| { | |
| "_name_or_path": "meta-llama/Llama-3.2-1B", | |
| "architectures": [ | |
| "LlamaForCausalLM" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "auto_map": { | |
| "AutoModelForCausalLM": "qpal_modelling_llama.QPalLlamaForCausalLM" | |
| }, | |
| "bos_token_id": 128000, | |
| "eos_token_id": 128001, | |
| "head_dim": 64, | |
| "hidden_act": "silu", | |
| "hidden_size": 2048, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 8192, | |
| "max_position_embeddings": 131072, | |
| "mlp_bias": false, | |
| "model_type": "llama", | |
| "num_attention_heads": 32, | |
| "num_hidden_layers": 16, | |
| "num_key_value_heads": 8, | |
| "pretraining_tp": 1, | |
| "qpal_quant_config": { | |
| "modules": { | |
| "model.layers.0.mlp.down_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 8192, | |
| "hadV": 2048, | |
| "in_features": 8192, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 8192, | |
| "in_part": [ | |
| 4096, | |
| 4096 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.0.mlp.gate_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.0.mlp.up_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.0.self_attn.k_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.0.self_attn.o_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.0.self_attn.q_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.0.self_attn.v_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.1.mlp.down_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 8192, | |
| "hadV": 2048, | |
| "in_features": 8192, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 8192, | |
| "in_part": [ | |
| 4096, | |
| 4096 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.1.mlp.gate_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.1.mlp.up_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.1.self_attn.k_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.1.self_attn.o_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.1.self_attn.q_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.1.self_attn.v_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.10.mlp.down_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 8192, | |
| "hadV": 2048, | |
| "in_features": 8192, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 8192, | |
| "in_part": [ | |
| 4096, | |
| 4096 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.10.mlp.gate_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.10.mlp.up_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.10.self_attn.k_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.10.self_attn.o_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.10.self_attn.q_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.10.self_attn.v_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.11.mlp.down_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 8192, | |
| "hadV": 2048, | |
| "in_features": 8192, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 8192, | |
| "in_part": [ | |
| 4096, | |
| 4096 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.11.mlp.gate_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.11.mlp.up_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.11.self_attn.k_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.11.self_attn.o_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.11.self_attn.q_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.11.self_attn.v_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.12.mlp.down_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 8192, | |
| "hadV": 2048, | |
| "in_features": 8192, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 8192, | |
| "in_part": [ | |
| 4096, | |
| 4096 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.12.mlp.gate_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.12.mlp.up_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.12.self_attn.k_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.12.self_attn.o_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.12.self_attn.q_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.12.self_attn.v_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.13.mlp.down_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 8192, | |
| "hadV": 2048, | |
| "in_features": 8192, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 8192, | |
| "in_part": [ | |
| 4096, | |
| 4096 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.13.mlp.gate_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.13.mlp.up_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.13.self_attn.k_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.13.self_attn.o_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.13.self_attn.q_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.13.self_attn.v_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.14.mlp.down_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 8192, | |
| "hadV": 2048, | |
| "in_features": 8192, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 8192, | |
| "in_part": [ | |
| 4096, | |
| 4096 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.14.mlp.gate_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.14.mlp.up_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.14.self_attn.k_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.14.self_attn.o_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.14.self_attn.q_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.14.self_attn.v_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.15.mlp.down_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 8192, | |
| "hadV": 2048, | |
| "in_features": 8192, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 8192, | |
| "in_part": [ | |
| 4096, | |
| 4096 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.15.mlp.gate_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.15.mlp.up_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.15.self_attn.k_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.15.self_attn.o_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.15.self_attn.q_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.15.self_attn.v_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.2.mlp.down_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 8192, | |
| "hadV": 2048, | |
| "in_features": 8192, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 8192, | |
| "in_part": [ | |
| 4096, | |
| 4096 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.2.mlp.gate_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.2.mlp.up_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.2.self_attn.k_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.2.self_attn.o_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.2.self_attn.q_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.2.self_attn.v_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.3.mlp.down_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 8192, | |
| "hadV": 2048, | |
| "in_features": 8192, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 8192, | |
| "in_part": [ | |
| 4096, | |
| 4096 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.3.mlp.gate_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.3.mlp.up_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.3.self_attn.k_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.3.self_attn.o_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.3.self_attn.q_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.3.self_attn.v_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.4.mlp.down_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 8192, | |
| "hadV": 2048, | |
| "in_features": 8192, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 8192, | |
| "in_part": [ | |
| 4096, | |
| 4096 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.4.mlp.gate_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.4.mlp.up_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.4.self_attn.k_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.4.self_attn.o_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.4.self_attn.q_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.4.self_attn.v_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.5.mlp.down_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 8192, | |
| "hadV": 2048, | |
| "in_features": 8192, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 8192, | |
| "in_part": [ | |
| 4096, | |
| 4096 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.5.mlp.gate_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.5.mlp.up_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.5.self_attn.k_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.5.self_attn.o_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.5.self_attn.q_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.5.self_attn.v_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.6.mlp.down_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 8192, | |
| "hadV": 2048, | |
| "in_features": 8192, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 8192, | |
| "in_part": [ | |
| 4096, | |
| 4096 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.6.mlp.gate_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.6.mlp.up_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.6.self_attn.k_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.6.self_attn.o_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.6.self_attn.q_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.6.self_attn.v_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.7.mlp.down_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 8192, | |
| "hadV": 2048, | |
| "in_features": 8192, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 8192, | |
| "in_part": [ | |
| 4096, | |
| 4096 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.7.mlp.gate_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.7.mlp.up_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.7.self_attn.k_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.7.self_attn.o_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.7.self_attn.q_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.7.self_attn.v_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.8.mlp.down_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 8192, | |
| "hadV": 2048, | |
| "in_features": 8192, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 8192, | |
| "in_part": [ | |
| 4096, | |
| 4096 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.8.mlp.gate_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.8.mlp.up_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.8.self_attn.k_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.8.self_attn.o_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.8.self_attn.q_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.8.self_attn.v_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.9.mlp.down_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 8192, | |
| "hadV": 2048, | |
| "in_features": 8192, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 8192, | |
| "in_part": [ | |
| 4096, | |
| 4096 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.9.mlp.gate_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.9.mlp.up_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 8192, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 8192, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 8192, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.9.self_attn.k_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.9.self_attn.o_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.9.self_attn.q_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 2048, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 2048, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 2048, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| }, | |
| "model.layers.9.self_attn.v_proj": { | |
| "bias": false, | |
| "dtype": "float32", | |
| "hadU": 2048, | |
| "hadV": 512, | |
| "in_features": 2048, | |
| "linear": { | |
| "KV": [ | |
| 6, | |
| 7 | |
| ], | |
| "L": 16, | |
| "V": 2, | |
| "bias": false, | |
| "in_features": 2048, | |
| "in_part": [ | |
| 1024, | |
| 1024 | |
| ], | |
| "linear_cls": "CombtLinearTCQ", | |
| "linear_dtype": "float32", | |
| "out_features": 512, | |
| "td_x": 16, | |
| "td_y": 16, | |
| "tlut_bits": 9 | |
| }, | |
| "module_type": "IncoherentLinear", | |
| "out_features": 512, | |
| "rot_info": "skip_r", | |
| "scale": 32.0 | |
| } | |
| } | |
| }, | |
| "rms_norm_eps": 1e-05, | |
| "rope_scaling": { | |
| "factor": 32.0, | |
| "high_freq_factor": 4.0, | |
| "low_freq_factor": 1.0, | |
| "original_max_position_embeddings": 8192, | |
| "rope_type": "llama3" | |
| }, | |
| "rope_theta": 500000.0, | |
| "tie_word_embeddings": true, | |
| "torch_dtype": "float16", | |
| "transformers_version": "4.45.2", | |
| "use_cache": true, | |
| "vocab_size": 128256 | |
| } | |