| .. |
|
async_aphrodite
|
055c8905a3
api: add sampling/engine option to return only deltas or final output (#1035)
|
1 개월 전 |
|
basic_correctness
|
7b6501bd05
tests: refactor model tests (#1078)
|
1 개월 전 |
|
benchmarks
|
86bf2cc4f3
core: rename `PromptInputs,inputs` -> `PromptType,prompt` (#1080)
|
1 개월 전 |
|
compile
|
239a8cae25
torch.compile: register all-reduce operations as custom ops (#1050)
|
1 개월 전 |
|
core
|
be0b0c13ca
tests: update scheduler tests (#1113)
|
2 주 전 |
|
distributed
|
c90abcc603
VLM: add pipeline parallelism support for Qwen2-VL (#1103)
|
2 주 전 |
|
encoder_decoder
|
a985143768
core: add cuda graph support for encoder-decoder models (#1051)
|
1 개월 전 |
|
endpoints
|
3e6addcc2c
LLM: enable batched inference for llm.chat() API (#1120)
|
1 주 전 |
|
engine
|
cc5e185795
VLM: support passing multimodal processor kwargs (#1102)
|
2 주 전 |
|
kernels
|
172dee2573
(2/N) Triton Backend: integrate Triton activation kernels (#1126)
|
6 일 전 |
|
lora
|
69cf654901
LoRA: add assertions for SGMV kernels to avoid incorrect results (#1104)
|
2 주 전 |
|
metrics
|
c6c91edab7
ci: update & overhaul test units (#769)
|
2 달 전 |
|
modeling
|
c6c91edab7
ci: update & overhaul test units (#769)
|
2 달 전 |
|
models
|
80be38ca6f
chore: expose phi3_v num_crops as an mm_processor_kwargs (#1117)
|
2 주 전 |
|
mq_aphrodite_engine
|
86bf2cc4f3
core: rename `PromptInputs,inputs` -> `PromptType,prompt` (#1080)
|
1 개월 전 |
|
multi_step
|
58aff3771d
core: support prompt logprobs in multi-step (#1060)
|
1 개월 전 |
|
multimodal
|
cc5e185795
VLM: support passing multimodal processor kwargs (#1102)
|
2 주 전 |
|
plugins
|
c6c91edab7
ci: update & overhaul test units (#769)
|
2 달 전 |
|
prefix_caching
|
6212072245
api: support LoRA lineage and base model metadata management (#1072)
|
1 개월 전 |
|
prompt_adapter
|
c6c91edab7
ci: update & overhaul test units (#769)
|
2 달 전 |
|
prompts
|
e1f3fd1e02
fix: test units (#201)
|
1 년 전 |
|
quantization
|
349a612338
chore: bump bitsandbytes version to latest; enable cuda graphs for 4bit bnb (#1123)
|
1 주 전 |
|
samplers
|
eb1ffacf74
Spec Decoding: fix typical acceptance sampler with correct recovered tok IDs (#1106)
|
2 주 전 |
|
spec_decode
|
1fac86c325
core: factor out common code in SequenceData (#1083)
|
1 개월 전 |
|
tensorizer_loader
|
673621a3d2
xpu: refactor the model runner for tensor parallelism (#910)
|
1 개월 전 |
|
tokenization
|
c6c91edab7
ci: update & overhaul test units (#769)
|
2 달 전 |
|
tool_use
|
0191c5efd1
tools: fix tool calls to more strictly follow OpenAI format (#1003)
|
1 개월 전 |
|
tpu
|
ea59784f59
tpu: remove torch._dynamo.reset() (#952)
|
1 개월 전 |
|
weight_loading
|
c6c91edab7
ci: update & overhaul test units (#769)
|
2 달 전 |
|
worker
|
1fac86c325
core: factor out common code in SequenceData (#1083)
|
1 개월 전 |
|
__init__.py
|
2755a48d51
merge dev branch into main (#153)
|
1 년 전 |
|
conftest.py
|
7b6501bd05
tests: refactor model tests (#1078)
|
1 개월 전 |
|
test_cache_block_hashing.py
|
c6c91edab7
ci: update & overhaul test units (#769)
|
2 달 전 |
|
test_config.py
|
c6c91edab7
ci: update & overhaul test units (#769)
|
2 달 전 |
|
test_embedded_commit.py
|
c6c91edab7
ci: update & overhaul test units (#769)
|
2 달 전 |
|
test_inputs.py
|
c6c91edab7
ci: update & overhaul test units (#769)
|
2 달 전 |
|
test_logits_processor.py
|
1fac86c325
core: factor out common code in SequenceData (#1083)
|
1 개월 전 |
|
test_regression.py
|
c6c91edab7
ci: update & overhaul test units (#769)
|
2 달 전 |
|
test_sampling_params.py
|
c6c91edab7
ci: update & overhaul test units (#769)
|
2 달 전 |
|
test_scalartype.py
|
c6c91edab7
ci: update & overhaul test units (#769)
|
2 달 전 |
|
test_sequence.py
|
1fac86c325
core: factor out common code in SequenceData (#1083)
|
1 개월 전 |
|
test_sharded_state_loader.py
|
c6c91edab7
ci: update & overhaul test units (#769)
|
2 달 전 |
|
test_utils.py
|
c6c91edab7
ci: update & overhaul test units (#769)
|
2 달 전 |
|
utils.py
|
5b03d67abb
Core: add output streaming support to multi-step + async (#1112)
|
2 주 전 |
AlpinDale
172dee2573
(2/N) Triton Backend: integrate Triton activation kernels (#1126)