File size: 5,386 Bytes
d67e821
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
{
  "retrieval": {
    "SIGMA_B": {
      "mrr10": 0.92306,
      "r1": 0.8856,
      "r5": 0.96956,
      "r10": 0.97886,
      "median_rank": 1.0,
      "pool_size": 22046
    },
    "CodeBERT_FT": {
      "mrr10": 0.96622,
      "r1": 0.94484,
      "r5": 0.99129,
      "r10": 0.99356,
      "median_rank": 1.0,
      "pool_size": 22046
    },
    "delta_mrr": -0.04316
  },
  "generation": {
    "sigma_gen_hist": {
      "tr_loss": [
        4.332911533272467,
        3.513244101578118,
        3.232151625534834,
        3.0188788402276594,
        2.876566329031201,
        2.774567126666569,
        2.7000688419086245,
        2.6459983390151094,
        2.61164868748713,
        2.5937092144469105
      ],
      "va_loss": [
        4.333631096980431,
        4.033719047488412,
        3.887238296438657,
        3.721125888444704,
        3.652845814161021,
        3.598194845310101,
        3.560626884405384,
        3.542548859641584,
        3.5358788442590083,
        3.5271108577827133
      ],
      "va_ppl": [
        76.2205489029526,
        56.470537811754056,
        48.77599550446955,
        41.31087937681806,
        38.58431374299519,
        36.53222856246678,
        35.18524731577227,
        34.5548825857037,
        34.32516793907655,
        34.02552090105466
      ]
    },
    "cbert_gen_hist": {
      "tr_loss": [
        4.357259073945283,
        3.5188005480055473,
        3.1955701593955923,
        2.988982406454519,
        2.858495755219526,
        2.7642144319277264,
        2.69454125002109,
        2.644378586365032,
        2.6121866509032623,
        2.5950699153240633
      ],
      "va_loss": [
        4.360623060350565,
        4.023048199407739,
        3.830441037275766,
        3.7134811096919123,
        3.6445059282146337,
        3.6011066965343077,
        3.5686047597707726,
        3.551744504028521,
        3.532704999526294,
        3.5293681702964337
      ],
      "va_ppl": [
        78.30590853372148,
        55.871152941678396,
        46.08285801115232,
        40.996270932086794,
        38.263863085402065,
        36.63876000351066,
        35.46707352597844,
        34.874102464608406,
        34.216397886548734,
        34.10241388827117
      ]
    },
    "gen_samples": [
      {
        "prompt": "Calculate the fibonacci number",
        "sigma": "def get_best(of, x):\n    \"\"\"Calculate the year of distribution.\"\"\"\n    if len(x, t.shape):\n     q = np.shape(cos_ind)\n    for t in range(x, n):\n        if s <= 0:\n ",
        "codebert": "def rot(self, X):\n        \"\"\"Return the increment and posterior of all details.\"\"\"\n        return self.bits(np(\"%s\")\n        return self.read_index_value(self._num)\n        else:\n   "
      },
      {
        "prompt": "Sort a list using quicksort algorithm",
        "sigma": "def _sqr_valid(self, p):\n    \"\"\"Build the input a region of shape,\n    a matrix in the grid(d = float)\n    # Find the lines to make a point.\n    # and '''\n\n\r\n    if isinstance(value, x1):\r\n   ",
        "codebert": "def _rc_words(self, node):\n    \"\"\"Returns the file when a dict.\"\"\"\n    if isinstance(log(\" not necessary\") == 'f'):\n     # self._match()\n  if _swith(\"The tag\" in self.format(self._case):\n   # Popon: d"
      },
      {
        "prompt": "Read a file and return its contents as string",
        "sigma": "def _format_to(path):\n    \"\"\"Helper a file from the doi into `file`.\n   Returns:\n   \"\"\"\n\n    # Note:\n    return raw_file['test__', 'f'\n    # Use ``object'\n    return (arg.strip)\n   # Runtime =",
        "codebert": "def _parse_file(self, filename):\n        \"\"\"\n        Render a single file object\n        \"\"\"\n        try:\n            for path in self.path['wargs']\n     "
      },
      {
        "prompt": "Convert a dictionary to a JSON string",
        "sigma": "def _code(self, placeholder):\n idd\"\"\"' if a string and return None. \"\"\"\n    if not None:\n    return '{_list':\n    return [int(node.join('OTO')\n   info = 'name['join(node', str)\n    data = '%s}\n ",
        "codebert": "def _read_names(self, type):\n    \"\"\"Convert a string with single style object.\n\n    Returns the value of `line` object and if not be a\n    \"\"\"\n\n    if self.values:\n     if not None:\n      # type in se"
      },
      {
        "prompt": "Find the maximum element in a list",
        "sigma": "def _info_field(self, value):\n    '''Returns a list for this is the packet\"\"\"\n    if not None:\n    return self._open(\n        return raw_value)\n    return not None:\n    # we coerree of the child\n   ",
        "codebert": "def get_field(self, node):\n    '''Return the raw tokens.\"\"\"\n    for key in value in node.items:\n    else = self._line_obj(index)\n    else: ast.value(self._text)\n    elif self.isfocol == 1:\n   "
      },
      {
        "prompt": "Calculate the factorial of a number",
        "sigma": "def calculate_p(self, width=None):\n    \"\"\"Calculate the water of a normalized time and\n    '''\n\n    n = np.term['sub1\n    return np.scplot_vON(x, x)\n    return self.fasttype(float)\n    return self",
        "codebert": "def compute(self):\n        \"\"\"\n        Updates the time-block over all one of a column\n        \"\"\"\n        s = self.build(self._coist, self.left_count)\n        if self.start"
      }
    ]
  }
}