skeskinen/bert.cpp

Issues

• Is it possible to make multiple socket connection at the same time from same machine.

  #44 opened 4 months ago by ZTAP0011
  1

• Segfault on large inputs?

  #37 opened 9 months ago by corani
  2

• How can I increase n_max_tokens?

  #43 opened 4 months ago by ztkg257
  1

• How do I properly build this on windows?

  #42 opened 5 months ago by qwbarch
  1

• Any plan to support GGUF?

  #41 opened 6 months ago by Jeevhi
  0

• Is this ever going to be updated?

  #39 opened 7 months ago by BBC-Esq
  6

• Does this support CUDA?

  #40 opened 7 months ago by SpaceCowboy850
  1

• Can this handle classification models, e.g. MNLI?

  #12 opened a year ago by redthing1
  3

• Issue reading ggml-model-q4_0.bin

  #38 opened 8 months ago by thijse
  4

• Tokenizer in bert.cpp is not good enough, how about `tokenizers-cpp`

  #36 opened 10 months ago by FFengIll
  11

• Error Running Server Example

  #28 opened 10 months ago by BradSmith2015
  1

• Classification Pipeline

  #11 opened a year ago by okpatil4u
  18

• subword `#` should be an option.

  #33 opened 10 months ago by FFengIll
  6

• Support for cjk. Please！and get an error on windows

  #35 opened 10 months ago by cgisky1980
  1

• converter does not work with the current ggml

  #23 opened a year ago by yunghoy
  4

• Error load model

  #34 opened 10 months ago by AgostinoA
  0

• Comparison of timing results

  #30 opened 10 months ago by mgoldenbe
  0

• throwing the error while running the model

  #27 opened 10 months ago by Jasheenshaik
  0

• [Regression] WASM alignment fault after ggml update

  #26 opened 10 months ago by ivanstepanovftw
  3

• update to current ggml

  #24 opened a year ago by manwantnosmoke
  2

• Can not infer quantized model, but fp32 works well.

  #20 opened a year ago by znsoftm
  9

• none

  #21 opened a year ago by znsoftm
  0

• About the calculation of overhead.

  #19 opened a year ago by znsoftm
  4

• Unable to build static library

  #17 opened a year ago by jjzazuet
  5

• why does the program ignore the three layers?

  #18 opened a year ago by znsoftm
  2

• I really appreciate your kind words about our project! May I ask which model and configuration of CPU you are using?

  #16 opened a year ago by zhengbohennuli
  0

• CodeBERT

  #7 opened a year ago by walking-octopus
  3

• Cannot build " set_target_properties Can not find target to add properties to: ggml"

  #14 opened a year ago by chrisbward
  1

• Running `bert_load_from_file` from code results with ` 'embeddings.word_embeddings.weight' has wrong shape in model file`

  #13 opened a year ago by NitayRabi
  1

• BrokenPipeError: [Errno 32] Broken pipe

  #8 opened a year ago by virajk31
  1

• convert 'chatglm-6b' model failed

  #5 opened a year ago by lindeer
  1

• implement do_handle_chinese_characters in tokenizing

  #1 opened a year ago by skeskinen
  1

• Provide C api in bert.h

  #2 opened a year ago by skeskinen
  0

• Q&A example?

  #3 opened a year ago by walking-octopus
  1