* Arm AArch64: Documentation updates
* Update docs/build.md to include information on how to enable the Arm optimized gemm/gemv kernels
* Update examples/quantize/README.md with information on the Q4_0_4_4, Q4_0_4_8 and Q4_0_8_8 formats
* Add newline to the end of docs/build.md
* ggml-quants : 1.625 bpw ternary packing for BitNet 1.58b
* ggml-quants : faster 1.625 bpw AVX2 vec_dot
Not using a lookup table anymore makes it match q4_0 speed.
* gguf-py : fix formatting
* llama : remove spaces on empty line
* ggml-quants : subtract 1 when back in epi8
This makes the 1.625 bpw type go faster than q4_0. Still not the fastest.
* ggml-quants : Q2_2 now faster than Q4_K on with AVX2
* ggml-quants : cleanup Q1_3 code formatting
* ggml-quants : ARM NEON vec_dot for q2_2 and q1_3
* ggml-quants : use ceiling division when quantizing q1_3
* convert-hf : simplify BitNet pre-quantization
This still results in the exact same tensor weights and scales,
but it reveals some weirdness in the current algorithm.
* convert-hf : allow converting the weird BitNet 1.3B
Its FFN size is 5460 which is not convenient.
The offending tensors are kept in F16,
which makes the final model 5.01 bpw.
* bitnet : replace 1.58b with b1.58, as in the paper
* ggml-quants : fix build failure on Windows
* ggml-quants : attempt to fix Arm 32-bit support
* ggml : add some informative comments in q1_3 vec_dot
* ggml : add TQ1_0 and TQ2_0 ternary quantization types
* ggml : even faster TQ2_0
* ggml : also faster TQ1_0
Same optimization as for TQ2_0 by offsetting the sum instead of the weights.
This makes TQ1_0 almost as fast as Q8_0 on AVX2.
* ggml : fix build issues in certain environments
* ggml : add NEON vec_dot implementation for TQ1_0 and TQ2_0
* ggml : avoid directly using vmlal_high_s8, for 32-bit ARM compat
The compiler seems smart enough to use the same instruction
even when using vget_high_s8 instead.
* ggml : remove q1_3 and q2_2
No more 1.625 bpw and 2.000 bpw,
now instead using 1.6875 bpw and 2.0625 bpw
with TQ1_0 and TQ2_0, respectively.
* llama : remove the separate scale tensors of BitNet b1.58
They won't be needed, since the remaining ternary quant types have
built-in scales.
* ggml-quants : rename fields of TQ1_0 and TQ2_0 structs for consistency
* ggml-quants : allow using vdotq_s32 in TQ2_0 vec_dot
Not yet tested on hardware which supports it,
might not work or might not even compile. But also it might.
It should make the performance better on recent ARM CPUs.
* ggml-quants : remove comment about possible format change of TQ2_0
Making it slightly more convenient for AVX512
but less convenient for everything else is not worth the trouble.
* gguf-py : Numpy (de)quantization for TQ1_0 and TQ2_0
* ggml-quants : use roundf instead of nearest_int for TQ1_0 and TQ2_0
This does not change anything for ternary models,
since their values should never end up being in halfway cases anyway.
* convert : allow direct conversion to TQ1_0 and TQ2_0
The token embeddings and output tensors are kept in F16
to allow quantizing them to Q4_K and Q6_K with llama-quantize.
* llama : handle fallback for TQ1_0 and TQ2_0 with Q4_0
Q4_0 is not completely symmetric (so not lossless for ternary models),
but it should be good enough.
* ggml-quants : allow using ARM dot product instructions for TQ1_0
* ggml-quants : deduplicate TQ1_0 and TQ2_0 __ARM_FEATURE_DOTPROD support
* ggml : remove unused ggml_mul special case
It would otherwise conflict with the more general
optimization coming with Mamba-2.
* ggml : handle TQ1_0 and TQ2_0 in dequantization-based operators
* test-backend-ops : add TQ1_0 and TQ2_0 comments for later
Not yet adding uncommented, because some backends like SYCL and Metal
do not properly handle unknown types in supports_op for GGML_OP_MUL_MAT.
(and Metal also doesn't handle it with GGML_OP_GET_ROWS)
Support for TQ1_0 and TQ2_0 for other backends than CPU
will be added in follow-up pull requests.
* Arm AArch64: optimized GEMV and GEMM kernels for q4_0_q8_0, and q8_0_q8_0 quantization
* Arm AArch64: add optimized GEMV and GEMM asm kernels for q4_0_q8_0 quantization and refactor code to address llama.cpp pr#5780 suggestions
* Arm AArch64: add optimized GEMV and GEMM asm kernels for q4_0_q8_0 quantization and refactor code to address llama.cpp pr#5780 suggestions
* Arm AArch64: add optimized GEMV and GEMM asm kernels for q4_0_q8_0 quantization and refactor code to address llama.cpp pr#5780 suggestions
* Arm AArch64: add optimized GEMV and GEMM asm kernels for q4_0_q8_0 quantization and refactor code to address llama.cpp pr#5780 suggestions
* Arm AArch64: add copyright claim only to ggml-aarch64.cpp and ggml-aarch64.h files
* Arm AArch64: minor code refactoring for rebase
* Arm AArch64: minor code refactoring for resolving a build issue with cmake
* Arm AArch64: minor code refactoring to split the Q4_0_AARC64 type into three separate types: Q4_0_4_4, Q4_0_4_8, and Q4_0_8_8
* Arm AArch64: minor code change for resolving a build issue with server-windows
* retrigger checks
* Arm AArch64: minor code changes for rebase
* Arm AArch64: minor changes to skip the pr#7433 vec_dot code for arm cpus with SVE VL not equal to 256 bits
* Arm AArch64: remove stale LLAMA_QKK_64 from CMakeLists.txt and delete build.zig
* Arm AArch64: add reference scalar gemm and gemv, and avoid dynamic memory allocations during quantization for Q4_0_4_4, Q4_0_4_8, and Q4_0_8_8
* Arm AArch64: add multithreaded quantization support for the new types: Q4_0_4_4, Q4_0_4_8, and Q4_0_8_8
* Arm AArch64: minor code refactoring
* Arm AArch64: simplify logic for calling gemm and gemv functions in ggml_compute_forward_mul_mat
* Arm AArch64: minimize changes in ggml_compute_forward_mul_mat
* Arm AArch64: minor code refactoring, and add reference scalar code to quantize routines for new quant types
* Arm AArch64: minor code refactoring
* Arm AArch64: minor code refactoring
* Arm AArch64: minor code refactoring
* rebase on the latest master commit 3fd62a6 and adapt to the new directory structure
* Arm AArch64: remove a redundant comment
* Arm AArch64: add pragma in ggml-aarch64.c to turn -Woverlength-strings warning off
* Arm AArch64: use __aarch64__ check to guard 64-bit neon kernels
* Arm AArch64: update docs/build.md README to include compile time flags for buiilding the Q4_0_4_4 quant type
* common : gpt_params_parse do not print usage
* common : rework usage print (wip)
* common : valign
* common : rework print_usage
* infill : remove cfg support
* common : reorder args
* server : deduplicate parameters
ggml-ci
* common : add missing header
ggml-ci
* common : remote --random-prompt usages
ggml-ci
* examples : migrate to gpt_params
ggml-ci
* batched-bench : migrate to gpt_params
* retrieval : migrate to gpt_params
* common : change defaults for escape and n_ctx
* common : remove chatml and instruct params
ggml-ci
* common : passkey use gpt_params
* 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
* imatrix: save the dataset file used in the output file
* llama: support kv overrides type string string
* common: factorize KV Overrides parsing between common and server
* quantize: add imatrix n entries and dataset KV metadata
quantize: factorize KV Overrides parsing between common
#6656
* llama: remove kv override str_value initialization as it does not compile on some toolchain
* quantize: add imatrix m_last_call as `quantize.imatrix.chunks_count`
* quantize: add imatrix filename in KV
* llama: add llama_model_kv_override_free
* common: add llama_model_kv_override_free
common: free kv override if used after model loading
* llama: finally move the string KV override value to the stack
* llama : minor
* no need to add a NUL to the std::vector, std::string can be initialized from a pair of iterators.
Co-authored-by: slaren <slarengh@gmail.com>
* kv override: ensure string termination
---------
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
Co-authored-by: slaren <slarengh@gmail.com>
* Implement '--keep-split' to quantize model into several shards
* Add test script
* Update examples/quantize/quantize.cpp
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
* Split model correctly even if tensor id is out-of-order
* Update llama_model_quantize_params
* Fix preci failures
---------
Co-authored-by: z5269887 <z5269887@unsw.edu.au>
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
* ggml : update mul_mat_id to use the same tensor for all the experts
* update cuda
* minor
* update metal
* update test-backend-ops
* fix cuda
* Update ggml-metal.m
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
* update convert.py
* update convert-hf-to-gguf.py
* update convert.py for mixtral hf models
* Update convert-hf-to-gguf.py
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
* cuda : support non-pow-2 number of experts
* allow quantize to work for split and merged experts models in the same way
* cleanup + disable mmap automatically with split tensors models
* update imatrix
* test-backend-ops : test qwen argsort
* update grok model loading
* llama : add merged experts tensors to the grok tensor map
* minor
* gguf : bump version
* fix quantizing of merged experts
* convert-hf-to-gguf.py : update grok (untested)
* make linter happy
* cuda/argsort : use shared memory instead of pool memory
* convert : fix grok tensor names
* metal : add support for non-pow-2 argsort
* llama : more loader cleanup, better error checking
* cuda : fix warning
* llama : still use mmap for loading old models, but copy the data to a host buffer
* add review note
* llama : remove ffn tensor counting + add sanity check
ggml-ci
* convert : fix handling of n_experts == None
ggml-ci
* imatrix : fix ncall counters
* llama : produce error if imatrix size does not match
* quantize : terminate on errors + trace logs
ggml-ci
* metal : pad shared memory to 16 bytes
---------
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
* iq1_m: basics
* iq1_m: basics-2
* iq1_m: CUDA dequantize works
Very 1st shot I get PPL = 9.76 for LLaMA-v2-7B.
* iq1_m: separate shifts for each group of 8 in a block
We get
PPL(LLaMA-v2-7B ) = 9.2810
PPL(LLaMA-v2-13B) = 6.8105
Not bad, but slightly higher than
sqrt(PPL(IQ1_S) * PPL(IQ2_XXS))
which is the expected outcome given that IQ1_M is
halfway between IQ1_S and IQ2_XXS in terms of bpw.
From this, we would expect
PPL = 9.14 for LLaMA-v2-7B
PPL = 6.63 for LLaMA-v2-13B
* iq1_m: go to 3-bit scales
There is slight increase in PPL, but the 0.0625 bpw reduction
in size is totally worth it.
We now have
PPL(LLaMA-v2-7B ) = 9.4469 at 1.96 bpw
PPL(LLaMA-v2-13B) = 6.8717 at 1.93 bpw
PPL(LLaMA-v2-70B) = 4.8568 at 1.85 bpw
* iq1_m: scalar dot product
* iq1_m: AVX2 dot product
* iq1_m: very slightly faster AVX2 dot product
* iq1_m: ARM_NEON dot product
Works, but very slow (10.5 t/s)
* iq1_m: Metal - dequantize works, dot product does not
* iq1_m: Metal now works
About the same performance as iq1_s.
* iq1_m: minor
* iq1_m: checking pure iq1_m quantization
It is pretty bad: PPL(LLaMA-v2-7B) = 34 if we quantize output.weight
with Q4_K.
* iiq1_m: slightly faster ARM_NEON dot product
10.5 t/s -> 11.65 t/s
* iq1_m: faster ARM_NEON dot product
11.65 t/s -> 14.9 t/s
* iq1_m: another minor ARM_NEON dot product improvement
14.9 -> 15.0 t/s
* iq1_m: small PPL improvement via super-block scale adjustment
After quantizing block scales redo the super-block scale fit.
PPL(LLaMA-v2-7B ) = 9.3346
PPL(LLaMA-v2-13B) = 6.8419
PPL(LLaMA-v2-70B) = 4.8294
PPL(Mistral-7B ) = 8.1624
* iq1_m: adapt to CUDA refactoring
* iq1_m: remove unused variable
We have progressed to warnings being errors.
* iq1_m: add to backend-ops tests
* iq1_m: fix Windows ARM
* iq1_m: use common definition of iq1m_scale_t
* cuda: assert -> NO_DEVICE_CODE
* iq1_M: PR comments
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* quantize: be able to override metadata by key
* minor : spacing
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
* quantize: be able to specify the output tensor type
* quantize: be able to specify the token embedding tensor type
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
