Default Branch

f6dd02fa67 · Fix for migration (#2183) · Updated 2025-11-24 05:02:02 -05:00

Branches

a53e714ddf · fix combine mlir for llama2 · Updated 2023-11-20 03:30:55 -05:00    github

79
1

ee0233e370 · Fix formatting. · Updated 2023-11-13 21:01:28 -05:00    github

89
2

489a858af1 · enforce fp32 accumulates for cpu · Updated 2023-10-29 14:59:00 -04:00    github

140
1

7a7252a8fd · Unpin timm. · Updated 2023-10-19 13:41:22 -04:00    github

133
5

d7399c8ee7 · Update nightly.yml · Updated 2023-09-28 12:45:54 -04:00    github

169
2

16daba99fe · wip script for lowering dlrm training · Updated 2023-09-05 23:48:20 -04:00    github

274
1

bce652767c · branch to run vicuna with 4 shards · Updated 2023-07-21 18:28:24 -04:00    github

340
1

dd40a3fafe · Add simple OPT tuning script. · Updated 2023-06-21 21:02:42 -04:00    github

393
3

493f776253 · add pretrain_models.csv · Updated 2023-06-15 15:52:18 -04:00    github

415
2

6c5705cf09 · fixed a bug where lm would crash with certain smaller models · Updated 2023-03-23 03:36:04 -04:00

613
3

604c0f81b9 · Update model_wrappers.py · Updated 2023-02-24 09:25:48 -05:00    github

704
4

6d6a9dcae8 · Revert "Revert "Enable --device_allocator=caching"" · Updated 2023-02-10 02:00:32 -05:00    github

775
1

d9c62e547c · Update releases. · Updated 2023-01-31 10:15:57 -05:00    github

1245
7

ce00c1c5e1 · [SharkInference] Make SharkInference compile the entire module · Updated 2022-12-24 04:05:06 -05:00

969
1

b466b51247 · Update opt_params.py · Updated 2022-12-19 13:09:04 -05:00    github

1004
1

a84e78a9cc · Revert "Add target triple selection for multiple cards" · Updated 2022-12-16 18:00:23 -05:00    github

1025
1

44a684b241 · Revert "[SD][WEB] Deduce vulkan-target-triple in the presence of multiple cards" · Updated 2022-12-16 17:56:43 -05:00    github

1026
1

dd2a5795a2 · Add stress test · Updated 2022-12-12 20:47:49 -05:00    github

1068
4

6eea111f5e · bert training wip · Updated 2022-11-17 12:24:46 -05:00    github

1247
1

4b1a0b43ff · [WEB] Remove long prompts support · Updated 2022-11-03 09:27:58 -04:00    github

1189
5