Standalone speculative decoding (#10090)
This commit is contained in:
@@ -72,6 +72,10 @@ DEFAULT_MODEL_NAME_FOR_TEST_W8A8_WITH_MOE = "nytopop/Qwen3-30B-A3B.w8a8"
|
||||
DEFAULT_EAGLE_TARGET_MODEL_FOR_TEST = "meta-llama/Llama-2-7b-chat-hf"
|
||||
DEFAULT_EAGLE_DRAFT_MODEL_FOR_TEST = "lmsys/sglang-EAGLE-llama2-chat-7B"
|
||||
DEFAULT_MODEL_NAME_FOR_TEST_EAGLE3 = "jamesliu1/sglang-EAGLE3-Llama-3.1-Instruct-8B"
|
||||
DEFAULT_STANDALONE_SPECULATIVE_TARGET_MODEL_FOR_TEST = (
|
||||
"meta-llama/Llama-3.1-8B-Instruct"
|
||||
)
|
||||
DEFAULT_STANDALONE_SPECULATIVE_DRAFT_MODEL_FOR_TEST = "meta-llama/Llama-3.2-1B-Instruct"
|
||||
|
||||
# Other use cases
|
||||
DEFAULT_MODEL_NAME_FOR_TEST_LOCAL_ATTENTION = (
|
||||
|
||||
Reference in New Issue
Block a user