ggerganov/llama.cpp

Pull Requests Commits

undo formatting

monatis committed 2 years ago

51105572

gitignore *.gguf

monatis committed 2 years ago

0c43a3b7

rm example.gguf

monatis committed 2 years ago

8e62d2b2

fix : fix errors upd writing example

monatis committed 2 years ago

62f4926b

refactor : rm unused import and upd todos

monatis committed 2 years ago

94112505

GGUF : Support writing tensors in Python

monatis committed 2 years ago

bb54d170

WIP: Write tensor

monatis committed 2 years ago

464192b9

convert.py : remove GGML vocab + other obsolete stuff

ggerganov committed 2 years ago

Verified d2bb3ac1

convert.py : start a new simplified implementation by removing old stuff

ggerganov committed 2 years ago

Verified 68f53485

gguf.py : some code style changes

ggerganov committed 2 years ago

Verified 158be8f7

gguf : add array support

ggerganov committed 2 years ago

Verified d2b6ca13

gguf : expose the gguf_type enum through the API for now

ggerganov committed 2 years ago

Verified d89533df

refactor : reduce code duplication and better API (#2415)

monatis committed 2 years ago

Verified c85d3178

gguf : add comments

ggerganov committed 2 years ago

Verified d8491fc7

gguf : read / write sample models

ggerganov committed 2 years ago

Verified 5628ec71

gguf : gguf.c is now part of ggml.c

ggerganov committed 2 years ago

Verified e46870f5

gguf : simplify gguf_get_val

ggerganov committed 2 years ago

Verified d313c0fa

gguf : do not support passing existing ggml_context to gguf_init

ggerganov committed 2 years ago

Verified cb871fa0

gguf : add gguf_get_tensor_name()

ggerganov committed 2 years ago

Verified 860c9c63

gguf : initial model loading - not tested

ggerganov committed 2 years ago

Verified 78b226a9

gguf : read tensor info

ggerganov committed 2 years ago

Verified d91b985d

gguf : read header + meta data

ggerganov committed 2 years ago

Verified 8d6acfec

gguf : first API pass

ggerganov committed 2 years ago

Verified 68731487

ci : disable CI temporary to not waste energy

ggerganov committed 2 years ago

Verified 7e82d25f

wip : implement GGUF (#2397)

monatis committed 2 years ago

Verified bae6b125

ggerganov committed 2 years ago

Verified 4d698495

ggml : allocate graphs in a context (#2392)

slaren committed 2 years ago

Verified 5488fb78

Add LLAMA_DEFAULT_RMS_EPS so we can change the default (#2384)

ikawrakow committed 2 years ago

Verified eb542d39

ggml : fix ggml_flash_attn to use op_params (#2387)

slaren committed 2 years ago

Verified 07aaa0f6

convert.py : support bpe tokenizer (#2228)

ftgreat committed 2 years ago

Verified fce48caf

Older