rbbt-dm 1.2.1 → 1.2.4

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: ab775c0224960820e5c62e294e6a183be49201da15710b66544762e1aaf97ebf
-  data.tar.gz: 8fffb47ba226f06d1f41a8893d085bdc12c33c021cf2f0152f4cc741db36e420
+  metadata.gz: abaea1fff82b5e14a84dc9afc966fc8dde6482d50769d196854c1d619adebaf3
+  data.tar.gz: 561b8864fc2c0ba271a2a658da0d3492c7481a2368b40c3b91fe6edb4ebca4cd
 SHA512:
-  metadata.gz: 8be084156063cd93c7fe905bc4b6248dd376bbfcff8e650cbb03a4cc5c28f29dbcdaa2801895a3663067d7660e8bc2cf96682829519ebd6511a7a74cec021da0
-  data.tar.gz: 9c8570722319caf5afe60c90778d0b8517e70064030e0081968d919a314cfe1af90f63d13a7b8ce8dd56da6644db89916a729a4005f1e20745c8d4b45d50394c
+  metadata.gz: f26f6b27f1beb2554fa78369d1d618cc13175e0c9bb0e789b9490dcae0f7f6df4449a3c72d183ae22c96324d4e2f1ab0352bde8068c1c18871d52c5f5b53c235
+  data.tar.gz: bb33d93cbe24ea974beedb0530f9af317dec06c7e76f32c37d724322ba05f241c6b79a706a88f1bbe703ac4bc78c53c220f28c3f38cf7939477274b8747c436e

data/lib/rbbt/vector/model/huggingface.rb

@@ -1,11 +1,12 @@
 require 'rbbt/vector/model'
 require 'rbbt/util/python'
 
+RbbtPython.add_path Rbbt.python.find(:lib)
 RbbtPython.init_rbbt
 
 class HuggingfaceModel < VectorModel
 
-  attr_accessor :checkpoint, :task, :locate_tokens, :class_labels
+  attr_accessor :checkpoint, :task, :locate_tokens, :class_labels, :class_weights
 
   def tsv_dataset(tsv_dataset_file, elements, labels = nil)
 
@@ -48,7 +49,7 @@ class HuggingfaceModel < VectorModel
 
       if labels
         training_args = call_method(:training_args, output_dir)
-        call_method(:train_model, @model, @tokenizer, training_args, tsv_dataset(tsv_file, elements, labels))
+        call_method(:train_model, @model, @tokenizer, training_args, tsv_dataset(tsv_file, elements, labels), @class_weights)
       else
         if Array === elements
           training_args = call_method(:training_args, output_dir)
@@ -135,6 +136,3 @@ class HuggingfaceModel < VectorModel
   end
 end
 
-if __FILE__ == $0
-
-end

data/lib/rbbt/vector/model.rb

@@ -436,7 +436,7 @@ cat(paste(label, sep="\\n", collapse="\\n"));
     ensure
       @features = orig_features
       @labels = orig_labels
-    end
+    end unless folds == -1
     self.train unless folds == 1
     res
   end

data/python/rbbt_dm/__init__.py

@@ -0,0 +1 @@
+# Keep

data/python/rbbt_dm/huggingface.py

@@ -0,0 +1,135 @@
+#{{{ LOAD MODEL
+
+def import_module_class(module, class_name):
+    exec(f"from {module} import {class_name}")
+    return eval(class_name)
+
+def load_model(task, checkpoint):
+    class_name = 'AutoModelFor' + task
+    return import_module_class('transformers', class_name).from_pretrained(checkpoint)
+
+def load_tokenizer(task, checkpoint):
+    class_name = 'AutoTokenizer'
+    return import_module_class('transformers', class_name).from_pretrained(checkpoint)
+
+def load_model_and_tokenizer(task, checkpoint):
+    model = load_model(task, checkpoint)
+    tokenizer = load_tokenizer(task, checkpoint)
+    return model, tokenizer
+
+#{{{ SIMPLE EVALUATE
+
+def forward(model, features):
+    return model(**features)
+
+def logits(predictions):
+    logits = predictions["logits"]
+    return [v.detach().cpu().numpy() for v in logits]
+
+def eval_model(model, tokenizer, texts, return_logits = True):
+    features = tokenizer(texts, return_tensors='pt', truncation=True).to(model.device)
+    predictions = forward(model, features)
+    if (return_logits):
+        return logits(predictions)
+    else:
+        return predictions
+
+#{{{ TRAIN AND PREDICT
+
+def load_tsv(tsv_file):
+    from datasets import load_dataset
+    return load_dataset('csv', data_files=[tsv_file], sep="\t")
+
+def tsv_dataset(tokenizer, tsv_file):
+    dataset = load_tsv(tsv_file)
+    tokenized_dataset = dataset.map(lambda example: tokenizer(example["text"], truncation=True) , batched=True)
+    return tokenized_dataset
+
+def training_args(*args, **kwargs):
+    from transformers import TrainingArguments
+    training_args = TrainingArguments(*args, **kwargs)
+    return training_args
+
+
+def train_model(model, tokenizer, training_args, tsv_file, class_weights=None):
+    from transformers import Trainer
+
+    tokenized_dataset = tsv_dataset(tokenizer, tsv_file)
+
+    if (not class_weights == None):
+      import torch
+      from torch import nn
+
+      class WeightTrainer(Trainer):
+          def compute_loss(self, model, inputs, return_outputs=False):
+              labels = inputs.get("labels")
+              # forward pass
+              outputs = model(**inputs)
+              logits = outputs.get('logits')
+              # compute custom loss
+              loss_fct = nn.CrossEntropyLoss(weight=torch.tensor(class_weights).to(model.device))
+              loss = loss_fct(logits.view(-1, self.model.config.num_labels), labels.view(-1))
+              return (loss, outputs) if return_outputs else loss
+
+      trainer = WeightTrainer(
+              model,
+              training_args,
+              train_dataset = tokenized_dataset["train"],
+              tokenizer = tokenizer
+              )
+    else:
+
+      trainer = Trainer(
+              model,
+              training_args,
+              train_dataset = tokenized_dataset["train"],
+              tokenizer = tokenizer
+              )
+
+    trainer.train()
+
+def find_tokens_in_input(dataset, token_ids):
+    position_rows = []
+
+    for row in dataset:
+        input_ids = row["input_ids"]
+
+        if (not hasattr(token_ids, "__len__")):
+            token_ids = [token_ids]
+
+        positions = []
+        for token_id in token_ids:
+
+            item_positions = []
+            for i in range(len(input_ids)):
+                if input_ids[i] == token_id:
+                    item_positions.append(i)
+
+            positions.append(item_positions)
+
+
+        position_rows.append(positions)
+
+    return position_rows
+
+
+def predict_model(model, tokenizer, training_args, tsv_file, locate_tokens = None):
+    from transformers import Trainer
+
+    tokenized_dataset = tsv_dataset(tokenizer, tsv_file)
+
+    trainer = Trainer(
+            model,
+            training_args,
+            tokenizer = tokenizer
+            )
+
+    result = trainer.predict(test_dataset = tokenized_dataset["train"])
+    if (locate_tokens != None):
+        token_ids = tokenizer.convert_tokens_to_ids(locate_tokens)
+        token_positions = find_tokens_in_input(tokenized_dataset["train"], token_ids)
+        return dict(result=result, token_positions=token_positions)
+    else:
+        return result
+
+

data/test/rbbt/vector/model/test_huggingface.rb

@@ -3,6 +3,22 @@ require 'rbbt/vector/model/huggingface'
 
 class TestHuggingface < Test::Unit::TestCase
 
+  def test_pipeline
+    require 'rbbt/util/python'
+    model = VectorModel.new
+    model.post_process do |elements|
+      elements.collect{|e| e['label'] }
+    end
+    model.eval_model do |file, elements|
+      RbbtPython.run :transformers do 
+        classifier ||= transformers.pipeline("sentiment-analysis")
+        classifier.call(elements)
+      end
+    end
+
+    assert_equal ["POSITIVE"], model.eval("I've been waiting for a HuggingFace course my whole life.")
+  end
+
   def test_sst_eval
     TmpFile.with_file do |dir|
       checkpoint = "distilbert-base-uncased-finetuned-sst-2-english"
@@ -12,7 +28,6 @@ class TestHuggingface < Test::Unit::TestCase
       model.class_labels = ["Bad", "Good"]
 
       assert_equal ["Bad", "Good"], model.eval(["This is dog", "This is cat"])
-
     end
   end
 

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: rbbt-dm
 version: !ruby/object:Gem::Version
-  version: 1.2.1
+  version: 1.2.4
 platform: ruby
 authors:
 - Miguel Vazquez
-autorequire: 
+autorequire:
 bindir: bin
 cert_chain: []
-date: 2023-02-04 00:00:00.000000000 Z
+date: 2023-02-07 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rbbt-util
@@ -113,6 +113,8 @@ files:
 - lib/rbbt/vector/model/svm.rb
 - lib/rbbt/vector/model/tensorflow.rb
 - lib/rbbt/vector/model/util.rb
+- python/rbbt_dm/__init__.py
+- python/rbbt_dm/huggingface.py
 - share/R/MA.R
 - share/R/barcode.R
 - share/R/heatmap.3.R
@@ -141,7 +143,7 @@ files:
 homepage: http://github.com/mikisvaz/rbbt-phgx
 licenses: []
 metadata: {}
-post_install_message: 
+post_install_message:
 rdoc_options: []
 require_paths:
 - lib
@@ -156,22 +158,22 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.1.2
-signing_key: 
+rubygems_version: 3.1.6
+signing_key:
 specification_version: 4
 summary: Data-mining and statistics
 test_files:
-- test/test_helper.rb
+- test/rbbt/statistics/test_hypergeometric.rb
+- test/rbbt/statistics/test_fisher.rb
+- test/rbbt/statistics/test_fdr.rb
+- test/rbbt/statistics/test_random_walk.rb
+- test/rbbt/test_ml_task.rb
 - test/rbbt/vector/test_model.rb
-- test/rbbt/vector/model/test_huggingface.rb
 - test/rbbt/vector/model/test_tensorflow.rb
 - test/rbbt/vector/model/test_spaCy.rb
+- test/rbbt/vector/model/test_huggingface.rb
 - test/rbbt/vector/model/test_svm.rb
-- test/rbbt/statistics/test_random_walk.rb
-- test/rbbt/statistics/test_fisher.rb
-- test/rbbt/statistics/test_fdr.rb
-- test/rbbt/statistics/test_hypergeometric.rb
-- test/rbbt/test_stan.rb
-- test/rbbt/matrix/test_barcode.rb
-- test/rbbt/test_ml_task.rb
 - test/rbbt/network/test_paths.rb
+- test/rbbt/matrix/test_barcode.rb
+- test/rbbt/test_stan.rb
+- test/test_helper.rb