tocmo0nlord/axolotl
1
0
Fork 0
Code Issues Pull Requests Actions 3 Packages Projects Releases Wiki Activity

drop old patches and code that are no longer needed (#3007) [skip ci]

Browse Source
This commit is contained in:
Wing Lian
2025-08-06 08:02:39 -04:00
committed by GitHub
parent 784f8c0e95
commit 97e86c6d47
11 changed files with 7 additions and 1330 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
.runpod/src/config/config.yaml
View File

@@ -296,7 +296,6 @@
# flash_attention:
# flash_attn_cross_entropy: # Whether to use flash-attention cross entropy implementation - advanced use only
# flash_attn_rms_norm: # Whether to use flash-attention rms norm implementation - advanced use only
# flash_attn_fuse_qkv: # Whether to fuse QKV into a single operation
# flash_attn_fuse_mlp: # Whether to fuse part of the MLP into a single operation
# # Whether to use scaled-dot-product attention
# # https://pytorch.org/docs/stable/generated/torch.nn.functional.scaled_dot_product_attention.html
@@ -541,7 +540,6 @@ xformers_attention: ${XFORMERS_ATTENTION}
flash_attention: ${FLASH_ATTENTION}
flash_attn_cross_entropy: ${FLASH_ATTN_CROSS_ENTROPY}
flash_attn_rms_norm: ${FLASH_ATTN_RMS_NORM}
flash_attn_fuse_qkv: ${FLASH_ATTN_FUSE_QKV}
flash_attn_fuse_mlp: ${FLASH_ATTN_FUSE_MLP}
sdp_attention: ${SDP_ATTENTION}
s2_attention: ${S2_ATTENTION}