AlpinDale 1efd0f89b7 feat: support FP8 for DeepSeekV2 MoE преди 6 месеца
..
adapter_commons 99680b2d23 feat: soft prompts (#589) преди 6 месеца
attention d8f9f0ec16 fix: prefix prefill kernels for fp32 data type преди 6 месеца
common bf15e1b4e8 chore: deprecation warning for beam search преди 6 месеца
distributed cc6399792f fix: keep consistent with how pytorch finds libcudart.so преди 6 месеца
endpoints a3b56353fa fix: another one missed преди 6 месеца
engine 0c17c2a8a7 chore: add commit hash, clean up engine logs преди 6 месеца
executor 23408b9b2b chore: skip the driver worker преди 6 месеца
inputs 4f7d212b70 feat: remove vision language config преди 6 месеца
kv_quant e42a78381a feat: switch from pylint to ruff (#322) преди 1 година
lora 99680b2d23 feat: soft prompts (#589) преди 6 месеца
modeling 1efd0f89b7 feat: support FP8 for DeepSeekV2 MoE преди 6 месеца
multimodal c11a8bdaad fix: calculate max number of multi-modal tokens automatically преди 6 месеца
platforms 1a40bf438b fix: incorrect gpu capability when used mixed gpus преди 6 месеца
processing 99680b2d23 feat: soft prompts (#589) преди 6 месеца
prompt_adapter 99680b2d23 feat: soft prompts (#589) преди 6 месеца
quantization 1efd0f89b7 feat: support FP8 for DeepSeekV2 MoE преди 6 месеца
spec_decode 16dff9babc chore: enable bonus token in spec decoding for KV cache based models преди 6 месеца
task_handler ddb28a80a3 fix: bump torch for rocm, unify CUDA_VISIBLE_DEVICES for cuda and rocm преди 6 месеца
transformers_utils 63becc67c0 fix: prompt logprob detokenization преди 6 месеца
__init__.py 0c17c2a8a7 chore: add commit hash, clean up engine logs преди 6 месеца
_custom_ops.py ad24e74a99 feat: FP8 weight-only quantization support for Ampere GPUs преди 6 месеца
_ipex_ops.py 6a57861fca feat: initial XPU support via intel_extension_for_pytorch (#571) преди 7 месеца
py.typed 1c988a48b2 fix logging and add py.typed преди 1 година
version.py 0c17c2a8a7 chore: add commit hash, clean up engine logs преди 6 месеца