Add Laguna XS.2.1 DFlash drafter support (#46853)
Signed-off-by: Adam Baranowski <adam.baranowski@poolside.ai>
This commit is contained in:
@@ -1410,6 +1410,13 @@ _SPECULATIVE_DECODING_EXAMPLE_MODELS = {
|
||||
max_model_len=8192, # Reduce max len to ensure test runs in low-VRAM CI env
|
||||
max_num_seqs=32,
|
||||
),
|
||||
"DFlashLagunaForCausalLM": _HfExamplesInfo(
|
||||
"poolside/Laguna-XS-2.1",
|
||||
speculative_model="poolside/Laguna-XS-2.1-DFlash",
|
||||
use_original_num_layers=True,
|
||||
max_model_len=8192, # Reduce max len to ensure test runs in low-VRAM CI env
|
||||
max_num_seqs=32,
|
||||
),
|
||||
"DFlashQwen3NextDraftModel": _HfExamplesInfo(
|
||||
"Qwen/Qwen3-Coder-Next",
|
||||
speculative_model="z-lab/Qwen3-Coder-Next-DFlash",
|
||||
|
||||
Reference in New Issue
Block a user