"神龙ai中chatGlm ,通过ptuning 创建自己的数据模型， 过web访问调试无应答为什么

"神龙ai中chatGlm , 通过ptuning 创建了自己的 数据模型， 在web demo进行测试，通过web访问调试，无应答 ，怎么排查？
日志显示：some weight 未设置， 系统自动初始化了 Explicitly passing a revision is encouraged when loading a model with custom code to ensure no malicious code has been contributed in a newer revision.Some weights of ChatGLMForConditionalGeneration were not initialized from the model checkpoint at........（此处省略提示信息）You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.操作流程：模型的生成 是按demo走的，只是自己建了一个 专业名词的 dev.json train.json , 通过 train.sh 生成模型，通过evaluate 评测， 然后修改 web demo 的模型为 自己生成的模型路径。
数据是参考 AdvertiseGen {content:"""",""summary"":""""} 键值对的形式 来建的 权重没有初始化，我可以参考 去组织？
streamlit run web_demoshoe2.py --server.address 0.0.0.0 --server.port 50003

You can now view your Streamlit app in your browser.

URL: http://0.0.0.0:50003

Explicitly passing a revision is encouraged when loading a model with custom code to ensure no malicious code has been contributed in a newer revision.
Explicitly passing a revision is encouraged when loading a configuration with custom code to ensure no malicious code has been contributed in a newer revision.
Explicitly passing a revision is encouraged when loading a model with custom code to ensure no malicious code has been contributed in a newer revision.
Some weights of ChatGLMForConditionalGeneration were not initialized from the model checkpoint at /root/chatglm/ChatGLM-6B/ptuning/output/shoetech-chatglm-6b-pt-128-2e-2/checkpoint-3000 and are newly initialized: ['transformer.layers.11.attention.query_key_value.weight', 'transformer.layers.23.mlp.dense_h_to_4h.bias', 'transformer.layers.3.attention.query_key_value.weight_scale', 'transformer.layers.7.mlp.dense_4h_to_h.weight', 'transformer.layers.16.post_attention_layernorm.bias', 'transformer.layers.3.mlp.dense_h_to_4h.weight_scale', 'transformer.layers.1.mlp.dense_4h_to_h.weight_scale', 'transformer.layers.4.attention.dense.bias', 'transformer.layers.7.mlp.dense_4h_to_h.bias', 'transformer.layers.11.input_layernorm.weight', 'transformer.layers.18.attention.dense.bias', 'transformer.layers.13.mlp.dense_h_to_4h.weight', 'transformer.layers.21.mlp.dense_h_to_4h.bias', 'transformer.layers.12.input_layernorm.weight', 'transformer.layers.4.mlp.dense_h_to_4h.weight_scale', 'transformer.layers.23.attention.dense.weight', 'transformer.layers.11.attention.query_key_value.bias', 'transformer.layers.4.post_attention_layernorm.weight', 'transformer.layers.12.attention.dense.weight_scale', 'transformer.layers.5.attention.rotary_emb.inv_freq', 'transformer.layers.12.attention.query_key_value.weight_scale', 'transformer.layers.20.post_attention_layernorm.weight', 'transformer.layers.13.mlp.dense_4h_to_h.weight_scale', 'transformer.layers.12.attention.dense.bias', 'transformer.layers.7.mlp.dense_4h_to_h.weight_scale', 'transformer.layers.8.input_layernorm.weight', 'transformer.layers.4.mlp.dense_4h_to_h.bias', 'transformer.layers.1.post_attention_layernorm.weight', 'transformer.layers.3.post_attention_layernorm.bias', 'transformer.layers.7.post_attention_layernorm.weight', 'transformer.layers.24.post_attention_layernorm.bias', 'transformer.layers.25.mlp.dense_h_to_4h.weight', 'transformer.layers.11.mlp.dense_4h_to_h.bias', 'transformer.layers.23.input_layernorm.bias', 'transformer.layers.10.input_layernorm.weight', 'transformer.layers.11.attention.rotary_emb.inv_freq', 'transformer.layers.25.mlp.dense_4h_to_h.weight_scale', 'transformer.layers.3.attention.dense.weight_scale', 'transformer.layers.21.mlp.dense_h_to_4h.weight_scale', 'transformer.layers.27.post_attention_layernorm.bias', 'transformer.layers.4.attention.query_key_value
"