← Back to report

jingyaogong/minimind

All events — 2026-03-27

Type AI Score Description Actor Reason Date
COMMIT 0.20 [fix] harmonize template Uses technical term 'harmonize' but in b 2025-11-06
COMMIT 0.15 fix: attn_forwad when is_causal=True assert attn_mask is Non Slightly more descriptive but remains te 2025-11-18
COMMIT 0.10 [update] params log Bare-bones commit messages typical of hu 2026-01-07
COMMIT 0.10 [update] mask log Bare-bones commit messages typical of hu 2026-01-07
COMMIT 0.10 [update] readme Bare-bones commit messages typical of hu 2026-01-06
COMMIT 0.10 [update] simplify loader Bare-bones commit messages typical of hu 2026-01-05
COMMIT 0.10 [update] readme Bare-bones commit messages typical of hu 2026-01-05
COMMIT 0.10 [update] rename train tokenizer Bare-bones commit messages typical of hu 2026-01-05
COMMIT 0.10 [update] readme Bare-bones commit messages typical of hu 2026-01-05
COMMIT 0.10 [fix] messages num Bare-bones commit messages typical of hu 2026-01-04
COMMIT 0.10 [fix] dist cleanup Bare-bones commit messages typical of hu 2026-01-02
COMMIT 0.10 [feat] update yarn Extremely terse commit-style messages wi 2025-12-01
COMMIT 0.10 [feat] release memory Very brief technical phrasing typical of 2025-11-27
COMMIT 0.10 [fix] ppo mask Minimal message uses domain abbreviation 2025-11-19
COMMIT 0.10 [fix] model attn_mask Concise technical fix reference with dom 2025-11-19
COMMIT 0.10 [fix] update model Simple two-word technical instruction wi 2025-11-18
COMMIT 0.10 [fix] prompt length calculate Brief domain-specific fix reference with 2025-11-15
COMMIT 0.10 [fix] model-name Minimal hyphenated fix reference typical 2025-11-07
COMMIT 0.10 [feat] update requirements Extremely terse technical commit message 2025-10-23
COMMIT 0.10 [feat] update readme Minimal commit message with specific tec 2025-10-23
COMMIT 0.10 [feat] update readme Brief, repetitive commit message typical 2025-10-23
COMMIT 0.10 [feat] repetition-penalty Specialized ML term with no polite or fo 2025-10-23
COMMIT 0.10 [feat] convert2llama Technical shorthand conversion label wit 2025-10-23
COMMIT 0.10 [feat] shuffle data Concise data operation command, not AI-s 2025-10-23
COMMIT 0.10 [fix] loss-issues-430 Issue-specific fix reference with techni 2025-10-23
COMMIT 0.10 [fix] restore Single-word commit message - too minimal 2025-10-23
COMMIT 0.10 [fix] issue-431 Issue number reference in terse human-st 2025-10-23
COMMIT 0.10 [fix] sampler-ddp Technical DDP sampler fix with domain-sp 2025-10-23
COMMIT 0.00 [update] random seed Very terse, standard commit message styl 2026-03-27
COMMIT 0.00 [update] fp16 inference Brief, domain-specific commit message. 2026-03-27
COMMIT 0.00 [update] image Terse, mechanical commit; typical human 2026-03-26
COMMIT 0.00 [update] change default seq_len Concise and technical; common human comm 2026-03-26
COMMIT 0.00 [update] minimind-3 Very brief, informal; no AI hallmarks pr 2026-03-24
COMMIT 0.00 [fix] align log/save last-step check and ETA with 1-indexed Specific technical fix; human-like jargo 2026-03-24
COMMIT 0.00 [fix] gradient accumulation step alignment Direct technical fix; human engineering 2026-03-24
COMMIT 0.00 [update] empty_think_ratio Minimal, to-the-point; lacks AI politene 2026-02-06
COMMIT 0.00 [update] empty_think_ratio Repetitive, brief update; likely human c 2026-02-05
COMMIT 0.00 [feat] data process Ambiguous but informal; typical human sh 2026-02-05
COMMIT 0.00 [update] save interval Concise and clear; common human commit m 2026-01-30
COMMIT 0.00 [update] safe half Very brief, technical term; no AI stylis 2026-01-30
COMMIT 0.00 [fix] data skip Extremely terse and informal phrasing ty 2026-01-18
COMMIT 0.00 [update] shuffle data Brief, informal update note with common 2026-01-18
COMMIT 0.00 [fix] max length Minimalist technical fix; lacks any poli 2026-01-17
COMMIT 0.00 [update] pretrain load Concise, technical update note with stan 2026-01-17
COMMIT 0.00 [update] align mask Very short and specific; common ML/engin 2026-01-15
COMMIT 0.00 [update] align loss Terse technical update; no hallmark AI p 2026-01-14
COMMIT 0.00 [fix] compile unpack Brief human-style fix message ('fix comp 2026-01-14
COMMIT 0.00 [feat] add compile Succinct feature addition; common Git co 2026-01-14
COMMIT 0.00 [update] prompt prefill Direct technical update; uses specific j 2026-01-13
COMMIT 0.00 [update] show speed Short, informal update typical of human 2026-01-07
COMMIT 0.00 [update] rename reason Bare-bones commit messages typical of hu 2026-01-05
COMMIT 0.00 [update] aux loss Terse commit-style messages with technic 2026-01-01
COMMIT 0.00 [fix] experts unused Concise fix notation typical of develope 2025-12-31
COMMIT 0.00 [fix] layers set 8 Very brief technical notation lacking AI 2025-12-31
COMMIT 0.00 [fix] moe unused Short technical fix message with domain- 2025-12-31
COMMIT 0.00 [feat] get params Simple feature description with no AI-st 2025-12-31
COMMIT 0.00 [feat] get params Identical to previous item; typical huma 2025-12-31
COMMIT 0.00 [feat] update config Brief config update notation without AI 2025-12-31
COMMIT 0.00 [feat] update lr Abbreviated technical term (lr) with no 2025-12-31
COMMIT 0.00 [feat] compatible tokenizer Concise feature description using domain 2025-12-31
COMMIT 0.00 [feat] stream load data Terse technical notation typical of deve 2025-12-28
COMMIT 0.00 [feat] remove empty_cache Terse, technical message with typical Gi 2025-12-26
COMMIT 0.00 [feat] explicit left padding Concise, technical message with standard 2025-12-23
COMMIT 0.00 [fix] lora weight Minimal, direct technical fix descriptio 2025-12-22
COMMIT 0.00 Fix: support loading DDP-saved LoRA weights for inference Direct technical description with specif 2025-12-22
COMMIT 0.00 [feat] adjust seq length Very brief, technical update typical of 2025-12-14
COMMIT 0.00 [feat] update readme Very brief update message, common for Gi 2025-12-11
COMMIT 0.00 [fix] dtype & lr Extremely concise technical fix with com 2025-12-09
COMMIT 0.00 [fix] Refactor get_lr function to include min_lr calculation Detailed, specific technical explanation 2025-12-06
COMMIT 0.00 [fix] reduce aux_loss_alpha Minimal technical message about a parame 2025-12-05
COMMIT 0.00 [fix] cuda memory #559 Terse, context-specific technical messag 2025-12-01
COMMIT 0.00 [feat] add MNN support to README. Specific feature addition but remains co 2025-11-10
COMMIT 0.00 [feat] clear cache Terse, informal commit style typical of 2025-11-06
COMMIT 0.00 [fix] harmonize template Concise technical wording, lacks AI form 2025-11-02
COMMIT 0.00 [feat] update import Brief and direct, common in human commit 2025-10-31
COMMIT 0.00 [feat] update readme Minimal update note, no AI stylistic mar 2025-10-30
COMMIT 0.00 [feat] update readme Repetitive simple update, typical human 2025-10-30
COMMIT 0.00 [feat] update readme Same as previous, no signs of AI generat 2025-10-30
COMMIT 0.00 [feat] update datasets Direct technical term, no AI phrasing. 2025-10-30
COMMIT 0.00 [feat] update args Short, informal, lacks any AI hallmarks. 2025-10-30
COMMIT 0.00 [feat] add args Minimalist and terse, characteristic of 2025-10-30
COMMIT 0.00 [feat] update readme Identical to other updates, clearly huma 2025-10-29
COMMIT 0.00 [fix] model device Very terse, technical commit format with 2025-10-29
COMMIT 0.00 [feat] update trainer Minimal, repetitive commit title typical 2025-10-28
COMMIT 0.00 [feat] update trainer Identical to previous; likely batch huma 2025-10-28
COMMIT 0.00 [feat] update readme Brief, standard commit to update documen 2025-10-27
COMMIT 0.00 [feat] update readme Repetitive format, no AI stylistic phras 2025-10-26
COMMIT 0.00 [feat] update eval-llm Abbreviated, informal module name like ' 2025-10-26
COMMIT 0.00 [feat] pause-training Hyphenated, descriptive feature name com 2025-10-26
COMMIT 0.00 [feat] update readme Repetitive, minimal content typical of m 2025-10-23
COMMIT 0.00 [feat] update readme Identical pattern; no AI hallmarks like 2025-10-23
COMMIT 0.00 [feat] update requirements Terse, technical commit typical of depen 2025-10-23
COMMIT 0.00 [fix] graph-oom & ddp-pos_cis Extremely terse and informal, typical of 2025-10-23
COMMIT 0.00 [fix] git track Short, minimal text with informal phrasi 2025-10-21
COMMIT 0.00 [feat] update readme Brief and typical human update to docume 2025-10-21
COMMIT 0.00 [feat] minimind-2510 Concise, likely a project-specific refer 2025-10-21
COMMIT 0.00 [feat] update eval Very terse; lacks any AI-style politenes 2025-10-17
COMMIT 0.00 [feat] update requirements Minimal description, common for dependen 2025-10-16
COMMIT 0.00 [fix] update model Too brief and direct to be AI-generated. 2025-10-16
COMMIT 0.00 [fix] update model Identical to previous; no AI hallmarks. 2025-10-16
PR 0.00 使用einops进一步提升代码可读性 Chinese text, concise and topic-specific 2026-03-27
PR 0.00 [feat] add dapo algorithm Detailed technical phrasing, human tone, 2026-03-27
PR 0.00 merge redundant forward passes for logps and aux_loss (in tr 2026-03-24
PR 0.00 添加数据集加载逻辑、网页内容抓取与数据预处理逻辑 2025-12-16
PR 0.00 [docs] Fix wording in RLHF section of README.md file 2026-01-27
PR 0.00 [docs]: clarify pretraining data format in README 2025-12-31
PR 0.00 refactor: optimize tensor wrapping in lm_dataset.py 2025-12-20
PR 0.00 [fix] 修复训练脚本中 1-indexed step 与 0-indexed 逻辑混用的问题 2026-03-24
PR 0.00 Fix SFT resume with torch.compile enabledfix sft resume with 2026-03-19
PR 0.00 新建test分支 2026-03-22
PR 0.00 [mod & add] fix spo algorithm, add dapo and cispo algorithm 2026-01-30
PR 0.00 Mega 2025-06-26
PR 0.00 Add dynamic growth pipeline, eval tooling, and overnight run 2026-02-22
PR 0.00 更新了 model / trainer 中的注释 & PyTorch 新版本 Automatic Mixed Preci 2026-02-04
PR 0.00 Update requirements.txt 2024-10-28
PR 0.00 Auto tokenizer name path fix 2024-10-03
PR 0.00 Update requirements 2024-10-01
PR 0.00 [add] add gating term on po algorithm 2026-02-03
PR 0.00 add muon optimizer 2026-02-02
PR 0.00 modified: .gitignore 2026-01-19
PR 0.00 Fix DPO loss_mask boundary (include first assistant token) 2026-01-07
PR 0.00 [feat] Support Minimind retrieval-augmented generation (RAG) 2025-11-16
PR 0.00 perf: merge LoRA weights into model for inference 2025-12-23
PR 0.00 Create RESOURCES.md for MiniMind project 2025-10-24
PR 0.00 Perf/merge lora weights 2025-12-23
PR 0.00 Fix: support loading DDP-saved LoRA weights for inference 2025-12-22
PR 0.00 [feat] add interactive notebook 2025-02-23
PR 0.00 [feat] Add Training Web UI 2025-11-06
PR 0.00 fix: 调整model_lora.py里面lora作用的对象 2025-12-05
PR 0.00 Add attention gate 2025-12-12
PR 0.00 feat: 增加 LoRA alpha 缩放系数及命令行支持 2025-12-12
PR 0.00 [fix] Refactor get_lr function to include min_lr calculation 2025-12-06
PR 0.00 feat: add merge_lora.py to support merging LoRA weights into 2025-12-05
PR 0.00 第一次尝试 2025-11-21
PR 0.00 [Security] Fix HIGH vulnerability: trailofbits.python.pickle 2025-11-19
PR 0.00 fix: attn_forwad when is_causal=True assert attn_mask is Non 2025-11-18
PR 0.00 Train_Grpo 添加注释 2025-10-26
PR 0.00 [feat] update model install method 2025-11-05
PR 0.00 [feat] add MNN support to README. 2025-11-10
PR 0.00 fix: Loading LoRA parameters which saved from multi-card tra 2025-11-06
PR 0.00 Update eval_llm.py 2025-11-03
PR 0.00 Merge pull request #1 from jingyaogong/master 2025-10-29
PR 0.00 Hope for Integrate swanlab希望集成SwanLab实验跟踪工具 2025-02-28
PR 0.00 新增注释,解释 Attention Trainer 细节 2025-08-15
PR 0.00 取消模型上下文限制,增加模型动态长度扩展机制,并保持前向兼容性 2025-07-09
PR 0.00 增加可选的MLA支持、修复模型内部精度一致,优化代码add mla, fix model dtype, improve 2025-02-28
PR 0.00 Improve training performance with torch.compile and torch.am 2025-05-21
PR 0.00 升腾NPU适配 2025-05-16
PR 0.00 完善注释及训练脚本 2025-05-08
PR 0.00 修改 serve_openai_api.py 的默认参数 2025-04-30
PR 0.00 Hotfix/issues 382 2025-04-29
PR 0.00 Update eval_model.py 2025-04-26
PR 0.00 sft should use pretrain model 2025-04-26
PR 0.00 完善 README 中关于加载已有模型的说明 2025-04-20
PR 0.00 chore: auto detect mps for pre train 2025-04-05
PR 0.00 Add Load ckpt 2025-04-03
PR 0.00 Little typo of readme 2025-03-09
PR 0.00 Add the interface testing interface for model API deployment 2025-02-21
PR 0.00 add smart gradient accumulation 2025-02-21
PR 0.00 Add ckp_dir and tokenizer path 2025-02-18
PR 0.00 修正了训练tokenizer中的chat_template中的逻辑,以及修正了tokenizer_config.json 2024-11-07
PR 0.00 Stabilize full SFT 2025-10-23
PR 0.00 接续训练 2025-05-28
PR 0.00 Fix Flash Attention attn_mask and is_causal conflict in Atte 2025-10-18
PR 0.00 Minimind 2025-10-10
PR 0.00 主要增加了直接使用huggingface模型的适配 2025-05-25
PR 0.00 update 2025-06-28
PR 0.00 Fix bug #329 top_p 参数由 int 类型调整为 float 2025-04-09
PR 0.00 123 2025-04-17
PR 0.00 移除构建输入文本时在开头和末尾重复添加的 2025-04-02
PR 0.00 feat: 优化导入和代码风格 2025-02-11
PR 0.00 Update README.md 2025-02-07
PR 0.00 fix weight initialization for residual block 2025-02-03
PR 0.00 fix 5-dpo train 2025-01-31
PR 0.00 Remove unnecessary code. 2024-12-03
PR 0.00 Update 5-dpo_train.py 2024-11-14
PR 0.00 fix 5-dpo_train.py bugs 2024-10-11
PR 0.00 修复wandb bug & 添加了argparse 2024-09-24
PR 0.00 添加了wandb 2024-09-23
PR 0.00 修复了data_process.py文件的bug 2024-09-23
PR 0.00 Update requirements.txt 2024-09-05