diff options
| author | Kawrakow <iwankawrakow@gmail.com> | 2025-02-09 19:48:44 +0200 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2025-02-09 19:48:44 +0200 |
| commit | c12f73ba6153d162f36434cb48e36dd3649b7701 (patch) | |
| tree | 3594697f680959c7130a3f109cd9e4abbc8d6e7d /gguf-py/gguf/tensor_mapping.py | |
| parent | cae2b81155fdad75b7beab3a835c438120412969 (diff) | |
Add optional MLA (#188)
* Deepseek MLA Optimizations
Co-authored-by: Stanisław Szymczyk <sszymczy@gmail.com>
* Make MLA optional
* Remove some unnecessary copies in the MLA attention
* Deepseek MLA Optimizations V2 (#195)
* Avoid allocating MHA KV cache when MLA is turned on
* Added missing gguf-py file
* Added final optimizations
Co-authored-by: Stanisław Szymczyk <sszymczy@gmail.com>
* Make sure we do have wk_b and wv_b before enabling MLA
---------
Co-authored-by: Stanisław Szymczyk <sszymczy@gmail.com>
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
* Use type_k and type_v to set the types of the MLA caches
They were hard-coded at f16.
On my Ryzen-7950X with native bf16 support I get a fairly
significant PP performance boost with bf16 KV-cache:
PP-4096 = 320 t/s up from 292 t/s with fp16 KV-cache.
* Better gemm strategy when nth > nhead
It gives a ~10% PP performance boost for DeepSeek-Lite with 32 threads
(with or without MLA).
Before this commit, when nth > nhead heads were processed
sequentially with all nth threads participating in each
matrix multiplication. Now we ind the gcd of nhead and
nth and split threads into nth/gcd groups, each group
processing nhead/gcd heads.
---------
Co-authored-by: Saood Karim <saood05@gmail.com>
Co-authored-by: Stanisław Szymczyk <sszymczy@gmail.com>
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
Diffstat (limited to 'gguf-py/gguf/tensor_mapping.py')
| -rw-r--r-- | gguf-py/gguf/tensor_mapping.py | 8 |
1 files changed, 8 insertions, 0 deletions
diff --git a/gguf-py/gguf/tensor_mapping.py b/gguf-py/gguf/tensor_mapping.py index a70b69c5..e8725426 100644 --- a/gguf-py/gguf/tensor_mapping.py +++ b/gguf-py/gguf/tensor_mapping.py @@ -446,6 +446,14 @@ class TensorNameMap: "model.layers.{bid}.self_attn.kv_b_proj", # deepseek2 ), + MODEL_TENSOR.ATTN_K_B: ( + "model.layers.{bid}.self_attn.k_b_proj", # deepseek2 + ), + + MODEL_TENSOR.ATTN_V_B: ( + "model.layers.{bid}.self_attn.v_b_proj", # deepseek2 + ), + MODEL_TENSOR.ATTN_Q_A_NORM: ( "model.layers.{bid}.self_attn.q_a_layernorm", # deepseek2 ), |