biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Feature] Support sequence parallelism for static fp8 quantization (#19181)

Browse Source 
Signed-off-by: cascade812 <cascade812@outlook.com>
This commit is contained in:
cascade
2025-06-23 13:09:02 -07:00
committed by GitHub
parent d0132f025d
commit e6327c9b3e
7 changed files with 531 additions and 195 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
tests/models/registry.py
View File

@@ -193,7 +193,8 @@ _TEXT_GENERATION_EXAMPLE_MODELS = {
extras={"tiny": "ai21labs/Jamba-tiny-dev"}), # noqa: E501
"LlamaForCausalLM": _HfExamplesInfo("meta-llama/Llama-3.2-1B-Instruct",
extras={"guard": "meta-llama/Llama-Guard-3-1B", # noqa: E501
"hermes": "NousResearch/Hermes-3-Llama-3.1-8B"}), # noqa: E501
"hermes": "NousResearch/Hermes-3-Llama-3.1-8B", # noqa: E501
"fp8": "RedHatAI/Meta-Llama-3.1-8B-Instruct-FP8"}), # noqa: E501
"LLaMAForCausalLM": _HfExamplesInfo("decapoda-research/llama-7b-hf",
is_available_online=False),
"MambaForCausalLM": _HfExamplesInfo("state-spaces/mamba-130m-hf"),