klosax
fb11dd3f92
common.h : .bin --> .gguf
2023-08-17 19:16:35 +02:00
Georgi Gerganov
e72c8c2124
ggml : fix bug in gguf_set_kv
...
ggml-ci
2023-08-17 20:13:48 +03:00
Georgi Gerganov
899f9a5350
llama : fix lambda capture
...
ggml-ci
2023-08-17 19:49:45 +03:00
Georgi Gerganov
93f285bdf1
gptneox : move as a WIP example
2023-08-17 19:49:45 +03:00
Georgi Gerganov
81a2c2a6f4
llama : fix llama_model_loader memory leak
2023-08-17 19:49:02 +03:00
Georgi Gerganov
dd9e2fc988
ci : update ".bin" to ".gguf" extension
...
ggml-ci
2023-08-17 19:32:14 +03:00
Georgi Gerganov
c3b739374e
editorconfig : ignore models folder
...
ggml-ci
2023-08-17 19:17:25 +03:00
Georgi Gerganov
6d66ef96eb
Merge branch 'master' into gguf
2023-08-17 19:04:59 +03:00
Georgi Gerganov
11bf4366c2
llama : sync with recent PRs on master
2023-08-17 19:03:15 +03:00
Georgi Gerganov
8ace03ad3d
convert.py : better always have n_head_kv and default it to n_head
2023-08-17 18:47:06 +03:00
klosax
d646c4efce
convert.py : n_head_kv optional and .gguf file extension
2023-08-17 17:20:36 +02:00
Georgi Gerganov
dd016cc246
Revert "ci : disable CI temporary to not waste energy"
...
This reverts commit 7e82d25f40
2023-08-17 17:23:16 +03:00
Georgi Gerganov
2ddd9681d6
convert.py : update to support GGUF output
2023-08-17 17:22:43 +03:00
Georgi Gerganov
e0429d38e4
convert-new.py : output gguf ( #2635 )
...
* convert-new.py : output gguf (WIP)
* convert-new.py : add gguf key-value pairs
* llama : add hparams.ctx_train + no longer print ftype
* convert-new.py : minor fixes
* convert-new.py : vocab-only option should work now
* llama : fix tokenizer to use llama_char_to_byte
* tests : add new ggml-vocab-llama.gguf
* convert-new.py : tensor name mapping
* convert-new.py : add map for skipping tensor serialization
* convert-new.py : convert script now works
* gguf.py : pick some of the refactoring from #2644
* convert-new.py : minor fixes
2023-08-17 17:19:52 +03:00
Kerfuffle
8dae7ce684
Add --cfg-negative-prompt-file option for examples ( #2591 )
...
Add --cfg-negative-prompt-file option for examples
2023-08-17 07:29:44 -06:00
klosax
d6fd53afd6
llama.cpp : use ggml_elements()
2023-08-17 15:24:35 +02:00
klosax
5a0a2c5685
llama.cpp : print actual model size
2023-08-17 15:18:16 +02:00
Georgi Gerganov
a73ccf1aa3
llama : replace (permute + reshape + view_1d) with (view_3d) ( #2538 )
...
ggml-ci
2023-08-17 10:47:09 +03:00
drbh
7cf54e1f74
tests : adds simple llama grammar tests ( #2618 )
...
* adds simple llama grammar tests
* fix lint and add Makefile
* 0 terminate code_points
* avoid dangling pointers in candidate cleanup
* cleanup grammar at end of test
2023-08-17 10:41:01 +03:00
Shouzheng Liu
a872a2b28e
ggml-alloc : fix discrepency between measure&eval ( #2639 )
...
The GGML memory allocator consistently places a tensor within the
optimal-fit memory block, which is the smallest block capable of
accommodating the tensor's size. During the measurement phase, the final
block is generously sized, ensuring it never qualifies as the
optimal-fit block as long as there exists another block capable of
accommodating the tensor. Nevertheless, in the evaluation phase, the
last block is constrained in size and could potentially qualify as the
optimal-fit block. Consequently, there exists the possibility of a
tensor being allocated to a different region during evaluation, leading
to more memory fragmentation in our scratch buffer.
This recent commit guarantees uniform behavior of the allocator across
both the measurement and evaluation phases, eliminating discrepancies
between the two.
2023-08-17 10:35:53 +03:00
M. Yusuf Sarıgöz
42f8fe1927
examples/gguf : no need to keep q option for quantization any more
2023-08-17 08:56:42 +03:00
Kolen Cheung
0919a0f73d
cmake : install ggml-meta.metal if LLAMA_METAL ( #2449 )
2023-08-16 23:09:49 +03:00
Jhen-Jie Hong
ed53db86c3
metal : print error of load pipeline state ( #2564 )
...
* metal : print error of load pipeline state
* metal : return null if load pipeline failed
2023-08-16 23:09:03 +03:00
Shouzheng Liu
fc8ef549e5
metal : enable ggml-alloc ( #2627 )
...
* metal: enable ggml-alloc
Make ggml-alloc work with concurrently dispatch.
* style-fix
Co-authored-by: slaren <slarengh@gmail.com>
---------
Co-authored-by: slaren <slarengh@gmail.com>
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
2023-08-16 23:08:28 +03:00
Shouzheng Liu
bf83bff674
metal : matrix-matrix multiplication kernel ( #2615 )
...
* metal: matrix-matrix multiplication kernel
This commit removes MPS and uses custom matrix-matrix multiplication
kernels for all quantization types. This commit also adds grouped-query
attention to support llama2 70B.
* metal: fix performance degradation from gqa
Integers are slow on the GPU, and 64-bit divides are extremely slow.
In the context of GQA, we introduce a 64-bit divide that cannot be
optimized out by the compiler, which results in a decrease of ~8% in
inference performance. This commit fixes that issue by calculating a
part of the offset with a 32-bit divide. Naturally, this limits the
size of a single matrix to ~4GB. However, this limitation should
suffice for the near future.
* metal: fix bugs for GQA and perplexity test.
I mixed up ne02 and nb02 in previous commit.
2023-08-16 23:07:04 +03:00
Georgi Gerganov
5ec18934ad
convert-new.py : pick #2427 for HF 70B support
2023-08-16 20:16:15 +03:00
Georgi Gerganov
c8ee87f141
gguf.py : merge all files in gguf.py
2023-08-16 19:55:49 +03:00
Georgi Gerganov
88b5769487
gguf : deduplicate ( #2629 )
...
* gguf : better type names
* dedup : CPU + Metal is working
* ggml : fix warnings about unused results
* llama.cpp : fix line feed and compiler warning
* llama : fix strncpy warning + note token_to_str does not write null
* llama : restore the original load/save session implementation
Will migrate this to GGUF in the future
* convert-llama-h5-to-gguf.py : support alt ctx param name
* ggml : assert when using ggml_mul with non-F32 src1
* examples : dedup simple
---------
Co-authored-by: klosax <131523366+klosax@users.noreply.github.com>
2023-08-16 19:25:29 +03:00
Georgi Gerganov
758ff1bbb5
llama : refactor model loading code ( #2620 )
...
* llama : style formatting + remove helper methods
* llama : fix quantization using gguf tool
* llama : simplify gguf_file_saver
* llama : fix method names
* llama : simplify write_header()
* llama : no need to pass full file loader to the file saver
just gguf_ctx
* llama : gguf_file_saver write I32
* llama : refactor tensor names (#2622 )
* gguf: update tensor names searched in quantization
* gguf : define tensor names as constants
* gguf : initial write API (not tested yet)
* gguf : write to file API (not tested)
* gguf : initial write API ready + example
* gguf : fix header write
* gguf : fixes + simplify example + add ggml_nbytes_pad()
* gguf : minor
* llama : replace gguf_file_saver with new gguf write API
* gguf : streaming support when writing files
* gguf : remove oboslete write methods
* gguf : remove obosolete gguf_get_arr_xxx API
* llama : simplify gguf_file_loader
* llama : move hparams and vocab from gguf_file_loader to llama_model_loader
* llama : merge gguf-util.h in llama.cpp
* llama : reorder definitions in .cpp to match .h
* llama : minor simplifications
* llama : refactor llama_model_loader (WIP)
wip : remove ggml_ctx from llama_model_loader
wip : merge gguf_file_loader in llama_model_loader
* llama : fix shape prints
* llama : fix Windows build + fix norm_rms_eps key
* llama : throw error on missing KV paris in model meta data
* llama : improve printing + log meta data
* llama : switch print order of meta data
---------
Co-authored-by: M. Yusuf Sarıgöz <yusufsarigoz@gmail.com>
2023-08-16 14:34:03 +03:00
klosax
ea5615a03a
convert-llama-h5-to-gguf.py : clarify the reverse permute
2023-08-16 11:23:15 +02:00
klosax
4a1741aa2d
gptneox-main.cpp : add tensor data layout
2023-08-15 19:56:19 +02:00
klosax
2ae0e985b3
convert-llama-7b-pth-to-gguf.py : add tensor data layout
2023-08-15 19:55:13 +02:00
klosax
66756c82af
convert-llama-h5-to-gguf.py : add tensor data layout
2023-08-15 19:54:33 +02:00
klosax
b6056c3db8
gguf.py : add tensor data layout
2023-08-15 19:53:44 +02:00
Georgi Gerganov
b5ffb2849d
scripts : add helper script to get wikitext
2023-08-15 10:05:25 +03:00
klosax
2dd5d2c92c
convert-llama-h5-to-gguf.py : add 70b gqa support
2023-08-15 00:43:10 +02:00
Jhen-Jie Hong
3ebb00935f
server : add missing /json-schema-to-grammar.mjs ( #2616 )
...
fixes #2611
2023-08-15 06:14:14 +08:00
klosax
ca4758290c
gguf-llama.cpp : fix n_head_kv
2023-08-14 23:18:41 +02:00
klosax
ab2cbd03ca
convert-llama-7b-pth-to-gguf.py : add token types
2023-08-14 22:10:50 +02:00
klosax
cedb4870c6
gguf.py : add token types
2023-08-14 22:08:40 +02:00
klosax
5d518d421f
constants.py : add token types
2023-08-14 22:07:53 +02:00
klosax
7ec125b1dc
convert-llama-h5-to-gguf.py : add token types
2023-08-14 22:06:33 +02:00
Georgi Gerganov
6c63550f63
llama : update tokenizer style
2023-08-14 22:11:57 +03:00
Georgi Gerganov
7494c78428
llama : sync gguf-llama with llama ( #2613 )
...
* llama : sync gguf-llama with llama
* tests : fix build + warnings (test-tokenizer-1 still fails)
* tests : fix wstring_convert
* convert : fix layer names
* llama : sync gguf-llama.cpp
* convert : update HF converter to new tokenizer voodoo magics
2023-08-14 21:33:33 +03:00
goerch
afc4ca2889
convert : update convert-new.py with tokenizer fixes ( #2614 )
...
* Merge tokenizer fixes into the gguf branch.
* Add test vocabularies
* Adapt convert-new.py (and fix a clang-cl compiler error on windows)
2023-08-14 20:20:04 +03:00
goerch
ec1b100720
llama : tokenizer fixes ( #2549 )
...
* Merge tokenizer fixes into the gguf branch.
* Add test vocabularies
2023-08-14 19:30:28 +03:00
Georgi Gerganov
8af3a99ff1
Merge branch 'master' into gguf
2023-08-14 16:39:18 +03:00
Georgi Gerganov
6f14854880
gitignore : add gptneox-main
2023-08-14 16:39:02 +03:00
Jhen-Jie Hong
d783f7982e
metal : return null instead of exit(1) ( #2573 )
2023-08-14 16:37:39 +03:00
Cheng Shao
d75561df20
server : add --numa support ( #2524 )
2023-08-14 16:36:42 +03:00
