ggerganov/llama.cpp

Pull Requests Commits

Check for reverse prompt by characters instead of tokens (#292) (#330)

tjohnman committed 2 years ago

Verified d5f56a5e

Fix convert script, warnings alpaca instructions, default params

ggerganov committed 2 years ago

Verified 3bfa3b43

Add OpenBSD support (#314)

kevlo committed 2 years ago

Verified 715d292e

fix typo in comment (#318)

eiz committed 2 years ago

Verified c98ae026

Makefile: slightly cleanup for Mac Intel; echo instead of run ./main -h (#335)

mqy committed 2 years ago

Verified c3b2306b

cmdline option for custom amount of model parts (--n_parts N) (#348)

anzz1 committed 2 years ago

Verified 975d2ceb

Update IPFS links to quantized alpaca with new tokenizer format (#352)

antimatter15 committed 2 years ago

Verified e0ffc861

Change default repeat_penalty to 1.0

ggerganov committed 2 years ago

Verified 8f644a0a

Add tokenizer test + revert to C++11 (#355)

ggerganov committed 2 years ago

Verified eb34620a

Add initial AVX512 support for dot product on Linux (#320)

Ameobea committed 2 years ago

Verified 2e664f1f

Adding missing features of CMakeLists.txt & Refactoring (#131)

nusu-github committed 2 years ago

Verified 8cf9f34e

Nix flake: set meta.mainProgram to llama

siraben committed 2 years ago

bd4b46d6

Fixed tokenizer.model not found error when model dir is symlink (#325)

mqy committed 2 years ago

Verified 6b6d5b50

move file magic/version to header, print expected version (#319)

eiz committed 2 years ago

Verified a791a68b

Docker - Fix publish docker image in GitHub Registry (#235)

bernatvadell committed 2 years ago

Verified 0f1b21cb

sentencepiece bpe compatible tokenizer (#252)

eiz committed 2 years ago

Verified 074bea2e

Add tqdm to Python requirements (#293)

sw committed 2 years ago

Verified 5cb63e24

bugfix: default should not be interactive (#304)

cocktailpeanut committed 2 years ago

Verified da5303c1

ggerganov committed 2 years ago

Verified 4545539d

Add temporary helper script for Alpaca chat

ggerganov committed 2 years ago

Verified edeba283

fix coloring of last `n_batch` of prompt, and refactor line input (#221)

bitRAKE committed 2 years ago

Verified 5c19c70b

Support for multiple reverse prompts. (#299)

tjohnman committed 2 years ago

Verified 24568371

Improved quantize script (#222)

SuajCarrot committed 2 years ago

Verified 7392f1cd

Make prompt randomization optional. (#300)

tjohnman committed 2 years ago

Verified ad5fd5b6

Respect the maximum number of tokens in interactive. (#298)

tjohnman committed 2 years ago

Verified 368d0c8a

Add --ignore-eos parameter (#181)

slaren committed 2 years ago

Verified 50fae10d

interactive mode: print '\n' in sigint_handler, this flush stdout thus ensure color reset. (#283)

mqy committed 2 years ago

Verified 084e2f0e

Command line switch to use F16 for memory_k and memory_v (refactor of #154) (#294)

Green-Sky committed 2 years ago

Verified 0b366e73

Update hot topics to mention Alpaca support

ggerganov committed 2 years ago

Verified 160bfb21

Fix off-by-one bug (#115)

ggerganov committed 2 years ago

Verified c494ed5b

Newer Older