| { |
| "model_a": "./base_model1", |
| "model_b": "./base_model4", |
| "output": "./merged_model_v6", |
| "total_keys": 404, |
| "alpha_config": { |
| "lm_head": 1.0, |
| "layernorm": 1.0, |
| "embedding": 1.0, |
| "other_frozen": 1.0, |
| "cp_mlp": 0.75, |
| "cp_attn": 0.62, |
| "talker_mlp": 0.35, |
| "talker_attn": 0.35 |
| }, |
| "layers": [ |
| { |
| "key": "talker.code_predictor.lm_head.0.weight", |
| "group": "lm_head", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.lm_head.1.weight", |
| "group": "lm_head", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.lm_head.10.weight", |
| "group": "lm_head", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.lm_head.11.weight", |
| "group": "lm_head", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.lm_head.12.weight", |
| "group": "lm_head", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.lm_head.13.weight", |
| "group": "lm_head", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.lm_head.14.weight", |
| "group": "lm_head", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.lm_head.2.weight", |
| "group": "lm_head", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.lm_head.3.weight", |
| "group": "lm_head", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.lm_head.4.weight", |
| "group": "lm_head", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.lm_head.5.weight", |
| "group": "lm_head", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.lm_head.6.weight", |
| "group": "lm_head", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.lm_head.7.weight", |
| "group": "lm_head", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.lm_head.8.weight", |
| "group": "lm_head", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.lm_head.9.weight", |
| "group": "lm_head", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.codec_embedding.0.weight", |
| "group": "other_frozen", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.codec_embedding.1.weight", |
| "group": "other_frozen", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.codec_embedding.10.weight", |
| "group": "other_frozen", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.codec_embedding.11.weight", |
| "group": "other_frozen", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.codec_embedding.12.weight", |
| "group": "other_frozen", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.codec_embedding.13.weight", |
| "group": "other_frozen", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.codec_embedding.14.weight", |
| "group": "other_frozen", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.codec_embedding.2.weight", |
| "group": "other_frozen", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.codec_embedding.3.weight", |
| "group": "other_frozen", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.codec_embedding.4.weight", |
| "group": "other_frozen", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.codec_embedding.5.weight", |
| "group": "other_frozen", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.codec_embedding.6.weight", |
| "group": "other_frozen", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.codec_embedding.7.weight", |
| "group": "other_frozen", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.codec_embedding.8.weight", |
| "group": "other_frozen", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.codec_embedding.9.weight", |
| "group": "other_frozen", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.0.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.0.mlp.down_proj.weight", |
| "group": "cp_mlp", |
| "alpha_a": 0.75, |
| "alpha_b": 0.25, |
| "shape": [ |
| 1024, |
| 3072 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.0.mlp.gate_proj.weight", |
| "group": "cp_mlp", |
| "alpha_a": 0.75, |
| "alpha_b": 0.25, |
| "shape": [ |
| 3072, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.0.mlp.up_proj.weight", |
| "group": "cp_mlp", |
| "alpha_a": 0.75, |
| "alpha_b": 0.25, |
| "shape": [ |
| 3072, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.0.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.0.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.0.self_attn.k_proj.weight", |
| "group": "cp_attn", |
| "alpha_a": 0.62, |
| "alpha_b": 0.38, |
| "shape": [ |
| 1024, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.0.self_attn.o_proj.weight", |
| "group": "cp_attn", |
| "alpha_a": 0.62, |
| "alpha_b": 0.38, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.0.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.0.self_attn.q_proj.weight", |
| "group": "cp_attn", |
| "alpha_a": 0.62, |
| "alpha_b": 0.38, |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.0.self_attn.v_proj.weight", |
| "group": "cp_attn", |
| "alpha_a": 0.62, |
| "alpha_b": 0.38, |
| "shape": [ |
| 1024, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.1.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.1.mlp.down_proj.weight", |
| "group": "cp_mlp", |
| "alpha_a": 0.75, |
| "alpha_b": 0.25, |
| "shape": [ |
| 1024, |
| 3072 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.1.mlp.gate_proj.weight", |
| "group": "cp_mlp", |
| "alpha_a": 0.75, |
| "alpha_b": 0.25, |
| "shape": [ |
| 3072, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.1.mlp.up_proj.weight", |
| "group": "cp_mlp", |
| "alpha_a": 0.75, |
| "alpha_b": 0.25, |
| "shape": [ |
| 3072, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.1.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.1.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.1.self_attn.k_proj.weight", |
| "group": "cp_attn", |
| "alpha_a": 0.62, |
| "alpha_b": 0.38, |
| "shape": [ |
| 1024, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.1.self_attn.o_proj.weight", |
| "group": "cp_attn", |
| "alpha_a": 0.62, |
| "alpha_b": 0.38, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.1.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.1.self_attn.q_proj.weight", |
| "group": "cp_attn", |
| "alpha_a": 0.62, |
| "alpha_b": 0.38, |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.1.self_attn.v_proj.weight", |
| "group": "cp_attn", |
| "alpha_a": 0.62, |
| "alpha_b": 0.38, |
| "shape": [ |
| 1024, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.2.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.2.mlp.down_proj.weight", |
| "group": "cp_mlp", |
| "alpha_a": 0.75, |
| "alpha_b": 0.25, |
| "shape": [ |
| 1024, |
| 3072 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.2.mlp.gate_proj.weight", |
| "group": "cp_mlp", |
| "alpha_a": 0.75, |
| "alpha_b": 0.25, |
| "shape": [ |
| 3072, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.2.mlp.up_proj.weight", |
| "group": "cp_mlp", |
| "alpha_a": 0.75, |
| "alpha_b": 0.25, |
| "shape": [ |
| 3072, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.2.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.2.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.2.self_attn.k_proj.weight", |
| "group": "cp_attn", |
| "alpha_a": 0.62, |
| "alpha_b": 0.38, |
| "shape": [ |
| 1024, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.2.self_attn.o_proj.weight", |
| "group": "cp_attn", |
| "alpha_a": 0.62, |
| "alpha_b": 0.38, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.2.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.2.self_attn.q_proj.weight", |
| "group": "cp_attn", |
| "alpha_a": 0.62, |
| "alpha_b": 0.38, |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.2.self_attn.v_proj.weight", |
| "group": "cp_attn", |
| "alpha_a": 0.62, |
| "alpha_b": 0.38, |
| "shape": [ |
| 1024, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.3.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.3.mlp.down_proj.weight", |
| "group": "cp_mlp", |
| "alpha_a": 0.75, |
| "alpha_b": 0.25, |
| "shape": [ |
| 1024, |
| 3072 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.3.mlp.gate_proj.weight", |
| "group": "cp_mlp", |
| "alpha_a": 0.75, |
| "alpha_b": 0.25, |
| "shape": [ |
| 3072, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.3.mlp.up_proj.weight", |
| "group": "cp_mlp", |
| "alpha_a": 0.75, |
| "alpha_b": 0.25, |
| "shape": [ |
| 3072, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.3.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.3.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.3.self_attn.k_proj.weight", |
| "group": "cp_attn", |
| "alpha_a": 0.62, |
| "alpha_b": 0.38, |
| "shape": [ |
| 1024, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.3.self_attn.o_proj.weight", |
| "group": "cp_attn", |
| "alpha_a": 0.62, |
| "alpha_b": 0.38, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.3.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.3.self_attn.q_proj.weight", |
| "group": "cp_attn", |
| "alpha_a": 0.62, |
| "alpha_b": 0.38, |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.3.self_attn.v_proj.weight", |
| "group": "cp_attn", |
| "alpha_a": 0.62, |
| "alpha_b": 0.38, |
| "shape": [ |
| 1024, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.4.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.4.mlp.down_proj.weight", |
| "group": "cp_mlp", |
| "alpha_a": 0.75, |
| "alpha_b": 0.25, |
| "shape": [ |
| 1024, |
| 3072 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.4.mlp.gate_proj.weight", |
| "group": "cp_mlp", |
| "alpha_a": 0.75, |
| "alpha_b": 0.25, |
| "shape": [ |
| 3072, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.4.mlp.up_proj.weight", |
| "group": "cp_mlp", |
| "alpha_a": 0.75, |
| "alpha_b": 0.25, |
| "shape": [ |
| 3072, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.4.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.4.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.4.self_attn.k_proj.weight", |
| "group": "cp_attn", |
| "alpha_a": 0.62, |
| "alpha_b": 0.38, |
| "shape": [ |
| 1024, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.4.self_attn.o_proj.weight", |
| "group": "cp_attn", |
| "alpha_a": 0.62, |
| "alpha_b": 0.38, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.4.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.4.self_attn.q_proj.weight", |
| "group": "cp_attn", |
| "alpha_a": 0.62, |
| "alpha_b": 0.38, |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.layers.4.self_attn.v_proj.weight", |
| "group": "cp_attn", |
| "alpha_a": 0.62, |
| "alpha_b": 0.38, |
| "shape": [ |
| 1024, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.model.norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.small_to_mtp_projection.bias", |
| "group": "other_frozen", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.code_predictor.small_to_mtp_projection.weight", |
| "group": "other_frozen", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.codec_head.weight", |
| "group": "other_frozen", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 3072, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.codec_embedding.weight", |
| "group": "embedding", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 3072, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.0.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.0.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.0.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.0.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.0.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.0.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.0.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.0.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.0.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.0.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.0.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.1.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.1.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.1.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.1.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.1.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.1.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.1.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.1.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.1.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.1.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.1.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.10.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.10.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.10.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.10.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.10.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.10.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.10.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.10.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.10.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.10.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.10.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.11.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.11.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.11.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.11.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.11.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.11.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.11.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.11.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.11.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.11.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.11.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.12.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.12.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.12.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.12.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.12.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.12.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.12.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.12.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.12.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.12.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.12.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.13.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.13.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.13.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.13.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.13.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.13.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.13.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.13.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.13.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.13.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.13.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.14.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.14.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.14.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.14.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.14.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.14.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.14.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.14.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.14.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.14.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.14.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.15.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.15.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.15.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.15.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.15.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.15.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.15.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.15.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.15.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.15.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.15.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.16.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.16.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.16.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.16.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.16.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.16.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.16.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.16.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.16.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.16.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.16.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.17.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.17.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.17.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.17.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.17.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.17.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.17.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.17.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.17.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.17.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.17.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.18.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.18.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.18.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.18.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.18.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.18.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.18.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.18.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.18.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.18.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.18.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.19.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.19.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.19.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.19.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.19.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.19.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.19.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.19.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.19.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.19.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.19.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.2.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.2.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.2.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.2.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.2.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.2.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.2.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.2.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.2.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.2.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.2.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.20.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.20.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.20.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.20.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.20.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.20.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.20.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.20.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.20.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.20.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.20.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.21.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.21.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.21.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.21.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.21.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.21.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.21.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.21.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.21.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.21.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.21.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.22.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.22.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.22.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.22.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.22.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.22.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.22.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.22.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.22.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.22.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.22.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.23.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.23.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.23.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.23.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.23.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.23.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.23.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.23.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.23.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.23.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.23.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.24.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.24.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.24.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.24.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.24.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.24.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.24.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.24.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.24.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.24.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.24.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.25.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.25.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.25.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.25.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.25.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.25.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.25.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.25.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.25.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.25.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.25.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.26.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.26.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.26.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.26.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.26.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.26.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.26.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.26.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.26.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.26.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.26.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.27.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.27.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.27.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.27.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.27.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.27.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.27.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.27.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.27.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.27.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.27.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.3.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.3.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.3.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.3.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.3.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.3.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.3.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.3.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.3.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.3.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.3.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.4.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.4.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.4.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.4.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.4.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.4.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.4.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.4.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.4.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.4.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.4.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.5.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.5.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.5.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.5.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.5.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.5.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.5.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.5.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.5.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.5.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.5.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.6.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.6.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.6.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.6.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.6.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.6.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.6.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.6.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.6.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.6.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.6.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.7.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.7.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.7.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.7.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.7.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.7.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.7.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.7.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.7.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.7.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.7.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.8.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.8.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.8.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.8.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.8.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.8.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.8.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.8.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.8.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.8.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.8.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.9.input_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.9.mlp.down_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 6144 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.9.mlp.gate_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.9.mlp.up_proj.weight", |
| "group": "talker_mlp", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.9.post_attention_layernorm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.9.self_attn.k_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.9.self_attn.k_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.9.self_attn.o_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.9.self_attn.q_norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.9.self_attn.q_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.layers.9.self_attn.v_proj.weight", |
| "group": "talker_attn", |
| "alpha_a": 0.35, |
| "alpha_b": 0.65, |
| "shape": [ |
| 1024, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.norm.weight", |
| "group": "layernorm", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.model.text_embedding.weight", |
| "group": "embedding", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 151936, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.text_projection.linear_fc1.bias", |
| "group": "other_frozen", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.text_projection.linear_fc1.weight", |
| "group": "other_frozen", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.text_projection.linear_fc2.bias", |
| "group": "other_frozen", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| { |
| "key": "talker.text_projection.linear_fc2.weight", |
| "group": "other_frozen", |
| "alpha_a": 1.0, |
| "alpha_b": 0.0, |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| } |
| ] |
| } |
