GLM 5.3 Flash not receiving arguments or stream ended
还没有人认领这个 Issue。
评估
调研方向
首先,在 Linux 上使用 Command Code 1.44.0,在报告的 80–100K 上下文范围附近复现使用 GLM 5.3 Flash 时的问题。对比报告的计划变更中反复出现的空 bash 参数和缺失的 finish_reason,并在工具调用收到有效参数且流正常结束、无需 auto-compact 时,将 issue 视为完成。
由索引模型根据 Issue 内容生成。
描述
Summary
I was trying to find alternative for dsv4f peak hours and I think GLM 5.3 flash is a good model
I was on Go model before, using 5.3 flash with no hassle, both in cmcd, opencode, pi, routed with 9router
Since my Go plan almost ended, I upgraded my plan to GOAT and started using GLM 5.3 Flash.
Weirdly, whenever the model almost reached 100K, It keep getting something like
Validation failed for tool "bash":
- command: must have required properties command
Received arguments:
{}
or
Stream ended without finish_reason
I've tried to ask dsv4f to fix this issue, most of the solution says to change the context window to somewhere between 80-90K. The solution sometimes working sometimes not that I reached annoyed state (and moved to dsv4f on peak hours 😭). I believe that this is neither the harness nor 9router's problem (also this is what ds4vf stated too). Just wondering why this happened when I switch plan and it was working fine before
Expected Behavior
GLM 5.3 Flash working fine like a few weeks ago
Actual Behavior
Weirdly, whenever the model almost reached 100K, It keep getting something like
Validation failed for tool "bash":
- command: must have required properties command
Received arguments:
{}
or
Stream ended without finish_reason
Steps to reproduce the issue
- Choose GLM 5.3 Flash as model
- Produce something with it that it almost reached 80-90K context
- The Received arguments error will continuously occurred if not auto compact
Command Code Version
1.44.0
Operating System
Linux
Terminal/IDE
Unknown
Shell
cmd.exe
Session file (optional)
No response
Fix prompt (optional)
No response
Additional context
No response
- 主要语言
- 没有语言数据
- 星标
- 4k
- 派生
- 350
- PR 合并指标
- 30 天内没有已合并 PR
贡献指南
这个仓库没有索引到贡献指南
从这里开始
- 先读完整个 Issue,再读项目的贡献指南。
- 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
- Fork 仓库,在一个分支上完成修改。
- 提交 Pull Request,并在描述里引用这个 Issue 编号。
CommandCodeAI/command-code 的其他 Issue
-
难度 2/5 1-3 小时 新手友好度 68/100
CommandCodeAI/command-code#855 ·
-
难度 2/5 1-3 小时 新手友好度 78/100
CommandCodeAI/command-code#841 · 1 条评论 ·
-
难度 2/5 1-3 小时 新手友好度 68/100
CommandCodeAI/command-code#655 · 1 条评论 ·
-
难度 2/5 1-3 小时 新手友好度 68/100
CommandCodeAI/command-code#608 ·
-
难度 3/5 1-2 天 新手友好度 70/100
CommandCodeAI/command-code#893 ·
查看 CommandCodeAI/command-code 的全部 Issue
相似的 Issue
-
enhancement
难度 2/5 1-3 小时 新手友好度 68/100
JuliusBrussee/caveman#1102 · 1 条评论 ·
-
难度 2/5 1-3 小时 新手友好度 78/100
use-agent-os/agent-os#3263 ·
-
[Bug]: context-limit error parsing has no pattern for llama.cpp's "context size (N tokens)" phrasing 未关闭area/compression area/local-models area/sessions comp/agent duplicate P2 sweeper:risk-session-state type/bug
难度 2/5 1-3 小时 新手友好度 82/100
NousResearch/hermes-agent#117793 · 1 条评论 ·
-
possible bug
难度 2/5 1-3 小时 新手友好度 88/100
Mintplex-Labs/anything-llm#6415 · 1 条评论 ·
-
难度 2/5 1-3 小时 新手友好度 82/100