jane

13B-chat量化之后推理报错

1

model = AutoPeftModelForCausalLM.from_pretrained("./output", device_map="auto", torch_dtype=torch.float16, trust_remote_code=True) model = model.quantize(4).cuda() model.save_pretrained('./quan') print("quan done") model = AutoModelForCausalLM.from_pretrained("./quan", device_map="auto", trust_remote_code=True) 报错： Traceback (most recent call last): File "/ossfs/node_38300464/workspace/Baichuan2-main/fine-tune/test.py", line 26, in #model =...

reranker为什么对大小写这么敏感！

6

from FlagEmbedding import FlagReranker reranker = FlagReranker('/bge-reranker-v2-m3/bge-reranker-v2-m3', use_fp16=True) score = reranker.compute_score(['Jc Sweet House', 'Candy House'], normalize=True) print(score)#0.14354750671079064 score = reranker.compute_score(['jc sweet house', 'candy house'], normalize=True) print(score)#0.8033281311678119

微调bge-reranker-v2-m3时loss一直变大

4

配置： !torchrun --nproc_per_node 1 \ -m FlagEmbedding.reranker.run \ --output_dir /bge-reranker-v2-m3-finetune \ --model_name_or_path /bge-reranker-v2-m3/bge-reranker-v2-m3 \ --train_data output.jsonl \ --learning_rate 6e-5 \ --fp16 \ --num_train_epochs 20 \ --per_device_train_batch_size 6 \ --gradient_accumulation_steps 4...

所有的清华的文档都是空的retriever-pretrained-checkpoint

1

{'error_msg': 'Share link bc96946dd9a14c84b8d4 not found.'}

IS there any wrong with the crawler_extraction?

1

 item_value = extract(html, xpath_rule[item][0]) should be item_value = extract(html, xpath_rule[item])