aboutsummaryrefslogtreecommitdiff
path: root/ggml.c
AgeCommit message (Collapse)Author
2023-04-22ggml : fix AVX build + update to new Q8_0 formatGeorgi Gerganov
2023-04-22ggml : alternative Q4_3 implementation using modified Q8_0 (#1109)Georgi Gerganov
* ggml : prefer vzip to vuzp This way we always use the same type of instruction across all quantizations * ggml : alternative Q4_3 implementation using modified Q8_0 * ggml : fix Q4_3 scalar imlpementation * ggml : slight improvement of Q4_3 - no need for loop unrolling * ggml : fix AVX paths for Q8_0 quantization
2023-04-22ggml : AVX2 optimization for vec_dot_q4_3_q8_0 and refactoring (#1099)Stephan Walter
* AVX2 optimization for vec_dot_q4_3_q8_0 and refactoring * finish AVX vectorization of quantize_row_q8_0 * Rename hsum_int_8 to hsum_i32_8
2023-04-21Improve cuBLAS performance by using a memory pool (#1094)slaren
* Improve cuBLAS performance by using a memory pool * Move cuda specific definitions to ggml-cuda.h/cu * Add CXX flags to nvcc * Change memory pool synchronization mechanism to a spin lock General code cleanup
2023-04-21ggml : a faster version for Q4_1 x Q8_0 dot products (#1083)Kawrakow
* A faster version for Q4_1 x Q8_0 dot products The idea nehind being that Q8_0 quantized values get used many times in the matrix multiplications where they are involved. In the current implementations, when we are evaluating the dot products, we need to compute the sum of the quants in the Q8_0 vector, so the same operation is repeated many times. Here we pre-compute the sum during Q8_0 quantization, store it in the now modified block_q8_0 struct, and then reuse this result in the subsequent dot products. In a synthetic benchmark (just compute a bunch of dot products), this change speeds up the Q4_1 * Q8_0 dot product by 80%, making the performance identical to Q4_0 * Q8_0. In practical application, I see a ~15% gain in speed for token prediction on M2, and ~5% gain on Ryzen 7950X. The speed gain in the prompt evaluation is much bigger (around 50%). I have only done the change for the scalar version, ARM_NEON, and AVX2, so we still need an AVX implementation. * Cleaning up --------- Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
2023-04-20ggml : sync ggml (add GPT-NeoX RoPE implementation)Georgi Gerganov
2023-04-20ggml : fix bug in ggml_compute_forward_dup_f32()Georgi Gerganov
2023-04-20ggml : do not break cuBLAS build (Q4_3 is not yet implemented)Georgi Gerganov
2023-04-20ggml : fix Q4_3 quantizationGeorgi Gerganov
Broke it during conflict resolution in last PR
2023-04-20llama : multi-threaded quantization (#1075)Kawrakow
* Multi-threading quantization. Not much gain for simple quantizations, bit it will be important for quantizations that require more CPU cycles. * Multi-threading for quantize-stats It now does the job in ~14 seconds on my Mac for Q4_0, Q4_1 and Q4_2. Single-threaded it was taking more than 2 minutes after adding the more elaborate version of Q4_2. * Reviewer comments * Avoiding compiler confusion After changing chunk_size to const int as suggested by @ggerganov, clang and GCC starting to warn me that I don't need to capture it in the lambda. So, I removed it from the capture list. But that makes the MSVC build fail. So, making it a constexpr to make every compiler happy. * Still fighting with lambda captures in MSVC --------- Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com> Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
2023-04-20ggml : add Q4_3 quantization (#1082)Georgi Gerganov
2023-04-20AVX2 optimization for vec_dot_q4_2_q8_0 (#1068)Stephan Walter
2023-04-20Improve cuBLAS performance by dequantizing on the GPU (#1065)slaren
2023-04-19Q4_2 quantization with rmse-optimized scale and quants (#1062)Kawrakow
* Q4_2 quantization with rmse-optimized scale and quants For quantize-stats we get q4_2: rmse 0.00159301, maxerr 0.17480469, 95pct<0.0030, median<0.0012 For 7B perplexity with BLAS enabled we get 6.2038 after 655 chunks. Quantization is slow (~90 seconds on my Mac for 7B) as not multi-threaded as in PR #896. * ggml : satisfy the sanitizer builds Not sure why this makes them fail * Better follow ggml conventions for function names * Fixed type as per reviewer comment --------- Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com> Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
2023-04-19ggml : use 8-bit precision for Q4_1 intermediate results (#1047)Georgi Gerganov
* ggml : use 8-bit precision for Q4_1 intermediate results (ARM) * ggml : optimize ggml_vec_dot_q4_1_q8_0() via vmalq_n_f32 56 ms/token with Q4_1 ! * ggml : AVX2 implementation of ggml_vec_dot_q4_1_q8_0 (#1051) * gitignore : ignore ppl-*.txt files --------- Co-authored-by: slaren <2141330+slaren@users.noreply.github.com>
2023-04-19ggml : Q4 cleanup - remove 4-bit dot product code (#1061)Stephan Walter
* Q4 cleanup * Remove unused AVX512 Q4_0 code
2023-04-19Add NVIDIA cuBLAS support (#1044)slaren
2023-04-19Multi-threaded ggml_cpy (#1035)slaren
* Multi-threaded ggml_cpy * Update ggml.c Co-authored-by: Georgi Gerganov <ggerganov@gmail.com> * Also fix wdata offset in ggml_compute_forward_add_q_f32 --------- Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
2023-04-18ggml : add new Q4_2 quantization (ARM only) (#1046)Georgi Gerganov
* ggml : Q4_2 ARM * ggml : add ggml_is_quantized() * llama : update llama_type_name() with Q4_2 entry * ggml : speed-up q4_2 - 4 threads: ~100ms -> ~90ms - 8 threads: ~55ms -> ~50ms * ggml : optimize q4_2 using vmlaq_n_f32 + vmulq_n_f32
2023-04-18ggml : scratch that - vmlaq_n_f32 is always betterGeorgi Gerganov
Had a background process that was messing with the timings
2023-04-18ggml : optimize ggml_vec_dot_q4_0_q8_0() using vectorized accumulatorsGeorgi Gerganov
2023-04-17Add LoRA support (#820)slaren
2023-04-17ggml : avoid using ggml_fp16_to_fp32() and ggml_fp32_to_fp16() in ggml.cGeorgi Gerganov
2023-04-17Speedup the AVX-512 implementation of ggml_vec_dot_q4_0() (#933)Ivan Komarov
2023-04-15Fix potential int8 overflow in non-SIMD vec_dot (#986)Stephan Walter
2023-04-15Refactor ggml.c for future tensor types (#1001)Stephan Walter
2023-04-15ggml : add Q8_0 quantization for intermediate results (#951)Georgi Gerganov
* ggml : add Q8_0 quantization for intermediate results * quantize-stats : fix test + add it to Makefile default * Q8: use int8_t, AVX/AVX2 optimizations * ggml : fix quantize_row_q8_0() ARM_NEON rounding * minor : updates after rebase to latest master * quantize-stats : delete obsolete strings * ggml : fix q4_1 dot func --------- Co-authored-by: Stephan Walter <stephan@walter.name>
2023-04-15ggml : use posix_memalign on non-Windows envGeorgi Gerganov
2023-04-14Expose type name from ggml (#970)Pavol Rusnak
Avoid duplication of type names in utils Co-authored-by: HÃ¥kon H. Hitland <haakon@likedan.net>
2023-04-14ggml : add unary and binary map operations (#874)Kerfuffle
* GGML map ops proof of concept. * Various cleanups. Add handling for task setting. Add handling for ggml_compute_backward. Rename functions to ggml_map_unary_f32 and ggml_map_binary_f32 Fix compiler warnings related to casting function pointers and `void *` Reorder functions and definitions based on the GGML op number. Use typedefs for map op function pointer types. * Fix position of map ops cases in ggml_compute_forward
2023-04-14ggml : minorGeorgi Gerganov
2023-04-14ggml : always allocate buffers with size multiple of GGML_MEM_ALIGNGeorgi Gerganov
2023-04-14ggml : fix q4_1 dot product typesGeorgi Gerganov
2023-04-14ggml : optimize rope function to avoid call powf in the tight loop (#807)Howard Su
2023-04-13ggml : add GGML_DEFAULT_N_THREADSGeorgi Gerganov
2023-04-13ggml : speed-up ggml_vec_dot_q4_1() ARM_NEON + 32-bit ARM support (#900)Georgi Gerganov
* ggml : speed-up q4_1 ARM_NEON by ~5% * ggml : implement vaddvq when missing * ggml : implement vminvq and vmaxvq when missing * ggml : implement vzip when missing * ggml : fix comment * ggml : try to use correct ifdef
2023-04-13ggml : optimize non-SIMD Q4_0 vector dot product (#703)Stephan Walter
2023-04-13ggml : introduce GGML_ALIGNED_MALLOC/GGML_ALIGNED_FREE macros (#884)Pavol Rusnak
which allows us to use aligned_alloc or _aligned_malloc functions
2023-04-13ggml : update cblas_sgemm columns var to be more reasonable (#838)Vladimir
2023-04-11Fix whitespace, add .editorconfig, add GitHub workflow (#883)Pavol Rusnak
2023-04-11Add enum llama_ftype, sync ggml_type to model files (#709)Stephan Walter
2023-04-11Windows fixes (#890)comex
Mostly for msys2 and mingw64 builds, which are different from each other and different from standard Visual Studio builds. Isn't Windows fun? - Define _GNU_SOURCE in more files (it's already used in ggml.c for Linux's sake). - Don't use PrefetchVirtualMemory if not building for Windows 8 or later (mingw64 doesn't by default). But warn the user about this situation since it's probably not intended. - Check for NOMINMAX already being defined, which it is on mingw64. - Actually use the `increment` variable (bug in my `pizza` PR). - Suppress unused variable warnings in the fake pthread_create and pthread_join implementations for Windows. - (not Windows-related) Remove mention of `asprintf` from comment; `asprintf` is no longer used. Fixes #871.
2023-04-10ggml : fix WASM buildGeorgi Gerganov
2023-04-10ggml : add ggml_cont() + optimize ggml_cpy() for contiguous dstGeorgi Gerganov
2023-04-10ggml : remove trailing whitespacesGeorgi Gerganov
2023-04-10Simplify to include lower-case windows.h always, fix compile on mingw32 (#747)Marco Matthies
2023-04-10ggml : fix quantize_row_q4_1() ARM_NEON (close #876)Georgi Gerganov
2023-04-10Rewrite loading code to try to satisfy everyone:comex
- Support all three formats (ggml, ggmf, ggjt). (However, I didn't include the hack needed to support GPT4All files without conversion. Those can still be used after converting them with convert.py from my other PR.) - Support both mmap and read (mmap is used by default, but can be disabled with `--no-mmap`, and is automatically disabled for pre-ggjt files or on platforms where mmap is not supported). - Support multi-file models like before, but automatically determine the number of parts rather than requiring `--n_parts`. - Improve validation and error checking. - Stop using the per-file type field (f16) entirely in favor of just relying on the per-tensor type/size fields. This has no immediate benefit, but makes it easier to experiment with different formats, and should make it easier to support the new GPTQ-for-LLaMa models in the future (I have some work in progress on that front). - Support VirtualLock on Windows (using the same `--mlock` option as on Unix). - Indicate loading progress when using mmap + mlock. (Which led me to the interesting observation that on my Linux machine, with a warm file cache, mlock actually takes some time, whereas mmap without mlock starts almost instantly...) - To help implement this, move mlock support from ggml to the loading code. - madvise/PrefetchVirtualMemory support (based on #740) - Switch from ifstream to the `fopen` family of functions to avoid unnecessary copying and, when mmap is enabled, allow reusing the same file descriptor for both metadata reads and mmap (whereas the existing implementation opens the file a second time to mmap). - Quantization now produces a single-file output even with multi-file inputs (not really a feature as much as 'it was easier this way'). Implementation notes: I tried to factor the code into more discrete pieces than before. Regarding code style: I tried to follow the code style, but I'm naughty and used a few advanced C++ features repeatedly: - Destructors to make it easier to ensure everything gets cleaned up. - Exceptions. I don't even usually use exceptions when writing C++, and I can remove them if desired... but here they make the loading code much more succinct while still properly handling a variety of errors, ranging from API calls failing to integer overflow and allocation failure. The exceptions are converted to error codes at the API boundary.) Co-authored-by: Pavol Rusnak <pavol@rusnak.io> (for the bit I copied from #740)
2023-04-08Add quantize-stats command for testing quantization (#728)unbounded
Command that calculates some statistics over the errors introduced by quantization, like mean square error, max error and some percentile errors for layer weights. Should be useful for testing quantization improvements. Exposes some internal state from ggml and llama for testing
2023-04-05ggml : multi-thread ggml_rope() (~3-4 times faster on M1) (#781)Georgi Gerganov