Commit Graph

1614 Commits

Author SHA1 Message Date
cebtenzzre
0eb332a10f
llama : fix llama_context_default_params after #2268 (#3893) 2023-11-01 19:29:14 -04:00
slaren
d02e98cde0
ggml-cuda : compute ptrs for cublasGemmBatchedEx in a kernel (#3891)
* ggml-cuda : compute ptrs for cublasGemmBatchedEx in a kernel

* fix warnings
2023-11-01 23:10:09 +01:00
cebtenzzre
898aeca90a
llama : implement YaRN RoPE scaling (#2268)
Co-authored-by: cebtenzzre <cebtenzzre@gmail.com>
Co-authored-by: Jeffrey Quesnelle <jquesnelle@gmail.com>
2023-11-01 18:04:33 -04:00
Georgi Gerganov
c43c2da8af
llm : fix llm_build_kqv taking unused tensor (benign, #3837) 2023-11-01 23:08:30 +02:00
Georgi Gerganov
523e49b111
llm : fix falcon norm after refactoring (#3837) 2023-11-01 23:00:50 +02:00
Georgi Gerganov
e16b9fa4ba
metal : multi-simd softmax (#3710)
ggml-ci
2023-11-01 21:25:00 +02:00
Georgi Gerganov
ff8f9a88da
common : minor (#3715) 2023-11-01 21:15:55 +02:00
Georgi Gerganov
50337961a6
llm : add llm_build_context (#3881)
* llm : add llm_build_context

* llm : deduce norm eps based on type + explict max_alibi_bias, clamp_kqv

* llm : restore the non-graph llm_build_ functional API

ggml-ci

* llm : cleanup + comments
2023-11-01 20:11:02 +02:00
bandoti
0e40806c1c
common : allow caller to handle help/argument exceptions (#3715)
* Allow caller to handle help/argument exceptions

* Prepend newline to usage output

* Add new gpt_params_parse_ex function to hide arg-parse impl

* Fix issue blocking success case

* exit instead of returning false

* Update common/common.h

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>

* Update common/common.cpp

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
2023-11-01 19:42:01 +02:00
staviq
a2758d08e4
log : make generating separate log files optional (#3787)
* impl --log-new, --log-append

* Update common/log.h

Co-authored-by: cebtenzzre <cebtenzzre@gmail.com>

* Update common/log.h

Co-authored-by: cebtenzzre <cebtenzzre@gmail.com>

* Apply suggestions from code review

Co-authored-by: cebtenzzre <cebtenzzre@gmail.com>

---------

Co-authored-by: cebtenzzre <cebtenzzre@gmail.com>
2023-11-01 16:18:27 +02:00
l3utterfly
e75dfdd31b
sampling : null grammar field after reset (#3885) 2023-11-01 15:40:43 +02:00
Georgi Gerganov
9a3b4f6c86
ggml : fix UNUSED macro (#3762) 2023-11-01 13:50:45 +02:00
Andrew Godfrey
73bdcb395e
finetune : add -ngl parameter (#3762)
* Add '-ngl' support to finetune.cpp

* Add fprintf in ggml_cuda_op_add

When I tried CUDA offloading during finetuning following the readme, I got an assert here.
This probably isn't an important case because inference later gives a warning saying you should use f16 or f32 instead when using lora

* Add 'finetune.sh', which currently fails when using GPU

"error: operator (): Finetuning on tensors with type 'f16' is not yet supported"

* tweak finetune.sh

* Suppress some warnings in ggml.c

* Add f16 implementation to ggml_compute_forward_add_f16_f32

* Add an f16 case to ggml_add_cast_impl and llama_build_lora_finetune_graphs

* finetune.sh: Edit comments

* Add "add_f16_f32_f32_cuda"

* Tweak an error message

* finetune.sh: Add an optional LLAMA_MODEL_DIR variable

* finetune.sh: Add an optional LLAMA_TRAINING_DIR variable

* train : minor

* tabs to spaces

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
Co-authored-by: cebtenzzre <cebtenzzre@gmail.com>
2023-11-01 13:49:04 +02:00
Georgi Gerganov
f0e209324a
scripts : add server-llm.sh (#3868)
* scripts : add deploy-server.sh

* scripts : rename to server-llm.sh

* scripts : working curl pipe
2023-11-01 11:29:07 +02:00
Adrian Hesketh
ca190bca8e
server : re-enable completion and embedded at the same time (#3876) 2023-11-01 11:28:28 +02:00
Georgi Gerganov
71e3718abd
llama : refactor graph build code (#3837)
* llama : factor out ggml-alloc from graph graph build functions

ggml-ci

* metal : disable kernel load log

* llama : factor out tensor offloading outside the build call (wip)

ggml-ci

* llama : offload rest of the models

ggml-ci

* llama : update offload log messages to print node index

* llama : comments

* llama : support offloading result_norm + comments

* llama : factor graph input into a function

* llama : do tensor offload only with CUDA

* llama : fix res_norm offloading

* llama : try to optimize offloading code

* llama : fix non-CUDA build

* llama : try to fix build

* llama : move refact in correct place + optimize graph input

* llama : refactor tensor offloading as callback

* llama : add layer index to all tensor names

* llama : add functional header

* llama : comment

ggml-ci

* llama : remove obsolete map for layer counting

* llama : add llm_build helper functions (#3848)

* llama : add llm_build_norm helper function

ggml-ci

* llama : add llm_build_ffn helper function (#3849)

ggml-ci

* llama : add llm_build_k_shift helper

ggml-ci

* llama : fix offloading after recent changes

* llama : add llm_build_kv_store helper

ggml-ci

* llama : remove obsolete offload names

* llama : fix llm_build_k_shift to use n_head_kv instead of n_head

* llama : simplify falcon Q, K, V computation

* llama : remove obsolete comments in build graphs

* llama : add llm_build_kqv helper

ggml-ci

* llama : minor

* llama : add LLAMA_OFFLOAD_DEBUG + fix starcoder offloading

* llama : fix input allocation logic

* llama : update offload functions for KQ tensors

* llama : normalize tensor names

ggml-ci

* llama : enable warning about not offloaded tensors

* llama : remove extra ; + deduplicate gate_b logic

* llama : add llm_build_inp_embd helper
2023-11-01 08:04:02 +02:00
kalomaze
238657db23
samplers : Min-P sampler implementation [alternative to Top P/Top K] (#3841)
* Introduce the new Min-P sampler by @kalomaze
   The Min-P sampling method was designed as an alternative to Top-P, and aims to ensure a balance of quality and variety. The parameter *p* represents the minimum probability for a token to be considered, relative to the probability of the most likely token.

* Min-P enabled and set to 0.05 default

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
Co-authored-by: cebtenzzre <cebtenzzre@gmail.com>
2023-10-31 20:44:49 +01:00
Tungsten842
07178c98e1
flake.nix: fix for rocm 5.7 (#3853) 2023-10-31 19:24:03 +02:00
Georgi Gerganov
207b51900e
ggml : move FP16 <-> FP32 code to ggml-impl.h (#3861)
* ggml : move FP16 <-> FP32 stuff to ggml-impl.h

ggml-ci

* tests : fix ARM build

* ggml : explicitly initialize deprecated type traits

* ggml : add math.h to ggml-impl.h

* ggml : remove duplicate static assert macros

* ggml : prefix lookup tables with ggml_

ggml-ci

* ggml-impl : move extern "C" to start of file
2023-10-30 19:19:15 +02:00
Kerfuffle
6e08281e58
Extend llama_kv_cache_seq_rm to allow matching any sequence (#3843)
* Extend llama_kv_cache_seq_rm to allow matichng any sequence

* Replace llama_kv_cache_tokens_rm with llama_kv_cache_clear

Use llama_kv_cache_clear for cache clearing

Change calls to llama_kv_cache_tokens_rm that want to delete by position to use llama_kv_cache_seq_rm functionality
2023-10-29 11:31:40 -06:00
cebtenzzre
2046eb4345
make : remove unnecessary dependency on build-info.h (#3842) 2023-10-29 18:33:47 +02:00
Georgi Gerganov
71a09da301
llama : fix kv shift bug (#3835)
ggml-ci
2023-10-29 18:32:51 +02:00
Georgi Gerganov
d69d777c02
ggml : quantization refactoring (#3833)
* ggml : factor all quantization code in ggml-quants

ggml-ci

* ggml-quants : fix Zig and Swift builds + quantize tool

ggml-ci

* quantize : --pure option for disabling k-quant mixtures

---------

Co-authored-by: cebtenzzre <cebtenzzre@gmail.com>
2023-10-29 18:32:28 +02:00
Erik Scholz
ff3bad83e2
flake : update flake.lock for newer transformers version + provide extra dev shell (#3797)
* flake : update flake.lock for newer transformers version + provide extra dev shell with torch and transformers (for most convert-xxx.py scripts)
2023-10-28 16:41:07 +02:00
Aarni Koskela
82a6646e02
metal : try cwd for ggml-metal.metal if bundle lookup fails (#3793)
* Try cwd for ggml-metal if bundle lookup fails

When building with `-DBUILD_SHARED_LIBS=ON -DLLAMA_METAL=ON -DLLAMA_BUILD_SERVER=ON`,
`server` would fail to load `ggml-metal.metal` because `[bundle pathForResource:...]`
returns `nil`.  In that case, fall back to `ggml-metal.metal` in the cwd instead of
passing `null` as a path.

Follows up on #1782

* Update ggml-metal.m

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
2023-10-28 15:43:01 +03:00
Georgi Gerganov
ba231e8a6d
issues : change label from bug to bug-unconfirmed (#3748) 2023-10-28 15:35:26 +03:00
Georgi Gerganov
8a2f2fea29
convert : ignore tokens if their IDs are within [0, vocab_size) (#3831) 2023-10-28 06:25:15 -06:00
Kerfuffle
bd6d9e2059
llama : allow quantizing k-quants to fall back when tensor size incompatible (#3747)
* Allow quantizing k-quants to fall back when tensor size incompatible

* quantizing: Add warning when tensors were incompatible with k-quants

Clean up k-quants state passing a bit
2023-10-28 14:54:24 +03:00
Georgi Gerganov
ee1a0ec9cb
llama : add option for greedy sampling with probs (#3813)
* llama : add option for greedy sampling with probs

* llama : add comment about llama_sample_token_greedy() missing probs

* sampling : temp == 0.0 -> no probs, temp < 0.0 -> probs
2023-10-28 14:23:11 +03:00
Henk Poley
177461104b
common : print that one line of the syntax help *also* to standard output (#3823) 2023-10-28 13:16:33 +03:00
Georgi Gerganov
fdee152e4e
starcoder : add GPU offloading (#3827)
* starcoder : do not GPU split 1D bias tensors

* starcoder : offload layers to GPU

ggml-ci
2023-10-28 12:06:08 +03:00
Kerfuffle
41aee4df82
speculative : ensure draft and target model vocab matches (#3812)
* speculative: Ensure draft and target model vocab matches

* Tolerate small differences when checking dft vs tgt vocab
2023-10-28 00:40:07 +03:00
cebtenzzre
6d459cbfbe
llama : correctly report GGUFv3 format (#3818) 2023-10-27 17:33:53 -04:00
Thibault Terrasson
c8d6a1f34a
simple : fix batch handling (#3803) 2023-10-27 08:37:41 -06:00
Georgi Gerganov
2f9ec7e271
cuda : improve text-generation and batched decoding performance (#3776)
* cuda : prints wip

* cuda : new cublas gemm branch for multi-batch quantized src0

* cuda : add F32 sgemm branch

* cuda : fine-tune >= VOLTA params + use MMQ only for small batches

* cuda : remove duplicated cuBLAS GEMM code

* cuda : add CUDA_USE_TENSOR_CORES and GGML_CUDA_FORCE_MMQ macros

* build : add compile option to force use of MMQ kernels
2023-10-27 17:01:23 +03:00
Georgi Gerganov
34b2a5e1ee
server : do not release slot on image input (#3798) 2023-10-26 22:54:17 +03:00
Georgi Gerganov
6961c4bd0b
batched-bench : print params at start 2023-10-25 10:26:27 +03:00
Georgi Gerganov
cc44877486
log : disable pid in log filenames 2023-10-25 10:09:16 +03:00
cebtenzzre
ad93962657
server : add parameter -tb N, --threads-batch N (#3584) (#3768)
Co-authored-by: Michael Coppola <m18coppola@gmail.com>
Co-authored-by: Michael Coppola <info@michaeljcoppola.com>
2023-10-24 23:10:43 +03:00
Georgi Gerganov
1717521cdb
server : do not block system prompt update (#3767)
* server : do not block system prompt update

* server : update state machine logic to process system prompts

* server : minor
2023-10-24 23:08:20 +03:00
Georgi Gerganov
b2f7e04bd3
sync : ggml (conv ops + cuda MSVC fixes) (#3765)
ggml-ci
2023-10-24 21:51:20 +03:00
John Smith
abd21fc99f
cmake : add missed dependencies (#3763) 2023-10-24 20:48:45 +03:00
Georgi Gerganov
2b4ea35e56
cuda : add batched cuBLAS GEMM for faster attention (#3749)
* cmake : add helper for faster CUDA builds

* batched : add NGL arg

* ggml : skip nops in compute_forward

* cuda : minor indentation

* cuda : batched cuBLAS GEMMs for src0 F16 and src1 F32 (attention ops)

* Apply suggestions from code review

These changes plus:

```c++
#define cublasGemmBatchedEx hipblasGemmBatchedEx
```

are needed to compile with ROCM. I haven't done performance testing, but it seems to work.

I couldn't figure out how to propose a change for lines outside what the pull changed, also this is the first time trying to create a multi-part review so please forgive me if I mess something up.

* cuda : add ROCm / hipBLAS cublasGemmBatchedEx define

* cuda : add cublasGemmStridedBatchedEx for non-broadcasted cases

* cuda : reduce mallocs in cublasGemmBatchedEx branch

* cuda : add TODO for calling cublas from kernel + using mem pool

---------

Co-authored-by: Kerfuffle <44031344+KerfuffleV2@users.noreply.github.com>
2023-10-24 16:48:37 +03:00
Galunid
daab3d7f45
Add more tokenizer tests (#3742)
* Add more tokenizer tests

* Add starcoder

* Update test vocab files

* Restrict bpe tokenizer tests to unicode planes

* Update comment

* Comment cosmetics

* Remove bloom vocab/test
2023-10-24 09:17:17 +02:00
Georgi Gerganov
469c9addef
metal : handle ggml_scale for n%4 != 0 (close #3754)
ggml-ci
2023-10-24 09:47:22 +03:00
Georgi Gerganov
e3932593d4
Revert "make : add optional CUDA_NATIVE_ARCH (#2482)"
This reverts commit 96981f37b1.

See:

https://github.com/ggerganov/llama.cpp/pull/2482#issuecomment-1775975866
2023-10-23 23:46:05 +03:00
M. Yusuf Sarıgöz
9d02956443
issues : separate bug and enhancement template + no default title (#3748) 2023-10-23 22:57:16 +03:00
Galunid
69a6735087
Update special token handling in conversion scripts for gpt2 derived tokenizers (#3746)
We still have the heads up in `README.md` regarding `bpe` tokenizers and this patch is needed for 

- a couple of tokenizer tests
- some more `special` and `non-special` added tokens handling (as far as I understand it)

* Update special token handling

* Add mpt
2023-10-23 21:46:00 +02:00
Marcus Dunn
5be6c803fa
llama : remove token functions with context args in favor of model (#3720)
* added `llama_model_token_*` variants to all the `llama_token_*` functions.

* added `LLAMA_API`

* formatting

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>

* removed old `llama_token` functions

* changed 3 more functions to take in model

- `llama_token_get_text`
- `llama_token_get_score`
- `llama_token_get_type`

* added back docs

* fixed main.cpp

* changed token functions to use new model variants

* changed token functions to use new model variants

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
2023-10-23 22:40:03 +03:00
Galunid
6336701c93
Fix baichuan convert script not detecing model (#3739)
It seems nobody objects.
2023-10-23 17:47:03 +02:00