ikawrakow/ik_llama.cpp
1
0
Fork 0
mirror of https://github.com/ikawrakow/ik_llama.cpp.git synced 2026-01-26 17:20:01 +00:00
Code Issues Packages Projects Releases Wiki Activity

Adding ability to have meta data per tensor row (#61)

Browse Source 
* POC: per row scale

This is a POC how to work around opinionated ggml to
have scales per row rather than per block.
Only implemened for Zen4 and only for iq2_tn.

* POC per row scale: iq2_tn on NEON

* POC per row scale: iq2_tn on Metal

* Per row scale Metal templates

* iq1_tn: shrink to 1.625 bpw (NEON and Metal)

* POC per row scale: CUDA

* POC per row scale: add CUDA TODOs

There are two places in ggml-cuda.cu left where it is assumed
that type_size * n_per_row / block_size is the way to compute
and handle row sizes. This does not affect simple usage,
but will lead to issues when tensors are split between GPUs.

* Per row scales - CUDA

The only place left where there are unnecessary assumptions being made
is in the Flash Attention code. As we are not using any quants that
use per row scales for quantized KV cache, it should be OK for now.

* Update IQ1_TN and IQ2_TN bpw shown to user

---------

Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
This commit is contained in:
Kawrakow
2024-09-27 08:16:06 +03:00
committed by GitHub
parent 1cdb6993ee
commit 733660accd
12 changed files with 171 additions and 138 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
examples/quantize/quantize.cpp
View File

@@ -28,8 +28,8 @@ static const std::vector<struct quant_option> QUANT_OPTIONS = {
{ "IQ1_M", LLAMA_FTYPE_MOSTLY_IQ1_M, " 1.75 bpw quantization", },
{ "IQ1_BN", LLAMA_FTYPE_MOSTLY_IQ1_BN, " 1.62 bpw quantization (Bitnet)", },
{ "IQ2_BN", LLAMA_FTYPE_MOSTLY_IQ2_BN, " 2.00 bpw quantization (Bitnet)", },
{ "IQ1_TN", LLAMA_FTYPE_MOSTLY_IQ1_TN, " 1.69 bpw quantization (TriLM)", },
{ "IQ2_TN", LLAMA_FTYPE_MOSTLY_IQ2_TN, " 2.06 bpw quantization (TriLM)", },
{ "IQ1_TN", LLAMA_FTYPE_MOSTLY_IQ1_TN, " 1.63 bpw quantization (TriLM)", },
{ "IQ2_TN", LLAMA_FTYPE_MOSTLY_IQ2_TN, " 2.00 bpw quantization (TriLM)", },
{ "Q2_K", LLAMA_FTYPE_MOSTLY_Q2_K, " 2.63G, +0.6717 ppl @ LLaMA-v1-7B", },
{ "Q2_K_S", LLAMA_FTYPE_MOSTLY_Q2_K_S, " 2.16G, +9.0634 ppl @ LLaMA-v1-7B", },
{ "IQ3_XXS", LLAMA_FTYPE_MOSTLY_IQ3_XXS, " 3.06 bpw quantization", },