-
Notifications
You must be signed in to change notification settings - Fork 750
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
LLama convert docs #6754
LLama convert docs #6754
Conversation
感谢你贡献飞桨文档,文档预览构建中,Docs-New 跑完后即可预览,预览链接:http://preview-pr-6754.paddle-docs-preview.paddlepaddle.org.cn/documentation/docs/zh/api/index_cn.html |
00dee87
to
c73e2e9
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
你直接手动维护markdown表格就行
docs/guides/model_convert/convert_from_pytorch/pytorch_api_mapping_cn.md
Outdated
Show resolved
Hide resolved
docs/guides/model_convert/convert_from_pytorch/apply_reference_from_api_difference.py
Show resolved
Hide resolved
docs/guides/model_convert/convert_from_pytorch/pytorch_api_mapping_cn.md
Show resolved
Hide resolved
docs/guides/model_convert/convert_from_pytorch/api_difference/others/torch.inference_mode.md
Outdated
Show resolved
Hide resolved
docs/guides/model_convert/convert_from_pytorch/api_difference/others/torch.inference_mode.md
Outdated
Show resolved
Hide resolved
...es/model_convert/convert_from_pytorch/api_difference/others/torch.set_default_tensor_type.md
Outdated
Show resolved
Hide resolved
...ence_third_party/fairscale/fairscale.nn.model_parallel.initialize.get_model_parallel_rank.md
Outdated
Show resolved
Hide resolved
...hird_party/fairscale/fairscale.nn.model_parallel.initialize.get_model_parallel_world_size.md
Show resolved
Hide resolved
..._difference_third_party/fairscale/fairscale.nn.model_parallel.layers.ColumnParallelLinear.md
Outdated
Show resolved
Hide resolved
..._difference_third_party/fairscale/fairscale.nn.model_parallel.layers.ColumnParallelLinear.md
Outdated
Show resolved
Hide resolved
| | mp_group| 模型并行组| | ||
| | name| 网络层名称| | ||
|
||
### 转写示例 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
转写示例 需要写明:对哪些参数进行转写,参考模板 https://github.com/PaddlePaddle/docs/blob/develop/docs/guides/model_convert/convert_from_pytorch/api_difference/pytorch_api_mapping_format_cn.md
#### var1: 简述功能
#### var2: 简述功能
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
torch多的参数从用途来看均可直接删除。故无需 转写示例
...api_difference_third_party/fairscale/fairscale.nn.model_parallel.layers.ParallelEmbedding.md
Show resolved
Hide resolved
...api_difference_third_party/fairscale/fairscale.nn.model_parallel.layers.RowParallelLinear.md
Outdated
Show resolved
Hide resolved
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
docs/guides/model_convert/convert_from_pytorch/api_difference/others/torch.inference_mode.md
Outdated
Show resolved
Hide resolved
|
||
| PyTorch | PaddlePaddle | 备注 | | ||
| ----------- | ------------ | -------------------------------------------------------------------------------------- | | ||
| d | d | 全局默认数据类型,均支持所有浮点类型| |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
输入数据类型不一致,要写: torch的输入数据类型,paddle的输入数据类型,需要转写
这个是不是有了,直接在原有基础上修改吧,这个 改成:[ 输入参数类型不一致 ] 分类更合理
...ce_third_party/fairscale/fairscale.nn.model_parallel.initialize.initialize_model_parallel.md
Outdated
Show resolved
Hide resolved
..._difference_third_party/fairscale/fairscale.nn.model_parallel.layers.ColumnParallelLinear.md
Outdated
Show resolved
Hide resolved
...api_difference_third_party/fairscale/fairscale.nn.model_parallel.layers.ParallelEmbedding.md
Outdated
Show resolved
Hide resolved
...api_difference_third_party/fairscale/fairscale.nn.model_parallel.layers.RowParallelLinear.md
Outdated
Show resolved
Hide resolved
...api_difference_third_party/fairscale/fairscale.nn.model_parallel.layers.RowParallelLinear.md
Outdated
Show resolved
Hide resolved
@@ -1,4 +1,4 @@ | |||
## [ 输入参数用法不一致 ]torch.set_default_tensor_type | |||
## [ 输入参数类型不一致 ]torch.set_default_tensor_type |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
这个目录应该放到 torch.*,也就是ops下面吧。
others是针对没有分类的,为啥要从toch.* 移到others下面去
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
set_default_dtype 是放在other下面,这个api好像也不是op操作,我理解ops下应该放置操作Tensor的api,即输入是Tensor的API
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM
PR types
Docs
PR changes
Docs
Description
Pcard-73263
补充llama使用的fairscaleAPI与paddle的映射文档。