Muennighoff commited on
Commit
4501ecf
1 Parent(s): e0718ff

Improved model trained with larger batch size

Browse files
added_tokens.json CHANGED
@@ -1 +1 @@
1
- {"<|extratoken_41|>": 50297, "<|extratoken_109|>": 50365, "<|extratoken_17|>": 50273, "<|extratoken_22|>": 50278, "<|extratoken_137|>": 50393, "<|extratoken_131|>": 50387, "<|extratoken_130|>": 50386, "<|extratoken_52|>": 50308, "<|extratoken_96|>": 50352, "<|extratoken_70|>": 50326, "<|extratoken_19|>": 50275, "<|extratoken_54|>": 50310, "<|extratoken_15|>": 50271, "<|extratoken_101|>": 50357, "<|extratoken_37|>": 50293, "<|extratoken_98|>": 50354, "<|extratoken_57|>": 50313, "<|extratoken_69|>": 50325, "<|extratoken_1|>": 50257, "<|extratoken_18|>": 50274, "<|extratoken_87|>": 50343, "<|extratoken_81|>": 50337, "<|extratoken_21|>": 50277, "<|extratoken_72|>": 50328, "<|extratoken_8|>": 50264, "<|extratoken_83|>": 50339, "<|extratoken_53|>": 50309, "<|extratoken_123|>": 50379, "<|extratoken_59|>": 50315, "<|extratoken_65|>": 50321, "<|extratoken_28|>": 50284, "<|extratoken_90|>": 50346, "<|extratoken_108|>": 50364, "<|extratoken_30|>": 50286, "<|extratoken_84|>": 50340, "<|extratoken_88|>": 50344, "<|extratoken_67|>": 50323, "<|extratoken_94|>": 50350, "<|extratoken_86|>": 50342, "<|extratoken_7|>": 50263, "<|extratoken_4|>": 50260, "<|extratoken_48|>": 50304, "<|extratoken_112|>": 50368, "<|extratoken_14|>": 50270, "<|extratoken_100|>": 50356, "<|extratoken_116|>": 50372, "<|extratoken_25|>": 50281, "<|extratoken_118|>": 50374, "<|extratoken_139|>": 50395, "<|extratoken_73|>": 50329, "<|extratoken_80|>": 50336, "<|extratoken_74|>": 50330, "<|extratoken_119|>": 50375, "<|extratoken_128|>": 50384, "<|extratoken_133|>": 50389, "<|extratoken_63|>": 50319, "<|extratoken_47|>": 50303, "<|extratoken_132|>": 50388, "<|extratoken_12|>": 50268, "<|extratoken_127|>": 50383, "<|extratoken_6|>": 50262, "<|extratoken_71|>": 50327, "<|extratoken_35|>": 50291, "<|extratoken_2|>": 50258, "<|extratoken_143|>": 50399, "<|extratoken_44|>": 50300, "<|extratoken_45|>": 50301, "<|extratoken_68|>": 50324, "<|extratoken_138|>": 50394, "<|extratoken_58|>": 50314, "<|extratoken_31|>": 50287, "<|extratoken_49|>": 50305, "<|extratoken_99|>": 50355, "<|extratoken_39|>": 50295, "<|extratoken_97|>": 50353, "<|extratoken_16|>": 50272, "<|extratoken_125|>": 50381, "<|extratoken_79|>": 50335, "<|extratoken_43|>": 50299, "<|extratoken_46|>": 50302, "<|extratoken_62|>": 50318, "<|extratoken_75|>": 50331, "<|extratoken_33|>": 50289, "<|extratoken_26|>": 50282, "<|extratoken_38|>": 50294, "<|extratoken_82|>": 50338, "<|extratoken_129|>": 50385, "<|extratoken_10|>": 50266, "<|extratoken_117|>": 50373, "<|extratoken_121|>": 50377, "<|extratoken_111|>": 50367, "<|extratoken_92|>": 50348, "<|extratoken_85|>": 50341, "<|extratoken_115|>": 50371, "<|extratoken_142|>": 50398, "<|extratoken_27|>": 50283, "<|extratoken_95|>": 50351, "<|extratoken_91|>": 50347, "<|extratoken_77|>": 50333, "<|extratoken_32|>": 50288, "<|extratoken_20|>": 50276, "<|extratoken_60|>": 50316, "<|extratoken_24|>": 50280, "<|extratoken_3|>": 50259, "<|extratoken_36|>": 50292, "<|extratoken_114|>": 50370, "<|extratoken_105|>": 50361, "<|extratoken_136|>": 50392, "<|extratoken_55|>": 50311, "<|extratoken_141|>": 50397, "<|extratoken_124|>": 50380, "<|extratoken_103|>": 50359, "<|extratoken_76|>": 50332, "<|extratoken_113|>": 50369, "<|extratoken_34|>": 50290, "<|extratoken_40|>": 50296, "<|extratoken_140|>": 50396, "<|extratoken_13|>": 50269, "<|extratoken_126|>": 50382, "<|extratoken_106|>": 50362, "<|extratoken_5|>": 50261, "<|extratoken_42|>": 50298, "<|extratoken_23|>": 50279, "<|extratoken_56|>": 50312, "<|extratoken_102|>": 50358, "<|extratoken_104|>": 50360, "<|extratoken_110|>": 50366, "<|extratoken_122|>": 50378, "<|extratoken_89|>": 50345, "<|extratoken_78|>": 50334, "<|extratoken_135|>": 50391, "<|extratoken_134|>": 50390, "<|extratoken_64|>": 50320, "<|extratoken_11|>": 50267, "<|extratoken_61|>": 50317, "<|extratoken_29|>": 50285, "<|extratoken_107|>": 50363, "<|extratoken_120|>": 50376, "<|extratoken_66|>": 50322, "<|extratoken_51|>": 50307, "<|extratoken_9|>": 50265, "<|extratoken_93|>": 50349, "<|extratoken_50|>": 50306}
1
+ {"<|extratoken_79|>": 50335, "<|extratoken_55|>": 50311, "<|extratoken_1|>": 50257, "<|extratoken_139|>": 50395, "<|extratoken_46|>": 50302, "<|extratoken_2|>": 50258, "<|extratoken_4|>": 50260, "<|extratoken_133|>": 50389, "<|extratoken_98|>": 50354, "<|extratoken_35|>": 50291, "<|extratoken_38|>": 50294, "<|extratoken_28|>": 50284, "<|extratoken_26|>": 50282, "<|extratoken_94|>": 50350, "<|extratoken_16|>": 50272, "<|extratoken_48|>": 50304, "<|extratoken_96|>": 50352, "<|extratoken_125|>": 50381, "<|extratoken_136|>": 50392, "<|extratoken_127|>": 50383, "<|extratoken_42|>": 50298, "<|extratoken_61|>": 50317, "<|extratoken_140|>": 50396, "<|extratoken_80|>": 50336, "<|extratoken_130|>": 50386, "<|extratoken_81|>": 50337, "<|extratoken_52|>": 50308, "<|extratoken_78|>": 50334, "<|extratoken_36|>": 50292, "<|extratoken_56|>": 50312, "<|extratoken_82|>": 50338, "<|extratoken_45|>": 50301, "<|extratoken_102|>": 50358, "<|extratoken_9|>": 50265, "<|extratoken_69|>": 50325, "<|extratoken_39|>": 50295, "<|extratoken_25|>": 50281, "<|extratoken_58|>": 50314, "<|extratoken_5|>": 50261, "<|extratoken_92|>": 50348, "<|extratoken_118|>": 50374, "<|extratoken_112|>": 50368, "<|extratoken_107|>": 50363, "<|extratoken_114|>": 50370, "<|extratoken_33|>": 50289, "<|extratoken_67|>": 50323, "<|extratoken_37|>": 50293, "<|extratoken_40|>": 50296, "<|extratoken_50|>": 50306, "<|extratoken_108|>": 50364, "<|extratoken_123|>": 50379, "<|extratoken_110|>": 50366, "<|extratoken_71|>": 50327, "<|extratoken_43|>": 50299, "<|extratoken_57|>": 50313, "<|extratoken_124|>": 50380, "<|extratoken_63|>": 50319, "<|extratoken_131|>": 50387, "<|extratoken_72|>": 50328, "<|extratoken_88|>": 50344, "<|extratoken_117|>": 50373, "<|extratoken_90|>": 50346, "<|extratoken_20|>": 50276, "<|extratoken_31|>": 50287, "<|extratoken_142|>": 50398, "<|extratoken_89|>": 50345, "<|extratoken_7|>": 50263, "<|extratoken_15|>": 50271, "<|extratoken_95|>": 50351, "<|extratoken_24|>": 50280, "<|extratoken_32|>": 50288, "<|extratoken_105|>": 50361, "<|extratoken_138|>": 50394, "<|extratoken_34|>": 50290, "<|extratoken_113|>": 50369, "<|extratoken_64|>": 50320, "<|extratoken_73|>": 50329, "<|extratoken_128|>": 50384, "<|extratoken_115|>": 50371, "<|extratoken_22|>": 50278, "<|extratoken_76|>": 50332, "<|extratoken_18|>": 50274, "<|extratoken_103|>": 50359, "<|extratoken_137|>": 50393, "<|extratoken_51|>": 50307, "<|extratoken_101|>": 50357, "<|extratoken_99|>": 50355, "<|extratoken_47|>": 50303, "<|extratoken_41|>": 50297, "<|extratoken_86|>": 50342, "<|extratoken_119|>": 50375, "<|extratoken_14|>": 50270, "<|extratoken_53|>": 50309, "<|extratoken_116|>": 50372, "<|extratoken_132|>": 50388, "<|extratoken_85|>": 50341, "<|extratoken_111|>": 50367, "<|extratoken_10|>": 50266, "<|extratoken_91|>": 50347, "<|extratoken_3|>": 50259, "<|extratoken_100|>": 50356, "<|extratoken_109|>": 50365, "<|extratoken_59|>": 50315, "<|extratoken_122|>": 50378, "<|extratoken_83|>": 50339, "<|extratoken_70|>": 50326, "<|extratoken_23|>": 50279, "<|extratoken_135|>": 50391, "<|extratoken_65|>": 50321, "<|extratoken_60|>": 50316, "<|extratoken_141|>": 50397, "<|extratoken_62|>": 50318, "<|extratoken_97|>": 50353, "<|extratoken_21|>": 50277, "<|extratoken_143|>": 50399, "<|extratoken_68|>": 50324, "<|extratoken_75|>": 50331, "<|extratoken_54|>": 50310, "<|extratoken_30|>": 50286, "<|extratoken_66|>": 50322, "<|extratoken_126|>": 50382, "<|extratoken_8|>": 50264, "<|extratoken_12|>": 50268, "<|extratoken_19|>": 50275, "<|extratoken_13|>": 50269, "<|extratoken_106|>": 50362, "<|extratoken_44|>": 50300, "<|extratoken_27|>": 50283, "<|extratoken_17|>": 50273, "<|extratoken_29|>": 50285, "<|extratoken_77|>": 50333, "<|extratoken_120|>": 50376, "<|extratoken_49|>": 50305, "<|extratoken_74|>": 50330, "<|extratoken_93|>": 50349, "<|extratoken_129|>": 50385, "<|extratoken_121|>": 50377, "<|extratoken_87|>": 50343, "<|extratoken_6|>": 50262, "<|extratoken_134|>": 50390, "<|extratoken_11|>": 50267, "<|extratoken_84|>": 50340, "<|extratoken_104|>": 50360}
config.json CHANGED
@@ -12,7 +12,6 @@
12
  "initializer_range": 0.02,
13
  "layer_norm_epsilon": 1e-05,
14
  "model_type": "gptj",
15
- "n_ctx": 2048,
16
  "n_embd": 4096,
17
  "n_head": 16,
18
  "n_inner": null,
@@ -34,9 +33,10 @@
34
  "temperature": 1.0
35
  }
36
  },
 
37
  "tokenizer_class": "GPT2Tokenizer",
38
  "torch_dtype": "float32",
39
- "transformers_version": "4.11.3",
40
  "use_cache": true,
41
  "vocab_size": 50400
42
  }
12
  "initializer_range": 0.02,
13
  "layer_norm_epsilon": 1e-05,
14
  "model_type": "gptj",
 
15
  "n_embd": 4096,
16
  "n_head": 16,
17
  "n_inner": null,
33
  "temperature": 1.0
34
  }
35
  },
36
+ "tie_word_embeddings": false,
37
  "tokenizer_class": "GPT2Tokenizer",
38
  "torch_dtype": "float32",
39
+ "transformers_version": "4.16.2",
40
  "use_cache": true,
41
  "vocab_size": 50400
42
  }
config_sentence_transformers.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
  "__version__": {
3
  "sentence_transformers": "2.1.0",
4
- "transformers": "4.11.3",
5
- "pytorch": "1.10.1"
6
  }
7
  }
1
  {
2
  "__version__": {
3
  "sentence_transformers": "2.1.0",
4
+ "transformers": "4.16.2",
5
+ "pytorch": "1.10.2"
6
  }
7
  }
eval/similarity_evaluation_sts-dev_results.csv CHANGED
@@ -1,12 +1,12 @@
1
  epoch,steps,cosine_pearson,cosine_spearman,euclidean_pearson,euclidean_spearman,manhattan_pearson,manhattan_spearman,dot_pearson,dot_spearman
2
- 0,9394,0.8667856977815679,0.8683593379338206,0.8295642056627444,0.8363884778489342,0.8297954592955538,0.8366900215048471,0.7273903009136995,0.7450463171103514
3
- 0,18788,0.8689264665368805,0.8730176667630773,0.8172574760443659,0.826681430241987,0.8174496565604266,0.826906202638698,0.7251340335813794,0.7494408699573307
4
- 0,28182,0.869626405230146,0.8746725718167937,0.8117418430598462,0.8233372731358232,0.8119217193234098,0.823456339054079,0.7156090695589337,0.745283587736306
5
- 0,37576,0.8721744243944177,0.876376158288002,0.8092960874069344,0.8208668853143776,0.8093684853534436,0.8207676236802162,0.7097850859694168,0.740015795854084
6
- 0,46970,0.8685973670816101,0.8735217151527351,0.7994400795758065,0.8141999661345705,0.7995773482164976,0.8142136245069611,0.7048865353912706,0.7374451001344114
7
- 0,56364,0.8670753990703621,0.8709915852972844,0.7950924657883385,0.8090357828645908,0.7952413489633985,0.809093359416678,0.7069143199588939,0.7401687532547431
8
- 0,65758,0.8680677166487298,0.872010642938237,0.7910244128044988,0.804831925878411,0.7911861404707381,0.8049586839727801,0.707650896879289,0.7411846721034997
9
- 0,75152,0.8671966956114382,0.8714106816711714,0.7916891431485983,0.8053307393890206,0.7918070185130721,0.8054928629924096,0.7031691980078598,0.7365621493487114
10
- 0,84546,0.8661075138525474,0.8700145205677993,0.7871377350626456,0.8006476471975226,0.7872536565066849,0.8006875885061543,0.7032941404752215,0.7383746132297418
11
- 0,93940,0.8669285782887518,0.8708788467855191,0.7882799653262015,0.8018519610168285,0.7884034221151361,0.8020195609532438,0.7023297544615408,0.7380559226189177
12
- 0,-1,0.8669294815942044,0.8708691243588981,0.7882804405199759,0.801845053715687,0.7884042369770401,0.8020289186109306,0.7023314160887122,0.7380687909551775
1
  epoch,steps,cosine_pearson,cosine_spearman,euclidean_pearson,euclidean_spearman,manhattan_pearson,manhattan_spearman,dot_pearson,dot_spearman
2
+ 0,440,0.8435273641017604,0.8377079784670888,0.82481463944944,0.8267013391986916,0.8245734655395294,0.8266982253175391,0.7118405287624352,0.7184243142976252
3
+ 0,880,0.869655877890635,0.8699497978705373,0.8475489850804985,0.851502012897531,0.8480305694810746,0.8519484438106,0.7605340581924763,0.7668055024902699
4
+ 0,1320,0.8737186584633364,0.8757787058077194,0.8447049801487592,0.8496915401178553,0.8451488935363771,0.8500748482115618,0.7587490112010467,0.7657501850360053
5
+ 0,1760,0.8758103158531061,0.8788304232162334,0.8443044082500283,0.8498558356049029,0.8446089524345526,0.8501245564309194,0.7573076594801015,0.7656188059417804
6
+ 0,2200,0.8757603773560414,0.8792368701159438,0.844592810525526,0.8505497222047579,0.8448456280424977,0.8506566564422554,0.7547551286308827,0.7642469420057187
7
+ 0,2640,0.8771726730697522,0.8804454225514022,0.8408614161951098,0.847412986943978,0.841027548237923,0.8473965581070175,0.7525579592233351,0.7633872474622564
8
+ 0,3080,0.8770552793721014,0.8801359383520588,0.8379448907941871,0.844976225475866,0.8380978554239149,0.8448671692120961,0.7530860299039706,0.7647709904100748
9
+ 0,3520,0.876965300656499,0.8801979277186818,0.8384943754109598,0.8457685534247953,0.8386650258071886,0.8459102670457251,0.749595497754286,0.761306105456643
10
+ 0,3960,0.8772147739649651,0.8802736644257527,0.837048131104366,0.8443104563331376,0.8371878911253698,0.8443443638298471,0.7497833919662191,0.7620948373176814
11
+ 0,4400,0.877205358988175,0.8803647143106513,0.8378002006921788,0.8450916401462623,0.8379340983093438,0.8450169763936685,0.7502390023903127,0.7623304300069568
12
+ 0,-1,0.8772025474537237,0.8803490221290035,0.837797030269612,0.8450876775100756,0.8379313728609171,0.8450205102039363,0.7502314765703098,0.762315748502163
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:51b02dde1fc8932ddb8abb409f7afb449bf889d7881f36eab6c4fab19dbd88d2
3
  size 23495139247
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:897e277fc169e3421eb8662b7445f61060b25349f854a5fb4ea102ce6360e85f
3
  size 23495139247
similarity_evaluation_sts-test_results.csv DELETED
@@ -1,2 +0,0 @@
1
- epoch,steps,cosine_pearson,cosine_spearman,euclidean_pearson,euclidean_spearman,manhattan_pearson,manhattan_spearman,dot_pearson,dot_spearman
2
- -1,-1,0.8545703732519089,0.8628560963867633,0.7953746046954051,0.803146877550554,0.7952608148592852,0.8032415734370247,0.5954996603580655,0.6054272543624561
 
 
tokenizer.json CHANGED
The diff for this file is too large to render. See raw diff