Kimi-K3-NVFP4 / hf_quant_config.json
zhiyucheng's picture
Initial public release snapshot
1ec10fa
Raw
History Blame Contribute Delete
324 kB
{
"producer": {
"name": "modelopt",
"version": "0.45.0"
},
"quantization": {
"quant_algo": "MIXED_PRECISION",
"kv_cache_quant_algo": null,
"group_size": 16,
"quantized_layers": {
"language_model.model.layers.1.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.1.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.1.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.1.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.1.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.1.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.10.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.10.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.10.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.10.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.10.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.10.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.11.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.11.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.11.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.11.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.11.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.11.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.12.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.12.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.12.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.12.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.12.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.12.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.13.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.13.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.13.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.13.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.13.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.13.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.14.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.14.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.14.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.14.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.14.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.14.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.15.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.15.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.15.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.15.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.15.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.15.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.16.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.16.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.16.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.16.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.16.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.16.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.17.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.17.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.17.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.17.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.17.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.17.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.18.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.18.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.18.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.18.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.18.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.18.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.19.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.19.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.19.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.19.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.19.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.19.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.2.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.2.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.2.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.2.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.2.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.2.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.20.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.20.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.20.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.20.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.20.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.20.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.21.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.21.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.21.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.21.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.21.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.21.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.22.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.22.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.22.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.22.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.22.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.22.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.23.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.23.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.23.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.23.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.23.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.23.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.24.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.24.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.24.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.24.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.24.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.24.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.25.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.25.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.25.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.25.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.25.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.25.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.26.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.26.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.26.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.26.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.26.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.26.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.27.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.27.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.27.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.27.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.27.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.27.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.28.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.28.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.28.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.28.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.28.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.28.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.29.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.29.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.29.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.29.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.29.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.29.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.3.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.3.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.3.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.3.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.3.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.3.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.30.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.30.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.30.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.30.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.30.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.30.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.31.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.31.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.31.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.31.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.31.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.31.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.32.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.32.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.32.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.32.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.32.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.32.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.33.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.33.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.33.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.33.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.33.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.33.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.34.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.34.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.34.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.34.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.34.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.34.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.35.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.35.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.35.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.35.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.35.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.35.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.36.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.36.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.36.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.36.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.36.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.36.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.37.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.37.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.37.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.37.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.37.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.37.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.38.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.38.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.38.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.38.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.38.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.38.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.39.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.39.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.39.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.39.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.39.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.39.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.4.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.4.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.4.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.4.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.4.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.4.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.40.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.40.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.40.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.40.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.40.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.40.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.41.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.41.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.41.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.41.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.41.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.41.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.42.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.42.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.42.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.42.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.42.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.42.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.43.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.43.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.43.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.43.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.43.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.43.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.44.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.44.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.44.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.44.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.44.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.44.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.45.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.45.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.45.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.45.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.45.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.45.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.46.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.46.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.46.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.46.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.46.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.46.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.47.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.47.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.47.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.47.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.47.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.47.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.48.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.48.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.48.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.48.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.48.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.48.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.49.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.49.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.49.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.49.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.49.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.49.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.5.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.5.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.5.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.5.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.5.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.5.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.50.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.50.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.50.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.50.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.50.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.50.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.51.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.51.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.51.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.51.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.51.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.51.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.52.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.52.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.52.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.52.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.52.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.52.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.53.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.53.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.53.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.53.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.53.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.53.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.54.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.54.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.54.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.54.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.54.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.54.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.55.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.55.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.55.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.55.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.55.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.55.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.56.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.56.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.56.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.56.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.56.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.56.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.57.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.57.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.57.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.57.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.57.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.57.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.58.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.58.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.58.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.58.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.58.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.58.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.59.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.59.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.59.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.59.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.59.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.59.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.6.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.6.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.6.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.6.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.6.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.6.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.60.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.60.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.60.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.60.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.60.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.60.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.61.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.61.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.61.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.61.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.61.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.61.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.62.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.62.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.62.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.62.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.62.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.62.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.63.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.63.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.63.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.63.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.63.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.63.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.64.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.64.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.64.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.64.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.64.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.64.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.65.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.65.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.65.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.65.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.65.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.65.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.66.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.66.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.66.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.66.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.66.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.66.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.67.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.67.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.67.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.67.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.67.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.67.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.68.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.68.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.68.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.68.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.68.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.68.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.69.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.69.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.69.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.69.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.69.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.69.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.7.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.7.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.7.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.7.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.7.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.7.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.70.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.70.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.70.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.70.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.70.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.70.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.71.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.71.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.71.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.71.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.71.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.71.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.72.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.72.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.72.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.72.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.72.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.72.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.73.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.73.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.73.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.73.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.73.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.73.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.74.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.74.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.74.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.74.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.74.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.74.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.75.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.75.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.75.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.75.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.75.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.75.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.76.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.76.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.76.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.76.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.76.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.76.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.77.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.77.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.77.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.77.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.77.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.77.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.78.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.78.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.78.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.78.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.78.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.78.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.79.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.79.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.79.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.79.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.79.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.79.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.8.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.8.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.8.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.8.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.8.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.8.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.80.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.80.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.80.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.80.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.80.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.80.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.81.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.81.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.81.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.81.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.81.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.81.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.82.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.82.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.82.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.82.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.82.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.82.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.83.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.83.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.83.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.83.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.83.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.83.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.84.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.84.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.84.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.84.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.84.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.84.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.85.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.85.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.85.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.85.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.85.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.85.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.86.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.86.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.86.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.86.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.86.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.86.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.87.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.87.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.87.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.87.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.87.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.87.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.88.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.88.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.88.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.88.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.88.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.88.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.89.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.89.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.89.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.89.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.89.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.89.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.9.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.9.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.9.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.9.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.9.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.9.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.90.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.90.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.90.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.90.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.90.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.90.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.91.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.91.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.91.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.91.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.91.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.91.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.92.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.92.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.92.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"layers.92.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.92.block_sparse_moe.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.layers.92.mlp.experts": {
"quant_algo": "NVFP4",
"group_size": 16
},
"language_model.model.layers.0.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.0.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.0.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.0.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.0.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.0.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.0.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.0.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.0.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.0.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.0.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.0.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.0.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.0.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.0.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.0.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.0.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.0.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.0.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.0.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.0.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.0.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.0.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.0.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.1.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.1.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.1.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.1.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.1.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.1.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.1.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.1.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.1.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.1.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.1.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.1.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.1.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.1.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.1.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.1.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.1.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.1.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.1.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.1.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.1.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.1.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.1.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.1.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.10.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.10.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.10.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.10.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.10.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.10.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.10.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.10.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.10.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.10.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.10.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.10.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.10.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.10.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.10.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.10.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.10.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.10.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.10.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.10.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.10.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.10.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.10.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.10.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.11.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.11.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.11.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.11.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.11.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.11.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.11.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.11.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.11.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.11.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.11.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.11.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.11.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.11.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.11.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.11.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.11.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.11.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.12.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.12.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.12.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.12.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.12.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.12.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.12.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.12.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.12.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.12.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.12.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.12.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.12.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.12.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.12.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.12.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.12.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.12.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.12.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.12.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.12.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.12.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.12.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.12.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.13.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.13.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.13.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.13.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.13.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.13.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.13.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.13.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.13.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.13.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.13.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.13.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.13.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.13.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.13.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.13.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.13.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.13.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.13.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.13.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.13.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.13.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.13.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.13.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.14.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.14.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.14.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.14.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.14.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.14.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.14.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.14.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.14.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.14.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.14.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.14.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.14.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.14.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.14.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.14.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.14.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.14.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.14.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.14.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.14.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.14.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.14.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.14.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.15.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.15.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.15.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.15.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.15.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.15.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.15.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.15.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.15.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.15.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.15.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.15.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.15.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.15.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.15.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.15.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.15.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.15.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.16.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.16.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.16.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.16.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.16.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.16.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.16.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.16.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.16.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.16.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.16.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.16.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.16.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.16.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.16.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.16.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.16.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.16.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.16.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.16.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.16.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.16.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.16.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.16.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.17.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.17.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.17.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.17.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.17.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.17.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.17.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.17.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.17.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.17.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.17.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.17.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.17.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.17.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.17.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.17.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.17.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.17.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.17.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.17.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.17.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.17.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.17.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.17.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.18.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.18.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.18.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.18.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.18.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.18.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.18.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.18.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.18.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.18.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.18.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.18.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.18.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.18.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.18.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.18.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.18.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.18.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.18.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.18.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.18.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.18.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.18.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.18.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.19.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.19.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.19.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.19.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.19.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.19.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.19.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.19.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.19.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.19.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.19.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.19.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.19.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.19.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.19.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.19.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.19.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.19.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.2.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.2.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.2.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.2.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.2.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.2.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.2.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.2.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.2.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.2.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.2.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.2.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.2.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.2.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.2.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.2.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.2.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.2.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.2.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.2.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.2.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.2.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.2.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.2.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.20.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.20.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.20.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.20.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.20.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.20.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.20.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.20.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.20.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.20.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.20.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.20.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.20.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.20.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.20.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.20.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.20.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.20.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.20.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.20.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.20.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.20.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.20.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.20.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.21.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.21.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.21.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.21.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.21.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.21.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.21.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.21.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.21.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.21.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.21.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.21.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.21.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.21.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.21.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.21.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.21.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.21.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.21.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.21.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.21.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.21.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.21.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.21.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.22.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.22.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.22.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.22.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.22.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.22.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.22.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.22.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.22.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.22.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.22.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.22.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.22.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.22.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.22.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.22.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.22.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.22.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.22.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.22.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.22.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.22.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.22.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.22.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.23.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.23.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.23.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.23.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.23.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.23.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.23.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.23.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.23.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.23.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.23.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.23.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.23.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.23.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.23.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.23.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.23.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.23.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.24.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.24.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.24.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.24.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.24.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.24.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.24.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.24.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.24.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.24.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.24.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.24.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.24.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.24.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.24.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.24.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.24.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.24.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.24.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.24.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.24.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.24.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.24.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.24.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.25.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.25.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.25.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.25.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.25.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.25.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.25.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.25.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.25.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.25.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.25.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.25.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.25.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.25.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.25.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.25.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.25.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.25.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.25.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.25.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.25.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.25.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.25.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.25.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.26.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.26.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.26.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.26.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.26.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.26.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.26.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.26.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.26.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.26.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.26.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.26.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.26.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.26.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.26.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.26.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.26.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.26.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.26.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.26.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.26.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.26.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.26.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.26.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.27.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.27.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.27.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.27.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.27.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.27.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.27.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.27.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.27.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.27.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.27.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.27.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.27.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.27.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.27.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.27.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.27.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.27.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.28.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.28.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.28.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.28.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.28.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.28.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.28.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.28.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.28.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.28.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.28.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.28.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.28.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.28.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.28.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.28.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.28.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.28.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.28.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.28.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.28.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.28.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.28.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.28.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.29.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.29.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.29.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.29.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.29.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.29.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.29.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.29.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.29.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.29.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.29.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.29.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.29.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.29.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.29.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.29.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.29.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.29.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.29.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.29.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.29.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.29.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.29.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.29.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.3.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.3.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.3.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.3.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.3.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.3.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.3.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.3.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.3.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.3.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.3.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.3.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.3.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.3.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.3.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.3.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.3.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.3.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.30.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.30.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.30.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.30.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.30.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.30.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.30.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.30.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.30.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.30.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.30.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.30.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.30.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.30.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.30.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.30.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.30.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.30.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.30.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.30.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.30.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.30.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.30.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.30.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.31.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.31.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.31.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.31.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.31.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.31.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.31.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.31.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.31.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.31.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.31.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.31.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.31.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.31.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.31.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.31.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.31.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.31.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.32.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.32.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.32.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.32.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.32.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.32.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.32.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.32.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.32.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.32.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.32.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.32.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.32.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.32.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.32.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.32.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.32.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.32.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.32.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.32.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.32.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.32.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.32.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.32.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.33.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.33.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.33.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.33.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.33.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.33.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.33.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.33.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.33.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.33.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.33.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.33.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.33.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.33.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.33.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.33.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.33.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.33.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.33.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.33.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.33.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.33.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.33.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.33.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.34.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.34.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.34.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.34.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.34.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.34.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.34.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.34.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.34.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.34.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.34.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.34.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.34.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.34.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.34.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.34.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.34.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.34.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.34.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.34.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.34.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.34.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.34.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.34.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.35.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.35.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.35.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.35.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.35.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.35.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.35.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.35.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.35.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.35.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.35.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.35.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.35.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.35.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.35.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.35.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.35.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.35.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.36.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.36.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.36.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.36.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.36.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.36.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.36.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.36.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.36.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.36.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.36.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.36.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.36.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.36.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.36.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.36.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.36.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.36.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.36.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.36.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.36.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.36.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.36.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.36.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.37.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.37.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.37.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.37.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.37.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.37.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.37.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.37.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.37.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.37.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.37.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.37.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.37.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.37.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.37.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.37.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.37.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.37.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.37.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.37.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.37.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.37.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.37.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.37.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.38.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.38.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.38.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.38.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.38.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.38.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.38.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.38.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.38.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.38.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.38.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.38.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.38.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.38.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.38.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.38.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.38.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.38.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.38.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.38.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.38.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.38.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.38.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.38.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.39.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.39.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.39.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.39.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.39.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.39.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.39.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.39.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.39.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.39.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.39.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.39.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.39.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.39.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.39.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.39.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.39.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.39.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.4.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.4.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.4.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.4.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.4.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.4.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.4.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.4.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.4.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.4.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.4.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.4.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.4.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.4.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.4.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.4.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.4.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.4.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.4.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.4.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.4.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.4.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.4.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.4.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.40.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.40.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.40.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.40.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.40.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.40.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.40.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.40.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.40.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.40.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.40.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.40.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.40.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.40.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.40.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.40.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.40.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.40.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.40.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.40.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.40.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.40.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.40.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.40.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.41.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.41.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.41.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.41.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.41.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.41.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.41.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.41.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.41.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.41.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.41.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.41.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.41.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.41.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.41.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.41.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.41.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.41.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.41.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.41.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.41.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.41.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.41.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.41.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.42.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.42.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.42.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.42.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.42.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.42.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.42.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.42.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.42.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.42.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.42.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.42.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.42.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.42.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.42.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.42.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.42.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.42.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.42.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.42.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.42.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.42.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.42.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.42.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.43.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.43.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.43.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.43.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.43.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.43.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.43.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.43.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.43.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.43.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.43.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.43.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.43.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.43.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.43.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.43.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.43.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.43.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.44.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.44.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.44.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.44.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.44.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.44.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.44.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.44.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.44.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.44.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.44.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.44.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.44.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.44.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.44.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.44.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.44.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.44.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.44.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.44.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.44.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.44.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.44.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.44.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.45.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.45.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.45.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.45.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.45.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.45.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.45.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.45.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.45.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.45.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.45.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.45.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.45.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.45.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.45.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.45.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.45.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.45.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.45.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.45.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.45.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.45.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.45.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.45.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.46.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.46.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.46.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.46.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.46.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.46.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.46.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.46.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.46.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.46.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.46.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.46.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.46.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.46.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.46.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.46.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.46.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.46.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.46.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.46.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.46.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.46.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.46.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.46.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.47.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.47.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.47.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.47.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.47.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.47.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.47.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.47.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.47.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.47.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.47.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.47.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.47.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.47.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.47.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.47.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.47.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.47.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.48.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.48.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.48.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.48.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.48.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.48.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.48.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.48.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.48.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.48.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.48.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.48.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.48.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.48.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.48.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.48.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.48.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.48.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.48.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.48.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.48.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.48.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.48.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.48.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.49.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.49.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.49.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.49.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.49.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.49.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.49.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.49.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.49.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.49.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.49.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.49.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.49.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.49.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.49.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.49.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.49.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.49.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.49.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.49.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.49.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.49.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.49.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.49.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.5.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.5.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.5.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.5.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.5.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.5.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.5.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.5.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.5.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.5.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.5.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.5.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.5.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.5.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.5.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.5.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.5.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.5.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.5.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.5.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.5.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.5.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.5.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.5.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.50.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.50.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.50.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.50.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.50.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.50.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.50.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.50.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.50.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.50.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.50.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.50.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.50.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.50.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.50.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.50.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.50.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.50.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.50.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.50.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.50.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.50.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.50.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.50.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.51.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.51.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.51.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.51.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.51.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.51.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.51.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.51.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.51.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.51.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.51.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.51.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.51.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.51.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.51.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.51.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.51.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.51.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.52.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.52.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.52.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.52.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.52.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.52.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.52.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.52.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.52.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.52.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.52.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.52.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.52.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.52.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.52.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.52.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.52.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.52.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.52.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.52.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.52.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.52.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.52.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.52.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.53.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.53.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.53.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.53.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.53.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.53.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.53.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.53.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.53.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.53.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.53.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.53.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.53.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.53.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.53.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.53.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.53.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.53.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.53.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.53.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.53.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.53.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.53.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.53.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.54.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.54.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.54.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.54.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.54.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.54.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.54.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.54.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.54.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.54.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.54.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.54.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.54.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.54.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.54.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.54.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.54.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.54.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.54.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.54.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.54.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.54.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.54.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.54.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.55.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.55.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.55.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.55.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.55.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.55.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.55.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.55.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.55.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.55.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.55.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.55.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.55.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.55.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.55.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.55.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.55.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.55.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.56.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.56.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.56.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.56.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.56.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.56.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.56.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.56.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.56.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.56.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.56.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.56.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.56.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.56.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.56.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.56.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.56.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.56.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.56.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.56.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.56.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.56.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.56.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.56.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.57.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.57.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.57.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.57.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.57.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.57.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.57.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.57.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.57.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.57.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.57.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.57.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.57.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.57.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.57.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.57.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.57.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.57.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.57.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.57.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.57.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.57.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.57.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.57.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.58.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.58.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.58.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.58.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.58.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.58.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.58.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.58.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.58.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.58.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.58.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.58.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.58.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.58.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.58.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.58.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.58.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.58.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.58.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.58.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.58.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.58.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.58.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.58.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.59.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.59.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.59.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.59.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.59.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.59.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.59.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.59.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.59.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.59.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.59.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.59.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.59.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.59.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.59.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.59.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.59.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.59.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.6.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.6.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.6.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.6.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.6.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.6.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.6.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.6.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.6.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.6.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.6.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.6.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.6.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.6.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.6.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.6.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.6.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.6.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.6.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.6.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.6.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.6.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.6.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.6.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.60.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.60.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.60.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.60.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.60.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.60.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.60.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.60.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.60.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.60.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.60.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.60.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.60.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.60.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.60.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.60.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.60.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.60.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.60.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.60.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.60.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.60.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.60.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.60.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.61.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.61.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.61.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.61.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.61.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.61.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.61.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.61.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.61.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.61.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.61.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.61.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.61.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.61.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.61.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.61.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.61.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.61.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.61.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.61.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.61.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.61.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.61.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.61.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.62.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.62.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.62.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.62.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.62.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.62.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.62.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.62.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.62.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.62.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.62.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.62.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.62.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.62.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.62.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.62.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.62.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.62.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.62.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.62.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.62.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.62.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.62.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.62.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.63.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.63.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.63.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.63.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.63.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.63.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.63.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.63.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.63.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.63.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.63.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.63.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.63.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.63.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.63.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.63.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.63.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.63.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.64.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.64.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.64.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.64.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.64.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.64.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.64.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.64.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.64.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.64.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.64.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.64.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.64.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.64.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.64.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.64.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.64.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.64.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.64.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.64.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.64.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.64.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.64.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.64.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.65.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.65.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.65.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.65.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.65.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.65.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.65.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.65.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.65.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.65.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.65.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.65.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.65.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.65.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.65.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.65.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.65.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.65.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.65.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.65.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.65.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.65.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.65.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.65.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.66.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.66.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.66.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.66.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.66.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.66.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.66.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.66.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.66.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.66.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.66.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.66.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.66.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.66.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.66.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.66.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.66.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.66.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.66.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.66.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.66.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.66.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.66.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.66.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.67.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.67.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.67.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.67.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.67.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.67.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.67.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.67.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.67.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.67.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.67.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.67.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.67.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.67.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.67.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.67.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.67.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.67.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.68.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.68.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.68.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.68.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.68.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.68.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.68.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.68.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.68.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.68.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.68.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.68.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.68.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.68.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.68.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.68.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.68.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.68.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.68.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.68.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.68.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.68.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.68.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.68.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.69.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.69.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.69.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.69.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.69.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.69.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.69.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.69.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.69.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.69.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.69.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.69.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.69.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.69.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.69.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.69.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.69.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.69.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.69.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.69.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.69.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.69.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.69.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.69.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.7.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.7.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.7.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.7.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.7.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.7.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.7.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.7.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.7.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.7.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.7.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.7.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.7.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.7.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.7.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.7.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.7.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.7.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.70.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.70.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.70.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.70.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.70.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.70.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.70.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.70.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.70.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.70.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.70.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.70.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.70.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.70.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.70.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.70.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.70.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.70.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.70.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.70.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.70.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.70.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.70.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.70.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.71.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.71.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.71.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.71.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.71.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.71.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.71.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.71.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.71.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.71.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.71.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.71.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.71.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.71.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.71.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.71.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.71.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.71.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.72.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.72.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.72.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.72.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.72.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.72.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.72.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.72.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.72.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.72.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.72.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.72.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.72.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.72.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.72.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.72.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.72.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.72.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.72.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.72.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.72.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.72.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.72.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.72.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.73.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.73.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.73.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.73.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.73.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.73.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.73.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.73.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.73.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.73.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.73.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.73.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.73.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.73.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.73.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.73.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.73.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.73.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.73.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.73.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.73.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.73.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.73.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.73.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.74.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.74.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.74.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.74.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.74.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.74.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.74.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.74.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.74.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.74.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.74.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.74.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.74.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.74.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.74.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.74.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.74.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.74.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.74.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.74.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.74.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.74.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.74.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.74.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.75.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.75.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.75.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.75.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.75.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.75.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.75.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.75.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.75.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.75.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.75.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.75.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.75.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.75.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.75.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.75.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.75.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.75.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.76.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.76.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.76.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.76.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.76.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.76.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.76.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.76.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.76.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.76.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.76.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.76.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.76.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.76.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.76.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.76.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.76.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.76.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.76.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.76.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.76.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.76.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.76.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.76.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.77.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.77.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.77.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.77.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.77.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.77.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.77.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.77.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.77.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.77.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.77.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.77.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.77.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.77.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.77.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.77.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.77.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.77.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.77.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.77.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.77.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.77.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.77.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.77.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.78.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.78.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.78.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.78.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.78.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.78.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.78.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.78.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.78.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.78.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.78.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.78.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.78.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.78.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.78.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.78.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.78.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.78.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.78.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.78.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.78.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.78.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.78.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.78.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.79.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.79.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.79.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.79.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.79.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.79.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.79.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.79.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.79.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.79.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.79.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.79.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.79.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.79.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.79.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.79.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.79.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.79.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.8.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.8.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.8.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.8.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.8.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.8.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.8.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.8.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.8.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.8.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.8.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.8.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.8.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.8.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.8.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.8.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.8.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.8.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.8.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.8.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.8.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.8.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.8.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.8.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.80.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.80.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.80.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.80.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.80.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.80.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.80.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.80.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.80.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.80.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.80.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.80.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.80.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.80.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.80.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.80.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.80.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.80.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.80.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.80.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.80.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.80.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.80.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.80.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.81.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.81.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.81.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.81.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.81.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.81.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.81.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.81.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.81.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.81.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.81.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.81.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.81.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.81.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.81.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.81.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.81.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.81.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.81.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.81.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.81.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.81.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.81.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.81.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.82.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.82.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.82.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.82.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.82.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.82.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.82.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.82.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.82.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.82.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.82.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.82.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.82.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.82.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.82.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.82.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.82.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.82.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.82.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.82.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.82.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.82.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.82.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.82.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.83.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.83.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.83.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.83.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.83.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.83.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.83.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.83.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.83.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.83.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.83.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.83.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.83.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.83.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.83.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.83.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.83.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.83.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.84.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.84.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.84.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.84.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.84.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.84.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.84.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.84.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.84.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.84.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.84.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.84.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.84.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.84.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.84.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.84.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.84.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.84.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.84.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.84.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.84.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.84.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.84.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.84.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.85.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.85.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.85.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.85.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.85.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.85.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.85.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.85.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.85.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.85.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.85.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.85.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.85.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.85.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.85.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.85.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.85.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.85.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.85.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.85.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.85.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.85.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.85.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.85.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.86.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.86.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.86.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.86.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.86.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.86.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.86.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.86.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.86.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.86.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.86.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.86.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.86.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.86.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.86.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.86.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.86.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.86.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.86.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.86.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.86.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.86.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.86.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.86.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.87.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.87.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.87.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.87.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.87.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.87.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.87.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.87.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.87.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.87.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.87.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.87.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.87.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.87.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.87.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.87.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.87.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.87.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.88.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.88.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.88.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.88.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.88.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.88.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.88.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.88.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.88.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.88.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.88.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.88.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.88.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.88.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.88.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.88.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.88.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.88.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.88.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.88.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.88.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.88.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.88.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.88.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.89.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.89.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.89.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.89.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.89.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.89.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.89.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.89.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.89.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.89.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.89.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.89.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.89.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.89.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.89.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.89.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.89.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.89.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.89.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.89.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.89.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.89.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.89.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.89.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.9.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.9.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.9.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.9.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.9.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.9.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.9.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.9.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.9.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.9.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.9.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.9.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.9.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.9.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.9.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.9.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.9.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.9.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.9.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.9.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.9.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.9.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.9.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.9.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.90.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.90.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.90.self_attn.b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.90.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.90.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.90.self_attn.f_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.90.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.90.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.90.self_attn.f_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.90.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.90.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.90.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.90.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.90.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.90.self_attn.k_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.90.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.90.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.90.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.90.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.90.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.90.self_attn.q_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.90.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.90.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.90.self_attn.v_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.91.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.91.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.91.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.91.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.91.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.91.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.91.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.91.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.91.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.91.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.91.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.91.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.91.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.91.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.91.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.91.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.91.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.91.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.92.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.92.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.92.self_attn.g_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.92.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.92.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.92.self_attn.kv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.92.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.92.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.92.self_attn.kv_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.92.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.92.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.92.self_attn.o_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.92.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.92.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.92.self_attn.q_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.92.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.92.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.92.self_attn.q_b_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.76.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.76.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.76.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.23.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.23.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.23.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.27.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.27.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.27.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.63.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.63.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.63.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.66.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.66.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.66.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.2.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.2.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.2.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.87.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.87.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.87.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.20.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.20.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.20.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.68.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.68.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.68.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.59.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.59.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.59.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.65.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.65.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.65.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.19.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.19.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.19.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.9.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.9.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.9.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.89.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.89.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.89.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.83.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.83.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.83.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.27.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.27.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.27.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.22.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.22.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.22.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.72.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.72.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.72.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.63.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.63.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.63.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.71.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.71.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.71.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.32.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.32.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.32.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.30.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.30.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.30.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.60.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.60.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.60.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.7.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.7.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.7.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.91.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.91.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.91.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.38.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.38.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.38.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.55.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.55.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.55.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.4.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.4.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.4.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.91.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.91.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.91.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.39.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.39.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.39.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.80.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.80.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.80.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.79.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.79.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.79.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.10.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.10.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.10.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.45.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.45.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.45.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.64.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.64.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.64.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.55.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.55.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.55.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.33.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.33.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.33.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.73.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.73.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.73.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.11.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.11.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.11.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.42.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.42.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.42.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.44.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.44.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.44.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.87.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.87.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.87.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.91.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.91.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.91.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.62.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.62.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.62.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.56.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.56.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.56.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.52.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.52.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.52.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.31.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.31.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.31.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.73.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.73.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.73.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.87.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.87.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.87.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.91.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.91.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.91.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.30.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.30.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.30.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.25.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.25.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.25.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.43.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.43.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.43.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.43.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.43.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.43.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.26.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.26.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.26.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.55.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.55.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.55.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.83.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.83.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.83.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.32.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.32.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.32.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.35.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.35.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.35.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.53.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.53.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.53.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.89.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.89.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.89.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.58.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.58.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.58.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.14.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.14.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.14.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.79.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.79.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.79.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.57.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.57.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.57.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.78.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.78.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.78.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.85.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.85.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.85.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.47.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.47.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.47.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.70.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.70.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.70.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.19.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.19.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.19.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.43.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.43.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.43.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.67.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.67.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.67.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.23.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.23.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.23.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.22.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.22.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.22.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.56.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.56.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.56.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.7.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.7.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.7.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.90.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.90.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.90.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.37.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.37.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.37.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.61.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.61.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.61.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.8.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.8.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.8.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.6.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.6.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.6.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.15.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.15.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.15.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.59.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.59.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.59.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.15.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.15.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.15.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.63.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.63.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.63.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.39.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.39.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.39.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.51.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.51.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.51.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.69.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.69.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.69.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.46.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.46.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.46.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.5.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.5.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.5.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.92.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.92.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.92.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.33.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.33.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.33.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.20.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.20.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.20.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.45.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.45.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.45.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.25.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.25.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.25.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.60.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.60.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.60.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.0.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.0.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.0.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.80.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.80.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.80.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.34.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.34.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.34.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.81.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.81.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.81.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.47.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.47.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.47.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.11.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.11.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.11.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.27.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.27.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.27.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.35.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.35.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.35.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.92.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.92.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.92.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.13.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.13.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.13.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.72.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.72.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.72.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.67.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.67.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.67.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.16.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.16.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.16.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.0.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.0.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.0.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.31.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.31.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.31.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.82.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.82.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.82.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.28.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.28.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.28.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.47.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.47.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.47.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.40.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.40.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.40.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.83.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.83.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.83.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.86.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.86.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.86.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.54.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.54.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.54.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.46.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.46.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.46.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.37.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.37.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.37.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.11.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.11.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.11.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.15.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.15.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.15.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.5.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.5.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.5.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.50.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.50.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.50.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.12.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.12.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.12.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.31.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.31.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.31.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.39.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.39.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.39.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.78.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.78.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.78.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.8.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.8.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.8.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.83.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.83.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.83.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.92.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.92.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.92.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.41.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.41.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.41.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.53.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.53.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.53.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.59.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.59.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.59.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.19.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.19.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.19.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.70.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.70.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.70.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.81.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.81.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.81.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.14.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.14.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.14.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.3.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.3.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.3.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.10.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.10.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.10.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.61.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.61.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.61.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.48.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.48.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.48.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.12.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.12.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.12.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.75.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.75.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.75.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.18.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.18.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.18.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.2.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.2.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.2.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.75.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.75.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.75.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.34.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.34.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.34.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.77.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.77.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.77.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.21.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.21.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.21.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.29.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.29.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.29.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.23.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.23.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.23.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.27.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.27.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.27.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.43.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.43.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.43.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.76.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.76.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.76.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.86.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.86.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.86.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.67.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.67.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.67.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.28.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.28.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.28.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.3.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.3.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.3.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.3.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.3.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.3.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.58.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.58.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.58.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.11.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.11.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.11.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.3.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.3.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.3.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.42.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.42.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.42.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.15.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.15.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.15.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.74.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.74.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.74.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.51.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.51.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.51.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.7.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.7.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.7.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.36.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.36.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.36.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.67.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.67.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.67.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.38.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.38.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.38.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.41.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.41.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.41.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.17.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.17.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.17.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.75.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.75.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.75.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.35.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.35.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.35.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.36.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.36.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.36.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.26.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.26.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.26.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.90.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.90.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.90.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.1.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.1.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.1.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.24.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.24.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.24.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.44.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.44.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.44.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.52.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.52.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.52.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.4.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.4.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.4.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.24.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.24.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.24.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.87.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.87.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.87.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.1.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.1.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.1.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.7.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.7.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.7.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.75.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.75.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.75.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.23.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.23.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.23.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.50.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.50.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.50.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.39.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.39.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.39.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.16.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.16.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.16.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.19.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.19.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.19.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.69.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.69.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.69.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.79.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.79.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.79.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.92.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.92.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.92.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.55.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.55.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.55.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.64.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.64.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.64.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.49.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.49.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.49.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.57.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.57.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.57.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.47.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.47.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.47.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.66.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.66.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.66.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.63.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.63.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.63.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.48.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.48.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.48.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.68.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.68.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.68.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.71.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.71.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.71.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.71.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.71.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.71.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.84.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.84.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.84.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.18.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.18.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.18.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.79.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.79.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.79.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.84.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.84.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.84.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.62.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.62.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.62.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.88.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.88.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.88.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.40.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.40.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.40.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.13.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.13.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.13.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.82.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.82.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.82.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.59.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.59.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.59.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.35.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.35.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.35.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.6.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.6.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.6.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.21.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.21.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.21.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.74.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.74.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.74.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.54.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.54.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.54.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.85.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.85.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.85.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.51.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.51.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.51.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.71.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.71.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.71.self_attn.fused_qkv_a_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.77.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.77.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.77.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.88.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.88.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.88.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.9.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.9.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.9.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.49.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.49.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.49.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.17.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.17.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.17.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.31.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"layers.31.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.31.self_attn.fused_qkvg_proj": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.65.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.65.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.65.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.51.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"layers.51.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.51.self_attn.fused_qkv_a_proj_with_mqa": {
"quant_algo": "FP8_PB_WO"
},
"language_model.model.layers.29.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"layers.29.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
},
"model.layers.29.self_attn.in_proj_qkvgfab": {
"quant_algo": "FP8_PB_WO"
}
},
"exclude_modules": [
"*block_sparse_moe.gate*",
"*mlp.gate*",
"*block_sparse_moe.routed_expert_up_proj*",
"*mlp.routed_expert_up_proj*",
"*block_sparse_moe.routed_expert_down_proj*",
"*mlp.routed_expert_down_proj*",
"*block_sparse_moe.routed_expert_norm*",
"*mlp.routed_expert_norm*",
"*block_sparse_moe.shared_experts*",
"*mlp.shared_experts*",
"*self_attn.q_conv1d*",
"*self_attn.k_conv1d*",
"*self_attn.v_conv1d*",
"*self_attn.o_norm*",
"*self_attn.q_a_layernorm*",
"*self_attn.kv_a_layernorm*",
"*_res_proj*",
"*_res_norm*",
"*layernorm*",
"language_model.model.layers.0.mlp.*",
"layers.0.mlp.*",
"model.layers.0.mlp.*",
"language_model.lm_head",
"vision_tower*",
"mm_projector*"
]
}
}