VikramPal's picture
Add files using upload-large-folder tool
11c9737 verified
Raw
History Blame Contribute Delete
15.5 kB
{
"architectures": [
"Lfm2MoeForCausalLM"
],
"bos_token_id": 124894,
"conv_L_cache": 3,
"conv_bias": false,
"dtype": "bfloat16",
"eos_token_id": 124900,
"hidden_size": 2048,
"initializer_range": 0.02,
"intermediate_size": 7168,
"layer_types": [
"conv",
"conv",
"full_attention",
"conv",
"conv",
"conv",
"full_attention",
"conv",
"conv",
"conv",
"full_attention",
"conv",
"conv",
"conv",
"full_attention",
"conv",
"conv",
"conv",
"full_attention",
"conv",
"conv",
"full_attention",
"conv",
"conv"
],
"max_position_embeddings": 128000,
"model_type": "lfm2_moe",
"moe_intermediate_size": 1792,
"norm_eps": 1e-05,
"norm_topk_prob": true,
"num_attention_heads": 32,
"num_dense_layers": 2,
"num_experts": 32,
"num_experts_per_tok": 4,
"num_hidden_layers": 24,
"num_key_value_heads": 8,
"pad_token_id": 124893,
"quantization_config": {
"checkpoint_format": "dynquant-packed",
"group_size": 128,
"lm_head_quantized": false,
"modules": {
"model.embed_tokens": {
"bits": 8,
"out_features": 128000
},
"model.layers.0.conv.in_proj": {
"bits": 4,
"out_features": 6144
},
"model.layers.0.conv.out_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.0.feed_forward.w1": {
"bits": 4,
"out_features": 7168
},
"model.layers.0.feed_forward.w2": {
"bits": 8,
"out_features": 2048
},
"model.layers.0.feed_forward.w3": {
"bits": 4,
"out_features": 7168
},
"model.layers.1.conv.in_proj": {
"bits": 4,
"out_features": 6144
},
"model.layers.1.conv.out_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.1.feed_forward.w1": {
"bits": 8,
"out_features": 7168
},
"model.layers.1.feed_forward.w2": {
"bits": 8,
"out_features": 2048
},
"model.layers.1.feed_forward.w3": {
"bits": 8,
"out_features": 7168
},
"model.layers.10.feed_forward.experts.down_proj": {
"bits": 2,
"out_features": 65536
},
"model.layers.10.feed_forward.experts.gate_up_proj": {
"bits": 4,
"out_features": 114688
},
"model.layers.10.feed_forward.gate": {
"bits": 8,
"out_features": 32
},
"model.layers.10.self_attn.k_proj": {
"bits": 8,
"out_features": 512
},
"model.layers.10.self_attn.out_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.10.self_attn.q_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.10.self_attn.v_proj": {
"bits": 8,
"out_features": 512
},
"model.layers.11.conv.in_proj": {
"bits": 4,
"out_features": 6144
},
"model.layers.11.conv.out_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.11.feed_forward.experts.down_proj": {
"bits": 3,
"out_features": 65536
},
"model.layers.11.feed_forward.experts.gate_up_proj": {
"bits": 4,
"out_features": 114688
},
"model.layers.11.feed_forward.gate": {
"bits": 8,
"out_features": 32
},
"model.layers.12.conv.in_proj": {
"bits": 4,
"out_features": 6144
},
"model.layers.12.conv.out_proj": {
"bits": 4,
"out_features": 2048
},
"model.layers.12.feed_forward.experts.down_proj": {
"bits": 2,
"out_features": 65536
},
"model.layers.12.feed_forward.experts.gate_up_proj": {
"bits": 4,
"out_features": 114688
},
"model.layers.12.feed_forward.gate": {
"bits": 8,
"out_features": 32
},
"model.layers.13.conv.in_proj": {
"bits": 4,
"out_features": 6144
},
"model.layers.13.conv.out_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.13.feed_forward.experts.down_proj": {
"bits": 2,
"out_features": 65536
},
"model.layers.13.feed_forward.experts.gate_up_proj": {
"bits": 4,
"out_features": 114688
},
"model.layers.13.feed_forward.gate": {
"bits": 8,
"out_features": 32
},
"model.layers.14.feed_forward.experts.down_proj": {
"bits": 3,
"out_features": 65536
},
"model.layers.14.feed_forward.experts.gate_up_proj": {
"bits": 4,
"out_features": 114688
},
"model.layers.14.feed_forward.gate": {
"bits": 8,
"out_features": 32
},
"model.layers.14.self_attn.k_proj": {
"bits": 8,
"out_features": 512
},
"model.layers.14.self_attn.out_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.14.self_attn.q_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.14.self_attn.v_proj": {
"bits": 8,
"out_features": 512
},
"model.layers.15.conv.in_proj": {
"bits": 4,
"out_features": 6144
},
"model.layers.15.conv.out_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.15.feed_forward.experts.down_proj": {
"bits": 2,
"out_features": 65536
},
"model.layers.15.feed_forward.experts.gate_up_proj": {
"bits": 4,
"out_features": 114688
},
"model.layers.15.feed_forward.gate": {
"bits": 8,
"out_features": 32
},
"model.layers.16.conv.in_proj": {
"bits": 4,
"out_features": 6144
},
"model.layers.16.conv.out_proj": {
"bits": 4,
"out_features": 2048
},
"model.layers.16.feed_forward.experts.down_proj": {
"bits": 3,
"out_features": 65536
},
"model.layers.16.feed_forward.experts.gate_up_proj": {
"bits": 4,
"out_features": 114688
},
"model.layers.16.feed_forward.gate": {
"bits": 8,
"out_features": 32
},
"model.layers.17.conv.in_proj": {
"bits": 4,
"out_features": 6144
},
"model.layers.17.conv.out_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.17.feed_forward.experts.down_proj": {
"bits": 3,
"out_features": 65536
},
"model.layers.17.feed_forward.experts.gate_up_proj": {
"bits": 4,
"out_features": 114688
},
"model.layers.17.feed_forward.gate": {
"bits": 8,
"out_features": 32
},
"model.layers.18.feed_forward.experts.down_proj": {
"bits": 3,
"out_features": 65536
},
"model.layers.18.feed_forward.experts.gate_up_proj": {
"bits": 4,
"out_features": 114688
},
"model.layers.18.feed_forward.gate": {
"bits": 8,
"out_features": 32
},
"model.layers.18.self_attn.k_proj": {
"bits": 8,
"out_features": 512
},
"model.layers.18.self_attn.out_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.18.self_attn.q_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.18.self_attn.v_proj": {
"bits": 8,
"out_features": 512
},
"model.layers.19.conv.in_proj": {
"bits": 4,
"out_features": 6144
},
"model.layers.19.conv.out_proj": {
"bits": 4,
"out_features": 2048
},
"model.layers.19.feed_forward.experts.down_proj": {
"bits": 4,
"out_features": 65536
},
"model.layers.19.feed_forward.experts.gate_up_proj": {
"bits": 4,
"out_features": 114688
},
"model.layers.19.feed_forward.gate": {
"bits": 8,
"out_features": 32
},
"model.layers.2.feed_forward.experts.down_proj": {
"bits": 4,
"out_features": 65536
},
"model.layers.2.feed_forward.experts.gate_up_proj": {
"bits": 4,
"out_features": 114688
},
"model.layers.2.feed_forward.gate": {
"bits": 8,
"out_features": 32
},
"model.layers.2.self_attn.k_proj": {
"bits": 8,
"out_features": 512
},
"model.layers.2.self_attn.out_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.2.self_attn.q_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.2.self_attn.v_proj": {
"bits": 8,
"out_features": 512
},
"model.layers.20.conv.in_proj": {
"bits": 4,
"out_features": 6144
},
"model.layers.20.conv.out_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.20.feed_forward.experts.down_proj": {
"bits": 3,
"out_features": 65536
},
"model.layers.20.feed_forward.experts.gate_up_proj": {
"bits": 4,
"out_features": 114688
},
"model.layers.20.feed_forward.gate": {
"bits": 8,
"out_features": 32
},
"model.layers.21.feed_forward.experts.down_proj": {
"bits": 3,
"out_features": 65536
},
"model.layers.21.feed_forward.experts.gate_up_proj": {
"bits": 4,
"out_features": 114688
},
"model.layers.21.feed_forward.gate": {
"bits": 8,
"out_features": 32
},
"model.layers.21.self_attn.k_proj": {
"bits": 8,
"out_features": 512
},
"model.layers.21.self_attn.out_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.21.self_attn.q_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.21.self_attn.v_proj": {
"bits": 8,
"out_features": 512
},
"model.layers.22.conv.in_proj": {
"bits": 4,
"out_features": 6144
},
"model.layers.22.conv.out_proj": {
"bits": 4,
"out_features": 2048
},
"model.layers.22.feed_forward.experts.down_proj": {
"bits": 3,
"out_features": 65536
},
"model.layers.22.feed_forward.experts.gate_up_proj": {
"bits": 4,
"out_features": 114688
},
"model.layers.22.feed_forward.gate": {
"bits": 8,
"out_features": 32
},
"model.layers.23.conv.in_proj": {
"bits": 4,
"out_features": 6144
},
"model.layers.23.conv.out_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.23.feed_forward.experts.down_proj": {
"bits": 4,
"out_features": 65536
},
"model.layers.23.feed_forward.experts.gate_up_proj": {
"bits": 4,
"out_features": 114688
},
"model.layers.23.feed_forward.gate": {
"bits": 8,
"out_features": 32
},
"model.layers.3.conv.in_proj": {
"bits": 4,
"out_features": 6144
},
"model.layers.3.conv.out_proj": {
"bits": 4,
"out_features": 2048
},
"model.layers.3.feed_forward.experts.down_proj": {
"bits": 3,
"out_features": 65536
},
"model.layers.3.feed_forward.experts.gate_up_proj": {
"bits": 4,
"out_features": 114688
},
"model.layers.3.feed_forward.gate": {
"bits": 8,
"out_features": 32
},
"model.layers.4.conv.in_proj": {
"bits": 4,
"out_features": 6144
},
"model.layers.4.conv.out_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.4.feed_forward.experts.down_proj": {
"bits": 3,
"out_features": 65536
},
"model.layers.4.feed_forward.experts.gate_up_proj": {
"bits": 4,
"out_features": 114688
},
"model.layers.4.feed_forward.gate": {
"bits": 8,
"out_features": 32
},
"model.layers.5.conv.in_proj": {
"bits": 4,
"out_features": 6144
},
"model.layers.5.conv.out_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.5.feed_forward.experts.down_proj": {
"bits": 4,
"out_features": 65536
},
"model.layers.5.feed_forward.experts.gate_up_proj": {
"bits": 4,
"out_features": 114688
},
"model.layers.5.feed_forward.gate": {
"bits": 8,
"out_features": 32
},
"model.layers.6.feed_forward.experts.down_proj": {
"bits": 3,
"out_features": 65536
},
"model.layers.6.feed_forward.experts.gate_up_proj": {
"bits": 4,
"out_features": 114688
},
"model.layers.6.feed_forward.gate": {
"bits": 8,
"out_features": 32
},
"model.layers.6.self_attn.k_proj": {
"bits": 8,
"out_features": 512
},
"model.layers.6.self_attn.out_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.6.self_attn.q_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.6.self_attn.v_proj": {
"bits": 8,
"out_features": 512
},
"model.layers.7.conv.in_proj": {
"bits": 4,
"out_features": 6144
},
"model.layers.7.conv.out_proj": {
"bits": 4,
"out_features": 2048
},
"model.layers.7.feed_forward.experts.down_proj": {
"bits": 3,
"out_features": 65536
},
"model.layers.7.feed_forward.experts.gate_up_proj": {
"bits": 4,
"out_features": 114688
},
"model.layers.7.feed_forward.gate": {
"bits": 8,
"out_features": 32
},
"model.layers.8.conv.in_proj": {
"bits": 4,
"out_features": 6144
},
"model.layers.8.conv.out_proj": {
"bits": 4,
"out_features": 2048
},
"model.layers.8.feed_forward.experts.down_proj": {
"bits": 3,
"out_features": 65536
},
"model.layers.8.feed_forward.experts.gate_up_proj": {
"bits": 4,
"out_features": 114688
},
"model.layers.8.feed_forward.gate": {
"bits": 8,
"out_features": 32
},
"model.layers.9.conv.in_proj": {
"bits": 8,
"out_features": 6144
},
"model.layers.9.conv.out_proj": {
"bits": 8,
"out_features": 2048
},
"model.layers.9.feed_forward.experts.down_proj": {
"bits": 3,
"out_features": 65536
},
"model.layers.9.feed_forward.experts.gate_up_proj": {
"bits": 4,
"out_features": 114688
},
"model.layers.9.feed_forward.gate": {
"bits": 8,
"out_features": 32
}
},
"modules_to_not_convert": [],
"quant_method": "dynquant",
"schema_version": 1,
"symmetric": false,
"version": "0.3.0"
},
"rope_parameters": {
"rope_theta": 5000000,
"rope_type": "default"
},
"routed_scaling_factor": 1.0,
"tie_word_embeddings": true,
"transformers_version": "5.14.1",
"use_cache": true,
"use_expert_bias": true,
"vocab_size": 128000
}