Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / THUDM/GLM issues and pull requests

#208 - 110M的模型有huggingface版本吗？

Issue - State: open - Opened by WzjCoder 3 months ago

#207 - 建议上架ollama

Issue - State: open - Opened by heimy2000 4 months ago

#205 - 模型的分词逻辑

Issue - State: open - Opened by loki-keroro 4 months ago

#204 - Add special_token

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

#203 - Few-shot tests on GLM-10B

Issue - State: open - Opened by Vispstar-V 6 months ago

#202 - What is the license of Pretrained Models?

Issue - State: open - Opened by phuchm 8 months ago

#201 - bug report！rouge-1 = 0.0000 rouge-2 = 0.0000 rouge-l = 0.0000

Issue - State: closed - Opened by runningabcd 8 months ago - 1 comment

#197 - mpi4py库

Issue - State: closed - Opened by XiaozhuLove 10 months ago

#196 - 使用glm-large-chinese微调分类任务

Issue - State: open - Opened by mechigonft 11 months ago

#193 - Create README_zh.md

Pull Request - State: open - Opened by great-wind 11 months ago

#192 - 使用GLM-2b推理时生成无意义内容

Issue - State: open - Opened by ChristLBUPT about 1 year ago

#191 - 运行 GLM-10B 的最低配置是多少？

Issue - State: open - Opened by nguyenvanhoangphuc about 1 year ago - 1 comment

#190 - 使用Zero-1+cpu_offload=true时，出现错误？

Issue - State: open - Opened by SkrDrag about 1 year ago - 2 comments

#187 - 使用glm-10b-chinese调用generate方法有时时会出错

Issue - State: open - Opened by adzhua about 1 year ago - 1 comment

#185 - 请问有人使用GLM跑通过Continual Pre-training么？

Issue - State: open - Opened by wjn1996 about 1 year ago

#184 - glm-10b-chinese原始模型推理报错

Issue - State: open - Opened by Mryangkaitong about 1 year ago

#183 - 有对glm-10b-chinese这个模型做过评测的吗？

Issue - State: open - Opened by hegang1-tal about 1 year ago - 1 comment

#182 - 如何将GLM10B封装成对话式API

Issue - State: open - Opened by yihuaxiang over 1 year ago - 3 comments

#180 - glm-10b / tokenization_glm.py

Issue - State: open - Opened by chenhaoenen over 1 year ago

#177 - 求问glm-10b-chinese推理所需最低配置

Issue - State: open - Opened by TianYangCai over 1 year ago

#176 - block_lm_ratio参数

Issue - State: closed - Opened by chenhaoenen over 1 year ago - 1 comment

#174 - Eligibility for Commercial Use

Issue - State: open - Opened by Hegelim over 1 year ago - 1 comment

#173 - 在没有InfiniBand情况下能训练glm-large吗

Issue - State: open - Opened by allendred over 1 year ago - 3 comments

#172 - GLM-10B中文版预训练权重下载后解压失败

Issue - State: open - Opened by echosyy over 1 year ago

#170 - 在预训练Pretrain时报no valid `self._rcvd_idx` is found错误

Issue - State: open - Opened by yt7589 over 1 year ago - 3 comments

#169 - parameter SCB

Issue - State: open - Opened by zhaoqf123 over 1 year ago - 1 comment

#167 - add comments and add ds_block_tiny for testing

Pull Request - State: open - Opened by LucienShui over 1 year ago

#166 - glm-10b-chinese模型的预训练数据量

Issue - State: open - Opened by wlike over 1 year ago

#165 - .

Issue - State: closed - Opened by xxzcc over 1 year ago

#163 - GLM 10B和ChatGLM 6B模型架构的差别

Issue - State: open - Opened by ccsquare over 1 year ago - 3 comments

#162 - 用"THUDM/glm-10b-chinese"做分类任务出错

Issue - State: closed - Opened by 18335100284 over 1 year ago - 2 comments

#161 - GLM-10B-Chinese模型文件太大无法解压

Issue - State: closed - Opened by wusi1590 over 1 year ago - 1 comment

#160 - 我基于10B模型做继续训练，loss只从11下降到5

Issue - State: open - Opened by TccccD over 1 year ago - 6 comments

#156 - 请问pretrain怎么控制训练的epoch数？

Issue - State: open - Opened by wusi1590 over 1 year ago - 7 comments

#155 - self.num_samples = 1000 * self.ds_len

Issue - State: open - Opened by superhg over 1 year ago

#154 - glm-large-chinese-335M重复生成

Issue - State: open - Opened by zh25714 over 1 year ago - 3 comments

#153 - 使用 kubeflow 启动分布式训练

Issue - State: open - Opened by EthanChen1234 over 1 year ago

#152 - 请问GLM_large_chinese的预训练语料在哪里找

Issue - State: open - Opened by zhangzai666 over 1 year ago

#151 - 请问10B-chinese的模型文件里为什么没有词表？

Issue - State: open - Opened by ggjge over 1 year ago - 1 comment

#150 - Accelerate support for GLM

Issue - State: open - Opened by larrylawl over 1 year ago

#145 - 使用p-tuning去finetune glm-large-chinese模型时 --continuous-prompt

Issue - State: closed - Opened by Chenchenwei over 1 year ago - 3 comments

#142 - accelerate 找不到模型

Issue - State: open - Opened by yata0 over 1 year ago - 7 comments
Labels: bug

#141 - GLM-10B 模型效率问题

Issue - State: open - Opened by tqjack over 1 year ago

#140 - BUG: GLM-10B-Chinese model generate " ⁇".

Issue - State: open - Opened by Tebmer over 1 year ago - 3 comments

#139 - 基于Prompt数据集如何微调模型？

Issue - State: open - Opened by xyjsjruiliu over 1 year ago

#138 - How to set hyperparameters during pretraining glm_doc?

Issue - State: open - Opened by ymr12 over 1 year ago

#137 - 模型并行训练结束后，如何将多个模型文件合并成一个？

Issue - State: closed - Opened by sduhkn over 1 year ago - 2 comments

#136 - 基于10B模型继续预训练，遇到world size 不一致导致报错

Issue - State: open - Opened by JinmingZhao over 1 year ago - 2 comments

#135 - cmrc数据集结果，预测结果都为空

Issue - State: open - Opened by yxk9810 over 1 year ago

#134 - chatglm-6b

Issue - State: open - Opened by mx8435 over 1 year ago - 1 comment

#133 - 单卡pretrain chinese-large模型

Issue - State: closed - Opened by yxk9810 over 1 year ago

#132 - impelement by megengine

Issue - State: open - Opened by chenqy4933 over 1 year ago

#130 - GLM 10B 模型零样本结果无法对齐

Issue - State: open - Opened by LemonNoel over 1 year ago

#129 - Update README.md

Pull Request - State: open - Opened by eltociear over 1 year ago

#128 - The attention mask and the pad token id were not set问题

Issue - State: open - Opened by taofennanhai over 1 year ago - 1 comment

#127 - Correct a typo in layers.py

Pull Request - State: open - Opened by felixonmars over 1 year ago

#125 - GPT2Dataset和BlockDataset

Issue - State: open - Opened by jiangix-paper over 1 year ago

#124 - 小数据finetune large-chinese rouge 为0

Issue - State: open - Opened by yxk9810 over 1 year ago - 2 comments

#122 - 50035 token id 报错

Issue - State: open - Opened by Mryangkaitong over 1 year ago - 1 comment

#120 - GLMForSequenceClassification的使用

Issue - State: open - Opened by Mryangkaitong over 1 year ago - 2 comments

#119 - AutoModelForCausalLM

Issue - State: open - Opened by beautifull4frank over 1 year ago - 1 comment

#101 - 运行 ds_finetune_superglue.sh key error "dev-0"

Issue - State: closed - Opened by lulia0228 over 1 year ago - 1 comment

#100 - 多机并行可以给点示例吗？

Issue - State: closed - Opened by lulia0228 over 1 year ago - 3 comments

#98 - WudaoCorpus-Dialog

Issue - State: closed - Opened by jiangix-paper over 1 year ago - 1 comment

#97 - where I can check the source code of model.generate()

Issue - State: closed - Opened by yuanjames over 1 year ago - 1 comment

#96 - GLM-10B怎么使用模型并行

Issue - State: open - Opened by Yiran-Zhu over 1 year ago - 2 comments

#95 - 如何使用在huggingface下载的离线模型推理glm-10b-chinese?

Issue - State: open - Opened by vicwer over 1 year ago - 2 comments

#94 - 使用ds_pretrain_nvidia.sh后模型生成异常

Issue - State: closed - Opened by yipintiancheng over 1 year ago - 3 comments