Load bug: slow and mistake?

#3
by XLH - opened

model = AutoModelForCausalLM.from_pretrained(
"Refs/Weights/Qwen-14B-Chat-Int4",
device_map='cuda:1',
trust_remote_code=True,
).eval()

Some weights of the model checkpoint at Refs/Weights/Qwen-14B-Chat-Int4 were not used when initializing QWenLMHeadModel: ['transformer.h.12.attn.c_proj.bias', 'transformer.h.10.attn.c_proj.bias', 'transformer.h.14.mlp.w2.qzeros', 'transformer.h.17.attn.c_proj.qweight', 'transformer.h.14.mlp.c_proj.bias', 'transformer.h.19.attn.c_proj.scales', 'transformer.h.20.mlp.w1.scales', 'transformer.h.22.mlp.c_proj.bias', 'transformer.h.37.attn.c_attn.scales', 'transformer.h.5.mlp.w1.scales', 'transformer.h.18.attn.c_proj.qzeros', 'transformer.h.5.mlp.w2.qweight', 'transformer.h.37.mlp.w1.qzeros', 'transformer.h.30.mlp.w1.g_idx', 'transformer.h.6.mlp.w2.scales', 'transformer.h.7.attn.c_proj.qzeros', 'transformer.h.28.mlp.w1.g_idx', 'transformer.h.23.mlp.w1.g_idx', 'transformer.h.34.mlp.w2.bias', 'transformer.h.36.mlp.w1.qweight', 'transformer.h.20.attn.c_proj.bias', 'transformer.h.37.mlp.w2.scales', 'transformer.h.5.mlp.w2.g_idx', 'transformer.h.6.attn.c_proj.bias', 'transformer.h.21.attn.c_proj.scales', 'transformer.h.11.mlp.c_proj.qweight', 'transformer.h.12.attn.c_proj.qweight', 'transformer.h.26.mlp.w1.qzeros', 'transformer.h.5.mlp.w2.bias', 'transformer.h.37.mlp.c_proj.qweight', 'transformer.h.38.mlp.w2.qzeros', 'transformer.h.0.attn.c_proj.bias', 'transformer.h.37.mlp.w2.qzeros', 'transformer.h.2.mlp.w2.qweight', 'transformer.h.19.attn.c_proj.g_idx', 'transformer.h.15.attn.c_attn.qzeros', 'transformer.h.20.mlp.c_proj.bias', 'transformer.h.27.mlp.c_proj.bias', 'transformer.h.6.attn.c_proj.qweight', 'transformer.h.26.attn.c_proj.scales', 'transformer.h.33.attn.c_attn.scales', 'transformer.h.12.mlp.c_proj.bias', 'transformer.h.12.attn.c_attn.g_idx', 'transformer.h.34.attn.c_attn.scales', 'transformer.h.34.attn.c_proj.bias', 'transformer.h.39.mlp.w1.scales', 'transformer.h.32.mlp.c_proj.qweight', 'transformer.h.39.mlp.c_proj.bias', 'transformer.h.8.mlp.w1.g_idx', 'transformer.h.24.attn.c_proj.bias', 'transformer.h.18.mlp.c_proj.scales', 'transformer.h.11.mlp.c_proj.scales', 'transformer.h.14.mlp.w2.scales', 'transformer.h.25.mlp.w2.qweight', 'transformer.h.6.attn.c_attn.qweight', 'transformer.h.18.attn.c_proj.g_idx', 'transformer.h.23.mlp.w2.qweight', 'transformer.h.17.mlp.w2.qzeros', 'transformer.h.29.mlp.c_proj.scales', 'transformer.h.17.mlp.c_proj.qweight', 'transformer.h.33.mlp.w1.scales', 'transformer.h.4.mlp.w1.qweight', 'transformer.h.0.mlp.w1.qweight', 'transformer.h.11.attn.c_attn.qweight', 'transformer.h.28.mlp.w2.qzeros', 'transformer.h.3.mlp.c_proj.g_idx', 'transformer.h.3.attn.c_attn.qzeros', 'transformer.h.4.mlp.c_proj.g_idx', 'transformer.h.12.mlp.w1.qzeros', 'transformer.h.31.mlp.c_proj.g_idx', 'transformer.h.16.attn.c_proj.qzeros', 'transformer.h.38.mlp.w1.scales', 'transformer.h.37.mlp.w2.qweight', 'transformer.h.29.attn.c_proj.scales', 'transformer.h.24.attn.c_attn.scales', 'transformer.h.5.mlp.c_proj.bias', 'transformer.h.19.mlp.w1.bias', 'transformer.h.13.mlp.c_proj.qzeros', 'transformer.h.36.mlp.w1.bias', 'transformer.h.7.mlp.w2.bias', 'transformer.h.36.attn.c_attn.qweight', 'transformer.h.18.attn.c_attn.scales', 'transformer.h.17.mlp.w2.g_idx', 'transformer.h.38.attn.c_attn.qzeros', 'transformer.h.0.mlp.w2.scales', 'transformer.h.18.attn.c_proj.scales', 'transformer.h.29.attn.c_proj.qweight', 'transformer.h.35.mlp.c_proj.g_idx', 'transformer.h.11.mlp.w2.qzeros', 'transformer.h.30.mlp.c_proj.bias', 'transformer.h.32.attn.c_proj.g_idx', 'transformer.h.37.mlp.w2.g_idx', 'transformer.h.32.attn.c_attn.qzeros', 'transformer.h.1.mlp.c_proj.qweight', 'transformer.h.15.mlp.w2.g_idx', 'transformer.h.31.mlp.w1.qzeros', 'transformer.h.28.attn.c_attn.qweight', 'transformer.h.3.mlp.w2.bias', 'transformer.h.35.attn.c_proj.qweight', 'transformer.h.16.mlp.c_proj.scales', 'transformer.h.38.attn.c_proj.qweight', 'transformer.h.5.mlp.w2.scales', 'transformer.h.22.mlp.w1.g_idx', 'transformer.h.31.mlp.w2.qzeros', 'transformer.h.0.mlp.w1.g_idx', 'transformer.h.17.mlp.c_proj.bias', 'transformer.h.19.mlp.w1.scales', 'transformer.h.34.mlp.w1.qzeros', 'transformer.h.4.attn.c_attn.qzeros', 'transformer.h.36.attn.c_attn.g_idx', 'transformer.h.28.mlp.w1.bias', 'transformer.h.39.attn.c_proj.bias', 'transformer.h.2.mlp.w1.scales', 'transformer.h.30.mlp.w1.qweight', 'transformer.h.34.attn.c_attn.qweight', 'transformer.h.7.attn.c_attn.g_idx', 'transformer.h.30.mlp.w1.bias', 'transformer.h.13.attn.c_proj.scales', 'transformer.h.19.mlp.c_proj.g_idx', 'transformer.h.13.mlp.w2.g_idx', 'transformer.h.24.mlp.w1.qweight', 'transformer.h.5.attn.c_proj.g_idx', 'transformer.h.16.mlp.w1.qzeros', 'transformer.h.38.mlp.w2.qweight', 'transformer.h.0.attn.c_proj.g_idx', 'transformer.h.22.attn.c_proj.qzeros', 'transformer.h.14.mlp.c_proj.qzeros', 'transformer.h.34.mlp.w2.qweight', 'transformer.h.13.attn.c_attn.qweight', 'transformer.h.35.attn.c_proj.g_idx', 'transformer.h.13.mlp.w1.qweight', 'transformer.h.3.mlp.w2.g_idx', 'transformer.h.36.attn.c_proj.qweight', 'transformer.h.24.mlp.w2.qzeros', 'transformer.h.39.mlp.c_proj.qweight', 'transformer.h.36.mlp.w2.g_idx', 'transformer.h.6.mlp.w2.bias', 'transformer.h.39.mlp.w2.qweight', 'transformer.h.5.attn.c_proj.scales', 'transformer.h.25.attn.c_attn.qzeros', 'transformer.h.17.mlp.w2.bias', 'transformer.h.6.mlp.c_proj.qzeros', 'transformer.h.15.mlp.w2.qweight', 'transformer.h.20.mlp.c_proj.g_idx', 'transformer.h.36.mlp.w1.qzeros', 'transformer.h.25.mlp.w1.scales', 'transformer.h.36.mlp.c_proj.qweight', 'transformer.h.25.mlp.c_proj.scales', 'transformer.h.36.mlp.w2.scales', 'transformer.h.32.mlp.w2.qzeros', 'transformer.h.20.attn.c_proj.scales', 'transformer.h.33.mlp.c_proj.scales', 'transformer.h.8.mlp.w1.scales', 'transformer.h.13.mlp.c_proj.g_idx', 'transformer.h.30.mlp.w2.bias', 'transformer.h.36.mlp.c_proj.bias', 'transformer.h.2.mlp.w2.g_idx', 'transformer.h.10.mlp.w1.scales', 'transformer.h.13.mlp.w2.qweight', 'transformer.h.23.mlp.c_proj.g_idx', 'transformer.h.38.mlp.w1.qzeros', 'transformer.h.18.mlp.w1.g_idx', 'transformer.h.11.attn.c_proj.g_idx', 'transformer.h.35.attn.c_proj.scales', 'transformer.h.8.attn.c_proj.scales', 'transformer.h.30.mlp.c_proj.qzeros', 'transformer.h.30.attn.c_attn.qweight', 'transformer.h.20.attn.c_attn.qweight', 'transformer.h.9.attn.c_proj.g_idx', 'transformer.h.6.mlp.c_proj.qweight', 'transformer.h.6.attn.c_proj.g_idx', 'transformer.h.12.mlp.w2.qzeros', 'transformer.h.1.attn.c_proj.qweight', 'transformer.h.19.attn.c_proj.qweight', 'transformer.h.25.mlp.w2.scales', 'transformer.h.18.mlp.w1.scales', 'transformer.h.14.mlp.c_proj.g_idx', 'transformer.h.29.attn.c_proj.g_idx', 'transformer.h.32.mlp.w1.qzeros', 'transformer.h.35.mlp.c_proj.qweight', 'transformer.h.24.mlp.w1.g_idx', 'transformer.h.9.attn.c_attn.g_idx', 'transformer.h.38.mlp.c_proj.scales', 'transformer.h.24.mlp.w1.scales', 'transformer.h.21.attn.c_attn.g_idx', 'transformer.h.17.mlp.c_proj.qzeros', 'transformer.h.23.mlp.w2.scales', 'transformer.h.21.attn.c_attn.qzeros', 'transformer.h.30.mlp.w2.g_idx', 'transformer.h.32.attn.c_attn.g_idx', 'transformer.h.31.mlp.w2.bias', 'transformer.h.19.mlp.w2.bias', 'transformer.h.37.mlp.w2.bias', 'transformer.h.38.mlp.w1.bias', 'transformer.h.35.mlp.w2.qweight', 'transformer.h.35.attn.c_attn.qzeros', 'transformer.h.35.mlp.w1.qzeros', 'transformer.h.6.mlp.c_proj.scales', 'transformer.h.9.attn.c_attn.scales', 'transformer.h.17.attn.c_attn.scales', 'transformer.h.21.mlp.w2.qzeros', 'transformer.h.24.mlp.w1.bias', 'transformer.h.22.attn.c_proj.scales', 'transformer.h.7.attn.c_attn.scales', 'transformer.h.32.mlp.w2.bias', 'transformer.h.24.mlp.w2.bias', 'transformer.h.12.mlp.w1.qweight', 'transformer.h.8.attn.c_attn.scales', 'transformer.h.21.attn.c_proj.g_idx', 'transformer.h.28.mlp.c_proj.g_idx', 'transformer.h.14.mlp.w2.qweight', 'transformer.h.29.mlp.c_proj.bias', 'transformer.h.29.mlp.w1.qzeros', 'transformer.h.37.attn.c_proj.g_idx', 'transformer.h.35.mlp.c_proj.bias', 'transformer.h.37.mlp.c_proj.bias', 'transformer.h.7.mlp.w1.qzeros', 'transformer.h.11.attn.c_proj.qweight', 'transformer.h.1.mlp.w1.g_idx', 'transformer.h.39.attn.c_proj.qweight', 'transformer.h.11.mlp.w1.scales', 'transformer.h.10.mlp.c_proj.bias', 'transformer.h.15.mlp.w2.scales', 'transformer.h.5.mlp.c_proj.g_idx', 'transformer.h.24.mlp.w2.g_idx', 'transformer.h.37.mlp.w1.bias', 'transformer.h.17.attn.c_attn.g_idx', 'transformer.h.34.mlp.w1.g_idx', 'transformer.h.2.mlp.w1.bias', 'transformer.h.10.attn.c_attn.scales', 'transformer.h.1.attn.c_proj.bias', 'transformer.h.1.mlp.c_proj.g_idx', 'transformer.h.20.mlp.w2.scales', 'transformer.h.22.mlp.c_proj.scales', 'transformer.h.39.mlp.w1.qzeros', 'transformer.h.8.mlp.w2.qweight', 'transformer.h.37.mlp.w1.g_idx', 'transformer.h.9.mlp.w1.bias', 'transformer.h.31.attn.c_attn.scales', 'transformer.h.35.attn.c_attn.scales', 'transformer.h.7.mlp.w2.qweight', 'transformer.h.28.attn.c_proj.scales', 'transformer.h.25.attn.c_attn.g_idx', 'transformer.h.5.attn.c_attn.qzeros', 'transformer.h.21.attn.c_proj.bias', 'transformer.h.24.attn.c_proj.scales', 'transformer.h.2.attn.c_proj.qzeros', 'transformer.h.2.mlp.c_proj.scales', 'transformer.h.18.mlp.w2.bias', 'transformer.h.22.mlp.w1.scales', 'transformer.h.24.mlp.c_proj.scales', 'transformer.h.36.mlp.w2.qzeros', 'transformer.h.2.mlp.c_proj.bias', 'transformer.h.16.mlp.w2.g_idx', 'transformer.h.21.mlp.w2.bias', 'transformer.h.24.mlp.w2.qweight', 'transformer.h.33.attn.c_proj.g_idx', 'transformer.h.0.attn.c_proj.scales', 'transformer.h.16.mlp.w1.qweight', 'transformer.h.30.mlp.c_proj.g_idx', 'transformer.h.10.mlp.w2.scales', 'transformer.h.14.mlp.w1.g_idx', 'transformer.h.4.attn.c_proj.g_idx', 'transformer.h.0.mlp.c_proj.qweight', 'transformer.h.1.attn.c_attn.g_idx', 'transformer.h.36.attn.c_proj.qzeros', 'transformer.h.10.mlp.c_proj.g_idx', 'transformer.h.33.mlp.w2.qzeros', 'transformer.h.33.mlp.c_proj.qweight', 'transformer.h.24.attn.c_proj.qzeros', 'transformer.h.33.attn.c_proj.qzeros', 'transformer.h.28.mlp.w2.qweight', 'transformer.h.38.attn.c_attn.scales', 'transformer.h.15.mlp.c_proj.qweight', 'transformer.h.8.mlp.c_proj.qweight', 'transformer.h.16.attn.c_proj.bias', 'transformer.h.26.mlp.c_proj.bias', 'transformer.h.11.mlp.w2.bias', 'transformer.h.9.mlp.w1.qzeros', 'transformer.h.27.attn.c_proj.qzeros', 'transformer.h.18.attn.c_attn.qzeros', 'transformer.h.38.mlp.c_proj.qzeros', 'transformer.h.2.attn.c_proj.qweight', 'transformer.h.34.mlp.c_proj.qzeros', 'transformer.h.10.attn.c_attn.qzeros', 'transformer.h.25.mlp.c_proj.g_idx', 'transformer.h.20.mlp.c_proj.qzeros', 'transformer.h.10.mlp.w1.g_idx', 'transformer.h.13.attn.c_proj.bias', 'transformer.h.17.mlp.w2.scales', 'transformer.h.12.mlp.w2.qweight', 'transformer.h.2.mlp.w2.scales', 'transformer.h.3.mlp.w1.g_idx', 'transformer.h.31.mlp.w2.qweight', 'transformer.h.3.attn.c_proj.g_idx', 'transformer.h.36.attn.c_proj.scales', 'transformer.h.14.mlp.w1.qzeros', 'transformer.h.16.attn.c_proj.qweight', 'transformer.h.25.mlp.w2.bias', 'transformer.h.27.mlp.c_proj.scales', 'transformer.h.22.attn.c_attn.qzeros', 'transformer.h.38.mlp.c_proj.g_idx', 'transformer.h.27.mlp.w1.bias', 'transformer.h.9.attn.c_proj.qweight', 'transformer.h.18.mlp.w2.g_idx', 'transformer.h.31.mlp.c_proj.bias', 'transformer.h.23.mlp.c_proj.scales', 'transformer.h.27.mlp.w1.g_idx', 'transformer.h.10.mlp.c_proj.scales', 'transformer.h.38.mlp.w2.bias', 'transformer.h.30.mlp.w1.qzeros', 'transformer.h.27.mlp.w2.bias', 'transformer.h.4.mlp.w1.bias', 'transformer.h.7.attn.c_proj.scales', 'transformer.h.22.attn.c_proj.qweight', 'transformer.h.26.mlp.w1.scales', 'transformer.h.27.mlp.w2.qzeros', 'transformer.h.28.attn.c_attn.g_idx', 'transformer.h.0.mlp.w2.qzeros', 'transformer.h.1.attn.c_proj.g_idx', 'transformer.h.32.mlp.w1.scales', 'transformer.h.31.mlp.c_proj.qweight', 'transformer.h.7.attn.c_proj.qweight', 'transformer.h.9.mlp.w1.g_idx', 'transformer.h.33.mlp.w2.scales', 'transformer.h.18.attn.c_proj.qweight', 'transformer.h.32.attn.c_proj.bias', 'transformer.h.0.attn.c_proj.qzeros', 'transformer.h.21.mlp.w2.qweight', 'transformer.h.3.attn.c_attn.scales', 'transformer.h.21.mlp.w1.qweight', 'transformer.h.29.mlp.c_proj.g_idx', 'transformer.h.7.mlp.w2.g_idx', 'transformer.h.17.mlp.w1.bias', 'transformer.h.20.mlp.w2.qzeros', 'transformer.h.0.mlp.w2.qweight', 'transformer.h.37.attn.c_attn.qzeros', 'transformer.h.10.attn.c_attn.qweight', 'transformer.h.3.mlp.w2.qzeros', 'transformer.h.4.attn.c_proj.bias', 'transformer.h.3.attn.c_attn.qweight', 'transformer.h.30.attn.c_proj.scales', 'transformer.h.8.mlp.w2.qzeros', 'transformer.h.10.mlp.w2.bias', 'transformer.h.28.mlp.c_proj.qweight', 'transformer.h.34.mlp.w1.qweight', 'transformer.h.6.mlp.w1.g_idx', 'transformer.h.6.mlp.w1.qweight', 'transformer.h.10.attn.c_attn.g_idx', 'transformer.h.25.attn.c_attn.qweight', 'transformer.h.3.attn.c_proj.qweight', 'transformer.h.4.mlp.w2.qzeros', 'transformer.h.17.attn.c_proj.scales', 'transformer.h.13.mlp.w1.scales', 'transformer.h.18.mlp.w2.qzeros', 'transformer.h.1.attn.c_proj.scales', 'transformer.h.19.mlp.w2.g_idx', 'transformer.h.33.mlp.c_proj.g_idx', 'transformer.h.5.mlp.w1.bias', 'transformer.h.31.attn.c_attn.qzeros', 'transformer.h.19.mlp.c_proj.bias', 'transformer.h.14.mlp.w2.bias', 'transformer.h.32.attn.c_proj.scales', 'transformer.h.16.attn.c_proj.g_idx', 'transformer.h.9.attn.c_attn.qweight', 'transformer.h.16.attn.c_attn.qzeros', 'transformer.h.32.mlp.w2.scales', 'transformer.h.9.attn.c_proj.bias', 'transformer.h.26.mlp.w1.qweight', 'transformer.h.27.mlp.c_proj.qweight', 'transformer.h.26.attn.c_proj.bias', 'transformer.h.34.mlp.c_proj.scales', 'transformer.h.8.mlp.c_proj.bias', 'transformer.h.23.mlp.w2.qzeros', 'transformer.h.14.mlp.c_proj.scales', 'transformer.h.36.attn.c_proj.bias', 'transformer.h.25.mlp.c_proj.qweight', 'transformer.h.26.attn.c_attn.qzeros', 'transformer.h.23.attn.c_proj.qzeros', 'transformer.h.2.mlp.w2.bias', 'transformer.h.38.attn.c_proj.bias', 'transformer.h.14.attn.c_attn.qzeros', 'transformer.h.2.mlp.w1.qzeros', 'transformer.h.12.mlp.w2.bias', 'transformer.h.20.mlp.w1.qweight', 'transformer.h.20.mlp.w1.bias', 'transformer.h.31.attn.c_attn.qweight', 'transformer.h.19.mlp.w2.qweight', 'transformer.h.2.mlp.w1.g_idx', 'transformer.h.38.mlp.c_proj.bias', 'transformer.h.1.mlp.c_proj.bias', 'transformer.h.1.attn.c_attn.scales', 'transformer.h.28.attn.c_proj.bias', 'transformer.h.25.mlp.w1.qzeros', 'transformer.h.18.mlp.w1.qzeros', 'transformer.h.27.mlp.c_proj.g_idx', 'transformer.h.25.attn.c_proj.g_idx', 'transformer.h.30.attn.c_proj.qweight', 'transformer.h.33.mlp.w1.bias', 'transformer.h.4.mlp.w2.g_idx', 'transformer.h.15.mlp.w2.qzeros', 'transformer.h.21.mlp.w1.qzeros', 'transformer.h.30.mlp.w2.qweight', 'transformer.h.30.attn.c_proj.qzeros', 'transformer.h.12.mlp.c_proj.qweight', 'transformer.h.15.mlp.c_proj.scales', 'transformer.h.16.mlp.w1.g_idx', 'transformer.h.33.mlp.c_proj.bias', 'transformer.h.10.mlp.w1.qzeros', 'transformer.h.26.attn.c_proj.qzeros', 'transformer.h.29.mlp.c_proj.qweight', 'transformer.h.1.mlp.w1.qweight', 'transformer.h.32.mlp.w1.qweight', 'transformer.h.32.mlp.w2.qweight', 'transformer.h.39.mlp.c_proj.scales', 'transformer.h.25.mlp.c_proj.qzeros', 'transformer.h.8.mlp.w1.qweight', 'transformer.h.28.attn.c_proj.qweight', 'transformer.h.23.mlp.w1.qzeros', 'transformer.h.32.mlp.w1.bias', 'transformer.h.5.mlp.c_proj.scales', 'transformer.h.23.attn.c_proj.bias', 'transformer.h.26.mlp.c_proj.scales', 'transformer.h.8.mlp.w2.scales', 'transformer.h.4.mlp.w1.g_idx', 'transformer.h.9.mlp.w1.qweight', 'transformer.h.38.attn.c_attn.qweight', 'transformer.h.7.attn.c_proj.g_idx', 'transformer.h.0.mlp.c_proj.qzeros', 'transformer.h.39.attn.c_proj.g_idx', 'transformer.h.8.attn.c_attn.qweight', 'transformer.h.38.attn.c_proj.scales', 'transformer.h.9.mlp.w2.scales', 'transformer.h.1.mlp.w2.g_idx', 'transformer.h.20.attn.c_proj.qweight', 'transformer.h.28.mlp.w2.scales', 'transformer.h.19.mlp.w2.scales', 'transformer.h.24.attn.c_attn.qzeros', 'transformer.h.37.mlp.c_proj.qzeros', 'transformer.h.12.mlp.c_proj.scales', 'transformer.h.36.mlp.w2.qweight', 'transformer.h.7.mlp.c_proj.qzeros', 'transformer.h.8.attn.c_proj.bias', 'transformer.h.1.mlp.w2.qweight', 'transformer.h.17.mlp.w1.scales', 'transformer.h.0.attn.c_attn.g_idx', 'transformer.h.12.attn.c_attn.scales', 'transformer.h.16.attn.c_attn.scales', 'transformer.h.9.mlp.w1.scales', 'transformer.h.27.attn.c_proj.g_idx', 'transformer.h.22.attn.c_proj.g_idx', 'transformer.h.17.mlp.w1.qzeros', 'transformer.h.12.mlp.c_proj.qzeros', 'transformer.h.25.mlp.w1.qweight', 'transformer.h.21.mlp.c_proj.g_idx', 'transformer.h.28.attn.c_attn.qzeros', 'transformer.h.29.mlp.w2.g_idx', 'transformer.h.10.mlp.c_proj.qweight', 'transformer.h.24.attn.c_attn.g_idx', 'transformer.h.20.mlp.w1.g_idx', 'transformer.h.0.mlp.w1.bias', 'transformer.h.16.mlp.w2.scales', 'transformer.h.13.mlp.w1.bias', 'transformer.h.18.mlp.w2.scales', 'transformer.h.30.attn.c_attn.scales', 'transformer.h.24.attn.c_attn.qweight', 'transformer.h.33.attn.c_attn.qzeros', 'transformer.h.12.attn.c_attn.qweight', 'transformer.h.19.mlp.w1.qzeros', 'transformer.h.28.mlp.c_proj.bias', 'transformer.h.33.attn.c_attn.qweight', 'transformer.h.35.mlp.w1.g_idx', 'transformer.h.36.attn.c_proj.g_idx', 'transformer.h.21.mlp.c_proj.bias', 'transformer.h.27.mlp.c_proj.qzeros', 'transformer.h.35.attn.c_attn.qweight', 'transformer.h.3.mlp.w1.qweight', 'transformer.h.10.attn.c_proj.qweight', 'transformer.h.21.mlp.c_proj.qzeros', 'transformer.h.20.attn.c_attn.g_idx', 'transformer.h.14.attn.c_attn.qweight', 'transformer.h.19.mlp.c_proj.scales', 'transformer.h.32.attn.c_attn.scales', 'transformer.h.31.attn.c_attn.g_idx', 'transformer.h.33.mlp.w2.bias', 'transformer.h.29.mlp.w1.bias', 'transformer.h.12.attn.c_proj.scales', 'transformer.h.1.attn.c_attn.qweight', 'transformer.h.18.attn.c_attn.qweight', 'transformer.h.26.mlp.w2.g_idx', 'transformer.h.27.attn.c_proj.bias', 'transformer.h.22.attn.c_proj.bias', 'transformer.h.26.attn.c_attn.g_idx', 'transformer.h.5.mlp.w1.qzeros', 'transformer.h.7.mlp.w1.qweight', 'transformer.h.22.attn.c_attn.qweight', 'transformer.h.13.attn.c_proj.qweight', 'transformer.h.13.mlp.w2.qzeros', 'transformer.h.21.attn.c_attn.qweight', 'transformer.h.26.mlp.w1.g_idx', 'transformer.h.17.mlp.w1.qweight', 'transformer.h.13.attn.c_attn.scales', 'transformer.h.17.attn.c_proj.qzeros', 'transformer.h.17.mlp.c_proj.g_idx', 'transformer.h.29.mlp.w2.scales', 'transformer.h.13.mlp.c_proj.scales', 'transformer.h.30.mlp.w2.scales', 'transformer.h.6.mlp.w1.bias', 'transformer.h.25.mlp.w2.g_idx', 'transformer.h.34.mlp.c_proj.bias', 'transformer.h.27.attn.c_proj.qweight', 'transformer.h.32.mlp.c_proj.scales', 'transformer.h.6.mlp.w1.qzeros', 'transformer.h.16.mlp.w1.bias', 'transformer.h.4.mlp.w2.scales', 'transformer.h.2.attn.c_attn.qweight', 'transformer.h.0.mlp.w1.scales', 'transformer.h.15.mlp.w1.scales', 'transformer.h.12.attn.c_proj.qzeros', 'transformer.h.11.mlp.c_proj.qzeros', 'transformer.h.5.mlp.c_proj.qzeros', 'transformer.h.8.mlp.c_proj.qzeros', 'transformer.h.9.mlp.w2.g_idx', 'transformer.h.12.attn.c_proj.g_idx', 'transformer.h.34.mlp.c_proj.qweight', 'transformer.h.22.mlp.w2.qweight', 'transformer.h.4.attn.c_attn.g_idx', 'transformer.h.12.mlp.w1.g_idx', 'transformer.h.13.attn.c_proj.g_idx', 'transformer.h.16.attn.c_attn.qweight', 'transformer.h.11.mlp.w1.qzeros', 'transformer.h.19.mlp.c_proj.qzeros', 'transformer.h.33.attn.c_proj.bias', 'transformer.h.7.mlp.c_proj.qweight', 'transformer.h.10.attn.c_proj.qzeros', 'transformer.h.34.attn.c_proj.scales', 'transformer.h.33.attn.c_proj.scales', 'transformer.h.15.attn.c_proj.g_idx', 'transformer.h.32.mlp.c_proj.qzeros', 'transformer.h.35.mlp.w1.qweight', 'transformer.h.15.attn.c_attn.qweight', 'transformer.h.26.attn.c_proj.g_idx', 'transformer.h.36.mlp.w1.scales', 'transformer.h.38.mlp.w2.g_idx', 'transformer.h.5.mlp.c_proj.qweight', 'transformer.h.1.mlp.w1.qzeros', 'transformer.h.29.attn.c_proj.qzeros', 'transformer.h.32.mlp.w2.g_idx', 'transformer.h.23.attn.c_attn.qzeros', 'transformer.h.23.mlp.w1.qweight', 'transformer.h.3.mlp.w1.bias', 'transformer.h.35.mlp.w2.g_idx', 'transformer.h.21.mlp.w2.scales', 'transformer.h.20.mlp.w1.qzeros', 'transformer.h.7.mlp.w2.qzeros', 'transformer.h.31.mlp.w2.scales', 'transformer.h.36.mlp.c_proj.qzeros', 'transformer.h.25.mlp.w1.bias', 'transformer.h.37.mlp.w1.qweight', 'transformer.h.35.mlp.c_proj.scales', 'transformer.h.4.mlp.w1.qzeros', 'transformer.h.11.attn.c_attn.scales', 'transformer.h.17.attn.c_attn.qweight', 'transformer.h.21.attn.c_proj.qzeros', 'transformer.h.14.attn.c_attn.scales', 'transformer.h.32.mlp.c_proj.bias', 'transformer.h.6.mlp.c_proj.bias', 'transformer.h.8.mlp.w2.bias', 'transformer.h.31.mlp.w1.bias', 'transformer.h.8.mlp.w2.g_idx', 'transformer.h.3.mlp.c_proj.bias', 'transformer.h.4.mlp.w2.bias', 'transformer.h.1.mlp.w1.scales', 'transformer.h.34.attn.c_proj.qzeros', 'transformer.h.7.mlp.w2.scales', 'transformer.h.7.mlp.c_proj.scales', 'transformer.h.9.attn.c_proj.qzeros', 'transformer.h.8.mlp.c_proj.g_idx', 'transformer.h.15.attn.c_attn.g_idx', 'transformer.h.27.mlp.w2.scales', 'transformer.h.2.mlp.c_proj.qweight', 'transformer.h.36.mlp.w2.bias', 'transformer.h.3.attn.c_proj.scales', 'transformer.h.1.attn.c_proj.qzeros', 'transformer.h.29.mlp.c_proj.qzeros', 'transformer.h.28.mlp.c_proj.qzeros', 'transformer.h.29.attn.c_attn.scales', 'transformer.h.17.attn.c_attn.qzeros', 'transformer.h.30.mlp.c_proj.scales', 'transformer.h.1.mlp.w1.bias', 'transformer.h.14.mlp.w1.qweight', 'transformer.h.6.mlp.w2.qzeros', 'transformer.h.12.mlp.w1.bias', 'transformer.h.29.mlp.w1.qweight', 'transformer.h.18.attn.c_proj.bias', 'transformer.h.25.mlp.w2.qzeros', 'transformer.h.30.mlp.w2.qzeros', 'transformer.h.13.attn.c_proj.qzeros', 'transformer.h.20.mlp.w2.qweight', 'transformer.h.25.attn.c_proj.scales', 'transformer.h.7.attn.c_attn.qzeros', 'transformer.h.39.mlp.w2.qzeros', 'transformer.h.26.mlp.c_proj.g_idx', 'transformer.h.25.attn.c_proj.qzeros', 'transformer.h.21.mlp.w1.g_idx', 'transformer.h.14.attn.c_proj.bias', 'transformer.h.36.attn.c_attn.qzeros', 'transformer.h.2.mlp.c_proj.qzeros', 'transformer.h.27.attn.c_attn.scales', 'transformer.h.16.mlp.c_proj.qzeros', 'transformer.h.38.mlp.w1.g_idx', 'transformer.h.23.attn.c_attn.scales', 'transformer.h.4.attn.c_proj.scales', 'transformer.h.38.attn.c_attn.g_idx', 'transformer.h.31.mlp.w2.g_idx', 'transformer.h.25.attn.c_proj.qweight', 'transformer.h.14.attn.c_proj.qweight', 'transformer.h.29.mlp.w2.bias', 'transformer.h.19.attn.c_attn.scales', 'transformer.h.27.mlp.w2.qweight', 'transformer.h.21.mlp.w1.bias', 'transformer.h.26.mlp.c_proj.qzeros', 'transformer.h.7.mlp.w1.scales', 'transformer.h.5.attn.c_attn.g_idx', 'transformer.h.30.attn.c_attn.g_idx', 'transformer.h.26.mlp.w2.qzeros', 'transformer.h.11.mlp.w2.g_idx', 'transformer.h.15.mlp.c_proj.bias', 'transformer.h.14.attn.c_proj.qzeros', 'transformer.h.11.attn.c_proj.bias', 'transformer.h.33.mlp.w1.g_idx', 'transformer.h.35.mlp.c_proj.qzeros', 'transformer.h.10.attn.c_proj.g_idx', 'transformer.h.37.attn.c_attn.qweight', 'transformer.h.2.attn.c_attn.g_idx', 'transformer.h.10.mlp.w2.g_idx', 'transformer.h.37.mlp.c_proj.g_idx', 'transformer.h.38.mlp.w1.qweight', 'transformer.h.12.attn.c_attn.qzeros', 'transformer.h.29.attn.c_attn.g_idx', 'transformer.h.37.attn.c_proj.qzeros', 'transformer.h.6.mlp.w2.g_idx', 'transformer.h.26.mlp.c_proj.qweight', 'transformer.h.10.mlp.w2.qzeros', 'transformer.h.16.mlp.w2.qweight', 'transformer.h.24.mlp.w1.qzeros', 'transformer.h.31.attn.c_proj.bias', 'transformer.h.5.attn.c_proj.bias', 'transformer.h.3.attn.c_proj.bias', 'transformer.h.26.attn.c_attn.scales', 'transformer.h.9.mlp.w2.qzeros', 'transformer.h.29.attn.c_attn.qweight', 'transformer.h.9.mlp.c_proj.scales', 'transformer.h.19.attn.c_attn.qweight', 'transformer.h.23.attn.c_proj.qweight', 'transformer.h.31.mlp.w1.g_idx', 'transformer.h.4.attn.c_proj.qzeros', 'transformer.h.11.mlp.w1.bias', 'transformer.h.14.attn.c_proj.g_idx', 'transformer.h.20.mlp.c_proj.qweight', 'transformer.h.1.mlp.c_proj.scales', 'transformer.h.15.attn.c_proj.scales', 'transformer.h.15.attn.c_proj.qzeros', 'transformer.h.37.attn.c_proj.qweight', 'transformer.h.26.mlp.w2.qweight', 'transformer.h.5.attn.c_attn.qweight', 'transformer.h.6.mlp.w1.scales', 'transformer.h.7.attn.c_attn.qweight', 'transformer.h.25.attn.c_proj.bias', 'transformer.h.0.mlp.c_proj.g_idx', 'transformer.h.39.mlp.w2.scales', 'transformer.h.0.mlp.c_proj.bias', 'transformer.h.10.mlp.w1.bias', 'transformer.h.15.attn.c_proj.bias', 'transformer.h.18.mlp.c_proj.bias', 'transformer.h.10.mlp.w2.qweight', 'transformer.h.22.mlp.w2.qzeros', 'transformer.h.27.mlp.w2.g_idx', 'transformer.h.14.mlp.c_proj.qweight', 'transformer.h.11.attn.c_proj.scales', 'transformer.h.16.mlp.c_proj.qweight', 'transformer.h.12.mlp.c_proj.g_idx', 'transformer.h.19.attn.c_attn.g_idx', 'transformer.h.33.mlp.w1.qweight', 'transformer.h.2.attn.c_proj.scales', 'transformer.h.9.attn.c_attn.qzeros', 'transformer.h.19.mlp.w1.qweight', 'transformer.h.7.mlp.c_proj.bias', 'transformer.h.24.mlp.c_proj.g_idx', 'transformer.h.18.mlp.w2.qweight', 'transformer.h.21.mlp.c_proj.scales', 'transformer.h.36.mlp.c_proj.g_idx', 'transformer.h.11.mlp.w1.g_idx', 'transformer.h.8.mlp.c_proj.scales', 'transformer.h.39.attn.c_proj.qzeros', 'transformer.h.11.attn.c_attn.qzeros', 'transformer.h.1.attn.c_attn.qzeros', 'transformer.h.22.mlp.w2.g_idx', 'transformer.h.34.mlp.c_proj.g_idx', 'transformer.h.11.attn.c_proj.qzeros', 'transformer.h.21.mlp.c_proj.qweight', 'transformer.h.36.mlp.c_proj.scales', 'transformer.h.20.attn.c_attn.qzeros', 'transformer.h.23.mlp.w1.bias', 'transformer.h.11.mlp.w1.qweight', 'transformer.h.3.mlp.w2.qweight', 'transformer.h.24.mlp.c_proj.qzeros', 'transformer.h.35.mlp.w1.scales', 'transformer.h.9.mlp.c_proj.g_idx', 'transformer.h.29.mlp.w2.qweight', 'transformer.h.18.mlp.c_proj.qzeros', 'transformer.h.33.mlp.c_proj.qzeros', 'transformer.h.35.attn.c_attn.g_idx', 'transformer.h.6.mlp.w2.qweight', 'transformer.h.31.attn.c_proj.qweight', 'transformer.h.30.mlp.w1.scales', 'transformer.h.38.mlp.w2.scales', 'transformer.h.13.mlp.w1.qzeros', 'transformer.h.18.mlp.c_proj.qweight', 'transformer.h.3.attn.c_attn.g_idx', 'transformer.h.9.mlp.c_proj.bias', 'transformer.h.4.mlp.w2.qweight', 'transformer.h.14.mlp.w2.g_idx', 'transformer.h.4.attn.c_attn.scales', 'transformer.h.15.mlp.w1.qzeros', 'transformer.h.18.mlp.c_proj.g_idx', 'transformer.h.34.attn.c_attn.g_idx', 'transformer.h.34.mlp.w2.g_idx', 'transformer.h.2.mlp.w2.qzeros', 'transformer.h.15.mlp.w2.bias', 'transformer.h.24.attn.c_proj.qweight', 'transformer.h.33.attn.c_proj.qweight', 'transformer.h.27.mlp.w1.scales', 'transformer.h.22.mlp.c_proj.qzeros', 'transformer.h.24.attn.c_proj.g_idx', 'transformer.h.35.mlp.w2.qzeros', 'transformer.h.4.mlp.c_proj.qzeros', 'transformer.h.16.mlp.c_proj.bias', 'transformer.h.13.mlp.c_proj.bias', 'transformer.h.4.attn.c_proj.qweight', 'transformer.h.25.mlp.w1.g_idx', 'transformer.h.2.mlp.w1.qweight', 'transformer.h.39.mlp.c_proj.qzeros', 'transformer.h.23.mlp.w2.g_idx', 'transformer.h.29.mlp.w2.qzeros', 'transformer.h.37.attn.c_proj.bias', 'transformer.h.23.attn.c_attn.g_idx', 'transformer.h.5.mlp.w1.qweight', 'transformer.h.19.attn.c_attn.qzeros', 'transformer.h.34.attn.c_attn.qzeros', 'transformer.h.20.mlp.c_proj.scales', 'transformer.h.3.attn.c_proj.qzeros', 'transformer.h.20.mlp.w2.bias', 'transformer.h.4.mlp.c_proj.bias', 'transformer.h.28.mlp.w1.qweight', 'transformer.h.20.attn.c_attn.scales', 'transformer.h.32.attn.c_attn.qweight', 'transformer.h.39.mlp.w2.g_idx', 'transformer.h.29.attn.c_attn.qzeros', 'transformer.h.31.attn.c_proj.scales', 'transformer.h.17.mlp.c_proj.scales', 'transformer.h.11.mlp.w2.qweight', 'transformer.h.30.attn.c_proj.bias', 'transformer.h.5.attn.c_proj.qweight', 'transformer.h.17.mlp.w2.qweight', 'transformer.h.16.attn.c_attn.g_idx', 'transformer.h.2.attn.c_attn.scales', 'transformer.h.24.mlp.w2.scales', 'transformer.h.0.mlp.w2.g_idx', 'transformer.h.26.attn.c_attn.qweight', 'transformer.h.4.mlp.c_proj.scales', 'transformer.h.29.attn.c_proj.bias', 'transformer.h.22.mlp.c_proj.qweight', 'transformer.h.5.mlp.w1.g_idx', 'transformer.h.28.mlp.c_proj.scales', 'transformer.h.8.attn.c_proj.qzeros', 'transformer.h.31.attn.c_proj.qzeros', 'transformer.h.1.mlp.c_proj.qzeros', 'transformer.h.31.mlp.c_proj.qzeros', 'transformer.h.39.attn.c_proj.scales', 'transformer.h.38.mlp.c_proj.qweight', 'transformer.h.15.attn.c_attn.scales', 'transformer.h.37.attn.c_attn.g_idx', 'transformer.h.34.mlp.w2.qzeros', 'transformer.h.39.attn.c_attn.qzeros', 'transformer.h.11.mlp.c_proj.bias', 'transformer.h.8.attn.c_attn.qzeros', 'transformer.h.25.attn.c_attn.scales', 'transformer.h.12.mlp.w2.scales', 'transformer.h.1.mlp.w2.qzeros', 'transformer.h.16.mlp.w2.bias', 'transformer.h.20.mlp.w2.g_idx', 'transformer.h.23.attn.c_attn.qweight', 'transformer.h.22.attn.c_attn.g_idx', 'transformer.h.20.attn.c_proj.g_idx', 'transformer.h.6.attn.c_attn.scales', 'transformer.h.14.attn.c_attn.g_idx', 'transformer.h.11.mlp.w2.scales', 'transformer.h.28.mlp.w2.bias', 'transformer.h.13.mlp.w1.g_idx', 'transformer.h.3.mlp.w2.scales', 'transformer.h.35.attn.c_proj.bias', 'transformer.h.14.mlp.w1.bias', 'transformer.h.31.mlp.c_proj.scales', 'transformer.h.0.attn.c_proj.qweight', 'transformer.h.31.attn.c_proj.g_idx', 'transformer.h.24.mlp.c_proj.bias', 'transformer.h.39.mlp.w1.g_idx', 'transformer.h.27.mlp.w1.qweight', 'transformer.h.12.mlp.w1.scales', 'transformer.h.34.attn.c_proj.qweight', 'transformer.h.17.attn.c_proj.bias', 'transformer.h.39.mlp.w2.bias', 'transformer.h.2.attn.c_proj.bias', 'transformer.h.16.attn.c_proj.scales', 'transformer.h.23.attn.c_proj.scales', 'transformer.h.3.mlp.c_proj.qzeros', 'transformer.h.13.attn.c_attn.qzeros', 'transformer.h.32.attn.c_proj.qzeros', 'transformer.h.35.mlp.w2.bias', 'transformer.h.32.attn.c_proj.qweight', 'transformer.h.19.attn.c_proj.bias', 'transformer.h.23.attn.c_proj.g_idx', 'transformer.h.15.mlp.w1.bias', 'transformer.h.22.attn.c_attn.scales', 'transformer.h.34.mlp.w2.scales', 'transformer.h.5.attn.c_proj.qzeros', 'transformer.h.26.mlp.w2.bias', 'transformer.h.22.mlp.w2.scales', 'transformer.h.14.attn.c_proj.scales', 'transformer.h.30.mlp.c_proj.qweight', 'transformer.h.3.mlp.c_proj.qweight', 'transformer.h.27.attn.c_attn.g_idx', 'transformer.h.33.mlp.w2.qweight', 'transformer.h.0.mlp.c_proj.scales', 'transformer.h.28.mlp.w1.scales', 'transformer.h.13.mlp.c_proj.qweight', 'transformer.h.22.mlp.w1.qweight', 'transformer.h.25.mlp.c_proj.bias', 'transformer.h.10.mlp.w1.qweight', 'transformer.h.17.mlp.w1.g_idx', 'transformer.h.0.mlp.w1.qzeros', 'transformer.h.2.mlp.c_proj.g_idx', 'transformer.h.5.mlp.w2.qzeros', 'transformer.h.8.mlp.w1.qzeros', 'transformer.h.31.mlp.w1.scales', 'transformer.h.8.mlp.w1.bias', 'transformer.h.10.mlp.c_proj.qzeros', 'transformer.h.20.attn.c_proj.qzeros', 'transformer.h.0.mlp.w2.bias', 'transformer.h.21.mlp.w1.scales', 'transformer.h.37.mlp.c_proj.scales', 'transformer.h.36.attn.c_attn.scales', 'transformer.h.9.mlp.c_proj.qweight', 'transformer.h.4.attn.c_attn.qweight', 'transformer.h.28.attn.c_attn.scales', 'transformer.h.1.mlp.w2.scales', 'transformer.h.7.attn.c_proj.bias', 'transformer.h.18.mlp.w1.qweight', 'transformer.h.9.mlp.w2.bias', 'transformer.h.33.mlp.w1.qzeros', 'transformer.h.33.mlp.w2.g_idx', 'transformer.h.23.mlp.c_proj.bias', 'transformer.h.35.attn.c_proj.qzeros', 'transformer.h.19.attn.c_proj.qzeros', 'transformer.h.21.attn.c_proj.qweight', 'transformer.h.16.mlp.w2.qzeros', 'transformer.h.19.mlp.c_proj.qweight', 'transformer.h.2.attn.c_proj.g_idx', 'transformer.h.27.attn.c_attn.qweight', 'transformer.h.28.mlp.w1.qzeros', 'transformer.h.8.attn.c_attn.g_idx', 'transformer.h.19.mlp.w2.qzeros', 'transformer.h.34.mlp.w1.bias', 'transformer.h.39.mlp.w1.bias', 'transformer.h.35.mlp.w1.bias', 'transformer.h.22.mlp.w1.bias', 'transformer.h.39.mlp.c_proj.g_idx', 'transformer.h.23.mlp.c_proj.qweight', 'transformer.h.28.attn.c_proj.qzeros', 'transformer.h.2.attn.c_attn.qzeros', 'transformer.h.38.attn.c_proj.g_idx', 'transformer.h.30.attn.c_attn.qzeros', 'transformer.h.37.mlp.w1.scales', 'transformer.h.26.attn.c_proj.qweight', 'transformer.h.39.attn.c_attn.qweight', 'transformer.h.12.mlp.w2.g_idx', 'transformer.h.37.attn.c_proj.scales', 'transformer.h.8.attn.c_proj.g_idx', 'transformer.h.6.attn.c_proj.qzeros', 'transformer.h.6.attn.c_proj.scales', 'transformer.h.4.mlp.w1.scales', 'transformer.h.0.attn.c_attn.qweight', 'transformer.h.11.mlp.c_proj.g_idx', 'transformer.h.15.attn.c_proj.qweight', 'transformer.h.16.mlp.c_proj.g_idx', 'transformer.h.29.mlp.w1.scales', 'transformer.h.6.mlp.c_proj.g_idx', 'transformer.h.32.mlp.c_proj.g_idx', 'transformer.h.3.mlp.w1.scales', 'transformer.h.32.mlp.w1.g_idx', 'transformer.h.39.mlp.w1.qweight', 'transformer.h.15.mlp.w1.g_idx', 'transformer.h.16.mlp.w1.scales', 'transformer.h.22.mlp.w2.bias', 'transformer.h.27.attn.c_proj.scales', 'transformer.h.14.mlp.w1.scales', 'transformer.h.34.attn.c_proj.g_idx', 'transformer.h.3.mlp.c_proj.scales', 'transformer.h.27.attn.c_attn.qzeros', 'transformer.h.33.attn.c_attn.g_idx', 'transformer.h.15.mlp.w1.qweight', 'transformer.h.23.mlp.w2.bias', 'transformer.h.11.attn.c_attn.g_idx', 'transformer.h.10.attn.c_proj.scales', 'transformer.h.7.mlp.c_proj.g_idx', 'transformer.h.15.mlp.c_proj.qzeros', 'transformer.h.22.mlp.w1.qzeros', 'transformer.h.23.mlp.w1.scales', 'transformer.h.21.mlp.w2.g_idx', 'transformer.h.7.mlp.w1.g_idx', 'transformer.h.9.attn.c_proj.scales', 'transformer.h.6.attn.c_attn.qzeros', 'transformer.h.34.mlp.w1.scales', 'transformer.h.36.mlp.w1.g_idx', 'transformer.h.0.attn.c_attn.scales', 'transformer.h.18.mlp.w1.bias', 'transformer.h.23.mlp.c_proj.qzeros', 'transformer.h.13.mlp.w2.scales', 'transformer.h.26.mlp.w2.scales', 'transformer.h.3.mlp.w1.qzeros', 'transformer.h.30.attn.c_proj.g_idx', 'transformer.h.9.mlp.c_proj.qzeros', 'transformer.h.27.mlp.w1.qzeros', 'transformer.h.31.mlp.w1.qweight', 'transformer.h.5.attn.c_attn.scales', 'transformer.h.7.mlp.w1.bias', 'transformer.h.26.mlp.w1.bias', 'transformer.h.13.attn.c_attn.g_idx', 'transformer.h.19.mlp.w1.g_idx', 'transformer.h.29.mlp.w1.g_idx', 'transformer.h.28.mlp.w2.g_idx', 'transformer.h.4.mlp.c_proj.qweight', 'transformer.h.9.mlp.w2.qweight', 'transformer.h.8.attn.c_proj.qweight', 'transformer.h.39.attn.c_attn.scales', 'transformer.h.0.attn.c_attn.qzeros', 'transformer.h.13.mlp.w2.bias', 'transformer.h.6.attn.c_attn.g_idx', 'transformer.h.22.mlp.c_proj.g_idx', 'transformer.h.18.attn.c_attn.g_idx', 'transformer.h.21.attn.c_attn.scales', 'transformer.h.17.attn.c_proj.g_idx', 'transformer.h.35.mlp.w2.scales', 'transformer.h.24.mlp.c_proj.qweight', 'transformer.h.1.mlp.w2.bias', 'transformer.h.15.mlp.c_proj.g_idx', 'transformer.h.38.attn.c_proj.qzeros', 'transformer.h.28.attn.c_proj.g_idx', 'transformer.h.39.attn.c_attn.g_idx']

  • This IS expected if you are initializing QWenLMHeadModel 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 QWenLMHeadModel from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).

Some weights of QWenLMHeadModel were not initialized from the model checkpoint at Refs/Weights/Qwen-14B-Chat-Int4 and are newly initialized: ['transformer.h.25.mlp.w2.weight', 'transformer.h.14.mlp.w1.weight', 'transformer.h.20.attn.c_attn.weight', 'transformer.h.14.mlp.w2.weight', 'transformer.h.37.mlp.c_proj.weight', 'transformer.h.18.mlp.w1.weight', 'transformer.h.7.mlp.w1.weight', 'transformer.h.2.mlp.w1.weight', 'transformer.h.19.attn.c_proj.weight', 'transformer.h.1.mlp.w2.weight', 'transformer.h.23.mlp.w2.weight', 'transformer.h.19.mlp.c_proj.weight', 'transformer.h.11.mlp.c_proj.weight', 'transformer.h.17.attn.c_proj.weight', 'transformer.h.2.mlp.c_proj.weight', 'transformer.h.34.mlp.w2.weight', 'transformer.h.36.mlp.w1.weight', 'transformer.h.26.attn.c_attn.weight', 'transformer.h.36.mlp.c_proj.weight', 'transformer.h.1.attn.c_proj.weight', 'transformer.h.22.attn.c_attn.weight', 'transformer.h.6.mlp.w2.weight', 'transformer.h.38.attn.c_attn.weight', 'transformer.h.33.mlp.w1.weight', 'transformer.h.19.mlp.w1.weight', 'transformer.h.27.mlp.c_proj.weight', 'transformer.h.38.mlp.w1.weight', 'transformer.h.32.mlp.w1.weight', 'transformer.h.19.attn.c_attn.weight', 'transformer.h.6.mlp.c_proj.weight', 'transformer.h.3.attn.c_attn.weight', 'transformer.h.2.attn.c_proj.weight', 'transformer.h.20.mlp.c_proj.weight', 'transformer.h.16.attn.c_attn.weight', 'transformer.h.27.attn.c_proj.weight', 'transformer.h.12.mlp.w2.weight', 'transformer.h.39.attn.c_proj.weight', 'transformer.h.16.mlp.w1.weight', 'transformer.h.28.mlp.w1.weight', 'transformer.h.6.attn.c_proj.weight', 'transformer.h.16.mlp.w2.weight', 'transformer.h.38.mlp.w2.weight', 'transformer.h.36.attn.c_proj.weight', 'transformer.h.21.mlp.w2.weight', 'transformer.h.33.mlp.c_proj.weight', 'transformer.h.29.attn.c_proj.weight', 'transformer.h.15.attn.c_attn.weight', 'transformer.h.29.attn.c_attn.weight', 'transformer.h.3.mlp.w2.weight', 'transformer.h.25.mlp.w1.weight', 'transformer.h.23.mlp.w1.weight', 'transformer.h.13.attn.c_attn.weight', 'transformer.h.37.attn.c_proj.weight', 'transformer.h.11.attn.c_proj.weight', 'transformer.h.24.mlp.w2.weight', 'transformer.h.0.mlp.w1.weight', 'transformer.h.21.mlp.c_proj.weight', 'transformer.h.32.mlp.c_proj.weight', 'transformer.h.17.mlp.c_proj.weight', 'transformer.h.39.mlp.c_proj.weight', 'transformer.h.14.attn.c_attn.weight', 'transformer.h.26.mlp.w1.weight', 'transformer.h.6.mlp.w1.weight', 'transformer.h.12.attn.c_proj.weight', 'transformer.h.27.attn.c_attn.weight', 'transformer.h.21.attn.c_attn.weight', 'transformer.h.17.attn.c_attn.weight', 'transformer.h.12.attn.c_attn.weight', 'transformer.h.3.attn.c_proj.weight', 'transformer.h.29.mlp.c_proj.weight', 'transformer.h.9.attn.c_attn.weight', 'transformer.h.28.attn.c_proj.weight', 'transformer.h.9.mlp.w1.weight', 'transformer.h.10.mlp.c_proj.weight', 'transformer.h.4.attn.c_proj.weight', 'transformer.h.4.mlp.c_proj.weight', 'transformer.h.7.mlp.w2.weight', 'transformer.h.18.mlp.c_proj.weight', 'transformer.h.34.mlp.c_proj.weight', 'transformer.h.39.mlp.w1.weight', 'transformer.h.2.mlp.w2.weight', 'transformer.h.38.attn.c_proj.weight', 'transformer.h.1.attn.c_attn.weight', 'transformer.h.20.mlp.w2.weight', 'transformer.h.30.attn.c_attn.weight', 'transformer.h.20.attn.c_proj.weight', 'transformer.h.19.mlp.w2.weight', 'transformer.h.35.mlp.c_proj.weight', 'transformer.h.17.mlp.w1.weight', 'transformer.h.14.attn.c_proj.weight', 'transformer.h.8.mlp.w2.weight', 'transformer.h.10.mlp.w1.weight', 'transformer.h.15.mlp.w2.weight', 'transformer.h.28.attn.c_attn.weight', 'transformer.h.24.mlp.w1.weight', 'transformer.h.30.mlp.c_proj.weight', 'transformer.h.23.mlp.c_proj.weight', 'transformer.h.13.mlp.w2.weight', 'transformer.h.35.attn.c_proj.weight', 'transformer.h.35.mlp.w2.weight', 'transformer.h.8.attn.c_proj.weight', 'transformer.h.24.attn.c_attn.weight', 'transformer.h.0.attn.c_attn.weight', 'transformer.h.33.attn.c_proj.weight', 'transformer.h.25.mlp.c_proj.weight', 'transformer.h.5.mlp.w1.weight', 'transformer.h.26.attn.c_proj.weight', 'transformer.h.34.attn.c_proj.weight', 'transformer.h.32.attn.c_proj.weight', 'transformer.h.8.mlp.c_proj.weight', 'transformer.h.11.mlp.w2.weight', 'transformer.h.34.attn.c_attn.weight', 'transformer.h.5.mlp.w2.weight', 'transformer.h.21.mlp.w1.weight', 'transformer.h.35.mlp.w1.weight', 'transformer.h.37.mlp.w2.weight', 'transformer.h.18.attn.c_proj.weight', 'transformer.h.3.mlp.w1.weight', 'transformer.h.31.attn.c_attn.weight', 'transformer.h.33.mlp.w2.weight', 'transformer.h.16.mlp.c_proj.weight', 'transformer.h.29.mlp.w1.weight', 'transformer.h.10.attn.c_attn.weight', 'transformer.h.7.mlp.c_proj.weight', 'transformer.h.13.mlp.w1.weight', 'transformer.h.9.mlp.w2.weight', 'transformer.h.3.mlp.c_proj.weight', 'transformer.h.37.mlp.w1.weight', 'transformer.h.8.mlp.w1.weight', 'transformer.h.15.mlp.w1.weight', 'transformer.h.18.mlp.w2.weight', 'transformer.h.4.mlp.w2.weight', 'transformer.h.4.attn.c_attn.weight', 'transformer.h.14.mlp.c_proj.weight', 'transformer.h.0.mlp.w2.weight', 'transformer.h.37.attn.c_attn.weight', 'transformer.h.31.attn.c_proj.weight', 'transformer.h.32.mlp.w2.weight', 'transformer.h.26.mlp.w2.weight', 'transformer.h.10.mlp.w2.weight', 'transformer.h.12.mlp.c_proj.weight', 'transformer.h.11.attn.c_attn.weight', 'transformer.h.17.mlp.w2.weight', 'transformer.h.30.mlp.w2.weight', 'transformer.h.23.attn.c_attn.weight', 'transformer.h.31.mlp.w1.weight', 'transformer.h.4.mlp.w1.weight', 'transformer.h.39.mlp.w2.weight', 'transformer.h.23.attn.c_proj.weight', 'transformer.h.11.mlp.w1.weight', 'transformer.h.39.attn.c_attn.weight', 'transformer.h.31.mlp.c_proj.weight', 'transformer.h.13.mlp.c_proj.weight', 'transformer.h.13.attn.c_proj.weight', 'transformer.h.36.attn.c_attn.weight', 'transformer.h.16.attn.c_proj.weight', 'transformer.h.15.attn.c_proj.weight', 'transformer.h.29.mlp.w2.weight', 'transformer.h.2.attn.c_attn.weight', 'transformer.h.9.attn.c_proj.weight', 'transformer.h.5.mlp.c_proj.weight', 'transformer.h.7.attn.c_proj.weight', 'transformer.h.30.mlp.w1.weight', 'transformer.h.28.mlp.w2.weight', 'transformer.h.26.mlp.c_proj.weight', 'transformer.h.7.attn.c_attn.weight', 'transformer.h.34.mlp.w1.weight', 'transformer.h.22.mlp.w2.weight', 'transformer.h.28.mlp.c_proj.weight', 'transformer.h.38.mlp.c_proj.weight', 'transformer.h.22.mlp.c_proj.weight', 'transformer.h.5.attn.c_proj.weight', 'transformer.h.20.mlp.w1.weight', 'transformer.h.10.attn.c_proj.weight', 'transformer.h.9.mlp.c_proj.weight', 'transformer.h.32.attn.c_attn.weight', 'transformer.h.0.mlp.c_proj.weight', 'transformer.h.18.attn.c_attn.weight', 'transformer.h.1.mlp.c_proj.weight', 'transformer.h.0.attn.c_proj.weight', 'transformer.h.22.attn.c_proj.weight', 'transformer.h.24.attn.c_proj.weight', 'transformer.h.35.attn.c_attn.weight', 'transformer.h.1.mlp.w1.weight', 'transformer.h.8.attn.c_attn.weight', 'transformer.h.27.mlp.w1.weight', 'transformer.h.36.mlp.w2.weight', 'transformer.h.25.attn.c_attn.weight', 'transformer.h.27.mlp.w2.weight', 'transformer.h.30.attn.c_proj.weight', 'transformer.h.5.attn.c_attn.weight', 'transformer.h.22.mlp.w1.weight', 'transformer.h.21.attn.c_proj.weight', 'transformer.h.6.attn.c_attn.weight', 'transformer.h.12.mlp.w1.weight', 'transformer.h.25.attn.c_proj.weight', 'transformer.h.24.mlp.c_proj.weight', 'transformer.h.33.attn.c_attn.weight', 'transformer.h.31.mlp.w2.weight', 'transformer.h.15.mlp.c_proj.weight']

You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.

Qwen org

Please check if it is the same problem as https://github.com/QwenLM/Qwen/issues/411

You need a newer version of transformers that support AutoGPTQ. Please refer to the GitHub README for more information.

jklj077 changed discussion status to closed

Sign up or log in to comment