SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

fix: "check out" as verb (#38678)

Browse Source 
"check out" as verb
This commit is contained in:
Anthony
2025-06-09 16:07:31 +02:00
committed by GitHub
parent 237ff80387
commit 19224c3642
75 changed files with 91 additions and 91 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
examples/modular-transformers/configuration_my_new_model.py
View File

@@ -36,7 +36,7 @@ class MyNewModelConfig(PretrainedConfig):
`num_key_value_heads=num_attention_heads`, the model will use Multi Head Attention (MHA), if
`num_key_value_heads=1` the model will use Multi Query Attention (MQA) otherwise GQA is used. When
converting a multi-head checkpoint to a GQA checkpoint, each group key and value head should be constructed
by meanpooling all the original heads within that group. For more details checkout [this
by meanpooling all the original heads within that group. For more details, check out [this
paper](https://arxiv.org/pdf/2305.13245.pdf). If it is not specified, will default to
`num_attention_heads`.
hidden_act (`str` or `function`, *optional*, defaults to `"silu"`):

2
examples/modular-transformers/configuration_new_model.py
View File

@@ -34,7 +34,7 @@ class NewModelConfig(PretrainedConfig):
`num_key_value_heads=num_attention_heads`, the model will use Multi Head Attention (MHA), if
`num_key_value_heads=1` the model will use Multi Query Attention (MQA) otherwise GQA is used. When
converting a multi-head checkpoint to a GQA checkpoint, each group key and value head should be constructed
by meanpooling all the original heads within that group. For more details checkout [this
by meanpooling all the original heads within that group. For more details, check out [this
paper](https://arxiv.org/pdf/2305.13245.pdf). If it is not specified, will default to
`num_attention_heads`.
head_dim (`int`, *optional*, defaults to 256):