Hugging Face's logo

mlx-community
/
Hunyuan-A13B-Instruct-mixed-6-8bit

Text Generation
MLX
Safetensors
hunyuan
conversational
custom_code
4-bit precision
Model card Files
xet
Community
Hunyuan-A13B-Instruct-mixed-6-8bit / config.json
m-i's picture
m-i
Add files using upload-large-folder tool
c7d9a3d verified
raw
history blame contribute delete
105 kB
 {
"add_classification_head": false,
"anyres_pooling_size": 2,
"anyres_vit_max_image_size": null,
"anyres_vit_two_views": false,
"architectures": [
"HunYuanMoEV1ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.1,
"attention_head_dim": 128,
"auto_map": {
"AutoConfig": "configuration_hunyuan.HunYuanConfig",
"AutoModel": "hunyuan.HunYuanModel",
"AutoModelForCausalLM": "hunyuan.HunYuanMoEV1ForCausalLM"
},
"bos_token_id": 1,
"cla_share_factor": 2,
"class_num": 0,
"dense_list": [
4096,
0
],
"eod_token_id": 127967,
"eos_token_id": 127960,
"group_limited_greedy": false,
"hidden_act": "silu",
"hidden_size": 4096,
"im_end_id": 6,
"im_newline_id": 12,
"im_start_id": 5,
"image_token_id": 9,
"initializer_range": 0.02,
"intermediate_size": 3072,
"kv_lora_rank": null,
"mask_init_id": 13,
"max_position_embeddings": 32768,
"mlp_bias": false,
"model_type": "hunyuan",
"moe_drop_tokens": false,
"moe_intermediate_size": [
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072,
3072
],
"moe_layer_num_skipped": 0,
"moe_random_routing_dropped_token": false,
"moe_topk": [
8,
8,
8,
8,
8,
8,
8,
8,
8,
8,
8,
8,
8,
8,
8,
8,
8,
8,
8,
8,
8,
8,
8,
8,
8,
8,
8,
8,
8,
8,
8,
8
],
"n_group": null,
"norm_topk_prob": true,
"norm_type": "rms",
"num_attention_heads": 32,
"num_experts": 64,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"num_media_embeds": 257,
"num_shared_expert": [
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1
],
"org_vocab_size": 128167,
"pad_id": 127961,
"pad_token_id": 127961,
"pool_type": "last",
"position_embedding_xdrope": false,
"pretraining_tp": 1,
"q_lora_rank": null,
"qk_nope_head_dim": null,
"qk_rope_head_dim": null,
"quantization": {
"group_size": 64,
"bits": 4,
"model.embed_tokens": {
"bits": 8,
"group_size": 64
},
"model.layers.0.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.0.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.0.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.0.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.0.self_attn.query_layernorm": false,
"model.layers.0.self_attn.key_layernorm": false,
"model.layers.0.self_attn.rope": false,
"model.layers.0.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.0.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.0.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.0.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.0.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.0.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.0.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.0.mlp.switch_mlp.activation": false,
"model.layers.0.input_layernorm": false,
"model.layers.0.post_attention_layernorm": false,
"model.layers.1.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.1.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.1.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.1.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.1.self_attn.query_layernorm": false,
"model.layers.1.self_attn.key_layernorm": false,
"model.layers.1.self_attn.rope": false,
"model.layers.1.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.1.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.1.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.1.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.1.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.1.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.1.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.1.mlp.switch_mlp.activation": false,
"model.layers.1.input_layernorm": false,
"model.layers.1.post_attention_layernorm": false,
"model.layers.2.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.2.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.2.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.2.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.2.self_attn.query_layernorm": false,
"model.layers.2.self_attn.key_layernorm": false,
"model.layers.2.self_attn.rope": false,
"model.layers.2.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.2.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.2.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.2.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.2.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.2.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.2.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.2.mlp.switch_mlp.activation": false,
"model.layers.2.input_layernorm": false,
"model.layers.2.post_attention_layernorm": false,
"model.layers.3.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.3.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.3.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.3.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.3.self_attn.query_layernorm": false,
"model.layers.3.self_attn.key_layernorm": false,
"model.layers.3.self_attn.rope": false,
"model.layers.3.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.3.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.3.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.3.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.3.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.3.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.3.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.3.mlp.switch_mlp.activation": false,
"model.layers.3.input_layernorm": false,
"model.layers.3.post_attention_layernorm": false,
"model.layers.4.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.4.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.4.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.4.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.4.self_attn.query_layernorm": false,
"model.layers.4.self_attn.key_layernorm": false,
"model.layers.4.self_attn.rope": false,
"model.layers.4.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.4.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.4.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.4.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.4.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.4.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.4.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.4.mlp.switch_mlp.activation": false,
"model.layers.4.input_layernorm": false,
"model.layers.4.post_attention_layernorm": false,
"model.layers.5.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.5.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.5.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.5.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.5.self_attn.query_layernorm": false,
"model.layers.5.self_attn.key_layernorm": false,
"model.layers.5.self_attn.rope": false,
"model.layers.5.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.5.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.5.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.5.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.5.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.5.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.5.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.5.mlp.switch_mlp.activation": false,
"model.layers.5.input_layernorm": false,
"model.layers.5.post_attention_layernorm": false,
"model.layers.6.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.6.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.6.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.6.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.6.self_attn.query_layernorm": false,
"model.layers.6.self_attn.key_layernorm": false,
"model.layers.6.self_attn.rope": false,
"model.layers.6.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.6.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.6.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.6.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.6.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.6.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.6.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.6.mlp.switch_mlp.activation": false,
"model.layers.6.input_layernorm": false,
"model.layers.6.post_attention_layernorm": false,
"model.layers.7.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.7.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.7.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.7.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.7.self_attn.query_layernorm": false,
"model.layers.7.self_attn.key_layernorm": false,
"model.layers.7.self_attn.rope": false,
"model.layers.7.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.7.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.7.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.7.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.7.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.7.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.7.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.7.mlp.switch_mlp.activation": false,
"model.layers.7.input_layernorm": false,
"model.layers.7.post_attention_layernorm": false,
"model.layers.8.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.8.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.8.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.8.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.8.self_attn.query_layernorm": false,
"model.layers.8.self_attn.key_layernorm": false,
"model.layers.8.self_attn.rope": false,
"model.layers.8.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.8.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.8.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.8.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.8.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.8.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.8.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.8.mlp.switch_mlp.activation": false,
"model.layers.8.input_layernorm": false,
"model.layers.8.post_attention_layernorm": false,
"model.layers.9.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.9.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.9.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.9.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.9.self_attn.query_layernorm": false,
"model.layers.9.self_attn.key_layernorm": false,
"model.layers.9.self_attn.rope": false,
"model.layers.9.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.9.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.9.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.9.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.9.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.9.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.9.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.9.mlp.switch_mlp.activation": false,
"model.layers.9.input_layernorm": false,
"model.layers.9.post_attention_layernorm": false,
"model.layers.10.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.10.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.10.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.10.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.10.self_attn.query_layernorm": false,
"model.layers.10.self_attn.key_layernorm": false,
"model.layers.10.self_attn.rope": false,
"model.layers.10.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.10.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.10.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.10.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.10.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.10.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.10.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.10.mlp.switch_mlp.activation": false,
"model.layers.10.input_layernorm": false,
"model.layers.10.post_attention_layernorm": false,
"model.layers.11.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.11.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.11.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.11.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.11.self_attn.query_layernorm": false,
"model.layers.11.self_attn.key_layernorm": false,
"model.layers.11.self_attn.rope": false,
"model.layers.11.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.11.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.11.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.11.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.11.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.11.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.11.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.11.mlp.switch_mlp.activation": false,
"model.layers.11.input_layernorm": false,
"model.layers.11.post_attention_layernorm": false,
"model.layers.12.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.12.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.12.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.12.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.12.self_attn.query_layernorm": false,
"model.layers.12.self_attn.key_layernorm": false,
"model.layers.12.self_attn.rope": false,
"model.layers.12.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.12.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.12.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.12.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.12.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.12.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.12.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.12.mlp.switch_mlp.activation": false,
"model.layers.12.input_layernorm": false,
"model.layers.12.post_attention_layernorm": false,
"model.layers.13.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.13.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.13.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.13.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.13.self_attn.query_layernorm": false,
"model.layers.13.self_attn.key_layernorm": false,
"model.layers.13.self_attn.rope": false,
"model.layers.13.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.13.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.13.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.13.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.13.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.13.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.13.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.13.mlp.switch_mlp.activation": false,
"model.layers.13.input_layernorm": false,
"model.layers.13.post_attention_layernorm": false,
"model.layers.14.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.14.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.14.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.14.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.14.self_attn.query_layernorm": false,
"model.layers.14.self_attn.key_layernorm": false,
"model.layers.14.self_attn.rope": false,
"model.layers.14.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.14.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.14.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.14.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.14.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.14.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.14.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.14.mlp.switch_mlp.activation": false,
"model.layers.14.input_layernorm": false,
"model.layers.14.post_attention_layernorm": false,
"model.layers.15.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.15.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.15.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.15.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.15.self_attn.query_layernorm": false,
"model.layers.15.self_attn.key_layernorm": false,
"model.layers.15.self_attn.rope": false,
"model.layers.15.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.15.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.15.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.15.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.15.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.15.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.15.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.15.mlp.switch_mlp.activation": false,
"model.layers.15.input_layernorm": false,
"model.layers.15.post_attention_layernorm": false,
"model.layers.16.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.16.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.16.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.16.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.16.self_attn.query_layernorm": false,
"model.layers.16.self_attn.key_layernorm": false,
"model.layers.16.self_attn.rope": false,
"model.layers.16.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.16.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.16.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.16.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.16.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.16.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.16.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.16.mlp.switch_mlp.activation": false,
"model.layers.16.input_layernorm": false,
"model.layers.16.post_attention_layernorm": false,
"model.layers.17.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.17.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.17.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.17.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.17.self_attn.query_layernorm": false,
"model.layers.17.self_attn.key_layernorm": false,
"model.layers.17.self_attn.rope": false,
"model.layers.17.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.17.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.17.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.17.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.17.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.17.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.17.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.17.mlp.switch_mlp.activation": false,
"model.layers.17.input_layernorm": false,
"model.layers.17.post_attention_layernorm": false,
"model.layers.18.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.18.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.18.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.18.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.18.self_attn.query_layernorm": false,
"model.layers.18.self_attn.key_layernorm": false,
"model.layers.18.self_attn.rope": false,
"model.layers.18.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.18.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.18.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.18.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.18.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.18.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.18.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.18.mlp.switch_mlp.activation": false,
"model.layers.18.input_layernorm": false,
"model.layers.18.post_attention_layernorm": false,
"model.layers.19.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.19.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.19.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.19.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.19.self_attn.query_layernorm": false,
"model.layers.19.self_attn.key_layernorm": false,
"model.layers.19.self_attn.rope": false,
"model.layers.19.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.19.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.19.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.19.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.19.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.19.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.19.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.19.mlp.switch_mlp.activation": false,
"model.layers.19.input_layernorm": false,
"model.layers.19.post_attention_layernorm": false,
"model.layers.20.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.20.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.20.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.20.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.20.self_attn.query_layernorm": false,
"model.layers.20.self_attn.key_layernorm": false,
"model.layers.20.self_attn.rope": false,
"model.layers.20.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.20.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.20.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.20.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.20.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.20.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.20.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.20.mlp.switch_mlp.activation": false,
"model.layers.20.input_layernorm": false,
"model.layers.20.post_attention_layernorm": false,
"model.layers.21.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.21.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.21.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.21.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.21.self_attn.query_layernorm": false,
"model.layers.21.self_attn.key_layernorm": false,
"model.layers.21.self_attn.rope": false,
"model.layers.21.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.21.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.21.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.21.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.21.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.21.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.21.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.21.mlp.switch_mlp.activation": false,
"model.layers.21.input_layernorm": false,
"model.layers.21.post_attention_layernorm": false,
"model.layers.22.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.22.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.22.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.22.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.22.self_attn.query_layernorm": false,
"model.layers.22.self_attn.key_layernorm": false,
"model.layers.22.self_attn.rope": false,
"model.layers.22.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.22.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.22.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.22.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.22.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.22.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.22.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.22.mlp.switch_mlp.activation": false,
"model.layers.22.input_layernorm": false,
"model.layers.22.post_attention_layernorm": false,
"model.layers.23.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.23.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.23.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.23.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.23.self_attn.query_layernorm": false,
"model.layers.23.self_attn.key_layernorm": false,
"model.layers.23.self_attn.rope": false,
"model.layers.23.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.23.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.23.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.23.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.23.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.23.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.23.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.23.mlp.switch_mlp.activation": false,
"model.layers.23.input_layernorm": false,
"model.layers.23.post_attention_layernorm": false,
"model.layers.24.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.24.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.24.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.24.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.24.self_attn.query_layernorm": false,
"model.layers.24.self_attn.key_layernorm": false,
"model.layers.24.self_attn.rope": false,
"model.layers.24.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.24.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.24.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.24.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.24.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.24.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.24.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.24.mlp.switch_mlp.activation": false,
"model.layers.24.input_layernorm": false,
"model.layers.24.post_attention_layernorm": false,
"model.layers.25.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.25.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.25.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.25.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.25.self_attn.query_layernorm": false,
"model.layers.25.self_attn.key_layernorm": false,
"model.layers.25.self_attn.rope": false,
"model.layers.25.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.25.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.25.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.25.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.25.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.25.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.25.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.25.mlp.switch_mlp.activation": false,
"model.layers.25.input_layernorm": false,
"model.layers.25.post_attention_layernorm": false,
"model.layers.26.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.26.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.26.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.26.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.26.self_attn.query_layernorm": false,
"model.layers.26.self_attn.key_layernorm": false,
"model.layers.26.self_attn.rope": false,
"model.layers.26.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.26.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.26.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.26.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.26.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.26.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.26.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.26.mlp.switch_mlp.activation": false,
"model.layers.26.input_layernorm": false,
"model.layers.26.post_attention_layernorm": false,
"model.layers.27.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.27.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.27.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.27.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.27.self_attn.query_layernorm": false,
"model.layers.27.self_attn.key_layernorm": false,
"model.layers.27.self_attn.rope": false,
"model.layers.27.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.27.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.27.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.27.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.27.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.27.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.27.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.27.mlp.switch_mlp.activation": false,
"model.layers.27.input_layernorm": false,
"model.layers.27.post_attention_layernorm": false,
"model.layers.28.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.28.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.28.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.28.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.28.self_attn.query_layernorm": false,
"model.layers.28.self_attn.key_layernorm": false,
"model.layers.28.self_attn.rope": false,
"model.layers.28.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.28.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.28.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.28.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.28.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.28.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.28.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.28.mlp.switch_mlp.activation": false,
"model.layers.28.input_layernorm": false,
"model.layers.28.post_attention_layernorm": false,
"model.layers.29.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.29.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.29.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.29.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.29.self_attn.query_layernorm": false,
"model.layers.29.self_attn.key_layernorm": false,
"model.layers.29.self_attn.rope": false,
"model.layers.29.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.29.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.29.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.29.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.29.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.29.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.29.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.29.mlp.switch_mlp.activation": false,
"model.layers.29.input_layernorm": false,
"model.layers.29.post_attention_layernorm": false,
"model.layers.30.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.30.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.30.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.30.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.30.self_attn.query_layernorm": false,
"model.layers.30.self_attn.key_layernorm": false,
"model.layers.30.self_attn.rope": false,
"model.layers.30.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.30.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.30.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.30.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.30.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.30.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.30.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.30.mlp.switch_mlp.activation": false,
"model.layers.30.input_layernorm": false,
"model.layers.30.post_attention_layernorm": false,
"model.layers.31.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.31.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.31.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.31.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.31.self_attn.query_layernorm": false,
"model.layers.31.self_attn.key_layernorm": false,
"model.layers.31.self_attn.rope": false,
"model.layers.31.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.31.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.31.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.31.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.31.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.31.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.31.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.31.mlp.switch_mlp.activation": false,
"model.layers.31.input_layernorm": false,
"model.layers.31.post_attention_layernorm": false,
"model.norm": false
},
"quantization_config": {
"group_size": 64,
"bits": 4,
"model.embed_tokens": {
"bits": 8,
"group_size": 64
},
"model.layers.0.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.0.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.0.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.0.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.0.self_attn.query_layernorm": false,
"model.layers.0.self_attn.key_layernorm": false,
"model.layers.0.self_attn.rope": false,
"model.layers.0.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.0.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.0.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.0.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.0.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.0.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.0.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.0.mlp.switch_mlp.activation": false,
"model.layers.0.input_layernorm": false,
"model.layers.0.post_attention_layernorm": false,
"model.layers.1.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.1.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.1.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.1.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.1.self_attn.query_layernorm": false,
"model.layers.1.self_attn.key_layernorm": false,
"model.layers.1.self_attn.rope": false,
"model.layers.1.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.1.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.1.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.1.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.1.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.1.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.1.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.1.mlp.switch_mlp.activation": false,
"model.layers.1.input_layernorm": false,
"model.layers.1.post_attention_layernorm": false,
"model.layers.2.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.2.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.2.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.2.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.2.self_attn.query_layernorm": false,
"model.layers.2.self_attn.key_layernorm": false,
"model.layers.2.self_attn.rope": false,
"model.layers.2.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.2.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.2.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.2.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.2.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.2.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.2.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.2.mlp.switch_mlp.activation": false,
"model.layers.2.input_layernorm": false,
"model.layers.2.post_attention_layernorm": false,
"model.layers.3.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.3.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.3.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.3.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.3.self_attn.query_layernorm": false,
"model.layers.3.self_attn.key_layernorm": false,
"model.layers.3.self_attn.rope": false,
"model.layers.3.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.3.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.3.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.3.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.3.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.3.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.3.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.3.mlp.switch_mlp.activation": false,
"model.layers.3.input_layernorm": false,
"model.layers.3.post_attention_layernorm": false,
"model.layers.4.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.4.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.4.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.4.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.4.self_attn.query_layernorm": false,
"model.layers.4.self_attn.key_layernorm": false,
"model.layers.4.self_attn.rope": false,
"model.layers.4.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.4.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.4.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.4.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.4.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.4.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.4.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.4.mlp.switch_mlp.activation": false,
"model.layers.4.input_layernorm": false,
"model.layers.4.post_attention_layernorm": false,
"model.layers.5.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.5.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.5.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.5.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.5.self_attn.query_layernorm": false,
"model.layers.5.self_attn.key_layernorm": false,
"model.layers.5.self_attn.rope": false,
"model.layers.5.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.5.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.5.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.5.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.5.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.5.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.5.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.5.mlp.switch_mlp.activation": false,
"model.layers.5.input_layernorm": false,
"model.layers.5.post_attention_layernorm": false,
"model.layers.6.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.6.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.6.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.6.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.6.self_attn.query_layernorm": false,
"model.layers.6.self_attn.key_layernorm": false,
"model.layers.6.self_attn.rope": false,
"model.layers.6.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.6.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.6.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.6.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.6.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.6.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.6.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.6.mlp.switch_mlp.activation": false,
"model.layers.6.input_layernorm": false,
"model.layers.6.post_attention_layernorm": false,
"model.layers.7.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.7.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.7.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.7.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.7.self_attn.query_layernorm": false,
"model.layers.7.self_attn.key_layernorm": false,
"model.layers.7.self_attn.rope": false,
"model.layers.7.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.7.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.7.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.7.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.7.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.7.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.7.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.7.mlp.switch_mlp.activation": false,
"model.layers.7.input_layernorm": false,
"model.layers.7.post_attention_layernorm": false,
"model.layers.8.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.8.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.8.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.8.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.8.self_attn.query_layernorm": false,
"model.layers.8.self_attn.key_layernorm": false,
"model.layers.8.self_attn.rope": false,
"model.layers.8.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.8.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.8.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.8.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.8.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.8.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.8.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.8.mlp.switch_mlp.activation": false,
"model.layers.8.input_layernorm": false,
"model.layers.8.post_attention_layernorm": false,
"model.layers.9.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.9.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.9.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.9.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.9.self_attn.query_layernorm": false,
"model.layers.9.self_attn.key_layernorm": false,
"model.layers.9.self_attn.rope": false,
"model.layers.9.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.9.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.9.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.9.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.9.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.9.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.9.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.9.mlp.switch_mlp.activation": false,
"model.layers.9.input_layernorm": false,
"model.layers.9.post_attention_layernorm": false,
"model.layers.10.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.10.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.10.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.10.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.10.self_attn.query_layernorm": false,
"model.layers.10.self_attn.key_layernorm": false,
"model.layers.10.self_attn.rope": false,
"model.layers.10.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.10.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.10.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.10.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.10.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.10.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.10.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.10.mlp.switch_mlp.activation": false,
"model.layers.10.input_layernorm": false,
"model.layers.10.post_attention_layernorm": false,
"model.layers.11.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.11.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.11.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.11.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.11.self_attn.query_layernorm": false,
"model.layers.11.self_attn.key_layernorm": false,
"model.layers.11.self_attn.rope": false,
"model.layers.11.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.11.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.11.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.11.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.11.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.11.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.11.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.11.mlp.switch_mlp.activation": false,
"model.layers.11.input_layernorm": false,
"model.layers.11.post_attention_layernorm": false,
"model.layers.12.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.12.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.12.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.12.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.12.self_attn.query_layernorm": false,
"model.layers.12.self_attn.key_layernorm": false,
"model.layers.12.self_attn.rope": false,
"model.layers.12.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.12.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.12.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.12.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.12.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.12.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.12.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.12.mlp.switch_mlp.activation": false,
"model.layers.12.input_layernorm": false,
"model.layers.12.post_attention_layernorm": false,
"model.layers.13.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.13.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.13.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.13.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.13.self_attn.query_layernorm": false,
"model.layers.13.self_attn.key_layernorm": false,
"model.layers.13.self_attn.rope": false,
"model.layers.13.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.13.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.13.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.13.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.13.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.13.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.13.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.13.mlp.switch_mlp.activation": false,
"model.layers.13.input_layernorm": false,
"model.layers.13.post_attention_layernorm": false,
"model.layers.14.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.14.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.14.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.14.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.14.self_attn.query_layernorm": false,
"model.layers.14.self_attn.key_layernorm": false,
"model.layers.14.self_attn.rope": false,
"model.layers.14.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.14.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.14.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.14.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.14.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.14.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.14.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.14.mlp.switch_mlp.activation": false,
"model.layers.14.input_layernorm": false,
"model.layers.14.post_attention_layernorm": false,
"model.layers.15.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.15.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.15.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.15.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.15.self_attn.query_layernorm": false,
"model.layers.15.self_attn.key_layernorm": false,
"model.layers.15.self_attn.rope": false,
"model.layers.15.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.15.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.15.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.15.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.15.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.15.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.15.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.15.mlp.switch_mlp.activation": false,
"model.layers.15.input_layernorm": false,
"model.layers.15.post_attention_layernorm": false,
"model.layers.16.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.16.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.16.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.16.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.16.self_attn.query_layernorm": false,
"model.layers.16.self_attn.key_layernorm": false,
"model.layers.16.self_attn.rope": false,
"model.layers.16.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.16.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.16.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.16.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.16.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.16.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.16.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.16.mlp.switch_mlp.activation": false,
"model.layers.16.input_layernorm": false,
"model.layers.16.post_attention_layernorm": false,
"model.layers.17.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.17.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.17.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.17.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.17.self_attn.query_layernorm": false,
"model.layers.17.self_attn.key_layernorm": false,
"model.layers.17.self_attn.rope": false,
"model.layers.17.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.17.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.17.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.17.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.17.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.17.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.17.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.17.mlp.switch_mlp.activation": false,
"model.layers.17.input_layernorm": false,
"model.layers.17.post_attention_layernorm": false,
"model.layers.18.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.18.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.18.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.18.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.18.self_attn.query_layernorm": false,
"model.layers.18.self_attn.key_layernorm": false,
"model.layers.18.self_attn.rope": false,
"model.layers.18.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.18.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.18.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.18.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.18.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.18.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.18.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.18.mlp.switch_mlp.activation": false,
"model.layers.18.input_layernorm": false,
"model.layers.18.post_attention_layernorm": false,
"model.layers.19.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.19.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.19.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.19.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.19.self_attn.query_layernorm": false,
"model.layers.19.self_attn.key_layernorm": false,
"model.layers.19.self_attn.rope": false,
"model.layers.19.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.19.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.19.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.19.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.19.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.19.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.19.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.19.mlp.switch_mlp.activation": false,
"model.layers.19.input_layernorm": false,
"model.layers.19.post_attention_layernorm": false,
"model.layers.20.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.20.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.20.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.20.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.20.self_attn.query_layernorm": false,
"model.layers.20.self_attn.key_layernorm": false,
"model.layers.20.self_attn.rope": false,
"model.layers.20.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.20.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.20.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.20.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.20.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.20.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.20.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.20.mlp.switch_mlp.activation": false,
"model.layers.20.input_layernorm": false,
"model.layers.20.post_attention_layernorm": false,
"model.layers.21.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.21.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.21.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.21.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.21.self_attn.query_layernorm": false,
"model.layers.21.self_attn.key_layernorm": false,
"model.layers.21.self_attn.rope": false,
"model.layers.21.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.21.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.21.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.21.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.21.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.21.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.21.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.21.mlp.switch_mlp.activation": false,
"model.layers.21.input_layernorm": false,
"model.layers.21.post_attention_layernorm": false,
"model.layers.22.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.22.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.22.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.22.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.22.self_attn.query_layernorm": false,
"model.layers.22.self_attn.key_layernorm": false,
"model.layers.22.self_attn.rope": false,
"model.layers.22.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.22.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.22.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.22.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.22.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.22.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.22.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.22.mlp.switch_mlp.activation": false,
"model.layers.22.input_layernorm": false,
"model.layers.22.post_attention_layernorm": false,
"model.layers.23.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.23.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.23.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.23.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.23.self_attn.query_layernorm": false,
"model.layers.23.self_attn.key_layernorm": false,
"model.layers.23.self_attn.rope": false,
"model.layers.23.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.23.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.23.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.23.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.23.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.23.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.23.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.23.mlp.switch_mlp.activation": false,
"model.layers.23.input_layernorm": false,
"model.layers.23.post_attention_layernorm": false,
"model.layers.24.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.24.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.24.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.24.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.24.self_attn.query_layernorm": false,
"model.layers.24.self_attn.key_layernorm": false,
"model.layers.24.self_attn.rope": false,
"model.layers.24.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.24.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.24.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.24.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.24.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.24.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.24.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.24.mlp.switch_mlp.activation": false,
"model.layers.24.input_layernorm": false,
"model.layers.24.post_attention_layernorm": false,
"model.layers.25.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.25.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.25.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.25.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.25.self_attn.query_layernorm": false,
"model.layers.25.self_attn.key_layernorm": false,
"model.layers.25.self_attn.rope": false,
"model.layers.25.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.25.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.25.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.25.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.25.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.25.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.25.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.25.mlp.switch_mlp.activation": false,
"model.layers.25.input_layernorm": false,
"model.layers.25.post_attention_layernorm": false,
"model.layers.26.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.26.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.26.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.26.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.26.self_attn.query_layernorm": false,
"model.layers.26.self_attn.key_layernorm": false,
"model.layers.26.self_attn.rope": false,
"model.layers.26.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.26.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.26.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.26.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.26.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.26.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.26.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.26.mlp.switch_mlp.activation": false,
"model.layers.26.input_layernorm": false,
"model.layers.26.post_attention_layernorm": false,
"model.layers.27.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.27.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.27.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.27.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.27.self_attn.query_layernorm": false,
"model.layers.27.self_attn.key_layernorm": false,
"model.layers.27.self_attn.rope": false,
"model.layers.27.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.27.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.27.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.27.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.27.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.27.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.27.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.27.mlp.switch_mlp.activation": false,
"model.layers.27.input_layernorm": false,
"model.layers.27.post_attention_layernorm": false,
"model.layers.28.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.28.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.28.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.28.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.28.self_attn.query_layernorm": false,
"model.layers.28.self_attn.key_layernorm": false,
"model.layers.28.self_attn.rope": false,
"model.layers.28.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.28.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.28.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.28.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.28.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.28.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.28.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.28.mlp.switch_mlp.activation": false,
"model.layers.28.input_layernorm": false,
"model.layers.28.post_attention_layernorm": false,
"model.layers.29.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.29.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.29.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.29.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.29.self_attn.query_layernorm": false,
"model.layers.29.self_attn.key_layernorm": false,
"model.layers.29.self_attn.rope": false,
"model.layers.29.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.29.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.29.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.29.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.29.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.29.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.29.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.29.mlp.switch_mlp.activation": false,
"model.layers.29.input_layernorm": false,
"model.layers.29.post_attention_layernorm": false,
"model.layers.30.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.30.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.30.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.30.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.30.self_attn.query_layernorm": false,
"model.layers.30.self_attn.key_layernorm": false,
"model.layers.30.self_attn.rope": false,
"model.layers.30.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.30.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.30.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.30.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.30.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.30.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.30.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.30.mlp.switch_mlp.activation": false,
"model.layers.30.input_layernorm": false,
"model.layers.30.post_attention_layernorm": false,
"model.layers.31.self_attn.q_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.31.self_attn.k_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.31.self_attn.v_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.31.self_attn.o_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.31.self_attn.query_layernorm": false,
"model.layers.31.self_attn.key_layernorm": false,
"model.layers.31.self_attn.rope": false,
"model.layers.31.mlp.shared_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.31.mlp.shared_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.31.mlp.shared_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.31.mlp.gate.wg": {
"bits": 6,
"group_size": 64
},
"model.layers.31.mlp.switch_mlp.gate_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.31.mlp.switch_mlp.up_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.31.mlp.switch_mlp.down_proj": {
"bits": 6,
"group_size": 64
},
"model.layers.31.mlp.switch_mlp.activation": false,
"model.layers.31.input_layernorm": false,
"model.layers.31.post_attention_layernorm": false,
"model.norm": false
},
"rms_norm_eps": 1e-05,
"rope_scaling": {
"alpha": 1000.0,
"beta_fast": 32,
"beta_slow": 1,
"factor": 1.0,
"mscale": 1.0,
"mscale_all_dim": 1.0,
"type": "dynamic"
},
"rope_theta": 10000.0,
"routed_scaling_factor": 1.0,
"sep_token_id": 127962,
"skip_cls_token": false,
"text_end_id": 8,
"text_start_id": 7,
"tie_word_embeddings": true,
"topk_group": null,
"torch_dtype": "bfloat16",
"transformers_version": "4.41.2",
"use_cache": true,
"use_cla": false,
"use_mixed_mlp_moe": true,
"use_mla": false,
"use_qk_norm": true,
"use_rotary_pos_emb": true,
"v_head_dim": null,
"video_end_id": 11,
"video_start_id": 10,
"vit_add_patchemb_bias": false,
"vit_input_resolution": 224,
"vit_mapping_type": "resampler",
"vit_norm_type": "fused",
"vit_patch": 1,
"vit_path": null,
"vit_remove_prenorm": false,
"vit_token": 64,
"vit_type": null,
"vit_used_rms_norm": false,
"vocab_size": 128167,
"xdrope_section": null
}