镜像地址来自官方文档 https://www.hiascend.com/developer/ascendhub/detail/07a016975cc341f3a5ae131f2b52399d
mis-tei:6.0.0-300I-Duo-aarch64 下载下来后docker镜像名称为 
服务器配置为昇腾910 + openEuler ，使用前提已满足
启动命令及报错截图：
因为下载下来镜像名称与官方示例不同，所以使用docker镜像id来启动
docker run -e ASCEND_VISIBLE_DEVICES=0 -itd --name=tei --net=host \
-v /home/data:/home/HwHiAiUser/model \
-u root \
--entrypoint /home/HwHiAiUser/start.sh \
7423ad7cd1fc BAAI/bge-reranker-large 127.0.0.1 8085
报错文本：
Model '/home/HwHiAiUser/model/bge-reranker-large' exists.
NPU_ID: 0, CHIP_ID: 0， CHIP_LOGIC_ID: 0 CHIP_TYPE: 310P3, MEMORY_TYPE: DDR, CAPACITY: 44280, USAGE_RATE: 3, AVAIL_CAPACITY: 42951
Using NPU: 0 start TEI service
Starting TEI service on 127.0.0.1:8086...
2025-03-31T02:54:07.490367Z INFO text_embeddings_router: router/src/main.rs:140: Args { model_id: "/hom*/**********/*****/***-********-**rge", revision: None, tokenization_workers: None, dtype: None, pooling: None, max_concurrent_requests: 512, max_batch_tokens: 16384, max_batch_requests: None, max_client_batch_size: 32, auto_truncate: false, hf_api_token: None, hostname: "127.0.0.1", port: 8086, uds_path: "/tmp/text-embeddings-inference-server", huggingface_hub_cache: None, payload_limit: 2000000, api_key: None, json_output: false, otlp_endpoint: None, cors_allow_origin: None }
thread 'main' panicked at /home/HwHiAiUser/text-embeddings-inference/router/src/lib.rs:137:62:
tokenizer.json not found. text-embeddings-inference only supports fast tokenizers: Error("expected value", line: 1, column: 1)
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
/home/HwHiAiUser/start.sh: line 69: 360 Aborted (core dumped) text-embeddings-router --model-id "${MODEL_DIR}/${MODEL_ID##*/}" --port "${LISTEN_PORT}" --hostname "${LISTEN_IP}"
点赞
收藏
回复
分享
举报
已解决
镜像地址来自官方文档 https://www.hiascend.com/developer/ascendhub/detail/07a016975cc341f3a5ae131f2b52399d
mis-tei:6.0.0-300I-Duo-aarch64 下载下来后docker镜像名称为
服务器配置为昇腾910 + openEuler ，使用前提已满足
启动命令及报错截图：
因为下载下来镜像名称与官方示例不同，所以使用docker镜像id来启动
docker run -e ASCEND_VISIBLE_DEVICES=0 -itd --name=tei --net=host \
-v /home/data:/home/HwHiAiUser/model \
-u root \
--entrypoint /home/HwHiAiUser/start.sh \
7423ad7cd1fc BAAI/bge-reranker-large 127.0.0.1 8085
报错文本：
Model '/home/HwHiAiUser/model/bge-reranker-large' exists.
NPU_ID: 0, CHIP_ID: 0， CHIP_LOGIC_ID: 0 CHIP_TYPE: 310P3, MEMORY_TYPE: DDR, CAPACITY: 44280, USAGE_RATE: 3, AVAIL_CAPACITY: 42951
Using NPU: 0 start TEI service
Starting TEI service on 127.0.0.1:8086...
2025-03-31T02:54:07.490367Z INFO text_embeddings_router: router/src/main.rs:140: Args { model_id: "/hom*/**********/*****/***-********-**rge", revision: None, tokenization_workers: None, dtype: None, pooling: None, max_concurrent_requests: 512, max_batch_tokens: 16384, max_batch_requests: None, max_client_batch_size: 32, auto_truncate: false, hf_api_token: None, hostname: "127.0.0.1", port: 8086, uds_path: "/tmp/text-embeddings-inference-server", huggingface_hub_cache: None, payload_limit: 2000000, api_key: None, json_output: false, otlp_endpoint: None, cors_allow_origin: None }
thread 'main' panicked at /home/HwHiAiUser/text-embeddings-inference/router/src/lib.rs:137:62:
tokenizer.json not found. text-embeddings-inference only supports fast tokenizers: Error("expected value", line: 1, column: 1)
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
/home/HwHiAiUser/start.sh: line 69: 360 Aborted (core dumped) text-embeddings-router --model-id "${MODEL_DIR}/${MODEL_ID##*/}" --port "${LISTEN_PORT}" --hostname "${LISTEN_IP}"