prod-jev-9b / config.json
Will-work-4-tokens's picture
Initial release of TheWhyAI/prod_jev (NVFP4)
b3087a8 verified
Raw History Blame Contribute Delete
4.63 kB
{
"kv_cache_clamp": {
"enabled": true,
"limits": {
"3": {"k": 7.375, "v": 2.984},
"7": {"k": 8.000, "v": 3.266},
"11": {"k": 8.938, "v": 3.516},
"15": {"k": 7.875, "v": 4.562},
"19": {"k": 7.562, "v": 7.812},
"23": {"k": 8.375, "v": 10.250},
"27": {"k": 7.688, "v": 28.875},
"31": {"k": 6.000, "v": 19.875}
}
},
"architectures": [
"Qwen3_5ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"attn_output_gate": true,
"bos_token_id": null,
"dtype": "bfloat16",
"eos_token_id": 248046,
"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": 262144,
"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,
"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": "default"
},
"tie_word_embeddings": false,
"transformers_version": "5.22.2",
"use_cache": true,
"vocab_size": 248320,
"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": [
"lm_head",
"model.layers.0.linear_attn.conv1d",
"model.layers.1.linear_attn.conv1d",
"model.layers.10.linear_attn.conv1d",
"model.layers.12.linear_attn.conv1d",
"model.layers.13.linear_attn.conv1d",
"model.layers.14.linear_attn.conv1d",
"model.layers.16.linear_attn.conv1d",
"model.layers.17.linear_attn.conv1d",
"model.layers.18.linear_attn.conv1d",
"model.layers.2.linear_attn.conv1d",
"model.layers.20.linear_attn.conv1d",
"model.layers.21.linear_attn.conv1d",
"model.layers.22.linear_attn.conv1d",
"model.layers.24.linear_attn.conv1d",
"model.layers.25.linear_attn.conv1d",
"model.layers.26.linear_attn.conv1d",
"model.layers.28.linear_attn.conv1d",
"model.layers.29.linear_attn.conv1d",
"model.layers.30.linear_attn.conv1d",
"model.layers.4.linear_attn.conv1d",
"model.layers.5.linear_attn.conv1d",
"model.layers.6.linear_attn.conv1d",
"model.layers.8.linear_attn.conv1d",
"model.layers.9.linear_attn.conv1d"
],
"quant_algo": "NVFP4",
"producer": {
"name": "modelopt",
"version": "0.42.0"
},
"quant_method": "modelopt"
}
}