Skip to content
Snippets Groups Projects
Select Git revision
  • GLM-130B
  • MoE
  • adamop
  • adapt_env
  • args
  • beamsearch
  • bert
  • bert_large
  • bert_new
  • cait
  • chatglm-rotary
  • clever_dataset
  • clip
  • cogvideo
  • config
  • cp_support
  • cross
  • dev
  • dev_fixnan
  • develop
  • v0.1.10
21 results
You can move around the graph by using the arrow keys.
Created with Raphaël 2.2.021Apr18121110973231Mar302928272210Jan25Dec17Nov113Sep228Aug22191816530Jul24222120191817161514111097230Jun272625222120181716151412109876543227May262523212019181715141312115428Apr2625222120191816141312109828Mar27241623Feb1918161427Jan24201918171514131211976531Dec3029282322212019181716151312111095432130Nov262523211914987653231Oct30292827262524232221201918141098765123Sep2216830Aug20181022Jun181716Merge pull request #85 from THUDM/optional_arch_args_savefix argstry & support cross_attn_hidden_sizesupport file with non-ascii charactersMerge branch 'v0.3' into mainfix deepspeed mpu simpleinit error and support 0.9Merge pull request #83 from THUDM/registrydynamic loadingregistryregistryMerge pull request #80 from zhangfanTJU/fix-chatmodel-orderUpdate chat_model.pyUpdate README.mdfix-chatmodel-ordersupport batch generation for chatglmMerge pull request #79 from THUDM/v0.3Merge branch 'main' of github.com:THUDM/SwissArmyTransformer into v0.3v0.3v0.3Merge pull request #75 from xloem/cross-attentionchange the pkg name back .change from_pretrained args-name orderMerge branch 'main' of github.com:THUDM/SwissArmyTransformer into v0.3v.0.3.1 refactor SwissArmyTransformer as satv0.3.0list info & model-only modefix chatglm pad mask and position bugchatglm batch generationupdate chatglm by official repotmpsavefix by fake some argsfix_iterable_va…fix_iterable_val_lastbatchfix chatglm finetune padupdate chatglm-6b inferenceupdate chatglm-6b finetunefinetune chatglm-6bupdate chatglm and eva2 inferenceeva2 modelchatglm-6b chat examplechatglm-6bgpt2 model (#77)gpt2 modelgptgptUpdate tokenizerSupport 2D position encoding for GLM130B modelRe-set cross attention forward parameter for T5
Loading