dev-2.0.T17.B010-800I-A2-py311-ubuntu22.04-aarch64 部署Qwen3-Coder 失败_MindIE_昇腾论坛

# dev-2.0.T17.B010-800I-A2-py311-ubuntu22.04-aarch64 部署Qwen3-Coder 失败_MindIE_昇腾论坛

## 概述

本文档基于昇腾社区论坛帖子生成的技术教程。

**原始链接**: https://www.hiascend.com/forum/thread-02116189153594124348-1-1.html

**生成时间**: 2025-08-27 10:33:57

---

## 问题描述

硬件环境昇腾910B Exception:weight model.layers.0.self_attn.q_proj.bias does not exist Traceback (most recent call last): File "/usr/local/lib/python3.11/site-packages/model_wrapper/model.py", line 57, in initialize return self.python_model.initialize(config)  File "/usr/local/lib/python3.11/site-packages/model_wrapper/standard_model.py", line 117, in initialize self.generator = Generator(  File "/usr/local/lib/python3.11/site-packages/mindie_llm/text_generator/generator.py", line 181, in __init__ self.generator_backend = get_generator_backend(model_config)  File "/usr/local/lib/python3.11/site-packages/mindie_llm/text_generator/adapter/__init__.py", line 31, in get_generator_backend return generator_cls(model_config)  File "/usr/local/lib/python3.11/site-packages/mindie_llm/text_generator/adapter/generator_torch.py", line 121, in __init__ super().__init__(model_config) File "/usr/local/lib/python3.11/site-packages/mindie_llm/text_generator/adapter/generator_backend.py", line 134, in __init__...

## 完整内容

硬件环境昇腾910B Exception:weight model.layers.0.self_attn.q_proj.bias does not exist Traceback (most recent call last): File "/usr/local/lib/python3.11/site-packages/model_wrapper/model.py", line 57, in initialize return self.python_model.initialize(config)  File "/usr/local/lib/python3.11/site-packages/model_wrapper/standard_model.py", line 117, in initialize self.generator = Generator(  File "/usr/local/lib/python3.11/site-packages/mindie_llm/text_generator/generator.py", line 181, in __init__ self.generator_backend = get_generator_backend(model_config)  File "/usr/local/lib/python3.11/site-packages/mindie_llm/text_generator/adapter/__init__.py", line 31, in get_generator_backend return generator_cls(model_config)  File "/usr/local/lib/python3.11/site-packages/mindie_llm/text_generator/adapter/generator_torch.py", line 121, in __init__ super().__init__(model_config) File "/usr/local/lib/python3.11/site-packages/mindie_llm/text_generator/adapter/generator_backend.py", line 134, in __init__ self.model_wrapper = get_model_wrapper(model_config, backend_type)  File "/usr/local/lib/python3.11/site-packages/mindie_llm/modeling/model_wrapper/__init__.py", line 15, in get_model_wrapper return wrapper_cls(**model_config)  File "/usr/local/lib/python3.11/site-packages/mindie_llm/modeling/model_wrapper/atb/atb_model_wrapper.py", line 78, in __init__ self.model_runner.load_weights() File "/usr/local/Ascend/atb-models/atb_llm/runner/model_runner.py", line 203, in load_weights self.model = self.model_cls(self.config,  File "/usr/local/Ascend/atb-models/atb_llm/models/qwen2/flash_causal_qwen2.py", line 47, in __init__ self.transformer = FlashQwenModel(  File "/usr/local/Ascend/atb-models/atb_llm/models/qwen2/modeling_qwen2.py", line 315, in __init__ [ File "/usr/local/Ascend/atb-models/atb_llm/models/qwen2/modeling_qwen2.py", line 316, in <listcomp> FlashQwenLayer( File "/usr/local/Ascend/atb-models/atb_llm/models/qwen2/modeling_qwen2.py", line 222, in __init__ self.attn = FlashQwenAttention(  File "/usr/local/Ascend/atb-models/atb_llm/models/qwen2/modeling_qwen2.py", line 147, in __init__ self.c_attn = load_column_multi(  File "/usr/local/Ascend/atb-models/atb_llm/utils/layers/__init__.py", line 50, in load_column_multi b = [weights.get_sharded(f"{p}.bias", dim=0, gqa_size=head_size) for p in prefixes]  File "/usr/local/Ascend/atb-models/atb_llm/utils/layers/__init__.py", line 50, in <listcomp> b = [weights.get_sharded(f"{p}.bias", dim=0, gqa_size=head_size) for p in prefixes]  File "/usr/local/Ascend/atb-models/atb_llm/utils/weights.py", line 343, in get_sharded slice_ = self._get_slice(tensor_name)  File "/usr/local/Ascend/atb-models/atb_llm/utils/weights.py", line 812, in _get_slice filename, tensor_name = self.get_filename(tensor_name)  File "/usr/local/Ascend/atb-models/atb_llm/utils/weights.py", line 158, in get_filename raise AssertionError(f"weight {tensor_name} does not exist") AssertionError: weight model.layers.0.self_attn.q_proj.bias does not exist [2025-07-29 13:52:45.029+0800] [64727] [281472552333536] [batchscheduler] [ERROR] [model.py:62] : [MIE04E13030A] [Model] >>> return initialize error result: {'status': 'error', 'npuBlockNum': '0', 'cpuBlockNum': '0', 'memPoolId': '-1'}

---

## 技术要点总结

基于以上内容，主要技术要点包括：

1. **问题类型**: 错误处理
2. **涉及技术**: 昇腾, NPU, AI, MindIE
3. **解决方案**: 请参考完整内容中的解决方案

## 相关资源

- 昇腾社区: https://www.hiascend.com/
- 昇腾论坛: https://www.hiascend.com/forum/

---

*本文档由AI自动生成，仅供参考。如有疑问，请参考原始帖子。*