ggerganov
Follow
Pinned
3,977 contributions in the last year
Less
No contributions.
Low contributions.
Medium-low contributions.
Medium-high contributions.
High contributions.
More
Contribution activity
March 2024
Created 116 commits in 3 repositories
Created a pull request in ggerganov/llama.cpp that received 21 comments
server : refactor
ref #4216
Moved the code around to logically similar things closer together and did some renaming. The cache_tokens management should be improved -…
+2,265
−2,711
lines changed
•
21
comments
Opened 28 other pull requests in 5 repositories
ggerganov/llama.cpp
22
merged
1
open
-
llama : fix integer overflow during quantization
This contribution was made on Mar 14
-
ggml : designate enum vals for integer types
This contribution was made on Mar 14
-
metal : build metallib + fix embed path
This contribution was made on Mar 12
-
ggml : fix UB in IQ2_S and IQ3_S
This contribution was made on Mar 12
-
sycl : try to fix SYCL after IQ1_S changes
This contribution was made on Mar 11
-
llama : more consistent names of count variables
This contribution was made on Mar 11
-
llama : refactor unicode stuff
This contribution was made on Mar 11
-
metal : move mm_id indices to shared mem
This contribution was made on Mar 10
-
nix: update flake.lock
This contribution was made on Mar 10
-
server : fix metrics init
This contribution was made on Mar 9
-
server : clarify some items in the readme
This contribution was made on Mar 9
-
server : simplify logic for empty prompts
This contribution was made on Mar 9
-
ggml : fix unnecessary f32 -> f16 -> f32 casts (mmla)
This contribution was made on Mar 9
-
ggml : reuse quantum structs across backends
This contribution was made on Mar 8
-
ggml : remove old quantization functions
This contribution was made on Mar 8
-
ggml : add ggml-common.h to deduplicate shared code
This contribution was made on Mar 8
-
server : fix EOS token detection with disabled cache
This contribution was made on Mar 8
-
llama-bench : add embeddings option
This contribution was made on Mar 7
-
llama : compute BERT graph with F16 K, V
This contribution was made on Mar 5
-
sync : ggml
This contribution was made on Mar 4
-
nix: update flake.lock
This contribution was made on Mar 3
-
ggml : fix IQ3_S AVX implementation
This contribution was made on Mar 2
-
server : remove api_like_OAI.py proxy script
This contribution was made on Mar 1
ggerganov/ggml
2
merged
-
spec : add GGUF diagram
This contribution was made on Mar 15
-
sync : llama.cpp
This contribution was made on Mar 14
NousResearch/nous-llama.cpp
1
merged
-
control-vectors : minor code style updates
This contribution was made on Mar 14
ggerganov/whisper.cpp
1
merged
-
ggml : try fix 32-bit arm compat
This contribution was made on Mar 8
pacman100/llama.cpp
1
merged
-
starcoder2 : change rope type to neox
This contribution was made on Mar 1
Reviewed 104 pull requests in 4 repositories
ggerganov/llama.cpp
25 pull requests
-
issues: ci - close inactive issue with workflow
This contribution was made on Mar 16
-
proposal: common: llama_load_model_from_url
This contribution was made on Mar 16
-
llama : fix Baichuan2 13B
This contribution was made on Mar 15
-
Add support for control vectors
This contribution was made on Mar 15
-
Add Command-R Model
This contribution was made on Mar 15
-
gritlm: add initial README.md to examples/gritlm
This contribution was made on Mar 15
-
llava: change llava API to pure C style for Rust FFI bindgen
This contribution was made on Mar 15
-
Add qwen2moe
This contribution was made on Mar 15
-
cuda : disable unused cudaLaunchHostFunc code
This contribution was made on Mar 15
-
[SYCL] Fix non-intel device selection
This contribution was made on Mar 15
-
[SYCL] fix set main gpu error, support single/mul gpu mode
This contribution was made on Mar 15
-
gguf : add support for I64 and F64 arrays
This contribution was made on Mar 15
-
Add Orion chat template
This contribution was made on Mar 15
-
llama-bench : use random tokens to improve accuracy with mixtral
This contribution was made on Mar 15
-
gguf-py : bump version to 0.8.0
This contribution was made on Mar 14
-
Add Unicode model filename support for Windows
This contribution was made on Mar 14
-
[SYCL] fix set main gpu error, support single/mul gpu mode
This contribution was made on Mar 14
-
readme: improve readme for Llava-1.6 example
This contribution was made on Mar 14
-
server: test: disable debug release type sanitizer, simplify trigger
This contribution was made on Mar 14
-
Fix: attempt to reduce the impact of a worst-case scenario on defragmentation
This contribution was made on Mar 14
-
gguf-py: add support for I8, I16 and I32
This contribution was made on Mar 14
-
Add ggml_cann backend
This contribution was made on Mar 14
-
README.md: Update details about running llama in Termux on Android
This contribution was made on Mar 13
-
Fix: GBNF missing "root" node crashing server
This contribution was made on Mar 13
-
llama : add pipeline parallelism support
This contribution was made on Mar 13
- Some pull request reviews not shown.
ggerganov/whisper.cpp
9 pull requests
-
whisper : document whisper_batch.n_seq_id
This contribution was made on Mar 10
-
whisper : improve beam search candidate diversity
This contribution was made on Mar 10
-
[DRAFT] Token level timestamps with DTW (#375)
This contribution was made on Mar 9
-
bindings/go : add linker flags to make metal work
This contribution was made on Mar 9
-
whisper : make beam candidate sort more stable
This contribution was made on Mar 9
-
Fix typo in source file whisper.cpp
This contribution was made on Mar 5
-
Fix SF(segment fault) issue in Android JNI
This contribution was made on Mar 5
-
Add library versioning
This contribution was made on Mar 4
-
Update README to Recommend MacOS Sonoma for Core ML to avoid hallucination
This contribution was made on Mar 4
ggerganov/ggml
3 pull requests
-
gguf : add Mamba keys and tensors
This contribution was made on Mar 13
-
ggml_status introduction
This contribution was made on Mar 4
-
add some new ops, fix some operators and add batch operations to certain operators.
This contribution was made on Mar 3
huggingface/huggingface.js
2 pull requests
-
GGUF parser: support big-endian files
This contribution was made on Mar 12
-
a GGUF parser that works on remotely hosted files (over HTTP range requests)
This contribution was made on Mar 12
Created an issue in ggerganov/llama.cpp that received 2 comments
llama : add Deepseek support
Creating this issue for more visibility The main problem is around tokenization support, since the models use some variation of the BPE pre-process…
2
comments
Opened 2 other issues in 2 repositories
ggerganov/whisper.cpp
1
open
-
whisper : adapt to latest ggml changes
This contribution was made on Mar 15
ggerganov/llama.cpp
1
open
-
llama : combine expert tensors into a single tensor
This contribution was made on Mar 15
Answered 1 discussion in 1 repository
ggerganov/llama.cpp
ggerganov/llama.cpp
-
Is llama.cpp designed to be consumed via CLI or C programs?
This contribution was made on Mar 10