in other PRs, closing this one.

Fix the file-paths for the auto-generated hip files through JIT

Closing as this targets a non-master branch and appears to be stale.

Error building extension 'cpu_adam'

Closing this issue as the original issue was resolved. If anyone is having issues with this, please open a new issue and link this one and we would be happy...

Add RTS and token masking to top-2 gating + configurable jitter epsilon

Hi @ykim362 - is this still an issue or a PR you'd like to see completed? If so we can fix the conflicts and review, otherwise we would like to...

Add RTS and token masking to top-2 gating + configurable jitter epsilon

Hi @ykim362 - I'm going to close this PR for now. If this is still something you'd like to merge, we'd be happy to review promptly next time, just re-open...

Add shared libs

Closing this as it is old and corresponding changes appear to have already been made.

Allow FixedSparseAttention with num_global_blocks = 0

Closing this PR as it appears to be stale and out of date - if this is still useful, please re-open or create a new PR. Thanks!

fix: remove layer_past storing in DeepSpeedTransformerInference

Hi @codertimo - Do you believe this PR is still useful? I'm not able to resolve the conflicts on your fork with our master branch. If so, could you resolve...

Add function `to_sequential` to PipelineModule

@sdtblck - just fixed some formatting issues that were preventing this - if the tests pass, would this be good to merge now?