可进入对比
档案摘要 / Decision snapshot
先判断能不能用于选型。
有真实任务证据
来自模型卡
边界优先
30 秒结论
可进入选型对比,但仍以案例证据为准。
来自公开模型资料库,并已补厂商/官方/模型家族证据;当前站内暂无可核验 A 类案例。
这页有来源、有风险说明、有公开可核验 A 类案例。
数据可信度 / 证据强度
强
A 类案例
5
AA 评分
9.55
官方来源
5 个补充入口
案例来源
5 条可核验
来自公开资料与模型数据库 + Artificial Analysis + 5 个厂商/官方/模型家族证据入口;AA 不是唯一事实来源。
Evidence Distribution
案例证据分布
Qwen1.5 Chat 110B 当前关联 5 条 A 类案例;这里按任务、来源和复核时间观察证据结构。
Evidence Snapshot
证据快照覆盖
Qwen1.5 Chat 110B 的 5 条 A 类案例中,已有 5 条完成本地证据快照;0 条仍在快照队列。
可用 manifest 复核
等待落盘
需补抓
需人工复核
证据完整度
5 / 5
Feishu Bitable
5 个入口
5 条
已绑定
已标注
能力边界
用标签表达信号,不用标签替代证据。
通用/待核验
暂无数据
待核验
开放/混合信号
中
| 厂商 | Qwen / Alibaba |
|---|---|
| 厂商页 | Qwen / Alibaba |
| 发布时间 | 2024-04-25 |
| 模型 ID | qwen1-5-chat-110b |
| 上下文 | 暂无数据 |
| 输出 | 暂无数据 |
| 模态 | 文本 / 多模态待核验 |
| 推理 | 否或官方未披露 |
| 价格 | 官方未披露 / 暂无数据 |
| 可用平台 | Qwen / Alibaba |
| AA 评分 | 9.55 |
| 基础模型发布时间 | Artificial Analysis 未披露 |
| 资料来源 | 公开资料、厂商信息和案例库 |
| A 类案例 | 5 |
谱系位置
它在厂商路线中的位置
此页将 Qwen1.5 Chat 110B 放入 Qwen / Alibaba 的当前 Atlas 路线中。若同厂前后代资料不足,先保留为路线追踪入口,不脑补谱系关系。
2024-04-25
Qwen1.5 Chat 110B 资料状态
来自公开资料与模型数据库 + Artificial Analysis + 5 个厂商/官方/模型家族证据入口;AA 不是唯一事实来源。
公开档案
模型状态
Qwen1.5 Chat 110B 当前标记为已有真实案例。
案例库
公开案例复核
已有 5 条可核验 A 类案例,达到完整补齐线
案例时间线
最近进入档案的 A 类案例
用采集时间呈现证据进入 Atlas 的顺序,方便复核来源新鲜度。
2026-06-27T07:46:08Z
DB-GPT integrates Qwen1.5-110B-Chat as a supported LLM backend
eosphoros-ai / DB-GPT / application_integration
2026-06-27T07:46:08Z
LMDeploy user serves Qwen1.5-110B-Chat as an OpenAI-compatible chat API on A800 GPUs
InternLM/lmdeploy GitHub user community / model_serving
2026-06-27T07:46:08Z
vLLM user runs Qwen1.5-110B-Chat-GPTQ-Int4 inference on A100 and measures request latency
vllm-project/vllm GitHub user community / batch_inference
2026-06-27T07:46:08Z
LLaMA-Factory user attempts SFT fine-tuning of Qwen1.5-110B-Chat with DeepSpeed ZeRO-3 on 4×A40
hiyouga/LLaMA-Factory GitHub user community / fine_tuning
2026-06-27T07:46:08Z
Xinference user launches replicated Qwen1.5-110B-Chat-GPTQ-Int4 serving on 8×4090 GPUs
xorbitsai/inference GitHub user community / model_serving
适合场景
- - 作为完整模型索引和厂商路线追踪入口
- - 通用模型对比
不适合
- - 缺少真实案例时不要包装成推荐
- - 缺字段时不要脑补价格、性能或上下文
真实案例
可核验 A 类案例精选
当前展示 5 条代表案例;排序优先看模型卡精选、证据可信度、快照状态和公开仓库/技术任务信号。案例总数仍为 5。
Representative Evidence
代表案例排序
当前第一代表案例是“DB-GPT integrates Qwen1.5-110B-Chat as a supported LLM backend”。它会优先出现在本页,是因为该案例同时进入模型卡精选、具备 100/100 的证据可信度,并带有 已快照证据。
模型卡精选优先
A+ 完整链路
archive/evidence/qwen1-5-chat-110b-db-gpt-integration-1474/manifest.json
Coding / repo / agent
eosphoros-ai / DB-GPT 使用 Qwen1.5 Chat 110B 处理代码审查和测试生成
eosphoros-ai / DB-GPT · Qwen1.5 Chat 110B
A+ 完整链路 · 代码仓库证据
eosphoros-ai / DB-GPT 公开的代码审查与测试案例,来源为 公开代码库,复核于 2026-06-27T07:46:08Z。 Model Atlas 将它标记为 A 类证据,因为它同时具备具体使用者、具体任务、公开原始证据和可访问产物。
任务:这是一个围绕代码审查和测试生成的真实任务,公开材料可以回溯到具体使用者和具体产物。 原始资料写作:DB-GPT maintainers added and tested Qwen1.5-110B-Chat support in the DB-GPT framework, with the PR explicitly testing LLM_MODEL=qwen1.5-110b-chat and attaching a working snapshot.
公开产物:公开材料提供公开代码、README 或项目配置,可用于核验任务结果、项目形态和模型绑定关系。 原始资料写作:The PR added Qwen110B support to DB-GPT and documented testing with LLM_MODEL=qwen1.5-110b-chat, making the model selectable as a backend for DB-GPT workflows.
模型作用:Qwen1.5 Chat 110B 在该案例中承担代码审查和测试生成相关的生成、分析、编排或实现角色。 原始资料写作:Qwen1.5-110B-Chat served as the large chat model backend DB-GPT integrated for conversational/database-oriented LLM application workflows.
A 类理由:有具体使用者、具体任务、公开原始证据和可访问产物,可公开核验。
风险边界:当前判断基于公开材料;若产物下线、仓库变更或模型参与比例仅来自作者自述,需要在引用前重新复核。 原始资料写作:Evidence is a product integration PR rather than an end-customer story; still includes concrete model identifier, maintainer organization, tested configuration, and public artifact.
原始记录:DB-GPT integrates Qwen1.5-110B-Chat as a supported LLM backend
hiyouga/LLaMA-Factory GitHub user c… 使用 Qwen1.5 Chat 110B 处理真实任务执行
hiyouga/LLaMA-Factory GitHub user community · Qwen1.5 Chat 110B
A+ 完整链路 · 代码仓库证据
hiyouga/LLaMA-Factory GitHub user community 公开的真实任务执行案例,来源为 公开代码库,复核于 2026-06-27T07:46:08Z。 Model Atlas 将它标记为 A 类证据,因为它同时具备具体使用者、具体任务、公开原始证据和可访问产物。
任务:这是一个围绕真实任务执行的真实任务,公开材料可以回溯到具体使用者和具体产物。 原始资料写作:A user configured LLaMA-Factory SFT with DeepSpeed ZeRO-3 offload to fine-tune ../../model/qwen/Qwen1.5-110B-Chat on a custom dataset named Kee_Instruction_NewEstabalish, using LoRA rank 128, cutoff_len 6000, and 4 A40 …
公开产物:公开材料提供公开代码、README 或项目配置,可用于核验任务结果、项目形态和模型绑定关系。 原始资料写作:The public issue includes the training command, dataset/configuration, expected goal of using 4×A40 plus DeepSpeed ZeRO-3 to fine-tune Qwen1.5-110B, and the encountered device placement runtime error.
模型作用:Qwen1.5 Chat 110B 在该案例中承担真实任务执行相关的生成、分析、编排或实现角色。 原始资料写作:Qwen1.5-110B-Chat was the base instruction/chat model being adapted through SFT/LoRA for the user's domain dataset.
A 类理由:有具体使用者、具体任务、公开原始证据和可访问产物,可公开核验。
风险边界:当前判断基于公开材料;若产物下线、仓库变更或模型参与比例仅来自作者自述,需要在引用前重新复核。 原始资料写作:This is a fine-tuning troubleshooting artifact, not a successful public demo; however it documents a concrete organization/user workflow with exact model path, task, dataset, and training setup.
原始记录:LLaMA-Factory user attempts SFT fine-tuning of Qwen1.5-110B-Chat with DeepSpeed ZeRO-3 on 4×A40
InternLM/lmdeploy GitHub user commu… 使用 Qwen1.5 Chat 110B 处理真实任务执行
InternLM/lmdeploy GitHub user community · Qwen1.5 Chat 110B
A+ 完整链路 · 代码仓库证据
InternLM/lmdeploy GitHub user community 公开的真实任务执行案例,来源为 公开代码库,复核于 2026-06-27T07:46:08Z。 Model Atlas 将它标记为 A 类证据,因为它同时具备具体使用者、具体任务、公开原始证据和可访问产物。
任务:这是一个围绕真实任务执行的真实任务,公开材料可以回溯到具体使用者和具体产物。 原始资料写作:A user deployed PATH/Qwen1.5-110B-Chat with lmdeploy serve api_server, model name qwen, tensor parallelism 8 or 4, then called /v1/chat/completions with a chat prompt.
公开产物:公开材料提供公开代码、README 或项目配置,可用于核验任务结果、项目形态和模型绑定关系。 原始资料写作:The issue records a concrete deployment attempt on four A800-SXM4-80GB GPUs: Qwen1.5-110B-Chat service started, but the chat completion request did not return promptly while the 72B service did.
模型作用:Qwen1.5 Chat 110B 在该案例中承担真实任务执行相关的生成、分析、编排或实现角色。 原始资料写作:Qwen1.5-110B-Chat was the hosted chat model behind the OpenAI-compatible API endpoint for answering user prompts.
A 类理由:有具体使用者、具体任务、公开原始证据和可访问产物,可公开核验。
风险边界:当前判断基于公开材料;若产物下线、仓库变更或模型参与比例仅来自作者自述,需要在引用前重新复核。 原始资料写作:This is a troubleshooting report with a negative serving outcome, not a polished showcase; it is still a real public deployment/use artifact with exact command, hardware, and model path.
原始记录:LMDeploy user serves Qwen1.5-110B-Chat as an OpenAI-compatible chat API on A800 GPUs
vllm-project/vllm GitHub user commu… 使用 Qwen1.5 Chat 110B 处理软件工程任务执行
vllm-project/vllm GitHub user community · Qwen1.5 Chat 110B
A+ 完整链路 · 代码仓库证据
vllm-project/vllm GitHub user community 公开的代码代理与软件工程案例,来源为 公开代码库,复核于 2026-06-27T07:46:08Z。 Model Atlas 将它标记为 A 类证据,因为它同时具备具体使用者、具体任务、公开原始证据和可访问产物。
任务:这是一个围绕软件工程任务执行的真实任务,公开材料可以回溯到具体使用者和具体产物。 原始资料写作:A user used vLLM 0.4.2 AsyncLLMEngine to run Qwen1.5-110B-Chat-GPTQ-Int4 from ModelScope on A100 80G, tuning gpu_memory_utilization, max_model_len, max_num_seqs, and sending 330-token prompts with short outputs.
公开产物:公开材料提供公开代码、README 或项目配置,可用于核验任务结果、项目形态和模型绑定关系。 原始资料写作:The report gives observed latency/QPS figures: about 0.45s per request at QPS=1, 0.70s at QPS=2, 1.4s at QPS=4, and 2.9s at QPS=8, with effective batch size around 2.
模型作用:Qwen1.5 Chat 110B 在该案例中承担软件工程任务执行相关的生成、分析、编排或实现角色。 原始资料写作:Qwen1.5-110B-Chat-GPTQ-Int4 was the inference model being served by vLLM for low-latency short-answer generation.
A 类理由:有具体使用者、具体任务、公开原始证据和可访问产物,可公开核验。
风险边界:当前判断基于公开材料;若产物下线、仓库变更或模型参与比例仅来自作者自述,需要在引用前重新复核。 原始资料写作:Evidence is a performance/usage issue rather than a customer case study; it includes exact model URL, serving framework, hardware, parameters, and measured outputs.
原始记录:vLLM user runs Qwen1.5-110B-Chat-GPTQ-Int4 inference on A100 and measures request latency
xorbitsai/inference GitHub user com… 使用 Qwen1.5 Chat 110B 处理真实任务执行
xorbitsai/inference GitHub user community · Qwen1.5 Chat 110B
A+ 完整链路 · 代码仓库证据
xorbitsai/inference GitHub user community 公开的真实任务执行案例,来源为 公开代码库,复核于 2026-06-27T07:46:08Z。 Model Atlas 将它标记为 A 类证据,因为它同时具备具体使用者、具体任务、公开原始证据和可访问产物。
任务:这是一个围绕真实任务执行的真实任务,公开材料可以回溯到具体使用者和具体产物。 原始资料写作:A user launched Qwen1.5-110B-Chat-GPTQ-Int4 through Xinference with model_uid qwen1.5-110b-4090, model_format gptq, quantization Int4, n_gpu=4, replica=2, gpu_memory_utilization=0.9, max_model_len=12000, and model_size_…
公开产物:公开材料提供公开代码、README 或项目配置,可用于核验任务结果、项目形态和模型绑定关系。 原始资料写作:The issue records the serving logs, including caching qwen/Qwen1.5-110B-Chat-GPTQ-Int4 from ModelScope and loading the model with vLLM configuration before an NCCL error when using replicas.
模型作用:Qwen1.5 Chat 110B 在该案例中承担真实任务执行相关的生成、分析、编排或实现角色。 原始资料写作:Qwen1.5-110B-Chat-GPTQ-Int4 was the large chat model being exposed through Xinference/vLLM for local multi-GPU serving.
A 类理由:有具体使用者、具体任务、公开原始证据和可访问产物,可公开核验。
风险边界:当前判断基于公开材料;若产物下线、仓库变更或模型参与比例仅来自作者自述,需要在引用前重新复核。 原始资料写作:This is an operational bug report and not a success story; it remains a real public deployment artifact with exact model identifier, serving parameters, hardware, and logs.
原始记录:Xinference user launches replicated Qwen1.5-110B-Chat-GPTQ-Int4 serving on 8×4090 GPUs
数据缺口
只基于现有字段判断缺口;缺失项不会被猜测填充。
- - 补充方向:核验 价格 的官方/API/案例来源。
- - 补充方向:核验 上下文 的官方/API/案例来源。
- - 补充方向:核验 输出 的官方/API/案例来源。
下一步验证建议
- - 继续保留 A 类案例的原始证据、产物页和版本快照。
- - 补官方发布/API/System Card 来源;若官方未披露,继续标注“官方未披露”。
- - 把 benchmark、教程、发布文保留为背景资料,不提升为真实案例。