423A35C7/LLaMA-Factory
1
0
Fork 0
mirror of https://github.com/hiyouga/LLaMA-Factory.git synced 2025-12-18 12:50:38 +08:00
Code Issues Packages Projects Releases Wiki Activity

[data] optimize qwen3 loss computation (#7923)

Browse Source
This commit is contained in:
hoshi-hiyouga
2025-04-30 16:18:00 +08:00
committed by GitHub
parent 73198a6645
commit 052ca871bd
11 changed files with 205 additions and 39 deletions
Download Patch File Download Diff File Expand all files Collapse all files

22
src/llamafactory/extras/constants.py
View File

@@ -533,6 +533,17 @@ register_model_group(
DownloadSource.DEFAULT: "deepseek-ai/DeepSeek-V3",
DownloadSource.MODELSCOPE: "deepseek-ai/DeepSeek-V3",
},
"DeepSeek-V3-671B-0324-Chat": {
DownloadSource.DEFAULT: "deepseek-ai/DeepSeek-V3-0324",
DownloadSource.MODELSCOPE: "deepseek-ai/DeepSeek-V3-0324",
},
},
template="deepseek3",
)
register_model_group(
models={
"DeepSeek-R1-1.5B-Distill": {
DownloadSource.DEFAULT: "deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B",
DownloadSource.MODELSCOPE: "deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B",
@@ -566,7 +577,7 @@ register_model_group(
DownloadSource.MODELSCOPE: "deepseek-ai/DeepSeek-R1",
},
},
template="deepseek3",
template="deepseekr1",
)
@@ -737,6 +748,13 @@ register_model_group(
DownloadSource.DEFAULT: "THUDM/GLM-4-32B-0414",
DownloadSource.MODELSCOPE: "ZhipuAI/GLM-4-32B-0414",
},
},
template="glm4",
)
register_model_group(
models={
"GLM-Z1-9B-0414-Chat": {
DownloadSource.DEFAULT: "THUDM/GLM-Z1-9B-0414",
DownloadSource.MODELSCOPE: "ZhipuAI/GLM-Z1-9B-0414",
@@ -746,7 +764,7 @@ register_model_group(
DownloadSource.MODELSCOPE: "ZhipuAI/GLM-Z1-32B-0414",
},
},
template="glm4",
template="glmz1",
)