[Misc] Remove deprecated calculate_kv_scales runtime KV scale calculation (#49389)

Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>
This commit is contained in:
wangxiyuan
2026-08-03 11:15:13 -04:00
committed by GitHub
parent 0cf49a5d15
commit dd11df04f3
19 changed files with 16 additions and 263 deletions
@@ -52,9 +52,6 @@ def test_per_token_head_kv_cache_accuracy(
) -> None:
"""Compare logprobs between bf16 baseline and per-token-head quantized KV
cache.
Uses calculate_kv_scales (dynamic scale computation) since there are
no per-token-head calibrated checkpoints available yet.
"""
with monkeypatch.context() as m:
m.setenv("TOKENIZERS_PARALLELISM", "true")
@@ -80,7 +77,6 @@ def test_per_token_head_kv_cache_accuracy(
tensor_parallel_size=tensor_parallel_size,
enforce_eager=enforce_eager,
kv_cache_dtype=kv_cache_dtype,
calculate_kv_scales=True,
attention_config={"backend": backend},
) as vllm_model:
test_outputs = vllm_model.generate_greedy_logprobs(