CISC
Follow
Popular repositories Loading
-
-
-
-
llama-cpp-python
llama-cpp-python PublicForked from abetlen/llama-cpp-python
Python bindings for llama.cpp
Python
-
410 contributions in the last year
Less
No contributions.
Low contributions.
Medium-low contributions.
Medium-high contributions.
High contributions.
More
Contribution activity
July 2025
Created 28 commits in 2 repositories
Created a pull request in ggml-org/llama.cpp that received 14 comments
llama : remove ggml_cont where possible
AFAIK all backends support non-contiguous norm/rope, therefore we can remove a lot of unnecessary ggml_cont()s by directly creating a 3D view befor…
+38
−48
lines changed
•
14
comments
Opened 15 other pull requests in 1 repository
ggml-org/llama.cpp
11
merged
3
closed
1
open
-
opencl : fix im2col when KW!=KH
This contribution was made on Jul 21
-
sycl : fix im2col when KW!=KH
This contribution was made on Jul 21
-
cuda : implement bf16 cpy ops and enable bf16 cont
This contribution was made on Jul 18
-
llama : fix exaone4 gf usage
This contribution was made on Jul 18
-
ci : disable failing vulkan crossbuilds
This contribution was made on Jul 16
-
convert : make hf token optional
This contribution was made on Jul 16
-
convert : only check for tokenizer folder if we need it
This contribution was made on Jul 15
-
convert : add pre-computed hashes first to prevent order mishaps
This contribution was made on Jul 15
-
cuda : add set rows for bf16
This contribution was made on Jul 13
-
llama : remove unintended whitespace
This contribution was made on Jul 9
-
llama : revert gptneox cont removal
This contribution was made on Jul 7
-
llama : fix incorrect minicpm3 v_states shape
This contribution was made on Jul 7
-
gguf-py : add support for chat template jinja files
This contribution was made on Jul 2
-
ci : add OpenCL to labeler workflow
This contribution was made on Jul 2
-
opencl : add GELU_ERF
This contribution was made on Jul 1
Reviewed 39 pull requests in 1 repository
ggml-org/llama.cpp
25 pull requests
-
sycl: Fix im2col
This contribution was made on Jul 21
-
cuda : implement bf16 cpy ops and enable bf16 cont
This contribution was made on Jul 21
-
model : jina-embeddings-v3 support
This contribution was made on Jul 20
-
Fix link for tools/perplexity in README.md
This contribution was made on Jul 20
-
Add LLaDA 8b Diffusion model
This contribution was made on Jul 20
-
Fix Ernie4.5 MoE without shared experts
This contribution was made on Jul 17
-
Add EXAONE 4.0 model architecture
This contribution was made on Jul 17
-
Fix parameter order issue for hybrid memory initialization
This contribution was made on Jul 16
-
Bug: fix inputs to conv1d in mamba layer of plamo2
This contribution was made on Jul 16
-
model : fix parallel processing for lfm2
This contribution was made on Jul 16
-
scripts: add bpw per layer and model
This contribution was made on Jul 15
-
Model : Add support for Kimi-K2
This contribution was made on Jul 15
-
Support diffusion models: Add Dream 7B
This contribution was made on Jul 15
-
model : add PLaMo-2 model
This contribution was made on Jul 15
-
bug fix: handle saving/loading null layers in recurrent memory
This contribution was made on Jul 14
-
Model: Add support for Ernie 4.5 MoE
This contribution was made on Jul 14
-
quantize: fix minor logic flaw in --tensor-type
This contribution was made on Jul 13
-
llama : add jinja template for rwkv-world
This contribution was made on Jul 13
-
Add ELU CUDA support
This contribution was made on Jul 13
-
test-backend-ops : cover lfm2 cases in test_ssm_conv
This contribution was made on Jul 12
-
readme : add LFM2 to models section
This contribution was made on Jul 12
-
HIP: Add HIP 7.0+ compatibility for hipBLAS compute types
This contribution was made on Jul 11
-
model : add Midm-2.0 model
This contribution was made on Jul 11
-
Granite Four
This contribution was made on Jul 10
-
llama : support LiquidAI LFM2 hybrid model family
This contribution was made on Jul 10
- Some pull request reviews not shown.
Answered 1 discussion in 1 repository
ggml-org/llama.cpp
ggml-org/llama.cpp
-
converting models not possible again
This contribution was made on Jul 12