| 文件 | 最后提交记录 | 最后更新时间 |
|---|---|---|
feat:Add adaptation support for GLM5.2 model Co-authored-by: LinShua<707894133@qq.com> # message auto-generated for no-merge-commit merge: !4610 merge master_glm5 into master feat:Add adaptation support for GLM5.2 model Created-by: LinShua Commit-by: LinShua Merged-by: ascend-robot Description: ## What this PR does / why we need it? Add adaptation support for GLM5.2 model: 1.Revise and extend the model configuration mapping, add the indexer_types field, and optimize parameter mapping for MoE shared experts. 2.Add dedicated weight conversion scripts for GLM5.2 to support bidirectional checkpoint conversion between HF and Mcore formats, with preconfigured parallelism, MoE and MLA parameters. 3.Implement the share indexer capability, which can be enabled via the arguments --index-topk-freq and --index-skip-topk-offset. 4.Integrate data preprocessing into training scripts and provide supporting environment variable scripts. No breaking changes to external APIs. Internal tests have been completed, with normal inference and precision aligned with vllm-ascend. 5.The MLA function requires enabling the parameters --apply-rope-no-in-complex and --no-use-sparse-c8-indexer. ## Does this PR introduce any user-facing change? NA. ## How was this patch tested? 见PR See merge request: Ascend/MindSpeed-LLM!4610 | 1 个月前 | |
feat:Add adaptation support for GLM5.2 model Co-authored-by: LinShua<707894133@qq.com> # message auto-generated for no-merge-commit merge: !4610 merge master_glm5 into master feat:Add adaptation support for GLM5.2 model Created-by: LinShua Commit-by: LinShua Merged-by: ascend-robot Description: ## What this PR does / why we need it? Add adaptation support for GLM5.2 model: 1.Revise and extend the model configuration mapping, add the indexer_types field, and optimize parameter mapping for MoE shared experts. 2.Add dedicated weight conversion scripts for GLM5.2 to support bidirectional checkpoint conversion between HF and Mcore formats, with preconfigured parallelism, MoE and MLA parameters. 3.Implement the share indexer capability, which can be enabled via the arguments --index-topk-freq and --index-skip-topk-offset. 4.Integrate data preprocessing into training scripts and provide supporting environment variable scripts. No breaking changes to external APIs. Internal tests have been completed, with normal inference and precision aligned with vllm-ascend. 5.The MLA function requires enabling the parameters --apply-rope-no-in-complex and --no-use-sparse-c8-indexer. ## Does this PR introduce any user-facing change? NA. ## How was this patch tested? 见PR See merge request: Ascend/MindSpeed-LLM!4610 | 1 个月前 | |
feat:Add adaptation support for GLM5.2 model Co-authored-by: LinShua<707894133@qq.com> # message auto-generated for no-merge-commit merge: !4610 merge master_glm5 into master feat:Add adaptation support for GLM5.2 model Created-by: LinShua Commit-by: LinShua Merged-by: ascend-robot Description: ## What this PR does / why we need it? Add adaptation support for GLM5.2 model: 1.Revise and extend the model configuration mapping, add the indexer_types field, and optimize parameter mapping for MoE shared experts. 2.Add dedicated weight conversion scripts for GLM5.2 to support bidirectional checkpoint conversion between HF and Mcore formats, with preconfigured parallelism, MoE and MLA parameters. 3.Implement the share indexer capability, which can be enabled via the arguments --index-topk-freq and --index-skip-topk-offset. 4.Integrate data preprocessing into training scripts and provide supporting environment variable scripts. No breaking changes to external APIs. Internal tests have been completed, with normal inference and precision aligned with vllm-ascend. 5.The MLA function requires enabling the parameters --apply-rope-no-in-complex and --no-use-sparse-c8-indexer. ## Does this PR introduce any user-facing change? NA. ## How was this patch tested? 见PR See merge request: Ascend/MindSpeed-LLM!4610 | 1 个月前 | |
fix: revert online weight conversion support for model scripts that do not support it. Co-authored-by: daixzh<daixuezheng@huawei.com> # message auto-generated for no-merge-commit merge: !4694 merge Train_from_HF into master fix: revert online weight conversion support for model scripts that do not support it. Created-by: daixzh Commit-by: daixzh Merged-by: ascend-robot Description: ## What this PR does / why we need it? Some training scripts currently use online weight conversion, but this feature is not yet supported for the corresponding models. Therefore, this PR reverts those scripts to use offline weight conversion instead. ## Does this PR introduce any user-facing change? The affected scripts switch from online weight conversion to offline weight conversion. ## How was this patch tested? By running the corresponding tests successfully. See merge request: Ascend/MindSpeed-LLM!4694 | 1 个月前 |
| 文件 | 最后提交记录 | 最后更新时间 |
|---|---|---|
| 1 个月前 | ||
| 1 个月前 | ||
| 1 个月前 | ||
| 1 个月前 |