Skip to content

[Klaud Cold] docs: deprecate DeepSeek-V4-Pro Single-turn 8k1k on 2026-09-08, keep AgentX / 弃用 DeepSeek-V4-Pro 单轮 8k1k（2026-09-08），保留 AgentX #2728

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
functionstackx merged 1 commit into from
Aug 25, 2026
Merged

[Klaud Cold] docs: deprecate DeepSeek-V4-Pro Single-turn 8k1k on 2026-09-08, keep AgentX / 弃用 DeepSeek-V4-Pro 单轮 8k1k（2026-09-08），保留 AgentX #2728

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 13 additions & 1 deletion MODELS.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -42,6 +42,18 @@ Speculative-decoding A/B retirements apply to each pair below. The spec-decode a

**Enacted on 2026-08-07** in [#2527](https://github.com/SemiAnalysisAI/InferenceX/pull/2527): 17 `kimik2.5` config keys were removed from the active master configs and archived under [`configs/deprecated/`](configs/deprecated/) as `nvidia-kimik2.5-8k1k-master.yaml` (10) and `amd-kimik2.5-8k1k-master.yaml` (7), and their 12 benchmark scripts were moved to the sibling `deprecated/` directories. `kimik2.5` now has **no active configuration in any master config** and is fully retired. The same PR archived `kimik2.5-int4-h100-vllm`, an agentic-coding key that #2493 left behind in `nvidia-master.yaml` after moving its script to `benchmarks/single_node/agentic/deprecated/`. It is now in `nvidia-kimik2.5-agentic-master.yaml` with its siblings. The SPEED-Bench acceptance-length script `benchmarks/single_node/speedbench/kimik2.5_fp4_b300_vllm.sh` is intentionally kept. Speedbench is driven by `speedbench-al.yml`, not the master configs, matching how #2493 treated MiniMax-M3.

### Tuesday, September 8, 2026

**Tuesday, September 8, 2026** is the last day for the **Single-turn 8k1k** scenario on **DeepSeek-V4-Pro 1.6T** (`dsv4`). The scenario is deprecated for this model after that date. **Agentic coding is unaffected and stays active for `dsv4`**, including its MTP and DSpark arms. The model is not retired: agentic coding becomes its only scenario, and it continues to run and publish.

| Model | Deprecated | Remains |
|---|---|---|
| DeepSeek-V4-Pro 1.6T (`dsv4`) | Single-turn 8k1k | Agentic coding, including the MTP and DSpark arms |
Comment on lines +47 to +51

@claude claude Bot Aug 25, 2026

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🔴 New Sept 8 2026 deprecation entry claims dsv4's agentic coding 'stays active... including its MTP and DSpark arms', but no dsv4 config key in configs/nvidia-master.yaml or configs/amd-master.yaml uses DSpark (grep for 'dspark' in either file returns zero dsv4 hits, unlike kimik3 which has multiple active -agentic-dspark- keys). The pre-existing spec-decode A/B table at MODELS.md lines 27-33 also lists only 'MTP' as dsv4's current agentic spec-decode arm, not DSpark; DSpark for dsv4 is only an 'agreed PoR' (line 123) and a SPEED-Bench golden-AL reference (golden_al_distribution/dsv4-pro-0813-dspark.yaml), not a running master-config arm.

Extended reasoning...

A reader trusts this new notice and expects published dsv4 DSpark agentic results after Sept 8 2026 deprecation of Single-turn 8k1k; no such active arm exists in either master config, so the claim is currently false and will stay false unless a real dsv4-agentic-dspark key is added. Same incorrect claim is duplicated in MODELS_zh.md lines 47 and 51.

Verification: normal — the new Sept 8 2026 entry introduces a false, self-contradictory claim not present on the base branch. MODELS.md:47 and the table row at MODELS.md:51 both state dsv4's agentic coding stays active "including the MTP and DSpark arms." No dsv4 config key uses DSpark: a grep for dsv4.*dspark/dspark.*dsv4 across configs/nvidia-master.yaml and configs/amd-master.yaml returns zero matches,


Rationale: `dsv4` carries the largest single-turn footprint in the repository. 45 active config keys use the 8k1k scenario, 32 in `configs/nvidia-master.yaml` and 13 in `configs/amd-master.yaml`, spanning H200, B200, B300, GB200, GB300, MI300X, MI325X, and MI355X across vLLM, SGLang, TensorRT-LLM, ATOM, Dynamo, and llm-d. That is a large share of every full sweep. AgentX trace replay is the scenario AI labs and the ML community ask about, and DeepSeek-V4-Pro's 19 agentic config keys are the part of `dsv4` that feeds the published North Star Pareto frontier. Retiring the fixed-sequence-length arm frees cluster hours for AgentX and for new frontier models such as Qwen3.8 2.4T without reducing what we publish for this model. Single-turn 8k1k stays active for the other models that still list it.

**Status: not yet enacted.** All 45 8k1k config keys still run. On enactment they are removed from the active master configs and archived under [`configs/deprecated/`](configs/deprecated/), with their benchmark scripts moved to the sibling `deprecated/` directories, matching how [#2493](https://github.com/SemiAnalysisAI/InferenceX/pull/2493) and [#2527](https://github.com/SemiAnalysisAI/InferenceX/pull/2527) were carried out. The SPEED-Bench acceptance-length scripts for `dsv4` are intentionally kept. Speedbench is driven by `speedbench-al.yml`, not the master configs.

## Scenarios

| Scenario | ISL/OSL | Status |
Expand Down Expand Up @@ -141,7 +153,7 @@ Other offloading tiers, including NVMe KV cache offloading, are outside the init
| Kimi-K3 | `kimik3` | 2026-07-27 ([#2391](https://github.com/SemiAnalysisAI/InferenceX/pull/2391)) | Agentic coding (DSpark only) | Agentic coding non-DSpark arm (deprecated from day 0) |
| GLM-5.2 | `glm5.2` | 2026-07-18 ([#2268](https://github.com/SemiAnalysisAI/InferenceX/pull/2268)) | Agentic coding (the non-MTP arm still runs while the MTP-only transition remains pending, as explained in the Deprecation Notice) | |
| MiniMax-M3 | `minimaxm3` | 2026-06-12 ([#1724](https://github.com/SemiAnalysisAI/InferenceX/pull/1724)) | Agentic coding | Single-turn 1k1k, Single-turn 8k1k (removed 2026-08-04, [#2493](https://github.com/SemiAnalysisAI/InferenceX/pull/2493)) |
| DeepSeek-V4-Pro | `dsv4` | 2026-04-24 ([#1130](https://github.com/SemiAnalysisAI/InferenceX/pull/1130)) | Single-turn 8k1k, Agentic coding (the non-MTP arm still runs while the MTP-only transition remains pending, as explained in the Deprecation Notice) | Single-turn 1k1k |
| DeepSeek-V4-Pro | `dsv4` | 2026-04-24 ([#1130](https://github.com/SemiAnalysisAI/InferenceX/pull/1130)) | Single-turn 8k1k (last day 2026-09-08, see [Deprecation Notice](#deprecation-notice)), Agentic coding (the non-MTP arm still runs while the MTP-only transition remains pending, as explained in the Deprecation Notice) | Single-turn 1k1k |
| GLM-5 / GLM-5.1 | `glm5`, `glm5.1` | 2026-03-06 ([#762](https://github.com/SemiAnalysisAI/InferenceX/pull/762)), with GLM-5.1 added 2026-04-21 ([#1098](https://github.com/SemiAnalysisAI/InferenceX/pull/1098)) | None (retired 2026-07-18, [#2276](https://github.com/SemiAnalysisAI/InferenceX/pull/2276)) | Single-turn 1k1k, Single-turn 1k8k (GLM-5 only), Single-turn 8k1k |
| MiniMax-M2.5/2.7 | `minimaxm2.5` | 2026-02-18 ([#755](https://github.com/SemiAnalysisAI/InferenceX/pull/755)) | None (retired 2026-06-20, [#1874](https://github.com/SemiAnalysisAI/InferenceX/pull/1874)) | Single-turn 1k1k, Single-turn 1k8k, Single-turn 8k1k |
| Kimi-K2.5/2.6/2.7-Code | `kimik2.5` | 2026-02-17 ([#734](https://github.com/SemiAnalysisAI/InferenceX/pull/734)) | None (fully retired 2026-08-07, [#2527](https://github.com/SemiAnalysisAI/InferenceX/pull/2527)) | Single-turn 1k1k, Single-turn 1k8k, Agentic coding (removed 2026-08-04, [#2493](https://github.com/SemiAnalysisAI/InferenceX/pull/2493)), Single-turn 8k1k (removed 2026-08-07, [#2527](https://github.com/SemiAnalysisAI/InferenceX/pull/2527)) |
Expand Down
12 changes: 12 additions & 0 deletions MODELS_zh.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -42,6 +42,18 @@ InferenceX-e2e 运行在数量固定且有限的 GPU 资源池上，并由一支

**已于 2026-08-07 执行**（[#2527](https://github.com/SemiAnalysisAI/InferenceX/pull/2527)）：从启用的主配置中移除 17 个 `kimik2.5` 配置项，归档至 [`configs/deprecated/`](configs/deprecated/)，分别为 `nvidia-kimik2.5-8k1k-master.yaml`（10 个）与 `amd-kimik2.5-8k1k-master.yaml`（7 个）。对应的 12 个基准测试脚本移入同级 `deprecated/` 目录。此后 `kimik2.5` 在所有主配置中**均无启用配置**，正式完全退役。同一 PR 还归档了 `kimik2.5-int4-h100-vllm`。#2493 将其脚本移入 `benchmarks/single_node/agentic/deprecated/` 时，该智能体编码配置项被遗留在 `nvidia-master.yaml` 中，现已与同类项一并归入 `nvidia-kimik2.5-agentic-master.yaml`。SPEED-Bench 接受长度脚本 `benchmarks/single_node/speedbench/kimik2.5_fp4_b300_vllm.sh` 予以保留。Speedbench 由 `speedbench-al.yml` 驱动，不经过主配置，与 #2493 处理 MiniMax-M3 的方式一致。

### 2026 年 9 月 8 日（星期二）

**2026 年 9 月 8 日（星期二）**为 **DeepSeek-V4-Pro 1.6T**（`dsv4`）**单轮 8k1k** 场景的最后运行日，此后该场景对该模型弃用。**智能体编码不受影响，`dsv4` 的该场景继续启用**，其 MTP 与 DSpark 分支均予保留。该模型不会退役：智能体编码将成为其唯一场景，并继续运行与发布。

| 模型 | 弃用内容 | 保留内容 |
|---|---|---|
| DeepSeek-V4-Pro 1.6T（`dsv4`） | 单轮 8k1k | 智能体编码，含 MTP 与 DSpark 分支 |

原因：`dsv4` 是本仓库中单轮场景占用最大的模型。当前有 45 个启用的配置项使用 8k1k 场景（`configs/nvidia-master.yaml` 32 个，`configs/amd-master.yaml` 13 个），覆盖 H200、B200、B300、GB200、GB300、MI300X、MI325X 与 MI355X，涉及 vLLM、SGLang、TensorRT-LLM、ATOM、Dynamo 与 llm-d，在每一轮完整 sweep 中占比可观。AgentX 轨迹回放才是 AI 实验室与 ML 社区真正关注的场景，而 DeepSeek-V4-Pro 的 19 个智能体编码配置项正是 `dsv4` 中支撑已发布北极星（North Star）帕累托前沿的部分。下线固定序列长度分支可为 AgentX 以及 Qwen3.8 2.4T 等新前沿模型腾出集群机时，同时不减少该模型对外发布的内容。对于仍列有该场景的其他模型，单轮 8k1k 保持启用。

**状态：尚未执行。** 全部 45 个 8k1k 配置项仍在运行。执行时将从启用的主配置中移除并归档至 [`configs/deprecated/`](configs/deprecated/)，对应基准测试脚本移入同级 `deprecated/` 目录，与 [#2493](https://github.com/SemiAnalysisAI/InferenceX/pull/2493) 和 [#2527](https://github.com/SemiAnalysisAI/InferenceX/pull/2527) 的做法一致。`dsv4` 的 SPEED-Bench 接受长度脚本予以保留。Speedbench 由 `speedbench-al.yml` 驱动，不经过主配置。

## 场景

| 场景 | ISL/OSL | 状态 |
Expand Down