Upload folder using huggingface_hub
Browse files- config.json +744 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +1175 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,744 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"CTCTransformerModel"
|
| 4 |
+
],
|
| 5 |
+
"blank_id": 0,
|
| 6 |
+
"hidden_size": 384,
|
| 7 |
+
"id2label": {
|
| 8 |
+
"0": "<pad>",
|
| 9 |
+
"1": "|",
|
| 10 |
+
"2": "AA0",
|
| 11 |
+
"3": "AA1",
|
| 12 |
+
"4": "AA2",
|
| 13 |
+
"5": "AE0",
|
| 14 |
+
"6": "AE1",
|
| 15 |
+
"7": "AE2",
|
| 16 |
+
"8": "AH0",
|
| 17 |
+
"9": "AH1",
|
| 18 |
+
"10": "AH2",
|
| 19 |
+
"11": "AO0",
|
| 20 |
+
"12": "AO1",
|
| 21 |
+
"13": "AO2",
|
| 22 |
+
"14": "AW0",
|
| 23 |
+
"15": "AW1",
|
| 24 |
+
"16": "AW2",
|
| 25 |
+
"17": "AY0",
|
| 26 |
+
"18": "AY1",
|
| 27 |
+
"19": "AY2",
|
| 28 |
+
"20": "B",
|
| 29 |
+
"21": "CH",
|
| 30 |
+
"22": "D",
|
| 31 |
+
"23": "DH",
|
| 32 |
+
"24": "EH0",
|
| 33 |
+
"25": "EH1",
|
| 34 |
+
"26": "EH2",
|
| 35 |
+
"27": "ER0",
|
| 36 |
+
"28": "ER1",
|
| 37 |
+
"29": "ER2",
|
| 38 |
+
"30": "EY0",
|
| 39 |
+
"31": "EY1",
|
| 40 |
+
"32": "EY2",
|
| 41 |
+
"33": "F",
|
| 42 |
+
"34": "G",
|
| 43 |
+
"35": "HH",
|
| 44 |
+
"36": "IH0",
|
| 45 |
+
"37": "IH1",
|
| 46 |
+
"38": "IH2",
|
| 47 |
+
"39": "IY0",
|
| 48 |
+
"40": "IY1",
|
| 49 |
+
"41": "IY2",
|
| 50 |
+
"42": "JH",
|
| 51 |
+
"43": "K",
|
| 52 |
+
"44": "L",
|
| 53 |
+
"45": "M",
|
| 54 |
+
"46": "N",
|
| 55 |
+
"47": "NG",
|
| 56 |
+
"48": "Ng",
|
| 57 |
+
"49": "OW0",
|
| 58 |
+
"50": "OW1",
|
| 59 |
+
"51": "OW2",
|
| 60 |
+
"52": "OY0",
|
| 61 |
+
"53": "OY1",
|
| 62 |
+
"54": "OY2",
|
| 63 |
+
"55": "P",
|
| 64 |
+
"56": "R",
|
| 65 |
+
"57": "S",
|
| 66 |
+
"58": "SH",
|
| 67 |
+
"59": "T",
|
| 68 |
+
"60": "TH",
|
| 69 |
+
"61": "UH0",
|
| 70 |
+
"62": "UH1",
|
| 71 |
+
"63": "UH2",
|
| 72 |
+
"64": "UW0",
|
| 73 |
+
"65": "UW1",
|
| 74 |
+
"66": "UW2",
|
| 75 |
+
"67": "V",
|
| 76 |
+
"68": "W",
|
| 77 |
+
"69": "Y",
|
| 78 |
+
"70": "Z",
|
| 79 |
+
"71": "ZH",
|
| 80 |
+
"72": "aa",
|
| 81 |
+
"73": "aa_1",
|
| 82 |
+
"74": "aa_2",
|
| 83 |
+
"75": "aa_3",
|
| 84 |
+
"76": "aa_4",
|
| 85 |
+
"77": "aa_5",
|
| 86 |
+
"78": "aa_6",
|
| 87 |
+
"79": "aai",
|
| 88 |
+
"80": "aai_1",
|
| 89 |
+
"81": "aai_2",
|
| 90 |
+
"82": "aai_3",
|
| 91 |
+
"83": "aai_4",
|
| 92 |
+
"84": "aai_5",
|
| 93 |
+
"85": "aai_6",
|
| 94 |
+
"86": "aak",
|
| 95 |
+
"87": "aak_1",
|
| 96 |
+
"88": "aak_2",
|
| 97 |
+
"89": "aak_3",
|
| 98 |
+
"90": "aak_6",
|
| 99 |
+
"91": "aam",
|
| 100 |
+
"92": "aam_1",
|
| 101 |
+
"93": "aam_2",
|
| 102 |
+
"94": "aam_3",
|
| 103 |
+
"95": "aam_4",
|
| 104 |
+
"96": "aam_5",
|
| 105 |
+
"97": "aam_6",
|
| 106 |
+
"98": "aan_1",
|
| 107 |
+
"99": "aan_2",
|
| 108 |
+
"100": "aan_3",
|
| 109 |
+
"101": "aan_4",
|
| 110 |
+
"102": "aan_5",
|
| 111 |
+
"103": "aan_6",
|
| 112 |
+
"104": "aang_1",
|
| 113 |
+
"105": "aang_2",
|
| 114 |
+
"106": "aang_3",
|
| 115 |
+
"107": "aang_4",
|
| 116 |
+
"108": "aang_5",
|
| 117 |
+
"109": "aang_6",
|
| 118 |
+
"110": "aap",
|
| 119 |
+
"111": "aap_2",
|
| 120 |
+
"112": "aap_3",
|
| 121 |
+
"113": "aap_6",
|
| 122 |
+
"114": "aat",
|
| 123 |
+
"115": "aat_1",
|
| 124 |
+
"116": "aat_2",
|
| 125 |
+
"117": "aat_3",
|
| 126 |
+
"118": "aat_6",
|
| 127 |
+
"119": "aau",
|
| 128 |
+
"120": "aau_1",
|
| 129 |
+
"121": "aau_2",
|
| 130 |
+
"122": "aau_3",
|
| 131 |
+
"123": "aau_4",
|
| 132 |
+
"124": "aau_5",
|
| 133 |
+
"125": "aau_6",
|
| 134 |
+
"126": "ai",
|
| 135 |
+
"127": "ai_1",
|
| 136 |
+
"128": "ai_2",
|
| 137 |
+
"129": "ai_3",
|
| 138 |
+
"130": "ai_4",
|
| 139 |
+
"131": "ai_5",
|
| 140 |
+
"132": "ai_6",
|
| 141 |
+
"133": "ak",
|
| 142 |
+
"134": "ak_1",
|
| 143 |
+
"135": "ak_6",
|
| 144 |
+
"136": "am",
|
| 145 |
+
"137": "am_1",
|
| 146 |
+
"138": "am_2",
|
| 147 |
+
"139": "am_3",
|
| 148 |
+
"140": "am_4",
|
| 149 |
+
"141": "am_6",
|
| 150 |
+
"142": "an_1",
|
| 151 |
+
"143": "an_2",
|
| 152 |
+
"144": "an_3",
|
| 153 |
+
"145": "an_4",
|
| 154 |
+
"146": "an_5",
|
| 155 |
+
"147": "an_6",
|
| 156 |
+
"148": "ang_1",
|
| 157 |
+
"149": "ang_2",
|
| 158 |
+
"150": "ang_3",
|
| 159 |
+
"151": "ang_4",
|
| 160 |
+
"152": "ang_6",
|
| 161 |
+
"153": "ap_1",
|
| 162 |
+
"154": "ap_2",
|
| 163 |
+
"155": "ap_6",
|
| 164 |
+
"156": "at_1",
|
| 165 |
+
"157": "at_2",
|
| 166 |
+
"158": "at_4",
|
| 167 |
+
"159": "at_6",
|
| 168 |
+
"160": "au",
|
| 169 |
+
"161": "au_1",
|
| 170 |
+
"162": "au_2",
|
| 171 |
+
"163": "au_3",
|
| 172 |
+
"164": "au_4",
|
| 173 |
+
"165": "au_5",
|
| 174 |
+
"166": "au_6",
|
| 175 |
+
"167": "b",
|
| 176 |
+
"168": "c",
|
| 177 |
+
"169": "d",
|
| 178 |
+
"170": "e_1",
|
| 179 |
+
"171": "e_2",
|
| 180 |
+
"172": "e_3",
|
| 181 |
+
"173": "e_4",
|
| 182 |
+
"174": "e_5",
|
| 183 |
+
"175": "e_6",
|
| 184 |
+
"176": "ei_1",
|
| 185 |
+
"177": "ei_2",
|
| 186 |
+
"178": "ei_3",
|
| 187 |
+
"179": "ei_4",
|
| 188 |
+
"180": "ei_5",
|
| 189 |
+
"181": "ei_6",
|
| 190 |
+
"182": "ek_1",
|
| 191 |
+
"183": "ek_3",
|
| 192 |
+
"184": "ek_6",
|
| 193 |
+
"185": "eng_1",
|
| 194 |
+
"186": "eng_2",
|
| 195 |
+
"187": "eng_3",
|
| 196 |
+
"188": "eng_4",
|
| 197 |
+
"189": "eng_5",
|
| 198 |
+
"190": "eng_6",
|
| 199 |
+
"191": "eoi_1",
|
| 200 |
+
"192": "eoi_2",
|
| 201 |
+
"193": "eoi_3",
|
| 202 |
+
"194": "eoi_4",
|
| 203 |
+
"195": "eoi_5",
|
| 204 |
+
"196": "eoi_6",
|
| 205 |
+
"197": "eon_1",
|
| 206 |
+
"198": "eon_2",
|
| 207 |
+
"199": "eon_3",
|
| 208 |
+
"200": "eon_4",
|
| 209 |
+
"201": "eon_5",
|
| 210 |
+
"202": "eon_6",
|
| 211 |
+
"203": "eot_1",
|
| 212 |
+
"204": "eot_2",
|
| 213 |
+
"205": "eot_6",
|
| 214 |
+
"206": "eu_6",
|
| 215 |
+
"207": "f",
|
| 216 |
+
"208": "g",
|
| 217 |
+
"209": "gw",
|
| 218 |
+
"210": "h",
|
| 219 |
+
"211": "i_1",
|
| 220 |
+
"212": "i_2",
|
| 221 |
+
"213": "i_3",
|
| 222 |
+
"214": "i_4",
|
| 223 |
+
"215": "i_5",
|
| 224 |
+
"216": "i_6",
|
| 225 |
+
"217": "ik_1",
|
| 226 |
+
"218": "ik_4",
|
| 227 |
+
"219": "ik_6",
|
| 228 |
+
"220": "im_1",
|
| 229 |
+
"221": "im_2",
|
| 230 |
+
"222": "im_3",
|
| 231 |
+
"223": "im_4",
|
| 232 |
+
"224": "im_5",
|
| 233 |
+
"225": "im_6",
|
| 234 |
+
"226": "in_1",
|
| 235 |
+
"227": "in_2",
|
| 236 |
+
"228": "in_3",
|
| 237 |
+
"229": "in_4",
|
| 238 |
+
"230": "in_5",
|
| 239 |
+
"231": "in_6",
|
| 240 |
+
"232": "ing_1",
|
| 241 |
+
"233": "ing_2",
|
| 242 |
+
"234": "ing_3",
|
| 243 |
+
"235": "ing_4",
|
| 244 |
+
"236": "ing_5",
|
| 245 |
+
"237": "ing_6",
|
| 246 |
+
"238": "ip_2",
|
| 247 |
+
"239": "ip_3",
|
| 248 |
+
"240": "ip_6",
|
| 249 |
+
"241": "it_1",
|
| 250 |
+
"242": "it_3",
|
| 251 |
+
"243": "it_6",
|
| 252 |
+
"244": "iu_1",
|
| 253 |
+
"245": "iu_2",
|
| 254 |
+
"246": "iu_3",
|
| 255 |
+
"247": "iu_4",
|
| 256 |
+
"248": "iu_5",
|
| 257 |
+
"249": "iu_6",
|
| 258 |
+
"250": "j",
|
| 259 |
+
"251": "k",
|
| 260 |
+
"252": "kw",
|
| 261 |
+
"253": "l",
|
| 262 |
+
"254": "m",
|
| 263 |
+
"255": "n",
|
| 264 |
+
"256": "ng",
|
| 265 |
+
"257": "o",
|
| 266 |
+
"258": "o_1",
|
| 267 |
+
"259": "o_2",
|
| 268 |
+
"260": "o_3",
|
| 269 |
+
"261": "o_4",
|
| 270 |
+
"262": "o_5",
|
| 271 |
+
"263": "o_6",
|
| 272 |
+
"264": "oe_1",
|
| 273 |
+
"265": "oe_2",
|
| 274 |
+
"266": "oe_3",
|
| 275 |
+
"267": "oek_2",
|
| 276 |
+
"268": "oek_3",
|
| 277 |
+
"269": "oek_4",
|
| 278 |
+
"270": "oek_6",
|
| 279 |
+
"271": "oeng_1",
|
| 280 |
+
"272": "oeng_2",
|
| 281 |
+
"273": "oeng_3",
|
| 282 |
+
"274": "oeng_4",
|
| 283 |
+
"275": "oeng_5",
|
| 284 |
+
"276": "oeng_6",
|
| 285 |
+
"277": "oi",
|
| 286 |
+
"278": "oi_1",
|
| 287 |
+
"279": "oi_2",
|
| 288 |
+
"280": "oi_3",
|
| 289 |
+
"281": "oi_4",
|
| 290 |
+
"282": "oi_5",
|
| 291 |
+
"283": "oi_6",
|
| 292 |
+
"284": "ok",
|
| 293 |
+
"285": "ok_1",
|
| 294 |
+
"286": "ok_2",
|
| 295 |
+
"287": "ok_3",
|
| 296 |
+
"288": "ok_6",
|
| 297 |
+
"289": "on",
|
| 298 |
+
"290": "on_1",
|
| 299 |
+
"291": "on_2",
|
| 300 |
+
"292": "on_3",
|
| 301 |
+
"293": "on_4",
|
| 302 |
+
"294": "on_6",
|
| 303 |
+
"295": "ong_1",
|
| 304 |
+
"296": "ong_2",
|
| 305 |
+
"297": "ong_3",
|
| 306 |
+
"298": "ong_4",
|
| 307 |
+
"299": "ong_5",
|
| 308 |
+
"300": "ong_6",
|
| 309 |
+
"301": "ot_3",
|
| 310 |
+
"302": "ou",
|
| 311 |
+
"303": "ou_1",
|
| 312 |
+
"304": "ou_2",
|
| 313 |
+
"305": "ou_3",
|
| 314 |
+
"306": "ou_4",
|
| 315 |
+
"307": "ou_5",
|
| 316 |
+
"308": "ou_6",
|
| 317 |
+
"309": "p",
|
| 318 |
+
"310": "s",
|
| 319 |
+
"311": "t",
|
| 320 |
+
"312": "u_1",
|
| 321 |
+
"313": "u_2",
|
| 322 |
+
"314": "u_3",
|
| 323 |
+
"315": "u_4",
|
| 324 |
+
"316": "u_5",
|
| 325 |
+
"317": "u_6",
|
| 326 |
+
"318": "ui_1",
|
| 327 |
+
"319": "ui_2",
|
| 328 |
+
"320": "ui_3",
|
| 329 |
+
"321": "ui_4",
|
| 330 |
+
"322": "ui_5",
|
| 331 |
+
"323": "ui_6",
|
| 332 |
+
"324": "uk",
|
| 333 |
+
"325": "uk_1",
|
| 334 |
+
"326": "uk_2",
|
| 335 |
+
"327": "uk_6",
|
| 336 |
+
"328": "un_1",
|
| 337 |
+
"329": "un_2",
|
| 338 |
+
"330": "un_3",
|
| 339 |
+
"331": "un_4",
|
| 340 |
+
"332": "un_5",
|
| 341 |
+
"333": "un_6",
|
| 342 |
+
"334": "ung",
|
| 343 |
+
"335": "ung_1",
|
| 344 |
+
"336": "ung_2",
|
| 345 |
+
"337": "ung_3",
|
| 346 |
+
"338": "ung_4",
|
| 347 |
+
"339": "ung_5",
|
| 348 |
+
"340": "ung_6",
|
| 349 |
+
"341": "ut_1",
|
| 350 |
+
"342": "ut_3",
|
| 351 |
+
"343": "ut_6",
|
| 352 |
+
"344": "w",
|
| 353 |
+
"345": "yu_1",
|
| 354 |
+
"346": "yu_2",
|
| 355 |
+
"347": "yu_3",
|
| 356 |
+
"348": "yu_4",
|
| 357 |
+
"349": "yu_5",
|
| 358 |
+
"350": "yu_6",
|
| 359 |
+
"351": "yun_1",
|
| 360 |
+
"352": "yun_2",
|
| 361 |
+
"353": "yun_3",
|
| 362 |
+
"354": "yun_4",
|
| 363 |
+
"355": "yun_5",
|
| 364 |
+
"356": "yun_6",
|
| 365 |
+
"357": "yut_1",
|
| 366 |
+
"358": "yut_2",
|
| 367 |
+
"359": "yut_3",
|
| 368 |
+
"360": "yut_4",
|
| 369 |
+
"361": "yut_6",
|
| 370 |
+
"362": "z"
|
| 371 |
+
},
|
| 372 |
+
"intermediate_size": 1024,
|
| 373 |
+
"label2id": {
|
| 374 |
+
"<pad>": 0,
|
| 375 |
+
"AA0": 2,
|
| 376 |
+
"AA1": 3,
|
| 377 |
+
"AA2": 4,
|
| 378 |
+
"AE0": 5,
|
| 379 |
+
"AE1": 6,
|
| 380 |
+
"AE2": 7,
|
| 381 |
+
"AH0": 8,
|
| 382 |
+
"AH1": 9,
|
| 383 |
+
"AH2": 10,
|
| 384 |
+
"AO0": 11,
|
| 385 |
+
"AO1": 12,
|
| 386 |
+
"AO2": 13,
|
| 387 |
+
"AW0": 14,
|
| 388 |
+
"AW1": 15,
|
| 389 |
+
"AW2": 16,
|
| 390 |
+
"AY0": 17,
|
| 391 |
+
"AY1": 18,
|
| 392 |
+
"AY2": 19,
|
| 393 |
+
"B": 20,
|
| 394 |
+
"CH": 21,
|
| 395 |
+
"D": 22,
|
| 396 |
+
"DH": 23,
|
| 397 |
+
"EH0": 24,
|
| 398 |
+
"EH1": 25,
|
| 399 |
+
"EH2": 26,
|
| 400 |
+
"ER0": 27,
|
| 401 |
+
"ER1": 28,
|
| 402 |
+
"ER2": 29,
|
| 403 |
+
"EY0": 30,
|
| 404 |
+
"EY1": 31,
|
| 405 |
+
"EY2": 32,
|
| 406 |
+
"F": 33,
|
| 407 |
+
"G": 34,
|
| 408 |
+
"HH": 35,
|
| 409 |
+
"IH0": 36,
|
| 410 |
+
"IH1": 37,
|
| 411 |
+
"IH2": 38,
|
| 412 |
+
"IY0": 39,
|
| 413 |
+
"IY1": 40,
|
| 414 |
+
"IY2": 41,
|
| 415 |
+
"JH": 42,
|
| 416 |
+
"K": 43,
|
| 417 |
+
"L": 44,
|
| 418 |
+
"M": 45,
|
| 419 |
+
"N": 46,
|
| 420 |
+
"NG": 47,
|
| 421 |
+
"Ng": 48,
|
| 422 |
+
"OW0": 49,
|
| 423 |
+
"OW1": 50,
|
| 424 |
+
"OW2": 51,
|
| 425 |
+
"OY0": 52,
|
| 426 |
+
"OY1": 53,
|
| 427 |
+
"OY2": 54,
|
| 428 |
+
"P": 55,
|
| 429 |
+
"R": 56,
|
| 430 |
+
"S": 57,
|
| 431 |
+
"SH": 58,
|
| 432 |
+
"T": 59,
|
| 433 |
+
"TH": 60,
|
| 434 |
+
"UH0": 61,
|
| 435 |
+
"UH1": 62,
|
| 436 |
+
"UH2": 63,
|
| 437 |
+
"UW0": 64,
|
| 438 |
+
"UW1": 65,
|
| 439 |
+
"UW2": 66,
|
| 440 |
+
"V": 67,
|
| 441 |
+
"W": 68,
|
| 442 |
+
"Y": 69,
|
| 443 |
+
"Z": 70,
|
| 444 |
+
"ZH": 71,
|
| 445 |
+
"aa": 72,
|
| 446 |
+
"aa_1": 73,
|
| 447 |
+
"aa_2": 74,
|
| 448 |
+
"aa_3": 75,
|
| 449 |
+
"aa_4": 76,
|
| 450 |
+
"aa_5": 77,
|
| 451 |
+
"aa_6": 78,
|
| 452 |
+
"aai": 79,
|
| 453 |
+
"aai_1": 80,
|
| 454 |
+
"aai_2": 81,
|
| 455 |
+
"aai_3": 82,
|
| 456 |
+
"aai_4": 83,
|
| 457 |
+
"aai_5": 84,
|
| 458 |
+
"aai_6": 85,
|
| 459 |
+
"aak": 86,
|
| 460 |
+
"aak_1": 87,
|
| 461 |
+
"aak_2": 88,
|
| 462 |
+
"aak_3": 89,
|
| 463 |
+
"aak_6": 90,
|
| 464 |
+
"aam": 91,
|
| 465 |
+
"aam_1": 92,
|
| 466 |
+
"aam_2": 93,
|
| 467 |
+
"aam_3": 94,
|
| 468 |
+
"aam_4": 95,
|
| 469 |
+
"aam_5": 96,
|
| 470 |
+
"aam_6": 97,
|
| 471 |
+
"aan_1": 98,
|
| 472 |
+
"aan_2": 99,
|
| 473 |
+
"aan_3": 100,
|
| 474 |
+
"aan_4": 101,
|
| 475 |
+
"aan_5": 102,
|
| 476 |
+
"aan_6": 103,
|
| 477 |
+
"aang_1": 104,
|
| 478 |
+
"aang_2": 105,
|
| 479 |
+
"aang_3": 106,
|
| 480 |
+
"aang_4": 107,
|
| 481 |
+
"aang_5": 108,
|
| 482 |
+
"aang_6": 109,
|
| 483 |
+
"aap": 110,
|
| 484 |
+
"aap_2": 111,
|
| 485 |
+
"aap_3": 112,
|
| 486 |
+
"aap_6": 113,
|
| 487 |
+
"aat": 114,
|
| 488 |
+
"aat_1": 115,
|
| 489 |
+
"aat_2": 116,
|
| 490 |
+
"aat_3": 117,
|
| 491 |
+
"aat_6": 118,
|
| 492 |
+
"aau": 119,
|
| 493 |
+
"aau_1": 120,
|
| 494 |
+
"aau_2": 121,
|
| 495 |
+
"aau_3": 122,
|
| 496 |
+
"aau_4": 123,
|
| 497 |
+
"aau_5": 124,
|
| 498 |
+
"aau_6": 125,
|
| 499 |
+
"ai": 126,
|
| 500 |
+
"ai_1": 127,
|
| 501 |
+
"ai_2": 128,
|
| 502 |
+
"ai_3": 129,
|
| 503 |
+
"ai_4": 130,
|
| 504 |
+
"ai_5": 131,
|
| 505 |
+
"ai_6": 132,
|
| 506 |
+
"ak": 133,
|
| 507 |
+
"ak_1": 134,
|
| 508 |
+
"ak_6": 135,
|
| 509 |
+
"am": 136,
|
| 510 |
+
"am_1": 137,
|
| 511 |
+
"am_2": 138,
|
| 512 |
+
"am_3": 139,
|
| 513 |
+
"am_4": 140,
|
| 514 |
+
"am_6": 141,
|
| 515 |
+
"an_1": 142,
|
| 516 |
+
"an_2": 143,
|
| 517 |
+
"an_3": 144,
|
| 518 |
+
"an_4": 145,
|
| 519 |
+
"an_5": 146,
|
| 520 |
+
"an_6": 147,
|
| 521 |
+
"ang_1": 148,
|
| 522 |
+
"ang_2": 149,
|
| 523 |
+
"ang_3": 150,
|
| 524 |
+
"ang_4": 151,
|
| 525 |
+
"ang_6": 152,
|
| 526 |
+
"ap_1": 153,
|
| 527 |
+
"ap_2": 154,
|
| 528 |
+
"ap_6": 155,
|
| 529 |
+
"at_1": 156,
|
| 530 |
+
"at_2": 157,
|
| 531 |
+
"at_4": 158,
|
| 532 |
+
"at_6": 159,
|
| 533 |
+
"au": 160,
|
| 534 |
+
"au_1": 161,
|
| 535 |
+
"au_2": 162,
|
| 536 |
+
"au_3": 163,
|
| 537 |
+
"au_4": 164,
|
| 538 |
+
"au_5": 165,
|
| 539 |
+
"au_6": 166,
|
| 540 |
+
"b": 167,
|
| 541 |
+
"c": 168,
|
| 542 |
+
"d": 169,
|
| 543 |
+
"e_1": 170,
|
| 544 |
+
"e_2": 171,
|
| 545 |
+
"e_3": 172,
|
| 546 |
+
"e_4": 173,
|
| 547 |
+
"e_5": 174,
|
| 548 |
+
"e_6": 175,
|
| 549 |
+
"ei_1": 176,
|
| 550 |
+
"ei_2": 177,
|
| 551 |
+
"ei_3": 178,
|
| 552 |
+
"ei_4": 179,
|
| 553 |
+
"ei_5": 180,
|
| 554 |
+
"ei_6": 181,
|
| 555 |
+
"ek_1": 182,
|
| 556 |
+
"ek_3": 183,
|
| 557 |
+
"ek_6": 184,
|
| 558 |
+
"eng_1": 185,
|
| 559 |
+
"eng_2": 186,
|
| 560 |
+
"eng_3": 187,
|
| 561 |
+
"eng_4": 188,
|
| 562 |
+
"eng_5": 189,
|
| 563 |
+
"eng_6": 190,
|
| 564 |
+
"eoi_1": 191,
|
| 565 |
+
"eoi_2": 192,
|
| 566 |
+
"eoi_3": 193,
|
| 567 |
+
"eoi_4": 194,
|
| 568 |
+
"eoi_5": 195,
|
| 569 |
+
"eoi_6": 196,
|
| 570 |
+
"eon_1": 197,
|
| 571 |
+
"eon_2": 198,
|
| 572 |
+
"eon_3": 199,
|
| 573 |
+
"eon_4": 200,
|
| 574 |
+
"eon_5": 201,
|
| 575 |
+
"eon_6": 202,
|
| 576 |
+
"eot_1": 203,
|
| 577 |
+
"eot_2": 204,
|
| 578 |
+
"eot_6": 205,
|
| 579 |
+
"eu_6": 206,
|
| 580 |
+
"f": 207,
|
| 581 |
+
"g": 208,
|
| 582 |
+
"gw": 209,
|
| 583 |
+
"h": 210,
|
| 584 |
+
"i_1": 211,
|
| 585 |
+
"i_2": 212,
|
| 586 |
+
"i_3": 213,
|
| 587 |
+
"i_4": 214,
|
| 588 |
+
"i_5": 215,
|
| 589 |
+
"i_6": 216,
|
| 590 |
+
"ik_1": 217,
|
| 591 |
+
"ik_4": 218,
|
| 592 |
+
"ik_6": 219,
|
| 593 |
+
"im_1": 220,
|
| 594 |
+
"im_2": 221,
|
| 595 |
+
"im_3": 222,
|
| 596 |
+
"im_4": 223,
|
| 597 |
+
"im_5": 224,
|
| 598 |
+
"im_6": 225,
|
| 599 |
+
"in_1": 226,
|
| 600 |
+
"in_2": 227,
|
| 601 |
+
"in_3": 228,
|
| 602 |
+
"in_4": 229,
|
| 603 |
+
"in_5": 230,
|
| 604 |
+
"in_6": 231,
|
| 605 |
+
"ing_1": 232,
|
| 606 |
+
"ing_2": 233,
|
| 607 |
+
"ing_3": 234,
|
| 608 |
+
"ing_4": 235,
|
| 609 |
+
"ing_5": 236,
|
| 610 |
+
"ing_6": 237,
|
| 611 |
+
"ip_2": 238,
|
| 612 |
+
"ip_3": 239,
|
| 613 |
+
"ip_6": 240,
|
| 614 |
+
"it_1": 241,
|
| 615 |
+
"it_3": 242,
|
| 616 |
+
"it_6": 243,
|
| 617 |
+
"iu_1": 244,
|
| 618 |
+
"iu_2": 245,
|
| 619 |
+
"iu_3": 246,
|
| 620 |
+
"iu_4": 247,
|
| 621 |
+
"iu_5": 248,
|
| 622 |
+
"iu_6": 249,
|
| 623 |
+
"j": 250,
|
| 624 |
+
"k": 251,
|
| 625 |
+
"kw": 252,
|
| 626 |
+
"l": 253,
|
| 627 |
+
"m": 254,
|
| 628 |
+
"n": 255,
|
| 629 |
+
"ng": 256,
|
| 630 |
+
"o": 257,
|
| 631 |
+
"o_1": 258,
|
| 632 |
+
"o_2": 259,
|
| 633 |
+
"o_3": 260,
|
| 634 |
+
"o_4": 261,
|
| 635 |
+
"o_5": 262,
|
| 636 |
+
"o_6": 263,
|
| 637 |
+
"oe_1": 264,
|
| 638 |
+
"oe_2": 265,
|
| 639 |
+
"oe_3": 266,
|
| 640 |
+
"oek_2": 267,
|
| 641 |
+
"oek_3": 268,
|
| 642 |
+
"oek_4": 269,
|
| 643 |
+
"oek_6": 270,
|
| 644 |
+
"oeng_1": 271,
|
| 645 |
+
"oeng_2": 272,
|
| 646 |
+
"oeng_3": 273,
|
| 647 |
+
"oeng_4": 274,
|
| 648 |
+
"oeng_5": 275,
|
| 649 |
+
"oeng_6": 276,
|
| 650 |
+
"oi": 277,
|
| 651 |
+
"oi_1": 278,
|
| 652 |
+
"oi_2": 279,
|
| 653 |
+
"oi_3": 280,
|
| 654 |
+
"oi_4": 281,
|
| 655 |
+
"oi_5": 282,
|
| 656 |
+
"oi_6": 283,
|
| 657 |
+
"ok": 284,
|
| 658 |
+
"ok_1": 285,
|
| 659 |
+
"ok_2": 286,
|
| 660 |
+
"ok_3": 287,
|
| 661 |
+
"ok_6": 288,
|
| 662 |
+
"on": 289,
|
| 663 |
+
"on_1": 290,
|
| 664 |
+
"on_2": 291,
|
| 665 |
+
"on_3": 292,
|
| 666 |
+
"on_4": 293,
|
| 667 |
+
"on_6": 294,
|
| 668 |
+
"ong_1": 295,
|
| 669 |
+
"ong_2": 296,
|
| 670 |
+
"ong_3": 297,
|
| 671 |
+
"ong_4": 298,
|
| 672 |
+
"ong_5": 299,
|
| 673 |
+
"ong_6": 300,
|
| 674 |
+
"ot_3": 301,
|
| 675 |
+
"ou": 302,
|
| 676 |
+
"ou_1": 303,
|
| 677 |
+
"ou_2": 304,
|
| 678 |
+
"ou_3": 305,
|
| 679 |
+
"ou_4": 306,
|
| 680 |
+
"ou_5": 307,
|
| 681 |
+
"ou_6": 308,
|
| 682 |
+
"p": 309,
|
| 683 |
+
"s": 310,
|
| 684 |
+
"t": 311,
|
| 685 |
+
"u_1": 312,
|
| 686 |
+
"u_2": 313,
|
| 687 |
+
"u_3": 314,
|
| 688 |
+
"u_4": 315,
|
| 689 |
+
"u_5": 316,
|
| 690 |
+
"u_6": 317,
|
| 691 |
+
"ui_1": 318,
|
| 692 |
+
"ui_2": 319,
|
| 693 |
+
"ui_3": 320,
|
| 694 |
+
"ui_4": 321,
|
| 695 |
+
"ui_5": 322,
|
| 696 |
+
"ui_6": 323,
|
| 697 |
+
"uk": 324,
|
| 698 |
+
"uk_1": 325,
|
| 699 |
+
"uk_2": 326,
|
| 700 |
+
"uk_6": 327,
|
| 701 |
+
"un_1": 328,
|
| 702 |
+
"un_2": 329,
|
| 703 |
+
"un_3": 330,
|
| 704 |
+
"un_4": 331,
|
| 705 |
+
"un_5": 332,
|
| 706 |
+
"un_6": 333,
|
| 707 |
+
"ung": 334,
|
| 708 |
+
"ung_1": 335,
|
| 709 |
+
"ung_2": 336,
|
| 710 |
+
"ung_3": 337,
|
| 711 |
+
"ung_4": 338,
|
| 712 |
+
"ung_5": 339,
|
| 713 |
+
"ung_6": 340,
|
| 714 |
+
"ut_1": 341,
|
| 715 |
+
"ut_3": 342,
|
| 716 |
+
"ut_6": 343,
|
| 717 |
+
"w": 344,
|
| 718 |
+
"yu_1": 345,
|
| 719 |
+
"yu_2": 346,
|
| 720 |
+
"yu_3": 347,
|
| 721 |
+
"yu_4": 348,
|
| 722 |
+
"yu_5": 349,
|
| 723 |
+
"yu_6": 350,
|
| 724 |
+
"yun_1": 351,
|
| 725 |
+
"yun_2": 352,
|
| 726 |
+
"yun_3": 353,
|
| 727 |
+
"yun_4": 354,
|
| 728 |
+
"yun_5": 355,
|
| 729 |
+
"yun_6": 356,
|
| 730 |
+
"yut_1": 357,
|
| 731 |
+
"yut_2": 358,
|
| 732 |
+
"yut_3": 359,
|
| 733 |
+
"yut_4": 360,
|
| 734 |
+
"yut_6": 361,
|
| 735 |
+
"z": 362,
|
| 736 |
+
"|": 1
|
| 737 |
+
},
|
| 738 |
+
"max_position_embeddings": 1024,
|
| 739 |
+
"num_attention_heads": 6,
|
| 740 |
+
"num_hidden_layers": 6,
|
| 741 |
+
"torch_dtype": "float32",
|
| 742 |
+
"transformers_version": "4.50.3",
|
| 743 |
+
"vocab_size": 6561
|
| 744 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:47947574cf1d0161d5e51a17465e3c5d6500449f7c2d421d04b8f3c2d241a750
|
| 3 |
+
size 43783580
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:042e96fb4c58ed8717496a4ebd1ed1ae01d8a45911590edb9a789e74209b6148
|
| 3 |
+
size 87614778
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9d613c42bc6269fa0dc958aa82cd8cd6dfce040534eaf9518e3193f2c7952dd0
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dfd9c5c96ae49d06833c053e6af6f37491082e4ff8aad12c8e68cdd6d2e4aa79
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1175 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 14.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 14728,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.09505703422053231,
|
| 14 |
+
"grad_norm": 0.9250678420066833,
|
| 15 |
+
"learning_rate": 3.95882818685669e-05,
|
| 16 |
+
"loss": 6.6663,
|
| 17 |
+
"step": 100
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.19011406844106463,
|
| 21 |
+
"grad_norm": 1.117205262184143,
|
| 22 |
+
"learning_rate": 7.91765637371338e-05,
|
| 23 |
+
"loss": 4.145,
|
| 24 |
+
"step": 200
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.28517110266159695,
|
| 28 |
+
"grad_norm": 1.3325448036193848,
|
| 29 |
+
"learning_rate": 0.00011876484560570071,
|
| 30 |
+
"loss": 3.8608,
|
| 31 |
+
"step": 300
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.38022813688212925,
|
| 35 |
+
"grad_norm": 0.9613128900527954,
|
| 36 |
+
"learning_rate": 0.0001583531274742676,
|
| 37 |
+
"loss": 3.768,
|
| 38 |
+
"step": 400
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.4752851711026616,
|
| 42 |
+
"grad_norm": 1.1198444366455078,
|
| 43 |
+
"learning_rate": 0.00019794140934283454,
|
| 44 |
+
"loss": 3.7093,
|
| 45 |
+
"step": 500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.5703422053231939,
|
| 49 |
+
"grad_norm": 1.0210366249084473,
|
| 50 |
+
"learning_rate": 0.00023752969121140142,
|
| 51 |
+
"loss": 3.6664,
|
| 52 |
+
"step": 600
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.6653992395437263,
|
| 56 |
+
"grad_norm": 0.9687440395355225,
|
| 57 |
+
"learning_rate": 0.00027711797307996834,
|
| 58 |
+
"loss": 3.5409,
|
| 59 |
+
"step": 700
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.7604562737642585,
|
| 63 |
+
"grad_norm": 1.4981633424758911,
|
| 64 |
+
"learning_rate": 0.0003167062549485352,
|
| 65 |
+
"loss": 3.3992,
|
| 66 |
+
"step": 800
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.8555133079847909,
|
| 70 |
+
"grad_norm": 0.8627603650093079,
|
| 71 |
+
"learning_rate": 0.00035629453681710216,
|
| 72 |
+
"loss": 3.1727,
|
| 73 |
+
"step": 900
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.9505703422053232,
|
| 77 |
+
"grad_norm": 0.9925593733787537,
|
| 78 |
+
"learning_rate": 0.0003958828186856691,
|
| 79 |
+
"loss": 2.9724,
|
| 80 |
+
"step": 1000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 1.0,
|
| 84 |
+
"eval_loss": 2.6960370540618896,
|
| 85 |
+
"eval_runtime": 3.7345,
|
| 86 |
+
"eval_samples_per_second": 1897.152,
|
| 87 |
+
"eval_steps_per_second": 118.622,
|
| 88 |
+
"step": 1052
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.0456273764258555,
|
| 92 |
+
"grad_norm": 0.9267619848251343,
|
| 93 |
+
"learning_rate": 0.00043547110055423594,
|
| 94 |
+
"loss": 2.7022,
|
| 95 |
+
"step": 1100
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.1406844106463878,
|
| 99 |
+
"grad_norm": 0.7666485905647278,
|
| 100 |
+
"learning_rate": 0.00047505938242280285,
|
| 101 |
+
"loss": 2.5641,
|
| 102 |
+
"step": 1200
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.2357414448669202,
|
| 106 |
+
"grad_norm": 0.5969619154930115,
|
| 107 |
+
"learning_rate": 0.0004990645699549983,
|
| 108 |
+
"loss": 2.5058,
|
| 109 |
+
"step": 1300
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.3307984790874525,
|
| 113 |
+
"grad_norm": 0.7782655358314514,
|
| 114 |
+
"learning_rate": 0.0004965363806441826,
|
| 115 |
+
"loss": 2.4665,
|
| 116 |
+
"step": 1400
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.4258555133079849,
|
| 120 |
+
"grad_norm": 0.8928040266036987,
|
| 121 |
+
"learning_rate": 0.000494008191333367,
|
| 122 |
+
"loss": 2.4311,
|
| 123 |
+
"step": 1500
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 1.5209125475285172,
|
| 127 |
+
"grad_norm": 0.8687949180603027,
|
| 128 |
+
"learning_rate": 0.0004914800020225515,
|
| 129 |
+
"loss": 2.3964,
|
| 130 |
+
"step": 1600
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.6159695817490496,
|
| 134 |
+
"grad_norm": 0.6245518922805786,
|
| 135 |
+
"learning_rate": 0.0004889518127117359,
|
| 136 |
+
"loss": 2.374,
|
| 137 |
+
"step": 1700
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.7110266159695817,
|
| 141 |
+
"grad_norm": 0.6903976202011108,
|
| 142 |
+
"learning_rate": 0.0004864236234009203,
|
| 143 |
+
"loss": 2.3606,
|
| 144 |
+
"step": 1800
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.806083650190114,
|
| 148 |
+
"grad_norm": 0.8996257781982422,
|
| 149 |
+
"learning_rate": 0.00048389543409010466,
|
| 150 |
+
"loss": 2.3376,
|
| 151 |
+
"step": 1900
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 1.9011406844106464,
|
| 155 |
+
"grad_norm": 0.734466016292572,
|
| 156 |
+
"learning_rate": 0.0004813672447792891,
|
| 157 |
+
"loss": 2.3226,
|
| 158 |
+
"step": 2000
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.9961977186311786,
|
| 162 |
+
"grad_norm": 0.6836825013160706,
|
| 163 |
+
"learning_rate": 0.0004788390554684735,
|
| 164 |
+
"loss": 2.3108,
|
| 165 |
+
"step": 2100
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.0,
|
| 169 |
+
"eval_loss": 2.285733461380005,
|
| 170 |
+
"eval_runtime": 3.623,
|
| 171 |
+
"eval_samples_per_second": 1955.579,
|
| 172 |
+
"eval_steps_per_second": 122.275,
|
| 173 |
+
"step": 2104
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 2.091254752851711,
|
| 177 |
+
"grad_norm": 0.5974160432815552,
|
| 178 |
+
"learning_rate": 0.0004763108661576579,
|
| 179 |
+
"loss": 2.2585,
|
| 180 |
+
"step": 2200
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 2.1863117870722433,
|
| 184 |
+
"grad_norm": 0.788093626499176,
|
| 185 |
+
"learning_rate": 0.0004737826768468423,
|
| 186 |
+
"loss": 2.264,
|
| 187 |
+
"step": 2300
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 2.2813688212927756,
|
| 191 |
+
"grad_norm": 0.7451100945472717,
|
| 192 |
+
"learning_rate": 0.00047125448753602674,
|
| 193 |
+
"loss": 2.2504,
|
| 194 |
+
"step": 2400
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 2.376425855513308,
|
| 198 |
+
"grad_norm": 0.6724629998207092,
|
| 199 |
+
"learning_rate": 0.0004687262982252111,
|
| 200 |
+
"loss": 2.2358,
|
| 201 |
+
"step": 2500
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 2.4714828897338403,
|
| 205 |
+
"grad_norm": 0.6606141924858093,
|
| 206 |
+
"learning_rate": 0.00046619810891439554,
|
| 207 |
+
"loss": 2.2301,
|
| 208 |
+
"step": 2600
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 2.5665399239543727,
|
| 212 |
+
"grad_norm": 0.6599621772766113,
|
| 213 |
+
"learning_rate": 0.0004636699196035799,
|
| 214 |
+
"loss": 2.2268,
|
| 215 |
+
"step": 2700
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 2.661596958174905,
|
| 219 |
+
"grad_norm": 0.6633493304252625,
|
| 220 |
+
"learning_rate": 0.00046114173029276434,
|
| 221 |
+
"loss": 2.2247,
|
| 222 |
+
"step": 2800
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 2.7566539923954374,
|
| 226 |
+
"grad_norm": 0.6308265328407288,
|
| 227 |
+
"learning_rate": 0.00045861354098194877,
|
| 228 |
+
"loss": 2.2221,
|
| 229 |
+
"step": 2900
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 2.8517110266159698,
|
| 233 |
+
"grad_norm": 0.6383451223373413,
|
| 234 |
+
"learning_rate": 0.00045608535167113314,
|
| 235 |
+
"loss": 2.2274,
|
| 236 |
+
"step": 3000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 2.9467680608365017,
|
| 240 |
+
"grad_norm": 0.61512291431427,
|
| 241 |
+
"learning_rate": 0.00045355716236031757,
|
| 242 |
+
"loss": 2.2067,
|
| 243 |
+
"step": 3100
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 3.0,
|
| 247 |
+
"eval_loss": 2.2008087635040283,
|
| 248 |
+
"eval_runtime": 3.5613,
|
| 249 |
+
"eval_samples_per_second": 1989.445,
|
| 250 |
+
"eval_steps_per_second": 124.393,
|
| 251 |
+
"step": 3156
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 3.041825095057034,
|
| 255 |
+
"grad_norm": 0.7461186647415161,
|
| 256 |
+
"learning_rate": 0.00045102897304950194,
|
| 257 |
+
"loss": 2.1882,
|
| 258 |
+
"step": 3200
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 3.1368821292775664,
|
| 262 |
+
"grad_norm": 0.6590662598609924,
|
| 263 |
+
"learning_rate": 0.00044850078373868637,
|
| 264 |
+
"loss": 2.1662,
|
| 265 |
+
"step": 3300
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 3.2319391634980987,
|
| 269 |
+
"grad_norm": 0.5832785964012146,
|
| 270 |
+
"learning_rate": 0.00044597259442787074,
|
| 271 |
+
"loss": 2.1603,
|
| 272 |
+
"step": 3400
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 3.326996197718631,
|
| 276 |
+
"grad_norm": 0.6356543898582458,
|
| 277 |
+
"learning_rate": 0.00044344440511705517,
|
| 278 |
+
"loss": 2.1601,
|
| 279 |
+
"step": 3500
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 3.4220532319391634,
|
| 283 |
+
"grad_norm": 0.7197031378746033,
|
| 284 |
+
"learning_rate": 0.0004409162158062396,
|
| 285 |
+
"loss": 2.1567,
|
| 286 |
+
"step": 3600
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 3.517110266159696,
|
| 290 |
+
"grad_norm": 0.5856086611747742,
|
| 291 |
+
"learning_rate": 0.00043838802649542397,
|
| 292 |
+
"loss": 2.1588,
|
| 293 |
+
"step": 3700
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 3.612167300380228,
|
| 297 |
+
"grad_norm": 0.6212655305862427,
|
| 298 |
+
"learning_rate": 0.00043585983718460834,
|
| 299 |
+
"loss": 2.1565,
|
| 300 |
+
"step": 3800
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 3.7072243346007605,
|
| 304 |
+
"grad_norm": 0.6765671968460083,
|
| 305 |
+
"learning_rate": 0.0004333316478737928,
|
| 306 |
+
"loss": 2.1667,
|
| 307 |
+
"step": 3900
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 3.802281368821293,
|
| 311 |
+
"grad_norm": 0.6720090508460999,
|
| 312 |
+
"learning_rate": 0.0004308034585629772,
|
| 313 |
+
"loss": 2.1675,
|
| 314 |
+
"step": 4000
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 3.897338403041825,
|
| 318 |
+
"grad_norm": 0.7150991559028625,
|
| 319 |
+
"learning_rate": 0.00042827526925216157,
|
| 320 |
+
"loss": 2.1474,
|
| 321 |
+
"step": 4100
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 3.9923954372623576,
|
| 325 |
+
"grad_norm": 0.5831249356269836,
|
| 326 |
+
"learning_rate": 0.00042574707994134605,
|
| 327 |
+
"loss": 2.1485,
|
| 328 |
+
"step": 4200
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 4.0,
|
| 332 |
+
"eval_loss": 2.15364408493042,
|
| 333 |
+
"eval_runtime": 3.644,
|
| 334 |
+
"eval_samples_per_second": 1944.292,
|
| 335 |
+
"eval_steps_per_second": 121.57,
|
| 336 |
+
"step": 4208
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 4.08745247148289,
|
| 340 |
+
"grad_norm": 0.6653150916099548,
|
| 341 |
+
"learning_rate": 0.0004232188906305304,
|
| 342 |
+
"loss": 2.0899,
|
| 343 |
+
"step": 4300
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 4.182509505703422,
|
| 347 |
+
"grad_norm": 0.7235066294670105,
|
| 348 |
+
"learning_rate": 0.0004206907013197148,
|
| 349 |
+
"loss": 2.0982,
|
| 350 |
+
"step": 4400
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 4.277566539923955,
|
| 354 |
+
"grad_norm": 0.7326545715332031,
|
| 355 |
+
"learning_rate": 0.0004181625120088992,
|
| 356 |
+
"loss": 2.1007,
|
| 357 |
+
"step": 4500
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 4.3726235741444865,
|
| 361 |
+
"grad_norm": 0.6236776113510132,
|
| 362 |
+
"learning_rate": 0.00041563432269808365,
|
| 363 |
+
"loss": 2.1031,
|
| 364 |
+
"step": 4600
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 4.467680608365019,
|
| 368 |
+
"grad_norm": 0.5669475197792053,
|
| 369 |
+
"learning_rate": 0.000413106133387268,
|
| 370 |
+
"loss": 2.1087,
|
| 371 |
+
"step": 4700
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 4.562737642585551,
|
| 375 |
+
"grad_norm": 0.5483006834983826,
|
| 376 |
+
"learning_rate": 0.00041057794407645245,
|
| 377 |
+
"loss": 2.1034,
|
| 378 |
+
"step": 4800
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 4.657794676806084,
|
| 382 |
+
"grad_norm": 0.5456926822662354,
|
| 383 |
+
"learning_rate": 0.0004080497547656369,
|
| 384 |
+
"loss": 2.1065,
|
| 385 |
+
"step": 4900
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 4.752851711026616,
|
| 389 |
+
"grad_norm": 0.9545803666114807,
|
| 390 |
+
"learning_rate": 0.00040552156545482125,
|
| 391 |
+
"loss": 2.1168,
|
| 392 |
+
"step": 5000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 4.847908745247148,
|
| 396 |
+
"grad_norm": 0.5378767251968384,
|
| 397 |
+
"learning_rate": 0.0004029933761440057,
|
| 398 |
+
"loss": 2.1107,
|
| 399 |
+
"step": 5100
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 4.942965779467681,
|
| 403 |
+
"grad_norm": 0.629880964756012,
|
| 404 |
+
"learning_rate": 0.00040046518683319005,
|
| 405 |
+
"loss": 2.0983,
|
| 406 |
+
"step": 5200
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 5.0,
|
| 410 |
+
"eval_loss": 2.132718801498413,
|
| 411 |
+
"eval_runtime": 3.6373,
|
| 412 |
+
"eval_samples_per_second": 1947.857,
|
| 413 |
+
"eval_steps_per_second": 121.793,
|
| 414 |
+
"step": 5260
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 5.038022813688213,
|
| 418 |
+
"grad_norm": 0.5900342464447021,
|
| 419 |
+
"learning_rate": 0.0003979369975223745,
|
| 420 |
+
"loss": 2.0758,
|
| 421 |
+
"step": 5300
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 5.133079847908745,
|
| 425 |
+
"grad_norm": 0.6181082129478455,
|
| 426 |
+
"learning_rate": 0.0003954088082115589,
|
| 427 |
+
"loss": 2.041,
|
| 428 |
+
"step": 5400
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 5.228136882129277,
|
| 432 |
+
"grad_norm": 0.6756412386894226,
|
| 433 |
+
"learning_rate": 0.0003928806189007433,
|
| 434 |
+
"loss": 2.0548,
|
| 435 |
+
"step": 5500
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 5.32319391634981,
|
| 439 |
+
"grad_norm": 0.6649320125579834,
|
| 440 |
+
"learning_rate": 0.0003903524295899277,
|
| 441 |
+
"loss": 2.0438,
|
| 442 |
+
"step": 5600
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 5.418250950570342,
|
| 446 |
+
"grad_norm": 0.5628513693809509,
|
| 447 |
+
"learning_rate": 0.00038782424027911214,
|
| 448 |
+
"loss": 2.0485,
|
| 449 |
+
"step": 5700
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 5.513307984790875,
|
| 453 |
+
"grad_norm": 0.6923677921295166,
|
| 454 |
+
"learning_rate": 0.0003852960509682965,
|
| 455 |
+
"loss": 2.063,
|
| 456 |
+
"step": 5800
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 5.608365019011407,
|
| 460 |
+
"grad_norm": 0.6819363236427307,
|
| 461 |
+
"learning_rate": 0.0003827678616574809,
|
| 462 |
+
"loss": 2.0618,
|
| 463 |
+
"step": 5900
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 5.7034220532319395,
|
| 467 |
+
"grad_norm": 0.6446284055709839,
|
| 468 |
+
"learning_rate": 0.00038023967234666537,
|
| 469 |
+
"loss": 2.0674,
|
| 470 |
+
"step": 6000
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 5.798479087452471,
|
| 474 |
+
"grad_norm": 0.6319680213928223,
|
| 475 |
+
"learning_rate": 0.00037771148303584974,
|
| 476 |
+
"loss": 2.061,
|
| 477 |
+
"step": 6100
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 5.893536121673003,
|
| 481 |
+
"grad_norm": 0.6318814754486084,
|
| 482 |
+
"learning_rate": 0.0003751832937250341,
|
| 483 |
+
"loss": 2.0656,
|
| 484 |
+
"step": 6200
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 5.988593155893536,
|
| 488 |
+
"grad_norm": 0.6261875033378601,
|
| 489 |
+
"learning_rate": 0.0003726551044142186,
|
| 490 |
+
"loss": 2.0663,
|
| 491 |
+
"step": 6300
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 6.0,
|
| 495 |
+
"eval_loss": 2.1098814010620117,
|
| 496 |
+
"eval_runtime": 3.698,
|
| 497 |
+
"eval_samples_per_second": 1915.889,
|
| 498 |
+
"eval_steps_per_second": 119.794,
|
| 499 |
+
"step": 6312
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 6.083650190114068,
|
| 503 |
+
"grad_norm": 0.6620230674743652,
|
| 504 |
+
"learning_rate": 0.00037012691510340297,
|
| 505 |
+
"loss": 1.9996,
|
| 506 |
+
"step": 6400
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 6.178707224334601,
|
| 510 |
+
"grad_norm": 1.0794607400894165,
|
| 511 |
+
"learning_rate": 0.00036759872579258734,
|
| 512 |
+
"loss": 2.0018,
|
| 513 |
+
"step": 6500
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 6.273764258555133,
|
| 517 |
+
"grad_norm": 1.372861385345459,
|
| 518 |
+
"learning_rate": 0.00036507053648177177,
|
| 519 |
+
"loss": 2.0059,
|
| 520 |
+
"step": 6600
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 6.3688212927756656,
|
| 524 |
+
"grad_norm": 0.5926664471626282,
|
| 525 |
+
"learning_rate": 0.0003625423471709562,
|
| 526 |
+
"loss": 2.012,
|
| 527 |
+
"step": 6700
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 6.4638783269961975,
|
| 531 |
+
"grad_norm": 0.7855852246284485,
|
| 532 |
+
"learning_rate": 0.00036001415786014057,
|
| 533 |
+
"loss": 2.0128,
|
| 534 |
+
"step": 6800
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 6.55893536121673,
|
| 538 |
+
"grad_norm": 0.6684075593948364,
|
| 539 |
+
"learning_rate": 0.000357485968549325,
|
| 540 |
+
"loss": 2.0221,
|
| 541 |
+
"step": 6900
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 6.653992395437262,
|
| 545 |
+
"grad_norm": 0.628013014793396,
|
| 546 |
+
"learning_rate": 0.00035495777923850937,
|
| 547 |
+
"loss": 2.0159,
|
| 548 |
+
"step": 7000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 6.749049429657795,
|
| 552 |
+
"grad_norm": 0.7943947911262512,
|
| 553 |
+
"learning_rate": 0.0003524295899276938,
|
| 554 |
+
"loss": 2.0223,
|
| 555 |
+
"step": 7100
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 6.844106463878327,
|
| 559 |
+
"grad_norm": 0.645799994468689,
|
| 560 |
+
"learning_rate": 0.0003499014006168782,
|
| 561 |
+
"loss": 2.0206,
|
| 562 |
+
"step": 7200
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 6.93916349809886,
|
| 566 |
+
"grad_norm": 0.6603648066520691,
|
| 567 |
+
"learning_rate": 0.0003473732113060626,
|
| 568 |
+
"loss": 2.0304,
|
| 569 |
+
"step": 7300
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 7.0,
|
| 573 |
+
"eval_loss": 2.099062919616699,
|
| 574 |
+
"eval_runtime": 3.631,
|
| 575 |
+
"eval_samples_per_second": 1951.251,
|
| 576 |
+
"eval_steps_per_second": 122.005,
|
| 577 |
+
"step": 7364
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 7.034220532319392,
|
| 581 |
+
"grad_norm": 0.6082973480224609,
|
| 582 |
+
"learning_rate": 0.000344845021995247,
|
| 583 |
+
"loss": 2.0039,
|
| 584 |
+
"step": 7400
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 7.129277566539924,
|
| 588 |
+
"grad_norm": 0.673995852470398,
|
| 589 |
+
"learning_rate": 0.0003423168326844314,
|
| 590 |
+
"loss": 1.9663,
|
| 591 |
+
"step": 7500
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 7.224334600760456,
|
| 595 |
+
"grad_norm": 0.675037682056427,
|
| 596 |
+
"learning_rate": 0.0003397886433736158,
|
| 597 |
+
"loss": 1.9696,
|
| 598 |
+
"step": 7600
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 7.319391634980988,
|
| 602 |
+
"grad_norm": 0.6488978266716003,
|
| 603 |
+
"learning_rate": 0.0003372604540628002,
|
| 604 |
+
"loss": 1.9701,
|
| 605 |
+
"step": 7700
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 7.414448669201521,
|
| 609 |
+
"grad_norm": 0.8255399465560913,
|
| 610 |
+
"learning_rate": 0.0003347322647519846,
|
| 611 |
+
"loss": 1.9654,
|
| 612 |
+
"step": 7800
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 7.509505703422053,
|
| 616 |
+
"grad_norm": 1.2661654949188232,
|
| 617 |
+
"learning_rate": 0.00033220407544116905,
|
| 618 |
+
"loss": 1.9736,
|
| 619 |
+
"step": 7900
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 7.604562737642586,
|
| 623 |
+
"grad_norm": 0.6545805335044861,
|
| 624 |
+
"learning_rate": 0.0003296758861303534,
|
| 625 |
+
"loss": 1.9783,
|
| 626 |
+
"step": 8000
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 7.699619771863118,
|
| 630 |
+
"grad_norm": 0.8890361189842224,
|
| 631 |
+
"learning_rate": 0.00032714769681953785,
|
| 632 |
+
"loss": 1.9807,
|
| 633 |
+
"step": 8100
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 7.79467680608365,
|
| 637 |
+
"grad_norm": 0.6547899842262268,
|
| 638 |
+
"learning_rate": 0.0003246195075087223,
|
| 639 |
+
"loss": 1.9723,
|
| 640 |
+
"step": 8200
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 7.889733840304182,
|
| 644 |
+
"grad_norm": 1.1239402294158936,
|
| 645 |
+
"learning_rate": 0.00032209131819790665,
|
| 646 |
+
"loss": 1.9734,
|
| 647 |
+
"step": 8300
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 7.984790874524715,
|
| 651 |
+
"grad_norm": 0.6624830961227417,
|
| 652 |
+
"learning_rate": 0.000319563128887091,
|
| 653 |
+
"loss": 1.9869,
|
| 654 |
+
"step": 8400
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 8.0,
|
| 658 |
+
"eval_loss": 2.1034328937530518,
|
| 659 |
+
"eval_runtime": 3.6013,
|
| 660 |
+
"eval_samples_per_second": 1967.337,
|
| 661 |
+
"eval_steps_per_second": 123.011,
|
| 662 |
+
"step": 8416
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 8.079847908745247,
|
| 666 |
+
"grad_norm": 0.6550971269607544,
|
| 667 |
+
"learning_rate": 0.0003170349395762755,
|
| 668 |
+
"loss": 1.9223,
|
| 669 |
+
"step": 8500
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 8.17490494296578,
|
| 673 |
+
"grad_norm": 0.660987138748169,
|
| 674 |
+
"learning_rate": 0.0003145067502654599,
|
| 675 |
+
"loss": 1.9245,
|
| 676 |
+
"step": 8600
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 8.269961977186313,
|
| 680 |
+
"grad_norm": 0.759884774684906,
|
| 681 |
+
"learning_rate": 0.00031197856095464425,
|
| 682 |
+
"loss": 1.9235,
|
| 683 |
+
"step": 8700
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 8.365019011406844,
|
| 687 |
+
"grad_norm": 0.9319919347763062,
|
| 688 |
+
"learning_rate": 0.00030945037164382874,
|
| 689 |
+
"loss": 1.9239,
|
| 690 |
+
"step": 8800
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 8.460076045627376,
|
| 694 |
+
"grad_norm": 0.6610597968101501,
|
| 695 |
+
"learning_rate": 0.0003069221823330131,
|
| 696 |
+
"loss": 1.928,
|
| 697 |
+
"step": 8900
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 8.55513307984791,
|
| 701 |
+
"grad_norm": 0.7076143622398376,
|
| 702 |
+
"learning_rate": 0.0003043939930221975,
|
| 703 |
+
"loss": 1.9289,
|
| 704 |
+
"step": 9000
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 8.65019011406844,
|
| 708 |
+
"grad_norm": 0.6368849873542786,
|
| 709 |
+
"learning_rate": 0.0003018658037113819,
|
| 710 |
+
"loss": 1.932,
|
| 711 |
+
"step": 9100
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 8.745247148288973,
|
| 715 |
+
"grad_norm": 0.7639185786247253,
|
| 716 |
+
"learning_rate": 0.00029933761440056634,
|
| 717 |
+
"loss": 1.9485,
|
| 718 |
+
"step": 9200
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 8.840304182509506,
|
| 722 |
+
"grad_norm": 1.0823330879211426,
|
| 723 |
+
"learning_rate": 0.0002968094250897507,
|
| 724 |
+
"loss": 1.9447,
|
| 725 |
+
"step": 9300
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 8.935361216730039,
|
| 729 |
+
"grad_norm": 0.8542035222053528,
|
| 730 |
+
"learning_rate": 0.00029428123577893514,
|
| 731 |
+
"loss": 1.942,
|
| 732 |
+
"step": 9400
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 9.0,
|
| 736 |
+
"eval_loss": 2.0947535037994385,
|
| 737 |
+
"eval_runtime": 3.6147,
|
| 738 |
+
"eval_samples_per_second": 1960.063,
|
| 739 |
+
"eval_steps_per_second": 122.556,
|
| 740 |
+
"step": 9468
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 9.03041825095057,
|
| 744 |
+
"grad_norm": 0.7601971626281738,
|
| 745 |
+
"learning_rate": 0.00029175304646811956,
|
| 746 |
+
"loss": 1.9243,
|
| 747 |
+
"step": 9500
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 9.125475285171103,
|
| 751 |
+
"grad_norm": 0.7461040019989014,
|
| 752 |
+
"learning_rate": 0.00028922485715730394,
|
| 753 |
+
"loss": 1.8704,
|
| 754 |
+
"step": 9600
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 9.220532319391635,
|
| 758 |
+
"grad_norm": 0.7719326019287109,
|
| 759 |
+
"learning_rate": 0.00028669666784648836,
|
| 760 |
+
"loss": 1.8832,
|
| 761 |
+
"step": 9700
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 9.315589353612168,
|
| 765 |
+
"grad_norm": 0.716136634349823,
|
| 766 |
+
"learning_rate": 0.00028416847853567274,
|
| 767 |
+
"loss": 1.8787,
|
| 768 |
+
"step": 9800
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 9.4106463878327,
|
| 772 |
+
"grad_norm": 0.6928532123565674,
|
| 773 |
+
"learning_rate": 0.00028164028922485717,
|
| 774 |
+
"loss": 1.8855,
|
| 775 |
+
"step": 9900
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 9.505703422053232,
|
| 779 |
+
"grad_norm": 0.7696681618690491,
|
| 780 |
+
"learning_rate": 0.0002791120999140416,
|
| 781 |
+
"loss": 1.8855,
|
| 782 |
+
"step": 10000
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 9.600760456273765,
|
| 786 |
+
"grad_norm": 0.8969391584396362,
|
| 787 |
+
"learning_rate": 0.00027658391060322597,
|
| 788 |
+
"loss": 1.9034,
|
| 789 |
+
"step": 10100
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 9.695817490494296,
|
| 793 |
+
"grad_norm": 0.8469530940055847,
|
| 794 |
+
"learning_rate": 0.00027405572129241034,
|
| 795 |
+
"loss": 1.8965,
|
| 796 |
+
"step": 10200
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 9.790874524714829,
|
| 800 |
+
"grad_norm": 0.7956866025924683,
|
| 801 |
+
"learning_rate": 0.0002715275319815948,
|
| 802 |
+
"loss": 1.9087,
|
| 803 |
+
"step": 10300
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 9.885931558935361,
|
| 807 |
+
"grad_norm": 0.8293343782424927,
|
| 808 |
+
"learning_rate": 0.0002689993426707792,
|
| 809 |
+
"loss": 1.9177,
|
| 810 |
+
"step": 10400
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 9.980988593155894,
|
| 814 |
+
"grad_norm": 0.7472631931304932,
|
| 815 |
+
"learning_rate": 0.00026647115335996357,
|
| 816 |
+
"loss": 1.9082,
|
| 817 |
+
"step": 10500
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 10.0,
|
| 821 |
+
"eval_loss": 2.097904920578003,
|
| 822 |
+
"eval_runtime": 3.5592,
|
| 823 |
+
"eval_samples_per_second": 1990.641,
|
| 824 |
+
"eval_steps_per_second": 124.468,
|
| 825 |
+
"step": 10520
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 10.076045627376425,
|
| 829 |
+
"grad_norm": 0.7787309288978577,
|
| 830 |
+
"learning_rate": 0.00026394296404914805,
|
| 831 |
+
"loss": 1.8393,
|
| 832 |
+
"step": 10600
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 10.171102661596958,
|
| 836 |
+
"grad_norm": 1.3328174352645874,
|
| 837 |
+
"learning_rate": 0.0002614147747383324,
|
| 838 |
+
"loss": 1.8283,
|
| 839 |
+
"step": 10700
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 10.26615969581749,
|
| 843 |
+
"grad_norm": 0.7740694284439087,
|
| 844 |
+
"learning_rate": 0.0002588865854275168,
|
| 845 |
+
"loss": 1.8422,
|
| 846 |
+
"step": 10800
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 10.361216730038024,
|
| 850 |
+
"grad_norm": 0.828940749168396,
|
| 851 |
+
"learning_rate": 0.0002563583961167012,
|
| 852 |
+
"loss": 1.8516,
|
| 853 |
+
"step": 10900
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 10.456273764258555,
|
| 857 |
+
"grad_norm": 0.751752495765686,
|
| 858 |
+
"learning_rate": 0.00025383020680588565,
|
| 859 |
+
"loss": 1.8624,
|
| 860 |
+
"step": 11000
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 10.551330798479087,
|
| 864 |
+
"grad_norm": 0.9940192103385925,
|
| 865 |
+
"learning_rate": 0.00025130201749507,
|
| 866 |
+
"loss": 1.8599,
|
| 867 |
+
"step": 11100
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 10.64638783269962,
|
| 871 |
+
"grad_norm": 0.8591569066047668,
|
| 872 |
+
"learning_rate": 0.00024877382818425445,
|
| 873 |
+
"loss": 1.8581,
|
| 874 |
+
"step": 11200
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 10.741444866920151,
|
| 878 |
+
"grad_norm": 0.7676281332969666,
|
| 879 |
+
"learning_rate": 0.0002462456388734388,
|
| 880 |
+
"loss": 1.8637,
|
| 881 |
+
"step": 11300
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 10.836501901140684,
|
| 885 |
+
"grad_norm": 0.7896871566772461,
|
| 886 |
+
"learning_rate": 0.00024371744956262325,
|
| 887 |
+
"loss": 1.8606,
|
| 888 |
+
"step": 11400
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 10.931558935361217,
|
| 892 |
+
"grad_norm": 0.8302274942398071,
|
| 893 |
+
"learning_rate": 0.00024118926025180765,
|
| 894 |
+
"loss": 1.8656,
|
| 895 |
+
"step": 11500
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 11.0,
|
| 899 |
+
"eval_loss": 2.0961618423461914,
|
| 900 |
+
"eval_runtime": 3.6362,
|
| 901 |
+
"eval_samples_per_second": 1948.473,
|
| 902 |
+
"eval_steps_per_second": 121.831,
|
| 903 |
+
"step": 11572
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 11.02661596958175,
|
| 907 |
+
"grad_norm": 0.8891871571540833,
|
| 908 |
+
"learning_rate": 0.00023866107094099208,
|
| 909 |
+
"loss": 1.8522,
|
| 910 |
+
"step": 11600
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 11.12167300380228,
|
| 914 |
+
"grad_norm": 0.7549653649330139,
|
| 915 |
+
"learning_rate": 0.00023613288163017645,
|
| 916 |
+
"loss": 1.7913,
|
| 917 |
+
"step": 11700
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 11.216730038022813,
|
| 921 |
+
"grad_norm": 0.8127674460411072,
|
| 922 |
+
"learning_rate": 0.00023360469231936088,
|
| 923 |
+
"loss": 1.8102,
|
| 924 |
+
"step": 11800
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 11.311787072243346,
|
| 928 |
+
"grad_norm": 0.841659426689148,
|
| 929 |
+
"learning_rate": 0.0002310765030085453,
|
| 930 |
+
"loss": 1.803,
|
| 931 |
+
"step": 11900
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"epoch": 11.406844106463879,
|
| 935 |
+
"grad_norm": 0.8460645079612732,
|
| 936 |
+
"learning_rate": 0.00022854831369772968,
|
| 937 |
+
"loss": 1.8201,
|
| 938 |
+
"step": 12000
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 11.50190114068441,
|
| 942 |
+
"grad_norm": 0.7932580709457397,
|
| 943 |
+
"learning_rate": 0.0002260201243869141,
|
| 944 |
+
"loss": 1.811,
|
| 945 |
+
"step": 12100
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 11.596958174904943,
|
| 949 |
+
"grad_norm": 0.8419378399848938,
|
| 950 |
+
"learning_rate": 0.0002234919350760985,
|
| 951 |
+
"loss": 1.8145,
|
| 952 |
+
"step": 12200
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 11.692015209125476,
|
| 956 |
+
"grad_norm": 0.8346748352050781,
|
| 957 |
+
"learning_rate": 0.0002209637457652829,
|
| 958 |
+
"loss": 1.8328,
|
| 959 |
+
"step": 12300
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 11.787072243346007,
|
| 963 |
+
"grad_norm": 1.019510269165039,
|
| 964 |
+
"learning_rate": 0.0002184355564544673,
|
| 965 |
+
"loss": 1.8257,
|
| 966 |
+
"step": 12400
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 11.88212927756654,
|
| 970 |
+
"grad_norm": 0.8175719976425171,
|
| 971 |
+
"learning_rate": 0.00021590736714365173,
|
| 972 |
+
"loss": 1.8274,
|
| 973 |
+
"step": 12500
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 11.977186311787072,
|
| 977 |
+
"grad_norm": 0.7476153373718262,
|
| 978 |
+
"learning_rate": 0.00021337917783283614,
|
| 979 |
+
"loss": 1.8361,
|
| 980 |
+
"step": 12600
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 12.0,
|
| 984 |
+
"eval_loss": 2.1029505729675293,
|
| 985 |
+
"eval_runtime": 3.5932,
|
| 986 |
+
"eval_samples_per_second": 1971.782,
|
| 987 |
+
"eval_steps_per_second": 123.289,
|
| 988 |
+
"step": 12624
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 12.072243346007605,
|
| 992 |
+
"grad_norm": 0.8637651205062866,
|
| 993 |
+
"learning_rate": 0.00021085098852202054,
|
| 994 |
+
"loss": 1.7684,
|
| 995 |
+
"step": 12700
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 12.167300380228136,
|
| 999 |
+
"grad_norm": 0.80800461769104,
|
| 1000 |
+
"learning_rate": 0.00020832279921120496,
|
| 1001 |
+
"loss": 1.7703,
|
| 1002 |
+
"step": 12800
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 12.262357414448669,
|
| 1006 |
+
"grad_norm": 1.0111021995544434,
|
| 1007 |
+
"learning_rate": 0.00020579460990038934,
|
| 1008 |
+
"loss": 1.7809,
|
| 1009 |
+
"step": 12900
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 12.357414448669202,
|
| 1013 |
+
"grad_norm": 0.8477798700332642,
|
| 1014 |
+
"learning_rate": 0.00020326642058957376,
|
| 1015 |
+
"loss": 1.7795,
|
| 1016 |
+
"step": 13000
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 12.452471482889734,
|
| 1020 |
+
"grad_norm": 0.8284028172492981,
|
| 1021 |
+
"learning_rate": 0.00020073823127875814,
|
| 1022 |
+
"loss": 1.7803,
|
| 1023 |
+
"step": 13100
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 12.547528517110266,
|
| 1027 |
+
"grad_norm": 0.7752136588096619,
|
| 1028 |
+
"learning_rate": 0.00019821004196794256,
|
| 1029 |
+
"loss": 1.7836,
|
| 1030 |
+
"step": 13200
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 12.642585551330798,
|
| 1034 |
+
"grad_norm": 0.8929184675216675,
|
| 1035 |
+
"learning_rate": 0.00019568185265712696,
|
| 1036 |
+
"loss": 1.7724,
|
| 1037 |
+
"step": 13300
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 12.737642585551331,
|
| 1041 |
+
"grad_norm": 0.8475900888442993,
|
| 1042 |
+
"learning_rate": 0.00019315366334631136,
|
| 1043 |
+
"loss": 1.7891,
|
| 1044 |
+
"step": 13400
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 12.832699619771864,
|
| 1048 |
+
"grad_norm": 0.9029939770698547,
|
| 1049 |
+
"learning_rate": 0.0001906254740354958,
|
| 1050 |
+
"loss": 1.7888,
|
| 1051 |
+
"step": 13500
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 12.927756653992395,
|
| 1055 |
+
"grad_norm": 0.841206967830658,
|
| 1056 |
+
"learning_rate": 0.0001880972847246802,
|
| 1057 |
+
"loss": 1.8005,
|
| 1058 |
+
"step": 13600
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 13.0,
|
| 1062 |
+
"eval_loss": 2.1176211833953857,
|
| 1063 |
+
"eval_runtime": 3.6226,
|
| 1064 |
+
"eval_samples_per_second": 1955.796,
|
| 1065 |
+
"eval_steps_per_second": 122.289,
|
| 1066 |
+
"step": 13676
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 13.022813688212928,
|
| 1070 |
+
"grad_norm": 0.786509096622467,
|
| 1071 |
+
"learning_rate": 0.0001855690954138646,
|
| 1072 |
+
"loss": 1.7784,
|
| 1073 |
+
"step": 13700
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 13.11787072243346,
|
| 1077 |
+
"grad_norm": 0.8644747734069824,
|
| 1078 |
+
"learning_rate": 0.000183040906103049,
|
| 1079 |
+
"loss": 1.7234,
|
| 1080 |
+
"step": 13800
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 13.212927756653992,
|
| 1084 |
+
"grad_norm": 0.8760172128677368,
|
| 1085 |
+
"learning_rate": 0.00018051271679223342,
|
| 1086 |
+
"loss": 1.7308,
|
| 1087 |
+
"step": 13900
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 13.307984790874524,
|
| 1091 |
+
"grad_norm": 0.7858941555023193,
|
| 1092 |
+
"learning_rate": 0.0001779845274814178,
|
| 1093 |
+
"loss": 1.7318,
|
| 1094 |
+
"step": 14000
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 13.403041825095057,
|
| 1098 |
+
"grad_norm": 0.8771238327026367,
|
| 1099 |
+
"learning_rate": 0.00017545633817060222,
|
| 1100 |
+
"loss": 1.7473,
|
| 1101 |
+
"step": 14100
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 13.49809885931559,
|
| 1105 |
+
"grad_norm": 0.8886803984642029,
|
| 1106 |
+
"learning_rate": 0.00017292814885978665,
|
| 1107 |
+
"loss": 1.7491,
|
| 1108 |
+
"step": 14200
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 13.593155893536121,
|
| 1112 |
+
"grad_norm": 0.8704127669334412,
|
| 1113 |
+
"learning_rate": 0.00017039995954897102,
|
| 1114 |
+
"loss": 1.7548,
|
| 1115 |
+
"step": 14300
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 13.688212927756654,
|
| 1119 |
+
"grad_norm": 1.2635705471038818,
|
| 1120 |
+
"learning_rate": 0.00016787177023815545,
|
| 1121 |
+
"loss": 1.7532,
|
| 1122 |
+
"step": 14400
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 13.783269961977187,
|
| 1126 |
+
"grad_norm": 0.9218750596046448,
|
| 1127 |
+
"learning_rate": 0.00016534358092733985,
|
| 1128 |
+
"loss": 1.7531,
|
| 1129 |
+
"step": 14500
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 13.87832699619772,
|
| 1133 |
+
"grad_norm": 0.9513919353485107,
|
| 1134 |
+
"learning_rate": 0.00016281539161652425,
|
| 1135 |
+
"loss": 1.7618,
|
| 1136 |
+
"step": 14600
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 13.97338403041825,
|
| 1140 |
+
"grad_norm": 1.010962963104248,
|
| 1141 |
+
"learning_rate": 0.00016028720230570865,
|
| 1142 |
+
"loss": 1.7646,
|
| 1143 |
+
"step": 14700
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 14.0,
|
| 1147 |
+
"eval_loss": 2.130631923675537,
|
| 1148 |
+
"eval_runtime": 3.6539,
|
| 1149 |
+
"eval_samples_per_second": 1938.998,
|
| 1150 |
+
"eval_steps_per_second": 121.239,
|
| 1151 |
+
"step": 14728
|
| 1152 |
+
}
|
| 1153 |
+
],
|
| 1154 |
+
"logging_steps": 100,
|
| 1155 |
+
"max_steps": 21040,
|
| 1156 |
+
"num_input_tokens_seen": 0,
|
| 1157 |
+
"num_train_epochs": 20,
|
| 1158 |
+
"save_steps": 500,
|
| 1159 |
+
"stateful_callbacks": {
|
| 1160 |
+
"TrainerControl": {
|
| 1161 |
+
"args": {
|
| 1162 |
+
"should_epoch_stop": false,
|
| 1163 |
+
"should_evaluate": false,
|
| 1164 |
+
"should_log": false,
|
| 1165 |
+
"should_save": true,
|
| 1166 |
+
"should_training_stop": false
|
| 1167 |
+
},
|
| 1168 |
+
"attributes": {}
|
| 1169 |
+
}
|
| 1170 |
+
},
|
| 1171 |
+
"total_flos": 5.6215157665850184e+16,
|
| 1172 |
+
"train_batch_size": 128,
|
| 1173 |
+
"trial_name": null,
|
| 1174 |
+
"trial_params": null
|
| 1175 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:83c747c7cd25c286914dbb5bbec4723cdf421e34ea221f411b85b150d0b96bd4
|
| 3 |
+
size 5304
|