From bcc2c1fd8fddc92b8d2e5feb912467d30486018d Mon Sep 17 00:00:00 2001 From: Yaowei Zheng Date: Thu, 2 Oct 2025 02:06:09 +0800 Subject: [PATCH] [misc] move wechat out (#9223) --- README.md | 2 +- README_zh.md | 2 +- examples/extras/fp8/llama3_fp8_deepspeed_sft.yaml | 6 +++--- examples/extras/fp8/llama3_fp8_fsdp_sft.yaml | 4 ++-- 4 files changed, 7 insertions(+), 7 deletions(-) diff --git a/README.md b/README.md index 92f24fe7..0e942232 100644 --- a/README.md +++ b/README.md @@ -36,7 +36,7 @@ -👋 Join our [WeChat](assets/wechat/main.jpg), [NPU](assets/wechat/npu.jpg), [Lab4AI](assets/wechat/lab4ai.jpg), [LLaMA Factory Online](assets/wechat/online.jpg) user group. +👋 Join our [WeChat](https://github.com/hiyouga/llamafactory-community/blob/main/wechat/main.jpg), [NPU](https://github.com/hiyouga/llamafactory-community/blob/main/wechat/npu.jpg), [Lab4AI](https://github.com/hiyouga/llamafactory-community/blob/main/wechat/lab4ai.jpg), [LLaMA Factory Online](https://github.com/hiyouga/llamafactory-community/blob/main/wechat/online.jpg) user group. \[ English | [中文](README_zh.md) \] diff --git a/README_zh.md b/README_zh.md index 204cd2c2..e9dbf6dd 100644 --- a/README_zh.md +++ b/README_zh.md @@ -36,7 +36,7 @@ -👋 加入我们的[微信群](assets/wechat/main.jpg)、[NPU 用户群](assets/wechat/npu.jpg)、[大模型实验室群](assets/wechat/lab4ai.jpg) 或 [LLaMA Factory Online 用户群](assets/wechat/online.png)。 +👋 加入我们的[微信群](https://github.com/hiyouga/llamafactory-community/blob/main/wechat/main.jpg)、[NPU 用户群](https://github.com/hiyouga/llamafactory-community/blob/main/wechat/npu.jpg)、[大模型实验室群](https://github.com/hiyouga/llamafactory-community/blob/main/wechat/lab4ai.jpg) 或 [LLaMA Factory Online 用户群](https://github.com/hiyouga/llamafactory-community/blob/main/wechat/online.png)。 \[ [English](README.md) | 中文 \] diff --git a/examples/extras/fp8/llama3_fp8_deepspeed_sft.yaml b/examples/extras/fp8/llama3_fp8_deepspeed_sft.yaml index 1f92c52b..555e884d 100644 --- a/examples/extras/fp8/llama3_fp8_deepspeed_sft.yaml +++ b/examples/extras/fp8/llama3_fp8_deepspeed_sft.yaml @@ -6,7 +6,7 @@ model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct trust_remote_code: true -### Method configuration +### Method configuration stage: sft do_train: true finetuning_type: full @@ -40,9 +40,9 @@ fp8: true fp8_backend: torchao # Use TorchAO backend for FP8 fp8_enable_fsdp_float8_all_gather: false # Not used with DeepSpeed -### DeepSpeed configuration +### DeepSpeed configuration deepspeed: examples/deepspeed/ds_z3_fp8_config.json ### Logging configuration report_to: wandb -run_name: llama3_fp8_deepspeed_sft \ No newline at end of file +run_name: llama3_fp8_deepspeed_sft diff --git a/examples/extras/fp8/llama3_fp8_fsdp_sft.yaml b/examples/extras/fp8/llama3_fp8_fsdp_sft.yaml index 7590140d..5983c3e8 100644 --- a/examples/extras/fp8/llama3_fp8_fsdp_sft.yaml +++ b/examples/extras/fp8/llama3_fp8_fsdp_sft.yaml @@ -1,4 +1,4 @@ -# FP8 training example with FSDP +# FP8 training example with FSDP # This config demonstrates FP8 mixed precision training using HuggingFace Accelerate # with FSDP for distributed training and float8 all-gather optimization @@ -48,4 +48,4 @@ fsdp_transformer_layer_cls_to_wrap: LlamaDecoderLayer ### Logging configuration report_to: wandb -run_name: llama3_fp8_fsdp_sft \ No newline at end of file +run_name: llama3_fp8_fsdp_sft