rustformers/llm

Issues

• Sub reddit is down

  #451 opened 4 months ago by Gamithon
  2

• Question: GPU support

  #456 opened a month ago by toastxc
  0

• Not enough space in the context's memory pool

  #455 opened 2 months ago by thecodechemist99
  0

• Out of context answers

  #453 opened 3 months ago by DravidVaishnav
  0

• Include bytes of model in program?

  #452 opened 3 months ago by Lamby777
  1

• Currently in dev any inference is broken

  #450 opened 4 months ago by gadLinux
  2

• When using tokio and HuggingFaceRemote it breaks dropping the runtime

  #449 opened 4 months ago by gadLinux
  1

• Disable tokenizers-remote support for the library by default

  #436 opened 7 months ago by philpax
  1

• Build fails: error: no such file or directory: 'ggml/src/ggml.c'

  #448 opened 5 months ago by yurivict
  0

• Behavior when missing quantization version

  #447 opened 6 months ago by Reichenbachian
  1

• EOS is not read from gguf format

  #446 opened 6 months ago by Alisa-lisa
  1

• Support Separate Loading of Vocabulary or Tensors

  #445 opened 6 months ago by skirodev
  0

• Support for Mistral-7b

  #434 opened 8 months ago by ranaya-formant
  5

• Why is the feed_prompt process so slow?

  #439 opened 7 months ago by zackshen
  5

• Video Tutorials

  #399 opened 10 months ago by okpatil4u
  3

• Reduce dependencies

  #437 opened 7 months ago by philpax
  0

• Supporting Llama-2 70B param

  #402 opened 10 months ago by AmineDiro
  6

• Api Server Example

  #414 opened 9 months ago by e253
  5

• How to disable ggml logging?

  #433 opened 9 months ago by mrwilby
  1

• Default String for ConfiguredSamplers

  #420 opened 9 months ago by JuliaMerz
  1

• Clarify MSRV policy

  #431 opened 9 months ago by chris-ha458
  2

• How do I use Huggingface tokenization to use a model on Huggingace in MODEL_PATH instead of my local machine?

  #427 opened 9 months ago by alfellati
  1

• Medusa Speculative Decoding

  #423 opened 9 months ago by someone13574
  1

• SIGTRAP triggered on MacOS

  #422 opened 9 months ago by jafioti
  2

• NaN logits on LLaMA 65B when using 2k+ token contexts

  #418 opened 9 months ago by hugoabonizio
  0

• Metal Prompt Feeding

  #403 opened 10 months ago by jafioti
  6

• WizardCoder llama assert failure

  #417 opened 9 months ago by jacohend
  3

• AMD ROCm support with HIPBLAS

  #415 opened 9 months ago by xangelix
  2

• GPT-2 load errors

  #397 opened 9 months ago by pabl-o-ce
  5

• LLaMA-2 GGML formats fail to generate any new token

  #413 opened 9 months ago by AnubhabB
  0

• System prompts

  #411 opened 10 months ago by kpcyrd
  1

• Issues using with whisper-rs

  #408 opened 10 months ago by jafioti
  8

• Proper Rewind+Refeed when stop token is detected.

  #407 opened 10 months ago by JuliaMerz
  1

• Cannot Compile llm-base on main

  #405 opened 10 months ago by carllippert
  4

• Installation of CLI on macOS

  #388 opened 10 months ago by twardoch
  2

• stack overflow after new merge

  #391 opened 10 months ago by fluffydolphin
  3

• Add classifier-free guidance

  #377 opened a year ago by philpax
  2

• Free tensors from RAM if they are offloaded to an Accelerator

  #390 opened 10 months ago by LLukas22
  0

• Build failure on Windows

  #396 opened 10 months ago by NuSkooler
  4

• WizardLM inference error: ggml-metal.m:773: false && "not implemented"

  #383 opened a year ago by clarkmcc
  6

• Halting inference works for session.infer but not for session.feed_prompt

  #393 opened 10 months ago by clarkmcc
  0

• Reddit community Wiki page is disabled

  #392 opened 10 months ago by viktor-ferenczi
  1

• Implement SuperHOT/interpolated RoPE support

  #378 opened 10 months ago by philpax
  1

• [compile fail] Undefined symbols for architecture x86_64: "_ggml_graph_plan", referenced from: ggml::GraphExecutionPlan:: new ::h19443c1d77cb997e in libggml-a3125eb72f014293.rlib(ggml-a3125eb72f014293.ggml.9c60abbf9368402d-cgu.1.rcgu.o)

  #387 opened 10 months ago by twardoch
  1

• Runtime GPU backend selection

  #386 opened a year ago by philpax
  4

• Invalid magic number while loading LoRAs

  #384 opened a year ago by clarkmcc
  2

• Any plans for being able to fine tune models in rustformers llm?

  #382 opened a year ago by moonstripe
  1

• I can't compile llm with nix

  #374 opened a year ago by GENDRAUD
  6

• Certain quantization levels produce garbage with CUDA acceleration

  #373 opened a year ago by philpax
  1

• Shared `ExecutionParameters` for all operations

  #372 opened a year ago by philpax
  0