sgl-project/sglang · error · ValueError

--linear-replayssm-cache-len must be >= 1, got {cfg.linear_r

Error message

--linear-replayssm-cache-len must be >= 1, got {cfg.linear_replayssm_cache_len}.

What it means

Validates that --linear-replayssm-cache-len is at least 1. The cache length sizes the ReplaySSM ring window (e.g. KDA keeps a cache-len window and folds via its fused verify ring-write), so zero or negative values are meaningless and rejected.

Source

Thrown at python/sglang/srt/server_args.py:6962

            if mamba_extra_buffer_of(resolved_view(self)):
                raise ValueError(
                    "--enable-linear-replayssm requires --mamba-radix-cache-strategy "
                    "no_buffer (the default); the extra_buffer ping-pong "
                    "donation path is not yet supported (follow-up). Got "
                    f"--mamba-radix-cache-strategy={cfg.mamba_radix_cache_strategy!r}."
                )
            if cfg.disaggregation_mode != "null":
                # The disaggregated decode pool (HybridMambaDecodeReqToTokenPool)
                # is not wired for the ReplaySSM ring, so the flag would silently
                # no-op there; disagg also runs a different cache/coordination
                # flow that is not yet validated for ReplaySSM (follow-up).
                raise ValueError(
                    "--enable-linear-replayssm is not supported under PD "
                    "disaggregation yet (follow-up). Got "
                    f"--disaggregation-mode={cfg.disaggregation_mode!r}."
                )
            if cfg.linear_replayssm_cache_len < 1:
                raise ValueError(
                    "--linear-replayssm-cache-len must be >= 1, got "
                    f"{cfg.linear_replayssm_cache_len}."
                )

        # ReplaySSM spec-verify (Part B of #28511): linear-chain target verify via
        # fold-every-commit -- the verify stores each draft step's raw inputs into
        # the per-slot (rawv, rawk, g, beta) window and the commit replays the
        # accepted prefix into the fp32 checkpoint. The intra-window interaction
        # uses a strictly-lower causal mask, so it is valid ONLY for a linear
        # draft chain (speculative_eagle_topk in {None, 1}, i.e. NEXTN / MTP);
        # EAGLE tree verify (topk > 1) must fall back to the recurrent verify.
        # GDN sizes the window to the draft maximum; KDA (kda_backend) keeps a
        # --linear-replayssm-cache-len window and folds via its own fused
        # verify ring-write + commit_kda_replayssm_after_verify.
        if cfg.enable_linear_replayssm_spec:
            if cfg.speculative_eagle_topk not in (None, 1):
                raise ValueError(
                    "--enable-linear-replayssm-spec requires a linear draft chain "

View on GitHub (pinned to 0132848349)

Solutions

1. Set --linear-replayssm-cache-len to a positive integer (e.g. 128)
2. If you intended the default, omit the flag entirely
3. Check any script computing this value for off-by-one/empty-input bugs

Example fix

# before
--enable-linear-replayssm --linear-replayssm-cache-len 0
# after
--enable-linear-replayssm --linear-replayssm-cache-len 128

def validate(cache_len, enable_replayssm):
    return not enable_replayssm or (isinstance(cache_len, int) and cache_len >= 1)

def valid_cache_len(n) -> bool: return isinstance(n, int) and n >= 1

When it happens

Trigger: Passing --linear-replayssm-cache-len 0 (or a negative number) together with --enable-linear-replayssm; e.g. someone setting 0 intending 'unlimited/default'.

Common situations: Treating 0 as 'auto'/'disabled' (common convention elsewhere); scripting loops that compute cache length and can emit 0; typos in launch scripts.

Understand the failure class

Background: "Must be a positive integer", "Invalid value", "Unsupported": the invalid-argument-value error family, when a library rejects the value you pass — this error's family across 35 libraries.

Related errors

• --speculative-draft-window-size must be positive, got {}.
• routed_dp_rank={routed_dp_rank} out of range [0, {dp_size})
• item_first is not supported when embeddings are supplied
• --dcp-replicate-q-proj requires --dcp-size > 1.
• --dcp-replicate-q-proj only applies to the a2a/fi_a2a DCP co

AI-assisted analysis of sgl-project/sglang@0132848349 (2026-08-28). Data as JSON: /api/errors/74a227be7c0cce26. Report an issue: GitHub.