光合开发者社区-论坛

Qwen

dkt25.04.1-ubuntu22.04-pytorch2.4.1基于K100_AI运行Qwen3-30B-2507-AWQ-int4报错

2025-09-05 22:43:03

334

环境：镜像是image.sourcefind.cn:5000/dcu/admin/base/pytorch:2.4.1-ubuntu22.04-dtk25.04.1-py3.10，vllm是0.8.5,运行server报下面的错误

Starting to load model /workspace/llm-models/Qwen3-30B-A3B-Instruct-2507-AWQ-4bit...

WARNING 09-05 22:37:55 [utils.py:207] The model class Qwen3MoeForCausalLM has not defined `packed_modules_mapping`, this may lead to incorrect mapping of quantized or ignored modules

INFO 09-05 22:37:55 [compressed_tensors_wNa16.py:94] Using MarlinLinearKernel for CompressedTensorsWNA16

Loading safetensors checkpoint shards: 0% Completed | 0/4 [00:00

Loading safetensors checkpoint shards: 25% Completed | 1/4 [00:07<00:23, 7.87s/it]

Loading safetensors checkpoint shards: 50% Completed | 2/4 [00:16<00:16, 8.34s/it]

Loading safetensors checkpoint shards: 75% Completed | 3/4 [00:25<00:08, 8.68s/it]

Loading safetensors checkpoint shards: 100% Completed | 4/4 [00:30<00:00, 7.08s/it]

Loading safetensors checkpoint shards: 100% Completed | 4/4 [00:30<00:00, 7.56s/it]

INFO 09-05 22:38:26 [loader.py:460] Loading weights took 30.48 seconds

ERROR 09-05 22:38:26 [engine.py:466] 'QKVParallelLinear' object has no attribute 'weight'

ERROR 09-05 22:38:26 [engine.py:466] Traceback (most recent call last):

ERROR 09-05 22:38:26 [engine.py:466] File "/usr/local/lib/python3.10/dist-packages/vllm/engine/multiprocessing/engine.py", line 454, in run_mp_engine

ERROR 09-05 22:38:26 [engine.py:466] engine = MQLLMEngine.from_vllm_config(

ERROR 09-05 22:38:26 [engine.py:466] File "/usr/local/lib/python3.10/dist-packages/vllm/engine/multiprocessing/engine.py", line 134, in from_vllm_config

ERROR 09-05 22:38:26 [engine.py:466] return cls(

ERROR 09-05 22:38:26 [engine.py:466] File "/usr/local/lib/python3.10/dist-packages/vllm/engine/multiprocessing/engine.py", line 88, in __init__

ERROR 09-05 22:38:26 [engine.py:466] self.engine = LLMEngine(*args, **kwargs)

ERROR 09-05 22:38:26 [engine.py:466] File "/usr/local/lib/python3.10/dist-packages/vllm/engine/llm_engine.py", line 283, in __init__

ERROR 09-05 22:38:26 [engine.py:466] self.model_executor = executor_class(vllm_config=vllm_config)

ERROR 09-05 22:38:26 [engine.py:466] File "/usr/local/lib/python3.10/dist-packages/vllm/executor/executor_base.py", line 52, in __init__

ERROR 09-05 22:38:26 [engine.py:466] self._init_executor()

ERROR 09-05 22:38:26 [engine.py:466] File "/usr/local/lib/python3.10/dist-packages/vllm/executor/uniproc_executor.py", line 47, in _init_executor

ERROR 09-05 22:38:26 [engine.py:466] self.collective_rpc("load_model")

ERROR 09-05 22:38:26 [engine.py:466] File "/usr/local/lib/python3.10/dist-packages/vllm/executor/uniproc_executor.py", line 56, in collective_rpc

ERROR 09-05 22:38:26 [engine.py:466] answer = run_method(self.driver_worker, method, args, kwargs)

ERROR 09-05 22:38:26 [engine.py:466] File "/usr/local/lib/python3.10/dist-packages/vllm/utils.py", line 2572, in run_method

ERROR 09-05 22:38:26 [engine.py:466] return func(*args, **kwargs)

ERROR 09-05 22:38:26 [engine.py:466] File "/usr/local/lib/python3.10/dist-packages/vllm/worker/worker.py", line 203, in load_model

ERROR 09-05 22:38:26 [engine.py:466] self.model_runner.load_model()

ERROR 09-05 22:38:26 [engine.py:466] File "/usr/local/lib/python3.10/dist-packages/vllm/worker/model_runner.py", line 1126, in load_model

ERROR 09-05 22:38:26 [engine.py:466] self.model = get_model(vllm_config=self.vllm_config)

ERROR 09-05 22:38:26 [engine.py:466] File "/usr/local/lib/python3.10/dist-packages/vllm/model_executor/model_loader/__init__.py", line 14, in get_model

ERROR 09-05 22:38:26 [engine.py:466] return loader.load_model(vllm_config=vllm_config)

ERROR 09-05 22:38:26 [engine.py:466] File "/usr/local/lib/python3.10/dist-packages/vllm/model_executor/model_loader/loader.py", line 473, in load_model

ERROR 09-05 22:38:26 [engine.py:466] _process_weights_after_loading(model, model_config, target_device)

ERROR 09-05 22:38:26 [engine.py:466] File "/usr/local/lib/python3.10/dist-packages/vllm/model_executor/model_loader/loader.py", line 182, in _process_weights_after_loading

ERROR 09-05 22:38:26 [engine.py:466] quant_method.process_weights_after_loading(module)

ERROR 09-05 22:38:26 [engine.py:466] File "/usr/local/lib/python3.10/dist-packages/vllm/model_executor/layers/quantization/compressed_tensors/compressed_tensors.py", line 550, in process_weights_after_loading

ERROR 09-05 22:38:26 [engine.py:466] n=layer.weight.shape[0]

ERROR 09-05 22:38:26 [engine.py:466] File "/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py", line 1931, in __getattr__

ERROR 09-05 22:38:26 [engine.py:466] raise AttributeError(

ERROR 09-05 22:38:26 [engine.py:466] AttributeError: 'QKVParallelLinear' object has no attribute 'weight'

test_mhd4l1wfbkam

5楼

楼主问下Qwen3-30B-A3B-Instruct-2507这个模型海光k100_ai卡支持嘛服务器是海康威视的x86

2025/10/30

张

回复 @test_mhd4l1wfbkam

我没去测2507版本，只跑了Qwen3-30-A3B和Qwen3-3B-A3B-AWQ,应该可以直接跑

2025/11/13

yangzhong

4楼

更换这个模型https://www.modelscope.cn/models/swift/Qwen3-30B-A3B-AWQ 试试，亲测可行

2025/9/28

张

回复 @yangzhong

感谢回复，是vllm起服务的时候需要加-q=moe_wna16这个参数

2025/10/9

wangs

3楼

目前还不支持 Qwen3-30B-A3B-AWQ 预计这个月底支持该模型

2025/9/6

wangs

2楼

麻烦贴一下Qwen3-30B-A3B-Instruct-2507-AWQ-4bit 的config文件

2025/9/6

张

回复 @wangs

{ "architectures": [ "Qwen3MoeForCausalLM" ], "attention_bias": false, "attention_dropout": 0.0, "bos_token_id": 151643, "decoder_sparse_step": 1, "dtype": "bfloat16", "eos_token_id": 151645, "head_dim": 128, "hidden_act": "silu", "hidden_size": 2048, "initializer_range": 0.02, "intermediate_size": 6144, "max_position_embeddings": 262144, "max_window_layers": 48, "mlp_only_layers": [], "model_type": "qwen3_moe", "moe_intermediate_size": 768, "norm_topk_prob": true, "num_attention_heads": 32, "num_experts": 128, "num_experts_per_tok": 8, "num_hidden_layers": 48, "num_key_value_heads": 4, "output_router_logits": false, "quantization_config": { "config_groups": { "group_0": { "format": "pack-quantized", "input_activations": null, "output_activations": null, "targets": [ "Linear" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": 32, "num_bits": 4, "observer": "mse", "observer_kwargs": {}, "strategy": "group", "symmetric": true, "type": "int" } } }, "format": "pack-quantized", "global_compression_ratio": null, "ignore": [ "model.layers.0.mlp.gate", "model.layers.1.mlp.gate", "model.layers.2.mlp.gate", "model.layers.3.mlp.gate", "model.layers.4.mlp.gate", "model.layers.5.mlp.gate", "model.layers.6.mlp.gate", "model.layers.7.mlp.gate", "model.layers.8.mlp.gate", "model.layers.9.mlp.gate", "model.layers.10.mlp.gate", "model.layers.11.mlp.gate", "model.layers.12.mlp.gate", "model.layers.13.mlp.gate", "model.layers.14.mlp.gate", "model.layers.15.mlp.gate", "model.layers.16.mlp.gate", "model.layers.17.mlp.gate", "model.layers.18.mlp.gate", "model.layers.19.mlp.gate", "model.layers.20.mlp.gate", "model.layers.21.mlp.gate", "model.layers.22.mlp.gate", "model.layers.23.mlp.gate", "model.layers.24.mlp.gate", "model.layers.25.mlp.gate", "model.layers.26.mlp.gate", "model.layers.27.mlp.gate", "model.layers.28.mlp.gate", "model.layers.29.mlp.gate", "model.layers.30.mlp.gate", "model.layers.31.mlp.gate", "model.layers.32.mlp.gate", "model.layers.33.mlp.gate", "model.layers.34.mlp.gate", "model.layers.35.mlp.gate", "model.layers.36.mlp.gate", "model.layers.37.mlp.gate", "model.layers.38.mlp.gate", "model.layers.39.mlp.gate", "model.layers.40.mlp.gate", "model.layers.41.mlp.gate", "model.layers.42.mlp.gate", "model.layers.43.mlp.gate", "model.layers.44.mlp.gate", "model.layers.45.mlp.gate", "model.layers.46.mlp.gate", "model.layers.47.mlp.gate", "lm_head" ], "kv_cache_scheme": null, "quant_method": "compressed-tensors", "quantization_status": "compressed", "sparsity_config": {}, "transform_config": {}, "version": "0.10.3.dev47+ge463fe6" }, "rms_norm_eps": 1e-06, "rope_scaling": null, "rope_theta": 10000000, "router_aux_loss_coef": 0.001, "sliding_window": null, "tie_word_embeddings": false, "transformers_version": "4.56.0.dev0", "use_cache": true, "use_sliding_window": false, "vocab_size": 151936 }

2025/9/6

张

回复 @wangs

模型地址：https://www.modelscope.cn/models/cpatonn-mirror/Qwen3-30B-A3B-Instruct-2507-AWQ-4bit/file/view/master/config.json?status=1

2025/9/6

张

楼主

1楼

另外，使用Qwen3-30B-A3B-AWQ报以下问题 INFO 09-06 00:02:25 [model_runner.py:1123] Starting to load model /workspace/llm-models/Qwen3-30B-A3B-AWQ... WARNING 09-06 00:02:26 [utils.py:207] The model class Qwen3MoeForCausalLM has not defined `packed_modules_mapping`, this may lead to incorrect mapping of quantized or ignored modules ERROR 09-06 00:02:26 [engine.py:466] Traceback (most recent call last): File "/usr/local/lib/python3.10/dist-packages/vllm/engine/multiprocessing/engine.py", line 454, in run_mp_engine engine = MQLLMEngine.from_vllm_config( File "/usr/local/lib/python3.10/dist-packages/vllm/engine/multiprocessing/engine.py", line 134, in from_vllm_config return cls( File "/usr/local/lib/python3.10/dist-packages/vllm/engine/multiprocessing/engine.py", line 88, in __init__ self.engine = LLMEngine(*args, **kwargs) File "/usr/local/lib/python3.10/dist-packages/vllm/engine/llm_engine.py", line 283, in __init__ self.model_executor = executor_class(vllm_config=vllm_config) File "/usr/local/lib/python3.10/dist-packages/vllm/executor/executor_base.py", line 52, in __init__ self._init_executor() File "/usr/local/lib/python3.10/dist-packages/vllm/executor/uniproc_executor.py", line 47, in _init_executor self.collective_rpc("load_model") File "/usr/local/lib/python3.10/dist-packages/vllm/executor/uniproc_executor.py", line 56, in collective_rpc answer = run_method(self.driver_worker, method, args, kwargs) File "/usr/local/lib/python3.10/dist-packages/vllm/utils.py", line 2572, in run_method return func(*args, **kwargs) File "/usr/local/lib/python3.10/dist-packages/vllm/worker/worker.py", line 203, in load_model self.model_runner.load_model() File "/usr/local/lib/python3.10/dist-packages/vllm/worker/model_runner.py", line 1126, in load_model self.model = get_model(vllm_config=self.vllm_config) File "/usr/local/lib/python3.10/dist-packages/vllm/model_executor/model_loader/__init__.py", line 14, in get_model return loader.load_model(vllm_config=vllm_config) File "/usr/local/lib/python3.10/dist-packages/vllm/model_executor/model_loader/loader.py", line 454, in load_model model = _initialize_model(vllm_config=vllm_config) File "/usr/local/lib/python3.10/dist-packages/vllm/model_executor/model_loader/loader.py", line 133, in _initialize_model return model_class(vllm_config=vllm_config, prefix=prefix) File "/usr/local/lib/python3.10/dist-packages/vllm/model_executor/models/qwen3_moe.py", line 552, in __init__ self.model = Qwen3MoeModel(vllm_config=vllm_config, File "/usr/local/lib/python3.10/dist-packages/vllm/compilation/decorators.py", line 151, in __init__ old_init(self, vllm_config=vllm_config, prefix=prefix, **kwargs) File "/usr/local/lib/python3.10/dist-packages/vllm/model_executor/models/qwen3_moe.py", line 346, in __init__ self.start_layer, self.end_layer, self.layers = make_layers( File "/usr/local/lib/python3.10/dist-packages/vllm/model_executor/models/utils.py", line 609, in make_layers [PPMissingLayer() for _ in range(start_layer)] + [ File "/usr/local/lib/python3.10/dist-packages/vllm/model_executor/models/utils.py", line 610, in maybe_offload_to_cpu(layer_fn(prefix=f"{prefix}.{idx}")) File "/usr/local/lib/python3.10/dist-packages/vllm/model_executor/models/qwen3_moe.py", line 348, in lambda prefix: Qwen3MoeDecoderLayer(config=config, File "/usr/local/lib/python3.10/dist-packages/vllm/model_executor/models/qwen3_moe.py", line 283, in __init__ self.mlp = Qwen3MoeSparseMoeBlock(config=config, File "/usr/local/lib/python3.10/dist-packages/vllm/model_executor/models/qwen3_moe.py", line 118, in __init__ self.experts = FusedMoE(num_experts=config.num_experts, File "/usr/local/lib/python3.10/dist-packages/vllm/model_executor/layers/fused_moe/layer.py", line 540, in __init__ assert self.quant_method is not None

2025/9/6