SmolLM2-135M-Instruct / variants /cpu-int4 /train /trainable_parameters.json
martinkorelic's picture
Upload folder using huggingface_hub
0442819 verified
Raw
History Blame Contribute Delete
11.6 kB
{
"baseLayers": [
"base_model.model.model.layers.0.self_attn.q_proj",
"base_model.model.model.layers.0.self_attn.v_proj",
"base_model.model.model.layers.1.self_attn.q_proj",
"base_model.model.model.layers.1.self_attn.v_proj",
"base_model.model.model.layers.10.self_attn.q_proj",
"base_model.model.model.layers.10.self_attn.v_proj",
"base_model.model.model.layers.11.self_attn.q_proj",
"base_model.model.model.layers.11.self_attn.v_proj",
"base_model.model.model.layers.12.self_attn.q_proj",
"base_model.model.model.layers.12.self_attn.v_proj",
"base_model.model.model.layers.13.self_attn.q_proj",
"base_model.model.model.layers.13.self_attn.v_proj",
"base_model.model.model.layers.14.self_attn.q_proj",
"base_model.model.model.layers.14.self_attn.v_proj",
"base_model.model.model.layers.15.self_attn.q_proj",
"base_model.model.model.layers.15.self_attn.v_proj",
"base_model.model.model.layers.16.self_attn.q_proj",
"base_model.model.model.layers.16.self_attn.v_proj",
"base_model.model.model.layers.17.self_attn.q_proj",
"base_model.model.model.layers.17.self_attn.v_proj",
"base_model.model.model.layers.18.self_attn.q_proj",
"base_model.model.model.layers.18.self_attn.v_proj",
"base_model.model.model.layers.19.self_attn.q_proj",
"base_model.model.model.layers.19.self_attn.v_proj",
"base_model.model.model.layers.2.self_attn.q_proj",
"base_model.model.model.layers.2.self_attn.v_proj",
"base_model.model.model.layers.20.self_attn.q_proj",
"base_model.model.model.layers.20.self_attn.v_proj",
"base_model.model.model.layers.21.self_attn.q_proj",
"base_model.model.model.layers.21.self_attn.v_proj",
"base_model.model.model.layers.22.self_attn.q_proj",
"base_model.model.model.layers.22.self_attn.v_proj",
"base_model.model.model.layers.23.self_attn.q_proj",
"base_model.model.model.layers.23.self_attn.v_proj",
"base_model.model.model.layers.24.self_attn.q_proj",
"base_model.model.model.layers.24.self_attn.v_proj",
"base_model.model.model.layers.25.self_attn.q_proj",
"base_model.model.model.layers.25.self_attn.v_proj",
"base_model.model.model.layers.26.self_attn.q_proj",
"base_model.model.model.layers.26.self_attn.v_proj",
"base_model.model.model.layers.27.self_attn.q_proj",
"base_model.model.model.layers.27.self_attn.v_proj",
"base_model.model.model.layers.28.self_attn.q_proj",
"base_model.model.model.layers.28.self_attn.v_proj",
"base_model.model.model.layers.29.self_attn.q_proj",
"base_model.model.model.layers.29.self_attn.v_proj",
"base_model.model.model.layers.3.self_attn.q_proj",
"base_model.model.model.layers.3.self_attn.v_proj",
"base_model.model.model.layers.4.self_attn.q_proj",
"base_model.model.model.layers.4.self_attn.v_proj",
"base_model.model.model.layers.5.self_attn.q_proj",
"base_model.model.model.layers.5.self_attn.v_proj",
"base_model.model.model.layers.6.self_attn.q_proj",
"base_model.model.model.layers.6.self_attn.v_proj",
"base_model.model.model.layers.7.self_attn.q_proj",
"base_model.model.model.layers.7.self_attn.v_proj",
"base_model.model.model.layers.8.self_attn.q_proj",
"base_model.model.model.layers.8.self_attn.v_proj",
"base_model.model.model.layers.9.self_attn.q_proj",
"base_model.model.model.layers.9.self_attn.v_proj"
],
"peftMethod": "lora",
"rank": 8,
"requiresGrad": [
"backbone.model.layers.0.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.0.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.0.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.0.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.1.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.1.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.1.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.1.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.10.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.10.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.10.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.10.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.11.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.11.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.11.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.11.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.12.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.12.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.12.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.12.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.13.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.13.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.13.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.13.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.14.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.14.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.14.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.14.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.15.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.15.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.15.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.15.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.16.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.16.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.16.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.16.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.17.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.17.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.17.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.17.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.18.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.18.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.18.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.18.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.19.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.19.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.19.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.19.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.2.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.2.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.2.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.2.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.20.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.20.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.20.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.20.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.21.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.21.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.21.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.21.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.22.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.22.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.22.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.22.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.23.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.23.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.23.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.23.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.24.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.24.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.24.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.24.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.25.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.25.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.25.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.25.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.26.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.26.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.26.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.26.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.27.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.27.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.27.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.27.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.28.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.28.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.28.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.28.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.29.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.29.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.29.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.29.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.3.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.3.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.3.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.3.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.4.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.4.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.4.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.4.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.5.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.5.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.5.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.5.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.6.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.6.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.6.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.6.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.7.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.7.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.7.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.7.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.8.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.8.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.8.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.8.self_attn.v_proj.lora_B.lora.weight",
"backbone.model.layers.9.self_attn.q_proj.lora_A.lora.weight",
"backbone.model.layers.9.self_attn.q_proj.lora_B.lora.weight",
"backbone.model.layers.9.self_attn.v_proj.lora_A.lora.weight",
"backbone.model.layers.9.self_attn.v_proj.lora_B.lora.weight"
],
"trainableParameterCount": 460800
}