Kaleto's picture
Quantize the LM head to NVFP4 (22 % faster decode, 2 GB smaller), add MTP and AWQ comparison to the card
659937b verified
Raw History Blame Contribute Delete
9.58 kB
{
"architectures": [
"SenseNovaLoopForConditionalGeneration"
],
"attn_implementation": "sdpa",
"bos_token_id": null,
"dtype": "bfloat16",
"eos_token_id": 248046,
"hidden_size": 4096,
"image_token_id": 248056,
"loop_settings": {
"counts": [
2
],
"end": 28,
"eval_count": 2,
"start": 4
},
"model_type": "sensenova_loop",
"pad_token_id": 248044,
"patch_conv3d": true,
"text_config": {
"attention_bias": false,
"attention_dropout": 0.0,
"attn_output_gate": true,
"bos_token_id": null,
"dtype": "bfloat16",
"eos_token_id": 248044,
"full_attention_interval": 4,
"head_dim": 256,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 12288,
"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_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": 32,
"linear_value_head_dim": 128,
"mamba_ssm_dtype": "float32",
"max_position_embeddings": 524288,
"mlp_only_layers": [],
"model_type": "qwen3_5_text",
"mtp_num_hidden_layers": 1,
"mtp_use_dedicated_embeddings": false,
"num_attention_heads": 16,
"num_hidden_layers": 32,
"num_key_value_heads": 4,
"pad_token_id": 248044,
"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,
"rope_type": "yarn",
"factor": 2.0,
"original_max_position_embeddings": 262144
},
"tie_word_embeddings": false,
"use_cache": true,
"vocab_size": 248320
},
"tie_word_embeddings": false,
"use_cache": true,
"video_token_id": 248057,
"vision_config": {
"deepstack_visual_indexes": [],
"depth": 27,
"dtype": "bfloat16",
"hidden_act": "gelu_pytorch_tanh",
"hidden_size": 1152,
"in_channels": 3,
"initializer_range": 0.02,
"intermediate_size": 4304,
"model_type": "qwen3_5_vision",
"num_heads": 16,
"num_position_embeddings": 2304,
"out_hidden_size": 4096,
"patch_size": 16,
"spatial_merge_size": 2,
"temporal_patch_size": 2
},
"vision_end_token_id": 248054,
"vision_start_token_id": 248053,
"mhc": {
"end": 28,
"fixed_identity": false,
"iterations": 20,
"pre_scale": 1.0,
"start": 4,
"streams": 2
},
"model_name": "SenseNova-Loop-9B",
"auto_map": {
"AutoConfig": "modeling_sensenova_loop.SenseNovaLoopConfig",
"AutoModel": "modeling_sensenova_loop.SenseNovaLoopModel",
"AutoModelForCausalLM": "modeling_sensenova_loop.SenseNovaLoopForConditionalGeneration",
"AutoModelForImageTextToText": "modeling_sensenova_loop.SenseNovaLoopForConditionalGeneration"
},
"quantization_config": {
"config_groups": {
"group_0": {
"input_activations": {
"dynamic": false,
"num_bits": 4,
"type": "float",
"group_size": 16
},
"weights": {
"dynamic": false,
"num_bits": 4,
"type": "float",
"group_size": 16
},
"targets": [
"Linear"
]
}
},
"ignore": [
"model.language_model.layers.0.linear_attn.in_proj_a",
"model.language_model.layers.0.linear_attn.in_proj_b",
"model.language_model.layers.1.linear_attn.in_proj_a",
"model.language_model.layers.1.linear_attn.in_proj_b",
"model.language_model.layers.10.linear_attn.in_proj_a",
"model.language_model.layers.10.linear_attn.in_proj_b",
"model.language_model.layers.10.mhc_attn",
"model.language_model.layers.10.mhc_ffn",
"model.language_model.layers.11.mhc_attn",
"model.language_model.layers.11.mhc_ffn",
"model.language_model.layers.12.linear_attn.in_proj_a",
"model.language_model.layers.12.linear_attn.in_proj_b",
"model.language_model.layers.12.mhc_attn",
"model.language_model.layers.12.mhc_ffn",
"model.language_model.layers.13.linear_attn.in_proj_a",
"model.language_model.layers.13.linear_attn.in_proj_b",
"model.language_model.layers.13.mhc_attn",
"model.language_model.layers.13.mhc_ffn",
"model.language_model.layers.14.linear_attn.in_proj_a",
"model.language_model.layers.14.linear_attn.in_proj_b",
"model.language_model.layers.14.mhc_attn",
"model.language_model.layers.14.mhc_ffn",
"model.language_model.layers.15.mhc_attn",
"model.language_model.layers.15.mhc_ffn",
"model.language_model.layers.16.linear_attn.in_proj_a",
"model.language_model.layers.16.linear_attn.in_proj_b",
"model.language_model.layers.16.mhc_attn",
"model.language_model.layers.16.mhc_ffn",
"model.language_model.layers.17.linear_attn.in_proj_a",
"model.language_model.layers.17.linear_attn.in_proj_b",
"model.language_model.layers.17.mhc_attn",
"model.language_model.layers.17.mhc_ffn",
"model.language_model.layers.18.linear_attn.in_proj_a",
"model.language_model.layers.18.linear_attn.in_proj_b",
"model.language_model.layers.18.mhc_attn",
"model.language_model.layers.18.mhc_ffn",
"model.language_model.layers.19.mhc_attn",
"model.language_model.layers.19.mhc_ffn",
"model.language_model.layers.2.linear_attn.in_proj_a",
"model.language_model.layers.2.linear_attn.in_proj_b",
"model.language_model.layers.20.linear_attn.in_proj_a",
"model.language_model.layers.20.linear_attn.in_proj_b",
"model.language_model.layers.20.mhc_attn",
"model.language_model.layers.20.mhc_ffn",
"model.language_model.layers.21.linear_attn.in_proj_a",
"model.language_model.layers.21.linear_attn.in_proj_b",
"model.language_model.layers.21.mhc_attn",
"model.language_model.layers.21.mhc_ffn",
"model.language_model.layers.22.linear_attn.in_proj_a",
"model.language_model.layers.22.linear_attn.in_proj_b",
"model.language_model.layers.22.mhc_attn",
"model.language_model.layers.22.mhc_ffn",
"model.language_model.layers.23.mhc_attn",
"model.language_model.layers.23.mhc_ffn",
"model.language_model.layers.24.linear_attn.in_proj_a",
"model.language_model.layers.24.linear_attn.in_proj_b",
"model.language_model.layers.24.mhc_attn",
"model.language_model.layers.24.mhc_ffn",
"model.language_model.layers.25.linear_attn.in_proj_a",
"model.language_model.layers.25.linear_attn.in_proj_b",
"model.language_model.layers.25.mhc_attn",
"model.language_model.layers.25.mhc_ffn",
"model.language_model.layers.26.linear_attn.in_proj_a",
"model.language_model.layers.26.linear_attn.in_proj_b",
"model.language_model.layers.26.mhc_attn",
"model.language_model.layers.26.mhc_ffn",
"model.language_model.layers.27.mhc_attn",
"model.language_model.layers.27.mhc_ffn",
"model.language_model.layers.28.linear_attn.in_proj_a",
"model.language_model.layers.28.linear_attn.in_proj_b",
"model.language_model.layers.29.linear_attn.in_proj_a",
"model.language_model.layers.29.linear_attn.in_proj_b",
"model.language_model.layers.30.linear_attn.in_proj_a",
"model.language_model.layers.30.linear_attn.in_proj_b",
"model.language_model.layers.4.linear_attn.in_proj_a",
"model.language_model.layers.4.linear_attn.in_proj_b",
"model.language_model.layers.4.mhc_attn",
"model.language_model.layers.4.mhc_ffn",
"model.language_model.layers.5.linear_attn.in_proj_a",
"model.language_model.layers.5.linear_attn.in_proj_b",
"model.language_model.layers.5.mhc_attn",
"model.language_model.layers.5.mhc_ffn",
"model.language_model.layers.6.linear_attn.in_proj_a",
"model.language_model.layers.6.linear_attn.in_proj_b",
"model.language_model.layers.6.mhc_attn",
"model.language_model.layers.6.mhc_ffn",
"model.language_model.layers.7.mhc_attn",
"model.language_model.layers.7.mhc_ffn",
"model.language_model.layers.8.linear_attn.in_proj_a",
"model.language_model.layers.8.linear_attn.in_proj_b",
"model.language_model.layers.8.mhc_attn",
"model.language_model.layers.8.mhc_ffn",
"model.language_model.layers.9.linear_attn.in_proj_a",
"model.language_model.layers.9.linear_attn.in_proj_b",
"model.language_model.layers.9.mhc_attn",
"model.language_model.layers.9.mhc_ffn",
"model.visual*",
"mtp*"
],
"quant_algo": "W4A16_NVFP4",
"producer": {
"name": "modelopt",
"version": "0.43.0"
},
"quant_method": "modelopt"
}
}