Lucas Pasqualin

[AmountOfMoney] Utterances that match the pattern `<number> $ <number>` return the wrong amount

3

The following requests replicate the issue the latest pull of `master`: * `i have two $ 500 checks` ``` [ { "body":"two $ 500", "start":7, "value": {"value":7,"type":"value","unit":"$"}, "end":16, "dim":"amount-of-money", "latent":false...

Utterance "winter 2020" not parsed correctly.

3

Utterance `winter 2020` is broken down into two entities - `winter` and `2020`. For all other seasons, the correct date range is returned (e.g. `summer 2020`, `fall 2021`, etc.). Having...

Implement's torch SDPA for FlashAttention Kernel

9

Implements torch sdpa for mem_efficient kernel support! Using the mem_efficient kernel results in a ~15.5% faster training time per batch, going from a ~154ms/batch baseline to ~130ms/batch. (Ran on 8...

[DCP] Adds storage reader and planner classes for online loading/sharding of models in torch.save format

6

Stack from [ghstack](https://github.com/ezyang/ghstack) (oldest at bottom): * #120128 * #120127 * __->__ #119816 * #119815 * #119814 * #119813 as title Differential Revision: [D53718041](https://our.internmc.facebook.com/intern/diff/D53718041/) cc @mrshenli @pritamdamania87 @zhaojuanmao @satgera @rohan-varma...

[DCP] Adds utility for converting dcp to torch save format

7

Stack from [ghstack](https://github.com/ezyang/ghstack) (oldest at bottom): * #120128 * #120127 * #119816 * #119815 * __->__ #119814 * #119813 as title Differential Revision: [D53718042](https://our.internmc.facebook.com/intern/diff/D53718042/) cc @mrshenli @pritamdamania87 @zhaojuanmao @satgera @rohan-varma...

[DCP] Adds utility for converting torch save to dcp

6

Stack from [ghstack](https://github.com/ezyang/ghstack) (oldest at bottom): * #120128 * #120127 * #119816 * __->__ #119815 * #119814 * #119813 as title Differential Revision: [D53718040](https://our.internmc.facebook.com/intern/diff/D53718040/) cc @mrshenli @pritamdamania87 @zhaojuanmao @satgera @rohan-varma...

[DCP] deletes legacy formatting test

3

Stack from [ghstack](https://github.com/ezyang/ghstack) (oldest at bottom): * #120128 * __->__ #120127 * #119816 * #119815 * #119814 * #119813 Should no longer be necessary Differential Revision: [D53791345](https://our.internmc.facebook.com/intern/diff/D53791345/)

[DCP] Adds main in format utils

3

Stack from [ghstack](https://github.com/ezyang/ghstack) (oldest at bottom): * __->__ #120128 * #120127 * #119816 * #119815 * #119814 * #119813 Adds main in format utils. Usage: `python -m torch.distributed.checkpoint.format_utils dcp_to_torch dcp_dir...

[DCP][TNT] Updates test to use multi-threading

1

Please read through our [contribution guide](https://github.com/pytorch/tnt/blob/main/CONTRIBUTING.md) prior to creating your pull request. Summary: Updates dcp test to use multithreading. This PR is showing an error in CI. This test passes...

Please read through our [contribution guide](https://github.com/pytorch/tnt/blob/main/CONTRIBUTING.md) prior to creating your pull request. Summary: Test plan: Fixes #{issue number}