v1-newq map3 0.6452

runs/v1-newq/config.json

@@ -0,0 +1,409 @@
+{
+ "backbone": "microsoft/deberta-v3-large",
+ "classes": [
+  "Correct",
+  "Neither",
+  "Adding_across",
+  "Adding_terms",
+  "Additive",
+  "Base_rate",
+  "Certainty",
+  "Definition",
+  "Denominator-only_change",
+  "Division",
+  "Duplication",
+  "Firstterm",
+  "FlipChange",
+  "Ignores_zeroes",
+  "Incomplete",
+  "Incorrect_equivalent_fraction_addition",
+  "Interior",
+  "Inverse_operation",
+  "Inversion",
+  "Irrelevant",
+  "Longer_is_bigger",
+  "Mult",
+  "Multiplying_by_4",
+  "Not_variable",
+  "Positive",
+  "Scale",
+  "Shorter_is_bigger",
+  "Subtraction",
+  "SwapDividend",
+  "Tacking",
+  "Unknowable",
+  "WNB",
+  "Whole_numbers_larger",
+  "Wrong_Fraction",
+  "Wrong_Operation",
+  "Wrong_fraction",
+  "Wrong_term"
+ ],
+ "max_len": 256,
+ "split": "questions [31772, 32835] held out",
+ "best_map3": 0.6452,
+ "log": [
+  {
+   "epoch": 0,
+   "train_loss": 0.9934,
+   "map3": 0.6452,
+   "acc": 0.5711,
+   "map3_masked": 0.6452,
+   "t": 790
+  },
+  {
+   "epoch": 1,
+   "train_loss": 0.4135,
+   "map3": 0.625,
+   "acc": 0.5306,
+   "map3_masked": 0.625,
+   "t": 1584
+  },
+  {
+   "epoch": 2,
+   "train_loss": 0.2981,
+   "map3": 0.6064,
+   "acc": 0.4919,
+   "map3_masked": 0.6064,
+   "t": 2376
+  }
+ ],
+ "questions": {
+  "31772": {
+   "text": "What fraction of the shape is not shaded? Give your answer in its simplest form. [Image: A triangle split into 9 equal smaller triangles. 6 of them are shaded.]",
+   "choices": [
+    "\\( \\frac{1}{3} \\)",
+    "\\( \\frac{3}{6} \\)",
+    "\\( \\frac{3}{8} \\)",
+    "\\( \\frac{3}{9} \\)"
+   ],
+   "right": "\\( \\frac{1}{3} \\)",
+   "classes": [
+    "Correct",
+    "Incomplete",
+    "Neither",
+    "WNB"
+   ],
+   "seen": []
+  },
+  "31774": {
+   "text": "Calculate \\( \\frac{1}{2} \\div 6 \\)",
+   "choices": [
+    "\\( 3 \\)",
+    "\\( \\frac{1}{12} \\)",
+    "\\( \\frac{1}{3} \\)",
+    "\\( \\frac{6}{2} \\)"
+   ],
+   "right": "\\( \\frac{1}{12} \\)",
+   "classes": [
+    "Correct",
+    "FlipChange",
+    "Mult",
+    "Neither",
+    "SwapDividend"
+   ],
+   "seen": [
+    "Correct",
+    "FlipChange",
+    "Mult",
+    "Neither",
+    "SwapDividend"
+   ]
+  },
+  "31777": {
+   "text": "A box contains \\( 120 \\) counters. The counters are red or blue. \\( \\frac{3}{5} \\) of the counters are red.\nHow many red counters are there?",
+   "choices": [
+    "\\( 24 \\)",
+    "\\( 48 \\)",
+    "\\( 60 \\)",
+    "\\( 72 \\)"
+   ],
+   "right": "\\( 72 \\)",
+   "classes": [
+    "Correct",
+    "Incomplete",
+    "Irrelevant",
+    "Neither",
+    "Wrong_Fraction"
+   ],
+   "seen": [
+    "Correct",
+    "Incomplete",
+    "Irrelevant",
+    "Neither",
+    "Wrong_Fraction"
+   ]
+  },
+  "31778": {
+   "text": "\\( \\frac{A}{10}=\\frac{9}{15} \\) What is the value of \\( A \\) ?",
+   "choices": [
+    "\\( 3 \\)",
+    "\\( 4 \\)",
+    "\\( 6 \\)",
+    "\\( 9 \\)"
+   ],
+   "right": "\\( 6 \\)",
+   "classes": [
+    "Additive",
+    "Correct",
+    "Irrelevant",
+    "Neither",
+    "WNB"
+   ],
+   "seen": [
+    "Additive",
+    "Correct",
+    "Irrelevant",
+    "Neither",
+    "WNB"
+   ]
+  },
+  "32829": {
+   "text": "\\( 2 y=24 \\) What is the value of \\( y \\) ?",
+   "choices": [
+    "\\( 12 \\)",
+    "\\( 22 \\)",
+    "\\( 4 \\)",
+    "\\( 48 \\)"
+   ],
+   "right": "\\( 12 \\)",
+   "classes": [
+    "Adding_terms",
+    "Correct",
+    "Inverse_operation",
+    "Neither",
+    "Not_variable"
+   ],
+   "seen": [
+    "Adding_terms",
+    "Correct",
+    "Inverse_operation",
+    "Neither",
+    "Not_variable"
+   ]
+  },
+  "32833": {
+   "text": "Calculate \\( \\frac{2}{3} \\times 5 \\)",
+   "choices": [
+    "\\( 3 \\frac{1}{3} \\)",
+    "\\( 5 \\frac{2}{3} \\)",
+    "\\( \\frac{10}{15} \\)",
+    "\\( \\frac{2}{15} \\)"
+   ],
+   "right": "\\( 3 \\frac{1}{3} \\)",
+   "classes": [
+    "Correct",
+    "Duplication",
+    "Inversion",
+    "Neither",
+    "Wrong_Operation"
+   ],
+   "seen": [
+    "Correct",
+    "Duplication",
+    "Inversion",
+    "Neither",
+    "Wrong_Operation"
+   ]
+  },
+  "32835": {
+   "text": "Which number is the greatest?",
+   "choices": [
+    "\\( 6 \\)",
+    "\\( 6.0001 \\)",
+    "\\( 6.079 \\)",
+    "\\( 6.2 \\)"
+   ],
+   "right": "\\( 6.2 \\)",
+   "classes": [
+    "Correct",
+    "Ignores_zeroes",
+    "Longer_is_bigger",
+    "Neither",
+    "Shorter_is_bigger",
+    "Whole_numbers_larger"
+   ],
+   "seen": []
+  },
+  "33471": {
+   "text": "A bag contains \\( 24 \\) yellow and green balls. \\( \\frac{3}{8} \\) of the balls are yellow. How many of the balls are green?",
+   "choices": [
+    "\\( 15 \\)",
+    "\\( 3 \\)",
+    "\\( 8 \\)",
+    "\\( 9 \\)"
+   ],
+   "right": "\\( 15 \\)",
+   "classes": [
+    "Correct",
+    "Incomplete",
+    "Neither",
+    "Wrong_fraction"
+   ],
+   "seen": [
+    "Correct",
+    "Incomplete",
+    "Neither",
+    "Wrong_fraction"
+   ]
+  },
+  "33472": {
+   "text": "\\( \\frac{1}{3}+\\frac{2}{5}= \\)",
+   "choices": [
+    "\\( \\frac{11}{15} \\)",
+    "\\( \\frac{11}{30} \\)",
+    "\\( \\frac{3}{15} \\)",
+    "\\( \\frac{3}{8} \\)"
+   ],
+   "right": "\\( \\frac{11}{15} \\)",
+   "classes": [
+    "Adding_across",
+    "Correct",
+    "Denominator-only_change",
+    "Incorrect_equivalent_fraction_addition",
+    "Neither"
+   ],
+   "seen": [
+    "Adding_across",
+    "Correct",
+    "Denominator-only_change",
+    "Incorrect_equivalent_fraction_addition",
+    "Neither"
+   ]
+  },
+  "33474": {
+   "text": "Sally has \\( \\frac{2}{3} \\) of a whole cake in the fridge. Robert eats \\( \\frac{1}{3} \\) of this piece. What fraction of the whole cake has Robert eaten?\nChoose the number sentence that would solve the word problem.",
+   "choices": [
+    "\\( \\frac{1}{3} \\times \\frac{2}{3} \\)",
+    "\\( \\frac{1}{3}+\\frac{2}{3} \\)",
+    "\\( \\frac{2}{3} \\div \\frac{1}{3} \\)",
+    "\\( \\frac{2}{3}-\\frac{1}{3} \\)"
+   ],
+   "right": "\\( \\frac{1}{3} \\times \\frac{2}{3} \\)",
+   "classes": [
+    "Correct",
+    "Division",
+    "Neither",
+    "Subtraction"
+   ],
+   "seen": [
+    "Correct",
+    "Division",
+    "Neither",
+    "Subtraction"
+   ]
+  },
+  "76870": {
+   "text": "This is part of a regular polygon. How many sides does it have? [Image: A diagram showing an obtuse angle labelled 144 degrees]",
+   "choices": [
+    "Not enough information",
+    "\\( 10 \\)",
+    "\\( 5 \\)",
+    "\\( 6 \\)"
+   ],
+   "right": "\\( 10 \\)",
+   "classes": [
+    "Correct",
+    "Definition",
+    "Interior",
+    "Neither",
+    "Unknowable"
+   ],
+   "seen": [
+    "Correct",
+    "Definition",
+    "Interior",
+    "Neither",
+    "Unknowable"
+   ]
+  },
+  "89443": {
+   "text": "What number belongs in the box?\n\\(\n(-8)-(-5)=\n\\square\\)",
+   "choices": [
+    "\\( -13 \\)",
+    "\\( -3 \\)",
+    "\\( 13 \\)",
+    "\\( 3 \\)"
+   ],
+   "right": "\\( -3 \\)",
+   "classes": [
+    "Correct",
+    "Neither",
+    "Positive",
+    "Tacking"
+   ],
+   "seen": [
+    "Correct",
+    "Neither",
+    "Positive",
+    "Tacking"
+   ]
+  },
+  "91695": {
+   "text": "Dots have been arranged in these patterns: [Image: Pattern 1 consists of 6 dots, Pattern 2 consists of 10 dots, Pattern 3 consists of 14 dots and Pattern 4 consists of 18 dots] How many dots would there be in Pattern \\( 6 \\) ?",
+   "choices": [
+    "\\( 20 \\)",
+    "\\( 22 \\)",
+    "\\( 26 \\)",
+    "\\( 36 \\)"
+   ],
+   "right": "\\( 26 \\)",
+   "classes": [
+    "Correct",
+    "Firstterm",
+    "Neither",
+    "Wrong_term"
+   ],
+   "seen": [
+    "Correct",
+    "Firstterm",
+    "Neither",
+    "Wrong_term"
+   ]
+  },
+  "104665": {
+   "text": "It takes \\( 3 \\) people a total of \\( 192 \\) hours to build a wall.\n\nHow long would it take if \\( 12 \\) people built the same wall?",
+   "choices": [
+    "\\( 192 \\) hours",
+    "\\( 48 \\) hours",
+    "\\( 64 \\) hours",
+    "\\( 768 \\) hours"
+   ],
+   "right": "\\( 48 \\) hours",
+   "classes": [
+    "Base_rate",
+    "Correct",
+    "Multiplying_by_4",
+    "Neither"
+   ],
+   "seen": [
+    "Base_rate",
+    "Correct",
+    "Multiplying_by_4",
+    "Neither"
+   ]
+  },
+  "109465": {
+   "text": "The probability of an event occurring is \\( 0.9 \\).\n\nWhich of the following most accurately describes the likelihood of the event occurring?",
+   "choices": [
+    "Certain",
+    "Impossible",
+    "Likely",
+    "Unlikely"
+   ],
+   "right": "Likely",
+   "classes": [
+    "Certainty",
+    "Correct",
+    "Neither",
+    "Scale"
+   ],
+   "seen": [
+    "Certainty",
+    "Correct",
+    "Neither",
+    "Scale"
+   ]
+  }
+ }
+}

runs/v1-newq/log.jsonl

@@ -0,0 +1,3 @@
+{"epoch": 0, "train_loss": 0.9934, "map3": 0.6452, "acc": 0.5711, "map3_masked": 0.6452, "t": 790}
+{"epoch": 1, "train_loss": 0.4135, "map3": 0.625, "acc": 0.5306, "map3_masked": 0.625, "t": 1584}
+{"epoch": 2, "train_loss": 0.2981, "map3": 0.6064, "acc": 0.4919, "map3_masked": 0.6064, "t": 2376}

runs/v1-newq/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:14e579bb53fc04e7b3276cd7c55102496ad64112259e0344fb741717cdf6d961
+size 1740447924

runs/v1-newq/tokenizer/added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "[MASK]": 128000
+}

runs/v1-newq/tokenizer/special_tokens_map.json

@@ -0,0 +1,15 @@
+{
+  "bos_token": "[CLS]",
+  "cls_token": "[CLS]",
+  "eos_token": "[SEP]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": {
+    "content": "[UNK]",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

runs/v1-newq/tokenizer/spm.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
+size 2464616

runs/v1-newq/tokenizer/tokenizer_config.json

@@ -0,0 +1,58 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128000": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "[CLS]",
+  "clean_up_tokenization_spaces": false,
+  "cls_token": "[CLS]",
+  "do_lower_case": false,
+  "eos_token": "[SEP]",
+  "mask_token": "[MASK]",
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "sp_model_kwargs": {},
+  "split_by_punct": false,
+  "tokenizer_class": "DebertaV2Tokenizer",
+  "unk_token": "[UNK]",
+  "vocab_type": "spm"
+}