diff --git a/mlc-chat-config.json b/mlc-chat-config.json new file mode 100644 index 0000000000000000000000000000000000000000..b2e47b5f91489d9182504b8d2fc3c9e077be02d3 --- /dev/null +++ b/mlc-chat-config.json @@ -0,0 +1,84 @@ +{ + "version": "0.1.0", + "model_type": "gemma2", + "quantization": "q4f32_1", + "model_config": { + "hidden_size": 3584, + "intermediate_size": 14336, + "attention_bias": false, + "num_attention_heads": 16, + "num_key_value_heads": 8, + "head_dim": 256, + "num_hidden_layers": 42, + "rms_norm_eps": 1e-06, + "vocab_size": 256000, + "hidden_activation": "gelu_pytorch_tanh", + "position_embedding_base": 10000.0, + "context_window_size": 4096, + "prefill_chunk_size": 4096, + "tensor_parallel_shards": 1, + "max_batch_size": 128, + "attn_logit_softcapping": 50.0, + "final_logit_softcapping": 30.0, + "query_pre_attn_scalar": 224, + "sliding_window": 4096 + }, + "vocab_size": 256000, + "context_window_size": 4096, + "sliding_window_size": -1, + "prefill_chunk_size": 4096, + "attention_sink_size": -1, + "tensor_parallel_shards": 1, + "pipeline_parallel_stages": 1, + "temperature": 1.0, + "presence_penalty": 0.0, + "frequency_penalty": 0.0, + "repetition_penalty": 1.0, + "top_p": 1.0, + "tokenizer_files": [ + "tokenizer.model", + "tokenizer.json", + "tokenizer_config.json" + ], + "tokenizer_info": { + "token_postproc_method": "byte_fallback", + "prepend_space_in_encode": false, + "strip_space_in_decode": false + }, + "conv_template": { + "name": "gemma_instruction", + "system_template": "{system_message}", + "system_message": "", + "system_prefix_token_ids": [ + 2 + ], + "add_role_after_system_message": true, + "roles": { + "user": "user", + "assistant": "model" + }, + "role_templates": { + "user": "{user_message}", + "assistant": "{assistant_message}", + "tool": "{tool_message}" + }, + "messages": [], + "seps": [ + "\n" + ], + "role_content_sep": "\n", + "role_empty_sep": "\n", + "stop_str": [ + "" + ], + "stop_token_ids": [ + 1, + 107 + ], + "function_string": "", + "use_function_calling": false + }, + "pad_token_id": 0, + "bos_token_id": 2, + "eos_token_id": 1 +} \ No newline at end of file diff --git a/ndarray-cache-b16.json b/ndarray-cache-b16.json new file mode 100644 index 0000000000000000000000000000000000000000..5dceb63b9662cf5bb6f3505de2cd378fdbb0081d --- /dev/null +++ b/ndarray-cache-b16.json @@ -0,0 +1,6449 @@ +{ + "metadata": { + "ParamSize": 507, + "ParamBytes": 5778110464.0, + "BitsPerParam": 4.55004707893633 + }, + "records": [ + { + "dataPath": "params_shard_0.bin", + "format": "raw-shard", + "nbytes": 458752000, + "records": [ + { + "name": "model.embed_tokens.q_weight", + "shape": [ + 256000, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 458752000, + "byteOffset": 0 + } + ], + "md5sum": "581f6c2d88b21fb583e43f7f6028cf67" + }, + { + "dataPath": "params_shard_1.bin", + "format": "raw-shard", + "nbytes": 57344000, + "records": [ + { + "name": "model.embed_tokens.q_scale", + "shape": [ + 256000, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 57344000, + "byteOffset": 0 + } + ], + "md5sum": "3613d9463ee55d16ab676080e0a4dcc2" + }, + { + "dataPath": "params_shard_2.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.0.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "9d77d9a915af93ce2d2bcf467a187315" + }, + { + "dataPath": "params_shard_3.bin", + "format": "raw-shard", + "nbytes": 28908544, + "records": [ + { + "name": "model.layers.0.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 0 + }, + { + "name": "model.layers.0.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 7168 + }, + { + "name": "model.layers.0.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 25697280 + } + ], + "md5sum": "73dcdd344a7804998745ad6bdf8efbf9" + }, + { + "dataPath": "params_shard_4.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.1.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "449f0b29fcc6fc603469532e08f7867e" + }, + { + "dataPath": "params_shard_5.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.0.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.0.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.0.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.0.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.0.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.0.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.0.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.0.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.1.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "cf946937f65b57651d0cd9dab5a18133" + }, + { + "dataPath": "params_shard_6.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.1.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "fcffd48425f63ddd92c316cf73747a84" + }, + { + "dataPath": "params_shard_7.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.1.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.1.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.1.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.1.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.1.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.1.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.1.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.1.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "75fa2858f084f8d981917bce055531e9" + }, + { + "dataPath": "params_shard_8.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.10.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "c8e5f798813da7a1b5d83f544b897bc6" + }, + { + "dataPath": "params_shard_9.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.1.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.10.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.10.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.10.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "7a63c7d93170ea80faeefdfa55facfec" + }, + { + "dataPath": "params_shard_10.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.11.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "82e504b71b9fdf94978a8380bb24aceb" + }, + { + "dataPath": "params_shard_11.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.10.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.10.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.10.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.10.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.10.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.10.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.10.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.10.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.11.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "df82d36cd85a5ee1e4d99995e3e671b9" + }, + { + "dataPath": "params_shard_12.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.11.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "9452f6ee26bc5b61ed950e974e45b398" + }, + { + "dataPath": "params_shard_13.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.11.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.11.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.11.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.11.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.11.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.11.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.11.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.11.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "8a04e974e90f8242f75c3c3ed86ebc88" + }, + { + "dataPath": "params_shard_14.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.12.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "2f883cc8a1975e556502768aab8aa1fe" + }, + { + "dataPath": "params_shard_15.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.11.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.12.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.12.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.12.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "dc10740f440de748772d30a2cf7f0e28" + }, + { + "dataPath": "params_shard_16.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.13.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "5d78bcc1fad887aae2608801b0ff49f4" + }, + { + "dataPath": "params_shard_17.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.12.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.12.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.12.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.12.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.12.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.12.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.12.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.12.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.13.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "156c0b8f10a699c9aa57149e3140c030" + }, + { + "dataPath": "params_shard_18.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.13.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "1e757436832e8e2cff7f6cb31b4d756e" + }, + { + "dataPath": "params_shard_19.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.13.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.13.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.13.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.13.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.13.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.13.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.13.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.13.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "97fd3b8333b5b7b2f0386e79a09d8de0" + }, + { + "dataPath": "params_shard_20.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.14.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "752138f9c4380234d8aee0664113dd8e" + }, + { + "dataPath": "params_shard_21.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.13.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.14.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.14.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.14.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "0a1754087fde872adf457971555fc1ba" + }, + { + "dataPath": "params_shard_22.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.15.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "a8a2bf61429b1c55964ebcdec651a3fd" + }, + { + "dataPath": "params_shard_23.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.14.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.14.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.14.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.14.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.14.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.14.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.14.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.14.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.15.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "130bff917c0fee604112d8161c3639b4" + }, + { + "dataPath": "params_shard_24.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.15.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "796d373866bd97e11bf70408805f028b" + }, + { + "dataPath": "params_shard_25.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.15.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.15.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.15.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.15.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.15.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.15.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.15.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.15.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "9474827d73a49d623cf8d7959444c8e0" + }, + { + "dataPath": "params_shard_26.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.16.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "351716e6742f2f60dcd9f1b2d7e44b24" + }, + { + "dataPath": "params_shard_27.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.15.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.16.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.16.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.16.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "f2dbda755098afa22a0932852fdd962d" + }, + { + "dataPath": "params_shard_28.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.17.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "940a302682f06a3eeb2ad192b94975d7" + }, + { + "dataPath": "params_shard_29.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.16.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.16.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.16.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.16.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.16.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.16.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.16.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.16.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.17.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "2b1a424d5be78d560ea73183a0c0c915" + }, + { + "dataPath": "params_shard_30.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.17.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "62684598adc77ae83375c413c071e42c" + }, + { + "dataPath": "params_shard_31.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.17.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.17.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.17.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.17.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.17.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.17.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.17.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.17.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "bce688b7f9a7a4f3057b96f7c78e0bbe" + }, + { + "dataPath": "params_shard_32.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.18.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "42d9afa7f667ff0719310729e240c0df" + }, + { + "dataPath": "params_shard_33.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.17.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.18.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.18.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.18.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "e0e02f8b8a2b2b6c2c5db2d050209d9a" + }, + { + "dataPath": "params_shard_34.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.19.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "15c42f6eb7fbc21984ad3937d752b59e" + }, + { + "dataPath": "params_shard_35.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.18.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.18.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.18.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.18.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.18.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.18.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.18.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.18.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.19.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "4f7a9b8dd4e46fb3666e29bea38d4505" + }, + { + "dataPath": "params_shard_36.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.19.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "6db73ecaab4ce736f561ede468249296" + }, + { + "dataPath": "params_shard_37.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.19.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.19.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.19.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.19.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.19.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.19.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.19.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.19.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "e6e6b8c064037f9a61784e862bd06f39" + }, + { + "dataPath": "params_shard_38.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.2.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "df26370d30899ef3a21931124c739f70" + }, + { + "dataPath": "params_shard_39.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.19.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.2.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.2.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.2.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "a32d3675e4bddd406e6ebcdcf613337a" + }, + { + "dataPath": "params_shard_40.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.20.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "7e4f821db7653ffcb961556a24d3ab0b" + }, + { + "dataPath": "params_shard_41.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.2.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.2.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.2.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.2.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.2.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.2.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.2.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.2.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.20.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "dbae62fdd15f2faa77232ea2f4f71b75" + }, + { + "dataPath": "params_shard_42.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.20.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "398f5212d8a294819e1f88329da80e96" + }, + { + "dataPath": "params_shard_43.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.20.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.20.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.20.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.20.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.20.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.20.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.20.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.20.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "32983ba180355cfcfe4b7d4250308d31" + }, + { + "dataPath": "params_shard_44.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.21.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "56334118d6b3c61e5cfc3e798cd5cc38" + }, + { + "dataPath": "params_shard_45.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.20.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.21.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.21.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.21.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "95cfee9bbf3655398a5235d60b3c17fa" + }, + { + "dataPath": "params_shard_46.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.22.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "0db14be904f016794c2362267293a49f" + }, + { + "dataPath": "params_shard_47.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.21.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.21.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.21.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.21.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.21.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.21.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.21.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.21.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.22.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "18764066077c2f3b804f24f7c4357993" + }, + { + "dataPath": "params_shard_48.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.22.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "1f6a5b382ace19548b43181c4ebbcb35" + }, + { + "dataPath": "params_shard_49.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.22.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.22.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.22.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.22.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.22.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.22.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.22.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.22.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "14843b07637321fcf8438319b411ec98" + }, + { + "dataPath": "params_shard_50.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.23.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "ad742e5014ff76c424b2bcf5764beffd" + }, + { + "dataPath": "params_shard_51.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.22.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.23.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.23.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.23.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "840965c1db059033784a1de6db0a5b8c" + }, + { + "dataPath": "params_shard_52.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.24.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "0773261fc9fb376fad9ce247d3d7f7aa" + }, + { + "dataPath": "params_shard_53.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.23.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.23.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.23.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.23.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.23.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.23.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.23.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.23.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.24.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "7ba1449744bf4b90d7a8410d951ae3bd" + }, + { + "dataPath": "params_shard_54.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.24.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "015c92d1a2259abd9783e9c09fd1948c" + }, + { + "dataPath": "params_shard_55.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.24.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.24.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.24.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.24.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.24.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.24.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.24.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.24.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "313a7980eb24ce27f4eaff95488bb16c" + }, + { + "dataPath": "params_shard_56.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.25.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "3daa3e78ba72975f65f1263f58672976" + }, + { + "dataPath": "params_shard_57.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.24.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.25.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.25.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.25.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "f4bbe8b4417dc18fd52a1f901d6903d4" + }, + { + "dataPath": "params_shard_58.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.26.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "16297be7b75400c9615262e78bb22e6f" + }, + { + "dataPath": "params_shard_59.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.25.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.25.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.25.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.25.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.25.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.25.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.25.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.25.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.26.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "4d0064c67970169057bc30df4362865c" + }, + { + "dataPath": "params_shard_60.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.26.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "07990ec64ce2cff1fc8802c0145a5533" + }, + { + "dataPath": "params_shard_61.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.26.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.26.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.26.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.26.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.26.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.26.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.26.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.26.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "a42dfbec720748392bcb28770dcfea5b" + }, + { + "dataPath": "params_shard_62.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.27.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "e0b346365bb21d2e1c22b8e005ed13db" + }, + { + "dataPath": "params_shard_63.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.26.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.27.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.27.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.27.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "4314d4e919559148ecfcad11309efaf5" + }, + { + "dataPath": "params_shard_64.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.28.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "cd6223c7f76df633c6702f9298c74973" + }, + { + "dataPath": "params_shard_65.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.27.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.27.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.27.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.27.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.27.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.27.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.27.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.27.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.28.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "83ef3471267866675b4384645cdc3d86" + }, + { + "dataPath": "params_shard_66.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.28.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "b40a0a35652eae64df9098fb61170e1d" + }, + { + "dataPath": "params_shard_67.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.28.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.28.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.28.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.28.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.28.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.28.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.28.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.28.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "50b896c731508d4db1866392e046b61f" + }, + { + "dataPath": "params_shard_68.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.29.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "acaae004e5fcc6b6df8f80d1dd336f75" + }, + { + "dataPath": "params_shard_69.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.28.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.29.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.29.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.29.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "4d072a21893238a66098950a428e506b" + }, + { + "dataPath": "params_shard_70.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.3.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "68df6a75f06aa4d60c8bb40fc57c15e4" + }, + { + "dataPath": "params_shard_71.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.29.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.29.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.29.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.29.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.29.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.29.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.29.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.29.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.3.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "59378843c469b9f67e5c6838421a6cd4" + }, + { + "dataPath": "params_shard_72.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.3.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "dabec2f9f17c3603d22b347dd0d283a2" + }, + { + "dataPath": "params_shard_73.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.3.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.3.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.3.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.3.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.3.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.3.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.3.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.3.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "2afa57068365721551cad660b8749c03" + }, + { + "dataPath": "params_shard_74.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.30.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "484491025b3d245bb4d01c8d76319a09" + }, + { + "dataPath": "params_shard_75.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.3.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.30.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.30.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.30.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "9799f54e85a2c4d740f7eb6ddd131323" + }, + { + "dataPath": "params_shard_76.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.31.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "404bceec731de46fe2931b549381e933" + }, + { + "dataPath": "params_shard_77.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.30.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.30.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.30.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.30.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.30.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.30.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.30.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.30.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.31.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "604cea7597e6d9620179f10e74ffbb2f" + }, + { + "dataPath": "params_shard_78.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.31.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "4f0d03da9ffe936a7f27acc9fa85b396" + }, + { + "dataPath": "params_shard_79.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.31.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.31.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.31.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.31.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.31.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.31.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.31.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.31.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "2a109d13c58db7ecb60e91f9dd20e77d" + }, + { + "dataPath": "params_shard_80.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.32.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "a3760a3254f9d99e159e17bb008e8388" + }, + { + "dataPath": "params_shard_81.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.31.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.32.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.32.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.32.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "957cf13ecb76c037a1c3d8511af623c8" + }, + { + "dataPath": "params_shard_82.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.33.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "27bbe2b513b7e96fe7b0d8678d3e5f5c" + }, + { + "dataPath": "params_shard_83.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.32.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.32.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.32.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.32.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.32.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.32.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.32.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.32.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.33.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "ef218c747796ec499d8c4879645f19d4" + }, + { + "dataPath": "params_shard_84.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.33.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "b00b944813a906f81077e627cd913bc4" + }, + { + "dataPath": "params_shard_85.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.33.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.33.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.33.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.33.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.33.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.33.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.33.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.33.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "61efbbdba707673e7a7c067330a2dd17" + }, + { + "dataPath": "params_shard_86.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.34.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "03395427be7b3b46dbb124428e969c10" + }, + { + "dataPath": "params_shard_87.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.33.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.34.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.34.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.34.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "7c06da46b686cbb6876e213d7a7f279a" + }, + { + "dataPath": "params_shard_88.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.35.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "0ecbf386a46750f266662cc1e2793526" + }, + { + "dataPath": "params_shard_89.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.34.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.34.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.34.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.34.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.34.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.34.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.34.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.34.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.35.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "c5de48b683a7076390b69697d90cc63a" + }, + { + "dataPath": "params_shard_90.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.35.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "c5c86adae657dbe83f910b5159330ec8" + }, + { + "dataPath": "params_shard_91.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.35.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.35.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.35.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.35.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.35.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.35.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.35.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.35.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "71db9d1796bbf1d39fd47364c82e7e78" + }, + { + "dataPath": "params_shard_92.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.36.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "81272fa33ae8e508745977d264ae41ce" + }, + { + "dataPath": "params_shard_93.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.35.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.36.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.36.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.36.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "87223323ad3e69c932233d6376953ca8" + }, + { + "dataPath": "params_shard_94.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.37.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "ed0f09bbcd695501a585da62e868781a" + }, + { + "dataPath": "params_shard_95.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.36.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.36.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.36.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.36.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.36.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.36.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.36.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.36.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.37.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "721af0c921f0d507bed441169c377d78" + }, + { + "dataPath": "params_shard_96.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.37.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "2557d8e1b348b67827e834ca42d0d22f" + }, + { + "dataPath": "params_shard_97.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.37.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.37.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.37.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.37.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.37.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.37.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.37.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.37.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "70dc7a4b3180f0db9f3eb707b8a30b6e" + }, + { + "dataPath": "params_shard_98.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.38.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "bbca11ea40ec53c7221ccb5a9f3250db" + }, + { + "dataPath": "params_shard_99.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.37.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.38.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.38.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.38.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "cd0a898a4a35eafa30f69d8c55666ec7" + }, + { + "dataPath": "params_shard_100.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.39.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "b5006c14282c975cc6ef1f66507a4d19" + }, + { + "dataPath": "params_shard_101.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.38.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.38.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.38.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.38.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.38.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.38.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.38.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.38.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.39.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "08ecdc8b63a8035706a12b1fbc77664e" + }, + { + "dataPath": "params_shard_102.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.39.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "957f24bf32d360ace5cfb9f51b1451fe" + }, + { + "dataPath": "params_shard_103.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.39.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.39.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.39.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.39.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.39.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.39.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.39.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.39.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "4f7c461d90072999ea9cf11aab8a38cb" + }, + { + "dataPath": "params_shard_104.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.4.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "23e270e726b907aa91c4e3c065acd39a" + }, + { + "dataPath": "params_shard_105.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.39.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.4.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.4.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.4.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "3e3ac8b52e514fbadfd8db34847aa798" + }, + { + "dataPath": "params_shard_106.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.40.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "8049d10a234bd075e6608253d177c5ee" + }, + { + "dataPath": "params_shard_107.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.4.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.4.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.4.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.4.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.4.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.4.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.4.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.4.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.40.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "f0db144e8417ec10ef3f68c38a3dbd34" + }, + { + "dataPath": "params_shard_108.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.40.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "b1f9ecfbcbb486dc5007ca4710d05f79" + }, + { + "dataPath": "params_shard_109.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.40.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.40.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.40.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.40.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.40.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.40.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.40.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.40.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "07c1b40bc7d57b818d0e2959355e03f5" + }, + { + "dataPath": "params_shard_110.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.41.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "4e755b65a605d4f8bfada7e00fdd2178" + }, + { + "dataPath": "params_shard_111.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.40.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.41.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.41.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.41.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "baedaa05248279d2fc4db2dcfa9b7971" + }, + { + "dataPath": "params_shard_112.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.5.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "4f6af57921dfb47d821f37aef5c28669" + }, + { + "dataPath": "params_shard_113.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.41.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.41.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.41.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.41.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.41.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.41.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.41.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.41.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.5.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "98e4865123ebe1b769035e8603a1cb3e" + }, + { + "dataPath": "params_shard_114.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.5.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "5c57d68b2981bc7abfc7b1bca436f2f1" + }, + { + "dataPath": "params_shard_115.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.5.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.5.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.5.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.5.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.5.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.5.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.5.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.5.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "1fb6240d7330e2d89bd7cd9602982dc4" + }, + { + "dataPath": "params_shard_116.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.6.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "85d9a33e8b538615f0919d970159cab2" + }, + { + "dataPath": "params_shard_117.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.5.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.6.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.6.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.6.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "e061ada5689bfc093a72c9a51498e734" + }, + { + "dataPath": "params_shard_118.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.7.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "1a270e7a25e7546ad9c92d36258652ae" + }, + { + "dataPath": "params_shard_119.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.6.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.6.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.6.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.6.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.6.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.6.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.6.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.6.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.7.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "53c0fb2f79bad56cc0d4d5988bcb21a5" + }, + { + "dataPath": "params_shard_120.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.7.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "8efb466707284864f76911509f7f8557" + }, + { + "dataPath": "params_shard_121.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.7.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.7.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.7.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.7.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.7.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.7.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.7.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.7.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "e1599620555fc631a40f8305336206d7" + }, + { + "dataPath": "params_shard_122.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.8.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "5bd7cfc1c4a7f3cc30dd2595a741f875" + }, + { + "dataPath": "params_shard_123.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.7.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.8.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.8.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.8.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "def4f59751241a114b2eced52edff5b1" + }, + { + "dataPath": "params_shard_124.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.9.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "d8c30599ef73a51613cc2c41394399ec" + }, + { + "dataPath": "params_shard_125.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.8.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.8.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.8.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.8.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.8.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.8.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.8.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.8.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.9.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "3684830f59f6755905f843adcc696789" + }, + { + "dataPath": "params_shard_126.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.9.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "7f028d14ee44583e6d8a160c58a6f422" + }, + { + "dataPath": "params_shard_127.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.9.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.9.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.9.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.9.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.9.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.9.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.9.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.9.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "22b66d4a54d6fcdc6c67c49376f3bc31" + }, + { + "dataPath": "params_shard_128.bin", + "format": "raw-shard", + "nbytes": 924672, + "records": [ + { + "name": "model.layers.9.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.norm.weight", + "shape": [ + 3584 + ], + "dtype": "bfloat16", + "format": "raw", + "nbytes": 7168, + "byteOffset": 917504 + } + ], + "md5sum": "3b5e874513a64b93d2437c958569ec45" + } + ] +} \ No newline at end of file diff --git a/ndarray-cache.json b/ndarray-cache.json new file mode 100644 index 0000000000000000000000000000000000000000..edd9ff15960a5fa20d57689caae77ae27baf1b51 --- /dev/null +++ b/ndarray-cache.json @@ -0,0 +1,6449 @@ +{ + "metadata": { + "ParamSize": 507, + "ParamBytes": 5778110464.0, + "BitsPerParam": 4.55004707893633 + }, + "records": [ + { + "dataPath": "params_shard_0.bin", + "format": "raw-shard", + "nbytes": 458752000, + "records": [ + { + "name": "model.embed_tokens.q_weight", + "shape": [ + 256000, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 458752000, + "byteOffset": 0 + } + ], + "md5sum": "581f6c2d88b21fb583e43f7f6028cf67" + }, + { + "dataPath": "params_shard_1.bin", + "format": "raw-shard", + "nbytes": 57344000, + "records": [ + { + "name": "model.embed_tokens.q_scale", + "shape": [ + 256000, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 57344000, + "byteOffset": 0 + } + ], + "md5sum": "3613d9463ee55d16ab676080e0a4dcc2" + }, + { + "dataPath": "params_shard_2.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.0.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "9d77d9a915af93ce2d2bcf467a187315" + }, + { + "dataPath": "params_shard_3.bin", + "format": "raw-shard", + "nbytes": 28908544, + "records": [ + { + "name": "model.layers.0.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 0 + }, + { + "name": "model.layers.0.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 7168 + }, + { + "name": "model.layers.0.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 25697280 + } + ], + "md5sum": "73dcdd344a7804998745ad6bdf8efbf9" + }, + { + "dataPath": "params_shard_4.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.1.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "449f0b29fcc6fc603469532e08f7867e" + }, + { + "dataPath": "params_shard_5.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.0.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.0.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.0.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.0.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.0.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.0.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.0.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.0.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.1.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "cf946937f65b57651d0cd9dab5a18133" + }, + { + "dataPath": "params_shard_6.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.1.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "fcffd48425f63ddd92c316cf73747a84" + }, + { + "dataPath": "params_shard_7.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.1.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.1.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.1.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.1.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.1.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.1.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.1.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.1.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "75fa2858f084f8d981917bce055531e9" + }, + { + "dataPath": "params_shard_8.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.10.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "c8e5f798813da7a1b5d83f544b897bc6" + }, + { + "dataPath": "params_shard_9.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.1.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.10.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.10.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.10.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "7a63c7d93170ea80faeefdfa55facfec" + }, + { + "dataPath": "params_shard_10.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.11.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "82e504b71b9fdf94978a8380bb24aceb" + }, + { + "dataPath": "params_shard_11.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.10.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.10.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.10.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.10.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.10.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.10.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.10.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.10.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.11.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "df82d36cd85a5ee1e4d99995e3e671b9" + }, + { + "dataPath": "params_shard_12.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.11.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "9452f6ee26bc5b61ed950e974e45b398" + }, + { + "dataPath": "params_shard_13.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.11.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.11.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.11.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.11.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.11.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.11.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.11.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.11.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "8a04e974e90f8242f75c3c3ed86ebc88" + }, + { + "dataPath": "params_shard_14.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.12.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "2f883cc8a1975e556502768aab8aa1fe" + }, + { + "dataPath": "params_shard_15.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.11.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.12.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.12.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.12.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "dc10740f440de748772d30a2cf7f0e28" + }, + { + "dataPath": "params_shard_16.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.13.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "5d78bcc1fad887aae2608801b0ff49f4" + }, + { + "dataPath": "params_shard_17.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.12.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.12.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.12.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.12.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.12.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.12.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.12.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.12.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.13.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "156c0b8f10a699c9aa57149e3140c030" + }, + { + "dataPath": "params_shard_18.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.13.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "1e757436832e8e2cff7f6cb31b4d756e" + }, + { + "dataPath": "params_shard_19.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.13.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.13.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.13.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.13.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.13.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.13.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.13.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.13.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "97fd3b8333b5b7b2f0386e79a09d8de0" + }, + { + "dataPath": "params_shard_20.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.14.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "752138f9c4380234d8aee0664113dd8e" + }, + { + "dataPath": "params_shard_21.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.13.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.14.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.14.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.14.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "0a1754087fde872adf457971555fc1ba" + }, + { + "dataPath": "params_shard_22.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.15.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "a8a2bf61429b1c55964ebcdec651a3fd" + }, + { + "dataPath": "params_shard_23.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.14.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.14.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.14.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.14.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.14.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.14.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.14.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.14.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.15.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "130bff917c0fee604112d8161c3639b4" + }, + { + "dataPath": "params_shard_24.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.15.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "796d373866bd97e11bf70408805f028b" + }, + { + "dataPath": "params_shard_25.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.15.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.15.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.15.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.15.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.15.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.15.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.15.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.15.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "9474827d73a49d623cf8d7959444c8e0" + }, + { + "dataPath": "params_shard_26.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.16.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "351716e6742f2f60dcd9f1b2d7e44b24" + }, + { + "dataPath": "params_shard_27.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.15.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.16.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.16.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.16.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "f2dbda755098afa22a0932852fdd962d" + }, + { + "dataPath": "params_shard_28.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.17.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "940a302682f06a3eeb2ad192b94975d7" + }, + { + "dataPath": "params_shard_29.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.16.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.16.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.16.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.16.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.16.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.16.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.16.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.16.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.17.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "2b1a424d5be78d560ea73183a0c0c915" + }, + { + "dataPath": "params_shard_30.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.17.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "62684598adc77ae83375c413c071e42c" + }, + { + "dataPath": "params_shard_31.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.17.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.17.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.17.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.17.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.17.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.17.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.17.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.17.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "bce688b7f9a7a4f3057b96f7c78e0bbe" + }, + { + "dataPath": "params_shard_32.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.18.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "42d9afa7f667ff0719310729e240c0df" + }, + { + "dataPath": "params_shard_33.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.17.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.18.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.18.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.18.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "e0e02f8b8a2b2b6c2c5db2d050209d9a" + }, + { + "dataPath": "params_shard_34.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.19.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "15c42f6eb7fbc21984ad3937d752b59e" + }, + { + "dataPath": "params_shard_35.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.18.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.18.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.18.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.18.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.18.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.18.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.18.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.18.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.19.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "4f7a9b8dd4e46fb3666e29bea38d4505" + }, + { + "dataPath": "params_shard_36.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.19.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "6db73ecaab4ce736f561ede468249296" + }, + { + "dataPath": "params_shard_37.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.19.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.19.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.19.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.19.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.19.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.19.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.19.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.19.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "e6e6b8c064037f9a61784e862bd06f39" + }, + { + "dataPath": "params_shard_38.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.2.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "df26370d30899ef3a21931124c739f70" + }, + { + "dataPath": "params_shard_39.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.19.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.2.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.2.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.2.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "a32d3675e4bddd406e6ebcdcf613337a" + }, + { + "dataPath": "params_shard_40.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.20.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "7e4f821db7653ffcb961556a24d3ab0b" + }, + { + "dataPath": "params_shard_41.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.2.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.2.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.2.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.2.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.2.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.2.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.2.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.2.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.20.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "dbae62fdd15f2faa77232ea2f4f71b75" + }, + { + "dataPath": "params_shard_42.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.20.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "398f5212d8a294819e1f88329da80e96" + }, + { + "dataPath": "params_shard_43.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.20.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.20.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.20.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.20.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.20.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.20.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.20.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.20.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "32983ba180355cfcfe4b7d4250308d31" + }, + { + "dataPath": "params_shard_44.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.21.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "56334118d6b3c61e5cfc3e798cd5cc38" + }, + { + "dataPath": "params_shard_45.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.20.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.21.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.21.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.21.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "95cfee9bbf3655398a5235d60b3c17fa" + }, + { + "dataPath": "params_shard_46.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.22.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "0db14be904f016794c2362267293a49f" + }, + { + "dataPath": "params_shard_47.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.21.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.21.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.21.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.21.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.21.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.21.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.21.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.21.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.22.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "18764066077c2f3b804f24f7c4357993" + }, + { + "dataPath": "params_shard_48.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.22.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "1f6a5b382ace19548b43181c4ebbcb35" + }, + { + "dataPath": "params_shard_49.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.22.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.22.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.22.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.22.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.22.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.22.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.22.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.22.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "14843b07637321fcf8438319b411ec98" + }, + { + "dataPath": "params_shard_50.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.23.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "ad742e5014ff76c424b2bcf5764beffd" + }, + { + "dataPath": "params_shard_51.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.22.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.23.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.23.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.23.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "840965c1db059033784a1de6db0a5b8c" + }, + { + "dataPath": "params_shard_52.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.24.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "0773261fc9fb376fad9ce247d3d7f7aa" + }, + { + "dataPath": "params_shard_53.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.23.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.23.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.23.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.23.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.23.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.23.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.23.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.23.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.24.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "7ba1449744bf4b90d7a8410d951ae3bd" + }, + { + "dataPath": "params_shard_54.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.24.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "015c92d1a2259abd9783e9c09fd1948c" + }, + { + "dataPath": "params_shard_55.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.24.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.24.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.24.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.24.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.24.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.24.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.24.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.24.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "313a7980eb24ce27f4eaff95488bb16c" + }, + { + "dataPath": "params_shard_56.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.25.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "3daa3e78ba72975f65f1263f58672976" + }, + { + "dataPath": "params_shard_57.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.24.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.25.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.25.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.25.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "f4bbe8b4417dc18fd52a1f901d6903d4" + }, + { + "dataPath": "params_shard_58.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.26.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "16297be7b75400c9615262e78bb22e6f" + }, + { + "dataPath": "params_shard_59.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.25.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.25.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.25.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.25.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.25.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.25.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.25.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.25.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.26.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "4d0064c67970169057bc30df4362865c" + }, + { + "dataPath": "params_shard_60.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.26.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "07990ec64ce2cff1fc8802c0145a5533" + }, + { + "dataPath": "params_shard_61.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.26.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.26.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.26.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.26.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.26.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.26.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.26.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.26.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "a42dfbec720748392bcb28770dcfea5b" + }, + { + "dataPath": "params_shard_62.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.27.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "e0b346365bb21d2e1c22b8e005ed13db" + }, + { + "dataPath": "params_shard_63.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.26.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.27.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.27.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.27.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "4314d4e919559148ecfcad11309efaf5" + }, + { + "dataPath": "params_shard_64.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.28.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "cd6223c7f76df633c6702f9298c74973" + }, + { + "dataPath": "params_shard_65.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.27.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.27.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.27.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.27.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.27.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.27.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.27.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.27.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.28.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "83ef3471267866675b4384645cdc3d86" + }, + { + "dataPath": "params_shard_66.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.28.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "b40a0a35652eae64df9098fb61170e1d" + }, + { + "dataPath": "params_shard_67.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.28.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.28.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.28.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.28.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.28.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.28.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.28.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.28.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "50b896c731508d4db1866392e046b61f" + }, + { + "dataPath": "params_shard_68.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.29.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "acaae004e5fcc6b6df8f80d1dd336f75" + }, + { + "dataPath": "params_shard_69.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.28.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.29.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.29.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.29.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "4d072a21893238a66098950a428e506b" + }, + { + "dataPath": "params_shard_70.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.3.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "68df6a75f06aa4d60c8bb40fc57c15e4" + }, + { + "dataPath": "params_shard_71.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.29.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.29.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.29.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.29.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.29.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.29.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.29.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.29.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.3.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "59378843c469b9f67e5c6838421a6cd4" + }, + { + "dataPath": "params_shard_72.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.3.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "dabec2f9f17c3603d22b347dd0d283a2" + }, + { + "dataPath": "params_shard_73.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.3.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.3.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.3.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.3.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.3.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.3.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.3.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.3.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "2afa57068365721551cad660b8749c03" + }, + { + "dataPath": "params_shard_74.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.30.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "484491025b3d245bb4d01c8d76319a09" + }, + { + "dataPath": "params_shard_75.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.3.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.30.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.30.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.30.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "9799f54e85a2c4d740f7eb6ddd131323" + }, + { + "dataPath": "params_shard_76.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.31.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "404bceec731de46fe2931b549381e933" + }, + { + "dataPath": "params_shard_77.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.30.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.30.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.30.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.30.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.30.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.30.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.30.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.30.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.31.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "604cea7597e6d9620179f10e74ffbb2f" + }, + { + "dataPath": "params_shard_78.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.31.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "4f0d03da9ffe936a7f27acc9fa85b396" + }, + { + "dataPath": "params_shard_79.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.31.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.31.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.31.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.31.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.31.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.31.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.31.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.31.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "2a109d13c58db7ecb60e91f9dd20e77d" + }, + { + "dataPath": "params_shard_80.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.32.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "a3760a3254f9d99e159e17bb008e8388" + }, + { + "dataPath": "params_shard_81.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.31.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.32.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.32.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.32.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "957cf13ecb76c037a1c3d8511af623c8" + }, + { + "dataPath": "params_shard_82.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.33.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "27bbe2b513b7e96fe7b0d8678d3e5f5c" + }, + { + "dataPath": "params_shard_83.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.32.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.32.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.32.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.32.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.32.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.32.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.32.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.32.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.33.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "ef218c747796ec499d8c4879645f19d4" + }, + { + "dataPath": "params_shard_84.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.33.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "b00b944813a906f81077e627cd913bc4" + }, + { + "dataPath": "params_shard_85.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.33.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.33.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.33.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.33.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.33.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.33.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.33.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.33.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "61efbbdba707673e7a7c067330a2dd17" + }, + { + "dataPath": "params_shard_86.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.34.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "03395427be7b3b46dbb124428e969c10" + }, + { + "dataPath": "params_shard_87.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.33.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.34.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.34.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.34.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "7c06da46b686cbb6876e213d7a7f279a" + }, + { + "dataPath": "params_shard_88.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.35.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "0ecbf386a46750f266662cc1e2793526" + }, + { + "dataPath": "params_shard_89.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.34.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.34.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.34.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.34.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.34.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.34.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.34.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.34.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.35.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "c5de48b683a7076390b69697d90cc63a" + }, + { + "dataPath": "params_shard_90.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.35.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "c5c86adae657dbe83f910b5159330ec8" + }, + { + "dataPath": "params_shard_91.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.35.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.35.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.35.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.35.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.35.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.35.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.35.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.35.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "71db9d1796bbf1d39fd47364c82e7e78" + }, + { + "dataPath": "params_shard_92.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.36.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "81272fa33ae8e508745977d264ae41ce" + }, + { + "dataPath": "params_shard_93.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.35.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.36.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.36.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.36.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "87223323ad3e69c932233d6376953ca8" + }, + { + "dataPath": "params_shard_94.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.37.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "ed0f09bbcd695501a585da62e868781a" + }, + { + "dataPath": "params_shard_95.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.36.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.36.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.36.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.36.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.36.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.36.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.36.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.36.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.37.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "721af0c921f0d507bed441169c377d78" + }, + { + "dataPath": "params_shard_96.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.37.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "2557d8e1b348b67827e834ca42d0d22f" + }, + { + "dataPath": "params_shard_97.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.37.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.37.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.37.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.37.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.37.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.37.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.37.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.37.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "70dc7a4b3180f0db9f3eb707b8a30b6e" + }, + { + "dataPath": "params_shard_98.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.38.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "bbca11ea40ec53c7221ccb5a9f3250db" + }, + { + "dataPath": "params_shard_99.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.37.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.38.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.38.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.38.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "cd0a898a4a35eafa30f69d8c55666ec7" + }, + { + "dataPath": "params_shard_100.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.39.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "b5006c14282c975cc6ef1f66507a4d19" + }, + { + "dataPath": "params_shard_101.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.38.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.38.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.38.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.38.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.38.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.38.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.38.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.38.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.39.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "08ecdc8b63a8035706a12b1fbc77664e" + }, + { + "dataPath": "params_shard_102.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.39.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "957f24bf32d360ace5cfb9f51b1451fe" + }, + { + "dataPath": "params_shard_103.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.39.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.39.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.39.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.39.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.39.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.39.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.39.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.39.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "4f7c461d90072999ea9cf11aab8a38cb" + }, + { + "dataPath": "params_shard_104.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.4.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "23e270e726b907aa91c4e3c065acd39a" + }, + { + "dataPath": "params_shard_105.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.39.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.4.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.4.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.4.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "3e3ac8b52e514fbadfd8db34847aa798" + }, + { + "dataPath": "params_shard_106.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.40.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "8049d10a234bd075e6608253d177c5ee" + }, + { + "dataPath": "params_shard_107.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.4.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.4.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.4.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.4.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.4.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.4.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.4.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.4.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.40.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "f0db144e8417ec10ef3f68c38a3dbd34" + }, + { + "dataPath": "params_shard_108.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.40.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "b1f9ecfbcbb486dc5007ca4710d05f79" + }, + { + "dataPath": "params_shard_109.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.40.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.40.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.40.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.40.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.40.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.40.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.40.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.40.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "07c1b40bc7d57b818d0e2959355e03f5" + }, + { + "dataPath": "params_shard_110.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.41.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "4e755b65a605d4f8bfada7e00fdd2178" + }, + { + "dataPath": "params_shard_111.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.40.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.41.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.41.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.41.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "baedaa05248279d2fc4db2dcfa9b7971" + }, + { + "dataPath": "params_shard_112.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.5.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "4f6af57921dfb47d821f37aef5c28669" + }, + { + "dataPath": "params_shard_113.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.41.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.41.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.41.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.41.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.41.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.41.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.41.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.41.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.5.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "98e4865123ebe1b769035e8603a1cb3e" + }, + { + "dataPath": "params_shard_114.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.5.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "5c57d68b2981bc7abfc7b1bca436f2f1" + }, + { + "dataPath": "params_shard_115.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.5.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.5.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.5.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.5.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.5.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.5.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.5.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.5.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "1fb6240d7330e2d89bd7cd9602982dc4" + }, + { + "dataPath": "params_shard_116.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.6.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "85d9a33e8b538615f0919d970159cab2" + }, + { + "dataPath": "params_shard_117.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.5.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.6.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.6.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.6.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "e061ada5689bfc093a72c9a51498e734" + }, + { + "dataPath": "params_shard_118.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.7.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "1a270e7a25e7546ad9c92d36258652ae" + }, + { + "dataPath": "params_shard_119.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.6.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.6.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.6.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.6.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.6.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.6.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.6.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.6.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.7.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "53c0fb2f79bad56cc0d4d5988bcb21a5" + }, + { + "dataPath": "params_shard_120.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.7.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "8efb466707284864f76911509f7f8557" + }, + { + "dataPath": "params_shard_121.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.7.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.7.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.7.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.7.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.7.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.7.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.7.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.7.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "e1599620555fc631a40f8305336206d7" + }, + { + "dataPath": "params_shard_122.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.8.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "5bd7cfc1c4a7f3cc30dd2595a741f875" + }, + { + "dataPath": "params_shard_123.bin", + "format": "raw-shard", + "nbytes": 29826048, + "records": [ + { + "name": "model.layers.7.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.layers.8.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 917504 + }, + { + "name": "model.layers.8.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 924672 + }, + { + "name": "model.layers.8.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 26614784 + } + ], + "md5sum": "def4f59751241a114b2eced52edff5b1" + }, + { + "dataPath": "params_shard_124.bin", + "format": "raw-shard", + "nbytes": 25690112, + "records": [ + { + "name": "model.layers.9.mlp.down_proj.q_weight", + "shape": [ + 3584, + 1792 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 25690112, + "byteOffset": 0 + } + ], + "md5sum": "d8c30599ef73a51613cc2c41394399ec" + }, + { + "dataPath": "params_shard_125.bin", + "format": "raw-shard", + "nbytes": 31223808, + "records": [ + { + "name": "model.layers.8.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 0 + }, + { + "name": "model.layers.8.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6422528 + }, + { + "name": "model.layers.8.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6429696 + }, + { + "name": "model.layers.8.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 6436864 + }, + { + "name": "model.layers.8.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 6444032 + }, + { + "name": "model.layers.8.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 21124096 + }, + { + "name": "model.layers.8.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 22959104 + }, + { + "name": "model.layers.8.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 30299136 + }, + { + "name": "model.layers.9.input_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 31216640 + } + ], + "md5sum": "3684830f59f6755905f843adcc696789" + }, + { + "dataPath": "params_shard_126.bin", + "format": "raw-shard", + "nbytes": 51380224, + "records": [ + { + "name": "model.layers.9.mlp.gate_up_proj.q_weight", + "shape": [ + 28672, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 51380224, + "byteOffset": 0 + } + ], + "md5sum": "7f028d14ee44583e6d8a160c58a6f422" + }, + { + "dataPath": "params_shard_127.bin", + "format": "raw-shard", + "nbytes": 33510400, + "records": [ + { + "name": "model.layers.9.mlp.down_proj.q_scale", + "shape": [ + 3584, + 448 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 3211264, + "byteOffset": 0 + }, + { + "name": "model.layers.9.mlp.gate_up_proj.q_scale", + "shape": [ + 28672, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 6422528, + "byteOffset": 3211264 + }, + { + "name": "model.layers.9.post_attention_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9633792 + }, + { + "name": "model.layers.9.post_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9640960 + }, + { + "name": "model.layers.9.pre_feedforward_layernorm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 9648128 + }, + { + "name": "model.layers.9.self_attn.qkv_proj.q_weight", + "shape": [ + 8192, + 448 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 14680064, + "byteOffset": 9655296 + }, + { + "name": "model.layers.9.self_attn.qkv_proj.q_scale", + "shape": [ + 8192, + 112 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 1835008, + "byteOffset": 24335360 + }, + { + "name": "model.layers.9.self_attn.o_proj.q_weight", + "shape": [ + 3584, + 512 + ], + "dtype": "uint32", + "format": "f32-to-bf16", + "nbytes": 7340032, + "byteOffset": 26170368 + } + ], + "md5sum": "22b66d4a54d6fcdc6c67c49376f3bc31" + }, + { + "dataPath": "params_shard_128.bin", + "format": "raw-shard", + "nbytes": 924672, + "records": [ + { + "name": "model.layers.9.self_attn.o_proj.q_scale", + "shape": [ + 3584, + 128 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 917504, + "byteOffset": 0 + }, + { + "name": "model.norm.weight", + "shape": [ + 3584 + ], + "dtype": "float32", + "format": "f32-to-bf16", + "nbytes": 7168, + "byteOffset": 917504 + } + ], + "md5sum": "3b5e874513a64b93d2437c958569ec45" + } + ] +} \ No newline at end of file diff --git a/params_shard_0.bin b/params_shard_0.bin new file mode 100644 index 0000000000000000000000000000000000000000..81075bb0b6f63b1b15657b3a9e9c7ba4d14e36e1 --- /dev/null +++ b/params_shard_0.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5cc9cce903165a1cba35f8aefc7d32d87b475cceba647f81a8ff019c3ecce4ab +size 458752000 diff --git a/params_shard_1.bin b/params_shard_1.bin new file mode 100644 index 0000000000000000000000000000000000000000..52691fd996d832ef03e1b3b0a8e7ef9f8719370b --- /dev/null +++ b/params_shard_1.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9ab63d132eb4c53fe500bafb5e572345256324f62921a271a51f067c8121a954 +size 57344000 diff --git a/params_shard_10.bin b/params_shard_10.bin new file mode 100644 index 0000000000000000000000000000000000000000..bab96a4149d594155f1f0c2d8c7b7a698b23d262 --- /dev/null +++ b/params_shard_10.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:606ad136bf7b80499912c09df2db7a13f3183c6e7696b69c572a2fa1b04ee7eb +size 25690112 diff --git a/params_shard_11.bin b/params_shard_11.bin new file mode 100644 index 0000000000000000000000000000000000000000..cd1329c38ebced07a9eb0a9da86dcf0216a36afc --- /dev/null +++ b/params_shard_11.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:718a2b70f09286809741c90817d3990851695ec5dc8190a66367198e7cc8b385 +size 31223808 diff --git a/params_shard_12.bin b/params_shard_12.bin new file mode 100644 index 0000000000000000000000000000000000000000..4cf93820b5cebcee683d1b748decdbe80d83c8ab --- /dev/null +++ b/params_shard_12.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e5ef2060a6e71a7ff506564abae13eda9b02c69473fe7b3f79b569c979a4bc90 +size 51380224 diff --git a/params_shard_13.bin b/params_shard_13.bin new file mode 100644 index 0000000000000000000000000000000000000000..7654e8cd3d19c055ecd26a46fd8abb2ef0108ee5 --- /dev/null +++ b/params_shard_13.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:08360b2080f574b205870a47846152095de065368a8b8a9713ce7400b31e237e +size 33510400 diff --git a/params_shard_14.bin b/params_shard_14.bin new file mode 100644 index 0000000000000000000000000000000000000000..a6e9a96c7340a1f4a0f77d02ac01483ed1d86831 --- /dev/null +++ b/params_shard_14.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f9b1b2c620f570861c0d3aa7fcecca68579991b7413824282bd59936b7351413 +size 51380224 diff --git a/params_shard_15.bin b/params_shard_15.bin new file mode 100644 index 0000000000000000000000000000000000000000..e55f46cd110fd0cc2d7ed0a561656170f3ada9b2 --- /dev/null +++ b/params_shard_15.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c650a94127db908387d95625846d896bf97671c1be5f6c1788de7cec1d3bf6c3 +size 29826048 diff --git a/params_shard_16.bin b/params_shard_16.bin new file mode 100644 index 0000000000000000000000000000000000000000..dbdca732aa87a14d40a561c53c2411fc8c8d36b1 --- /dev/null +++ b/params_shard_16.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a5b0f76a5b5fff6ebdbb70e6949cdfdc9444e2ba5b89120a9845de0f2e05a822 +size 25690112 diff --git a/params_shard_17.bin b/params_shard_17.bin new file mode 100644 index 0000000000000000000000000000000000000000..872f59aaebb61958294f4cfc41d0e903e4807c93 --- /dev/null +++ b/params_shard_17.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4f44f325c8a699cea3c00bf8373492e3c2c0d44a14b4de913d0fc263c59237ff +size 31223808 diff --git a/params_shard_18.bin b/params_shard_18.bin new file mode 100644 index 0000000000000000000000000000000000000000..f81652a6f446cae4c4794e907c52b803fbb50127 --- /dev/null +++ b/params_shard_18.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9b466077a499e5b5d3d299b44173ae5841bf7b6ace388b5cc0ef521da88e00ab +size 51380224 diff --git a/params_shard_19.bin b/params_shard_19.bin new file mode 100644 index 0000000000000000000000000000000000000000..1a3bde92f1e5794001f93ef2dd08167f8458ab86 --- /dev/null +++ b/params_shard_19.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:037486ca1e0fad342aaf2362e38d9adf409ead3e15af64022fc461202aed4f7c +size 33510400 diff --git a/params_shard_2.bin b/params_shard_2.bin new file mode 100644 index 0000000000000000000000000000000000000000..5eccd8ad2e2d6a8e87a98eb8a5d1470f679c72ef --- /dev/null +++ b/params_shard_2.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e2b330192cd3be213e49a6bcbeb76fff37cdb5d32d699f2dc6a8150d79e8f374 +size 51380224 diff --git a/params_shard_20.bin b/params_shard_20.bin new file mode 100644 index 0000000000000000000000000000000000000000..89bc7e75b853f729b8ae29081b0be2a4c974d443 --- /dev/null +++ b/params_shard_20.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b7d08989bbfd63b122dbce5167496a138717f55701217512a36598476d72454b +size 51380224 diff --git a/params_shard_21.bin b/params_shard_21.bin new file mode 100644 index 0000000000000000000000000000000000000000..dcef4ca31d1cd107ae51c32e576dc90028817551 --- /dev/null +++ b/params_shard_21.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d54a304ccb823c2ee6ba894f50e2bee1d00ec84343c4ad36183fce9f959320c5 +size 29826048 diff --git a/params_shard_22.bin b/params_shard_22.bin new file mode 100644 index 0000000000000000000000000000000000000000..4f7453c0e43bb9ac9d276a125fba971b30d988fb --- /dev/null +++ b/params_shard_22.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d1dfbdcc36a11c41609be1ee92feed61d3a273481075225168615948639bb9f9 +size 25690112 diff --git a/params_shard_23.bin b/params_shard_23.bin new file mode 100644 index 0000000000000000000000000000000000000000..329ccbd6c88c9f9b18a8dea87cc1474f302f8b9d --- /dev/null +++ b/params_shard_23.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e32ea25678de8bd3e85169e7a59b3b2401fdfcec1c3f2ab28fb20fe37b3e16fc +size 31223808 diff --git a/params_shard_24.bin b/params_shard_24.bin new file mode 100644 index 0000000000000000000000000000000000000000..01e5e4c0946105c2eeb52a6c723155b07dcb5c0f --- /dev/null +++ b/params_shard_24.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:03f33443e858c88fa619adcafbaf3a1c1d4e368386202ecf43cd969240521390 +size 51380224 diff --git a/params_shard_25.bin b/params_shard_25.bin new file mode 100644 index 0000000000000000000000000000000000000000..04e5ec9dc66f217df44def8394f2e3fc048e2cef --- /dev/null +++ b/params_shard_25.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:611526f3f1dec29010c57d3a207c36f970ae662a68a8ab3a640360ca5613cc4f +size 33510400 diff --git a/params_shard_26.bin b/params_shard_26.bin new file mode 100644 index 0000000000000000000000000000000000000000..1d2f1730c088880c87edf2ba6abe758aa7053f8b --- /dev/null +++ b/params_shard_26.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6432f50a904db46e95586437aa00f837f523767dfaa8bd66bfef469e356c378b +size 51380224 diff --git a/params_shard_27.bin b/params_shard_27.bin new file mode 100644 index 0000000000000000000000000000000000000000..4468073461cbebc7df794d1ca111934e1eb6bcd2 --- /dev/null +++ b/params_shard_27.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cdba7dd079ffe0284b2de3e8dcfea76f47ab4242ea4e954817099a0b89025df7 +size 29826048 diff --git a/params_shard_28.bin b/params_shard_28.bin new file mode 100644 index 0000000000000000000000000000000000000000..6a1dabdcb3dd0974f8ac9000b03b2f61fe056ef8 --- /dev/null +++ b/params_shard_28.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:16800f69196a7f1690c9f5e4ecd24aeb7e5fe25161aa9407e69103861f177057 +size 25690112 diff --git a/params_shard_29.bin b/params_shard_29.bin new file mode 100644 index 0000000000000000000000000000000000000000..9a51f2a28c3ed0c6f999f126e3dfc66b11ec0eca --- /dev/null +++ b/params_shard_29.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:66ce1b5b42ed92c1d0c0d83fa565aa9c7daa1b98323996982a28d3a6ff511487 +size 31223808 diff --git a/params_shard_3.bin b/params_shard_3.bin new file mode 100644 index 0000000000000000000000000000000000000000..c771faa423fc009dcda0d8c2f6ec1d86075373d9 --- /dev/null +++ b/params_shard_3.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b008e64289417cf329a27704dd615ea06eb1a2412df327f00247b5bd420b4b82 +size 28908544 diff --git a/params_shard_30.bin b/params_shard_30.bin new file mode 100644 index 0000000000000000000000000000000000000000..106aaf7fc37e09cc4170bb9aaede53e622ac99a6 --- /dev/null +++ b/params_shard_30.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:25318baaf258a2ac76f4d160f79603c05311ab99cf22efc8e114282bf7837235 +size 51380224 diff --git a/params_shard_31.bin b/params_shard_31.bin new file mode 100644 index 0000000000000000000000000000000000000000..fa4d2fe451d789c21ee20a0b0efd73b64e82b216 --- /dev/null +++ b/params_shard_31.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d2cb8a2efae44deb6dc7084fecbd370e73f9fb423d4a43e69f8afc1780327ac6 +size 33510400 diff --git a/params_shard_32.bin b/params_shard_32.bin new file mode 100644 index 0000000000000000000000000000000000000000..3fe8dd8c43ffff10c908f90c559bbb495a7cd293 --- /dev/null +++ b/params_shard_32.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0a50adf772d83580ad541fd4f6cdb5df6a2cc004edb9214f1ed61926510b969a +size 51380224 diff --git a/params_shard_33.bin b/params_shard_33.bin new file mode 100644 index 0000000000000000000000000000000000000000..1708753f47212387d588ed1577d94df134d568f4 --- /dev/null +++ b/params_shard_33.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5f44a1e3cc8d2446af87cc7f1c0ac594347855eacadc9b538f572aabe837faec +size 29826048 diff --git a/params_shard_34.bin b/params_shard_34.bin new file mode 100644 index 0000000000000000000000000000000000000000..06f890cff24c43bbe6622fcacf1a8bc687ee036f --- /dev/null +++ b/params_shard_34.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ae74fa2d5a66c0b2a2a0f7489aeea2ddf05384a7238437cb97c40ea7a5daa1ba +size 25690112 diff --git a/params_shard_35.bin b/params_shard_35.bin new file mode 100644 index 0000000000000000000000000000000000000000..4042c170bf57056dc45c45f0a1c8458d37b72620 --- /dev/null +++ b/params_shard_35.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f5da61aa6bf76099d726237a80ae181c3d21a4a5c0c35fd3a043708aaa377ec8 +size 31223808 diff --git a/params_shard_36.bin b/params_shard_36.bin new file mode 100644 index 0000000000000000000000000000000000000000..be12007dc9d2a3328b78a85cd5e8ee25815dbc57 --- /dev/null +++ b/params_shard_36.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:58942ca8da0b9e4adda18d93b4521c176707f270f4fb5c32b7139800fa2e6b24 +size 51380224 diff --git a/params_shard_37.bin b/params_shard_37.bin new file mode 100644 index 0000000000000000000000000000000000000000..3ed911176de4aff6d45a399998f419f9b62b30b0 --- /dev/null +++ b/params_shard_37.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:667f31f9e36420c7ea0425afb3bf8d39f6f27250bd773d07d02017784766b41a +size 33510400 diff --git a/params_shard_38.bin b/params_shard_38.bin new file mode 100644 index 0000000000000000000000000000000000000000..660d8641cbe3a35816c6497f602c9bda42ab7988 --- /dev/null +++ b/params_shard_38.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:951f19e9656a8010267cd5c3338931379fbdd8cc5e1726f9e6dc36834abc5808 +size 51380224 diff --git a/params_shard_39.bin b/params_shard_39.bin new file mode 100644 index 0000000000000000000000000000000000000000..5f715ab3f61d7bdaedb44c7c7f58374002540bf2 --- /dev/null +++ b/params_shard_39.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a6a59a4ee8ae09169cd3f8a0cf9ca51b969d5c1d7ecce43d027f61ef80d9ca0f +size 29826048 diff --git a/params_shard_4.bin b/params_shard_4.bin new file mode 100644 index 0000000000000000000000000000000000000000..0639b2aed37bb290f02f4768ae9524ec5a61363f --- /dev/null +++ b/params_shard_4.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:33699e1c05fb5fdc260a8ea9a8fd17f09a89afc1c319dcfb1e55f6558c682498 +size 25690112 diff --git a/params_shard_40.bin b/params_shard_40.bin new file mode 100644 index 0000000000000000000000000000000000000000..0e04bf2ec039d4a8ac478acf57c1fe48fcdfd8c3 --- /dev/null +++ b/params_shard_40.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ee544cd4f59586db6494d90d07137e8c9c5eac6844df3a0d9c8f86ebaa563549 +size 25690112 diff --git a/params_shard_41.bin b/params_shard_41.bin new file mode 100644 index 0000000000000000000000000000000000000000..8c014bd4e327ba96c1e5a096d5678898a97d34ac --- /dev/null +++ b/params_shard_41.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1b6705a7e2cde3f49235977f49b2e45acc216a8855a6a81e638778ff1e50f14e +size 31223808 diff --git a/params_shard_42.bin b/params_shard_42.bin new file mode 100644 index 0000000000000000000000000000000000000000..32db24cebf8546e7566ee2b53c420f02230a955e --- /dev/null +++ b/params_shard_42.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:69d42d89470b77b32c6eda7a4d32a85a1239366acf30963fd2fe0fe3b84c88c1 +size 51380224 diff --git a/params_shard_43.bin b/params_shard_43.bin new file mode 100644 index 0000000000000000000000000000000000000000..de0452debe989acaba7c1bb68573e7ca74e37408 --- /dev/null +++ b/params_shard_43.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:40c46f7d37ab8e7047ecb7d54f5d547d3be12dc90dd583de368b3742e7cc8486 +size 33510400 diff --git a/params_shard_44.bin b/params_shard_44.bin new file mode 100644 index 0000000000000000000000000000000000000000..699a2e19597ead45c2d189d2ace7873fcacf7b5a --- /dev/null +++ b/params_shard_44.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c7d30e776b3452111ef02bee8d88d615c78501e28b5b7013ebedbcc7de15e929 +size 51380224 diff --git a/params_shard_45.bin b/params_shard_45.bin new file mode 100644 index 0000000000000000000000000000000000000000..b21cbaad4823f6f2c0c8055ccd82f4384701d8c1 --- /dev/null +++ b/params_shard_45.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:08c3b15712c4106b382720cc658f9c8717a203df93373c0c8d4b6e0483fddc9b +size 29826048 diff --git a/params_shard_46.bin b/params_shard_46.bin new file mode 100644 index 0000000000000000000000000000000000000000..ec31957485013a38224ea02e866e65bdbbc75c5c --- /dev/null +++ b/params_shard_46.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6642535da204f18df3532eec3892bb97255814d5fc0cf68be0b5bcb2cfb0d04d +size 25690112 diff --git a/params_shard_47.bin b/params_shard_47.bin new file mode 100644 index 0000000000000000000000000000000000000000..06576fdbc50ed6099fa639b01189595838a1adea --- /dev/null +++ b/params_shard_47.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ef3dfccc588262296f2869053796592c662389e8f50d1196f15fa4dfdb5855bf +size 31223808 diff --git a/params_shard_48.bin b/params_shard_48.bin new file mode 100644 index 0000000000000000000000000000000000000000..8be55aeec8f305ace2f29359ebd63d719e679c6a --- /dev/null +++ b/params_shard_48.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:67994ca8b490bf458b00bb54fcdca5456be6285f068f0d456711d5d7494d0a24 +size 51380224 diff --git a/params_shard_49.bin b/params_shard_49.bin new file mode 100644 index 0000000000000000000000000000000000000000..38ddba8c95ab72c418e5554b2970af9ade17115f --- /dev/null +++ b/params_shard_49.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7627a3f9c58756542c201af3cca3cae5588a04bdee3c43b6e8d1b119d1cd2dc6 +size 33510400 diff --git a/params_shard_5.bin b/params_shard_5.bin new file mode 100644 index 0000000000000000000000000000000000000000..397aa87b9ec30764751bf9dab985075d7db42c28 --- /dev/null +++ b/params_shard_5.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:711486797c97763c2d4b5e20f8dddde5660faf293be0c4c2327c64ebd5564270 +size 31223808 diff --git a/params_shard_50.bin b/params_shard_50.bin new file mode 100644 index 0000000000000000000000000000000000000000..baf5e8861e39a761fed5f2b53af5ebf8be0fa00f --- /dev/null +++ b/params_shard_50.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e7b50cf9e6d8bab4250d588a0292757b265d3afa1ecffdc073c6850ff554045b +size 51380224 diff --git a/params_shard_6.bin b/params_shard_6.bin new file mode 100644 index 0000000000000000000000000000000000000000..bfb014763bf9380507b737ce73c47cd5b0e5e0f9 --- /dev/null +++ b/params_shard_6.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7970b49a525d77aeebe624caf817f1a3e5488e8df753bae67860b17fe2faa52f +size 51380224 diff --git a/params_shard_7.bin b/params_shard_7.bin new file mode 100644 index 0000000000000000000000000000000000000000..37501ea6529cc80680a293a612499e0618defe53 --- /dev/null +++ b/params_shard_7.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:67f31656cadaee7fc9b2ea72905cbccb922808f7b621d0ddb7ff47dbc6d6cdad +size 33510400 diff --git a/params_shard_8.bin b/params_shard_8.bin new file mode 100644 index 0000000000000000000000000000000000000000..02af46cd6e33350a1ae038875e67934838613a10 --- /dev/null +++ b/params_shard_8.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:95064ee0987526ece1c26013c612fa9d7a34008026d36b9b94fe8c5597020077 +size 51380224 diff --git a/params_shard_9.bin b/params_shard_9.bin new file mode 100644 index 0000000000000000000000000000000000000000..27b81bad39537569f273c2d194dc74793e2b4eb1 --- /dev/null +++ b/params_shard_9.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0ede3ad626638e2f411648b32ad5ea2ec01592ddfa90f85b37f70deb319aa940 +size 29826048