Skip to content

Actions: winglian/transformers

Actions

Secret Leaks

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
trufflehog.yml
66 workflow runs
66 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

fix-copies Secret Leaks #66: Commit e0a731d pushed by ArthurZucker
December 23, 2024 11:56 30s torch-compile-llama-attn
torch-compile-llama-attn
December 23, 2024 11:56 30s
chore: lint (make fix-copies) Secret Leaks #65: Commit b40f041 pushed by ArthurZucker
December 23, 2024 11:50 30s torch-compile-llama-attn
torch-compile-llama-attn
December 23, 2024 11:50 30s
avoid un-necessary unwrapping of the model on each train step Secret Leaks #64: Commit 7185dd5 pushed by winglian
December 18, 2024 18:10 31s no-unwrap-in-training
no-unwrap-in-training
December 18, 2024 18:10 31s
chore: lint Secret Leaks #63: Commit 647eccb pushed by winglian
December 13, 2024 18:22 30s deepspeed-no-sync-zero
deepspeed-no-sync-zero
December 13, 2024 18:22 30s
chore: lint (make fix-copies) Secret Leaks #62: Commit b999ea3 pushed by winglian
December 11, 2024 19:18 42s torch-compile-llama-attn
torch-compile-llama-attn
December 11, 2024 19:18 42s
fix no_sync context for deepspeed across all zero types Secret Leaks #61: Commit 41434e6 pushed by winglian
December 11, 2024 18:01 34s deepspeed-no-sync-zero
deepspeed-no-sync-zero
December 11, 2024 18:01 34s
make LlamaModel._update_causal_mask torch compilable Secret Leaks #60: Commit e60aa6b pushed by winglian
December 10, 2024 19:43 29s torch-compile-llama-attn
torch-compile-llama-attn
December 10, 2024 19:43 29s
flash attention forward doesn't play well with torch.compile Secret Leaks #59: Commit 92df994 pushed by winglian
December 9, 2024 23:35 29s flash-attn-torch-compile-no-graph-break
flash-attn-torch-compile-no-graph-break
December 9, 2024 23:35 29s
chore: lint Secret Leaks #58: Commit 37773bd pushed by winglian
December 9, 2024 04:06 29s deepspeed-no-sync-zero
deepspeed-no-sync-zero
December 9, 2024 04:06 29s
don't use no_sync when deepspeed doesn't support it for certain zero … Secret Leaks #57: Commit 63acec4 pushed by winglian
December 9, 2024 03:52 28s deepspeed-no-sync-zero
deepspeed-no-sync-zero
December 9, 2024 03:52 28s
don't use no_sync when deepspeed doesn't support it for certain zero … Secret Leaks #56: Commit 7ef3da1 pushed by winglian
December 9, 2024 03:46 33s deepspeed-no-sync-zero
deepspeed-no-sync-zero
December 9, 2024 03:46 33s
fix delay optimizer creation for non-fp8 cases Secret Leaks #55: Commit b7621f7 pushed by winglian
December 8, 2024 13:05 28s fsdp-optimizer-prepare
fsdp-optimizer-prepare
December 8, 2024 13:05 28s
use default in pop in case it doesn't exist Secret Leaks #54: Commit 6f69870 pushed by winglian
December 6, 2024 22:35 32s fix-num_items_in_batch-check
fix-num_items_in_batch-check
December 6, 2024 22:35 32s
don't pass ga loss kwargs to flash_attention Secret Leaks #53: Commit c605d68 pushed by winglian
December 6, 2024 21:12 28s fix-num_items_in_batch-check
fix-num_items_in_batch-check
December 6, 2024 21:12 28s
don't pass ga loss kwargs to flash_attention Secret Leaks #52: Commit 4dce79b pushed by winglian
December 6, 2024 21:12 1m 22s fix-num_items_in_batch-check
fix-num_items_in_batch-check
December 6, 2024 21:12 1m 22s
logic was inverted Secret Leaks #51: Commit 56811fd pushed by winglian
December 6, 2024 20:54 29s fix-num_items_in_batch-check
fix-num_items_in_batch-check
December 6, 2024 20:54 29s
fix check for loss_kwargs/GA Secret Leaks #50: Commit 935127b pushed by winglian
December 6, 2024 20:37 27s fix-num_items_in_batch-check
fix-num_items_in_batch-check
December 6, 2024 20:37 27s
Update I-JEPA checkpoints path (#35120) Secret Leaks #49: Commit c8c8dff pushed by winglian
December 6, 2024 20:34 35s main
main
December 6, 2024 20:34 35s
invert no_sync context to fix training for fsdp Secret Leaks #48: Commit edd102f pushed by winglian
November 7, 2024 17:28 29s fsdp-grad-accum-fix
fsdp-grad-accum-fix
November 7, 2024 17:28 29s
fix(dvclive): pass fake dataset to avoid exception in trainer init (#… Secret Leaks #47: Commit 6938524 pushed by winglian
November 7, 2024 17:20 28s main
main
November 7, 2024 17:20 28s
make sure to disable gradients for integer tensor Secret Leaks #46: Commit 130828f pushed by winglian
October 31, 2024 16:08 28s 8bit-deepspeed-zero3
8bit-deepspeed-zero3
October 31, 2024 16:08 28s
fix(DPT,Depth-Anything) Address expected_slice errors inside inferenc… Secret Leaks #45: Commit b5919e1 pushed by winglian
October 31, 2024 16:05 35s main
main
October 31, 2024 16:05 35s
Only cast logits to float when computing loss (#34147) Secret Leaks #44: Commit 816f442 pushed by winglian
October 19, 2024 16:10 36s main
main
October 19, 2024 16:10 36s
fix cross attn and vision attn Secret Leaks #43: Commit 674ad4d pushed by winglian
October 1, 2024 17:35 30s mllama-flash-attention-2
mllama-flash-attention-2
October 1, 2024 17:35 30s
fixes for fa2 Secret Leaks #42: Commit 6c9a312 pushed by winglian
September 30, 2024 20:25 26s mllama-flash-attention-2
mllama-flash-attention-2
September 30, 2024 20:25 26s