madlag's picture
Adding model, graphs and metadata.
e3e88c8
{"total": 334064578, "nnz": 109815770, "linear_total": 301989888, "linear_nnz": 77822976, "layers": {"0": {"total": 12594880, "nnz": 2631960, "linear_total": 12582912, "linear_nnz": 2623488, "linear_attention_total": 4194304, "linear_attention_nnz": 1329152, "linear_dense_total": 8388608, "linear_dense_nnz": 1294336}, "1": {"total": 12594880, "nnz": 2977018, "linear_total": 12582912, "linear_nnz": 2968576, "linear_attention_total": 4194304, "linear_attention_nnz": 1539072, "linear_dense_total": 8388608, "linear_dense_nnz": 1429504}, "2": {"total": 12593920, "nnz": 2545251, "linear_total": 12582912, "linear_nnz": 2537472, "linear_attention_total": 4194304, "linear_attention_nnz": 761856, "linear_dense_total": 8388608, "linear_dense_nnz": 1775616}, "3": {"total": 12594304, "nnz": 3200094, "linear_total": 12582912, "linear_nnz": 3191808, "linear_attention_total": 4194304, "linear_attention_nnz": 1164288, "linear_dense_total": 8388608, "linear_dense_nnz": 2027520}, "4": {"total": 12594112, "nnz": 2879362, "linear_total": 12582912, "linear_nnz": 2871296, "linear_attention_total": 4194304, "linear_attention_nnz": 835584, "linear_dense_total": 8388608, "linear_dense_nnz": 2035712}, "5": {"total": 12594112, "nnz": 2866029, "linear_total": 12582912, "linear_nnz": 2857984, "linear_attention_total": 4194304, "linear_attention_nnz": 734208, "linear_dense_total": 8388608, "linear_dense_nnz": 2123776}, "6": {"total": 12593920, "nnz": 3136361, "linear_total": 12582912, "linear_nnz": 3128320, "linear_attention_total": 4194304, "linear_attention_nnz": 816128, "linear_dense_total": 8388608, "linear_dense_nnz": 2312192}, "7": {"total": 12594688, "nnz": 3850774, "linear_total": 12582912, "linear_nnz": 3842048, "linear_attention_total": 4194304, "linear_attention_nnz": 1503232, "linear_dense_total": 8388608, "linear_dense_nnz": 2338816}, "8": {"total": 12594880, "nnz": 4122385, "linear_total": 12582912, "linear_nnz": 4113408, "linear_attention_total": 4194304, "linear_attention_nnz": 1784832, "linear_dense_total": 8388608, "linear_dense_nnz": 2328576}, "9": {"total": 12594688, "nnz": 3869363, "linear_total": 12582912, "linear_nnz": 3860480, "linear_attention_total": 4194304, "linear_attention_nnz": 1396736, "linear_dense_total": 8388608, "linear_dense_nnz": 2463744}, "10": {"total": 12594688, "nnz": 3717580, "linear_total": 12582912, "linear_nnz": 3708928, "linear_attention_total": 4194304, "linear_attention_nnz": 1718272, "linear_dense_total": 8388608, "linear_dense_nnz": 1990656}, "11": {"total": 12595072, "nnz": 4307894, "linear_total": 12582912, "linear_nnz": 4298752, "linear_attention_total": 4194304, "linear_attention_nnz": 2091008, "linear_dense_total": 8388608, "linear_dense_nnz": 2207744}, "12": {"total": 12595264, "nnz": 4008939, "linear_total": 12582912, "linear_nnz": 3999744, "linear_attention_total": 4194304, "linear_attention_nnz": 2011136, "linear_dense_total": 8388608, "linear_dense_nnz": 1988608}, "13": {"total": 12595072, "nnz": 4534288, "linear_total": 12582912, "linear_nnz": 4525056, "linear_attention_total": 4194304, "linear_attention_nnz": 2132992, "linear_dense_total": 8388608, "linear_dense_nnz": 2392064}, "14": {"total": 12595840, "nnz": 4441620, "linear_total": 12582912, "linear_nnz": 4431872, "linear_attention_total": 4194304, "linear_attention_nnz": 2490368, "linear_dense_total": 8388608, "linear_dense_nnz": 1941504}, "15": {"total": 12595840, "nnz": 4399635, "linear_total": 12582912, "linear_nnz": 4389888, "linear_attention_total": 4194304, "linear_attention_nnz": 2515968, "linear_dense_total": 8388608, "linear_dense_nnz": 1873920}, "16": {"total": 12596224, "nnz": 4989801, "linear_total": 12582912, "linear_nnz": 4979712, "linear_attention_total": 4194304, "linear_attention_nnz": 2995200, "linear_dense_total": 8388608, "linear_dense_nnz": 1984512}, "17": {"total": 12595648, "nnz": 4203737, "linear_total": 12582912, "linear_nnz": 4194304, "linear_attention_total": 4194304, "linear_attention_nnz": 2373632, "linear_dense_total": 8388608, "linear_dense_nnz": 1820672}, "18": {"total": 12595840, "nnz": 3897575, "linear_total": 12582912, "linear_nnz": 3888128, "linear_attention_total": 4194304, "linear_attention_nnz": 2628608, "linear_dense_total": 8388608, "linear_dense_nnz": 1259520}, "19": {"total": 12596032, "nnz": 2920315, "linear_total": 12582912, "linear_nnz": 2911232, "linear_attention_total": 4194304, "linear_attention_nnz": 2135040, "linear_dense_total": 8388608, "linear_dense_nnz": 776192}, "20": {"total": 12595648, "nnz": 2061560, "linear_total": 12582912, "linear_nnz": 2053120, "linear_attention_total": 4194304, "linear_attention_nnz": 1545216, "linear_dense_total": 8388608, "linear_dense_nnz": 507904}, "21": {"total": 12594496, "nnz": 875551, "linear_total": 12582912, "linear_nnz": 868352, "linear_attention_total": 4194304, "linear_attention_nnz": 608256, "linear_dense_total": 8388608, "linear_dense_nnz": 260096}, "22": {"total": 12594688, "nnz": 774281, "linear_total": 12582912, "linear_nnz": 766976, "linear_attention_total": 4194304, "linear_attention_nnz": 617472, "linear_dense_total": 8388608, "linear_dense_nnz": 149504}, "23": {"total": 12594880, "nnz": 819435, "linear_total": 12582912, "linear_nnz": 812032, "linear_attention_total": 4194304, "linear_attention_nnz": 527360, "linear_dense_total": 8388608, "linear_dense_nnz": 284672}}, "total_sparsity": 67.12738277806874, "linear_sparsity": 74.22993977864583, "pruned_heads": {"0": [0, 3, 4, 5, 9, 13, 15], "1": [0, 1, 7, 9, 10, 13, 14], "2": [0, 1, 4, 6, 7, 8, 9, 11, 12, 13, 14, 15], "3": [0, 3, 5, 6, 7, 8, 9, 13, 14, 15], "4": [0, 2, 3, 5, 6, 7, 9, 10, 11, 12, 14], "5": [3, 4, 5, 6, 7, 8, 9, 12, 13, 14, 15], "6": [0, 1, 2, 3, 4, 7, 8, 9, 10, 11, 12, 15], "7": [3, 5, 8, 9, 10, 11, 12, 13], "8": [3, 5, 7, 8, 9, 11, 12], "9": [0, 1, 2, 5, 6, 7, 13, 14], "10": [1, 2, 4, 5, 6, 8, 11, 13], "11": [2, 5, 6, 8, 12, 15], "12": [0, 2, 6, 9, 13], "13": [2, 3, 6, 10, 12, 15], "14": [10, 5], "15": [0, 9], "16": [], "17": [1, 4, 12], "18": [11, 4], "19": [5], "20": [0, 1, 12], "21": [0, 2, 3, 4, 8, 10, 11, 12, 15], "22": [0, 1, 3, 7, 9, 10, 11, 13], "23": [2, 4, 7, 8, 9, 10, 13]}}