| Age | Commit message (Collapse) | Author |
|---|
|
K*Q and KQ*V are n_kv_embed x n_token x n_head matrix multiplications.
Before this PR, this meant n_head calls to iqk_mul_mat to perform
n_kv_embed x n_token 2D multiplications, each using nth threads.
Instead, in this PR, if n_head is a multiple of nth, each thread
does n_head/nth multiplications of the n_kv_embed x n_token 2D matrices.
This improves PP-512(32 threads) for Bitnet-3B to 433 t/s up from
409 t/s. It is beneficial in other cases too. E.g., for LLaMA-7B,
we go to 201 t/s up from 193 t/s for q4_K_S, and to 144 t/s up from
139 t/s for fp16. All these numbers are for the Ryzen-7950X CPU.
|
|
|
|
Also save one scale operation in the ffn network by adjusting
rms_eps. We gain up to 3% in performance by doing this, but it
is a bit of a hack (we store the tensor scales in op_params
while loading the model).
|
|
|
|
The scalar dot product already chieves 37 t/s for TG!
|
|
|
|
* Remove iqk_mul_mat from llamafile_sgemm
* Pass tensor types and strides to iqk_mul_mat
It is marked WIP because only tested on __aarch64__
|
|
|
|
|
|
|
|
|
|
|
|
|
|
* move BLAS to a separate backend
* rename GGML_USE_OPENBLAS to GGML_USE_BLAS
* alloc : reuse same buffer when the same buffer type if used multiple times
* set number of threads automatically for openblas and blis
* sched : print assignments when GGML_SCHED_DEBUG env variable is set
* sched : allow ops with weights on an incompatible buffer type
This will cause the weight to be copied to a backend that supports the
op, which is very costly. The weight should have been stored in a buffer
of a backend that can run the op, but llama.cpp cannot do this
automatically at the moment.
---------
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
|
|
* tests : add non-cont unary tests
* ggml : update unary asserts and "supports_op"
ggml-ci
|
|
* ggml : improve ggml_is_contiguous logic
ggml-ci
* ggml : support more contiguous cases
ggml-ci
|
|
* ggml : unify rope norm/neox (CPU)
* ggml : fix compile warning
* ggml : remove GLM rope mode
ggml-ci
* metal : better rope implementation
ggml-ci
* cuda : better rope implementation
ggml-ci
* naming : n_orig_ctx -> n_ctx_orig
ggml-ci
* dev : add reminders to update backends
ggml-ci
* vulkan : fix ggml_rope_ext() usage
* cuda : fix array size + indents
ggml-ci
|
|
ggml-ci
|
|
This enforces a check that -fno-finite-math-only was set and that the operating
compiling mode is not in finite maths mode. This is because during rewriting of
silu and softmax for cpu #7154 there emerged an issue where the result that was
observed when >1 slot was nondeterministic as found by @JohannesGaessler.
@LostRuins narrowed the problem down to -ffinite-math-only which was theorised
to be due to SiLU, instead of flushing small values to 0, returns NaN or some
other garbage. @jart proposed a fix that @ggerganov then implemented in this fix
ref https://github.com/ggerganov/llama.cpp/pull/7154#issuecomment-2145661825
|
|
* ggml: Added OpenMP for multi-threads processing
* ggml : Limit the number of threads used to avoid deadlock
* update shared state n_threads in parallel region
* clear numa affinity for main thread even with openmp
* enable openmp by default
* fix msvc build
* disable openmp on macos
* ci : disable openmp with thread sanitizer
* Update ggml.c
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
---------
Co-authored-by: slaren <slarengh@gmail.com>
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
|
|
* ggml : fix loongson compile warnings
ggml-ci
* Fix loongarch quantize test fail.
Fix unexpected error introduced during rebase code.
* tests : disable json test due to lack of python on the CI node
ggml-ci
---------
Co-authored-by: junchao-loongson <zhaojunchao@loongson.cn>
|
|
* faster avx512 exp implementation
* x->r
* improve accuracy, handle special cases
* remove `e`
|
|
|
|
* tests : add rope tests
ggml-ci
* ggml : fixes (hopefully)
ggml-ci
* tests : add non-cont tests
ggml-ci
* cuda : add asserts for rope/norm + fix DS2
ggml-ci
* ggml : assert contiguousness
* tests : reduce RoPE tests
ggml-ci
|
|
|
|
* ggml : use atomic_flag for critical section
* add windows shims
|
|
ggml-ci
|
|
|
|
* ggml : generalize GGML_OP_CONCAT (WIP)
ggml-ci
* tests : add dim != 2 tests
* metal : generalize concat kernel
* tests : naming
* cuda : generalize concat kernel
ggml-ci
* sycl : add warning and assert
* ggml : fix op params handling
* metal : bugfix kernel
ggml-ci
* ggml : reimplement CPU and Metal
* cuda : add asserts
ggml-ci
* ggml : fix ptrs
ggml-ci
|
|
* Add SVE support for q4_0_q8_0 q8_0_q8_0
* remove ifdef
|
|
ggml-ci
|
|
* ggml : drop support for QK_K=64
ggml-ci
* opencl : restore QK_K=256 define
|
|
* cuda : fix rope pos data
ggml-ci
* ggml : drop mode & 1 == 1 support for ggml_rope
ggml-ci
* ggml : support freq_factors for f16 rope (CPU)
ggml-ci
* tests : add rope tests using frequency factors
ggml-ci
|
|
* add phi3 128k support in convert-hf-to-gguf
* add phi3 128k support in cuda
* address build warnings on llama.cpp
* adjust index value in cuda long rope freq factors
* add long rope support in ggml cpu backend
* make freq factors only depend on ctx size
* remove unused rope scaling type 'su' frin gguf converter
* fix flint warnings on convert-hf-to-gguf.py
* set to the short freq factor when context size is small than trained context size
* add one line of comments
* metal : support rope freq_factors
* ggml : update ggml_rope_ext API to support freq. factors
* backends : add dev messages to support rope freq. factors
* minor : style
* tests : update to use new rope API
* backends : fix pragma semicolons
* minor : cleanup
* llama : move rope factors from KV header to tensors
* llama : remove tmp assert
* cuda : fix compile warning
* convert : read/write n_head_kv
* llama : fix uninitialized tensors
---------
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
|
|
* add loongarch lsx and lasx optimize code
* Add loongarch compilation support to makefile
* revert stb_image.h
* opt bytes_from_nibbles_32 and sum_i16_pairs_float
* fix undeclared
* format code
* update
* update 2
---------
Co-authored-by: Jinyang He <hejinyang@loongson.cn>
|
|
for enabling AVX512_BF16 (#7258)
|
|
|
|
* android : use "ci-android" branch for CI
* ggml : disable SIMD exp and silu for 32-bit ARM
ggml-ci
* android : do not fetch, use add_subdirectory instead
* cmake : provide binary dir
|
|
This change upstreams llamafile's vectorized expf() functions. This lets
us compute softmax and silu more accurately than the short[65536] lookup
table that GGML previously used to make this operation go faster. We can
support aarch64 and sse2+ with the worst case rounding error of 2ulp. It
makes make -j8 tests && ./tests/test-backend-ops -o SOFT_MAX -b CPU perf
go 1.5x faster for SSE2+FMA, 1.9x faster for AVX2+FMA and 2.1x on AVX512
|
|
* Just reordering some structs.
* Adding in the calls to mm_pause
* Passing around the state
* Renaming and moving a bunch of variables around.
* Extracting the logic to it's own function.
* Moving some variable definitions into the chunk function.
* Moving some variables around
* moving src1_cont inside
* Moving row_size
* adding the current_chunk
* Reorg the code.
* Formatting to match the orig patch
* starting to setup the chunking variables
* Starting the buildup of the loop
* The yield shouldn't be necessary.
* adding the looping structure based on the chunk configuration.
* Add in the re-chunking code.
* Making it much more likely to rechunk.
* disable resizing if numa is enabled.
* Updating comments with what we've learned.
* Fix formatting
* Couple more formatting fixes.
* More style fixes.
* Fix Warnings
* Going with unused because there's conditional logic that needs it.
* Update ggml.c
* Update ggml.c
---------
|
|
|
|
* initial commit with CPU implementation of upscale to shape and test, cuda implementation next
* experimental commit to see if dst shape is correct
* test version
* test
* removed unnecessary params
* refactor
* fixed tests
* ggml : metal impl + cleanup + sycl dev warnings
* patched ggml_upscale cuda op to handle non-contiguous tensors, added test for non-contiguous behavior
* metal : fix upsacle op to support nb00 + style
---------
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
|
|
* ggml : fa without mask + add asserts
ggml-ci
* metal : support non-contiguous KV
ggml-ci
|
|
|
|
ggml-ci
|
|
* added sigmoid function
* implemented metal kernel for sigmoid
* implemented cuda kernel for sigmoid
* added sigmoid unary op and incremented count
|
|
* ggml : full ALiBi support
* ggml : update ggml_soft_max_ext() CUDA, SYCL
* ggml : ggml_flash_attn_ext() support ALiBi (CPU)
* ggml : ggml_flash_attn_ext() support ALiBi (Metal)
* ggml : fix warning
* ggml : ggml_flash_attn_ext() support ALiBi (CUDA)
ggml-ci
* ggml : fix assert message
* vulkan : add dev notes
* ggml : require mask when using ALiBi
ggml-ci
* convert : fix convert for refact models
|
|
* Introduce bfloat16 support
Many models on Hugging Face (e.g. Mistral, TinyLLaMA) use bfloat16 as
their canonical floating point format.
┌sign
│
│ ┌exponent
│ │
│ │ ┌mantissa
│ │ │
│┌──┴───┐┌─┴───┐
0b0000000000000000 brain16
This encoding has the same number of exponent bits as float32. That
makes conversion relatively straightforward, even in the absence of
hardware support. For example, converting brain16 to binary32 means
simply shifting 16 bits to the left.
┌sign
│
│ ┌exponent
│ │
│ │ ┌mantissa
│ │ │
│┌──┴───┐┌─┴───────────────────┐
0b00000000000000000000000000000000 IEEE binary32
The issue is that converting bf16 to fp16 can result in information
loss. Only 13% of bf16 numbers can be precisely represented in fp16
which in practice ends up being 99.71% of Mistral 7b v0.2's weights
however there is currently no way other than fp32 to get the others
┌sign
│
│ ┌exponent
│ │
│ │ ┌mantissa
│ │ │
│┌─┴─┐┌─┴──────┐
0b0000000000000000 IEEE binary16
This change fixes that, by adding a bf16 data type to GGML. Support
for CPU inference has been implemented along with optimizations for
the AVX2, AVX512, and AVX512BF16 ISAs. Perplexity on Mistral 7b 0.2
improves somewhere around -0.0024 to -0.0046 compared to using fp16
* Remove GGML code that's not needed
* Minimize the GGML API surface area for BF16
* Remove bf16 luts
* Make the GGML header look nicer
* Fix documentation
* Apply ggerganov's fixes for test-backend-ops
* Add BF16 code for new ggml_validate_row_data() function
|
|
|
|
* ggml : add ggml_flash_attn_ext API
* ggml : fix GQA support in ggml_flash_attn_ext
* ggml : online attention (CPU)
* metal : initial implementation
* metal : f16 precision
* metal : reduce branches
* metal : specialize for head size
* wip : 8 rows per simd group
* wip : 4 rows per simd group
* wip : template for rows per warp
* metal : parallelize across KV size
* metal : parallel reduce across heads
* metal : efficient flash_attn_f16 implementation
* metal : avoid redundant loads of the attention
* metal : scale and mask in matrix form
* metal : fix comment
* llama : avoid ggml_cast, use F32 query
* metal : add parallel reduce version (disabled)
* metal : move output into local memory + optimize
- the result from each simdgroup now stays in the registers
- significantly reduced SRAM usage
- more efficient skipping of -INF blocks
- avoid simdgroup barrier in hot loop
- add comments
* metal : add tests, fix scaling, support C > 32
* metal : improve precision
* ggml : fix f16 mad
* metal : minor
* metal : support Q > 8
* tests : add ATTN tests
* metal : disable buffer allocation logs
* tests : more
* metal : faster inner loop for C == 32
* metal : fix array initialization
* tests : ifdef
* ggml : switch to padded F16 mask for ggml_soft_max, ggml_flash_attn_ext
* ggml : fix ggml_soft_max mask requirement
* cuda : fix soft_max to use correct mask size
* cuda : add flash_attn kernel (wip)
* metal : optimize softmax for C > 32
* metal : optimize softmax
* tests : minor fix
* cuda : avoid zeroing fragments
* tests : update dims
* cuda : fix __hisinf() result check
* cuda : avoid warp_reduce for smax
* cuda : use int instead of int64_t
Noticeably improves performance (thanks to Johannes)
* cuda : make loops use the same loop values
Thanks Johannes again for the tip
* cuda : unroll some of the loops
* cuda : avoid __hisinf branches
* cuda : use half2 in softmax
* cuda : switch to 1 warp for bs > 16
* cuda : speed-up reduce part of the kernel
* cuda : unroll Q*K^T loop
* cuda : fix -INF block check
* cuda : simplify softmax
* cuda : fix matrix names
* cuda : minor
* llama : adapt to F16 KQ_pos
* llama : adapt new models to F16 KQ_mask
* ggml : fix F16 store (ARM NEON)
* llama : fix type of KQ_mask and KQ_pos
* ggml : fix CPU soft_max
* tests : add hs=256
* cuda : fix build
* metal : improve perf via smaller int registers
* cuda : adapt soft_max to F16 mask and pos
* CUDA: faster FlashAttention, kernel for bs == 1
* 16 cols for Phi-2
* no vec for hs, no hs==256 ncols==32 for Volta
* adjust kernel selection logic
* 4 warps, 256 stride for all D
* no ncols == 64
* Multiple parallel blocks for batch size 1
* fix compile warnings
* fix excessive KQ_b loads
* fix cmake build
* fix KV cache padding, NaN from INFINITY (#6438)
* llama : flash_attn cparam + fix defrag
* server: support flash_attn param
* server: bench: enable flash_attn param
* CUDA: refactor host code, dyn. par. blocks
* fix flash_attn_vec_f16 race condition
* flush softmax exp below threshold to 0
* store temp KQ in registers
* Calculate KQ as FP32 if KQV has GGML_PREC_F32
* Add __hgt2_mask implementation for CUDA 11
* fix KQ FP32 precision fpr parallel_blocks > 1
* llama-bench : add -fa,--flash-attn arg
* metal : add BS=1 kernel for flash attention (#6508)
* metal : add BS=1 kernel for flash attention (wip)
* metal : support more than 1 warps
* metal : opts
* metal : opt
* metal : switch to parallel reduce
* metal : reduce registers
* metal : simplify
* metal : initial FA vec kernel
* metal : use F32 attention accumulators
* batched-bench : add fattn arg
* llama : simplify llama_build_kv_store
ggml-ci
* llama : adapt build_olmo to changes
* ggml : fix arm fp16 store on windows
* metal : clean-up
* metal : clean-up kernel code
* metal : minor
* tests : remove benchmarks
ggml-ci
* ggml : fix avx512 const correctness
ggml-ci
* ggml : fix soft_max with bias on CPU
ggml-ci
* common : print --flash-attn in help
* ggml : fix num dimensions in ggml_flash_attn_ext
* llama : force disable flash attention for incompatible models
* ggml : ggml_soft_max support F16/F32 mask/pos
ggml-ci
* cuda : uint -> uint32_t
* cuda : "constexpr dim3" -> "const dim3"
ggml-ci
* cuda : try to fix __hgt2_mask
ggml-ci
* ggml : add TODO's for F16/F32 mask/pos support in other backends
* llama : replace bool need_kq_pos with use_alibi
* llama : prep ALiBi support for BERT models
ggml-ci
* llama : fix n_batch requirements
ggml-ci
* cont
* server : add help for --flash-attn arg
* llama : disable FA for AMD
* tests : remove TMP_ATTN_BENCH
ggml-ci
* llama : support save/load state with FA enabled
ggml-ci
* ci : add CUDA save-load-state tests
ggml-ci
* llama : llama_kv_cache_clear zeroes data + fix save-load seq
ggml-ci
* llama : fix copy-paste errors, add TODO
* llama : disallow incompatible states
* llama : update llama_state_get_size after v_trans field
* metal : remove tmp log
* llama : add static reminder for llama_state_get_size
* metal : fix max nsg
ggml-ci
* ci : fix arg order
ggml-ci
---------
Co-authored-by: Johannes Gäßler <johannesg@5d6.de>
Co-authored-by: Pierrick HYMBERT <pierrick.hymbert@gmail.com>
|
