diff --git "a/image-large/imageEncoder/manifest.json" "b/image-large/imageEncoder/manifest.json" new file mode 100644--- /dev/null +++ "b/image-large/imageEncoder/manifest.json" @@ -0,0 +1,8406 @@ +{ + "schema": "trellis-webgpu-pack", + "version": 1, + "stage": "imageEncoder", + "upstreamRevision": "e4c89a4e05589de9b3e188688a303d0f3c04d0f3", + "shards": [ + { + "file": "stage-000.bin", + "bytes": 66911232, + "sha256": "b0cd6c96e03ddb04012451d17f9f205d3475ba9ebb86aef05c0a69144f8b79a0" + }, + { + "file": "stage-001.bin", + "bytes": 66691072, + "sha256": "53ad5c083824c97100cb87f2b8b89f17f8a01e2b5602cfe6712cabdac69ed43a" + }, + { + "file": "stage-002.bin", + "bytes": 65630208, + "sha256": "a22b8a11c24a41949fc3a584b2f5cf5b8b3e274e19694915cb6869a1e83983a5" + }, + { + "file": "stage-003.bin", + "bytes": 66234368, + "sha256": "bc4b8433aaedbf356896ba19adfc50b124d220348c50934c5c72b2cca71bfd70" + }, + { + "file": "stage-004.bin", + "bytes": 65050624, + "sha256": "729a6b912725037dbb818277efabbd2c2ee326c97c051287b47d8699251e00b3" + }, + { + "file": "stage-005.bin", + "bytes": 65630208, + "sha256": "ee5f7b513c7a11ec970707f76291f281372aa101a77d98526a0f0c888ef911eb" + }, + { + "file": "stage-006.bin", + "bytes": 66234368, + "sha256": "2ae93178c535a3b29fdd28ac91b0cee8e9483f561b68da20eb02c309bcbf23bb" + }, + { + "file": "stage-007.bin", + "bytes": 34463744, + "sha256": "2d40d3bb21a58145168e6b08000ccfa82b045e8a0263d8eebf465cbf6776a150" + } + ], + "tensors": { + "embeddings.cls_token": { + "shape": [ + 1, + 1, + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 0, + "bytes": 2048 + } + }, + "embeddings.mask_token": { + "shape": [ + 1, + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 2048, + "bytes": 2048 + } + }, + "embeddings.patch_embeddings.projection.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 4096, + "bytes": 2048 + } + }, + "embeddings.patch_embeddings.projection.weight": { + "shape": [ + 1024, + 3, + 14, + 14 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 588, + "padded_rows": 1024, + "padded_columns": 640, + "base_data": { + "shard": 0, + "offset": 6144, + "bytes": 655360 + }, + "base_scales": { + "shard": 0, + "offset": 661504, + "bytes": 20480 + }, + "residual_data": { + "shard": 0, + "offset": 681984, + "bytes": 327680 + }, + "residual_scales": { + "shard": 0, + "offset": 1009664, + "bytes": 20480 + } + }, + "embeddings.position_embeddings": { + "shape": [ + 1, + 1370, + 1024 + ], + "encoding": "symmetric-quantized", + "bits": 8, + "group_size": 64, + "output_tile": 16, + "rows": 1, + "columns": 1402880, + "padded_rows": 16, + "padded_columns": 1402880, + "data": { + "shard": 0, + "offset": 1030144, + "bytes": 22446080 + }, + "scales": { + "shard": 0, + "offset": 23476224, + "bytes": 701440 + } + }, + "embeddings.register_tokens": { + "shape": [ + 1, + 4, + 1024 + ], + "encoding": "symmetric-quantized", + "bits": 8, + "group_size": 64, + "output_tile": 16, + "rows": 1, + "columns": 4096, + "padded_rows": 16, + "padded_columns": 4096, + "data": { + "shard": 0, + "offset": 24177664, + "bytes": 65536 + }, + "scales": { + "shard": 0, + "offset": 24243200, + "bytes": 2048 + } + }, + "encoder.layer.0.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 24245248, + "bytes": 2048 + } + }, + "encoder.layer.0.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 0, + "offset": 24247296, + "bytes": 1048576 + }, + "base_scales": { + "shard": 0, + "offset": 25295872, + "bytes": 32768 + }, + "residual_data": { + "shard": 0, + "offset": 25328640, + "bytes": 524288 + }, + "residual_scales": { + "shard": 0, + "offset": 25852928, + "bytes": 32768 + } + }, + "encoder.layer.0.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 25885696, + "bytes": 2048 + } + }, + "encoder.layer.0.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 0, + "offset": 25887744, + "bytes": 1048576 + }, + "base_scales": { + "shard": 0, + "offset": 26936320, + "bytes": 32768 + }, + "residual_data": { + "shard": 0, + "offset": 26969088, + "bytes": 524288 + }, + "residual_scales": { + "shard": 0, + "offset": 27493376, + "bytes": 32768 + } + }, + "encoder.layer.0.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 27526144, + "bytes": 2048 + } + }, + "encoder.layer.0.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 0, + "offset": 27528192, + "bytes": 1048576 + }, + "base_scales": { + "shard": 0, + "offset": 28576768, + "bytes": 32768 + }, + "residual_data": { + "shard": 0, + "offset": 28609536, + "bytes": 524288 + }, + "residual_scales": { + "shard": 0, + "offset": 29133824, + "bytes": 32768 + } + }, + "encoder.layer.0.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 29166592, + "bytes": 2048 + } + }, + "encoder.layer.0.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 0, + "offset": 29168640, + "bytes": 1048576 + }, + "base_scales": { + "shard": 0, + "offset": 30217216, + "bytes": 32768 + }, + "residual_data": { + "shard": 0, + "offset": 30249984, + "bytes": 524288 + }, + "residual_scales": { + "shard": 0, + "offset": 30774272, + "bytes": 32768 + } + }, + "encoder.layer.0.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 30807040, + "bytes": 2048 + } + }, + "encoder.layer.0.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 30809088, + "bytes": 2048 + } + }, + "encoder.layer.0.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 30811136, + "bytes": 8192 + } + }, + "encoder.layer.0.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 0, + "offset": 30819328, + "bytes": 4194304 + }, + "base_scales": { + "shard": 0, + "offset": 35013632, + "bytes": 131072 + }, + "residual_data": { + "shard": 0, + "offset": 35144704, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 0, + "offset": 37241856, + "bytes": 131072 + } + }, + "encoder.layer.0.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 37372928, + "bytes": 2048 + } + }, + "encoder.layer.0.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 0, + "offset": 37374976, + "bytes": 4194304 + }, + "base_scales": { + "shard": 0, + "offset": 41569280, + "bytes": 131072 + }, + "residual_data": { + "shard": 0, + "offset": 41700352, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 0, + "offset": 43797504, + "bytes": 131072 + } + }, + "encoder.layer.0.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 43928576, + "bytes": 2048 + } + }, + "encoder.layer.0.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 43930624, + "bytes": 2048 + } + }, + "encoder.layer.0.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 43932672, + "bytes": 2048 + } + }, + "encoder.layer.0.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 43934720, + "bytes": 2048 + } + }, + "encoder.layer.1.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 43936768, + "bytes": 2048 + } + }, + "encoder.layer.1.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 0, + "offset": 43938816, + "bytes": 1048576 + }, + "base_scales": { + "shard": 0, + "offset": 44987392, + "bytes": 32768 + }, + "residual_data": { + "shard": 0, + "offset": 45020160, + "bytes": 524288 + }, + "residual_scales": { + "shard": 0, + "offset": 45544448, + "bytes": 32768 + } + }, + "encoder.layer.1.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 45577216, + "bytes": 2048 + } + }, + "encoder.layer.1.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 0, + "offset": 45579264, + "bytes": 1048576 + }, + "base_scales": { + "shard": 0, + "offset": 46627840, + "bytes": 32768 + }, + "residual_data": { + "shard": 0, + "offset": 46660608, + "bytes": 524288 + }, + "residual_scales": { + "shard": 0, + "offset": 47184896, + "bytes": 32768 + } + }, + "encoder.layer.1.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 47217664, + "bytes": 2048 + } + }, + "encoder.layer.1.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 0, + "offset": 47219712, + "bytes": 1048576 + }, + "base_scales": { + "shard": 0, + "offset": 48268288, + "bytes": 32768 + }, + "residual_data": { + "shard": 0, + "offset": 48301056, + "bytes": 524288 + }, + "residual_scales": { + "shard": 0, + "offset": 48825344, + "bytes": 32768 + } + }, + "encoder.layer.1.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 48858112, + "bytes": 2048 + } + }, + "encoder.layer.1.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 0, + "offset": 48860160, + "bytes": 1048576 + }, + "base_scales": { + "shard": 0, + "offset": 49908736, + "bytes": 32768 + }, + "residual_data": { + "shard": 0, + "offset": 49941504, + "bytes": 524288 + }, + "residual_scales": { + "shard": 0, + "offset": 50465792, + "bytes": 32768 + } + }, + "encoder.layer.1.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 50498560, + "bytes": 2048 + } + }, + "encoder.layer.1.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 50500608, + "bytes": 2048 + } + }, + "encoder.layer.1.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 50502656, + "bytes": 8192 + } + }, + "encoder.layer.1.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 0, + "offset": 50510848, + "bytes": 4194304 + }, + "base_scales": { + "shard": 0, + "offset": 54705152, + "bytes": 131072 + }, + "residual_data": { + "shard": 0, + "offset": 54836224, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 0, + "offset": 56933376, + "bytes": 131072 + } + }, + "encoder.layer.1.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 57064448, + "bytes": 2048 + } + }, + "encoder.layer.1.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 0, + "offset": 57066496, + "bytes": 4194304 + }, + "base_scales": { + "shard": 0, + "offset": 61260800, + "bytes": 131072 + }, + "residual_data": { + "shard": 0, + "offset": 61391872, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 0, + "offset": 63489024, + "bytes": 131072 + } + }, + "encoder.layer.1.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 63620096, + "bytes": 2048 + } + }, + "encoder.layer.1.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 63622144, + "bytes": 2048 + } + }, + "encoder.layer.1.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 63624192, + "bytes": 2048 + } + }, + "encoder.layer.1.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 63626240, + "bytes": 2048 + } + }, + "encoder.layer.10.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 63628288, + "bytes": 2048 + } + }, + "encoder.layer.10.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 0, + "offset": 63630336, + "bytes": 1048576 + }, + "base_scales": { + "shard": 0, + "offset": 64678912, + "bytes": 32768 + }, + "residual_data": { + "shard": 0, + "offset": 64711680, + "bytes": 524288 + }, + "residual_scales": { + "shard": 0, + "offset": 65235968, + "bytes": 32768 + } + }, + "encoder.layer.10.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 65268736, + "bytes": 2048 + } + }, + "encoder.layer.10.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 0, + "offset": 65270784, + "bytes": 1048576 + }, + "base_scales": { + "shard": 0, + "offset": 66319360, + "bytes": 32768 + }, + "residual_data": { + "shard": 0, + "offset": 66352128, + "bytes": 524288 + }, + "residual_scales": { + "shard": 0, + "offset": 66876416, + "bytes": 32768 + } + }, + "encoder.layer.10.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 0, + "offset": 66909184, + "bytes": 2048 + } + }, + "encoder.layer.10.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 1, + "offset": 0, + "bytes": 1048576 + }, + "base_scales": { + "shard": 1, + "offset": 1048576, + "bytes": 32768 + }, + "residual_data": { + "shard": 1, + "offset": 1081344, + "bytes": 524288 + }, + "residual_scales": { + "shard": 1, + "offset": 1605632, + "bytes": 32768 + } + }, + "encoder.layer.10.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 1638400, + "bytes": 2048 + } + }, + "encoder.layer.10.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 1, + "offset": 1640448, + "bytes": 1048576 + }, + "base_scales": { + "shard": 1, + "offset": 2689024, + "bytes": 32768 + }, + "residual_data": { + "shard": 1, + "offset": 2721792, + "bytes": 524288 + }, + "residual_scales": { + "shard": 1, + "offset": 3246080, + "bytes": 32768 + } + }, + "encoder.layer.10.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 3278848, + "bytes": 2048 + } + }, + "encoder.layer.10.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 3280896, + "bytes": 2048 + } + }, + "encoder.layer.10.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 3282944, + "bytes": 8192 + } + }, + "encoder.layer.10.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 1, + "offset": 3291136, + "bytes": 4194304 + }, + "base_scales": { + "shard": 1, + "offset": 7485440, + "bytes": 131072 + }, + "residual_data": { + "shard": 1, + "offset": 7616512, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 1, + "offset": 9713664, + "bytes": 131072 + } + }, + "encoder.layer.10.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 9844736, + "bytes": 2048 + } + }, + "encoder.layer.10.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 1, + "offset": 9846784, + "bytes": 4194304 + }, + "base_scales": { + "shard": 1, + "offset": 14041088, + "bytes": 131072 + }, + "residual_data": { + "shard": 1, + "offset": 14172160, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 1, + "offset": 16269312, + "bytes": 131072 + } + }, + "encoder.layer.10.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 16400384, + "bytes": 2048 + } + }, + "encoder.layer.10.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 16402432, + "bytes": 2048 + } + }, + "encoder.layer.10.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 16404480, + "bytes": 2048 + } + }, + "encoder.layer.10.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 16406528, + "bytes": 2048 + } + }, + "encoder.layer.11.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 16408576, + "bytes": 2048 + } + }, + "encoder.layer.11.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 1, + "offset": 16410624, + "bytes": 1048576 + }, + "base_scales": { + "shard": 1, + "offset": 17459200, + "bytes": 32768 + }, + "residual_data": { + "shard": 1, + "offset": 17491968, + "bytes": 524288 + }, + "residual_scales": { + "shard": 1, + "offset": 18016256, + "bytes": 32768 + } + }, + "encoder.layer.11.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 18049024, + "bytes": 2048 + } + }, + "encoder.layer.11.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 1, + "offset": 18051072, + "bytes": 1048576 + }, + "base_scales": { + "shard": 1, + "offset": 19099648, + "bytes": 32768 + }, + "residual_data": { + "shard": 1, + "offset": 19132416, + "bytes": 524288 + }, + "residual_scales": { + "shard": 1, + "offset": 19656704, + "bytes": 32768 + } + }, + "encoder.layer.11.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 19689472, + "bytes": 2048 + } + }, + "encoder.layer.11.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 1, + "offset": 19691520, + "bytes": 1048576 + }, + "base_scales": { + "shard": 1, + "offset": 20740096, + "bytes": 32768 + }, + "residual_data": { + "shard": 1, + "offset": 20772864, + "bytes": 524288 + }, + "residual_scales": { + "shard": 1, + "offset": 21297152, + "bytes": 32768 + } + }, + "encoder.layer.11.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 21329920, + "bytes": 2048 + } + }, + "encoder.layer.11.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 1, + "offset": 21331968, + "bytes": 1048576 + }, + "base_scales": { + "shard": 1, + "offset": 22380544, + "bytes": 32768 + }, + "residual_data": { + "shard": 1, + "offset": 22413312, + "bytes": 524288 + }, + "residual_scales": { + "shard": 1, + "offset": 22937600, + "bytes": 32768 + } + }, + "encoder.layer.11.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 22970368, + "bytes": 2048 + } + }, + "encoder.layer.11.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 22972416, + "bytes": 2048 + } + }, + "encoder.layer.11.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 22974464, + "bytes": 8192 + } + }, + "encoder.layer.11.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 1, + "offset": 22982656, + "bytes": 4194304 + }, + "base_scales": { + "shard": 1, + "offset": 27176960, + "bytes": 131072 + }, + "residual_data": { + "shard": 1, + "offset": 27308032, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 1, + "offset": 29405184, + "bytes": 131072 + } + }, + "encoder.layer.11.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 29536256, + "bytes": 2048 + } + }, + "encoder.layer.11.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 1, + "offset": 29538304, + "bytes": 4194304 + }, + "base_scales": { + "shard": 1, + "offset": 33732608, + "bytes": 131072 + }, + "residual_data": { + "shard": 1, + "offset": 33863680, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 1, + "offset": 35960832, + "bytes": 131072 + } + }, + "encoder.layer.11.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 36091904, + "bytes": 2048 + } + }, + "encoder.layer.11.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 36093952, + "bytes": 2048 + } + }, + "encoder.layer.11.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 36096000, + "bytes": 2048 + } + }, + "encoder.layer.11.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 36098048, + "bytes": 2048 + } + }, + "encoder.layer.12.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 36100096, + "bytes": 2048 + } + }, + "encoder.layer.12.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 1, + "offset": 36102144, + "bytes": 1048576 + }, + "base_scales": { + "shard": 1, + "offset": 37150720, + "bytes": 32768 + }, + "residual_data": { + "shard": 1, + "offset": 37183488, + "bytes": 524288 + }, + "residual_scales": { + "shard": 1, + "offset": 37707776, + "bytes": 32768 + } + }, + "encoder.layer.12.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 37740544, + "bytes": 2048 + } + }, + "encoder.layer.12.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 1, + "offset": 37742592, + "bytes": 1048576 + }, + "base_scales": { + "shard": 1, + "offset": 38791168, + "bytes": 32768 + }, + "residual_data": { + "shard": 1, + "offset": 38823936, + "bytes": 524288 + }, + "residual_scales": { + "shard": 1, + "offset": 39348224, + "bytes": 32768 + } + }, + "encoder.layer.12.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 39380992, + "bytes": 2048 + } + }, + "encoder.layer.12.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 1, + "offset": 39383040, + "bytes": 1048576 + }, + "base_scales": { + "shard": 1, + "offset": 40431616, + "bytes": 32768 + }, + "residual_data": { + "shard": 1, + "offset": 40464384, + "bytes": 524288 + }, + "residual_scales": { + "shard": 1, + "offset": 40988672, + "bytes": 32768 + } + }, + "encoder.layer.12.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 41021440, + "bytes": 2048 + } + }, + "encoder.layer.12.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 1, + "offset": 41023488, + "bytes": 1048576 + }, + "base_scales": { + "shard": 1, + "offset": 42072064, + "bytes": 32768 + }, + "residual_data": { + "shard": 1, + "offset": 42104832, + "bytes": 524288 + }, + "residual_scales": { + "shard": 1, + "offset": 42629120, + "bytes": 32768 + } + }, + "encoder.layer.12.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 42661888, + "bytes": 2048 + } + }, + "encoder.layer.12.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 42663936, + "bytes": 2048 + } + }, + "encoder.layer.12.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 42665984, + "bytes": 8192 + } + }, + "encoder.layer.12.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 1, + "offset": 42674176, + "bytes": 4194304 + }, + "base_scales": { + "shard": 1, + "offset": 46868480, + "bytes": 131072 + }, + "residual_data": { + "shard": 1, + "offset": 46999552, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 1, + "offset": 49096704, + "bytes": 131072 + } + }, + "encoder.layer.12.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 49227776, + "bytes": 2048 + } + }, + "encoder.layer.12.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 1, + "offset": 49229824, + "bytes": 4194304 + }, + "base_scales": { + "shard": 1, + "offset": 53424128, + "bytes": 131072 + }, + "residual_data": { + "shard": 1, + "offset": 53555200, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 1, + "offset": 55652352, + "bytes": 131072 + } + }, + "encoder.layer.12.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 55783424, + "bytes": 2048 + } + }, + "encoder.layer.12.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 55785472, + "bytes": 2048 + } + }, + "encoder.layer.12.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 55787520, + "bytes": 2048 + } + }, + "encoder.layer.12.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 55789568, + "bytes": 2048 + } + }, + "encoder.layer.13.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 55791616, + "bytes": 2048 + } + }, + "encoder.layer.13.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 1, + "offset": 55793664, + "bytes": 1048576 + }, + "base_scales": { + "shard": 1, + "offset": 56842240, + "bytes": 32768 + }, + "residual_data": { + "shard": 1, + "offset": 56875008, + "bytes": 524288 + }, + "residual_scales": { + "shard": 1, + "offset": 57399296, + "bytes": 32768 + } + }, + "encoder.layer.13.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 57432064, + "bytes": 2048 + } + }, + "encoder.layer.13.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 1, + "offset": 57434112, + "bytes": 1048576 + }, + "base_scales": { + "shard": 1, + "offset": 58482688, + "bytes": 32768 + }, + "residual_data": { + "shard": 1, + "offset": 58515456, + "bytes": 524288 + }, + "residual_scales": { + "shard": 1, + "offset": 59039744, + "bytes": 32768 + } + }, + "encoder.layer.13.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 59072512, + "bytes": 2048 + } + }, + "encoder.layer.13.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 1, + "offset": 59074560, + "bytes": 1048576 + }, + "base_scales": { + "shard": 1, + "offset": 60123136, + "bytes": 32768 + }, + "residual_data": { + "shard": 1, + "offset": 60155904, + "bytes": 524288 + }, + "residual_scales": { + "shard": 1, + "offset": 60680192, + "bytes": 32768 + } + }, + "encoder.layer.13.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 60712960, + "bytes": 2048 + } + }, + "encoder.layer.13.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 1, + "offset": 60715008, + "bytes": 1048576 + }, + "base_scales": { + "shard": 1, + "offset": 61763584, + "bytes": 32768 + }, + "residual_data": { + "shard": 1, + "offset": 61796352, + "bytes": 524288 + }, + "residual_scales": { + "shard": 1, + "offset": 62320640, + "bytes": 32768 + } + }, + "encoder.layer.13.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 62353408, + "bytes": 2048 + } + }, + "encoder.layer.13.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 62355456, + "bytes": 2048 + } + }, + "encoder.layer.13.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 1, + "offset": 62357504, + "bytes": 8192 + } + }, + "encoder.layer.13.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 1, + "offset": 62365696, + "bytes": 4194304 + }, + "base_scales": { + "shard": 1, + "offset": 66560000, + "bytes": 131072 + }, + "residual_data": { + "shard": 2, + "offset": 0, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 2, + "offset": 2097152, + "bytes": 131072 + } + }, + "encoder.layer.13.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 2228224, + "bytes": 2048 + } + }, + "encoder.layer.13.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 2, + "offset": 2230272, + "bytes": 4194304 + }, + "base_scales": { + "shard": 2, + "offset": 6424576, + "bytes": 131072 + }, + "residual_data": { + "shard": 2, + "offset": 6555648, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 2, + "offset": 8652800, + "bytes": 131072 + } + }, + "encoder.layer.13.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 8783872, + "bytes": 2048 + } + }, + "encoder.layer.13.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 8785920, + "bytes": 2048 + } + }, + "encoder.layer.13.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 8787968, + "bytes": 2048 + } + }, + "encoder.layer.13.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 8790016, + "bytes": 2048 + } + }, + "encoder.layer.14.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 8792064, + "bytes": 2048 + } + }, + "encoder.layer.14.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 2, + "offset": 8794112, + "bytes": 1048576 + }, + "base_scales": { + "shard": 2, + "offset": 9842688, + "bytes": 32768 + }, + "residual_data": { + "shard": 2, + "offset": 9875456, + "bytes": 524288 + }, + "residual_scales": { + "shard": 2, + "offset": 10399744, + "bytes": 32768 + } + }, + "encoder.layer.14.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 10432512, + "bytes": 2048 + } + }, + "encoder.layer.14.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 2, + "offset": 10434560, + "bytes": 1048576 + }, + "base_scales": { + "shard": 2, + "offset": 11483136, + "bytes": 32768 + }, + "residual_data": { + "shard": 2, + "offset": 11515904, + "bytes": 524288 + }, + "residual_scales": { + "shard": 2, + "offset": 12040192, + "bytes": 32768 + } + }, + "encoder.layer.14.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 12072960, + "bytes": 2048 + } + }, + "encoder.layer.14.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 2, + "offset": 12075008, + "bytes": 1048576 + }, + "base_scales": { + "shard": 2, + "offset": 13123584, + "bytes": 32768 + }, + "residual_data": { + "shard": 2, + "offset": 13156352, + "bytes": 524288 + }, + "residual_scales": { + "shard": 2, + "offset": 13680640, + "bytes": 32768 + } + }, + "encoder.layer.14.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 13713408, + "bytes": 2048 + } + }, + "encoder.layer.14.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 2, + "offset": 13715456, + "bytes": 1048576 + }, + "base_scales": { + "shard": 2, + "offset": 14764032, + "bytes": 32768 + }, + "residual_data": { + "shard": 2, + "offset": 14796800, + "bytes": 524288 + }, + "residual_scales": { + "shard": 2, + "offset": 15321088, + "bytes": 32768 + } + }, + "encoder.layer.14.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 15353856, + "bytes": 2048 + } + }, + "encoder.layer.14.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 15355904, + "bytes": 2048 + } + }, + "encoder.layer.14.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 15357952, + "bytes": 8192 + } + }, + "encoder.layer.14.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 2, + "offset": 15366144, + "bytes": 4194304 + }, + "base_scales": { + "shard": 2, + "offset": 19560448, + "bytes": 131072 + }, + "residual_data": { + "shard": 2, + "offset": 19691520, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 2, + "offset": 21788672, + "bytes": 131072 + } + }, + "encoder.layer.14.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 21919744, + "bytes": 2048 + } + }, + "encoder.layer.14.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 2, + "offset": 21921792, + "bytes": 4194304 + }, + "base_scales": { + "shard": 2, + "offset": 26116096, + "bytes": 131072 + }, + "residual_data": { + "shard": 2, + "offset": 26247168, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 2, + "offset": 28344320, + "bytes": 131072 + } + }, + "encoder.layer.14.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 28475392, + "bytes": 2048 + } + }, + "encoder.layer.14.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 28477440, + "bytes": 2048 + } + }, + "encoder.layer.14.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 28479488, + "bytes": 2048 + } + }, + "encoder.layer.14.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 28481536, + "bytes": 2048 + } + }, + "encoder.layer.15.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 28483584, + "bytes": 2048 + } + }, + "encoder.layer.15.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 2, + "offset": 28485632, + "bytes": 1048576 + }, + "base_scales": { + "shard": 2, + "offset": 29534208, + "bytes": 32768 + }, + "residual_data": { + "shard": 2, + "offset": 29566976, + "bytes": 524288 + }, + "residual_scales": { + "shard": 2, + "offset": 30091264, + "bytes": 32768 + } + }, + "encoder.layer.15.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 30124032, + "bytes": 2048 + } + }, + "encoder.layer.15.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 2, + "offset": 30126080, + "bytes": 1048576 + }, + "base_scales": { + "shard": 2, + "offset": 31174656, + "bytes": 32768 + }, + "residual_data": { + "shard": 2, + "offset": 31207424, + "bytes": 524288 + }, + "residual_scales": { + "shard": 2, + "offset": 31731712, + "bytes": 32768 + } + }, + "encoder.layer.15.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 31764480, + "bytes": 2048 + } + }, + "encoder.layer.15.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 2, + "offset": 31766528, + "bytes": 1048576 + }, + "base_scales": { + "shard": 2, + "offset": 32815104, + "bytes": 32768 + }, + "residual_data": { + "shard": 2, + "offset": 32847872, + "bytes": 524288 + }, + "residual_scales": { + "shard": 2, + "offset": 33372160, + "bytes": 32768 + } + }, + "encoder.layer.15.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 33404928, + "bytes": 2048 + } + }, + "encoder.layer.15.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 2, + "offset": 33406976, + "bytes": 1048576 + }, + "base_scales": { + "shard": 2, + "offset": 34455552, + "bytes": 32768 + }, + "residual_data": { + "shard": 2, + "offset": 34488320, + "bytes": 524288 + }, + "residual_scales": { + "shard": 2, + "offset": 35012608, + "bytes": 32768 + } + }, + "encoder.layer.15.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 35045376, + "bytes": 2048 + } + }, + "encoder.layer.15.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 35047424, + "bytes": 2048 + } + }, + "encoder.layer.15.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 35049472, + "bytes": 8192 + } + }, + "encoder.layer.15.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 2, + "offset": 35057664, + "bytes": 4194304 + }, + "base_scales": { + "shard": 2, + "offset": 39251968, + "bytes": 131072 + }, + "residual_data": { + "shard": 2, + "offset": 39383040, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 2, + "offset": 41480192, + "bytes": 131072 + } + }, + "encoder.layer.15.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 41611264, + "bytes": 2048 + } + }, + "encoder.layer.15.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 2, + "offset": 41613312, + "bytes": 4194304 + }, + "base_scales": { + "shard": 2, + "offset": 45807616, + "bytes": 131072 + }, + "residual_data": { + "shard": 2, + "offset": 45938688, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 2, + "offset": 48035840, + "bytes": 131072 + } + }, + "encoder.layer.15.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 48166912, + "bytes": 2048 + } + }, + "encoder.layer.15.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 48168960, + "bytes": 2048 + } + }, + "encoder.layer.15.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 48171008, + "bytes": 2048 + } + }, + "encoder.layer.15.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 48173056, + "bytes": 2048 + } + }, + "encoder.layer.16.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 48175104, + "bytes": 2048 + } + }, + "encoder.layer.16.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 2, + "offset": 48177152, + "bytes": 1048576 + }, + "base_scales": { + "shard": 2, + "offset": 49225728, + "bytes": 32768 + }, + "residual_data": { + "shard": 2, + "offset": 49258496, + "bytes": 524288 + }, + "residual_scales": { + "shard": 2, + "offset": 49782784, + "bytes": 32768 + } + }, + "encoder.layer.16.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 49815552, + "bytes": 2048 + } + }, + "encoder.layer.16.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 2, + "offset": 49817600, + "bytes": 1048576 + }, + "base_scales": { + "shard": 2, + "offset": 50866176, + "bytes": 32768 + }, + "residual_data": { + "shard": 2, + "offset": 50898944, + "bytes": 524288 + }, + "residual_scales": { + "shard": 2, + "offset": 51423232, + "bytes": 32768 + } + }, + "encoder.layer.16.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 51456000, + "bytes": 2048 + } + }, + "encoder.layer.16.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 2, + "offset": 51458048, + "bytes": 1048576 + }, + "base_scales": { + "shard": 2, + "offset": 52506624, + "bytes": 32768 + }, + "residual_data": { + "shard": 2, + "offset": 52539392, + "bytes": 524288 + }, + "residual_scales": { + "shard": 2, + "offset": 53063680, + "bytes": 32768 + } + }, + "encoder.layer.16.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 53096448, + "bytes": 2048 + } + }, + "encoder.layer.16.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 2, + "offset": 53098496, + "bytes": 1048576 + }, + "base_scales": { + "shard": 2, + "offset": 54147072, + "bytes": 32768 + }, + "residual_data": { + "shard": 2, + "offset": 54179840, + "bytes": 524288 + }, + "residual_scales": { + "shard": 2, + "offset": 54704128, + "bytes": 32768 + } + }, + "encoder.layer.16.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 54736896, + "bytes": 2048 + } + }, + "encoder.layer.16.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 54738944, + "bytes": 2048 + } + }, + "encoder.layer.16.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 54740992, + "bytes": 8192 + } + }, + "encoder.layer.16.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 2, + "offset": 54749184, + "bytes": 4194304 + }, + "base_scales": { + "shard": 2, + "offset": 58943488, + "bytes": 131072 + }, + "residual_data": { + "shard": 2, + "offset": 59074560, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 2, + "offset": 61171712, + "bytes": 131072 + } + }, + "encoder.layer.16.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 2, + "offset": 61302784, + "bytes": 2048 + } + }, + "encoder.layer.16.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 2, + "offset": 61304832, + "bytes": 4194304 + }, + "base_scales": { + "shard": 2, + "offset": 65499136, + "bytes": 131072 + }, + "residual_data": { + "shard": 3, + "offset": 0, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 3, + "offset": 2097152, + "bytes": 131072 + } + }, + "encoder.layer.16.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 2228224, + "bytes": 2048 + } + }, + "encoder.layer.16.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 2230272, + "bytes": 2048 + } + }, + "encoder.layer.16.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 2232320, + "bytes": 2048 + } + }, + "encoder.layer.16.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 2234368, + "bytes": 2048 + } + }, + "encoder.layer.17.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 2236416, + "bytes": 2048 + } + }, + "encoder.layer.17.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 3, + "offset": 2238464, + "bytes": 1048576 + }, + "base_scales": { + "shard": 3, + "offset": 3287040, + "bytes": 32768 + }, + "residual_data": { + "shard": 3, + "offset": 3319808, + "bytes": 524288 + }, + "residual_scales": { + "shard": 3, + "offset": 3844096, + "bytes": 32768 + } + }, + "encoder.layer.17.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 3876864, + "bytes": 2048 + } + }, + "encoder.layer.17.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 3, + "offset": 3878912, + "bytes": 1048576 + }, + "base_scales": { + "shard": 3, + "offset": 4927488, + "bytes": 32768 + }, + "residual_data": { + "shard": 3, + "offset": 4960256, + "bytes": 524288 + }, + "residual_scales": { + "shard": 3, + "offset": 5484544, + "bytes": 32768 + } + }, + "encoder.layer.17.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 5517312, + "bytes": 2048 + } + }, + "encoder.layer.17.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 3, + "offset": 5519360, + "bytes": 1048576 + }, + "base_scales": { + "shard": 3, + "offset": 6567936, + "bytes": 32768 + }, + "residual_data": { + "shard": 3, + "offset": 6600704, + "bytes": 524288 + }, + "residual_scales": { + "shard": 3, + "offset": 7124992, + "bytes": 32768 + } + }, + "encoder.layer.17.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 7157760, + "bytes": 2048 + } + }, + "encoder.layer.17.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 3, + "offset": 7159808, + "bytes": 1048576 + }, + "base_scales": { + "shard": 3, + "offset": 8208384, + "bytes": 32768 + }, + "residual_data": { + "shard": 3, + "offset": 8241152, + "bytes": 524288 + }, + "residual_scales": { + "shard": 3, + "offset": 8765440, + "bytes": 32768 + } + }, + "encoder.layer.17.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 8798208, + "bytes": 2048 + } + }, + "encoder.layer.17.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 8800256, + "bytes": 2048 + } + }, + "encoder.layer.17.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 8802304, + "bytes": 8192 + } + }, + "encoder.layer.17.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 3, + "offset": 8810496, + "bytes": 4194304 + }, + "base_scales": { + "shard": 3, + "offset": 13004800, + "bytes": 131072 + }, + "residual_data": { + "shard": 3, + "offset": 13135872, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 3, + "offset": 15233024, + "bytes": 131072 + } + }, + "encoder.layer.17.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 15364096, + "bytes": 2048 + } + }, + "encoder.layer.17.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 3, + "offset": 15366144, + "bytes": 4194304 + }, + "base_scales": { + "shard": 3, + "offset": 19560448, + "bytes": 131072 + }, + "residual_data": { + "shard": 3, + "offset": 19691520, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 3, + "offset": 21788672, + "bytes": 131072 + } + }, + "encoder.layer.17.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 21919744, + "bytes": 2048 + } + }, + "encoder.layer.17.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 21921792, + "bytes": 2048 + } + }, + "encoder.layer.17.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 21923840, + "bytes": 2048 + } + }, + "encoder.layer.17.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 21925888, + "bytes": 2048 + } + }, + "encoder.layer.18.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 21927936, + "bytes": 2048 + } + }, + "encoder.layer.18.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 3, + "offset": 21929984, + "bytes": 1048576 + }, + "base_scales": { + "shard": 3, + "offset": 22978560, + "bytes": 32768 + }, + "residual_data": { + "shard": 3, + "offset": 23011328, + "bytes": 524288 + }, + "residual_scales": { + "shard": 3, + "offset": 23535616, + "bytes": 32768 + } + }, + "encoder.layer.18.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 23568384, + "bytes": 2048 + } + }, + "encoder.layer.18.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 3, + "offset": 23570432, + "bytes": 1048576 + }, + "base_scales": { + "shard": 3, + "offset": 24619008, + "bytes": 32768 + }, + "residual_data": { + "shard": 3, + "offset": 24651776, + "bytes": 524288 + }, + "residual_scales": { + "shard": 3, + "offset": 25176064, + "bytes": 32768 + } + }, + "encoder.layer.18.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 25208832, + "bytes": 2048 + } + }, + "encoder.layer.18.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 3, + "offset": 25210880, + "bytes": 1048576 + }, + "base_scales": { + "shard": 3, + "offset": 26259456, + "bytes": 32768 + }, + "residual_data": { + "shard": 3, + "offset": 26292224, + "bytes": 524288 + }, + "residual_scales": { + "shard": 3, + "offset": 26816512, + "bytes": 32768 + } + }, + "encoder.layer.18.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 26849280, + "bytes": 2048 + } + }, + "encoder.layer.18.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 3, + "offset": 26851328, + "bytes": 1048576 + }, + "base_scales": { + "shard": 3, + "offset": 27899904, + "bytes": 32768 + }, + "residual_data": { + "shard": 3, + "offset": 27932672, + "bytes": 524288 + }, + "residual_scales": { + "shard": 3, + "offset": 28456960, + "bytes": 32768 + } + }, + "encoder.layer.18.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 28489728, + "bytes": 2048 + } + }, + "encoder.layer.18.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 28491776, + "bytes": 2048 + } + }, + "encoder.layer.18.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 28493824, + "bytes": 8192 + } + }, + "encoder.layer.18.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 3, + "offset": 28502016, + "bytes": 4194304 + }, + "base_scales": { + "shard": 3, + "offset": 32696320, + "bytes": 131072 + }, + "residual_data": { + "shard": 3, + "offset": 32827392, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 3, + "offset": 34924544, + "bytes": 131072 + } + }, + "encoder.layer.18.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 35055616, + "bytes": 2048 + } + }, + "encoder.layer.18.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 3, + "offset": 35057664, + "bytes": 4194304 + }, + "base_scales": { + "shard": 3, + "offset": 39251968, + "bytes": 131072 + }, + "residual_data": { + "shard": 3, + "offset": 39383040, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 3, + "offset": 41480192, + "bytes": 131072 + } + }, + "encoder.layer.18.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 41611264, + "bytes": 2048 + } + }, + "encoder.layer.18.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 41613312, + "bytes": 2048 + } + }, + "encoder.layer.18.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 41615360, + "bytes": 2048 + } + }, + "encoder.layer.18.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 41617408, + "bytes": 2048 + } + }, + "encoder.layer.19.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 41619456, + "bytes": 2048 + } + }, + "encoder.layer.19.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 3, + "offset": 41621504, + "bytes": 1048576 + }, + "base_scales": { + "shard": 3, + "offset": 42670080, + "bytes": 32768 + }, + "residual_data": { + "shard": 3, + "offset": 42702848, + "bytes": 524288 + }, + "residual_scales": { + "shard": 3, + "offset": 43227136, + "bytes": 32768 + } + }, + "encoder.layer.19.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 43259904, + "bytes": 2048 + } + }, + "encoder.layer.19.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 3, + "offset": 43261952, + "bytes": 1048576 + }, + "base_scales": { + "shard": 3, + "offset": 44310528, + "bytes": 32768 + }, + "residual_data": { + "shard": 3, + "offset": 44343296, + "bytes": 524288 + }, + "residual_scales": { + "shard": 3, + "offset": 44867584, + "bytes": 32768 + } + }, + "encoder.layer.19.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 44900352, + "bytes": 2048 + } + }, + "encoder.layer.19.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 3, + "offset": 44902400, + "bytes": 1048576 + }, + "base_scales": { + "shard": 3, + "offset": 45950976, + "bytes": 32768 + }, + "residual_data": { + "shard": 3, + "offset": 45983744, + "bytes": 524288 + }, + "residual_scales": { + "shard": 3, + "offset": 46508032, + "bytes": 32768 + } + }, + "encoder.layer.19.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 46540800, + "bytes": 2048 + } + }, + "encoder.layer.19.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 3, + "offset": 46542848, + "bytes": 1048576 + }, + "base_scales": { + "shard": 3, + "offset": 47591424, + "bytes": 32768 + }, + "residual_data": { + "shard": 3, + "offset": 47624192, + "bytes": 524288 + }, + "residual_scales": { + "shard": 3, + "offset": 48148480, + "bytes": 32768 + } + }, + "encoder.layer.19.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 48181248, + "bytes": 2048 + } + }, + "encoder.layer.19.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 48183296, + "bytes": 2048 + } + }, + "encoder.layer.19.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 48185344, + "bytes": 8192 + } + }, + "encoder.layer.19.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 3, + "offset": 48193536, + "bytes": 4194304 + }, + "base_scales": { + "shard": 3, + "offset": 52387840, + "bytes": 131072 + }, + "residual_data": { + "shard": 3, + "offset": 52518912, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 3, + "offset": 54616064, + "bytes": 131072 + } + }, + "encoder.layer.19.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 54747136, + "bytes": 2048 + } + }, + "encoder.layer.19.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 3, + "offset": 54749184, + "bytes": 4194304 + }, + "base_scales": { + "shard": 3, + "offset": 58943488, + "bytes": 131072 + }, + "residual_data": { + "shard": 3, + "offset": 59074560, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 3, + "offset": 61171712, + "bytes": 131072 + } + }, + "encoder.layer.19.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 61302784, + "bytes": 2048 + } + }, + "encoder.layer.19.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 61304832, + "bytes": 2048 + } + }, + "encoder.layer.19.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 61306880, + "bytes": 2048 + } + }, + "encoder.layer.19.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 61308928, + "bytes": 2048 + } + }, + "encoder.layer.2.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 61310976, + "bytes": 2048 + } + }, + "encoder.layer.2.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 3, + "offset": 61313024, + "bytes": 1048576 + }, + "base_scales": { + "shard": 3, + "offset": 62361600, + "bytes": 32768 + }, + "residual_data": { + "shard": 3, + "offset": 62394368, + "bytes": 524288 + }, + "residual_scales": { + "shard": 3, + "offset": 62918656, + "bytes": 32768 + } + }, + "encoder.layer.2.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 62951424, + "bytes": 2048 + } + }, + "encoder.layer.2.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 3, + "offset": 62953472, + "bytes": 1048576 + }, + "base_scales": { + "shard": 3, + "offset": 64002048, + "bytes": 32768 + }, + "residual_data": { + "shard": 3, + "offset": 64034816, + "bytes": 524288 + }, + "residual_scales": { + "shard": 3, + "offset": 64559104, + "bytes": 32768 + } + }, + "encoder.layer.2.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 64591872, + "bytes": 2048 + } + }, + "encoder.layer.2.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 3, + "offset": 64593920, + "bytes": 1048576 + }, + "base_scales": { + "shard": 3, + "offset": 65642496, + "bytes": 32768 + }, + "residual_data": { + "shard": 3, + "offset": 65675264, + "bytes": 524288 + }, + "residual_scales": { + "shard": 3, + "offset": 66199552, + "bytes": 32768 + } + }, + "encoder.layer.2.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 3, + "offset": 66232320, + "bytes": 2048 + } + }, + "encoder.layer.2.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 4, + "offset": 0, + "bytes": 1048576 + }, + "base_scales": { + "shard": 4, + "offset": 1048576, + "bytes": 32768 + }, + "residual_data": { + "shard": 4, + "offset": 1081344, + "bytes": 524288 + }, + "residual_scales": { + "shard": 4, + "offset": 1605632, + "bytes": 32768 + } + }, + "encoder.layer.2.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 1638400, + "bytes": 2048 + } + }, + "encoder.layer.2.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 1640448, + "bytes": 2048 + } + }, + "encoder.layer.2.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 1642496, + "bytes": 8192 + } + }, + "encoder.layer.2.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 4, + "offset": 1650688, + "bytes": 4194304 + }, + "base_scales": { + "shard": 4, + "offset": 5844992, + "bytes": 131072 + }, + "residual_data": { + "shard": 4, + "offset": 5976064, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 4, + "offset": 8073216, + "bytes": 131072 + } + }, + "encoder.layer.2.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 8204288, + "bytes": 2048 + } + }, + "encoder.layer.2.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 4, + "offset": 8206336, + "bytes": 4194304 + }, + "base_scales": { + "shard": 4, + "offset": 12400640, + "bytes": 131072 + }, + "residual_data": { + "shard": 4, + "offset": 12531712, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 4, + "offset": 14628864, + "bytes": 131072 + } + }, + "encoder.layer.2.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 14759936, + "bytes": 2048 + } + }, + "encoder.layer.2.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 14761984, + "bytes": 2048 + } + }, + "encoder.layer.2.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 14764032, + "bytes": 2048 + } + }, + "encoder.layer.2.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 14766080, + "bytes": 2048 + } + }, + "encoder.layer.20.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 14768128, + "bytes": 2048 + } + }, + "encoder.layer.20.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 4, + "offset": 14770176, + "bytes": 1048576 + }, + "base_scales": { + "shard": 4, + "offset": 15818752, + "bytes": 32768 + }, + "residual_data": { + "shard": 4, + "offset": 15851520, + "bytes": 524288 + }, + "residual_scales": { + "shard": 4, + "offset": 16375808, + "bytes": 32768 + } + }, + "encoder.layer.20.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 16408576, + "bytes": 2048 + } + }, + "encoder.layer.20.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 4, + "offset": 16410624, + "bytes": 1048576 + }, + "base_scales": { + "shard": 4, + "offset": 17459200, + "bytes": 32768 + }, + "residual_data": { + "shard": 4, + "offset": 17491968, + "bytes": 524288 + }, + "residual_scales": { + "shard": 4, + "offset": 18016256, + "bytes": 32768 + } + }, + "encoder.layer.20.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 18049024, + "bytes": 2048 + } + }, + "encoder.layer.20.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 4, + "offset": 18051072, + "bytes": 1048576 + }, + "base_scales": { + "shard": 4, + "offset": 19099648, + "bytes": 32768 + }, + "residual_data": { + "shard": 4, + "offset": 19132416, + "bytes": 524288 + }, + "residual_scales": { + "shard": 4, + "offset": 19656704, + "bytes": 32768 + } + }, + "encoder.layer.20.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 19689472, + "bytes": 2048 + } + }, + "encoder.layer.20.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 4, + "offset": 19691520, + "bytes": 1048576 + }, + "base_scales": { + "shard": 4, + "offset": 20740096, + "bytes": 32768 + }, + "residual_data": { + "shard": 4, + "offset": 20772864, + "bytes": 524288 + }, + "residual_scales": { + "shard": 4, + "offset": 21297152, + "bytes": 32768 + } + }, + "encoder.layer.20.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 21329920, + "bytes": 2048 + } + }, + "encoder.layer.20.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 21331968, + "bytes": 2048 + } + }, + "encoder.layer.20.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 21334016, + "bytes": 8192 + } + }, + "encoder.layer.20.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 4, + "offset": 21342208, + "bytes": 4194304 + }, + "base_scales": { + "shard": 4, + "offset": 25536512, + "bytes": 131072 + }, + "residual_data": { + "shard": 4, + "offset": 25667584, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 4, + "offset": 27764736, + "bytes": 131072 + } + }, + "encoder.layer.20.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 27895808, + "bytes": 2048 + } + }, + "encoder.layer.20.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 4, + "offset": 27897856, + "bytes": 4194304 + }, + "base_scales": { + "shard": 4, + "offset": 32092160, + "bytes": 131072 + }, + "residual_data": { + "shard": 4, + "offset": 32223232, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 4, + "offset": 34320384, + "bytes": 131072 + } + }, + "encoder.layer.20.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 34451456, + "bytes": 2048 + } + }, + "encoder.layer.20.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 34453504, + "bytes": 2048 + } + }, + "encoder.layer.20.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 34455552, + "bytes": 2048 + } + }, + "encoder.layer.20.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 34457600, + "bytes": 2048 + } + }, + "encoder.layer.21.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 34459648, + "bytes": 2048 + } + }, + "encoder.layer.21.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 4, + "offset": 34461696, + "bytes": 1048576 + }, + "base_scales": { + "shard": 4, + "offset": 35510272, + "bytes": 32768 + }, + "residual_data": { + "shard": 4, + "offset": 35543040, + "bytes": 524288 + }, + "residual_scales": { + "shard": 4, + "offset": 36067328, + "bytes": 32768 + } + }, + "encoder.layer.21.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 36100096, + "bytes": 2048 + } + }, + "encoder.layer.21.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 4, + "offset": 36102144, + "bytes": 1048576 + }, + "base_scales": { + "shard": 4, + "offset": 37150720, + "bytes": 32768 + }, + "residual_data": { + "shard": 4, + "offset": 37183488, + "bytes": 524288 + }, + "residual_scales": { + "shard": 4, + "offset": 37707776, + "bytes": 32768 + } + }, + "encoder.layer.21.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 37740544, + "bytes": 2048 + } + }, + "encoder.layer.21.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 4, + "offset": 37742592, + "bytes": 1048576 + }, + "base_scales": { + "shard": 4, + "offset": 38791168, + "bytes": 32768 + }, + "residual_data": { + "shard": 4, + "offset": 38823936, + "bytes": 524288 + }, + "residual_scales": { + "shard": 4, + "offset": 39348224, + "bytes": 32768 + } + }, + "encoder.layer.21.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 39380992, + "bytes": 2048 + } + }, + "encoder.layer.21.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 4, + "offset": 39383040, + "bytes": 1048576 + }, + "base_scales": { + "shard": 4, + "offset": 40431616, + "bytes": 32768 + }, + "residual_data": { + "shard": 4, + "offset": 40464384, + "bytes": 524288 + }, + "residual_scales": { + "shard": 4, + "offset": 40988672, + "bytes": 32768 + } + }, + "encoder.layer.21.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 41021440, + "bytes": 2048 + } + }, + "encoder.layer.21.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 41023488, + "bytes": 2048 + } + }, + "encoder.layer.21.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 41025536, + "bytes": 8192 + } + }, + "encoder.layer.21.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 4, + "offset": 41033728, + "bytes": 4194304 + }, + "base_scales": { + "shard": 4, + "offset": 45228032, + "bytes": 131072 + }, + "residual_data": { + "shard": 4, + "offset": 45359104, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 4, + "offset": 47456256, + "bytes": 131072 + } + }, + "encoder.layer.21.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 47587328, + "bytes": 2048 + } + }, + "encoder.layer.21.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 4, + "offset": 47589376, + "bytes": 4194304 + }, + "base_scales": { + "shard": 4, + "offset": 51783680, + "bytes": 131072 + }, + "residual_data": { + "shard": 4, + "offset": 51914752, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 4, + "offset": 54011904, + "bytes": 131072 + } + }, + "encoder.layer.21.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 54142976, + "bytes": 2048 + } + }, + "encoder.layer.21.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 54145024, + "bytes": 2048 + } + }, + "encoder.layer.21.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 54147072, + "bytes": 2048 + } + }, + "encoder.layer.21.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 54149120, + "bytes": 2048 + } + }, + "encoder.layer.22.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 54151168, + "bytes": 2048 + } + }, + "encoder.layer.22.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 4, + "offset": 54153216, + "bytes": 1048576 + }, + "base_scales": { + "shard": 4, + "offset": 55201792, + "bytes": 32768 + }, + "residual_data": { + "shard": 4, + "offset": 55234560, + "bytes": 524288 + }, + "residual_scales": { + "shard": 4, + "offset": 55758848, + "bytes": 32768 + } + }, + "encoder.layer.22.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 55791616, + "bytes": 2048 + } + }, + "encoder.layer.22.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 4, + "offset": 55793664, + "bytes": 1048576 + }, + "base_scales": { + "shard": 4, + "offset": 56842240, + "bytes": 32768 + }, + "residual_data": { + "shard": 4, + "offset": 56875008, + "bytes": 524288 + }, + "residual_scales": { + "shard": 4, + "offset": 57399296, + "bytes": 32768 + } + }, + "encoder.layer.22.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 57432064, + "bytes": 2048 + } + }, + "encoder.layer.22.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 4, + "offset": 57434112, + "bytes": 1048576 + }, + "base_scales": { + "shard": 4, + "offset": 58482688, + "bytes": 32768 + }, + "residual_data": { + "shard": 4, + "offset": 58515456, + "bytes": 524288 + }, + "residual_scales": { + "shard": 4, + "offset": 59039744, + "bytes": 32768 + } + }, + "encoder.layer.22.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 59072512, + "bytes": 2048 + } + }, + "encoder.layer.22.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 4, + "offset": 59074560, + "bytes": 1048576 + }, + "base_scales": { + "shard": 4, + "offset": 60123136, + "bytes": 32768 + }, + "residual_data": { + "shard": 4, + "offset": 60155904, + "bytes": 524288 + }, + "residual_scales": { + "shard": 4, + "offset": 60680192, + "bytes": 32768 + } + }, + "encoder.layer.22.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 60712960, + "bytes": 2048 + } + }, + "encoder.layer.22.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 60715008, + "bytes": 2048 + } + }, + "encoder.layer.22.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 4, + "offset": 60717056, + "bytes": 8192 + } + }, + "encoder.layer.22.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 4, + "offset": 60725248, + "bytes": 4194304 + }, + "base_scales": { + "shard": 4, + "offset": 64919552, + "bytes": 131072 + }, + "residual_data": { + "shard": 5, + "offset": 0, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 5, + "offset": 2097152, + "bytes": 131072 + } + }, + "encoder.layer.22.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 2228224, + "bytes": 2048 + } + }, + "encoder.layer.22.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 5, + "offset": 2230272, + "bytes": 4194304 + }, + "base_scales": { + "shard": 5, + "offset": 6424576, + "bytes": 131072 + }, + "residual_data": { + "shard": 5, + "offset": 6555648, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 5, + "offset": 8652800, + "bytes": 131072 + } + }, + "encoder.layer.22.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 8783872, + "bytes": 2048 + } + }, + "encoder.layer.22.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 8785920, + "bytes": 2048 + } + }, + "encoder.layer.22.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 8787968, + "bytes": 2048 + } + }, + "encoder.layer.22.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 8790016, + "bytes": 2048 + } + }, + "encoder.layer.23.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 8792064, + "bytes": 2048 + } + }, + "encoder.layer.23.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 5, + "offset": 8794112, + "bytes": 1048576 + }, + "base_scales": { + "shard": 5, + "offset": 9842688, + "bytes": 32768 + }, + "residual_data": { + "shard": 5, + "offset": 9875456, + "bytes": 524288 + }, + "residual_scales": { + "shard": 5, + "offset": 10399744, + "bytes": 32768 + } + }, + "encoder.layer.23.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 10432512, + "bytes": 2048 + } + }, + "encoder.layer.23.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 5, + "offset": 10434560, + "bytes": 1048576 + }, + "base_scales": { + "shard": 5, + "offset": 11483136, + "bytes": 32768 + }, + "residual_data": { + "shard": 5, + "offset": 11515904, + "bytes": 524288 + }, + "residual_scales": { + "shard": 5, + "offset": 12040192, + "bytes": 32768 + } + }, + "encoder.layer.23.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 12072960, + "bytes": 2048 + } + }, + "encoder.layer.23.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 5, + "offset": 12075008, + "bytes": 1048576 + }, + "base_scales": { + "shard": 5, + "offset": 13123584, + "bytes": 32768 + }, + "residual_data": { + "shard": 5, + "offset": 13156352, + "bytes": 524288 + }, + "residual_scales": { + "shard": 5, + "offset": 13680640, + "bytes": 32768 + } + }, + "encoder.layer.23.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 13713408, + "bytes": 2048 + } + }, + "encoder.layer.23.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 5, + "offset": 13715456, + "bytes": 1048576 + }, + "base_scales": { + "shard": 5, + "offset": 14764032, + "bytes": 32768 + }, + "residual_data": { + "shard": 5, + "offset": 14796800, + "bytes": 524288 + }, + "residual_scales": { + "shard": 5, + "offset": 15321088, + "bytes": 32768 + } + }, + "encoder.layer.23.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 15353856, + "bytes": 2048 + } + }, + "encoder.layer.23.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 15355904, + "bytes": 2048 + } + }, + "encoder.layer.23.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 15357952, + "bytes": 8192 + } + }, + "encoder.layer.23.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 5, + "offset": 15366144, + "bytes": 4194304 + }, + "base_scales": { + "shard": 5, + "offset": 19560448, + "bytes": 131072 + }, + "residual_data": { + "shard": 5, + "offset": 19691520, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 5, + "offset": 21788672, + "bytes": 131072 + } + }, + "encoder.layer.23.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 21919744, + "bytes": 2048 + } + }, + "encoder.layer.23.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 5, + "offset": 21921792, + "bytes": 4194304 + }, + "base_scales": { + "shard": 5, + "offset": 26116096, + "bytes": 131072 + }, + "residual_data": { + "shard": 5, + "offset": 26247168, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 5, + "offset": 28344320, + "bytes": 131072 + } + }, + "encoder.layer.23.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 28475392, + "bytes": 2048 + } + }, + "encoder.layer.23.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 28477440, + "bytes": 2048 + } + }, + "encoder.layer.23.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 28479488, + "bytes": 2048 + } + }, + "encoder.layer.23.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 28481536, + "bytes": 2048 + } + }, + "encoder.layer.3.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 28483584, + "bytes": 2048 + } + }, + "encoder.layer.3.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 5, + "offset": 28485632, + "bytes": 1048576 + }, + "base_scales": { + "shard": 5, + "offset": 29534208, + "bytes": 32768 + }, + "residual_data": { + "shard": 5, + "offset": 29566976, + "bytes": 524288 + }, + "residual_scales": { + "shard": 5, + "offset": 30091264, + "bytes": 32768 + } + }, + "encoder.layer.3.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 30124032, + "bytes": 2048 + } + }, + "encoder.layer.3.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 5, + "offset": 30126080, + "bytes": 1048576 + }, + "base_scales": { + "shard": 5, + "offset": 31174656, + "bytes": 32768 + }, + "residual_data": { + "shard": 5, + "offset": 31207424, + "bytes": 524288 + }, + "residual_scales": { + "shard": 5, + "offset": 31731712, + "bytes": 32768 + } + }, + "encoder.layer.3.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 31764480, + "bytes": 2048 + } + }, + "encoder.layer.3.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 5, + "offset": 31766528, + "bytes": 1048576 + }, + "base_scales": { + "shard": 5, + "offset": 32815104, + "bytes": 32768 + }, + "residual_data": { + "shard": 5, + "offset": 32847872, + "bytes": 524288 + }, + "residual_scales": { + "shard": 5, + "offset": 33372160, + "bytes": 32768 + } + }, + "encoder.layer.3.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 33404928, + "bytes": 2048 + } + }, + "encoder.layer.3.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 5, + "offset": 33406976, + "bytes": 1048576 + }, + "base_scales": { + "shard": 5, + "offset": 34455552, + "bytes": 32768 + }, + "residual_data": { + "shard": 5, + "offset": 34488320, + "bytes": 524288 + }, + "residual_scales": { + "shard": 5, + "offset": 35012608, + "bytes": 32768 + } + }, + "encoder.layer.3.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 35045376, + "bytes": 2048 + } + }, + "encoder.layer.3.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 35047424, + "bytes": 2048 + } + }, + "encoder.layer.3.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 35049472, + "bytes": 8192 + } + }, + "encoder.layer.3.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 5, + "offset": 35057664, + "bytes": 4194304 + }, + "base_scales": { + "shard": 5, + "offset": 39251968, + "bytes": 131072 + }, + "residual_data": { + "shard": 5, + "offset": 39383040, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 5, + "offset": 41480192, + "bytes": 131072 + } + }, + "encoder.layer.3.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 41611264, + "bytes": 2048 + } + }, + "encoder.layer.3.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 5, + "offset": 41613312, + "bytes": 4194304 + }, + "base_scales": { + "shard": 5, + "offset": 45807616, + "bytes": 131072 + }, + "residual_data": { + "shard": 5, + "offset": 45938688, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 5, + "offset": 48035840, + "bytes": 131072 + } + }, + "encoder.layer.3.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 48166912, + "bytes": 2048 + } + }, + "encoder.layer.3.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 48168960, + "bytes": 2048 + } + }, + "encoder.layer.3.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 48171008, + "bytes": 2048 + } + }, + "encoder.layer.3.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 48173056, + "bytes": 2048 + } + }, + "encoder.layer.4.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 48175104, + "bytes": 2048 + } + }, + "encoder.layer.4.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 5, + "offset": 48177152, + "bytes": 1048576 + }, + "base_scales": { + "shard": 5, + "offset": 49225728, + "bytes": 32768 + }, + "residual_data": { + "shard": 5, + "offset": 49258496, + "bytes": 524288 + }, + "residual_scales": { + "shard": 5, + "offset": 49782784, + "bytes": 32768 + } + }, + "encoder.layer.4.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 49815552, + "bytes": 2048 + } + }, + "encoder.layer.4.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 5, + "offset": 49817600, + "bytes": 1048576 + }, + "base_scales": { + "shard": 5, + "offset": 50866176, + "bytes": 32768 + }, + "residual_data": { + "shard": 5, + "offset": 50898944, + "bytes": 524288 + }, + "residual_scales": { + "shard": 5, + "offset": 51423232, + "bytes": 32768 + } + }, + "encoder.layer.4.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 51456000, + "bytes": 2048 + } + }, + "encoder.layer.4.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 5, + "offset": 51458048, + "bytes": 1048576 + }, + "base_scales": { + "shard": 5, + "offset": 52506624, + "bytes": 32768 + }, + "residual_data": { + "shard": 5, + "offset": 52539392, + "bytes": 524288 + }, + "residual_scales": { + "shard": 5, + "offset": 53063680, + "bytes": 32768 + } + }, + "encoder.layer.4.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 53096448, + "bytes": 2048 + } + }, + "encoder.layer.4.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 5, + "offset": 53098496, + "bytes": 1048576 + }, + "base_scales": { + "shard": 5, + "offset": 54147072, + "bytes": 32768 + }, + "residual_data": { + "shard": 5, + "offset": 54179840, + "bytes": 524288 + }, + "residual_scales": { + "shard": 5, + "offset": 54704128, + "bytes": 32768 + } + }, + "encoder.layer.4.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 54736896, + "bytes": 2048 + } + }, + "encoder.layer.4.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 54738944, + "bytes": 2048 + } + }, + "encoder.layer.4.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 54740992, + "bytes": 8192 + } + }, + "encoder.layer.4.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 5, + "offset": 54749184, + "bytes": 4194304 + }, + "base_scales": { + "shard": 5, + "offset": 58943488, + "bytes": 131072 + }, + "residual_data": { + "shard": 5, + "offset": 59074560, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 5, + "offset": 61171712, + "bytes": 131072 + } + }, + "encoder.layer.4.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 5, + "offset": 61302784, + "bytes": 2048 + } + }, + "encoder.layer.4.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 5, + "offset": 61304832, + "bytes": 4194304 + }, + "base_scales": { + "shard": 5, + "offset": 65499136, + "bytes": 131072 + }, + "residual_data": { + "shard": 6, + "offset": 0, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 6, + "offset": 2097152, + "bytes": 131072 + } + }, + "encoder.layer.4.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 2228224, + "bytes": 2048 + } + }, + "encoder.layer.4.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 2230272, + "bytes": 2048 + } + }, + "encoder.layer.4.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 2232320, + "bytes": 2048 + } + }, + "encoder.layer.4.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 2234368, + "bytes": 2048 + } + }, + "encoder.layer.5.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 2236416, + "bytes": 2048 + } + }, + "encoder.layer.5.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 6, + "offset": 2238464, + "bytes": 1048576 + }, + "base_scales": { + "shard": 6, + "offset": 3287040, + "bytes": 32768 + }, + "residual_data": { + "shard": 6, + "offset": 3319808, + "bytes": 524288 + }, + "residual_scales": { + "shard": 6, + "offset": 3844096, + "bytes": 32768 + } + }, + "encoder.layer.5.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 3876864, + "bytes": 2048 + } + }, + "encoder.layer.5.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 6, + "offset": 3878912, + "bytes": 1048576 + }, + "base_scales": { + "shard": 6, + "offset": 4927488, + "bytes": 32768 + }, + "residual_data": { + "shard": 6, + "offset": 4960256, + "bytes": 524288 + }, + "residual_scales": { + "shard": 6, + "offset": 5484544, + "bytes": 32768 + } + }, + "encoder.layer.5.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 5517312, + "bytes": 2048 + } + }, + "encoder.layer.5.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 6, + "offset": 5519360, + "bytes": 1048576 + }, + "base_scales": { + "shard": 6, + "offset": 6567936, + "bytes": 32768 + }, + "residual_data": { + "shard": 6, + "offset": 6600704, + "bytes": 524288 + }, + "residual_scales": { + "shard": 6, + "offset": 7124992, + "bytes": 32768 + } + }, + "encoder.layer.5.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 7157760, + "bytes": 2048 + } + }, + "encoder.layer.5.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 6, + "offset": 7159808, + "bytes": 1048576 + }, + "base_scales": { + "shard": 6, + "offset": 8208384, + "bytes": 32768 + }, + "residual_data": { + "shard": 6, + "offset": 8241152, + "bytes": 524288 + }, + "residual_scales": { + "shard": 6, + "offset": 8765440, + "bytes": 32768 + } + }, + "encoder.layer.5.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 8798208, + "bytes": 2048 + } + }, + "encoder.layer.5.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 8800256, + "bytes": 2048 + } + }, + "encoder.layer.5.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 8802304, + "bytes": 8192 + } + }, + "encoder.layer.5.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 6, + "offset": 8810496, + "bytes": 4194304 + }, + "base_scales": { + "shard": 6, + "offset": 13004800, + "bytes": 131072 + }, + "residual_data": { + "shard": 6, + "offset": 13135872, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 6, + "offset": 15233024, + "bytes": 131072 + } + }, + "encoder.layer.5.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 15364096, + "bytes": 2048 + } + }, + "encoder.layer.5.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 6, + "offset": 15366144, + "bytes": 4194304 + }, + "base_scales": { + "shard": 6, + "offset": 19560448, + "bytes": 131072 + }, + "residual_data": { + "shard": 6, + "offset": 19691520, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 6, + "offset": 21788672, + "bytes": 131072 + } + }, + "encoder.layer.5.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 21919744, + "bytes": 2048 + } + }, + "encoder.layer.5.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 21921792, + "bytes": 2048 + } + }, + "encoder.layer.5.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 21923840, + "bytes": 2048 + } + }, + "encoder.layer.5.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 21925888, + "bytes": 2048 + } + }, + "encoder.layer.6.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 21927936, + "bytes": 2048 + } + }, + "encoder.layer.6.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 6, + "offset": 21929984, + "bytes": 1048576 + }, + "base_scales": { + "shard": 6, + "offset": 22978560, + "bytes": 32768 + }, + "residual_data": { + "shard": 6, + "offset": 23011328, + "bytes": 524288 + }, + "residual_scales": { + "shard": 6, + "offset": 23535616, + "bytes": 32768 + } + }, + "encoder.layer.6.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 23568384, + "bytes": 2048 + } + }, + "encoder.layer.6.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 6, + "offset": 23570432, + "bytes": 1048576 + }, + "base_scales": { + "shard": 6, + "offset": 24619008, + "bytes": 32768 + }, + "residual_data": { + "shard": 6, + "offset": 24651776, + "bytes": 524288 + }, + "residual_scales": { + "shard": 6, + "offset": 25176064, + "bytes": 32768 + } + }, + "encoder.layer.6.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 25208832, + "bytes": 2048 + } + }, + "encoder.layer.6.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 6, + "offset": 25210880, + "bytes": 1048576 + }, + "base_scales": { + "shard": 6, + "offset": 26259456, + "bytes": 32768 + }, + "residual_data": { + "shard": 6, + "offset": 26292224, + "bytes": 524288 + }, + "residual_scales": { + "shard": 6, + "offset": 26816512, + "bytes": 32768 + } + }, + "encoder.layer.6.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 26849280, + "bytes": 2048 + } + }, + "encoder.layer.6.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 6, + "offset": 26851328, + "bytes": 1048576 + }, + "base_scales": { + "shard": 6, + "offset": 27899904, + "bytes": 32768 + }, + "residual_data": { + "shard": 6, + "offset": 27932672, + "bytes": 524288 + }, + "residual_scales": { + "shard": 6, + "offset": 28456960, + "bytes": 32768 + } + }, + "encoder.layer.6.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 28489728, + "bytes": 2048 + } + }, + "encoder.layer.6.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 28491776, + "bytes": 2048 + } + }, + "encoder.layer.6.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 28493824, + "bytes": 8192 + } + }, + "encoder.layer.6.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 6, + "offset": 28502016, + "bytes": 4194304 + }, + "base_scales": { + "shard": 6, + "offset": 32696320, + "bytes": 131072 + }, + "residual_data": { + "shard": 6, + "offset": 32827392, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 6, + "offset": 34924544, + "bytes": 131072 + } + }, + "encoder.layer.6.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 35055616, + "bytes": 2048 + } + }, + "encoder.layer.6.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 6, + "offset": 35057664, + "bytes": 4194304 + }, + "base_scales": { + "shard": 6, + "offset": 39251968, + "bytes": 131072 + }, + "residual_data": { + "shard": 6, + "offset": 39383040, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 6, + "offset": 41480192, + "bytes": 131072 + } + }, + "encoder.layer.6.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 41611264, + "bytes": 2048 + } + }, + "encoder.layer.6.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 41613312, + "bytes": 2048 + } + }, + "encoder.layer.6.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 41615360, + "bytes": 2048 + } + }, + "encoder.layer.6.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 41617408, + "bytes": 2048 + } + }, + "encoder.layer.7.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 41619456, + "bytes": 2048 + } + }, + "encoder.layer.7.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 6, + "offset": 41621504, + "bytes": 1048576 + }, + "base_scales": { + "shard": 6, + "offset": 42670080, + "bytes": 32768 + }, + "residual_data": { + "shard": 6, + "offset": 42702848, + "bytes": 524288 + }, + "residual_scales": { + "shard": 6, + "offset": 43227136, + "bytes": 32768 + } + }, + "encoder.layer.7.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 43259904, + "bytes": 2048 + } + }, + "encoder.layer.7.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 6, + "offset": 43261952, + "bytes": 1048576 + }, + "base_scales": { + "shard": 6, + "offset": 44310528, + "bytes": 32768 + }, + "residual_data": { + "shard": 6, + "offset": 44343296, + "bytes": 524288 + }, + "residual_scales": { + "shard": 6, + "offset": 44867584, + "bytes": 32768 + } + }, + "encoder.layer.7.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 44900352, + "bytes": 2048 + } + }, + "encoder.layer.7.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 6, + "offset": 44902400, + "bytes": 1048576 + }, + "base_scales": { + "shard": 6, + "offset": 45950976, + "bytes": 32768 + }, + "residual_data": { + "shard": 6, + "offset": 45983744, + "bytes": 524288 + }, + "residual_scales": { + "shard": 6, + "offset": 46508032, + "bytes": 32768 + } + }, + "encoder.layer.7.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 46540800, + "bytes": 2048 + } + }, + "encoder.layer.7.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 6, + "offset": 46542848, + "bytes": 1048576 + }, + "base_scales": { + "shard": 6, + "offset": 47591424, + "bytes": 32768 + }, + "residual_data": { + "shard": 6, + "offset": 47624192, + "bytes": 524288 + }, + "residual_scales": { + "shard": 6, + "offset": 48148480, + "bytes": 32768 + } + }, + "encoder.layer.7.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 48181248, + "bytes": 2048 + } + }, + "encoder.layer.7.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 48183296, + "bytes": 2048 + } + }, + "encoder.layer.7.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 48185344, + "bytes": 8192 + } + }, + "encoder.layer.7.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 6, + "offset": 48193536, + "bytes": 4194304 + }, + "base_scales": { + "shard": 6, + "offset": 52387840, + "bytes": 131072 + }, + "residual_data": { + "shard": 6, + "offset": 52518912, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 6, + "offset": 54616064, + "bytes": 131072 + } + }, + "encoder.layer.7.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 54747136, + "bytes": 2048 + } + }, + "encoder.layer.7.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 6, + "offset": 54749184, + "bytes": 4194304 + }, + "base_scales": { + "shard": 6, + "offset": 58943488, + "bytes": 131072 + }, + "residual_data": { + "shard": 6, + "offset": 59074560, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 6, + "offset": 61171712, + "bytes": 131072 + } + }, + "encoder.layer.7.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 61302784, + "bytes": 2048 + } + }, + "encoder.layer.7.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 61304832, + "bytes": 2048 + } + }, + "encoder.layer.7.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 61306880, + "bytes": 2048 + } + }, + "encoder.layer.7.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 61308928, + "bytes": 2048 + } + }, + "encoder.layer.8.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 61310976, + "bytes": 2048 + } + }, + "encoder.layer.8.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 6, + "offset": 61313024, + "bytes": 1048576 + }, + "base_scales": { + "shard": 6, + "offset": 62361600, + "bytes": 32768 + }, + "residual_data": { + "shard": 6, + "offset": 62394368, + "bytes": 524288 + }, + "residual_scales": { + "shard": 6, + "offset": 62918656, + "bytes": 32768 + } + }, + "encoder.layer.8.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 62951424, + "bytes": 2048 + } + }, + "encoder.layer.8.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 6, + "offset": 62953472, + "bytes": 1048576 + }, + "base_scales": { + "shard": 6, + "offset": 64002048, + "bytes": 32768 + }, + "residual_data": { + "shard": 6, + "offset": 64034816, + "bytes": 524288 + }, + "residual_scales": { + "shard": 6, + "offset": 64559104, + "bytes": 32768 + } + }, + "encoder.layer.8.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 64591872, + "bytes": 2048 + } + }, + "encoder.layer.8.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 6, + "offset": 64593920, + "bytes": 1048576 + }, + "base_scales": { + "shard": 6, + "offset": 65642496, + "bytes": 32768 + }, + "residual_data": { + "shard": 6, + "offset": 65675264, + "bytes": 524288 + }, + "residual_scales": { + "shard": 6, + "offset": 66199552, + "bytes": 32768 + } + }, + "encoder.layer.8.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 6, + "offset": 66232320, + "bytes": 2048 + } + }, + "encoder.layer.8.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 7, + "offset": 0, + "bytes": 1048576 + }, + "base_scales": { + "shard": 7, + "offset": 1048576, + "bytes": 32768 + }, + "residual_data": { + "shard": 7, + "offset": 1081344, + "bytes": 524288 + }, + "residual_scales": { + "shard": 7, + "offset": 1605632, + "bytes": 32768 + } + }, + "encoder.layer.8.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 7, + "offset": 1638400, + "bytes": 2048 + } + }, + "encoder.layer.8.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 7, + "offset": 1640448, + "bytes": 2048 + } + }, + "encoder.layer.8.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 7, + "offset": 1642496, + "bytes": 8192 + } + }, + "encoder.layer.8.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 7, + "offset": 1650688, + "bytes": 4194304 + }, + "base_scales": { + "shard": 7, + "offset": 5844992, + "bytes": 131072 + }, + "residual_data": { + "shard": 7, + "offset": 5976064, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 7, + "offset": 8073216, + "bytes": 131072 + } + }, + "encoder.layer.8.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 7, + "offset": 8204288, + "bytes": 2048 + } + }, + "encoder.layer.8.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 7, + "offset": 8206336, + "bytes": 4194304 + }, + "base_scales": { + "shard": 7, + "offset": 12400640, + "bytes": 131072 + }, + "residual_data": { + "shard": 7, + "offset": 12531712, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 7, + "offset": 14628864, + "bytes": 131072 + } + }, + "encoder.layer.8.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 7, + "offset": 14759936, + "bytes": 2048 + } + }, + "encoder.layer.8.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 7, + "offset": 14761984, + "bytes": 2048 + } + }, + "encoder.layer.8.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 7, + "offset": 14764032, + "bytes": 2048 + } + }, + "encoder.layer.8.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 7, + "offset": 14766080, + "bytes": 2048 + } + }, + "encoder.layer.9.attention.attention.key.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 7, + "offset": 14768128, + "bytes": 2048 + } + }, + "encoder.layer.9.attention.attention.key.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 7, + "offset": 14770176, + "bytes": 1048576 + }, + "base_scales": { + "shard": 7, + "offset": 15818752, + "bytes": 32768 + }, + "residual_data": { + "shard": 7, + "offset": 15851520, + "bytes": 524288 + }, + "residual_scales": { + "shard": 7, + "offset": 16375808, + "bytes": 32768 + } + }, + "encoder.layer.9.attention.attention.query.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 7, + "offset": 16408576, + "bytes": 2048 + } + }, + "encoder.layer.9.attention.attention.query.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 7, + "offset": 16410624, + "bytes": 1048576 + }, + "base_scales": { + "shard": 7, + "offset": 17459200, + "bytes": 32768 + }, + "residual_data": { + "shard": 7, + "offset": 17491968, + "bytes": 524288 + }, + "residual_scales": { + "shard": 7, + "offset": 18016256, + "bytes": 32768 + } + }, + "encoder.layer.9.attention.attention.value.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 7, + "offset": 18049024, + "bytes": 2048 + } + }, + "encoder.layer.9.attention.attention.value.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 7, + "offset": 18051072, + "bytes": 1048576 + }, + "base_scales": { + "shard": 7, + "offset": 19099648, + "bytes": 32768 + }, + "residual_data": { + "shard": 7, + "offset": 19132416, + "bytes": 524288 + }, + "residual_scales": { + "shard": 7, + "offset": 19656704, + "bytes": 32768 + } + }, + "encoder.layer.9.attention.output.dense.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 7, + "offset": 19689472, + "bytes": 2048 + } + }, + "encoder.layer.9.attention.output.dense.weight": { + "shape": [ + 1024, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 1024, + "padded_rows": 1024, + "padded_columns": 1024, + "base_data": { + "shard": 7, + "offset": 19691520, + "bytes": 1048576 + }, + "base_scales": { + "shard": 7, + "offset": 20740096, + "bytes": 32768 + }, + "residual_data": { + "shard": 7, + "offset": 20772864, + "bytes": 524288 + }, + "residual_scales": { + "shard": 7, + "offset": 21297152, + "bytes": 32768 + } + }, + "encoder.layer.9.layer_scale1.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 7, + "offset": 21329920, + "bytes": 2048 + } + }, + "encoder.layer.9.layer_scale2.lambda1": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 7, + "offset": 21331968, + "bytes": 2048 + } + }, + "encoder.layer.9.mlp.fc1.bias": { + "shape": [ + 4096 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 7, + "offset": 21334016, + "bytes": 8192 + } + }, + "encoder.layer.9.mlp.fc1.weight": { + "shape": [ + 4096, + 1024 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 4096, + "columns": 1024, + "padded_rows": 4096, + "padded_columns": 1024, + "base_data": { + "shard": 7, + "offset": 21342208, + "bytes": 4194304 + }, + "base_scales": { + "shard": 7, + "offset": 25536512, + "bytes": 131072 + }, + "residual_data": { + "shard": 7, + "offset": 25667584, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 7, + "offset": 27764736, + "bytes": 131072 + } + }, + "encoder.layer.9.mlp.fc2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 7, + "offset": 27895808, + "bytes": 2048 + } + }, + "encoder.layer.9.mlp.fc2.weight": { + "shape": [ + 1024, + 4096 + ], + "encoding": "residual-quantized", + "group_size": 64, + "output_tile": 16, + "rows": 1024, + "columns": 4096, + "padded_rows": 1024, + "padded_columns": 4096, + "base_data": { + "shard": 7, + "offset": 27897856, + "bytes": 4194304 + }, + "base_scales": { + "shard": 7, + "offset": 32092160, + "bytes": 131072 + }, + "residual_data": { + "shard": 7, + "offset": 32223232, + "bytes": 2097152 + }, + "residual_scales": { + "shard": 7, + "offset": 34320384, + "bytes": 131072 + } + }, + "encoder.layer.9.norm1.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 7, + "offset": 34451456, + "bytes": 2048 + } + }, + "encoder.layer.9.norm1.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 7, + "offset": 34453504, + "bytes": 2048 + } + }, + "encoder.layer.9.norm2.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 7, + "offset": 34455552, + "bytes": 2048 + } + }, + "encoder.layer.9.norm2.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 7, + "offset": 34457600, + "bytes": 2048 + } + }, + "layernorm.bias": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 7, + "offset": 34459648, + "bytes": 2048 + } + }, + "layernorm.weight": { + "shape": [ + 1024 + ], + "encoding": "raw", + "dtype": "f16", + "data": { + "shard": 7, + "offset": 34461696, + "bytes": 2048 + } + } + } +}