Commits

Branch selector

 master

User selector

All users

All time

Commit History

Commits on Apr 9, 2023

• Merge pull request #6 from DaniAsh551/master

  Noeda

  committedApr 9, 2023
  1e1131f

Commits on Apr 8, 2023

• docker:nvidia opencl support with instructions

  DaniAsh551

  committedApr 8, 2023
  faf98bd

• readme: nvidia RTX GPU naming

  DaniAsh551

  committedApr 8, 2023
  32eac9e

Commits on Apr 7, 2023

• Name a CPU-only Dockerfile since the NVidia turned out to be complicated to make actually work due to NVidia's own shenanigans.

  Noeda

  committedApr 7, 2023
  059c948

• Merge branch 'master' of github.com:DaniAsh551/rllama

  DaniAsh551

  committedApr 7, 2023
  5b76ef4

• opencl docker support for nvidia GPUs

  DaniAsh551

  committedApr 7, 2023
  3f35a84

• Fix interactive system prompt overwriting --prompt.

  Noeda

  committedApr 7, 2023
  8d6897e

Commits on Apr 6, 2023

• Small fix to README.md on the Vicuna13B interactive thing.

  Noeda

  committedApr 6, 2023
  61b68c6

• Adjust default interactive settings so they match what Vicuna-13B expects.

  Noeda

  committedApr 6, 2023
  7aa1051

• Update benchmark numbers in README.md after new OpenCL changes.

  Noeda

  committedApr 6, 2023
  4faf07a

• Accidentally broke build. Unbreak it.

  Noeda

  committedApr 6, 2023
  6751f59

• Force f16 if OpenCL is on, because otherwise we will crash.

  Noeda

  committedApr 6, 2023
  f5328ab

Commits on Apr 5, 2023

• Write special OpenCL kernel when left side matrix has only 1 row.

  Noeda

  committedApr 5, 2023
  746fc56

• Update goals, as Stanford's model is no longer that relevant in the face of Vicuna-13B and Vicuna-13B works.

  Noeda

  committedApr 5, 2023
  4a1e295

• Add support to only load a model partially to GPU.

  Noeda

  committedApr 5, 2023
  e4af9d9

• Mention that Vicuna-13B works in the README.md

  Noeda

  committedApr 5, 2023
  35a3a5c

• Fix warnings.

  Noeda

  committedApr 5, 2023
  88cb876

• Add Huggingface model loading (Vicuna-13B) support.

  Noeda

  committedApr 5, 2023
  e27e649

Commits on Apr 2, 2023

• Fix some awkward wording in README.md

  Noeda

  committedApr 2, 2023
  48244ab

• Don't report interactive stop token sequence if we are not interactive mode.

  Noeda

  committedApr 2, 2023
  4c99e7f

• Add --interactive-prompt-postfix and update README.md a bit on this stuff.

  Noeda

  committedApr 2, 2023
  f5f8547

• Erase debug println!() I had.

  Noeda

  committedApr 2, 2023
  19e552e

• Document interactive mode, run rustfmt on rllama_main.rs

  Noeda

  committedApr 2, 2023
  39f69f2

Commits on Apr 1, 2023

• Add interactive mode in cli

  kazord

  committedApr 1, 2023
  ec514ec

Commits on Mar 21, 2023

• Add skeleton code for 4-bit quantization.

  Noeda

  committedMar 21, 2023
  f6249e8

• Add a flag that will exit the HTTP server after just one query.

  Noeda

  committedMar 21, 2023
  26f343a

• Mention that server feature must be turned on to use the inference API.

  Noeda

  committedMar 21, 2023
  957a8f9

• Fix compilation when opencl feature is being used.

  Noeda

  committedMar 21, 2023
  5e24172

• Mention HTTP server in features in README.md

  Noeda

  committedMar 21, 2023
  d85ed7f

• Fix some things in README.md after proofreading it and removing lies.

  Noeda

  committedMar 21, 2023
  a832061

• Add simple HTTP API support.

  Noeda

  committedMar 21, 2023
  b9be485

Commits on Mar 18, 2023

• Refactor all SIMD to one file, simd_support.rs

  Noeda

  committedMar 18, 2023
  9c86c17

• Update README.md on LLaMA-65B benchmark result.

  Noeda

  committedMar 18, 2023
  25e3e12

• Forgot to mark last benchmark at March 17

  Noeda

  committedMar 18, 2023
  f233f8a

• Update README.md, add a nice animation.

  Noeda

  committedMar 18, 2023
  db0f22e

Pagination

PreviousNext