Text Generation
MLX
Safetensors
English
qwen3_moe
programming
code generation
code
coding
coder
chat
brainstorm
qwen
qwen3
qwencoder
brainstorm 20x
creative
all uses cases
Jan-V1
horror
science fiction
fantasy
Star Trek
Star Trek Original
Star Trek The Next Generation
Star Trek Deep Space Nine
Star Trek Voyager
Star Trek Enterprise
Star Trek Discovery.
finetune
thinking
reasoning
unsloth
6x6B
Mixture of Experts
mixture of experts
conversational
6-bit
| { | |
| "architectures": [ | |
| "Qwen3MoeForCausalLM" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 151643, | |
| "decoder_sparse_step": 1, | |
| "dtype": "bfloat16", | |
| "eos_token_id": 151645, | |
| "head_dim": 128, | |
| "hidden_act": "silu", | |
| "hidden_size": 2560, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 9728, | |
| "layer_types": [ | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention" | |
| ], | |
| "max_position_embeddings": 262144, | |
| "max_window_layers": 55, | |
| "mlp_only_layers": [], | |
| "model_type": "qwen3_moe", | |
| "moe_intermediate_size": 9728, | |
| "norm_topk_prob": true, | |
| "num_attention_heads": 32, | |
| "num_experts": 6, | |
| "num_experts_per_tok": 2, | |
| "num_hidden_layers": 55, | |
| "num_key_value_heads": 8, | |
| "output_router_logits": false, | |
| "quantization": { | |
| "group_size": 64, | |
| "bits": 6, | |
| "mode": "affine", | |
| "model.embed_tokens": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.0.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.0.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.0.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.0.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.0.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.0.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.0.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.0.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.1.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.1.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.1.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.1.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.1.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.1.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.1.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.1.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.2.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.2.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.2.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.2.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.2.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.2.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.2.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.2.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.3.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.3.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.3.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.3.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.3.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.3.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.3.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.3.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.4.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.4.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.4.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.4.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.4.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.4.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.4.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.4.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.5.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.5.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.5.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.5.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.5.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.5.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.5.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.5.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.6.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.6.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.6.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.6.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.6.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.6.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.6.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.6.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.7.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.7.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.7.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.7.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.7.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.7.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.7.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.7.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.8.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.8.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.8.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.8.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.8.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.8.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.8.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.8.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.9.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.9.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.9.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.9.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.9.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.9.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.9.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.9.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.10.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.10.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.10.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.10.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.10.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.10.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.10.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.10.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.11.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.11.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.11.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.11.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.11.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.11.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.11.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.11.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.12.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.12.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.12.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.12.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.12.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.12.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.12.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.12.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.13.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.13.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.13.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.13.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.13.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.13.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.13.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.13.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.14.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.14.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.14.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.14.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.14.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.14.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.14.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.14.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.15.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.15.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.15.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.15.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.15.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.15.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.15.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.15.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.16.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.16.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.16.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.16.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.16.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.16.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.16.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.16.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.17.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.17.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.17.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.17.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.17.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.17.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.17.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.17.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.18.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.18.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.18.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.18.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.18.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.18.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.18.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.18.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.19.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.19.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.19.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.19.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.19.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.19.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.19.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.19.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.20.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.20.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.20.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.20.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.20.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.20.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.20.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.20.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.21.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.21.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.21.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.21.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.21.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.21.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.21.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.21.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.22.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.22.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.22.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.22.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.22.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.22.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.22.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.22.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.23.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.23.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.23.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.23.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.23.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.23.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.23.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.23.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.24.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.24.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.24.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.24.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.24.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.24.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.24.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.24.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.25.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.25.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.25.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.25.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.25.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.25.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.25.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.25.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.26.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.26.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.26.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.26.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.26.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.26.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.26.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.26.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.27.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.27.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.27.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.27.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.27.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.27.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.27.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.27.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.28.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.28.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.28.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.28.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.28.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.28.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.28.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.28.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.29.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.29.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.29.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.29.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.29.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.29.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.29.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.29.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.30.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.30.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.30.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.30.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.30.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.30.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.30.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.30.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.31.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.31.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.31.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.31.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.31.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.31.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.31.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.31.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.32.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.32.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.32.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.32.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.32.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.32.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.32.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.32.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.33.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.33.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.33.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.33.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.33.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.33.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.33.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.33.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.34.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.34.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.34.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.34.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.34.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.34.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.34.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.34.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.35.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.35.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.35.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.35.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.35.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.35.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.35.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.35.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.36.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.36.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.36.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.36.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.36.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.36.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.36.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.36.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.37.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.37.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.37.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.37.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.37.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.37.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.37.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.37.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.38.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.38.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.38.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.38.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.38.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.38.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.38.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.38.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.39.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.39.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.39.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.39.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.39.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.39.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.39.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.39.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.40.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.40.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.40.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.40.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.40.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.40.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.40.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.40.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.41.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.41.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.41.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.41.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.41.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.41.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.41.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.41.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.42.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.42.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.42.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.42.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.42.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.42.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.42.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.42.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.43.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.43.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.43.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.43.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.43.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.43.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.43.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.43.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.44.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.44.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.44.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.44.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.44.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.44.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.44.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.44.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.45.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.45.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.45.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.45.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.45.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.45.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.45.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.45.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.46.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.46.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.46.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.46.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.46.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.46.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.46.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.46.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.47.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.47.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.47.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.47.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.47.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.47.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.47.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.47.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.48.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.48.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.48.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.48.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.48.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.48.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.48.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.48.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.49.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.49.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.49.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.49.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.49.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.49.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.49.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.49.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.50.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.50.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.50.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.50.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.50.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.50.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.50.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.50.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.51.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.51.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.51.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.51.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.51.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.51.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.51.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.51.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.52.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.52.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.52.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.52.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.52.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.52.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.52.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.52.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.53.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.53.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.53.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.53.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.53.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.53.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.53.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.53.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.54.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.54.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.54.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.54.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.54.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.54.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.54.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.54.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "lm_head": { | |
| "group_size": 64, | |
| "bits": 6 | |
| } | |
| }, | |
| "quantization_config": { | |
| "group_size": 64, | |
| "bits": 6, | |
| "mode": "affine", | |
| "model.embed_tokens": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.0.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.0.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.0.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.0.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.0.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.0.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.0.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.0.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.1.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.1.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.1.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.1.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.1.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.1.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.1.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.1.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.2.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.2.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.2.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.2.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.2.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.2.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.2.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.2.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.3.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.3.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.3.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.3.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.3.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.3.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.3.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.3.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.4.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.4.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.4.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.4.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.4.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.4.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.4.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.4.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.5.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.5.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.5.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.5.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.5.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.5.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.5.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.5.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.6.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.6.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.6.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.6.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.6.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.6.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.6.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.6.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.7.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.7.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.7.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.7.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.7.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.7.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.7.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.7.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.8.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.8.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.8.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.8.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.8.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.8.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.8.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.8.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.9.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.9.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.9.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.9.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.9.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.9.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.9.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.9.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.10.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.10.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.10.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.10.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.10.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.10.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.10.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.10.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.11.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.11.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.11.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.11.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.11.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.11.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.11.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.11.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.12.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.12.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.12.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.12.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.12.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.12.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.12.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.12.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.13.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.13.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.13.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.13.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.13.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.13.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.13.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.13.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.14.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.14.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.14.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.14.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.14.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.14.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.14.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.14.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.15.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.15.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.15.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.15.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.15.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.15.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.15.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.15.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.16.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.16.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.16.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.16.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.16.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.16.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.16.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.16.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.17.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.17.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.17.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.17.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.17.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.17.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.17.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.17.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.18.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.18.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.18.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.18.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.18.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.18.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.18.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.18.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.19.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.19.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.19.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.19.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.19.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.19.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.19.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.19.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.20.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.20.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.20.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.20.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.20.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.20.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.20.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.20.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.21.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.21.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.21.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.21.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.21.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.21.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.21.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.21.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.22.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.22.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.22.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.22.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.22.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.22.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.22.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.22.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.23.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.23.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.23.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.23.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.23.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.23.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.23.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.23.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.24.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.24.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.24.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.24.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.24.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.24.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.24.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.24.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.25.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.25.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.25.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.25.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.25.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.25.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.25.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.25.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.26.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.26.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.26.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.26.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.26.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.26.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.26.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.26.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.27.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.27.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.27.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.27.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.27.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.27.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.27.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.27.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.28.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.28.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.28.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.28.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.28.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.28.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.28.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.28.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.29.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.29.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.29.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.29.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.29.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.29.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.29.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.29.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.30.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.30.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.30.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.30.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.30.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.30.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.30.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.30.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.31.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.31.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.31.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.31.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.31.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.31.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.31.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.31.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.32.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.32.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.32.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.32.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.32.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.32.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.32.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.32.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.33.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.33.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.33.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.33.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.33.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.33.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.33.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.33.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.34.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.34.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.34.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.34.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.34.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.34.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.34.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.34.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.35.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.35.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.35.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.35.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.35.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.35.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.35.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.35.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.36.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.36.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.36.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.36.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.36.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.36.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.36.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.36.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.37.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.37.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.37.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.37.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.37.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.37.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.37.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.37.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.38.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.38.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.38.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.38.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.38.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.38.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.38.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.38.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.39.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.39.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.39.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.39.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.39.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.39.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.39.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.39.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.40.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.40.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.40.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.40.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.40.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.40.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.40.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.40.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.41.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.41.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.41.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.41.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.41.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.41.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.41.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.41.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.42.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.42.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.42.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.42.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.42.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.42.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.42.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.42.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.43.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.43.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.43.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.43.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.43.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.43.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.43.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.43.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.44.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.44.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.44.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.44.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.44.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.44.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.44.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.44.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.45.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.45.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.45.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.45.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.45.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.45.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.45.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.45.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.46.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.46.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.46.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.46.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.46.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.46.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.46.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.46.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.47.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.47.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.47.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.47.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.47.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.47.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.47.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.47.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.48.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.48.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.48.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.48.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.48.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.48.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.48.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.48.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.49.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.49.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.49.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.49.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.49.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.49.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.49.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.49.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.50.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.50.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.50.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.50.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.50.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.50.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.50.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.50.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.51.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.51.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.51.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.51.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.51.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.51.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.51.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.51.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.52.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.52.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.52.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.52.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.52.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.52.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.52.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.52.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.53.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.53.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.53.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.53.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.53.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.53.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.53.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.53.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.54.self_attn.q_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.54.self_attn.k_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.54.self_attn.v_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "model.layers.54.self_attn.o_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.54.mlp.gate": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.54.mlp.switch_mlp.gate_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.54.mlp.switch_mlp.up_proj": { | |
| "group_size": 32, | |
| "bits": 5 | |
| }, | |
| "model.layers.54.mlp.switch_mlp.down_proj": { | |
| "group_size": 32, | |
| "bits": 6 | |
| }, | |
| "lm_head": { | |
| "group_size": 64, | |
| "bits": 6 | |
| } | |
| }, | |
| "rms_norm_eps": 1e-06, | |
| "rope_scaling": null, | |
| "rope_theta": 5000000, | |
| "router_aux_loss_coef": 0.001, | |
| "sliding_window": null, | |
| "tie_word_embeddings": true, | |
| "transformers_version": "4.56.1", | |
| "use_cache": false, | |
| "use_sliding_window": false, | |
| "vocab_size": 151936 | |
| } |