diff --git "a/tensor-cache.json" "b/tensor-cache.json" --- "a/tensor-cache.json" +++ "b/tensor-cache.json" @@ -1,5153 +1,5153 @@ { - "metadata": { - "ParamSize": 471, - "ParamBytes": 562628864.0, - "BitsPerParam": 4.5015443041247964 + "metadata": { + "ParamSize": 471, + "ParamBytes": 625381888, + "BitsPerParam": 5.003625757510392 + }, + "records": [ + { + "dataPath": "params_shard_0.bin", + "format": "raw-shard", + "nbytes": 150994944, + "records": [ + { + "name": "language_model.model.embed_tokens.q_weight", + "shape": [ + 262144, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 150994944, + "byteOffset": 0 + } + ], + "md5sum": "b42c6271a3061a65b31cf50acc2512cc" + }, + { + "dataPath": "params_shard_1.bin", + "format": "raw-shard", + "nbytes": 53620736, + "records": [ + { + "name": "language_model.model.embed_tokens.q_scale", + "shape": [ + 262144, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 37748736, + "byteOffset": 0 + }, + { + "name": "language_model.model.layers.0.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 37748736 + }, + { + "name": "language_model.model.layers.0.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 37753344 + }, + { + "name": "language_model.model.layers.0.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 41734656 + }, + { + "name": "language_model.model.layers.0.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 42729984 + }, + { + "name": "language_model.model.layers.0.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 50692608 + }, + { + "name": "language_model.model.layers.0.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 52683264 + }, + { + "name": "language_model.model.layers.0.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 52687872 + }, + { + "name": "language_model.model.layers.0.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 52692480 + }, + { + "name": "language_model.model.layers.0.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 52697088 + }, + { + "name": "language_model.model.layers.0.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 52698112 + }, + { + "name": "language_model.model.layers.0.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 52845568 + }, + { + "name": "language_model.model.layers.0.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 52882432 + }, + { + "name": "language_model.model.layers.0.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 53472256 + }, + { + "name": "language_model.model.layers.0.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 53619712 + } + ], + "md5sum": "f110625f9d2eec72ca0e02d9bf3ee399" + }, + { + "dataPath": "params_shard_2.bin", + "format": "raw-shard", + "nbytes": 34513408, + "records": [ + { + "name": "language_model.model.layers.0.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 0 + }, + { + "name": "language_model.model.layers.0.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 589824 + }, + { + "name": "language_model.model.layers.0.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 737280 + }, + { + "name": "language_model.model.layers.0.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 884736 + }, + { + "name": "language_model.model.layers.1.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 921600 + }, + { + "name": "language_model.model.layers.1.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 926208 + }, + { + "name": "language_model.model.layers.1.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 4907520 + }, + { + "name": "language_model.model.layers.1.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 5902848 + }, + { + "name": "language_model.model.layers.1.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 13865472 + }, + { + "name": "language_model.model.layers.1.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 15856128 + }, + { + "name": "language_model.model.layers.1.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 15860736 + }, + { + "name": "language_model.model.layers.1.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 15865344 + }, + { + "name": "language_model.model.layers.1.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 15869952 + }, + { + "name": "language_model.model.layers.1.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 15870976 + }, + { + "name": "language_model.model.layers.1.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 16018432 + }, + { + "name": "language_model.model.layers.1.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 16055296 + }, + { + "name": "language_model.model.layers.1.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16645120 + }, + { + "name": "language_model.model.layers.1.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 16792576 + }, + { + "name": "language_model.model.layers.1.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 16793600 + }, + { + "name": "language_model.model.layers.1.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 17383424 + }, + { + "name": "language_model.model.layers.1.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 17530880 + }, + { + "name": "language_model.model.layers.1.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 17678336 + }, + { + "name": "language_model.model.layers.10.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 17715200 + }, + { + "name": "language_model.model.layers.10.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 17719808 + }, + { + "name": "language_model.model.layers.10.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 21701120 + }, + { + "name": "language_model.model.layers.10.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 22696448 + }, + { + "name": "language_model.model.layers.10.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 30659072 + }, + { + "name": "language_model.model.layers.10.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 32649728 + }, + { + "name": "language_model.model.layers.10.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 32654336 + }, + { + "name": "language_model.model.layers.10.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 32658944 + }, + { + "name": "language_model.model.layers.10.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 32663552 + }, + { + "name": "language_model.model.layers.10.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 32664576 + }, + { + "name": "language_model.model.layers.10.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 32812032 + }, + { + "name": "language_model.model.layers.10.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 32848896 + }, + { + "name": "language_model.model.layers.10.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33438720 + }, + { + "name": "language_model.model.layers.10.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 33586176 + }, + { + "name": "language_model.model.layers.10.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 33587200 + }, + { + "name": "language_model.model.layers.10.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 34177024 + }, + { + "name": "language_model.model.layers.10.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 34324480 + }, + { + "name": "language_model.model.layers.10.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 34471936 + }, + { + "name": "language_model.model.layers.11.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 34508800 + } + ], + "md5sum": "106d022ee2801b44c63903c4faca905a" + }, + { + "dataPath": "params_shard_3.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "language_model.model.layers.11.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 0 + }, + { + "name": "language_model.model.layers.11.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 3981312 + }, + { + "name": "language_model.model.layers.11.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 4976640 + }, + { + "name": "language_model.model.layers.11.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 12939264 + }, + { + "name": "language_model.model.layers.11.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14929920 + }, + { + "name": "language_model.model.layers.11.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14934528 + }, + { + "name": "language_model.model.layers.11.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14939136 + }, + { + "name": "language_model.model.layers.11.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 14943744 + }, + { + "name": "language_model.model.layers.11.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 14944768 + }, + { + "name": "language_model.model.layers.11.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 15092224 + }, + { + "name": "language_model.model.layers.11.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15129088 + }, + { + "name": "language_model.model.layers.11.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 15718912 + }, + { + "name": "language_model.model.layers.11.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 15866368 + }, + { + "name": "language_model.model.layers.11.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15867392 + }, + { + "name": "language_model.model.layers.11.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16457216 + }, + { + "name": "language_model.model.layers.11.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16604672 + }, + { + "name": "language_model.model.layers.11.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 16752128 + }, + { + "name": "language_model.model.layers.12.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 16788992 + }, + { + "name": "language_model.model.layers.12.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 16793600 + }, + { + "name": "language_model.model.layers.12.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 20774912 + }, + { + "name": "language_model.model.layers.12.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 21770240 + }, + { + "name": "language_model.model.layers.12.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 29732864 + }, + { + "name": "language_model.model.layers.12.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31723520 + }, + { + "name": "language_model.model.layers.12.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31728128 + }, + { + "name": "language_model.model.layers.12.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31732736 + }, + { + "name": "language_model.model.layers.12.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 31737344 + }, + { + "name": "language_model.model.layers.12.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 31738368 + }, + { + "name": "language_model.model.layers.12.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 31885824 + }, + { + "name": "language_model.model.layers.12.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 31922688 + }, + { + "name": "language_model.model.layers.12.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 32512512 + }, + { + "name": "language_model.model.layers.12.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 32659968 + }, + { + "name": "language_model.model.layers.12.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 32660992 + }, + { + "name": "language_model.model.layers.12.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33250816 + }, + { + "name": "language_model.model.layers.12.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33398272 + }, + { + "name": "language_model.model.layers.12.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 33545728 + }, + { + "name": "language_model.model.layers.13.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 33582592 + } + ], + "md5sum": "7110c6726a8a068fde0776890d9f01d9" + }, + { + "dataPath": "params_shard_4.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "language_model.model.layers.13.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 0 + }, + { + "name": "language_model.model.layers.13.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 3981312 + }, + { + "name": "language_model.model.layers.13.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 4976640 + }, + { + "name": "language_model.model.layers.13.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 12939264 + }, + { + "name": "language_model.model.layers.13.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14929920 + }, + { + "name": "language_model.model.layers.13.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14934528 + }, + { + "name": "language_model.model.layers.13.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14939136 + }, + { + "name": "language_model.model.layers.13.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 14943744 + }, + { + "name": "language_model.model.layers.13.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 14944768 + }, + { + "name": "language_model.model.layers.13.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 15092224 + }, + { + "name": "language_model.model.layers.13.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15129088 + }, + { + "name": "language_model.model.layers.13.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 15718912 + }, + { + "name": "language_model.model.layers.13.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 15866368 + }, + { + "name": "language_model.model.layers.13.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15867392 + }, + { + "name": "language_model.model.layers.13.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16457216 + }, + { + "name": "language_model.model.layers.13.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16604672 + }, + { + "name": "language_model.model.layers.13.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 16752128 + }, + { + "name": "language_model.model.layers.14.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 16788992 + }, + { + "name": "language_model.model.layers.14.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 16793600 + }, + { + "name": "language_model.model.layers.14.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 20774912 + }, + { + "name": "language_model.model.layers.14.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 21770240 + }, + { + "name": "language_model.model.layers.14.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 29732864 + }, + { + "name": "language_model.model.layers.14.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31723520 + }, + { + "name": "language_model.model.layers.14.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31728128 + }, + { + "name": "language_model.model.layers.14.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31732736 + }, + { + "name": "language_model.model.layers.14.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 31737344 + }, + { + "name": "language_model.model.layers.14.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 31738368 + }, + { + "name": "language_model.model.layers.14.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 31885824 + }, + { + "name": "language_model.model.layers.14.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 31922688 + }, + { + "name": "language_model.model.layers.14.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 32512512 + }, + { + "name": "language_model.model.layers.14.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 32659968 + }, + { + "name": "language_model.model.layers.14.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 32660992 + }, + { + "name": "language_model.model.layers.14.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33250816 + }, + { + "name": "language_model.model.layers.14.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33398272 + }, + { + "name": "language_model.model.layers.14.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 33545728 + }, + { + "name": "language_model.model.layers.15.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 33582592 + } + ], + "md5sum": "ffddd1711c3fa7f8ffd106a15749b5aa" + }, + { + "dataPath": "params_shard_5.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "language_model.model.layers.15.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 0 + }, + { + "name": "language_model.model.layers.15.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 3981312 + }, + { + "name": "language_model.model.layers.15.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 4976640 + }, + { + "name": "language_model.model.layers.15.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 12939264 + }, + { + "name": "language_model.model.layers.15.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14929920 + }, + { + "name": "language_model.model.layers.15.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14934528 + }, + { + "name": "language_model.model.layers.15.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14939136 + }, + { + "name": "language_model.model.layers.15.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 14943744 + }, + { + "name": "language_model.model.layers.15.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 14944768 + }, + { + "name": "language_model.model.layers.15.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 15092224 + }, + { + "name": "language_model.model.layers.15.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15129088 + }, + { + "name": "language_model.model.layers.15.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 15718912 + }, + { + "name": "language_model.model.layers.15.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 15866368 + }, + { + "name": "language_model.model.layers.15.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15867392 + }, + { + "name": "language_model.model.layers.15.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16457216 + }, + { + "name": "language_model.model.layers.15.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16604672 + }, + { + "name": "language_model.model.layers.15.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 16752128 + }, + { + "name": "language_model.model.layers.16.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 16788992 + }, + { + "name": "language_model.model.layers.16.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 16793600 + }, + { + "name": "language_model.model.layers.16.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 20774912 + }, + { + "name": "language_model.model.layers.16.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 21770240 + }, + { + "name": "language_model.model.layers.16.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 29732864 + }, + { + "name": "language_model.model.layers.16.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31723520 + }, + { + "name": "language_model.model.layers.16.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31728128 + }, + { + "name": "language_model.model.layers.16.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31732736 + }, + { + "name": "language_model.model.layers.16.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 31737344 + }, + { + "name": "language_model.model.layers.16.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 31738368 + }, + { + "name": "language_model.model.layers.16.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 31885824 + }, + { + "name": "language_model.model.layers.16.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 31922688 + }, + { + "name": "language_model.model.layers.16.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 32512512 + }, + { + "name": "language_model.model.layers.16.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 32659968 + }, + { + "name": "language_model.model.layers.16.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 32660992 + }, + { + "name": "language_model.model.layers.16.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33250816 + }, + { + "name": "language_model.model.layers.16.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33398272 + }, + { + "name": "language_model.model.layers.16.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 33545728 + }, + { + "name": "language_model.model.layers.17.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 33582592 + } + ], + "md5sum": "612a32dd61b90d306199f3b42d584fdd" + }, + { + "dataPath": "params_shard_6.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "language_model.model.layers.17.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 0 + }, + { + "name": "language_model.model.layers.17.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 3981312 + }, + { + "name": "language_model.model.layers.17.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 4976640 + }, + { + "name": "language_model.model.layers.17.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 12939264 + }, + { + "name": "language_model.model.layers.17.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14929920 + }, + { + "name": "language_model.model.layers.17.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14934528 + }, + { + "name": "language_model.model.layers.17.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14939136 + }, + { + "name": "language_model.model.layers.17.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 14943744 + }, + { + "name": "language_model.model.layers.17.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 14944768 + }, + { + "name": "language_model.model.layers.17.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 15092224 + }, + { + "name": "language_model.model.layers.17.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15129088 + }, + { + "name": "language_model.model.layers.17.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 15718912 + }, + { + "name": "language_model.model.layers.17.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 15866368 + }, + { + "name": "language_model.model.layers.17.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15867392 + }, + { + "name": "language_model.model.layers.17.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16457216 + }, + { + "name": "language_model.model.layers.17.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16604672 + }, + { + "name": "language_model.model.layers.17.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 16752128 + }, + { + "name": "language_model.model.layers.18.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 16788992 + }, + { + "name": "language_model.model.layers.18.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 16793600 + }, + { + "name": "language_model.model.layers.18.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 20774912 + }, + { + "name": "language_model.model.layers.18.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 21770240 + }, + { + "name": "language_model.model.layers.18.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 29732864 + }, + { + "name": "language_model.model.layers.18.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31723520 + }, + { + "name": "language_model.model.layers.18.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31728128 + }, + { + "name": "language_model.model.layers.18.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31732736 + }, + { + "name": "language_model.model.layers.18.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 31737344 + }, + { + "name": "language_model.model.layers.18.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 31738368 + }, + { + "name": "language_model.model.layers.18.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 31885824 + }, + { + "name": "language_model.model.layers.18.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 31922688 + }, + { + "name": "language_model.model.layers.18.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 32512512 + }, + { + "name": "language_model.model.layers.18.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 32659968 + }, + { + "name": "language_model.model.layers.18.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 32660992 + }, + { + "name": "language_model.model.layers.18.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33250816 + }, + { + "name": "language_model.model.layers.18.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33398272 + }, + { + "name": "language_model.model.layers.18.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 33545728 + }, + { + "name": "language_model.model.layers.19.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 33582592 + } + ], + "md5sum": "c835bae133926437be61f1fab7c4d2e0" + }, + { + "dataPath": "params_shard_7.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "language_model.model.layers.19.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 0 + }, + { + "name": "language_model.model.layers.19.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 3981312 + }, + { + "name": "language_model.model.layers.19.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 4976640 + }, + { + "name": "language_model.model.layers.19.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 12939264 + }, + { + "name": "language_model.model.layers.19.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14929920 + }, + { + "name": "language_model.model.layers.19.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14934528 + }, + { + "name": "language_model.model.layers.19.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14939136 + }, + { + "name": "language_model.model.layers.19.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 14943744 + }, + { + "name": "language_model.model.layers.19.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 14944768 + }, + { + "name": "language_model.model.layers.19.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 15092224 + }, + { + "name": "language_model.model.layers.19.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15129088 + }, + { + "name": "language_model.model.layers.19.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 15718912 + }, + { + "name": "language_model.model.layers.19.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 15866368 + }, + { + "name": "language_model.model.layers.19.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15867392 + }, + { + "name": "language_model.model.layers.19.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16457216 + }, + { + "name": "language_model.model.layers.19.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16604672 + }, + { + "name": "language_model.model.layers.19.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 16752128 + }, + { + "name": "language_model.model.layers.2.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 16788992 + }, + { + "name": "language_model.model.layers.2.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 16793600 + }, + { + "name": "language_model.model.layers.2.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 20774912 + }, + { + "name": "language_model.model.layers.2.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 21770240 + }, + { + "name": "language_model.model.layers.2.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 29732864 + }, + { + "name": "language_model.model.layers.2.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31723520 + }, + { + "name": "language_model.model.layers.2.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31728128 + }, + { + "name": "language_model.model.layers.2.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31732736 + }, + { + "name": "language_model.model.layers.2.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 31737344 + }, + { + "name": "language_model.model.layers.2.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 31738368 + }, + { + "name": "language_model.model.layers.2.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 31885824 + }, + { + "name": "language_model.model.layers.2.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 31922688 + }, + { + "name": "language_model.model.layers.2.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 32512512 + }, + { + "name": "language_model.model.layers.2.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 32659968 + }, + { + "name": "language_model.model.layers.2.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 32660992 + }, + { + "name": "language_model.model.layers.2.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33250816 + }, + { + "name": "language_model.model.layers.2.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33398272 + }, + { + "name": "language_model.model.layers.2.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 33545728 + }, + { + "name": "language_model.model.layers.20.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 33582592 + } + ], + "md5sum": "0a2cddd3c2865940660ff9951793b308" + }, + { + "dataPath": "params_shard_8.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "language_model.model.layers.20.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 0 + }, + { + "name": "language_model.model.layers.20.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 3981312 + }, + { + "name": "language_model.model.layers.20.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 4976640 + }, + { + "name": "language_model.model.layers.20.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 12939264 + }, + { + "name": "language_model.model.layers.20.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14929920 + }, + { + "name": "language_model.model.layers.20.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14934528 + }, + { + "name": "language_model.model.layers.20.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14939136 + }, + { + "name": "language_model.model.layers.20.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 14943744 + }, + { + "name": "language_model.model.layers.20.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 14944768 + }, + { + "name": "language_model.model.layers.20.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 15092224 + }, + { + "name": "language_model.model.layers.20.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15129088 + }, + { + "name": "language_model.model.layers.20.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 15718912 + }, + { + "name": "language_model.model.layers.20.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 15866368 + }, + { + "name": "language_model.model.layers.20.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15867392 + }, + { + "name": "language_model.model.layers.20.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16457216 + }, + { + "name": "language_model.model.layers.20.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16604672 + }, + { + "name": "language_model.model.layers.20.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 16752128 + }, + { + "name": "language_model.model.layers.21.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 16788992 + }, + { + "name": "language_model.model.layers.21.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 16793600 + }, + { + "name": "language_model.model.layers.21.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 20774912 + }, + { + "name": "language_model.model.layers.21.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 21770240 + }, + { + "name": "language_model.model.layers.21.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 29732864 + }, + { + "name": "language_model.model.layers.21.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31723520 + }, + { + "name": "language_model.model.layers.21.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31728128 + }, + { + "name": "language_model.model.layers.21.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31732736 + }, + { + "name": "language_model.model.layers.21.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 31737344 + }, + { + "name": "language_model.model.layers.21.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 31738368 + }, + { + "name": "language_model.model.layers.21.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 31885824 + }, + { + "name": "language_model.model.layers.21.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 31922688 + }, + { + "name": "language_model.model.layers.21.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 32512512 + }, + { + "name": "language_model.model.layers.21.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 32659968 + }, + { + "name": "language_model.model.layers.21.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 32660992 + }, + { + "name": "language_model.model.layers.21.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33250816 + }, + { + "name": "language_model.model.layers.21.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33398272 + }, + { + "name": "language_model.model.layers.21.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 33545728 + }, + { + "name": "language_model.model.layers.22.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 33582592 + } + ], + "md5sum": "b0724f909cd2b9536140a28c856ada85" + }, + { + "dataPath": "params_shard_9.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "language_model.model.layers.22.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 0 + }, + { + "name": "language_model.model.layers.22.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 3981312 + }, + { + "name": "language_model.model.layers.22.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 4976640 + }, + { + "name": "language_model.model.layers.22.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 12939264 + }, + { + "name": "language_model.model.layers.22.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14929920 + }, + { + "name": "language_model.model.layers.22.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14934528 + }, + { + "name": "language_model.model.layers.22.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14939136 + }, + { + "name": "language_model.model.layers.22.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 14943744 + }, + { + "name": "language_model.model.layers.22.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 14944768 + }, + { + "name": "language_model.model.layers.22.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 15092224 + }, + { + "name": "language_model.model.layers.22.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15129088 + }, + { + "name": "language_model.model.layers.22.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 15718912 + }, + { + "name": "language_model.model.layers.22.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 15866368 + }, + { + "name": "language_model.model.layers.22.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15867392 + }, + { + "name": "language_model.model.layers.22.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16457216 + }, + { + "name": "language_model.model.layers.22.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16604672 + }, + { + "name": "language_model.model.layers.22.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 16752128 + }, + { + "name": "language_model.model.layers.23.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 16788992 + }, + { + "name": "language_model.model.layers.23.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 16793600 + }, + { + "name": "language_model.model.layers.23.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 20774912 + }, + { + "name": "language_model.model.layers.23.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 21770240 + }, + { + "name": "language_model.model.layers.23.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 29732864 + }, + { + "name": "language_model.model.layers.23.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31723520 + }, + { + "name": "language_model.model.layers.23.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31728128 + }, + { + "name": "language_model.model.layers.23.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31732736 + }, + { + "name": "language_model.model.layers.23.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 31737344 + }, + { + "name": "language_model.model.layers.23.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 31738368 + }, + { + "name": "language_model.model.layers.23.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 31885824 + }, + { + "name": "language_model.model.layers.23.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 31922688 + }, + { + "name": "language_model.model.layers.23.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 32512512 + }, + { + "name": "language_model.model.layers.23.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 32659968 + }, + { + "name": "language_model.model.layers.23.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 32660992 + }, + { + "name": "language_model.model.layers.23.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33250816 + }, + { + "name": "language_model.model.layers.23.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33398272 + }, + { + "name": "language_model.model.layers.23.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 33545728 + }, + { + "name": "language_model.model.layers.24.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 33582592 + } + ], + "md5sum": "d223ea2ad0407394aa7156ca6b22b1db" + }, + { + "dataPath": "params_shard_10.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "language_model.model.layers.24.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 0 + }, + { + "name": "language_model.model.layers.24.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 3981312 + }, + { + "name": "language_model.model.layers.24.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 4976640 + }, + { + "name": "language_model.model.layers.24.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 12939264 + }, + { + "name": "language_model.model.layers.24.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14929920 + }, + { + "name": "language_model.model.layers.24.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14934528 + }, + { + "name": "language_model.model.layers.24.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14939136 + }, + { + "name": "language_model.model.layers.24.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 14943744 + }, + { + "name": "language_model.model.layers.24.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 14944768 + }, + { + "name": "language_model.model.layers.24.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 15092224 + }, + { + "name": "language_model.model.layers.24.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15129088 + }, + { + "name": "language_model.model.layers.24.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 15718912 + }, + { + "name": "language_model.model.layers.24.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 15866368 + }, + { + "name": "language_model.model.layers.24.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15867392 + }, + { + "name": "language_model.model.layers.24.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16457216 + }, + { + "name": "language_model.model.layers.24.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16604672 + }, + { + "name": "language_model.model.layers.24.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 16752128 + }, + { + "name": "language_model.model.layers.25.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 16788992 + }, + { + "name": "language_model.model.layers.25.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 16793600 + }, + { + "name": "language_model.model.layers.25.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 20774912 + }, + { + "name": "language_model.model.layers.25.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 21770240 + }, + { + "name": "language_model.model.layers.25.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 29732864 + }, + { + "name": "language_model.model.layers.25.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31723520 + }, + { + "name": "language_model.model.layers.25.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31728128 + }, + { + "name": "language_model.model.layers.25.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31732736 + }, + { + "name": "language_model.model.layers.25.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 31737344 + }, + { + "name": "language_model.model.layers.25.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 31738368 + }, + { + "name": "language_model.model.layers.25.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 31885824 + }, + { + "name": "language_model.model.layers.25.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 31922688 + }, + { + "name": "language_model.model.layers.25.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 32512512 + }, + { + "name": "language_model.model.layers.25.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 32659968 + }, + { + "name": "language_model.model.layers.25.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 32660992 + }, + { + "name": "language_model.model.layers.25.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33250816 + }, + { + "name": "language_model.model.layers.25.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33398272 + }, + { + "name": "language_model.model.layers.25.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 33545728 + }, + { + "name": "language_model.model.layers.3.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 33582592 + } + ], + "md5sum": "670ae54e28351d282d502de1729726dc" + }, + { + "dataPath": "params_shard_11.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "language_model.model.layers.3.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 0 + }, + { + "name": "language_model.model.layers.3.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 3981312 + }, + { + "name": "language_model.model.layers.3.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 4976640 + }, + { + "name": "language_model.model.layers.3.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 12939264 + }, + { + "name": "language_model.model.layers.3.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14929920 + }, + { + "name": "language_model.model.layers.3.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14934528 + }, + { + "name": "language_model.model.layers.3.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14939136 + }, + { + "name": "language_model.model.layers.3.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 14943744 + }, + { + "name": "language_model.model.layers.3.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 14944768 + }, + { + "name": "language_model.model.layers.3.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 15092224 + }, + { + "name": "language_model.model.layers.3.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15129088 + }, + { + "name": "language_model.model.layers.3.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 15718912 + }, + { + "name": "language_model.model.layers.3.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 15866368 + }, + { + "name": "language_model.model.layers.3.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15867392 + }, + { + "name": "language_model.model.layers.3.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16457216 + }, + { + "name": "language_model.model.layers.3.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16604672 + }, + { + "name": "language_model.model.layers.3.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 16752128 + }, + { + "name": "language_model.model.layers.4.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 16788992 + }, + { + "name": "language_model.model.layers.4.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 16793600 + }, + { + "name": "language_model.model.layers.4.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 20774912 + }, + { + "name": "language_model.model.layers.4.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 21770240 + }, + { + "name": "language_model.model.layers.4.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 29732864 + }, + { + "name": "language_model.model.layers.4.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31723520 + }, + { + "name": "language_model.model.layers.4.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31728128 + }, + { + "name": "language_model.model.layers.4.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31732736 + }, + { + "name": "language_model.model.layers.4.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 31737344 + }, + { + "name": "language_model.model.layers.4.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 31738368 + }, + { + "name": "language_model.model.layers.4.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 31885824 + }, + { + "name": "language_model.model.layers.4.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 31922688 + }, + { + "name": "language_model.model.layers.4.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 32512512 + }, + { + "name": "language_model.model.layers.4.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 32659968 + }, + { + "name": "language_model.model.layers.4.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 32660992 + }, + { + "name": "language_model.model.layers.4.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33250816 + }, + { + "name": "language_model.model.layers.4.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33398272 + }, + { + "name": "language_model.model.layers.4.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 33545728 + }, + { + "name": "language_model.model.layers.5.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 33582592 + } + ], + "md5sum": "f9ea8bd1381ad2fa5997c62d31f3969a" + }, + { + "dataPath": "params_shard_12.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "language_model.model.layers.5.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 0 + }, + { + "name": "language_model.model.layers.5.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 3981312 + }, + { + "name": "language_model.model.layers.5.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 4976640 + }, + { + "name": "language_model.model.layers.5.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 12939264 + }, + { + "name": "language_model.model.layers.5.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14929920 + }, + { + "name": "language_model.model.layers.5.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14934528 + }, + { + "name": "language_model.model.layers.5.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14939136 + }, + { + "name": "language_model.model.layers.5.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 14943744 + }, + { + "name": "language_model.model.layers.5.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 14944768 + }, + { + "name": "language_model.model.layers.5.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 15092224 + }, + { + "name": "language_model.model.layers.5.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15129088 + }, + { + "name": "language_model.model.layers.5.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 15718912 + }, + { + "name": "language_model.model.layers.5.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 15866368 + }, + { + "name": "language_model.model.layers.5.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15867392 + }, + { + "name": "language_model.model.layers.5.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16457216 + }, + { + "name": "language_model.model.layers.5.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16604672 + }, + { + "name": "language_model.model.layers.5.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 16752128 + }, + { + "name": "language_model.model.layers.6.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 16788992 + }, + { + "name": "language_model.model.layers.6.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 16793600 + }, + { + "name": "language_model.model.layers.6.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 20774912 + }, + { + "name": "language_model.model.layers.6.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 21770240 + }, + { + "name": "language_model.model.layers.6.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 29732864 + }, + { + "name": "language_model.model.layers.6.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31723520 + }, + { + "name": "language_model.model.layers.6.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31728128 + }, + { + "name": "language_model.model.layers.6.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31732736 + }, + { + "name": "language_model.model.layers.6.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 31737344 + }, + { + "name": "language_model.model.layers.6.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 31738368 + }, + { + "name": "language_model.model.layers.6.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 31885824 + }, + { + "name": "language_model.model.layers.6.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 31922688 + }, + { + "name": "language_model.model.layers.6.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 32512512 + }, + { + "name": "language_model.model.layers.6.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 32659968 + }, + { + "name": "language_model.model.layers.6.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 32660992 + }, + { + "name": "language_model.model.layers.6.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33250816 + }, + { + "name": "language_model.model.layers.6.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33398272 + }, + { + "name": "language_model.model.layers.6.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 33545728 + }, + { + "name": "language_model.model.layers.7.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 33582592 + } + ], + "md5sum": "0b676e6ca1160b15e4a9177e6884589d" + }, + { + "dataPath": "params_shard_13.bin", + "format": "raw-shard", + "nbytes": 33587200, + "records": [ + { + "name": "language_model.model.layers.7.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 0 + }, + { + "name": "language_model.model.layers.7.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 3981312 + }, + { + "name": "language_model.model.layers.7.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 4976640 + }, + { + "name": "language_model.model.layers.7.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 12939264 + }, + { + "name": "language_model.model.layers.7.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14929920 + }, + { + "name": "language_model.model.layers.7.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14934528 + }, + { + "name": "language_model.model.layers.7.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14939136 + }, + { + "name": "language_model.model.layers.7.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 14943744 + }, + { + "name": "language_model.model.layers.7.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 14944768 + }, + { + "name": "language_model.model.layers.7.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 15092224 + }, + { + "name": "language_model.model.layers.7.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15129088 + }, + { + "name": "language_model.model.layers.7.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 15718912 + }, + { + "name": "language_model.model.layers.7.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 15866368 + }, + { + "name": "language_model.model.layers.7.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15867392 + }, + { + "name": "language_model.model.layers.7.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16457216 + }, + { + "name": "language_model.model.layers.7.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16604672 + }, + { + "name": "language_model.model.layers.7.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 16752128 + }, + { + "name": "language_model.model.layers.8.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 16788992 + }, + { + "name": "language_model.model.layers.8.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 16793600 + }, + { + "name": "language_model.model.layers.8.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 20774912 + }, + { + "name": "language_model.model.layers.8.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 21770240 + }, + { + "name": "language_model.model.layers.8.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 29732864 + }, + { + "name": "language_model.model.layers.8.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31723520 + }, + { + "name": "language_model.model.layers.8.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31728128 + }, + { + "name": "language_model.model.layers.8.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 31732736 + }, + { + "name": "language_model.model.layers.8.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 31737344 + }, + { + "name": "language_model.model.layers.8.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 31738368 + }, + { + "name": "language_model.model.layers.8.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 31885824 + }, + { + "name": "language_model.model.layers.8.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 31922688 + }, + { + "name": "language_model.model.layers.8.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 32512512 + }, + { + "name": "language_model.model.layers.8.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 32659968 + }, + { + "name": "language_model.model.layers.8.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 32660992 + }, + { + "name": "language_model.model.layers.8.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33250816 + }, + { + "name": "language_model.model.layers.8.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 33398272 + }, + { + "name": "language_model.model.layers.8.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 33545728 + }, + { + "name": "language_model.model.layers.9.input_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 33582592 + } + ], + "md5sum": "97277aec8da312d218f134655974f23d" }, - "records": [ - { - "dataPath": "params_shard_0.bin", - "format": "raw-shard", - "nbytes": 150994944, - "records": [ - { - "name": "language_model.model.embed_tokens.q_weight", - "shape": [ - 262144, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 150994944, - "byteOffset": 0 - } - ], - "md5sum": "b42c6271a3061a65b31cf50acc2512cc" - }, - { - "dataPath": "params_shard_1.bin", - "format": "raw-shard", - "nbytes": 33150976, - "records": [ - { - "name": "language_model.model.embed_tokens.q_scale", - "shape": [ - 262144, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18874368, - "byteOffset": 0 - }, - { - "name": "language_model.model.layers.0.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 18874368 - }, - { - "name": "language_model.model.layers.0.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 18876672 - }, - { - "name": "language_model.model.layers.0.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 22857984 - }, - { - "name": "language_model.model.layers.0.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 23355648 - }, - { - "name": "language_model.model.layers.0.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 31318272 - }, - { - "name": "language_model.model.layers.0.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 32313600 - }, - { - "name": "language_model.model.layers.0.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 32315904 - }, - { - "name": "language_model.model.layers.0.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 32318208 - }, - { - "name": "language_model.model.layers.0.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 32320512 - }, - { - "name": "language_model.model.layers.0.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 32321024 - }, - { - "name": "language_model.model.layers.0.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 32468480 - }, - { - "name": "language_model.model.layers.0.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 32486912 - }, - { - "name": "language_model.model.layers.0.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 33076736 - }, - { - "name": "language_model.model.layers.0.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 33150464 - } - ], - "md5sum": "627e5bd8d43434b9dc5a63461640e6a4" - }, - { - "dataPath": "params_shard_2.bin", - "format": "raw-shard", - "nbytes": 31043840, - "records": [ - { - "name": "language_model.model.layers.0.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 0 - }, - { - "name": "language_model.model.layers.0.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 589824 - }, - { - "name": "language_model.model.layers.0.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 663552 - }, - { - "name": "language_model.model.layers.0.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 811008 - }, - { - "name": "language_model.model.layers.1.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 829440 - }, - { - "name": "language_model.model.layers.1.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 831744 - }, - { - "name": "language_model.model.layers.1.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 4813056 - }, - { - "name": "language_model.model.layers.1.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 5310720 - }, - { - "name": "language_model.model.layers.1.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 13273344 - }, - { - "name": "language_model.model.layers.1.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 14268672 - }, - { - "name": "language_model.model.layers.1.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 14270976 - }, - { - "name": "language_model.model.layers.1.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 14273280 - }, - { - "name": "language_model.model.layers.1.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 14275584 - }, - { - "name": "language_model.model.layers.1.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 14276096 - }, - { - "name": "language_model.model.layers.1.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 14423552 - }, - { - "name": "language_model.model.layers.1.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 14441984 - }, - { - "name": "language_model.model.layers.1.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 15031808 - }, - { - "name": "language_model.model.layers.1.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 15105536 - }, - { - "name": "language_model.model.layers.1.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 15106048 - }, - { - "name": "language_model.model.layers.1.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 15695872 - }, - { - "name": "language_model.model.layers.1.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 15769600 - }, - { - "name": "language_model.model.layers.1.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 15917056 - }, - { - "name": "language_model.model.layers.10.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 15935488 - }, - { - "name": "language_model.model.layers.10.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 15937792 - }, - { - "name": "language_model.model.layers.10.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 19919104 - }, - { - "name": "language_model.model.layers.10.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 20416768 - }, - { - "name": "language_model.model.layers.10.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 28379392 - }, - { - "name": "language_model.model.layers.10.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 29374720 - }, - { - "name": "language_model.model.layers.10.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 29377024 - }, - { - "name": "language_model.model.layers.10.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 29379328 - }, - { - "name": "language_model.model.layers.10.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 29381632 - }, - { - "name": "language_model.model.layers.10.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 29382144 - }, - { - "name": "language_model.model.layers.10.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 29529600 - }, - { - "name": "language_model.model.layers.10.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 29548032 - }, - { - "name": "language_model.model.layers.10.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 30137856 - }, - { - "name": "language_model.model.layers.10.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 30211584 - }, - { - "name": "language_model.model.layers.10.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 30212096 - }, - { - "name": "language_model.model.layers.10.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 30801920 - }, - { - "name": "language_model.model.layers.10.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 30875648 - }, - { - "name": "language_model.model.layers.10.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 31023104 - }, - { - "name": "language_model.model.layers.11.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 31041536 - } - ], - "md5sum": "7a3908b48c7d395479b71fe1b0956fd6" - }, - { - "dataPath": "params_shard_3.bin", - "format": "raw-shard", - "nbytes": 30212096, - "records": [ - { - "name": "language_model.model.layers.11.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 0 - }, - { - "name": "language_model.model.layers.11.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 3981312 - }, - { - "name": "language_model.model.layers.11.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 4478976 - }, - { - "name": "language_model.model.layers.11.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 12441600 - }, - { - "name": "language_model.model.layers.11.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13436928 - }, - { - "name": "language_model.model.layers.11.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13439232 - }, - { - "name": "language_model.model.layers.11.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13441536 - }, - { - "name": "language_model.model.layers.11.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 13443840 - }, - { - "name": "language_model.model.layers.11.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 13444352 - }, - { - "name": "language_model.model.layers.11.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 13591808 - }, - { - "name": "language_model.model.layers.11.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 13610240 - }, - { - "name": "language_model.model.layers.11.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14200064 - }, - { - "name": "language_model.model.layers.11.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 14273792 - }, - { - "name": "language_model.model.layers.11.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 14274304 - }, - { - "name": "language_model.model.layers.11.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14864128 - }, - { - "name": "language_model.model.layers.11.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 14937856 - }, - { - "name": "language_model.model.layers.11.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 15085312 - }, - { - "name": "language_model.model.layers.12.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 15103744 - }, - { - "name": "language_model.model.layers.12.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 15106048 - }, - { - "name": "language_model.model.layers.12.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 19087360 - }, - { - "name": "language_model.model.layers.12.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 19585024 - }, - { - "name": "language_model.model.layers.12.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 27547648 - }, - { - "name": "language_model.model.layers.12.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28542976 - }, - { - "name": "language_model.model.layers.12.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28545280 - }, - { - "name": "language_model.model.layers.12.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28547584 - }, - { - "name": "language_model.model.layers.12.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 28549888 - }, - { - "name": "language_model.model.layers.12.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 28550400 - }, - { - "name": "language_model.model.layers.12.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 28697856 - }, - { - "name": "language_model.model.layers.12.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 28716288 - }, - { - "name": "language_model.model.layers.12.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 29306112 - }, - { - "name": "language_model.model.layers.12.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 29379840 - }, - { - "name": "language_model.model.layers.12.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 29380352 - }, - { - "name": "language_model.model.layers.12.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 29970176 - }, - { - "name": "language_model.model.layers.12.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 30043904 - }, - { - "name": "language_model.model.layers.12.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 30191360 - }, - { - "name": "language_model.model.layers.13.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 30209792 - } - ], - "md5sum": "f783ebb1aa2abe2948a1e8b327f2b95b" - }, - { - "dataPath": "params_shard_4.bin", - "format": "raw-shard", - "nbytes": 30212096, - "records": [ - { - "name": "language_model.model.layers.13.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 0 - }, - { - "name": "language_model.model.layers.13.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 3981312 - }, - { - "name": "language_model.model.layers.13.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 4478976 - }, - { - "name": "language_model.model.layers.13.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 12441600 - }, - { - "name": "language_model.model.layers.13.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13436928 - }, - { - "name": "language_model.model.layers.13.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13439232 - }, - { - "name": "language_model.model.layers.13.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13441536 - }, - { - "name": "language_model.model.layers.13.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 13443840 - }, - { - "name": "language_model.model.layers.13.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 13444352 - }, - { - "name": "language_model.model.layers.13.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 13591808 - }, - { - "name": "language_model.model.layers.13.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 13610240 - }, - { - "name": "language_model.model.layers.13.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14200064 - }, - { - "name": "language_model.model.layers.13.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 14273792 - }, - { - "name": "language_model.model.layers.13.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 14274304 - }, - { - "name": "language_model.model.layers.13.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14864128 - }, - { - "name": "language_model.model.layers.13.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 14937856 - }, - { - "name": "language_model.model.layers.13.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 15085312 - }, - { - "name": "language_model.model.layers.14.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 15103744 - }, - { - "name": "language_model.model.layers.14.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 15106048 - }, - { - "name": "language_model.model.layers.14.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 19087360 - }, - { - "name": "language_model.model.layers.14.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 19585024 - }, - { - "name": "language_model.model.layers.14.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 27547648 - }, - { - "name": "language_model.model.layers.14.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28542976 - }, - { - "name": "language_model.model.layers.14.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28545280 - }, - { - "name": "language_model.model.layers.14.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28547584 - }, - { - "name": "language_model.model.layers.14.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 28549888 - }, - { - "name": "language_model.model.layers.14.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 28550400 - }, - { - "name": "language_model.model.layers.14.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 28697856 - }, - { - "name": "language_model.model.layers.14.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 28716288 - }, - { - "name": "language_model.model.layers.14.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 29306112 - }, - { - "name": "language_model.model.layers.14.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 29379840 - }, - { - "name": "language_model.model.layers.14.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 29380352 - }, - { - "name": "language_model.model.layers.14.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 29970176 - }, - { - "name": "language_model.model.layers.14.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 30043904 - }, - { - "name": "language_model.model.layers.14.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 30191360 - }, - { - "name": "language_model.model.layers.15.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 30209792 - } - ], - "md5sum": "49fc0f8bcb5d2d9797ab1a8f17678c2b" - }, - { - "dataPath": "params_shard_5.bin", - "format": "raw-shard", - "nbytes": 30212096, - "records": [ - { - "name": "language_model.model.layers.15.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 0 - }, - { - "name": "language_model.model.layers.15.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 3981312 - }, - { - "name": "language_model.model.layers.15.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 4478976 - }, - { - "name": "language_model.model.layers.15.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 12441600 - }, - { - "name": "language_model.model.layers.15.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13436928 - }, - { - "name": "language_model.model.layers.15.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13439232 - }, - { - "name": "language_model.model.layers.15.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13441536 - }, - { - "name": "language_model.model.layers.15.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 13443840 - }, - { - "name": "language_model.model.layers.15.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 13444352 - }, - { - "name": "language_model.model.layers.15.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 13591808 - }, - { - "name": "language_model.model.layers.15.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 13610240 - }, - { - "name": "language_model.model.layers.15.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14200064 - }, - { - "name": "language_model.model.layers.15.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 14273792 - }, - { - "name": "language_model.model.layers.15.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 14274304 - }, - { - "name": "language_model.model.layers.15.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14864128 - }, - { - "name": "language_model.model.layers.15.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 14937856 - }, - { - "name": "language_model.model.layers.15.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 15085312 - }, - { - "name": "language_model.model.layers.16.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 15103744 - }, - { - "name": "language_model.model.layers.16.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 15106048 - }, - { - "name": "language_model.model.layers.16.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 19087360 - }, - { - "name": "language_model.model.layers.16.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 19585024 - }, - { - "name": "language_model.model.layers.16.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 27547648 - }, - { - "name": "language_model.model.layers.16.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28542976 - }, - { - "name": "language_model.model.layers.16.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28545280 - }, - { - "name": "language_model.model.layers.16.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28547584 - }, - { - "name": "language_model.model.layers.16.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 28549888 - }, - { - "name": "language_model.model.layers.16.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 28550400 - }, - { - "name": "language_model.model.layers.16.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 28697856 - }, - { - "name": "language_model.model.layers.16.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 28716288 - }, - { - "name": "language_model.model.layers.16.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 29306112 - }, - { - "name": "language_model.model.layers.16.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 29379840 - }, - { - "name": "language_model.model.layers.16.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 29380352 - }, - { - "name": "language_model.model.layers.16.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 29970176 - }, - { - "name": "language_model.model.layers.16.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 30043904 - }, - { - "name": "language_model.model.layers.16.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 30191360 - }, - { - "name": "language_model.model.layers.17.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 30209792 - } - ], - "md5sum": "cdea0ad27111986aaa5ad650a090647d" - }, - { - "dataPath": "params_shard_6.bin", - "format": "raw-shard", - "nbytes": 30212096, - "records": [ - { - "name": "language_model.model.layers.17.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 0 - }, - { - "name": "language_model.model.layers.17.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 3981312 - }, - { - "name": "language_model.model.layers.17.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 4478976 - }, - { - "name": "language_model.model.layers.17.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 12441600 - }, - { - "name": "language_model.model.layers.17.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13436928 - }, - { - "name": "language_model.model.layers.17.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13439232 - }, - { - "name": "language_model.model.layers.17.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13441536 - }, - { - "name": "language_model.model.layers.17.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 13443840 - }, - { - "name": "language_model.model.layers.17.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 13444352 - }, - { - "name": "language_model.model.layers.17.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 13591808 - }, - { - "name": "language_model.model.layers.17.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 13610240 - }, - { - "name": "language_model.model.layers.17.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14200064 - }, - { - "name": "language_model.model.layers.17.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 14273792 - }, - { - "name": "language_model.model.layers.17.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 14274304 - }, - { - "name": "language_model.model.layers.17.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14864128 - }, - { - "name": "language_model.model.layers.17.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 14937856 - }, - { - "name": "language_model.model.layers.17.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 15085312 - }, - { - "name": "language_model.model.layers.18.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 15103744 - }, - { - "name": "language_model.model.layers.18.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 15106048 - }, - { - "name": "language_model.model.layers.18.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 19087360 - }, - { - "name": "language_model.model.layers.18.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 19585024 - }, - { - "name": "language_model.model.layers.18.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 27547648 - }, - { - "name": "language_model.model.layers.18.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28542976 - }, - { - "name": "language_model.model.layers.18.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28545280 - }, - { - "name": "language_model.model.layers.18.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28547584 - }, - { - "name": "language_model.model.layers.18.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 28549888 - }, - { - "name": "language_model.model.layers.18.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 28550400 - }, - { - "name": "language_model.model.layers.18.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 28697856 - }, - { - "name": "language_model.model.layers.18.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 28716288 - }, - { - "name": "language_model.model.layers.18.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 29306112 - }, - { - "name": "language_model.model.layers.18.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 29379840 - }, - { - "name": "language_model.model.layers.18.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 29380352 - }, - { - "name": "language_model.model.layers.18.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 29970176 - }, - { - "name": "language_model.model.layers.18.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 30043904 - }, - { - "name": "language_model.model.layers.18.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 30191360 - }, - { - "name": "language_model.model.layers.19.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 30209792 - } - ], - "md5sum": "6e104e78ee962d1129aecb7ef9c6e9a9" - }, - { - "dataPath": "params_shard_7.bin", - "format": "raw-shard", - "nbytes": 30212096, - "records": [ - { - "name": "language_model.model.layers.19.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 0 - }, - { - "name": "language_model.model.layers.19.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 3981312 - }, - { - "name": "language_model.model.layers.19.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 4478976 - }, - { - "name": "language_model.model.layers.19.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 12441600 - }, - { - "name": "language_model.model.layers.19.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13436928 - }, - { - "name": "language_model.model.layers.19.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13439232 - }, - { - "name": "language_model.model.layers.19.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13441536 - }, - { - "name": "language_model.model.layers.19.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 13443840 - }, - { - "name": "language_model.model.layers.19.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 13444352 - }, - { - "name": "language_model.model.layers.19.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 13591808 - }, - { - "name": "language_model.model.layers.19.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 13610240 - }, - { - "name": "language_model.model.layers.19.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14200064 - }, - { - "name": "language_model.model.layers.19.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 14273792 - }, - { - "name": "language_model.model.layers.19.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 14274304 - }, - { - "name": "language_model.model.layers.19.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14864128 - }, - { - "name": "language_model.model.layers.19.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 14937856 - }, - { - "name": "language_model.model.layers.19.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 15085312 - }, - { - "name": "language_model.model.layers.2.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 15103744 - }, - { - "name": "language_model.model.layers.2.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 15106048 - }, - { - "name": "language_model.model.layers.2.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 19087360 - }, - { - "name": "language_model.model.layers.2.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 19585024 - }, - { - "name": "language_model.model.layers.2.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 27547648 - }, - { - "name": "language_model.model.layers.2.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28542976 - }, - { - "name": "language_model.model.layers.2.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28545280 - }, - { - "name": "language_model.model.layers.2.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28547584 - }, - { - "name": "language_model.model.layers.2.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 28549888 - }, - { - "name": "language_model.model.layers.2.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 28550400 - }, - { - "name": "language_model.model.layers.2.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 28697856 - }, - { - "name": "language_model.model.layers.2.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 28716288 - }, - { - "name": "language_model.model.layers.2.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 29306112 - }, - { - "name": "language_model.model.layers.2.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 29379840 - }, - { - "name": "language_model.model.layers.2.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 29380352 - }, - { - "name": "language_model.model.layers.2.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 29970176 - }, - { - "name": "language_model.model.layers.2.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 30043904 - }, - { - "name": "language_model.model.layers.2.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 30191360 - }, - { - "name": "language_model.model.layers.20.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 30209792 - } - ], - "md5sum": "ef9aa948e2584d367a1b9a2f22c0a906" - }, - { - "dataPath": "params_shard_8.bin", - "format": "raw-shard", - "nbytes": 30212096, - "records": [ - { - "name": "language_model.model.layers.20.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 0 - }, - { - "name": "language_model.model.layers.20.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 3981312 - }, - { - "name": "language_model.model.layers.20.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 4478976 - }, - { - "name": "language_model.model.layers.20.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 12441600 - }, - { - "name": "language_model.model.layers.20.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13436928 - }, - { - "name": "language_model.model.layers.20.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13439232 - }, - { - "name": "language_model.model.layers.20.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13441536 - }, - { - "name": "language_model.model.layers.20.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 13443840 - }, - { - "name": "language_model.model.layers.20.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 13444352 - }, - { - "name": "language_model.model.layers.20.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 13591808 - }, - { - "name": "language_model.model.layers.20.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 13610240 - }, - { - "name": "language_model.model.layers.20.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14200064 - }, - { - "name": "language_model.model.layers.20.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 14273792 - }, - { - "name": "language_model.model.layers.20.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 14274304 - }, - { - "name": "language_model.model.layers.20.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14864128 - }, - { - "name": "language_model.model.layers.20.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 14937856 - }, - { - "name": "language_model.model.layers.20.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 15085312 - }, - { - "name": "language_model.model.layers.21.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 15103744 - }, - { - "name": "language_model.model.layers.21.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 15106048 - }, - { - "name": "language_model.model.layers.21.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 19087360 - }, - { - "name": "language_model.model.layers.21.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 19585024 - }, - { - "name": "language_model.model.layers.21.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 27547648 - }, - { - "name": "language_model.model.layers.21.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28542976 - }, - { - "name": "language_model.model.layers.21.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28545280 - }, - { - "name": "language_model.model.layers.21.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28547584 - }, - { - "name": "language_model.model.layers.21.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 28549888 - }, - { - "name": "language_model.model.layers.21.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 28550400 - }, - { - "name": "language_model.model.layers.21.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 28697856 - }, - { - "name": "language_model.model.layers.21.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 28716288 - }, - { - "name": "language_model.model.layers.21.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 29306112 - }, - { - "name": "language_model.model.layers.21.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 29379840 - }, - { - "name": "language_model.model.layers.21.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 29380352 - }, - { - "name": "language_model.model.layers.21.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 29970176 - }, - { - "name": "language_model.model.layers.21.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 30043904 - }, - { - "name": "language_model.model.layers.21.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 30191360 - }, - { - "name": "language_model.model.layers.22.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 30209792 - } - ], - "md5sum": "65affc90a51f223504f0814739a9ffb2" - }, - { - "dataPath": "params_shard_9.bin", - "format": "raw-shard", - "nbytes": 30212096, - "records": [ - { - "name": "language_model.model.layers.22.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 0 - }, - { - "name": "language_model.model.layers.22.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 3981312 - }, - { - "name": "language_model.model.layers.22.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 4478976 - }, - { - "name": "language_model.model.layers.22.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 12441600 - }, - { - "name": "language_model.model.layers.22.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13436928 - }, - { - "name": "language_model.model.layers.22.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13439232 - }, - { - "name": "language_model.model.layers.22.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13441536 - }, - { - "name": "language_model.model.layers.22.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 13443840 - }, - { - "name": "language_model.model.layers.22.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 13444352 - }, - { - "name": "language_model.model.layers.22.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 13591808 - }, - { - "name": "language_model.model.layers.22.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 13610240 - }, - { - "name": "language_model.model.layers.22.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14200064 - }, - { - "name": "language_model.model.layers.22.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 14273792 - }, - { - "name": "language_model.model.layers.22.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 14274304 - }, - { - "name": "language_model.model.layers.22.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14864128 - }, - { - "name": "language_model.model.layers.22.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 14937856 - }, - { - "name": "language_model.model.layers.22.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 15085312 - }, - { - "name": "language_model.model.layers.23.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 15103744 - }, - { - "name": "language_model.model.layers.23.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 15106048 - }, - { - "name": "language_model.model.layers.23.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 19087360 - }, - { - "name": "language_model.model.layers.23.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 19585024 - }, - { - "name": "language_model.model.layers.23.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 27547648 - }, - { - "name": "language_model.model.layers.23.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28542976 - }, - { - "name": "language_model.model.layers.23.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28545280 - }, - { - "name": "language_model.model.layers.23.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28547584 - }, - { - "name": "language_model.model.layers.23.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 28549888 - }, - { - "name": "language_model.model.layers.23.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 28550400 - }, - { - "name": "language_model.model.layers.23.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 28697856 - }, - { - "name": "language_model.model.layers.23.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 28716288 - }, - { - "name": "language_model.model.layers.23.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 29306112 - }, - { - "name": "language_model.model.layers.23.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 29379840 - }, - { - "name": "language_model.model.layers.23.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 29380352 - }, - { - "name": "language_model.model.layers.23.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 29970176 - }, - { - "name": "language_model.model.layers.23.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 30043904 - }, - { - "name": "language_model.model.layers.23.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 30191360 - }, - { - "name": "language_model.model.layers.24.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 30209792 - } - ], - "md5sum": "4cf4d2ad29f6359863ccc49a12998f44" - }, - { - "dataPath": "params_shard_10.bin", - "format": "raw-shard", - "nbytes": 30212096, - "records": [ - { - "name": "language_model.model.layers.24.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 0 - }, - { - "name": "language_model.model.layers.24.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 3981312 - }, - { - "name": "language_model.model.layers.24.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 4478976 - }, - { - "name": "language_model.model.layers.24.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 12441600 - }, - { - "name": "language_model.model.layers.24.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13436928 - }, - { - "name": "language_model.model.layers.24.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13439232 - }, - { - "name": "language_model.model.layers.24.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13441536 - }, - { - "name": "language_model.model.layers.24.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 13443840 - }, - { - "name": "language_model.model.layers.24.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 13444352 - }, - { - "name": "language_model.model.layers.24.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 13591808 - }, - { - "name": "language_model.model.layers.24.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 13610240 - }, - { - "name": "language_model.model.layers.24.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14200064 - }, - { - "name": "language_model.model.layers.24.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 14273792 - }, - { - "name": "language_model.model.layers.24.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 14274304 - }, - { - "name": "language_model.model.layers.24.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14864128 - }, - { - "name": "language_model.model.layers.24.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 14937856 - }, - { - "name": "language_model.model.layers.24.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 15085312 - }, - { - "name": "language_model.model.layers.25.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 15103744 - }, - { - "name": "language_model.model.layers.25.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 15106048 - }, - { - "name": "language_model.model.layers.25.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 19087360 - }, - { - "name": "language_model.model.layers.25.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 19585024 - }, - { - "name": "language_model.model.layers.25.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 27547648 - }, - { - "name": "language_model.model.layers.25.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28542976 - }, - { - "name": "language_model.model.layers.25.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28545280 - }, - { - "name": "language_model.model.layers.25.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28547584 - }, - { - "name": "language_model.model.layers.25.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 28549888 - }, - { - "name": "language_model.model.layers.25.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 28550400 - }, - { - "name": "language_model.model.layers.25.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 28697856 - }, - { - "name": "language_model.model.layers.25.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 28716288 - }, - { - "name": "language_model.model.layers.25.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 29306112 - }, - { - "name": "language_model.model.layers.25.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 29379840 - }, - { - "name": "language_model.model.layers.25.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 29380352 - }, - { - "name": "language_model.model.layers.25.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 29970176 - }, - { - "name": "language_model.model.layers.25.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 30043904 - }, - { - "name": "language_model.model.layers.25.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 30191360 - }, - { - "name": "language_model.model.layers.3.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 30209792 - } - ], - "md5sum": "260789f8c9c953247e02bd2abd3cd6db" - }, - { - "dataPath": "params_shard_11.bin", - "format": "raw-shard", - "nbytes": 30212096, - "records": [ - { - "name": "language_model.model.layers.3.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 0 - }, - { - "name": "language_model.model.layers.3.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 3981312 - }, - { - "name": "language_model.model.layers.3.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 4478976 - }, - { - "name": "language_model.model.layers.3.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 12441600 - }, - { - "name": "language_model.model.layers.3.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13436928 - }, - { - "name": "language_model.model.layers.3.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13439232 - }, - { - "name": "language_model.model.layers.3.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13441536 - }, - { - "name": "language_model.model.layers.3.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 13443840 - }, - { - "name": "language_model.model.layers.3.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 13444352 - }, - { - "name": "language_model.model.layers.3.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 13591808 - }, - { - "name": "language_model.model.layers.3.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 13610240 - }, - { - "name": "language_model.model.layers.3.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14200064 - }, - { - "name": "language_model.model.layers.3.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 14273792 - }, - { - "name": "language_model.model.layers.3.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 14274304 - }, - { - "name": "language_model.model.layers.3.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14864128 - }, - { - "name": "language_model.model.layers.3.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 14937856 - }, - { - "name": "language_model.model.layers.3.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 15085312 - }, - { - "name": "language_model.model.layers.4.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 15103744 - }, - { - "name": "language_model.model.layers.4.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 15106048 - }, - { - "name": "language_model.model.layers.4.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 19087360 - }, - { - "name": "language_model.model.layers.4.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 19585024 - }, - { - "name": "language_model.model.layers.4.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 27547648 - }, - { - "name": "language_model.model.layers.4.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28542976 - }, - { - "name": "language_model.model.layers.4.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28545280 - }, - { - "name": "language_model.model.layers.4.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28547584 - }, - { - "name": "language_model.model.layers.4.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 28549888 - }, - { - "name": "language_model.model.layers.4.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 28550400 - }, - { - "name": "language_model.model.layers.4.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 28697856 - }, - { - "name": "language_model.model.layers.4.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 28716288 - }, - { - "name": "language_model.model.layers.4.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 29306112 - }, - { - "name": "language_model.model.layers.4.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 29379840 - }, - { - "name": "language_model.model.layers.4.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 29380352 - }, - { - "name": "language_model.model.layers.4.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 29970176 - }, - { - "name": "language_model.model.layers.4.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 30043904 - }, - { - "name": "language_model.model.layers.4.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 30191360 - }, - { - "name": "language_model.model.layers.5.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 30209792 - } - ], - "md5sum": "c0e98b13ed191e46d62643bb712a3f65" - }, - { - "dataPath": "params_shard_12.bin", - "format": "raw-shard", - "nbytes": 30212096, - "records": [ - { - "name": "language_model.model.layers.5.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 0 - }, - { - "name": "language_model.model.layers.5.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 3981312 - }, - { - "name": "language_model.model.layers.5.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 4478976 - }, - { - "name": "language_model.model.layers.5.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 12441600 - }, - { - "name": "language_model.model.layers.5.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13436928 - }, - { - "name": "language_model.model.layers.5.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13439232 - }, - { - "name": "language_model.model.layers.5.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13441536 - }, - { - "name": "language_model.model.layers.5.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 13443840 - }, - { - "name": "language_model.model.layers.5.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 13444352 - }, - { - "name": "language_model.model.layers.5.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 13591808 - }, - { - "name": "language_model.model.layers.5.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 13610240 - }, - { - "name": "language_model.model.layers.5.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14200064 - }, - { - "name": "language_model.model.layers.5.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 14273792 - }, - { - "name": "language_model.model.layers.5.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 14274304 - }, - { - "name": "language_model.model.layers.5.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14864128 - }, - { - "name": "language_model.model.layers.5.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 14937856 - }, - { - "name": "language_model.model.layers.5.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 15085312 - }, - { - "name": "language_model.model.layers.6.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 15103744 - }, - { - "name": "language_model.model.layers.6.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 15106048 - }, - { - "name": "language_model.model.layers.6.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 19087360 - }, - { - "name": "language_model.model.layers.6.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 19585024 - }, - { - "name": "language_model.model.layers.6.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 27547648 - }, - { - "name": "language_model.model.layers.6.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28542976 - }, - { - "name": "language_model.model.layers.6.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28545280 - }, - { - "name": "language_model.model.layers.6.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28547584 - }, - { - "name": "language_model.model.layers.6.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 28549888 - }, - { - "name": "language_model.model.layers.6.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 28550400 - }, - { - "name": "language_model.model.layers.6.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 28697856 - }, - { - "name": "language_model.model.layers.6.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 28716288 - }, - { - "name": "language_model.model.layers.6.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 29306112 - }, - { - "name": "language_model.model.layers.6.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 29379840 - }, - { - "name": "language_model.model.layers.6.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 29380352 - }, - { - "name": "language_model.model.layers.6.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 29970176 - }, - { - "name": "language_model.model.layers.6.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 30043904 - }, - { - "name": "language_model.model.layers.6.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 30191360 - }, - { - "name": "language_model.model.layers.7.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 30209792 - } - ], - "md5sum": "e5e50df9866142d2acd2e765a8becd33" - }, - { - "dataPath": "params_shard_13.bin", - "format": "raw-shard", - "nbytes": 30212096, - "records": [ - { - "name": "language_model.model.layers.7.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 0 - }, - { - "name": "language_model.model.layers.7.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 3981312 - }, - { - "name": "language_model.model.layers.7.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 4478976 - }, - { - "name": "language_model.model.layers.7.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 12441600 - }, - { - "name": "language_model.model.layers.7.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13436928 - }, - { - "name": "language_model.model.layers.7.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13439232 - }, - { - "name": "language_model.model.layers.7.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13441536 - }, - { - "name": "language_model.model.layers.7.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 13443840 - }, - { - "name": "language_model.model.layers.7.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 13444352 - }, - { - "name": "language_model.model.layers.7.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 13591808 - }, - { - "name": "language_model.model.layers.7.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 13610240 - }, - { - "name": "language_model.model.layers.7.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14200064 - }, - { - "name": "language_model.model.layers.7.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 14273792 - }, - { - "name": "language_model.model.layers.7.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 14274304 - }, - { - "name": "language_model.model.layers.7.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14864128 - }, - { - "name": "language_model.model.layers.7.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 14937856 - }, - { - "name": "language_model.model.layers.7.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 15085312 - }, - { - "name": "language_model.model.layers.8.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 15103744 - }, - { - "name": "language_model.model.layers.8.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 15106048 - }, - { - "name": "language_model.model.layers.8.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 19087360 - }, - { - "name": "language_model.model.layers.8.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 19585024 - }, - { - "name": "language_model.model.layers.8.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 27547648 - }, - { - "name": "language_model.model.layers.8.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28542976 - }, - { - "name": "language_model.model.layers.8.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28545280 - }, - { - "name": "language_model.model.layers.8.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 28547584 - }, - { - "name": "language_model.model.layers.8.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 28549888 - }, - { - "name": "language_model.model.layers.8.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 28550400 - }, - { - "name": "language_model.model.layers.8.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 28697856 - }, - { - "name": "language_model.model.layers.8.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 28716288 - }, - { - "name": "language_model.model.layers.8.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 29306112 - }, - { - "name": "language_model.model.layers.8.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 29379840 - }, - { - "name": "language_model.model.layers.8.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 29380352 - }, - { - "name": "language_model.model.layers.8.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 29970176 - }, - { - "name": "language_model.model.layers.8.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 30043904 - }, - { - "name": "language_model.model.layers.8.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 30191360 - }, - { - "name": "language_model.model.layers.9.input_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 30209792 - } - ], - "md5sum": "491d6185a6bf5fd694f9f15393cc2e8d" - }, - { - "dataPath": "params_shard_14.bin", - "format": "raw-shard", - "nbytes": 15106048, - "records": [ - { - "name": "language_model.model.layers.9.mlp.down_proj.q_weight", - "shape": [ - 1152, - 864 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 3981312, - "byteOffset": 0 - }, - { - "name": "language_model.model.layers.9.mlp.down_proj.q_scale", - "shape": [ - 1152, - 216 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 497664, - "byteOffset": 3981312 - }, - { - "name": "language_model.model.layers.9.mlp.gate_up_proj.q_weight", - "shape": [ - 13824, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 7962624, - "byteOffset": 4478976 - }, - { - "name": "language_model.model.layers.9.mlp.gate_up_proj.q_scale", - "shape": [ - 13824, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 995328, - "byteOffset": 12441600 - }, - { - "name": "language_model.model.layers.9.post_attention_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13436928 - }, - { - "name": "language_model.model.layers.9.post_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13439232 - }, - { - "name": "language_model.model.layers.9.pre_feedforward_layernorm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 13441536 - }, - { - "name": "language_model.model.layers.9.self_attn.k_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 13443840 - }, - { - "name": "language_model.model.layers.9.self_attn.k_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 13444352 - }, - { - "name": "language_model.model.layers.9.self_attn.k_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 13591808 - }, - { - "name": "language_model.model.layers.9.self_attn.o_proj.q_weight", - "shape": [ - 1152, - 128 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 13610240 - }, - { - "name": "language_model.model.layers.9.self_attn.o_proj.q_scale", - "shape": [ - 1152, - 32 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14200064 - }, - { - "name": "language_model.model.layers.9.self_attn.q_norm.weight", - "shape": [ - 256 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 512, - "byteOffset": 14273792 - }, - { - "name": "language_model.model.layers.9.self_attn.q_proj.q_weight", - "shape": [ - 1024, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 589824, - "byteOffset": 14274304 - }, - { - "name": "language_model.model.layers.9.self_attn.q_proj.q_scale", - "shape": [ - 1024, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 73728, - "byteOffset": 14864128 - }, - { - "name": "language_model.model.layers.9.self_attn.v_proj.q_weight", - "shape": [ - 256, - 144 - ], - "dtype": "uint32", - "format": "f32-to-bf16", - "nbytes": 147456, - "byteOffset": 14937856 - }, - { - "name": "language_model.model.layers.9.self_attn.v_proj.q_scale", - "shape": [ - 256, - 36 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 18432, - "byteOffset": 15085312 - }, - { - "name": "language_model.model.norm.weight", - "shape": [ - 1152 - ], - "dtype": "float16", - "format": "f32-to-bf16", - "nbytes": 2304, - "byteOffset": 15103744 - } - ], - "md5sum": "ce6f27a52869a4b98a296484fbb22c27" + { + "dataPath": "params_shard_14.bin", + "format": "raw-shard", + "nbytes": 16793600, + "records": [ + { + "name": "language_model.model.layers.9.mlp.down_proj.q_weight", + "shape": [ + 1152, + 864 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 3981312, + "byteOffset": 0 + }, + { + "name": "language_model.model.layers.9.mlp.down_proj.q_scale", + "shape": [ + 1152, + 216 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 995328, + "byteOffset": 3981312 + }, + { + "name": "language_model.model.layers.9.mlp.gate_up_proj.q_weight", + "shape": [ + 13824, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 7962624, + "byteOffset": 4976640 + }, + { + "name": "language_model.model.layers.9.mlp.gate_up_proj.q_scale", + "shape": [ + 13824, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1990656, + "byteOffset": 12939264 + }, + { + "name": "language_model.model.layers.9.post_attention_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14929920 + }, + { + "name": "language_model.model.layers.9.post_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14934528 + }, + { + "name": "language_model.model.layers.9.pre_feedforward_layernorm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 14939136 + }, + { + "name": "language_model.model.layers.9.self_attn.k_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 14943744 + }, + { + "name": "language_model.model.layers.9.self_attn.k_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 14944768 + }, + { + "name": "language_model.model.layers.9.self_attn.k_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 15092224 + }, + { + "name": "language_model.model.layers.9.self_attn.o_proj.q_weight", + "shape": [ + 1152, + 128 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15129088 + }, + { + "name": "language_model.model.layers.9.self_attn.o_proj.q_scale", + "shape": [ + 1152, + 32 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 15718912 + }, + { + "name": "language_model.model.layers.9.self_attn.q_norm.weight", + "shape": [ + 256 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 1024, + "byteOffset": 15866368 + }, + { + "name": "language_model.model.layers.9.self_attn.q_proj.q_weight", + "shape": [ + 1024, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 589824, + "byteOffset": 15867392 + }, + { + "name": "language_model.model.layers.9.self_attn.q_proj.q_scale", + "shape": [ + 1024, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16457216 + }, + { + "name": "language_model.model.layers.9.self_attn.v_proj.q_weight", + "shape": [ + 256, + 144 + ], + "dtype": "uint32", + "format": "raw", + "nbytes": 147456, + "byteOffset": 16604672 + }, + { + "name": "language_model.model.layers.9.self_attn.v_proj.q_scale", + "shape": [ + 256, + 36 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 36864, + "byteOffset": 16752128 + }, + { + "name": "language_model.model.norm.weight", + "shape": [ + 1152 + ], + "dtype": "float32", + "format": "raw", + "nbytes": 4608, + "byteOffset": 16788992 } - ] + ], + "md5sum": "20055e6cb5a5c1c1e45e5bab0555cbdf" + } + ] } \ No newline at end of file