0
代码
代码
Issues
Pull Requests
讨论
Wiki
分析
0
ascend-robotascend-robotfix: fix the kimi2 script error
9b6e36c5创建于 13 天前历史提交
文件最后提交记录最后更新时间
deepseek2_lite
test(pytorch): update deepseek2_lite scripts Co-authored-by: yanzhixiao<yanzhixiao@h-partners.com> # message auto-generated for no-merge-commit merge: !4326 merge fix-v2-lite into master test(pytorch): update deepseek2_lite scripts Created-by: yanzhixiao23 Commit-by: yanzhixiao Merged-by: ascend-robot Description: ## What this PR does / why we need it? Added CPU_AFFINITY_CONF and TASK_QUEUE_ENABLE to the DeepSeek2-Lite pretraining script to eliminate performance degradation caused by host instability. ## Does this PR introduce any user-facing change? NA ## How was this patch tested? Test case performance optimized. See merge request: Ascend/MindSpeed-LLM!43262 个月前
deepseek3
docs(pytorch): docs updated Co-authored-by: LQ1206<liuqian164@h-partners.com> # message auto-generated for no-merge-commit merge: !4418 merge 26.0.0 into 26.0.0 docs(pytorch): docs updated Created-by: LQ1206 Commit-by: LQ1206 Merged-by: ascend-robot Description: ## What this PR does / why we need it? Update the link to the latest version，and fixing the AIDD tool report issues. ## Does this PR introduce any user-facing change? Updated installation-related information to the latest version. ## How was this patch tested? NA See merge request: Ascend/MindSpeed-LLM!44181 个月前
deepseek32
[pytorch][sh]update deepseek32 pretrain script Co-authored-by: guozhihua2<guozhihua2@huawei.com> # message auto-generated for no-merge-commit merge: !4225 merge update_deepseek32_pretrain_master into master [pytorch][sh]update deepseek32 pretrain script Created-by: guozhihua2 Commit-by: guozhihua2 Merged-by: ascend-robot Description: 1. 增加deepseek32的通信构建和执行时间，防止出现通信执行超时 See merge request: Ascend/MindSpeed-LLM!42253 个月前
gemma2
modify ti config Co-authored-by: 乌兰娜仁<wulannarenzhao@gmail.com> # message auto-generated for no-merge-commit merge: !4135 merge add_ti_config2 into master modify ti config Created-by: hid88941705 Commit-by: 乌兰娜仁 Merged-by: ascend-robot Description: modify ti config See merge request: Ascend/MindSpeed-LLM!41353 个月前
glm45-moe
modify ti config Co-authored-by: 乌兰娜仁<wulannarenzhao@gmail.com> # message auto-generated for no-merge-commit merge: !4135 merge add_ti_config2 into master modify ti config Created-by: hid88941705 Commit-by: 乌兰娜仁 Merged-by: ascend-robot Description: modify ti config See merge request: Ascend/MindSpeed-LLM!41353 个月前
glm5
chore(pytorch):optimize GLM5 pretraining script Co-authored-by: HanhuiChen<chenhanhui1@h-partners.com> # message auto-generated for no-merge-commit merge: !4340 merge glm into master chore(pytorch):optimize GLM5 pretraining script Created-by: HANHU1CHEN Commit-by: HanhuiChen Merged-by: ascend-robot Description: ## What this PR does / why we need it? This PR updates the GLM5 pretraining script to improve training performance and efficiency. ## Does this PR introduce any user-facing change? No major user-facing changes, but minor adjustments to configs or launch commands may be required. ## How was this patch tested? Validated through end-to-end pretraining runs, confirming improved performance and stable training behavior. See merge request: Ascend/MindSpeed-LLM!43402 个月前
internlm3
modify ti config Co-authored-by: 乌兰娜仁<wulannarenzhao@gmail.com> # message auto-generated for no-merge-commit merge: !4135 merge add_ti_config2 into master modify ti config Created-by: hid88941705 Commit-by: 乌兰娜仁 Merged-by: ascend-robot Description: modify ti config See merge request: Ascend/MindSpeed-LLM!41353 个月前
kimi2
fix: fix the kimi2 script error Co-authored-by: downtiser<wangchenyang52@huawei.com> # message auto-generated for no-merge-commit merge: !4486 merge 26.0.0 into 26.0.0 fix: fix the kimi2 script error Created-by: downtiser Commit-by: downtiser Merged-by: ascend-robot Description: ## What this PR does / why we need it? fix the kimi2 script error ## Does this PR introduce any user-facing change? Please describe whether the PR will result in any user-facing usage changes. If there is related documentation, please specify its path. ## How was this patch tested? Please explain how to verify the correctness and effectiveness of this feature, as well as its usage constraints and limitations. See merge request: Ascend/MindSpeed-LLM!448613 天前
ling_v2
fix(python): fix ling-mini bug when enabling mtp Co-authored-by: yanzhixiao<yanzhixiao@h-partners.com> # message auto-generated for no-merge-commit merge: !4328 merge ling-mini-mtp into master fix(python): fix ling-mini bug when enabling mtp Created-by: yanzhixiao23 Commit-by: yanzhixiao Merged-by: ascend-robot Description: ## What this PR does / why we need it? Fix the precision bug when enabling MTP for the ling-mini model. ## Does this PR introduce any user-facing change? NA ## How was this patch tested? The bug has been fixed. See merge request: Ascend/MindSpeed-LLM!43282 个月前
llama2
refactor(pytorch): Model Sunset Plan II Co-authored-by: HanhuiChen<chenhanhui1@h-partners.com> # message auto-generated for no-merge-commit merge: !4337 merge deprecated into master refactor(pytorch): Model Sunset Plan II Created-by: HANHU1CHEN Commit-by: HanhuiChen Merged-by: ascend-robot Description: ## What this PR does / why we need it? **This PR is for removing deprecated model scripts, please refer associated issue for details.** ## Does this PR introduce any user-facing change? This will not affect users; they can continue to use the deprecated model for training in the previous supported version. ## How was this patch tested? Following internal discussions and the public announcement of the issue, we have decided to remove these models. See merge request: Ascend/MindSpeed-LLM!43372 个月前
llama31
refactor(pytorch): Model Sunset Plan II Co-authored-by: HanhuiChen<chenhanhui1@h-partners.com> # message auto-generated for no-merge-commit merge: !4337 merge deprecated into master refactor(pytorch): Model Sunset Plan II Created-by: HANHU1CHEN Commit-by: HanhuiChen Merged-by: ascend-robot Description: ## What this PR does / why we need it? **This PR is for removing deprecated model scripts, please refer associated issue for details.** ## Does this PR introduce any user-facing change? This will not affect users; they can continue to use the deprecated model for training in the previous supported version. ## How was this patch tested? Following internal discussions and the public announcement of the issue, we have decided to remove these models. See merge request: Ascend/MindSpeed-LLM!43372 个月前
longcat
[pytorch][model]longcat model fix Co-authored-by: guihaowen666<guihaowen@huawei.com> # message auto-generated for no-merge-commit merge: !4251 merge br_master_longcat_fix into master [pytorch][model]longcat model fix Created-by: guihaowen666 Commit-by: guihaowen666 Merged-by: ascend-robot Description: longcat model fix See merge request: Ascend/MindSpeed-LLM!42513 个月前
magistral
modify ti config Co-authored-by: 乌兰娜仁<wulannarenzhao@gmail.com> # message auto-generated for no-merge-commit merge: !4135 merge add_ti_config2 into master modify ti config Created-by: hid88941705 Commit-by: 乌兰娜仁 Merged-by: ascend-robot Description: modify ti config See merge request: Ascend/MindSpeed-LLM!41353 个月前
mamba2
modify ti config Co-authored-by: 乌兰娜仁<wulannarenzhao@gmail.com> # message auto-generated for no-merge-commit merge: !4135 merge add_ti_config2 into master modify ti config Created-by: hid88941705 Commit-by: 乌兰娜仁 Merged-by: ascend-robot Description: modify ti config See merge request: Ascend/MindSpeed-LLM!41353 个月前
phi35
modify ti config Co-authored-by: 乌兰娜仁<wulannarenzhao@gmail.com> # message auto-generated for no-merge-commit merge: !4135 merge add_ti_config2 into master modify ti config Created-by: hid88941705 Commit-by: 乌兰娜仁 Merged-by: ascend-robot Description: modify ti config See merge request: Ascend/MindSpeed-LLM!41353 个月前
plm
modify ti config Co-authored-by: 乌兰娜仁<wulannarenzhao@gmail.com> # message auto-generated for no-merge-commit merge: !4135 merge add_ti_config2 into master modify ti config Created-by: hid88941705 Commit-by: 乌兰娜仁 Merged-by: ascend-robot Description: modify ti config See merge request: Ascend/MindSpeed-LLM!41353 个月前
qwen25
refactor(pytorch): Model Sunset Plan II Co-authored-by: HanhuiChen<chenhanhui1@h-partners.com> # message auto-generated for no-merge-commit merge: !4337 merge deprecated into master refactor(pytorch): Model Sunset Plan II Created-by: HANHU1CHEN Commit-by: HanhuiChen Merged-by: ascend-robot Description: ## What this PR does / why we need it? **This PR is for removing deprecated model scripts, please refer associated issue for details.** ## Does this PR introduce any user-facing change? This will not affect users; they can continue to use the deprecated model for training in the previous supported version. ## How was this patch tested? Following internal discussions and the public announcement of the issue, we have decided to remove these models. See merge request: Ascend/MindSpeed-LLM!43372 个月前
qwen3
refactor(pytorch): Model Sunset Plan II Co-authored-by: HanhuiChen<chenhanhui1@h-partners.com> # message auto-generated for no-merge-commit merge: !4337 merge deprecated into master refactor(pytorch): Model Sunset Plan II Created-by: HANHU1CHEN Commit-by: HanhuiChen Merged-by: ascend-robot Description: ## What this PR does / why we need it? **This PR is for removing deprecated model scripts, please refer associated issue for details.** ## Does this PR introduce any user-facing change? This will not affect users; they can continue to use the deprecated model for training in the previous supported version. ## How was this patch tested? Following internal discussions and the public announcement of the issue, we have decided to remove these models. See merge request: Ascend/MindSpeed-LLM!43372 个月前
qwen3_coder_next
[pytorch][bugfix]qwen3-coder-next support Co-authored-by: mhh111<mahonghao1@huawei.com> # message auto-generated for no-merge-commit merge: !4176 merge master333 into master [pytorch][bugfix]qwen3-coder-next support Created-by: mhh111 Commit-by: mhh111 Merged-by: ascend-robot Description: [pytorch][bugfix]qwen3-coder-next support See merge request: Ascend/MindSpeed-LLM!41763 个月前
qwen3_moe
fix(pytorch): fix weight of optimizer save problem Co-authored-by: HanhuiChen<chenhanhui1@h-partners.com> # message auto-generated for no-merge-commit merge: !4352 merge master into master fix(pytorch): fix weight of optimizer save problem Created-by: HANHU1CHEN Commit-by: HanhuiChen Merged-by: ascend-robot Description: ## What this PR does / why we need it? Disable gloo communication to avoid the issue of weight save in optimizers. ## Does this PR introduce any user-facing change? No. ## How was this patch tested? This patch have been on the cluster. See merge request: Ascend/MindSpeed-LLM!43522 个月前
qwen3_next
[pytorch][bugfix]qwen3-coder-next support Co-authored-by: mhh111<mahonghao1@huawei.com> # message auto-generated for no-merge-commit merge: !4176 merge master333 into master [pytorch][bugfix]qwen3-coder-next support Created-by: mhh111 Commit-by: mhh111 Merged-by: ascend-robot Description: [pytorch][bugfix]qwen3-coder-next support See merge request: Ascend/MindSpeed-LLM!41763 个月前
seed_oss
modify ti config Co-authored-by: 乌兰娜仁<wulannarenzhao@gmail.com> # message auto-generated for no-merge-commit merge: !4135 merge add_ti_config2 into master modify ti config Created-by: hid88941705 Commit-by: 乌兰娜仁 Merged-by: ascend-robot Description: modify ti config See merge request: Ascend/MindSpeed-LLM!41353 个月前