[Chore] Use upstream Transformers DeepseekV4Config (#61392)

Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
Co-authored-by: Kevin H. Luu <khluu000@gmail.com>
This commit is contained in:
Harry Mellor
2026-10-15 01:08:08 -07:00
committed by GitHub
co-authored by Claude Opus 5.5 mergify[bot] <37929162+mergify[bot]@users.noreply.github.com> Kevin H. Luu
parent 1c4c319792
commit b965ee564d
31 changed files with 155 additions and 120 deletions
+2 -2
View File
@@ -7,12 +7,12 @@ from types import SimpleNamespace
import lm_eval
import pytest
import torch
from transformers import DeepseekV4Config
from tests.utils import large_gpu_mark
from vllm.distributed.eplb.eplb_state import EplbState
from vllm.models.deepseek_v4.common.eplb_util import dspark_draft_supports_eplb
from vllm.platforms import current_platform
from vllm.transformers_utils.configs.deepseek_v4 import DeepseekV4Config
from vllm.v1.worker.gpu.eplb_utils import (
EPLBController,
draft_model_supports_eplb,
@@ -105,7 +105,7 @@ def _make_dsv4_dspark_hf_config() -> DeepseekV4Config:
vocab_size=256,
n_mtp_layers=2,
enable_confidence_head=False,
compress_ratios=[1, 1],
layer_types=["sliding_attention"] * 2,
)