UbiquitousLearning/mllm issues and pull requests

#141 - Compile .so for Android

Issue - State: open - Opened by KAIWEILIUCC 5 days ago - 1 comment

#140 - The Llama 7B model works on my Android phone, but other models do not.

Issue - State: open - Opened by siz0001 5 days ago - 2 comments

#139 - I converted custom model, llama-ko-7b, and when play demo get this error message.

Issue - State: closed - Opened by siz0001 6 days ago - 4 comments

#138 - where is bin folder?

Issue - State: closed - Opened by siz0001 6 days ago - 6 comments

#137 - Using Qwen-2.0

Issue - State: open - Opened by KAIWEILIUCC 9 days ago - 2 comments

#136 - refactor:: remove Layer Class `Split`, replace it with `Tensor::split`

Pull Request - State: closed - Opened by yirongjie 17 days ago

#135 - APk BUild issues

Issue - State: open - Opened by Vinaysukhesh98 19 days ago

#134 - Failed to allocate memory error on Galaxy S24 NPU

Issue - State: open - Opened by gingerly 20 days ago

#133 - 运行时出现merge file is broken

Issue - State: closed - Opened by mailonghua 21 days ago - 2 comments

#132 - feat: add MiniCPM 2B demo

Pull Request - State: closed - Opened by yirongjie 21 days ago

#131 - 关于OpPackage-LLaMAAdd中的Q6_V_valign_VVR计算方式的疑惑

Issue - State: open - Opened by mailonghua 21 days ago

#130 - refactor: `Tensor::run` &`Layer::getFunc`

Pull Request - State: closed - Opened by yirongjie 23 days ago

#129 - fix: +-*/ for old front end

Pull Request - State: closed - Opened by yirongjie 24 days ago

#128 - run “run_qwen_npu.sh “ fail. chip is Snapdragon 870

Issue - State: closed - Opened by yangh0597 24 days ago - 1 comment

#127 - fix 修复windows环境

Pull Request - State: closed - Opened by WhiteNight123 25 days ago

#126 - How did you obtain the two model files, qwen-1.5-1.8b-chat-int8.mllm and qwen-1.5-1.8b-chat-q4k.mllm?

Issue - State: open - Opened by yhwang-hub 27 days ago - 3 comments

#125 - ./bin/main_qwen_npu fail

Issue - State: open - Opened by zcxo 27 days ago - 2 comments

#124 - ./run_qwen-npu.sh failed

Issue - State: closed - Opened by zcxo 27 days ago - 2 comments

#123 - ./run_qwen_npu.sh失败

Issue - State: closed - Opened by zcxo 27 days ago

#121 - refactor: `Tensor::run` &`Layer::getFunc`: Tensor& -> Tensor

Pull Request - State: closed - Opened by yirongjie about 1 month ago

#120 - refactor: `Layer::run` & `Tensor::getStaticFunc`

Pull Request - State: closed - Opened by yirongjie about 1 month ago

#119 - feat: add Phi-3-mini model

Pull Request - State: closed - Opened by WhiteNight123 about 1 month ago

#118 - 为什么预填充和解码不能都在 NPU 上运行？

Issue - State: open - Opened by yhwang-hub about 1 month ago - 4 comments

#117 - Segmentation fault on OPPO FindX7 Ultra (Snapdragon8Gen3)

Issue - State: open - Opened by bingo787 about 1 month ago - 1 comment

#116 - Prefill speed is approximately 4~6 tokens/s for Qwen1.5-1.8B

Issue - State: open - Opened by mengllm about 1 month ago - 5 comments

#115 - Crash on Xiaomi 14(8gen3) with QNN

Issue - State: open - Opened by zhuipiaochen about 1 month ago - 1 comment

#114 - Android crashed and forcely rebooted when executing main_qwen_npu

Issue - State: open - Opened by taegeonum about 1 month ago - 9 comments

#113 - doc: Update README.md

Pull Request - State: closed - Opened by xumengwei about 1 month ago

#112 - feat: Preliminary implementation on Qualcomm NPU (QNN) backend.

Pull Request - State: closed - Opened by liang1232018 about 1 month ago
Labels: good first issue

#111 - feat: llamafile_sgemm bias support

Pull Request - State: closed - Opened by chenghuaWang about 1 month ago

#110 - chore: Disable OpenMP for Mac.

Pull Request - State: closed - Opened by lx200916 about 1 month ago

#109 - feat: GEMV + Bias mixed precision support for ARM Devices

Pull Request - State: closed - Opened by chenghuaWang about 2 months ago

#108 - feat: add clear_kvcache && fix: BUG in quantize.

Pull Request - State: closed - Opened by yirongjie about 2 months ago

#107 - fix: Qwen v1.5 Tokenizer bug

Pull Request - State: closed - Opened by chenghuaWang about 2 months ago - 1 comment

#106 - feat: add DEBUGSAVETENSOR & DEBUGOPTIME

Pull Request - State: closed - Opened by yirongjie about 2 months ago

#105 - feat: topk/topp sampling

Pull Request - State: closed - Opened by chenghuaWang about 2 months ago - 1 comment

#104 - perf: add AArch64 GEMM/GEMV for q4_0.

Pull Request - State: closed - Opened by yirongjie about 2 months ago

#101 - perf: Use `vector<shared_ptr<Tensor>> Tensor::graphs`

Pull Request - State: closed - Opened by yirongjie about 2 months ago

#100 - feat: add Qwen 1.8B demo

Pull Request - State: closed - Opened by yirongjie 2 months ago

#99 - feat:Add OPT support

Pull Request - State: closed - Opened by yirongjie 2 months ago

#98 - feat: add elastic llama

Pull Request - State: closed - Opened by yirongjie 2 months ago

#97 - Is Subgraph Heterogeneous Compute Available in MLLM?

Issue - State: open - Opened by MaTwickenham 2 months ago - 2 comments

#96 - [Question] Does this demo support mediatek neuropilot?

Issue - State: closed - Opened by AkaneTan 2 months ago - 2 comments

#95 - doc: Update README.md

Pull Request - State: closed - Opened by yirongjie 2 months ago

#94 - feat: Stablelm 2 1.6b support

Pull Request - State: closed - Opened by emt0re0 2 months ago

#93 - Layer not found when trying mllm in Google Colab

Issue - State: closed - Opened by gsamaras 2 months ago - 9 comments

#92 - feat: Add new demo: demo_imagebind_1mod

Pull Request - State: closed - Opened by yirongjie 2 months ago

#91 - feat: Inference speed(tokens/s) profiling

Pull Request - State: closed - Opened by yirongjie 2 months ago

#90 - feat: Add Multi-Head Latent Attention(MLA) support.

Pull Request - State: closed - Opened by yirongjie 2 months ago

#89 - doc: Update README.md

Pull Request - State: closed - Opened by xumengwei 2 months ago

#88 - feat: Yi-1.5-6B support

Pull Request - State: closed - Opened by chenghuaWang 3 months ago

#87 - Update requirements.txt

Pull Request - State: closed - Opened by lx200916 3 months ago

#86 - feat: add sparse inference like powerinfer

Pull Request - State: closed - Opened by XieWeikai 3 months ago

#85 - feat: add sparse inference of sparse models like ReLU Llama

Pull Request - State: closed - Opened by XieWeikai 3 months ago

#84 - App download link cannot be successfully accessed

Issue - State: closed - Opened by jzyztzn 3 months ago - 4 comments

#83 - feat: mistral v0.2 7B support

Pull Request - State: closed - Opened by chenghuaWang 4 months ago

#82 - Detailed Android smart phone model and configuration

Issue - State: closed - Opened by baimuchu 4 months ago - 2 comments

#81 - CANNOT LINK EXECUTABLE "./demo_fuyu": library "libomp.so" not found: needed by main executable

Issue - State: closed - Opened by sharmilamani 5 months ago - 3 comments

#80 - improve examples

Issue - State: closed - Opened by zhehaoxu 5 months ago - 1 comment

#79 - Support for the QWen1.5-0.5B model

Pull Request - State: closed - Opened by chenghuaWang 5 months ago - 2 comments

#78 - Trying to add custom models but cannot run properly on Android

Issue - State: closed - Opened by TiAmoLip 5 months ago - 9 comments

#77 - How to check how many tokens/sec it was generated in android app

Issue - State: open - Opened by Vinaysukhesh98 5 months ago

#76 - Update README.md

Pull Request - State: closed - Opened by yirongjie 5 months ago

#75 - Single precision inference support for the gemma-2B model

Pull Request - State: closed - Opened by chenghuaWang 5 months ago

#74 - Android app crash while Image Reading

Issue - State: open - Opened by Vinaysukhesh98 6 months ago - 12 comments
Labels: bug

#72 - Feat: Fill in input Tensor

Pull Request - State: closed - Opened by yirongjie 6 months ago

#71 - 请教llava推理教程

Issue - State: closed - Opened by lijianxing123 6 months ago - 3 comments

#70 - Fix: Tensor:: mm(): reference not passed in as input

Pull Request - State: closed - Opened by yirongjie 6 months ago

#69 - Feat: Optimize the operation process

Pull Request - State: closed - Opened by yirongjie 6 months ago

#68 - Feat: Add new FrontEnd and model demos.

Pull Request - State: closed - Opened by yirongjie 6 months ago

#67 - how to add tinyllama to app

Issue - State: closed - Opened by alexliyang 7 months ago - 1 comment
Labels: enhancement

#66 - feat: Add OPT Tokenizer.

Pull Request - State: closed - Opened by lx200916 7 months ago

#65 - about test models

Issue - State: closed - Opened by alexliyang 7 months ago - 2 comments

#64 - fix: remove unused file

Pull Request - State: closed - Opened by yirongjie 8 months ago

#63 - Update README.md

Pull Request - State: closed - Opened by UbiquitousLearning 8 months ago

#62 - fix

Pull Request - State: closed - Opened by yirongjie 8 months ago

#61 - fix: app

Pull Request - State: closed - Opened by yirongjie 8 months ago

#60 - doc: README

Pull Request - State: closed - Opened by yirongjie 8 months ago

#59 - feat: New Model LLaVA

Pull Request - State: closed - Opened by yirongjie 8 months ago

#58 - Merge develop to main

Pull Request - State: closed - Opened by yirongjie 8 months ago - 1 comment

#57 - fix: fix vocab.py.

Pull Request - State: closed - Opened by lx200916 8 months ago

#56 - Add Video.

Pull Request - State: closed - Opened by lx200916 8 months ago

#55 - Merge pull request #54 from UbiquitousLearning/develop

Pull Request - State: closed - Opened by yirongjie 8 months ago

#54 - Backup NNAPI

Pull Request - State: closed - Opened by yirongjie 8 months ago

#53 - fix: qnn linear quantize tensor duplicate

Pull Request - State: closed - Opened by liang1232018 8 months ago

#52 - Develop qnn zh

Pull Request - State: closed - Opened by liang1232018 8 months ago

#51 - Develop qnn zh

Pull Request - State: closed - Opened by liang1232018 8 months ago