mirror of
https://github.com/ikawrakow/ik_llama.cpp.git
synced 2026-01-26 17:20:01 +00:00
25d34e3d2f
* WIP * WIP * WIP * Testing Trellis quantization Using 12 bits per 8 weights I get a better rmse than iq2_xxs. I still need to see how quantizing the group-of-8 scales will affect accuracy. By AVX2 SIMDifying the search for the best code, LLaMA-3.1-8B gets quantized in 130 seconds on the Ryzen-7950X CPU - sluggish but still acceptable. * Testing Trellis quantization: 4-bit quantized block scales rmse increases by just 3%, so this is beating iq2_xss in terms of rmse at the same 2.0625 bpw. * Testing Trellis quantization: playing with scales and generators * iq2_kt: quantize / dequantize I now see that I was comparing apples to oranges: iq2_xxs was using a weight of sigma^2/4 + x^2, while the Trellis approach wasn't (weight = 1). Once I use the same weight, iq2_kt is actually slightly worse than iq2_xxs in terms of rmse, so does not look promising at this point. Also, once each group of 8 Trellis values no longer has a constant sum(q^2) that we can precompute, quantization becomes significantly slower (476 seconds for LLaMA-3.1-8B). * iq2_kt: CUDA dequantize so we can run perplexity calcs. As already indicated by rmse, the 2-bit trellis approach is quite a bit worse than iq2_xxs. * WIP * WIP * WIP - try larger blocks With blocks of 32 and 16 bits per groups of 8 the brute force seach becomes prohibitive in terms of CPU time (30+ minutes for 8B LLaMA after SIMDifying with AVX2). The trick is to group the points in clusters, find the nearest cluster, and only search within the cluster. * iq2_kt - this is better Using blocks of 32 and 16 bits per group of 8 weights it beats iq2_xxs in terms of PPL by a significant margin. It is 0.0625 bpw larger, but even if we go to 15 bits per group od 8 (so 0.0625 bpw less than iq2_xxs), PPL is still lower. * iq2_kt - even better Re-quantize after determining block scales (at the epxense of much longer quantization time). * iq2_kt: CUDA dot product Implemented as DMMV. Very slow - just 81 t/s for LLaMA-3.1-8B. Then again, Q2_K_S with forced to use DMMV only gets 112 t/s vs 145 t/s via MMVQ. My memory is that when the DMMV kernels were properly maintained/used, DMMV was about on par with MMVQ for k-quants on my GPU. * iq2_kt: very slightly faster CUDA dot product * iq2_kt: f16 CUDA dot product We arrive at 112 t/s. * iq2_kt: faster f16 CUDA dot product We arrive at 139 t/s (no FA), and 149 t/s (FA). My RTX-4080 is ~20% slower than the RTX-6000 quoted in the QTIP repository, so with FA (which I'm sure they also used) we are at around ~180 t/s on their GPU, so almost matching their performance. * iq2_kt: faster f16 CUDA dot product We arrive at 146 t/s (no FA), and 158 t/s (FA). This is measured for LLaMA-3.1-8B with output.weight left as f16. * Minor * Adding iq3_kt 3.125 bpw. So far does not look good on the PPL vs bpw plot. * Forgotten change * WIP * WIP * iq3_kt WIP: slowly improving PPL(LLaMA-3.1-8B-Instruct, 8192) is now 6.8322, which is starting to be competitive/slightly better than other quants. * WIP * iq3_kt WIP: slowly improving PPL(LLaMA-3.1-8B-Instruct, 8192) is now 6.7892 * iq3_kt WIP: slowly improving PPL(LLaMA-3.1-8B-Instruct, 8192) is now 6.7689 after shrinking by 0.015 bpw by using iq4_k instead of q5_k for attn_v. * iq3_kt WIP: speed up quantization Nearly 60% improvement of quantization speed by having the points nelonging to a cluster copied to contiguous memory during initialization, and then accessed sequantially while searching for the closest point. LLaMA-3.1-8B now gets quantized in ~150 seconds on the Ryzen-5975WX. * iq3_kt speed up quantization Same trick as last commit applied to iq2_kt. Here we get an even larger speedup: quantization time on the Ryzen-5975WX for LLaMA-3.1-8B drops to 195 seconds from 375 seconds! * iq3_kt: CUDA dot product * iq2_kt: SOTA We arrive at PPL(LLaMA-3.1-8B-Instruct, 8192) = 9.2406 PPL(LLaMA-2-7B, 4096) = 6.4179 * iq2_kt: SOTA We arrive at PPL(LLaMA-3.1-8B-Instruct, 8192) = 9.1642 PPL(LLaMA-2-7B, 4096) = 6.3920 * Adding iq4_kt - not competitive at this point * WIP * WIP * iq4_kt: CUDA dot product * iq4_kt: minor tweaks * iq2_kt: SOTA We arrive at PPL(LLaMA-3.1-8B-Instruct, 8192) = 9.1642 PPL(LLaMA-2-7B, 4096) = 6.3920 * iq2_kt: SOTA We arrive at PPL(LLaMA-3.1-8B-Instruct, 8192) = 9.0297 PPL(LLaMA-2-7B, 4096) = 6.3913 Ah, quantization is faster too. About 20% faster. * iq3_kt: small improvements and faster quantization * iq2_kt: SOTA We arrive at PPL(LLaMA-3.1-8B-Instruct, 8192) = 8.9627 PPL(LLaMA-2-7B, 4096) = 6.3825 Quantization is faster too: ~200 seconds for LLaMA-3.1-8B on Ryzen-5975WX. * iq3_kt: small progress * WIP * iq4_kt: go to 4.0 bpw 15 bits per group of 4, plus 8 bit scales ifor blocks of 32. This gives a slightly better PPL than iq4_kss. * iq4_kt: very slightly better at the expense of much longer quantization time. * iq4_kt: failed attemt to adjust CUDA dot product It was working for 4.125 bpw. But after changing to 4.0 bpw there is something wrong and I don't see the bug. * DRY * DRY * iq4_kt: CUDA dot product works * DRY * Report actual bpw * Minor tweaks * Checkpoint Go to groups of 8 for iq3_kt. 2 x 8 = 16 bits for the magnitude plus 1 bpw for the sign. It goves a visible improvement in the PPL vs bpw plot, but that comes at the expense of much longer quantization time (7.5 minutes for LLaMA-3.1-8B on the Ryzen-5975WX). I also notices that the 3INST generator is not actually generating a Gaussian distribution. But going to a better generator means readjusting all the hyper-parameters, so leaving it for later. * WIP for IQ2_KT * WIP - working basic iq2_kt * still super slow (0.17t/s eval) * flatten 3inst iters + avx2 (0.3t/s eval) * iq3_kt (0.3t/s eval) and renames * wip buggy iq4_KT * fix (0.22t/s eval) * naming and remove unused fn * cleanup * more cleanup * delete unused and noncompiling mmvq functions * Some performance tweaks * Slighty faster iq2_kt * port Trellis struct to iq3_kt, iq4_kt * oops untracked files --------- Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
quantize
You can also use the GGUF-my-repo space on Hugging Face to build your own quants without any setup.
Note: It is synced from llama.cpp main every 6 hours.
Example usage:
# obtain the official LLaMA model weights and place them in ./models
ls ./models
llama-2-7b tokenizer_checklist.chk tokenizer.model
# [Optional] for models using BPE tokenizers
ls ./models
<folder containing weights and tokenizer json> vocab.json
# [Optional] for PyTorch .bin models like Mistral-7B
ls ./models
<folder containing weights and tokenizer json>
# install Python dependencies
python3 -m pip install -r requirements.txt
# convert the model to ggml FP16 format
python3 convert_hf_to_gguf.py models/mymodel/
# quantize the model to 4-bits (using Q4_K_M method)
./llama-quantize ./models/mymodel/ggml-model-f16.gguf ./models/mymodel/ggml-model-Q4_K_M.gguf Q4_K_M
# update the gguf filetype to current version if older version is now unsupported
./llama-quantize ./models/mymodel/ggml-model-Q4_K_M.gguf ./models/mymodel/ggml-model-Q4_K_M-v2.gguf COPY
Run the quantized model:
# start inference on a gguf model
./llama-cli -m ./models/mymodel/ggml-model-Q4_K_M.gguf -n 128
When running the larger models, make sure you have enough disk space to store all the intermediate files.
Memory/Disk Requirements
As the models are currently fully loaded into memory, you will need adequate disk space to save them and sufficient RAM to load them. At the moment, memory and disk requirements are the same.
| Model | Original size | Quantized size (Q4_0) |
|---|---|---|
| 7B | 13 GB | 3.9 GB |
| 13B | 24 GB | 7.8 GB |
| 30B | 60 GB | 19.5 GB |
| 65B | 120 GB | 38.5 GB |
Quantization
Several quantization methods are supported. They differ in the resulting model disk size and inference speed.
(outdated)
| Model | Measure | F16 | Q4_0 | Q4_1 | Q5_0 | Q5_1 | Q8_0 |
|---|---|---|---|---|---|---|---|
| 7B | perplexity | 5.9066 | 6.1565 | 6.0912 | 5.9862 | 5.9481 | 5.9070 |
| 7B | file size | 13.0G | 3.5G | 3.9G | 4.3G | 4.7G | 6.7G |
| 7B | ms/tok @ 4th | 127 | 55 | 54 | 76 | 83 | 72 |
| 7B | ms/tok @ 8th | 122 | 43 | 45 | 52 | 56 | 67 |
| 7B | bits/weight | 16.0 | 4.5 | 5.0 | 5.5 | 6.0 | 8.5 |
| 13B | perplexity | 5.2543 | 5.3860 | 5.3608 | 5.2856 | 5.2706 | 5.2548 |
| 13B | file size | 25.0G | 6.8G | 7.6G | 8.3G | 9.1G | 13G |
| 13B | ms/tok @ 4th | - | 103 | 105 | 148 | 160 | 131 |
| 13B | ms/tok @ 8th | - | 73 | 82 | 98 | 105 | 128 |
| 13B | bits/weight | 16.0 | 4.5 | 5.0 | 5.5 | 6.0 | 8.5 |
- k-quants
- recent k-quants improvements and new i-quants
- #2707
- #2807
- #4773 - 2-bit i-quants (inference)
- #4856 - 2-bit i-quants (inference)
- #4861 - importance matrix
- #4872 - MoE models
- #4897 - 2-bit quantization
- #4930 - imatrix for all k-quants
- #4951 - imatrix on the GPU
- #4969 - imatrix for legacy quants
- #4996 - k-qunats tuning
- #5060 - Q3_K_XS
- #5196 - 3-bit i-quants
- quantization tuning, another one, and another one
Llama 2 7B
| Quantization | Bits per Weight (BPW) |
|---|---|
| Q2_K | 3.35 |
| Q3_K_S | 3.50 |
| Q3_K_M | 3.91 |
| Q3_K_L | 4.27 |
| Q4_K_S | 4.58 |
| Q4_K_M | 4.84 |
| Q5_K_S | 5.52 |
| Q5_K_M | 5.68 |
| Q6_K | 6.56 |
Llama 2 13B
| Quantization | Bits per Weight (BPW) |
|---|---|
| Q2_K | 3.34 |
| Q3_K_S | 3.48 |
| Q3_K_M | 3.89 |
| Q3_K_L | 4.26 |
| Q4_K_S | 4.56 |
| Q4_K_M | 4.83 |
| Q5_K_S | 5.51 |
| Q5_K_M | 5.67 |
| Q6_K | 6.56 |
Llama 2 70B
| Quantization | Bits per Weight (BPW) |
|---|---|
| Q2_K | 3.40 |
| Q3_K_S | 3.47 |
| Q3_K_M | 3.85 |
| Q3_K_L | 4.19 |
| Q4_K_S | 4.53 |
| Q4_K_M | 4.80 |
| Q5_K_S | 5.50 |
| Q5_K_M | 5.65 |
| Q6_K | 6.56 |