model could not load after finetuning

#2
by zguo0525 - opened

Hey I wonder if anyone has experienced this issue before when loading a fine-tuned llama 2 model

Some weights of the model checkpoint at ../../vicuna4tools/output/llama-2-7b-hf/total_IC_train_clean were not used when initializing LlamaForCausalLM: ['_flat_param']

  • This IS expected if you are initializing LlamaForCausalLM from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).
  • This IS NOT expected if you are initializing LlamaForCausalLM from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).
    Some weights of LlamaForCausalLM were not initialized from the model checkpoint at ../../vicuna4tools/output/llama-2-7b-hf/total_IC_train_clean and are newly initialized: ['layers.15.self_attn.rotary_emb.inv_freq', 'layers.2.self_attn.k_proj.weight',
    You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.
NousResearch org

Hey I wonder if anyone has experienced this issue before when loading a fine-tuned llama 2 model

Some weights of the model checkpoint at ../../vicuna4tools/output/llama-2-7b-hf/total_IC_train_clean were not used when initializing LlamaForCausalLM: ['_flat_param']

  • This IS expected if you are initializing LlamaForCausalLM from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).
  • This IS NOT expected if you are initializing LlamaForCausalLM from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).
    Some weights of LlamaForCausalLM were not initialized from the model checkpoint at ../../vicuna4tools/output/llama-2-7b-hf/total_IC_train_clean and are newly initialized: ['layers.15.self_attn.rotary_emb.inv_freq', 'layers.2.self_attn.k_proj.weight',
    You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.

I have no idea

Sign up or log in to comment