|
{ |
|
"metadata": { |
|
"total_size": 5648487312 |
|
}, |
|
"weight_map": { |
|
"lm_head.weight": "model-00002-of-00002.safetensors", |
|
"transformer.embeddings.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.0.attn.A_log": "model-00001-of-00002.safetensors", |
|
"transformer.h.0.attn.D": "model-00001-of-00002.safetensors", |
|
"transformer.h.0.attn.b_proj.bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.0.attn.b_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.0.attn.dt_bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.0.attn.g_norm_swish_gate.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.0.attn.g_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.0.attn.gk_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.0.attn.k_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.0.attn.k_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.0.attn.o_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.0.attn.q_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.0.attn.q_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.0.attn.v_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.0.attn.v_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.0.mlp.swiglu.w1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.0.mlp.swiglu.w2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.0.mlp.swiglu.w3.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.0.norm_1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.0.norm_2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.1.attn.A_log": "model-00001-of-00002.safetensors", |
|
"transformer.h.1.attn.D": "model-00001-of-00002.safetensors", |
|
"transformer.h.1.attn.b_proj.bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.1.attn.b_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.1.attn.dt_bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.1.attn.g_norm_swish_gate.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.1.attn.g_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.1.attn.gk_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.1.attn.k_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.1.attn.k_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.1.attn.o_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.1.attn.q_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.1.attn.q_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.1.attn.v_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.1.attn.v_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.1.mlp.swiglu.w1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.1.mlp.swiglu.w2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.1.mlp.swiglu.w3.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.1.norm_1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.1.norm_2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.10.attn.A_log": "model-00001-of-00002.safetensors", |
|
"transformer.h.10.attn.D": "model-00001-of-00002.safetensors", |
|
"transformer.h.10.attn.b_proj.bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.10.attn.b_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.10.attn.dt_bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.10.attn.g_norm_swish_gate.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.10.attn.g_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.10.attn.gk_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.10.attn.k_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.10.attn.k_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.10.attn.o_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.10.attn.q_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.10.attn.q_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.10.attn.v_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.10.attn.v_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.10.mlp.swiglu.w1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.10.mlp.swiglu.w2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.10.mlp.swiglu.w3.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.10.norm_1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.10.norm_2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.11.attn.A_log": "model-00001-of-00002.safetensors", |
|
"transformer.h.11.attn.D": "model-00001-of-00002.safetensors", |
|
"transformer.h.11.attn.b_proj.bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.11.attn.b_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.11.attn.dt_bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.11.attn.g_norm_swish_gate.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.11.attn.g_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.11.attn.gk_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.11.attn.k_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.11.attn.k_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.11.attn.o_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.11.attn.q_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.11.attn.q_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.11.attn.v_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.11.attn.v_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.11.mlp.swiglu.w1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.11.mlp.swiglu.w2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.11.mlp.swiglu.w3.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.11.norm_1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.11.norm_2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.12.attn.A_log": "model-00001-of-00002.safetensors", |
|
"transformer.h.12.attn.D": "model-00001-of-00002.safetensors", |
|
"transformer.h.12.attn.b_proj.bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.12.attn.b_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.12.attn.dt_bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.12.attn.g_norm_swish_gate.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.12.attn.g_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.12.attn.gk_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.12.attn.k_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.12.attn.k_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.12.attn.o_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.12.attn.q_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.12.attn.q_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.12.attn.v_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.12.attn.v_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.12.mlp.swiglu.w1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.12.mlp.swiglu.w2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.12.mlp.swiglu.w3.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.12.norm_1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.12.norm_2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.13.attn.A_log": "model-00001-of-00002.safetensors", |
|
"transformer.h.13.attn.D": "model-00001-of-00002.safetensors", |
|
"transformer.h.13.attn.b_proj.bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.13.attn.b_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.13.attn.dt_bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.13.attn.g_norm_swish_gate.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.13.attn.g_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.13.attn.gk_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.13.attn.k_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.13.attn.k_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.13.attn.o_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.13.attn.q_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.13.attn.q_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.13.attn.v_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.13.attn.v_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.13.mlp.swiglu.w1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.13.mlp.swiglu.w2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.13.mlp.swiglu.w3.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.13.norm_1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.13.norm_2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.14.attn.A_log": "model-00001-of-00002.safetensors", |
|
"transformer.h.14.attn.D": "model-00001-of-00002.safetensors", |
|
"transformer.h.14.attn.b_proj.bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.14.attn.b_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.14.attn.dt_bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.14.attn.g_norm_swish_gate.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.14.attn.g_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.14.attn.gk_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.14.attn.k_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.14.attn.k_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.14.attn.o_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.14.attn.q_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.14.attn.q_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.14.attn.v_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.14.attn.v_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.14.mlp.swiglu.w1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.14.mlp.swiglu.w2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.14.mlp.swiglu.w3.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.14.norm_1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.14.norm_2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.15.attn.A_log": "model-00001-of-00002.safetensors", |
|
"transformer.h.15.attn.D": "model-00001-of-00002.safetensors", |
|
"transformer.h.15.attn.b_proj.bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.15.attn.b_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.15.attn.dt_bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.15.attn.g_norm_swish_gate.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.15.attn.g_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.15.attn.gk_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.15.attn.k_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.15.attn.k_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.15.attn.o_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.15.attn.q_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.15.attn.q_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.15.attn.v_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.15.attn.v_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.15.mlp.swiglu.w1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.15.mlp.swiglu.w2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.15.mlp.swiglu.w3.weight": "model-00002-of-00002.safetensors", |
|
"transformer.h.15.norm_1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.15.norm_2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.16.attn.A_log": "model-00002-of-00002.safetensors", |
|
"transformer.h.16.attn.D": "model-00002-of-00002.safetensors", |
|
"transformer.h.16.attn.b_proj.bias": "model-00002-of-00002.safetensors", |
|
"transformer.h.16.attn.b_proj.weight": "model-00002-of-00002.safetensors", |
|
"transformer.h.16.attn.dt_bias": "model-00002-of-00002.safetensors", |
|
"transformer.h.16.attn.g_norm_swish_gate.weight": "model-00002-of-00002.safetensors", |
|
"transformer.h.16.attn.g_proj.weight": "model-00002-of-00002.safetensors", |
|
"transformer.h.16.attn.gk_proj.weight": "model-00002-of-00002.safetensors", |
|
"transformer.h.16.attn.k_conv1d.weight": "model-00002-of-00002.safetensors", |
|
"transformer.h.16.attn.k_proj.weight": "model-00002-of-00002.safetensors", |
|
"transformer.h.16.attn.o_proj.weight": "model-00002-of-00002.safetensors", |
|
"transformer.h.16.attn.q_conv1d.weight": "model-00002-of-00002.safetensors", |
|
"transformer.h.16.attn.q_proj.weight": "model-00002-of-00002.safetensors", |
|
"transformer.h.16.attn.v_conv1d.weight": "model-00002-of-00002.safetensors", |
|
"transformer.h.16.attn.v_proj.weight": "model-00002-of-00002.safetensors", |
|
"transformer.h.16.mlp.swiglu.w1.weight": "model-00002-of-00002.safetensors", |
|
"transformer.h.16.mlp.swiglu.w2.weight": "model-00002-of-00002.safetensors", |
|
"transformer.h.16.mlp.swiglu.w3.weight": "model-00002-of-00002.safetensors", |
|
"transformer.h.16.norm_1.weight": "model-00002-of-00002.safetensors", |
|
"transformer.h.16.norm_2.weight": "model-00002-of-00002.safetensors", |
|
"transformer.h.2.attn.A_log": "model-00001-of-00002.safetensors", |
|
"transformer.h.2.attn.D": "model-00001-of-00002.safetensors", |
|
"transformer.h.2.attn.b_proj.bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.2.attn.b_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.2.attn.dt_bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.2.attn.g_norm_swish_gate.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.2.attn.g_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.2.attn.gk_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.2.attn.k_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.2.attn.k_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.2.attn.o_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.2.attn.q_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.2.attn.q_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.2.attn.v_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.2.attn.v_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.2.mlp.swiglu.w1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.2.mlp.swiglu.w2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.2.mlp.swiglu.w3.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.2.norm_1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.2.norm_2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.3.attn.A_log": "model-00001-of-00002.safetensors", |
|
"transformer.h.3.attn.D": "model-00001-of-00002.safetensors", |
|
"transformer.h.3.attn.b_proj.bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.3.attn.b_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.3.attn.dt_bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.3.attn.g_norm_swish_gate.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.3.attn.g_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.3.attn.gk_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.3.attn.k_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.3.attn.k_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.3.attn.o_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.3.attn.q_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.3.attn.q_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.3.attn.v_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.3.attn.v_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.3.mlp.swiglu.w1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.3.mlp.swiglu.w2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.3.mlp.swiglu.w3.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.3.norm_1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.3.norm_2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.4.attn.A_log": "model-00001-of-00002.safetensors", |
|
"transformer.h.4.attn.D": "model-00001-of-00002.safetensors", |
|
"transformer.h.4.attn.b_proj.bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.4.attn.b_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.4.attn.dt_bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.4.attn.g_norm_swish_gate.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.4.attn.g_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.4.attn.gk_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.4.attn.k_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.4.attn.k_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.4.attn.o_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.4.attn.q_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.4.attn.q_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.4.attn.v_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.4.attn.v_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.4.mlp.swiglu.w1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.4.mlp.swiglu.w2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.4.mlp.swiglu.w3.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.4.norm_1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.4.norm_2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.5.attn.A_log": "model-00001-of-00002.safetensors", |
|
"transformer.h.5.attn.D": "model-00001-of-00002.safetensors", |
|
"transformer.h.5.attn.b_proj.bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.5.attn.b_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.5.attn.dt_bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.5.attn.g_norm_swish_gate.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.5.attn.g_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.5.attn.gk_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.5.attn.k_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.5.attn.k_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.5.attn.o_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.5.attn.q_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.5.attn.q_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.5.attn.v_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.5.attn.v_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.5.mlp.swiglu.w1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.5.mlp.swiglu.w2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.5.mlp.swiglu.w3.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.5.norm_1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.5.norm_2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.6.attn.A_log": "model-00001-of-00002.safetensors", |
|
"transformer.h.6.attn.D": "model-00001-of-00002.safetensors", |
|
"transformer.h.6.attn.b_proj.bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.6.attn.b_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.6.attn.dt_bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.6.attn.g_norm_swish_gate.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.6.attn.g_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.6.attn.gk_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.6.attn.k_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.6.attn.k_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.6.attn.o_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.6.attn.q_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.6.attn.q_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.6.attn.v_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.6.attn.v_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.6.mlp.swiglu.w1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.6.mlp.swiglu.w2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.6.mlp.swiglu.w3.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.6.norm_1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.6.norm_2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.7.attn.A_log": "model-00001-of-00002.safetensors", |
|
"transformer.h.7.attn.D": "model-00001-of-00002.safetensors", |
|
"transformer.h.7.attn.b_proj.bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.7.attn.b_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.7.attn.dt_bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.7.attn.g_norm_swish_gate.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.7.attn.g_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.7.attn.gk_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.7.attn.k_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.7.attn.k_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.7.attn.o_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.7.attn.q_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.7.attn.q_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.7.attn.v_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.7.attn.v_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.7.mlp.swiglu.w1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.7.mlp.swiglu.w2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.7.mlp.swiglu.w3.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.7.norm_1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.7.norm_2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.8.attn.A_log": "model-00001-of-00002.safetensors", |
|
"transformer.h.8.attn.D": "model-00001-of-00002.safetensors", |
|
"transformer.h.8.attn.b_proj.bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.8.attn.b_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.8.attn.dt_bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.8.attn.g_norm_swish_gate.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.8.attn.g_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.8.attn.gk_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.8.attn.k_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.8.attn.k_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.8.attn.o_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.8.attn.q_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.8.attn.q_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.8.attn.v_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.8.attn.v_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.8.mlp.swiglu.w1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.8.mlp.swiglu.w2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.8.mlp.swiglu.w3.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.8.norm_1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.8.norm_2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.9.attn.A_log": "model-00001-of-00002.safetensors", |
|
"transformer.h.9.attn.D": "model-00001-of-00002.safetensors", |
|
"transformer.h.9.attn.b_proj.bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.9.attn.b_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.9.attn.dt_bias": "model-00001-of-00002.safetensors", |
|
"transformer.h.9.attn.g_norm_swish_gate.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.9.attn.g_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.9.attn.gk_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.9.attn.k_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.9.attn.k_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.9.attn.o_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.9.attn.q_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.9.attn.q_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.9.attn.v_conv1d.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.9.attn.v_proj.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.9.mlp.swiglu.w1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.9.mlp.swiglu.w2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.9.mlp.swiglu.w3.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.9.norm_1.weight": "model-00001-of-00002.safetensors", |
|
"transformer.h.9.norm_2.weight": "model-00001-of-00002.safetensors", |
|
"transformer.ln_f.weight": "model-00002-of-00002.safetensors" |
|
} |
|
} |
|
|