Models
Docs
Pricing
Sign in
Download
Models
Download
Docs
Pricing
Sign in
gemma4
:e4b-mlx
16.3M
Downloads
Updated
5 hours ago
Gemma 4 models are designed to deliver frontier-level performance at each size. They are well-suited for reasoning, agentic workflows, coding, and multimodal understanding.
Gemma 4 models are designed to deliver frontier-level performance at each size. They are well-suited for reasoning, agentic workflows, coding, and multimodal understanding.
Cancel
vision
tools
thinking
audio
cloud
e2b
e4b
12b
26b
31b
gemma4:e4b-mlx
...
/
json
33cfd94b42d3 · 98kB
{
"producer": {
"name": "modelopt",
"version": "0.45.0.dev110+g04f58166a"
},
"quantization": {
"quant_algo": "MIXED_PRECISION",
"kv_cache_quant_algo": "FP8",
"quantized_layers": {
"model.vision_tower.patch_embedder.input_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.0.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.0.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.0.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.0.self_attn.o_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.0.mlp.gate_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.0.mlp.up_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.0.mlp.down_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.1.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.1.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.1.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.1.self_attn.o_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.1.mlp.gate_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.1.mlp.up_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.1.mlp.down_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.2.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.2.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.2.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.2.self_attn.o_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.2.mlp.gate_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.2.mlp.up_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.2.mlp.down_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.3.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.3.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.3.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.3.self_attn.o_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.3.mlp.gate_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.3.mlp.up_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.3.mlp.down_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.4.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.4.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.4.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.4.self_attn.o_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.4.mlp.gate_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.4.mlp.up_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.4.mlp.down_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.5.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.5.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.5.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.5.self_attn.o_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.5.mlp.gate_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.5.mlp.up_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.5.mlp.down_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.6.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.6.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.6.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.6.self_attn.o_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.6.mlp.gate_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.6.mlp.up_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.6.mlp.down_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.7.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.7.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.7.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.7.self_attn.o_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.7.mlp.gate_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.7.mlp.up_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.7.mlp.down_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.8.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.8.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.8.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.8.self_attn.o_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.8.mlp.gate_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.8.mlp.up_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.8.mlp.down_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.9.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.9.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.9.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.9.self_attn.o_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.9.mlp.gate_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.9.mlp.up_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.9.mlp.down_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.10.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.10.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.10.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.10.self_attn.o_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.10.mlp.gate_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.10.mlp.up_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.10.mlp.down_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.11.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.11.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.11.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.11.self_attn.o_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.11.mlp.gate_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.11.mlp.up_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.11.mlp.down_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.12.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.12.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.12.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.12.self_attn.o_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.12.mlp.gate_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.12.mlp.up_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.12.mlp.down_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.13.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.13.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.13.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.13.self_attn.o_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.13.mlp.gate_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.13.mlp.up_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.13.mlp.down_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.14.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.14.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.14.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.14.self_attn.o_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.14.mlp.gate_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.14.mlp.up_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.14.mlp.down_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.15.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.15.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.15.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.15.self_attn.o_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.15.mlp.gate_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.15.mlp.up_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.vision_tower.encoder.layers.15.mlp.down_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.0.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.0.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.0.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.0.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.0.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.0.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.0.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.0.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.0.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.1.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.1.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.1.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.1.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.1.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.1.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.1.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.1.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.1.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.2.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.2.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.2.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.2.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.2.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.2.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.2.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.2.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.2.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.3.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.3.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.3.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.3.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.3.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.3.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.3.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.3.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.3.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.4.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.4.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.4.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.4.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.4.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.4.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.4.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.4.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.4.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.5.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.5.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.5.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.5.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.5.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.5.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.5.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.5.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.5.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.6.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.6.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.6.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.6.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.6.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.6.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.6.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.6.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.6.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.7.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.7.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.7.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.7.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.7.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.7.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.7.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.7.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.7.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.8.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.8.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.8.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.8.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.8.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.8.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.8.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.8.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.8.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.9.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.9.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.9.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.9.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.9.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.9.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.9.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.9.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.9.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.10.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.10.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.10.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.10.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.10.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.10.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.10.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.10.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.10.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.11.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.11.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.11.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.11.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.11.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.11.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.11.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.11.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.11.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.12.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.12.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.12.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.12.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.12.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.12.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.12.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.12.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.12.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.13.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.13.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.13.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.13.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.13.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.13.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.13.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.13.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.13.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.14.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.14.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.14.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.14.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.14.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.14.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.14.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.14.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.14.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.15.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.15.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.15.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.15.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.15.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.15.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.15.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.15.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.15.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.16.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.16.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.16.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.16.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.16.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.16.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.16.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.16.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.16.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.17.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.17.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.17.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.17.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.17.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.17.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.17.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.17.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.17.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.18.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.18.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.18.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.18.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.18.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.18.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.18.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.18.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.18.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.19.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.19.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.19.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.19.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.19.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.19.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.19.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.19.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.19.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.20.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.20.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.20.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.20.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.20.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.20.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.20.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.20.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.20.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.21.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.21.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.21.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.21.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.21.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.21.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.21.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.21.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.21.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.22.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.22.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.22.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.22.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.22.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.22.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.22.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.22.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.22.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.23.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.23.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.23.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.23.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.23.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.23.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.23.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.23.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.23.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.24.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.24.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.24.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.24.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.24.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.24.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.24.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.24.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.24.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.25.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.25.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.25.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.25.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.25.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.25.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.25.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.25.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.25.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.26.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.26.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.26.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.26.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.26.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.26.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.26.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.26.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.26.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.27.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.27.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.27.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.27.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.27.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.27.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.27.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.27.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.27.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.28.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.28.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.28.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.28.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.28.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.28.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.28.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.28.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.28.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.29.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.29.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.29.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.29.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.29.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.29.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.29.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.29.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.29.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.30.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.30.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.30.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.30.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.30.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.30.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.30.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.30.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.30.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.31.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.31.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.31.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.31.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.31.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.31.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.31.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.31.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.31.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.32.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.32.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.32.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.32.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.32.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.32.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.32.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.32.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.32.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.33.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.33.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.33.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.33.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.33.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.33.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.33.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.33.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.33.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.34.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.34.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.34.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.34.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.34.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.34.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.34.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.34.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.34.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.35.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.35.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.35.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.35.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.35.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.35.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.35.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.35.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.35.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.36.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.36.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.36.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.36.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.36.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.36.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.36.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.36.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.36.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.37.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.37.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.37.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.37.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.37.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.37.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.37.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.37.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.37.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.38.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.38.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.38.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.38.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.38.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.38.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.38.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.38.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.38.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.39.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.39.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.39.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.39.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.39.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.39.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.39.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.39.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.39.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.40.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.40.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.40.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.40.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.40.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.40.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.40.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.40.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.40.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.41.self_attn.q_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.41.self_attn.k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.41.self_attn.v_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.41.self_attn.o_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.41.mlp.gate_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.41.mlp.up_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.41.mlp.down_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.41.per_layer_input_gate": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.layers.41.per_layer_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.per_layer_model_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.subsample_conv_projection.layer0.conv": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.subsample_conv_projection.layer1.conv": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.subsample_conv_projection.input_proj_linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.0.feed_forward1.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.0.feed_forward1.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.0.feed_forward2.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.0.feed_forward2.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.0.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.0.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.0.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.0.self_attn.post.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.0.self_attn.relative_k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.0.lconv1d.linear_start.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.0.lconv1d.linear_end.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.1.feed_forward1.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.1.feed_forward1.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.1.feed_forward2.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.1.feed_forward2.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.1.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.1.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.1.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.1.self_attn.post.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.1.self_attn.relative_k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.1.lconv1d.linear_start.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.1.lconv1d.linear_end.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.2.feed_forward1.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.2.feed_forward1.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.2.feed_forward2.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.2.feed_forward2.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.2.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.2.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.2.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.2.self_attn.post.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.2.self_attn.relative_k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.2.lconv1d.linear_start.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.2.lconv1d.linear_end.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.3.feed_forward1.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.3.feed_forward1.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.3.feed_forward2.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.3.feed_forward2.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.3.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.3.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.3.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.3.self_attn.post.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.3.self_attn.relative_k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.3.lconv1d.linear_start.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.3.lconv1d.linear_end.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.4.feed_forward1.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.4.feed_forward1.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.4.feed_forward2.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.4.feed_forward2.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.4.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.4.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.4.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.4.self_attn.post.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.4.self_attn.relative_k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.4.lconv1d.linear_start.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.4.lconv1d.linear_end.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.5.feed_forward1.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.5.feed_forward1.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.5.feed_forward2.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.5.feed_forward2.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.5.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.5.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.5.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.5.self_attn.post.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.5.self_attn.relative_k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.5.lconv1d.linear_start.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.5.lconv1d.linear_end.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.6.feed_forward1.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.6.feed_forward1.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.6.feed_forward2.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.6.feed_forward2.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.6.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.6.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.6.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.6.self_attn.post.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.6.self_attn.relative_k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.6.lconv1d.linear_start.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.6.lconv1d.linear_end.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.7.feed_forward1.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.7.feed_forward1.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.7.feed_forward2.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.7.feed_forward2.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.7.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.7.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.7.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.7.self_attn.post.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.7.self_attn.relative_k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.7.lconv1d.linear_start.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.7.lconv1d.linear_end.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.8.feed_forward1.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.8.feed_forward1.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.8.feed_forward2.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.8.feed_forward2.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.8.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.8.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.8.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.8.self_attn.post.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.8.self_attn.relative_k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.8.lconv1d.linear_start.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.8.lconv1d.linear_end.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.9.feed_forward1.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.9.feed_forward1.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.9.feed_forward2.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.9.feed_forward2.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.9.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.9.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.9.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.9.self_attn.post.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.9.self_attn.relative_k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.9.lconv1d.linear_start.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.9.lconv1d.linear_end.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.10.feed_forward1.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.10.feed_forward1.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.10.feed_forward2.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.10.feed_forward2.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.10.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.10.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.10.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.10.self_attn.post.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.10.self_attn.relative_k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.10.lconv1d.linear_start.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.10.lconv1d.linear_end.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.11.feed_forward1.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.11.feed_forward1.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.11.feed_forward2.ffw_layer_1.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.11.feed_forward2.ffw_layer_2.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.11.self_attn.q_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.11.self_attn.k_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.11.self_attn.v_proj.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.11.self_attn.post.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.11.self_attn.relative_k_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.11.lconv1d.linear_start.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.layers.11.lconv1d.linear_end.linear": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.audio_tower.output_proj": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.embed_vision.embedding_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.embed_audio.embedding_projection": {
"quant_algo": "NVFP4",
"group_size": 16
},
"model.language_model.embed_tokens": {
"quant_algo": "W4A16_NVFP4",
"group_size": 16
}
}
}
}