* Remove kv_l, kvt_l and just use k_l and v_l
* Hopefully take care of missing V cache (MLA)
* Fix save and restore when there is no V cache
* Fix double print
* Update write_kv_cache_data and read_kv_cache_data to be MLA aware
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* Adding top-n-sigma sampler
* Fix typos in XTC PR
* Update README.md for main and server
* More README
* More README
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* Remove kv_l, kvt_l and just use k_l and v_l
* Hopefully take care of missing V cache (MLA)
* Replace MLA-specific KV cache with the standard KV cache V2 (#473)
* Fix save and restore when there is no V cache
* Fix double print
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
Co-authored-by: saood06 <saood05@gmail.com>
* Legacy quants conversion schemes in convert_hf_to_gguf.py
This, notably in order to make smaller conversions to generate an iMatrix file.
`Q4_0`,`Q4_1` are here using embeddings, output, attn_k and attn_v in q5_0.
`Q5_0`,`Q5_1` are here using embeddings, output, attn_k and attn_v in q8_0.
Adapted from the following llama.cpp mainline PR : https://github.com/ggml-org/llama.cpp/pull/9022
Original author @chentyjpm
Also, 2 forgotten mentions of FTYPE IQ3_KL in llama.cpp file.
* forgotten IQ5_KS case mention
* 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>
* Streamline a bit the quant strategies
No change over the existing patterns, except for the bump for attn_k and attn_v for the models with 4 and 6 experts (several frankensteins seen on HF, and which also use GQA).
The rest is applying the existing patterns to the new IQ_K quants.
Also, a Q8_0 for attn_q slipped into the MOEs 8 experts rule, I removed it, because that tensor is much bigger than attn_k or attn_v.
* remove <=8 experts condition.
* iq5_ks: basics
* iq5_ks: quantize
* iq5_ks: CUDA dequantize works
* iq5_ks: dot product works on CUDA
* iq5_ks: MMQ works
* iq5_ks: Zen4
* iq5_ks: AVX2
But is is not quite right, just like iq4_k, iq5_k, iq6_k, iq4_ks.
All these need fixing on AVX2.
* iq5_ks: NEON
* iq5_ks: Metal dequantize
* iq5_ks: Metal dot product
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* Enable MLA-3 in crippled GGUFs: WIP
* Enable MLA-3 in crippled GGUFs: seems to work
* Add newly created tensors to model.tensors_by_name
Else they don't get run-time repacked.
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* New DeepSeek FlashMLA
Does not work because the RoPE portion is stored at the end
in our case, while in mainline it is stored at the beginning,
and the FA kernel assumes that.
* Rearrange MLA K cache so it first new CUDA FA implementation
* constexpr and minor changes
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* conflict resolution
* Changes to make work and add longrope support
* Changes to n_attention_wv rule
* Untested support of 253B
* DeciLMCausalModel now reads rope_theta from config.json properly
* Remove errant Granite mentions
* Better n_attention_vw rule
* Update vocab.py
---------
Co-authored-by: Yee Man Chan <ymchan@gmail.com>
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* Add GLM-4-0414 model support
Based on zRzRzRzRzRzRzR's PR on mainline llama.cpp.
Still some issues where it doesn't work:
* offloading >=60 layers to GPU
* no flash attention
* Remove seemingly unused llm_tensor enums
Both of these seem unused and LLM_TENSOR_ATTN_POST_NORM already
existed which seems pretty similar? Don't think they were used in the
python code either...
So removed these as possibly just cruft:
* LLM_TENSOR_POST_ATTN_NORM
* LLM_TENSOR_POST_MLP_NORM
* Set flash attention precision to f32 on GLM4 arch
* Set non flash attention precision to f32 on GLM4
* Remove reshape_3d() for Vcur in build_glm4()
This fixes the non-flash-attention inferencing on both CPU and CUDA.
* Make sure no interleaved quants are being used for token embeddings
also with `--pure` and/or `--custom-q`.
* Simplify
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* WIP - not working
* q8_0 without bells and wistles works
* It works for q8_0
* Use bf16 instead of f16,int16
* q4_0_r8
* q5_0_r4
* q6_0_r4
* Also q4_1 and q5_1
* Add check if selected type is possible with --pure
I often want to quantize with --pure to see quantization performance
without quantization mixes. But for models where there qre tensors
with row sizes that are not multiple of 256, this results in a crash
for k- and i-quants. Hence, lets add a check if the quant selected
via --pure is applicable, and change it if not.
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* Improve DeepSeek batched processing speed
* Revert the commented out section in iqk_mul_mat.cpp
It does have some benefit at long contexts.
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* Adding ability to use THP on Linux
* Use the actual page size4 used for mmap also in munmap
* Add -thp to llama-bench
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* WIP Gemma3: not working
* gemma3: build_gemma3 seems to be working now
* Revert changes to convert_hf_to_gguf.py
It wasn't working, so I guess, it is better to leave the
conversion up tp upstream.
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* Repack a model with the quantize tool
* WIP
* Fixed various issues
As we don't have a way to tell if a repacked quant has been modified,
I had to remove the modification at the expense of a slight decrease
in performance. This affects q8_0_r8, q8_KV_r8, q8_k_r8 on Zen4, and
q4_0_r8 on ARM.
* Create wk_b and wv_b as Q8_0_R8 if the wkv_b type is interleaved
* Fix GCC 13.3 compilation error
* Another one
* Add missing include
* FlashMLA-3: the best of both worlds - CPU only
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* Repack a model with the quantize tool
* WIP
* Fixed various issues
As we don't have a way to tell if a repacked quant has been modified,
I had to remove the modification at the expense of a slight decrease
in performance. This affects q8_0_r8, q8_KV_r8, q8_k_r8 on Zen4, and
q4_0_r8 on ARM.
* Create wk_b and wv_b as Q8_0_R8 if the wkv_b type is interleaved
* Fix GCC 13.3 compilation error
* Another one
* Add missing include
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* FlashMLA-2: eliminate intermediate f32 tensors
This works on the CPU. PP performance is ~13% better for 16k tokens
and compute buffer is quite a bit smaller.
* FlashMLA-2: enable fast path only on the CPU for now
I did implement the necessary ops on CUDA, but something is
still wrong there, so for now we only use it when running
CPU-only.
* FlashMLA-2: slightly smaller computer buffer size
* Prepare wk_b when loading DeepSeek models (if wk_b is missing)
* Add some comments
* Fix case where wkv_b is quantized with k- or i-quants.
* Fix CUDA
There is an issue with quantized GEMV on CUDA when the left operand
(the matrix) is not contiguous. So, for now, we also create wv_b
during model loading and use that instead of the 3D view of wkv_b.
* FlashMLA-2: avoid conversions to f32 also on CUDA
* Be able to compute for more than 65535 tokens
On CUDA just a quick hack that allows us to cancatenate tensors
with more than 65535 rows along zroth dimension as needed by
FlashMLA-2. Also needed some care in the perplexity tool to
avoid int overflows when evaluating the computed logits.
* Reduce memory usage for FlashMLA-2
Oh, also fix int overflow in the CUDA concat implementation.
It is funny how the llama.cpp 64-bit police has gone (almost) everywhere
and replaced 32-bit ints with 64-bit ints, needed or not,
but hasn't done it where it is actually needed.
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* FlashMLA-2: eliminate intermediate f32 tensors
This works on the CPU. PP performance is ~13% better for 16k tokens
and compute buffer is quite a bit smaller.
* FlashMLA-2: enable fast path only on the CPU for now
I did implement the necessary ops on CUDA, but something is
still wrong there, so for now we only use it when running
CPU-only.
* FlashMLA-2: slightly smaller computer buffer size
* Prepare wk_b when loading DeepSeek models (if wk_b is missing)
* Add some comments
* Fix case where wkv_b is quantized with k- or i-quants.
* Fix CUDA
There is an issue with quantized GEMV on CUDA when the left operand
(the matrix) is not contiguous. So, for now, we also create wv_b
during model loading and use that instead of the 3D view of wkv_b.
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* FlashMLA-2: eliminate intermediate f32 tensors
This works on the CPU. PP performance is ~13% better for 16k tokens
and compute buffer is quite a bit smaller.
* FlashMLA-2: enable fast path only on the CPU for now
I did implement the necessary ops on CUDA, but something is
still wrong there, so for now we only use it when running
CPU-only.
* FlashMLA-2: slightly smaller computer buffer size
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* This gives us ~20% TG speedup for DeepSeek on CUDA
* Slightly better
* Also do it for plain (not fused) mul_mat_id
* Guard against numerical precision issues for MLA on CUDA
* imatrix: wv_b <-> wkv_b
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* This gives us ~20% TG speedup for DeepSeek on CUDA
* Slightly better
* Also do it for plain (not fused) mul_mat_id
* Guard against numerical precision issues for MLA on CUDA
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
PP speed is great, almost on par with standard FA.
But TG speed is pathetic. The strangest thing is that
the slowdown is not due to FA, but due to the ffn_gate_exps
gemm, which somehow becomes very slow. WTF?
As I'm unable the resolve the slow ffn_gate_exps GEMM mystery,
for now TG goes via mla=2, PP is via FA.
Also discovered the ggml_cast op, so we don't need the aux
tensors that I had added to the KV cache.
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* FlashMLA-2: faster prompt processing
The current MLA implementation computes
wv_b * (k_cache * softmax(k_cache * (wk_b*q)))
This leads to 3.4X more multiply-adds (madds)
compared to standard attention. Due to the resulting
tensor shapes, TG is still faster than standard attention
because the k_cache*(wk_b*q) and k_cache*(softmax(k_cache * (wk_b*q)))
multiplications become GEMMs, so the additional madds are
more than compensated for due to the much higher performance
of GEMMs compared to GEMVs. But for PP, where we are dealing
with GEMMs in both cases, the additional madds needed for MLA
lead to lower performance, with the performance gap increasing
with context length.
So, then, when we are dealing with PP, we can rearrange the
above to (wv_b * k_cache) * softmax( (wk_b^T*k_cache) * q),
thus transforming it into the standard attention mechanism.
We do need two additional matrix multiplications (which in practice
is done as a single wkv_b * k_cache GEMM) with the *entire*
K cache. But this is still cheaper than MLA, as we end up with
1.8X the madds required by standard attention. Oh, these figures
are for the DeepSeek-V3/R1/Lite attention architecture.
This leads to a significant PP performance increase compared
to standard MLA with FA.
There are many upsides to this:
* If we only apply the above trick when we are processing more than
X tokens (with suitable chosen X), TG performance stays the same
as MLA with FA
* We still need to store just the K-cache, so 576 entries per layer
for DeepSeek-V3/R1/Lite
* We get significantly better PP performance
* We can use MLA+FA on CUDA. It works already with this commit
for PP, something is not yet quite right for TG.
The downside is that it only works with fp16 cache (for now).
This is so because we need to convert the cache to fp32,
else we cannot do the wkv_b * k_cache matrix multiplication
(which in ggml requires the second operand to be fp32).
But converting (copying) to fp32 only works for f16, bf16 and
f32 tensors, so no luck with quantized cache. Another reason
that we need to convert to fp32 is that the cache contains the
RoPE'd portion, which we need to concatenate to the result of
the wkv_b * k_cache matrix multiplication. Also this op
works only when the tensors being concatenated are both fp32.
So much about ggml being a general purpose ML library.
* FlashMLA-2: on the CPU it now works for quantized cache
except for q8_KV (q8_KV has row meta data, and there is still
some confusion with row sizes because of that).
* FlashMLA-2: on the CPU it now works also with q8_KV
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* Custom quantization rules with regular expressions
* Add the --custom-q option to the help
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* WIP CUDA FA with Dk != Dv
* WIP
* CUDA FA WIP - It actually works!
No TG yet, but for PP I can run FA with fp16 cache and it gets
the same answer.
* CUDA FA WIP - it now works for Q8_0 + Q8_0 for KV cache
* CUDA FA WIP - TG, not working yet.
* CUDA FA with Dk != Dv: it works now for DeepSeek
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* FlashMLA - it finally works (on the CPU)
* FlashMLA: allow for f16 and bf16 cache in addition to q8_0
* It works with ggml FA, not with iqk FA
* WIP
* FlashMLA: it now works with iqk
I had forgotten to divide the Q stride by sizeof(float) and
that's why, very cobfusingly, it was working for TG but not for PP.
* WIP
* FlashMLA: that should be it for now
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* A better way to measure the cost of ggml_barrier
* Smart expert selection
* Add ser option to llama-bench
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* This reduces compute buffer size for MLA
* This should accomplish it for standard attention
* Much better
* Better concat for contiguous tensors
If all the op does is to concatenate the second tensor
to the first, why would we want to have a loop?
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
The `-mla` command line option turns into an int from a bool.
mla = 0: use standard attention
mla = 1: use MLA with transposed cache
mla > 1: use MLA without transposed cache
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
