huggingface/text-generation-inference

Pull Requests Commits

fix: ensure latest requirements exported

drbh committed 2 years ago

fc86dba7

feat: bump lock and requirement files

drbh committed 2 years ago

400a3c68

GPTQ support on ROCm (#1489)

fxmarty committed 2 years ago

Verified 650fea18

Update the docs to include newer models. (#1492)

Narsil committed 2 years ago

Verified ebecc061

fix: launcher doc typos (#1462)

thelinuxkid committed 2 years ago

Verified 50a20a83

Trying to fix that flaky test. (#1491)

Narsil committed 2 years ago

Verified 4c7315dd

Add sealion mpt support (#1477)

Narsil committed 2 years ago

Verified ac499727

Reinstate exl2 with tp (#1490)

Narsil committed 2 years ago

Verified b9573218

fix: launcher doc typos (#1473)

Narsil committed 2 years ago

Verified 16958fe3

fix: show warning with tokenizer config parsing error (#1488)

drbh committed 2 years ago

Verified 13dd8e23

Update the docs

Narsil committed 2 years ago

17b7b75e

fix: read stderr in download (#1486)

OlivierDehaene committed 2 years ago

Verified 9c320e26

feat: adds phi model (#1442)

drbh committed 2 years ago

Verified 7e2a7433

Add a new `/tokenize` route to get the tokenized input (#1471)

Narsil committed 2 years ago

Verified 86c8335f

Add messages api compatibility docs (#1478)

drbh committed 2 years ago

Verified 7872b8c5

Fixing non divisible embeddings. (#1476)

Narsil committed 2 years ago

Verified 7e542d4d

Disable `decoder_input_details` on OpenAI-compatible chat streaming, pass temp and top-k from API (#1470)

EndlessReform committed 2 years ago

Verified 82f87ada

feat: conditionally toggle chat on invocations route (#1454)

drbh committed 2 years ago

Verified 98e5faff

chore: bump rust version and annotate/fix all clippy warnings (#1455)

drbh committed 2 years ago

Verified becd0997

feat: support raise_exception, bos and eos tokens (#1450)

drbh committed 2 years ago

Verified 3ccb3bb0

feat: supports openai chat completions API (#1427)

drbh committed 2 years ago

Verified 0eabc835

Return prompt vs generated tokens. (#1436)

Narsil committed 2 years ago

Verified ac08b4ef

Fix local load for Medusa (#1420)

PYNing committed 2 years ago

Verified da27fbdf

fix: follow base model for tokenizer in router (#1424)

OlivierDehaene committed 2 years ago

Verified fbeb1c44

Fix missing make target platform for local install: 'install-flash-attention-v2' (#1414)

deepily committed 2 years ago

Verified 91d72675

fix: fix local loading for .bin models (#1419)

OlivierDehaene committed 2 years ago

Verified 564f2a3b

docs: update required CUDA version to 12.2

OlivierDehaene committed 2 years ago

3f9b3f45

OlivierDehaene committed 2 years ago

630800ee

Fix local load for peft (#1373)

Narsil committed 2 years ago

Verified 529d7c25

feat: update exllamav2 kernels (#1370)

OlivierDehaene committed 2 years ago

Verified 564199ba

Older