LookupError

[frownia37]

I am trying to apply MEMIT to Qwen2.5-7B-Instruct-1M model, and there is a problem like this:

               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

File "/root/xuetou/memit/experiments/py/demo.py", line 50, in demo_model_editing
model_new, orig_weights = apply_method(
^^^^^^^^^^^^^
File "/root/xuetou/memit/memit/memit_main.py", line 44, in apply_memit_to_model
deltas = execute_memit(model, tok, requests, hparams, cache_template=cache_template)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/root/xuetou/memit/memit/memit_main.py", line 101, in execute_memit
param = nethook.get_parameter(model, key)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/root/xuetou/memit/util/nethook.py", line 372, in get_parameter
raise LookupError(name)
LookupError: models.layers.3.mlp.gate_proj.weight

This should indicate a discrepancy between the layer name written in /root/xuetou/memit/hparams/MEMIT/my.json, and the actual name of the layer in the model. And this is my json:

{
"layers": [3, 4, 5, 6, 7],
"clamp_norm_factor": 0.75,
"layer_selection": "all",
"fact_token": "subject_last",
"v_num_grad_steps": 20,
"v_lr": 5e-1,
"v_loss_layer": 47,
"v_weight_decay": 0.5,
"kl_factor": 0.0625,
"mom2_adjustment": true,
"mom2_update_weight": 20000,
"rewrite_module_tmp": "models.layers.{}.mlp.gate_proj",
"layer_module_tmp": "model.layers.{}",
"mlp_module_tmp": "model.layers.{}.mlp",
"attn_module_tmp": "model.layers.{}.self_attn",
"ln_f_module": "model.norm",
"lm_head_module": "lm_head",
"mom2_dataset": "wikipedia",
"mom2_n_samples": 100000,
"mom2_dtype": "float32"
}

There should not be any difference now, since I acutally printed out the name of each layer in my model, and checked each one carefully, but the problem still exists.