ALTICDEV's picture
Publish Fluid-1 Mini Mixed Q4/Q6 with DFlash v2
566b7c5 verified
Raw
History Blame Contribute Delete
51.1 kB
{
"architectures": [
"Qwen3_5ForConditionalGeneration"
],
"eos_token_id": 98280,
"image_token_id": 98290,
"model_name": "/home/bsubramaniam/ws/dictation-training/training/runs/qwen35-2b-vocab96k-list-professional-v1-stage1/serveable-checkpoint-1600",
"model_type": "qwen3_5",
"pad_token_id": 98278,
"quantization": {
"bits": 4,
"group_size": 64,
"language_model.model.embed_tokens": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.0.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.0.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.0.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.0.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.0.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.0.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.0.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.0.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.1.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.1.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.1.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.1.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.1.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.1.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.1.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.1.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.10.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.10.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.10.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.10.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.10.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.10.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.10.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.10.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.11.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.11.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.11.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.11.self_attn.k_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.11.self_attn.o_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.11.self_attn.q_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.11.self_attn.v_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.12.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.12.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.12.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.12.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.12.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.12.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.12.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.12.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.13.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.13.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.13.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.13.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.13.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.13.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.13.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.13.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.14.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.14.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.14.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.14.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.14.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.14.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.14.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.14.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.15.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.15.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.15.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.15.self_attn.k_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.15.self_attn.o_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.15.self_attn.q_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.15.self_attn.v_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.16.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.16.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.16.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.16.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.16.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.16.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.16.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.16.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.17.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.17.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.17.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.17.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.17.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.17.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.17.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.17.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.18.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.18.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.18.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.18.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.18.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.18.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.18.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.18.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.19.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.19.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.19.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.19.self_attn.k_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.19.self_attn.o_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.19.self_attn.q_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.19.self_attn.v_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.2.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.2.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.2.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.2.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.2.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.2.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.2.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.2.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.20.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.20.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.20.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.20.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.20.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.20.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.20.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.20.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.21.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.21.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.21.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.21.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.21.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.21.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.21.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.21.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.22.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.22.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.22.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.22.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.22.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.22.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.22.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.22.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.23.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.23.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.23.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.23.self_attn.k_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.23.self_attn.o_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.23.self_attn.q_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.23.self_attn.v_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.3.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.3.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.3.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.3.self_attn.k_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.3.self_attn.o_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.3.self_attn.q_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.3.self_attn.v_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.4.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.4.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.4.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.4.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.4.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.4.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.4.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.4.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.5.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.5.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.5.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.5.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.5.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.5.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.5.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.5.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.6.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.6.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.6.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.6.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.6.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.6.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.6.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.6.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.7.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.7.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.7.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.7.self_attn.k_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.7.self_attn.o_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.7.self_attn.q_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.7.self_attn.v_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.8.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.8.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.8.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.8.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.8.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.8.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.8.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.8.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.9.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.9.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.9.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.9.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.9.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.9.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.9.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.9.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"mode": "affine"
},
"quantization_config": {
"bits": 4,
"group_size": 64,
"language_model.model.embed_tokens": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.0.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.0.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.0.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.0.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.0.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.0.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.0.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.0.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.1.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.1.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.1.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.1.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.1.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.1.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.1.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.1.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.10.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.10.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.10.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.10.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.10.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.10.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.10.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.10.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.11.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.11.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.11.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.11.self_attn.k_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.11.self_attn.o_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.11.self_attn.q_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.11.self_attn.v_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.12.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.12.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.12.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.12.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.12.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.12.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.12.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.12.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.13.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.13.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.13.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.13.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.13.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.13.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.13.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.13.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.14.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.14.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.14.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.14.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.14.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.14.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.14.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.14.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.15.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.15.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.15.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.15.self_attn.k_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.15.self_attn.o_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.15.self_attn.q_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.15.self_attn.v_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.16.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.16.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.16.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.16.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.16.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.16.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.16.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.16.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.17.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.17.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.17.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.17.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.17.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.17.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.17.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.17.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.18.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.18.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.18.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.18.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.18.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.18.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.18.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.18.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.19.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.19.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.19.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.19.self_attn.k_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.19.self_attn.o_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.19.self_attn.q_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.19.self_attn.v_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.2.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.2.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.2.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.2.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.2.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.2.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.2.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.2.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.20.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.20.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.20.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.20.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.20.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.20.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.20.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.20.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.21.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.21.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.21.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.21.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.21.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.21.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.21.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.21.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.22.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.22.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.22.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.22.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.22.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.22.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.22.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.22.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.23.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.23.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.23.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.23.self_attn.k_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.23.self_attn.o_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.23.self_attn.q_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.23.self_attn.v_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.3.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.3.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.3.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.3.self_attn.k_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.3.self_attn.o_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.3.self_attn.q_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.3.self_attn.v_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.4.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.4.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.4.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.4.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.4.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.4.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.4.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.4.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.5.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.5.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.5.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.5.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.5.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.5.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.5.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.5.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.6.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.6.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.6.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.6.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.6.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.6.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.6.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.6.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.7.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.7.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.7.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.7.self_attn.k_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.7.self_attn.o_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.7.self_attn.q_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.7.self_attn.v_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.8.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.8.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.8.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.8.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.8.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.8.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.8.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.8.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.9.linear_attn.in_proj_a": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.9.linear_attn.in_proj_b": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.9.linear_attn.in_proj_qkv": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.9.linear_attn.in_proj_z": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.9.linear_attn.out_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.9.mlp.down_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.9.mlp.gate_proj": {
"bits": 4,
"group_size": 64,
"mode": "affine"
},
"language_model.model.layers.9.mlp.up_proj": {
"bits": 6,
"group_size": 64,
"mode": "affine"
},
"mode": "affine"
},
"text_config": {
"attention_bias": false,
"attention_dropout": 0.0,
"attn_output_gate": true,
"bos_token_id": null,
"eos_token_id": 98280,
"full_attention_interval": 4,
"head_dim": 256,
"hidden_act": "silu",
"hidden_size": 2048,
"initializer_range": 0.02,
"intermediate_size": 6144,
"layer_types": [
"linear_attention",
"linear_attention",
"linear_attention",
"full_attention",
"linear_attention",
"linear_attention",
"linear_attention",
"full_attention",
"linear_attention",
"linear_attention",
"linear_attention",
"full_attention",
"linear_attention",
"linear_attention",
"linear_attention",
"full_attention",
"linear_attention",
"linear_attention",
"linear_attention",
"full_attention",
"linear_attention",
"linear_attention",
"linear_attention",
"full_attention"
],
"linear_conv_kernel_dim": 4,
"linear_key_head_dim": 128,
"linear_num_key_heads": 16,
"linear_num_value_heads": 16,
"linear_value_head_dim": 128,
"mamba_ssm_dtype": "float32",
"max_position_embeddings": 262144,
"mlp_only_layers": [],
"model_type": "qwen3_5_text",
"mtp_num_hidden_layers": 1,
"mtp_use_dedicated_embeddings": false,
"num_attention_heads": 8,
"num_hidden_layers": 24,
"num_key_value_heads": 2,
"pad_token_id": 98278,
"partial_rotary_factor": 0.25,
"rms_norm_eps": 1e-06,
"rope_parameters": {
"mrope_interleaved": true,
"mrope_section": [
11,
11,
10
],
"partial_rotary_factor": 0.25,
"rope_theta": 10000000,
"type": "default"
},
"tie_word_embeddings": true,
"torch_dtype": "bfloat16",
"use_cache": true,
"vocab_size": 98304
},
"tie_word_embeddings": true,
"torch_dtype": "bfloat16",
"unsloth_version": "2026.4.6",
"video_token_id": 98291,
"vision_end_token_id": 98288,
"vision_start_token_id": 98287,
"dflash_config": {
"block_size": 8,
"mask_token_id": 98278,
"target_layer_ids": [
2,
12,
21
],
"draft_vocab_size": 8192,
"num_layers": 5,
"vocabulary_source": "sidecar",
"attention_head_count": 8,
"attention_kv_head_count": 2,
"head_dimension": 256,
"sliding_window": 0,
"quantization": {
"bits": 6,
"group_size": 64,
"mode": "affine"
}
}
}