| { | |
| "metadata": { | |
| "ParamSize": 110, | |
| "ParamBytes": 10024689664.0, | |
| "BitsPerParam": 32.0 | |
| }, | |
| "records": [ | |
| { | |
| "dataPath": "params_shard_0.bin", | |
| "format": "raw-shard", | |
| "nbytes": 1048576000, | |
| "records": [ | |
| { | |
| "name": "model.embed_tokens.weight", | |
| "shape": [ | |
| 256000, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 1048576000, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "21128c642a829dff80b86de466616d9d" | |
| }, | |
| { | |
| "dataPath": "params_shard_1.bin", | |
| "format": "raw-shard", | |
| "nbytes": 67108864, | |
| "records": [ | |
| { | |
| "name": "model.layers.0.mlp.down_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 16384 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 67108864, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "e3daa1938df1d775e3448e4fdd33527e" | |
| }, | |
| { | |
| "dataPath": "params_shard_2.bin", | |
| "format": "raw-shard", | |
| "nbytes": 134217728, | |
| "records": [ | |
| { | |
| "name": "model.layers.0.mlp.gate_up_proj.weight", | |
| "shape": [ | |
| 32768, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 134217728, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "5ebf170cde97e9025d2c3473b8357c13" | |
| }, | |
| { | |
| "dataPath": "params_shard_3.bin", | |
| "format": "raw-shard", | |
| "nbytes": 67108864, | |
| "records": [ | |
| { | |
| "name": "model.layers.1.mlp.down_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 16384 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 67108864, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "b11672a46d1224e14cd2a5c8590a02ea" | |
| }, | |
| { | |
| "dataPath": "params_shard_4.bin", | |
| "format": "raw-shard", | |
| "nbytes": 134217728, | |
| "records": [ | |
| { | |
| "name": "model.layers.1.mlp.gate_up_proj.weight", | |
| "shape": [ | |
| 32768, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 134217728, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "9970e7f9c5a4be1cb1acbb075fe39a60" | |
| }, | |
| { | |
| "dataPath": "params_shard_5.bin", | |
| "format": "raw-shard", | |
| "nbytes": 29376512, | |
| "records": [ | |
| { | |
| "name": "model.layers.0.input_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 0 | |
| }, | |
| { | |
| "name": "model.layers.0.post_attention_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 4096 | |
| }, | |
| { | |
| "name": "model.layers.0.self_attn.qkv_proj.weight", | |
| "shape": [ | |
| 2560, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 10485760, | |
| "byteOffset": 8192 | |
| }, | |
| { | |
| "name": "model.layers.0.self_attn.o_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 8388608, | |
| "byteOffset": 10493952 | |
| }, | |
| { | |
| "name": "model.layers.1.input_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 18882560 | |
| }, | |
| { | |
| "name": "model.layers.1.post_attention_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 18886656 | |
| }, | |
| { | |
| "name": "model.layers.1.self_attn.qkv_proj.weight", | |
| "shape": [ | |
| 2560, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 10485760, | |
| "byteOffset": 18890752 | |
| } | |
| ], | |
| "md5sum": "690bd198a683ab8104715e9fa33e3eaf" | |
| }, | |
| { | |
| "dataPath": "params_shard_6.bin", | |
| "format": "raw-shard", | |
| "nbytes": 67108864, | |
| "records": [ | |
| { | |
| "name": "model.layers.10.mlp.down_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 16384 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 67108864, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "71c4717b75b1547985ea79fb4451908d" | |
| }, | |
| { | |
| "dataPath": "params_shard_7.bin", | |
| "format": "raw-shard", | |
| "nbytes": 134217728, | |
| "records": [ | |
| { | |
| "name": "model.layers.10.mlp.gate_up_proj.weight", | |
| "shape": [ | |
| 32768, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 134217728, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "0385b2a5124f03324c0cbbf8f8afb4d5" | |
| }, | |
| { | |
| "dataPath": "params_shard_8.bin", | |
| "format": "raw-shard", | |
| "nbytes": 67108864, | |
| "records": [ | |
| { | |
| "name": "model.layers.11.mlp.down_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 16384 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 67108864, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "9d7bbb224840a3c5f24b33f029fc8d3d" | |
| }, | |
| { | |
| "dataPath": "params_shard_9.bin", | |
| "format": "raw-shard", | |
| "nbytes": 134217728, | |
| "records": [ | |
| { | |
| "name": "model.layers.11.mlp.gate_up_proj.weight", | |
| "shape": [ | |
| 32768, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 134217728, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "babd2c92010bfd919b25fbd9daaaa8d9" | |
| }, | |
| { | |
| "dataPath": "params_shard_10.bin", | |
| "format": "raw-shard", | |
| "nbytes": 27279360, | |
| "records": [ | |
| { | |
| "name": "model.layers.1.self_attn.o_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 8388608, | |
| "byteOffset": 0 | |
| }, | |
| { | |
| "name": "model.layers.10.input_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 8388608 | |
| }, | |
| { | |
| "name": "model.layers.10.post_attention_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 8392704 | |
| }, | |
| { | |
| "name": "model.layers.10.self_attn.qkv_proj.weight", | |
| "shape": [ | |
| 2560, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 10485760, | |
| "byteOffset": 8396800 | |
| }, | |
| { | |
| "name": "model.layers.10.self_attn.o_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 8388608, | |
| "byteOffset": 18882560 | |
| }, | |
| { | |
| "name": "model.layers.11.input_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 27271168 | |
| }, | |
| { | |
| "name": "model.layers.11.post_attention_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 27275264 | |
| } | |
| ], | |
| "md5sum": "73e3bd854f6acf1969dae0923e8bb11b" | |
| }, | |
| { | |
| "dataPath": "params_shard_11.bin", | |
| "format": "raw-shard", | |
| "nbytes": 67108864, | |
| "records": [ | |
| { | |
| "name": "model.layers.12.mlp.down_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 16384 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 67108864, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "113f6d9072ff3bd46cb356305c3d6435" | |
| }, | |
| { | |
| "dataPath": "params_shard_12.bin", | |
| "format": "raw-shard", | |
| "nbytes": 134217728, | |
| "records": [ | |
| { | |
| "name": "model.layers.12.mlp.gate_up_proj.weight", | |
| "shape": [ | |
| 32768, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 134217728, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "2c179b332604d046c25fdb8fa7e8a38d" | |
| }, | |
| { | |
| "dataPath": "params_shard_13.bin", | |
| "format": "raw-shard", | |
| "nbytes": 29368320, | |
| "records": [ | |
| { | |
| "name": "model.layers.11.self_attn.qkv_proj.weight", | |
| "shape": [ | |
| 2560, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 10485760, | |
| "byteOffset": 0 | |
| }, | |
| { | |
| "name": "model.layers.11.self_attn.o_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 8388608, | |
| "byteOffset": 10485760 | |
| }, | |
| { | |
| "name": "model.layers.12.input_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 18874368 | |
| }, | |
| { | |
| "name": "model.layers.12.post_attention_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 18878464 | |
| }, | |
| { | |
| "name": "model.layers.12.self_attn.qkv_proj.weight", | |
| "shape": [ | |
| 2560, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 10485760, | |
| "byteOffset": 18882560 | |
| } | |
| ], | |
| "md5sum": "b155622e091c8a222b1f4de1814cbb49" | |
| }, | |
| { | |
| "dataPath": "params_shard_14.bin", | |
| "format": "raw-shard", | |
| "nbytes": 67108864, | |
| "records": [ | |
| { | |
| "name": "model.layers.13.mlp.down_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 16384 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 67108864, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "fb8e2063f991dd07a06fe38ea7365789" | |
| }, | |
| { | |
| "dataPath": "params_shard_15.bin", | |
| "format": "raw-shard", | |
| "nbytes": 134217728, | |
| "records": [ | |
| { | |
| "name": "model.layers.13.mlp.gate_up_proj.weight", | |
| "shape": [ | |
| 32768, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 134217728, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "9d5ceb3198e4924fccc31e3f2421ff14" | |
| }, | |
| { | |
| "dataPath": "params_shard_16.bin", | |
| "format": "raw-shard", | |
| "nbytes": 67108864, | |
| "records": [ | |
| { | |
| "name": "model.layers.14.mlp.down_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 16384 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 67108864, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "9fb12a797891bebed74f0e65eefb538b" | |
| }, | |
| { | |
| "dataPath": "params_shard_17.bin", | |
| "format": "raw-shard", | |
| "nbytes": 134217728, | |
| "records": [ | |
| { | |
| "name": "model.layers.14.mlp.gate_up_proj.weight", | |
| "shape": [ | |
| 32768, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 134217728, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "59630bc8e808b139cf23c4a4d048a81f" | |
| }, | |
| { | |
| "dataPath": "params_shard_18.bin", | |
| "format": "raw-shard", | |
| "nbytes": 27279360, | |
| "records": [ | |
| { | |
| "name": "model.layers.12.self_attn.o_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 8388608, | |
| "byteOffset": 0 | |
| }, | |
| { | |
| "name": "model.layers.13.input_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 8388608 | |
| }, | |
| { | |
| "name": "model.layers.13.post_attention_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 8392704 | |
| }, | |
| { | |
| "name": "model.layers.13.self_attn.qkv_proj.weight", | |
| "shape": [ | |
| 2560, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 10485760, | |
| "byteOffset": 8396800 | |
| }, | |
| { | |
| "name": "model.layers.13.self_attn.o_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 8388608, | |
| "byteOffset": 18882560 | |
| }, | |
| { | |
| "name": "model.layers.14.input_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 27271168 | |
| }, | |
| { | |
| "name": "model.layers.14.post_attention_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 27275264 | |
| } | |
| ], | |
| "md5sum": "17759e68d24e2eb936f9129a722522f5" | |
| }, | |
| { | |
| "dataPath": "params_shard_19.bin", | |
| "format": "raw-shard", | |
| "nbytes": 67108864, | |
| "records": [ | |
| { | |
| "name": "model.layers.15.mlp.down_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 16384 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 67108864, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "321ba4846c90abaa26d7179a514625dd" | |
| }, | |
| { | |
| "dataPath": "params_shard_20.bin", | |
| "format": "raw-shard", | |
| "nbytes": 134217728, | |
| "records": [ | |
| { | |
| "name": "model.layers.15.mlp.gate_up_proj.weight", | |
| "shape": [ | |
| 32768, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 134217728, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "f8665cbb955f5b9470e4fb4f2bebfd3a" | |
| }, | |
| { | |
| "dataPath": "params_shard_21.bin", | |
| "format": "raw-shard", | |
| "nbytes": 29368320, | |
| "records": [ | |
| { | |
| "name": "model.layers.14.self_attn.qkv_proj.weight", | |
| "shape": [ | |
| 2560, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 10485760, | |
| "byteOffset": 0 | |
| }, | |
| { | |
| "name": "model.layers.14.self_attn.o_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 8388608, | |
| "byteOffset": 10485760 | |
| }, | |
| { | |
| "name": "model.layers.15.input_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 18874368 | |
| }, | |
| { | |
| "name": "model.layers.15.post_attention_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 18878464 | |
| }, | |
| { | |
| "name": "model.layers.15.self_attn.qkv_proj.weight", | |
| "shape": [ | |
| 2560, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 10485760, | |
| "byteOffset": 18882560 | |
| } | |
| ], | |
| "md5sum": "1d686c17f8ed473437ce9dbb37f66f5a" | |
| }, | |
| { | |
| "dataPath": "params_shard_22.bin", | |
| "format": "raw-shard", | |
| "nbytes": 67108864, | |
| "records": [ | |
| { | |
| "name": "model.layers.16.mlp.down_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 16384 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 67108864, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "2218e355df7616289c17f7cf571c21e4" | |
| }, | |
| { | |
| "dataPath": "params_shard_23.bin", | |
| "format": "raw-shard", | |
| "nbytes": 134217728, | |
| "records": [ | |
| { | |
| "name": "model.layers.16.mlp.gate_up_proj.weight", | |
| "shape": [ | |
| 32768, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 134217728, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "5ecbb6372848f97fae1bdef305c623d0" | |
| }, | |
| { | |
| "dataPath": "params_shard_24.bin", | |
| "format": "raw-shard", | |
| "nbytes": 134217728, | |
| "records": [ | |
| { | |
| "name": "model.layers.17.mlp.gate_up_proj.weight", | |
| "shape": [ | |
| 32768, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 134217728, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "466ae14bce7c0f6efa9c939f132ea082" | |
| }, | |
| { | |
| "dataPath": "params_shard_25.bin", | |
| "format": "raw-shard", | |
| "nbytes": 27271168, | |
| "records": [ | |
| { | |
| "name": "model.layers.15.self_attn.o_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 8388608, | |
| "byteOffset": 0 | |
| }, | |
| { | |
| "name": "model.layers.16.input_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 8388608 | |
| }, | |
| { | |
| "name": "model.layers.16.post_attention_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 8392704 | |
| }, | |
| { | |
| "name": "model.layers.16.self_attn.qkv_proj.weight", | |
| "shape": [ | |
| 2560, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 10485760, | |
| "byteOffset": 8396800 | |
| }, | |
| { | |
| "name": "model.layers.16.self_attn.o_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 8388608, | |
| "byteOffset": 18882560 | |
| } | |
| ], | |
| "md5sum": "517de03093360d322e5b3152bd5c1ef0" | |
| }, | |
| { | |
| "dataPath": "params_shard_26.bin", | |
| "format": "raw-shard", | |
| "nbytes": 67108864, | |
| "records": [ | |
| { | |
| "name": "model.layers.2.mlp.down_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 16384 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 67108864, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "7a6c37d60d558c0a9d650a59c4fc5c68" | |
| }, | |
| { | |
| "dataPath": "params_shard_27.bin", | |
| "format": "raw-shard", | |
| "nbytes": 134217728, | |
| "records": [ | |
| { | |
| "name": "model.layers.2.mlp.gate_up_proj.weight", | |
| "shape": [ | |
| 32768, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 134217728, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "1f51d337520bbbf5a56530c60e6b4e2b" | |
| }, | |
| { | |
| "dataPath": "params_shard_28.bin", | |
| "format": "raw-shard", | |
| "nbytes": 29368320, | |
| "records": [ | |
| { | |
| "name": "model.layers.17.self_attn.qkv_proj.weight", | |
| "shape": [ | |
| 2560, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 10485760, | |
| "byteOffset": 0 | |
| }, | |
| { | |
| "name": "model.layers.17.self_attn.o_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 8388608, | |
| "byteOffset": 10485760 | |
| }, | |
| { | |
| "name": "model.layers.2.input_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 18874368 | |
| }, | |
| { | |
| "name": "model.layers.2.post_attention_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 18878464 | |
| }, | |
| { | |
| "name": "model.layers.2.self_attn.qkv_proj.weight", | |
| "shape": [ | |
| 2560, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 10485760, | |
| "byteOffset": 18882560 | |
| } | |
| ], | |
| "md5sum": "81402425b99d3caf95d4a94ea8616bf0" | |
| }, | |
| { | |
| "dataPath": "params_shard_29.bin", | |
| "format": "raw-shard", | |
| "nbytes": 67108864, | |
| "records": [ | |
| { | |
| "name": "model.layers.3.mlp.down_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 16384 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 67108864, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "c3300e44f9c90f83867efd3d6aef1d8d" | |
| }, | |
| { | |
| "dataPath": "params_shard_30.bin", | |
| "format": "raw-shard", | |
| "nbytes": 134217728, | |
| "records": [ | |
| { | |
| "name": "model.layers.3.mlp.gate_up_proj.weight", | |
| "shape": [ | |
| 32768, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 134217728, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "538c26b83e0c2ab4bdf48cd10df6bd53" | |
| }, | |
| { | |
| "dataPath": "params_shard_31.bin", | |
| "format": "raw-shard", | |
| "nbytes": 67108864, | |
| "records": [ | |
| { | |
| "name": "model.layers.4.mlp.down_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 16384 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 67108864, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "88501c646188da831d14676a256ac84f" | |
| }, | |
| { | |
| "dataPath": "params_shard_32.bin", | |
| "format": "raw-shard", | |
| "nbytes": 134217728, | |
| "records": [ | |
| { | |
| "name": "model.layers.4.mlp.gate_up_proj.weight", | |
| "shape": [ | |
| 32768, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 134217728, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "074d0db9b1d459438705b37b6670c06b" | |
| }, | |
| { | |
| "dataPath": "params_shard_33.bin", | |
| "format": "raw-shard", | |
| "nbytes": 27279360, | |
| "records": [ | |
| { | |
| "name": "model.layers.2.self_attn.o_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 8388608, | |
| "byteOffset": 0 | |
| }, | |
| { | |
| "name": "model.layers.3.input_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 8388608 | |
| }, | |
| { | |
| "name": "model.layers.3.post_attention_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 8392704 | |
| }, | |
| { | |
| "name": "model.layers.3.self_attn.qkv_proj.weight", | |
| "shape": [ | |
| 2560, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 10485760, | |
| "byteOffset": 8396800 | |
| }, | |
| { | |
| "name": "model.layers.3.self_attn.o_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 8388608, | |
| "byteOffset": 18882560 | |
| }, | |
| { | |
| "name": "model.layers.4.input_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 27271168 | |
| }, | |
| { | |
| "name": "model.layers.4.post_attention_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 27275264 | |
| } | |
| ], | |
| "md5sum": "895ac1ceb3e21746d03374268ff4244e" | |
| }, | |
| { | |
| "dataPath": "params_shard_34.bin", | |
| "format": "raw-shard", | |
| "nbytes": 67108864, | |
| "records": [ | |
| { | |
| "name": "model.layers.5.mlp.down_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 16384 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 67108864, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "291f8fffa0d9b51370bf90286b0118fc" | |
| }, | |
| { | |
| "dataPath": "params_shard_35.bin", | |
| "format": "raw-shard", | |
| "nbytes": 134217728, | |
| "records": [ | |
| { | |
| "name": "model.layers.5.mlp.gate_up_proj.weight", | |
| "shape": [ | |
| 32768, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 134217728, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "8469971b7c6bca2a6ae1eb38c269a898" | |
| }, | |
| { | |
| "dataPath": "params_shard_36.bin", | |
| "format": "raw-shard", | |
| "nbytes": 29368320, | |
| "records": [ | |
| { | |
| "name": "model.layers.4.self_attn.qkv_proj.weight", | |
| "shape": [ | |
| 2560, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 10485760, | |
| "byteOffset": 0 | |
| }, | |
| { | |
| "name": "model.layers.4.self_attn.o_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 8388608, | |
| "byteOffset": 10485760 | |
| }, | |
| { | |
| "name": "model.layers.5.input_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 18874368 | |
| }, | |
| { | |
| "name": "model.layers.5.post_attention_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 18878464 | |
| }, | |
| { | |
| "name": "model.layers.5.self_attn.qkv_proj.weight", | |
| "shape": [ | |
| 2560, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 10485760, | |
| "byteOffset": 18882560 | |
| } | |
| ], | |
| "md5sum": "a03a45c1809bebe8da1d4c28dd8af86f" | |
| }, | |
| { | |
| "dataPath": "params_shard_37.bin", | |
| "format": "raw-shard", | |
| "nbytes": 67108864, | |
| "records": [ | |
| { | |
| "name": "model.layers.6.mlp.down_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 16384 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 67108864, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "ede3975fcb55174d920a3f6c5331cbdd" | |
| }, | |
| { | |
| "dataPath": "params_shard_38.bin", | |
| "format": "raw-shard", | |
| "nbytes": 134217728, | |
| "records": [ | |
| { | |
| "name": "model.layers.6.mlp.gate_up_proj.weight", | |
| "shape": [ | |
| 32768, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 134217728, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "81113b1b64021ec2a7d680376b1340c4" | |
| }, | |
| { | |
| "dataPath": "params_shard_39.bin", | |
| "format": "raw-shard", | |
| "nbytes": 67108864, | |
| "records": [ | |
| { | |
| "name": "model.layers.7.mlp.down_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 16384 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 67108864, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "6b65ffb4a139d0fc34ee600b71a15984" | |
| }, | |
| { | |
| "dataPath": "params_shard_40.bin", | |
| "format": "raw-shard", | |
| "nbytes": 134217728, | |
| "records": [ | |
| { | |
| "name": "model.layers.7.mlp.gate_up_proj.weight", | |
| "shape": [ | |
| 32768, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 134217728, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "7907574e8d6fe13a58bf8b4f2cdf5540" | |
| }, | |
| { | |
| "dataPath": "params_shard_41.bin", | |
| "format": "raw-shard", | |
| "nbytes": 27279360, | |
| "records": [ | |
| { | |
| "name": "model.layers.5.self_attn.o_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 8388608, | |
| "byteOffset": 0 | |
| }, | |
| { | |
| "name": "model.layers.6.input_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 8388608 | |
| }, | |
| { | |
| "name": "model.layers.6.post_attention_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 8392704 | |
| }, | |
| { | |
| "name": "model.layers.6.self_attn.qkv_proj.weight", | |
| "shape": [ | |
| 2560, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 10485760, | |
| "byteOffset": 8396800 | |
| }, | |
| { | |
| "name": "model.layers.6.self_attn.o_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 8388608, | |
| "byteOffset": 18882560 | |
| }, | |
| { | |
| "name": "model.layers.7.input_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 27271168 | |
| }, | |
| { | |
| "name": "model.layers.7.post_attention_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 27275264 | |
| } | |
| ], | |
| "md5sum": "c4d742181e56a57f389cdba6fe1115d4" | |
| }, | |
| { | |
| "dataPath": "params_shard_42.bin", | |
| "format": "raw-shard", | |
| "nbytes": 67108864, | |
| "records": [ | |
| { | |
| "name": "model.layers.8.mlp.down_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 16384 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 67108864, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "2b3708ef965db08042cfbb26753a31a0" | |
| }, | |
| { | |
| "dataPath": "params_shard_43.bin", | |
| "format": "raw-shard", | |
| "nbytes": 134217728, | |
| "records": [ | |
| { | |
| "name": "model.layers.8.mlp.gate_up_proj.weight", | |
| "shape": [ | |
| 32768, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 134217728, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "396069673d0ae19fb75be4bbc0752a25" | |
| }, | |
| { | |
| "dataPath": "params_shard_44.bin", | |
| "format": "raw-shard", | |
| "nbytes": 29368320, | |
| "records": [ | |
| { | |
| "name": "model.layers.7.self_attn.qkv_proj.weight", | |
| "shape": [ | |
| 2560, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 10485760, | |
| "byteOffset": 0 | |
| }, | |
| { | |
| "name": "model.layers.7.self_attn.o_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 8388608, | |
| "byteOffset": 10485760 | |
| }, | |
| { | |
| "name": "model.layers.8.input_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 18874368 | |
| }, | |
| { | |
| "name": "model.layers.8.post_attention_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 18878464 | |
| }, | |
| { | |
| "name": "model.layers.8.self_attn.qkv_proj.weight", | |
| "shape": [ | |
| 2560, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 10485760, | |
| "byteOffset": 18882560 | |
| } | |
| ], | |
| "md5sum": "a02b4f22cd694b60d34e8889f76be3cb" | |
| }, | |
| { | |
| "dataPath": "params_shard_45.bin", | |
| "format": "raw-shard", | |
| "nbytes": 67108864, | |
| "records": [ | |
| { | |
| "name": "model.layers.9.mlp.down_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 16384 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 67108864, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "ab38a774e2e25e749bf80c4df50d8253" | |
| }, | |
| { | |
| "dataPath": "params_shard_46.bin", | |
| "format": "raw-shard", | |
| "nbytes": 134217728, | |
| "records": [ | |
| { | |
| "name": "model.layers.9.mlp.gate_up_proj.weight", | |
| "shape": [ | |
| 32768, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 134217728, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "2ee6fc965e7edf0b4354af2d7bee5001" | |
| }, | |
| { | |
| "dataPath": "params_shard_47.bin", | |
| "format": "raw-shard", | |
| "nbytes": 67108864, | |
| "records": [ | |
| { | |
| "name": "model.layers.17.mlp.down_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 16384 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 67108864, | |
| "byteOffset": 0 | |
| } | |
| ], | |
| "md5sum": "e04f27c840be79a43646c85d3bb6e1e6" | |
| }, | |
| { | |
| "dataPath": "params_shard_48.bin", | |
| "format": "raw-shard", | |
| "nbytes": 27283456, | |
| "records": [ | |
| { | |
| "name": "model.layers.8.self_attn.o_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 8388608, | |
| "byteOffset": 0 | |
| }, | |
| { | |
| "name": "model.layers.9.input_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 8388608 | |
| }, | |
| { | |
| "name": "model.layers.9.post_attention_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 8392704 | |
| }, | |
| { | |
| "name": "model.layers.9.self_attn.qkv_proj.weight", | |
| "shape": [ | |
| 2560, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 10485760, | |
| "byteOffset": 8396800 | |
| }, | |
| { | |
| "name": "model.layers.9.self_attn.o_proj.weight", | |
| "shape": [ | |
| 2048, | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 8388608, | |
| "byteOffset": 18882560 | |
| }, | |
| { | |
| "name": "model.layers.17.input_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 27271168 | |
| }, | |
| { | |
| "name": "model.layers.17.post_attention_layernorm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 27275264 | |
| }, | |
| { | |
| "name": "model.norm.weight", | |
| "shape": [ | |
| 2048 | |
| ], | |
| "dtype": "float32", | |
| "format": "f32-to-bf16", | |
| "nbytes": 4096, | |
| "byteOffset": 27279360 | |
| } | |
| ], | |
| "md5sum": "a63f5ec6dc518cc6bd4b80ac101aa028" | |
| } | |
| ] | |
| } |