index
:
ffwd
master
Fastforward embedding inference engine
about
summary
refs
log
tree
commit
diff
log msg
author
committer
range
Commit message (
Collapse
)
Author
Age
Files
Lines
...
*
api: dedup scalar bf16-to-f32 into dtype.h
Onur Gündüz
2026-06-16
3
-18
/
+30
|
*
chore: move sources to src/ with proper names, drop qwen_ prefix
Onur Gündüz
2026-06-16
39
-625
/
+580
|
*
chore: document supported models and backends in README
Onur Gündüz
2026-06-16
1
-8
/
+53
|
*
cpu: size default threads to cgroup cpu quota, not host core count
Onur Gündüz
2026-06-16
3
-4
/
+57
|
*
cuda: memory-only bf16 for BERT (f32 compute on bf16 weights)
Onur Gündüz
2026-06-16
1
-6
/
+46
|
*
server: register bge-base and bge-large encoders
Onur Gündüz
2026-06-16
1
-0
/
+6
|
*
all: support BERT tanh-GeLU (gelu_new / gelu_pytorch_tanh)
Onur Gündüz
2026-06-16
8
-9
/
+109
|
*
cli: tokenize BERT models with WordPiece and [CLS]/[SEP]
Onur Gündüz
2026-06-16
2
-37
/
+94
|
*
mlx: upcast BERT norm/bias/embedding fields for BF16 models
Onur Gündüz
2026-06-16
1
-0
/
+13
|
*
tokenizer: Unicode case-fold and accent strip for multilingual WordPiece
Onur Gündüz
2026-06-16
3
-108
/
+5037
|
*
tokenizer: full-Unicode punctuation and control classification
Onur Gündüz
2026-06-16
4
-47
/
+198
|
*
cuda: add the BERT encoder family forward and loader
Onur Gündüz
2026-06-16
1
-9
/
+238
|
*
server: serve BERT encoders (MiniLM, bge) on /v1/embeddings
Onur Gündüz
2026-06-15
2
-42
/
+123
|
*
mlx: add the BERT encoder family forward and loader
Onur Gündüz
2026-06-15
1
-7
/
+269
|
*
gitignore: add test_wordpiece to ignored files
Onur Gündüz
2026-06-15
1
-0
/
+1
|
*
cpu: add the BERT encoder family forward and loader
Onur Gündüz
2026-06-15
4
-17
/
+276
|
*
kernels: add LayerNorm and exact-GeLU for the BERT family
Onur Gündüz
2026-06-15
3
-0
/
+105
|
*
tokenizer: add WordPiece backend for BERT-family models
Onur Gündüz
2026-06-15
4
-0
/
+969
|
*
all: add CLS first-token pooling
Onur Gündüz
2026-06-15
9
-27
/
+165
|
*
gitignore: add test_qwen2 and test_qwen3 to ignored files
Onur Gündüz
2026-06-15
1
-0
/
+2
|
*
server: guard empty-chunk memcpy and check self-pipe writes
Onur Gündüz
2026-06-15
1
-4
/
+12
|
*
all: support GTE-Qwen2 embeddings
Onur Gündüz
2026-06-15
11
-85
/
+574
|
*
api: share config parsing across backends
Onur Gündüz
2026-06-15
5
-280
/
+166
|
*
api: accept Hugging Face model weight prefix
Onur Gündüz
2026-06-15
6
-65
/
+134
|
*
server: add Qwen3 text_type query/document instruction hint
Onur Gündüz
2026-06-15
2
-2
/
+163
|
*
late: add CUDA late-interaction rerank path (GPU encode, host MaxSim)
Onur Gündüz
2026-06-15
5
-13
/
+346
|
*
late: restore MLX device select/concat used by interaction check
Onur Gündüz
2026-06-15
2
-0
/
+111
|
*
late: batch CPU rerank candidate encoding in one packed forward
Onur Gündüz
2026-06-15
4
-21
/
+219
|
*
late: batch MLX rerank candidate encoding in one padded pass
Onur Gündüz
2026-06-15
3
-111
/
+254
|
*
chore: run clang-format
Onur Gündüz
2026-06-15
35
-4024
/
+4742
|
*
cli: update print format in build info
Onur Gündüz
2026-06-15
1
-5
/
+5
|
*
server: version flags, drop cost block, tidy batch-wait note
Onur Gündüz
2026-06-15
2
-37
/
+35
|
*
cli: add --json output, plain matrix, and version flags
Onur Gündüz
2026-06-15
3
-38
/
+122
|
*
build: backend-selectable targets, server-only cjson, version flags
Onur Gündüz
2026-06-15
1
-66
/
+51
|
*
api: add embed_build.h for version and build identity
Onur Gündüz
2026-06-15
1
-0
/
+64
|
*
ci: add github release workflow
Onur Gündüz
2026-06-15
2
-3
/
+231
|
*
server: document HTTP API and add openapi.yaml spec
Onur Gündüz
2026-06-15
2
-44
/
+434
|
*
server: per-family embedding encoding (Qwen3 float, pplx int8)
Onur Gündüz
2026-06-14
3
-57
/
+170
|
*
server: add Qwen3-Embedding-8B model slot
Onur Gündüz
2026-06-14
2
-2
/
+5
|
*
all: add Qwen3-Embedding-4B + fix terminal token
Onur Gündüz
2026-06-14
7
-25
/
+29
|
*
chore: add MIT license
Onur Gündüz
2026-06-14
1
-0
/
+20
|
*
docs: add acknowledgements to README
Onur Gündüz
2026-06-14
1
-0
/
+15
|
*
all: add support for Qwen3-Embedding + causal GQA attention
Onur Gündüz
2026-06-14
16
-132
/
+901
|
*
docs: correct path to rank_embeddings.py in usage example
Onur Gündüz
2026-06-14
1
-1
/
+1
|
*
docs: update README with the model support and usage examples
Onur Gündüz
2026-06-14
1
-3
/
+62
|
*
docs: fix CLI name in comments
Onur Gündüz
2026-06-14
2
-5
/
+2
|
*
cli: scope GPU options to their respective build targets
Onur Gündüz
2026-06-14
3
-122
/
+130
|
*
docs: fix benchstat.py path in comments
Onur Gündüz
2026-06-14
2
-3
/
+3
|
*
server: remove CORS support
Onur Gündüz
2026-06-14
4
-16
/
+5
|
*
cli: remove --backend, use --mlx and --cuda consistently
Onur Gündüz
2026-06-14
2
-66
/
+11
|
[prev]
[next]