init

transformers/tests/models/gpt_neox/test_modeling_gpt_neox.py

@@ -0,0 +1,482 @@
+# Copyright 2022 The HuggingFace Inc. team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""Testing suite for the PyTorch GPTNeoX model."""
+
+import unittest
+
+from parameterized import parameterized
+
+from transformers import AutoTokenizer, DynamicCache, GPTNeoXConfig, is_torch_available, set_seed
+from transformers.testing_utils import require_torch, slow, torch_device
+
+from ...generation.test_utils import GenerationTesterMixin
+from ...test_configuration_common import ConfigTester
+from ...test_modeling_common import ModelTesterMixin, ids_tensor, random_attention_mask
+from ...test_pipeline_mixin import PipelineTesterMixin
+
+
+if is_torch_available():
+    import torch
+
+    from transformers import (
+        GPTNeoXForCausalLM,
+        GPTNeoXForQuestionAnswering,
+        GPTNeoXForSequenceClassification,
+        GPTNeoXForTokenClassification,
+        GPTNeoXModel,
+    )
+    from transformers.models.gpt_neox.modeling_gpt_neox import GPTNeoXRotaryEmbedding
+
+
+class GPTNeoXModelTester:
+    def __init__(
+        self,
+        parent,
+        batch_size=13,
+        seq_length=7,
+        is_training=True,
+        use_input_mask=True,
+        use_token_type_ids=True,
+        use_labels=True,
+        vocab_size=99,
+        hidden_size=64,
+        num_hidden_layers=2,
+        num_attention_heads=4,
+        intermediate_size=37,
+        hidden_act="gelu",
+        hidden_dropout_prob=0.1,
+        attention_probs_dropout_prob=0.1,
+        max_position_embeddings=512,
+        type_vocab_size=16,
+        type_sequence_label_size=2,
+        initializer_range=0.02,
+        num_labels=3,
+        num_choices=4,
+        scope=None,
+    ):
+        self.parent = parent
+        self.batch_size = batch_size
+        self.seq_length = seq_length
+        self.is_training = is_training
+        self.use_input_mask = use_input_mask
+        self.use_token_type_ids = use_token_type_ids
+        self.use_labels = use_labels
+        self.vocab_size = vocab_size
+        self.hidden_size = hidden_size
+        self.num_hidden_layers = num_hidden_layers
+        self.num_attention_heads = num_attention_heads
+        self.intermediate_size = intermediate_size
+        self.hidden_act = hidden_act
+        self.hidden_dropout_prob = hidden_dropout_prob
+        self.attention_probs_dropout_prob = attention_probs_dropout_prob
+        self.max_position_embeddings = max_position_embeddings
+        self.type_vocab_size = type_vocab_size
+        self.type_sequence_label_size = type_sequence_label_size
+        self.initializer_range = initializer_range
+        self.num_labels = num_labels
+        self.num_choices = num_choices
+        self.scope = scope
+        self.pad_token_id = vocab_size - 1
+
+    def prepare_config_and_inputs(self):
+        input_ids = ids_tensor([self.batch_size, self.seq_length], self.vocab_size)
+
+        input_mask = None
+        if self.use_input_mask:
+            input_mask = random_attention_mask([self.batch_size, self.seq_length])
+
+        token_labels = None
+        if self.use_labels:
+            token_labels = ids_tensor([self.batch_size, self.seq_length], self.num_labels)
+
+        config = self.get_config()
+
+        return config, input_ids, input_mask, token_labels
+
+    def get_config(self):
+        return GPTNeoXConfig(
+            vocab_size=self.vocab_size,
+            hidden_size=self.hidden_size,
+            num_hidden_layers=self.num_hidden_layers,
+            num_attention_heads=self.num_attention_heads,
+            intermediate_size=self.intermediate_size,
+            hidden_act=self.hidden_act,
+            hidden_dropout_prob=self.hidden_dropout_prob,
+            attention_probs_dropout_prob=self.attention_probs_dropout_prob,
+            max_position_embeddings=self.max_position_embeddings,
+            type_vocab_size=self.type_vocab_size,
+            is_decoder=False,
+            initializer_range=self.initializer_range,
+            pad_token_id=self.pad_token_id,
+        )
+
+    def prepare_config_and_inputs_for_decoder(self):
+        config, input_ids, input_mask, token_labels = self.prepare_config_and_inputs()
+
+        config.is_decoder = True
+
+        return config, input_ids, input_mask, token_labels
+
+    def create_and_check_model(self, config, input_ids, input_mask):
+        model = GPTNeoXModel(config=config)
+        model.to(torch_device)
+        model.eval()
+        _ = model(input_ids, attention_mask=input_mask)
+        result = model(input_ids)
+        self.parent.assertEqual(result.last_hidden_state.shape, (self.batch_size, self.seq_length, self.hidden_size))
+
+    def create_and_check_model_as_decoder(self, config, input_ids, input_mask):
+        config.add_cross_attention = True
+        model = GPTNeoXModel(config)
+        model.to(torch_device)
+        model.eval()
+        result = model(input_ids, attention_mask=input_mask)
+        self.parent.assertEqual(result.last_hidden_state.shape, (self.batch_size, self.seq_length, self.hidden_size))
+
+    def create_and_check_for_causal_lm(self, config, input_ids, input_mask, token_labels):
+        model = GPTNeoXForCausalLM(config=config)
+        model.to(torch_device)
+        model.eval()
+        result = model(input_ids, attention_mask=input_mask, labels=token_labels)
+        self.parent.assertEqual(result.logits.shape, (self.batch_size, self.seq_length, self.vocab_size))
+
+    def create_and_check_for_question_answering(self, config, input_ids, input_mask, token_labels):
+        config.num_labels = self.num_labels
+        model = GPTNeoXForQuestionAnswering(config)
+        model.to(torch_device)
+        model.eval()
+        result = model(input_ids, attention_mask=input_mask)
+        self.parent.assertEqual(result.start_logits.shape, (self.batch_size, self.seq_length))
+        self.parent.assertEqual(result.end_logits.shape, (self.batch_size, self.seq_length))
+
+    def create_and_check_for_sequence_classification(self, config, input_ids, input_mask, token_labels):
+        config.num_labels = self.num_labels
+        model = GPTNeoXForSequenceClassification(config)
+        model.to(torch_device)
+        model.eval()
+        sequence_labels = ids_tensor([self.batch_size], self.type_sequence_label_size)
+        result = model(input_ids, attention_mask=input_mask, labels=sequence_labels)
+        self.parent.assertEqual(result.logits.shape, (self.batch_size, self.num_labels))
+
+    def create_and_check_for_token_classification(self, config, input_ids, input_mask, token_labels):
+        config.num_labels = self.num_labels
+        model = GPTNeoXForTokenClassification(config)
+        model.to(torch_device)
+        model.eval()
+        result = model(input_ids, attention_mask=input_mask, labels=token_labels)
+        self.parent.assertEqual(result.logits.shape, (self.batch_size, self.seq_length, self.num_labels))
+
+    def create_and_check_decoder_model_past_large_inputs(self, config, input_ids, input_mask):
+        config.is_decoder = True
+        model = GPTNeoXForCausalLM(config=config)
+        model.to(torch_device)
+        model.eval()
+
+        # first forward pass
+        outputs = model(input_ids, attention_mask=input_mask, use_cache=True)
+        past_key_values = outputs.past_key_values
+
+        # create hypothetical multiple next token and extent to next_input_ids
+        next_tokens = ids_tensor((self.batch_size, 3), config.vocab_size)
+        next_mask = ids_tensor((self.batch_size, 3), vocab_size=2)
+
+        # append to next input_ids and
+        next_input_ids = torch.cat([input_ids, next_tokens], dim=-1)
+        next_attention_mask = torch.cat([input_mask, next_mask], dim=-1)
+
+        output_from_no_past = model(next_input_ids, attention_mask=next_attention_mask, output_hidden_states=True)
+        output_from_no_past = output_from_no_past["hidden_states"][0]
+        output_from_past = model(
+            next_tokens,
+            attention_mask=next_attention_mask,
+            past_key_values=past_key_values,
+            output_hidden_states=True,
+        )["hidden_states"][0]
+
+        # select random slice
+        random_slice_idx = ids_tensor((1,), output_from_past.shape[-1]).item()
+        output_from_no_past_slice = output_from_no_past[:, -3:, random_slice_idx].detach()
+        output_from_past_slice = output_from_past[:, :, random_slice_idx].detach()
+
+        self.parent.assertTrue(output_from_past_slice.shape[1] == next_tokens.shape[1])
+
+        # test that outputs are equal for slice
+        self.parent.assertTrue(torch.allclose(output_from_past_slice, output_from_no_past_slice, atol=1e-3))
+
+    def create_and_check_cached_forward_with_and_without_attention_mask(self, config, input_ids, *args):
+        # Relevant issue: https://github.com/huggingface/transformers/issues/31943
+        model = GPTNeoXModel(config)
+        model.to(torch_device)
+        model.eval()
+
+        # We want this for SDPA, eager works with a `None` attention mask
+        assert model.config._attn_implementation == "sdpa", (
+            "This test assumes the model to have the SDPA implementation for its attention calculations."
+        )
+
+        # Prepare cache and non_cache input, needs a full attention mask
+        cached_len = input_ids.shape[-1] // 2
+        input_mask = torch.ones(size=input_ids.size()).to(torch_device)
+        cache_inputs = {"input_ids": input_ids[:, :cached_len], "attention_mask": input_mask[:, :cached_len]}
+        non_cache_inputs = {"input_ids": input_ids[:, cached_len:], "attention_mask": input_mask}
+
+        def copy_cache(cache: DynamicCache):
+            """Deep copy a DynamicCache to reuse the same one multiple times."""
+            new_cache = cache
+            for i in range(len(cache)):
+                new_cache.layers[i].keys = cache.layers[i].keys.clone()
+                new_cache.layers[i].values = cache.layers[i].values.clone()
+
+        # Cached forward once with the attention mask provided and the other time without it (which should assume full attention)
+        # We need to run both on a copy of the cache, otherwise it is modified in-place
+        cache_outputs = model(**cache_inputs)
+        cache = cache_outputs.past_key_values
+        full_outputs_with_attention_mask = model(
+            **non_cache_inputs, past_key_values=copy_cache(cache)
+        ).last_hidden_state
+        full_outputs_without_attention_mask = model(
+            non_cache_inputs["input_ids"], past_key_values=copy_cache(cache)
+        ).last_hidden_state
+
+        self.parent.assertTrue(
+            torch.allclose(full_outputs_with_attention_mask, full_outputs_without_attention_mask, atol=1e-5)
+        )
+
+    def prepare_config_and_inputs_for_common(self):
+        config_and_inputs = self.prepare_config_and_inputs()
+        config, input_ids, input_mask, token_labels = config_and_inputs
+        inputs_dict = {"input_ids": input_ids, "attention_mask": input_mask}
+        return config, inputs_dict
+
+
+@require_torch
+class GPTNeoXModelTest(ModelTesterMixin, GenerationTesterMixin, PipelineTesterMixin, unittest.TestCase):
+    all_model_classes = (
+        (
+            GPTNeoXModel,
+            GPTNeoXForCausalLM,
+            GPTNeoXForQuestionAnswering,
+            GPTNeoXForSequenceClassification,
+            GPTNeoXForTokenClassification,
+        )
+        if is_torch_available()
+        else ()
+    )
+    pipeline_model_mapping = (
+        {
+            "feature-extraction": GPTNeoXModel,
+            "question-answering": GPTNeoXForQuestionAnswering,
+            "text-classification": GPTNeoXForSequenceClassification,
+            "text-generation": GPTNeoXForCausalLM,
+            "token-classification": GPTNeoXForTokenClassification,
+            "zero-shot": GPTNeoXForSequenceClassification,
+        }
+        if is_torch_available()
+        else {}
+    )
+    test_pruning = False
+    test_missing_keys = False
+    test_model_parallel = False
+    test_head_masking = False
+
+    def setUp(self):
+        self.model_tester = GPTNeoXModelTester(self)
+        self.config_tester = ConfigTester(self, config_class=GPTNeoXConfig, hidden_size=64, num_attention_heads=8)
+
+    def test_config(self):
+        self.config_tester.run_common_tests()
+
+    def test_model(self):
+        config, input_ids, input_mask, token_labels = self.model_tester.prepare_config_and_inputs()
+        self.model_tester.create_and_check_model(config, input_ids, input_mask)
+
+    def test_model_as_decoder(self):
+        config, input_ids, input_mask, token_labels = self.model_tester.prepare_config_and_inputs_for_decoder()
+        self.model_tester.create_and_check_model_as_decoder(config, input_ids, input_mask)
+
+    def test_model_as_decoder_with_default_input_mask(self):
+        config, input_ids, input_mask, token_labels = self.model_tester.prepare_config_and_inputs_for_decoder()
+
+        input_mask = None
+
+        self.model_tester.create_and_check_model_as_decoder(config, input_ids, input_mask)
+
+    def test_decoder_model_past_large_inputs(self):
+        config, input_ids, input_mask, token_labels = self.model_tester.prepare_config_and_inputs()
+        self.model_tester.create_and_check_decoder_model_past_large_inputs(config, input_ids, input_mask)
+
+    def test_model_for_causal_lm(self):
+        config_and_inputs = self.model_tester.prepare_config_and_inputs()
+        self.model_tester.create_and_check_for_causal_lm(*config_and_inputs)
+
+    def test_model_for_question_answering(self):
+        config_and_inputs = self.model_tester.prepare_config_and_inputs()
+        self.model_tester.create_and_check_for_question_answering(*config_and_inputs)
+
+    def test_model_for_sequence_classification(self):
+        config_and_inputs = self.model_tester.prepare_config_and_inputs()
+        self.model_tester.create_and_check_for_sequence_classification(*config_and_inputs)
+
+    def test_model_for_token_classification(self):
+        config_and_inputs = self.model_tester.prepare_config_and_inputs()
+        self.model_tester.create_and_check_for_token_classification(*config_and_inputs)
+
+    def test_cached_forward_with_and_without_attention_mask(self):
+        config_and_inputs = self.model_tester.prepare_config_and_inputs()
+        self.model_tester.create_and_check_cached_forward_with_and_without_attention_mask(*config_and_inputs)
+
+    @unittest.skip(reason="Feed forward chunking is not implemented")
+    def test_feed_forward_chunking(self):
+        pass
+
+    @parameterized.expand([("linear",), ("dynamic",)])
+    def test_model_rope_scaling_from_config(self, scaling_type):
+        config, _ = self.model_tester.prepare_config_and_inputs_for_common()
+        short_input = ids_tensor([1, 10], config.vocab_size)
+        long_input = ids_tensor([1, int(config.max_position_embeddings * 1.5)], config.vocab_size)
+
+        set_seed(42)  # Fixed seed at init time so the two models get the same random weights
+        original_model = GPTNeoXModel(config)
+        original_model.to(torch_device)
+        original_model.eval()
+        original_short_output = original_model(short_input).last_hidden_state
+        original_long_output = original_model(long_input).last_hidden_state
+
+        set_seed(42)  # Fixed seed at init time so the two models get the same random weights
+        config.rope_scaling = {"type": scaling_type, "factor": 10.0}
+        scaled_model = GPTNeoXModel(config)
+        scaled_model.to(torch_device)
+        scaled_model.eval()
+        scaled_short_output = scaled_model(short_input).last_hidden_state
+        scaled_long_output = scaled_model(long_input).last_hidden_state
+
+        # Dynamic scaling does not change the RoPE embeddings until it receives an input longer than the original
+        # maximum sequence length, so the outputs for the short input should match.
+        if scaling_type == "dynamic":
+            torch.testing.assert_close(original_short_output, scaled_short_output, rtol=1e-5, atol=1e-5)
+        else:
+            self.assertFalse(torch.allclose(original_short_output, scaled_short_output, atol=1e-5))
+
+        # The output should be different for long inputs
+        self.assertFalse(torch.allclose(original_long_output, scaled_long_output, atol=1e-5))
+
+    def test_model_rope_scaling(self):
+        config, _ = self.model_tester.prepare_config_and_inputs_for_common()
+        scaling_factor = 10
+        short_input_length = 10
+        long_input_length = int(config.max_position_embeddings * 1.5)
+
+        # Inputs
+        x = torch.randn(
+            1, dtype=torch.float32, device=torch_device
+        )  # used exclusively to get the dtype and the device
+        position_ids_short = torch.arange(short_input_length, dtype=torch.long, device=torch_device)
+        position_ids_short = position_ids_short.unsqueeze(0)
+        position_ids_long = torch.arange(long_input_length, dtype=torch.long, device=torch_device)
+        position_ids_long = position_ids_long.unsqueeze(0)
+
+        # Sanity check original RoPE
+        original_rope = GPTNeoXRotaryEmbedding(config).to(torch_device)
+        original_cos_short, original_sin_short = original_rope(x, position_ids_short)
+        original_cos_long, original_sin_long = original_rope(x, position_ids_long)
+        torch.testing.assert_close(original_cos_short, original_cos_long[:, :short_input_length, :])
+        torch.testing.assert_close(original_sin_short, original_sin_long[:, :short_input_length, :])
+
+        # Sanity check linear RoPE scaling
+        # New position "x" should match original position with index "x/scaling_factor"
+        config.rope_scaling = {"type": "linear", "factor": scaling_factor}
+        linear_scaling_rope = GPTNeoXRotaryEmbedding(config).to(torch_device)
+        linear_cos_short, linear_sin_short = linear_scaling_rope(x, position_ids_short)
+        linear_cos_long, linear_sin_long = linear_scaling_rope(x, position_ids_long)
+        torch.testing.assert_close(linear_cos_short, linear_cos_long[:, :short_input_length, :])
+        torch.testing.assert_close(linear_sin_short, linear_sin_long[:, :short_input_length, :])
+        for new_position in range(0, long_input_length, scaling_factor):
+            original_position = int(new_position // scaling_factor)
+            torch.testing.assert_close(linear_cos_long[:, new_position, :], original_cos_long[:, original_position, :])
+            torch.testing.assert_close(linear_sin_long[:, new_position, :], original_sin_long[:, original_position, :])
+
+        # Sanity check Dynamic NTK RoPE scaling
+        # Scaling should only be observed after a long input is fed. We can observe that the frequencies increase
+        # with scaling_factor (or that `inv_freq` decreases)
+        config.rope_scaling = {"type": "dynamic", "factor": scaling_factor}
+        ntk_scaling_rope = GPTNeoXRotaryEmbedding(config).to(torch_device)
+        ntk_cos_short, ntk_sin_short = ntk_scaling_rope(x, position_ids_short)
+        ntk_cos_long, ntk_sin_long = ntk_scaling_rope(x, position_ids_long)
+        torch.testing.assert_close(ntk_cos_short, original_cos_short)
+        torch.testing.assert_close(ntk_sin_short, original_sin_short)
+        with self.assertRaises(AssertionError):
+            torch.testing.assert_close(ntk_cos_long, original_cos_long)
+        with self.assertRaises(AssertionError):
+            torch.testing.assert_close(ntk_sin_long, original_sin_long)
+        self.assertTrue((ntk_scaling_rope.inv_freq <= original_rope.inv_freq).all())
+
+
+@require_torch
+class GPTNeoXLanguageGenerationTest(unittest.TestCase):
+    @slow
+    def test_lm_generate_gptneox(self):
+        tokenizer = AutoTokenizer.from_pretrained("EleutherAI/pythia-410m-deduped")
+        for checkpointing in [True, False]:
+            model = GPTNeoXForCausalLM.from_pretrained("EleutherAI/pythia-410m-deduped")
+
+            if checkpointing:
+                model.gradient_checkpointing_enable()
+            else:
+                model.gradient_checkpointing_disable()
+            model.to(torch_device)
+
+            inputs = tokenizer("My favorite food is", return_tensors="pt").to(torch_device)
+            # The hub repo. is updated on 2023-04-04, resulting in poor outputs.
+            # See: https://github.com/huggingface/transformers/pull/24193
+            expected_output = "My favorite food is a good old-fashioned, old-fashioned, old-fashioned.\n\nI'm not sure"
+
+            output_ids = model.generate(**inputs, do_sample=False, max_new_tokens=20)
+            output_str = tokenizer.batch_decode(output_ids)[0]
+
+            self.assertEqual(output_str, expected_output)
+
+    @slow
+    def test_lm_generate_flex_attn_gptneox(self):
+        tokenizer = AutoTokenizer.from_pretrained("EleutherAI/pythia-410m-deduped")
+        for checkpointing in [True, False]:
+            model = GPTNeoXForCausalLM.from_pretrained(
+                "EleutherAI/pythia-410m-deduped", attn_implementation="flex_attention"
+            )
+            self.assertTrue(model.config._attn_implementation == "flex_attention")
+
+            if checkpointing:
+                model.gradient_checkpointing_enable()
+            else:
+                model.gradient_checkpointing_disable()
+            model.to(torch_device)
+
+            inputs = tokenizer("My favorite food is", return_tensors="pt").to(torch_device)
+            # The hub repo. is updated on 2023-04-04, resulting in poor outputs.
+            # See: https://github.com/huggingface/transformers/pull/24193
+            expected_output = "My favorite food is a good old-fashioned, old-fashioned, old-fashioned.\n\nI'm not sure"
+
+            output_ids = model.generate(**inputs, do_sample=False, max_new_tokens=20)
+            output_str = tokenizer.batch_decode(output_ids)[0]
+
+            self.assertEqual(output_str, expected_output)
+
+    def pythia_integration_test(self):
+        model_name_or_path = "EleutherAI/pythia-70m"
+        model = GPTNeoXForCausalLM.from_pretrained(model_name_or_path, dtype=torch.float16).to(torch_device)
+        EXPECTED_LOGITS = torch.tensor([1069.0000,  228.7500, 1072.0000, 1072.0000, 1069.0000, 1068.0000, 1068.0000, 1071.0000, 1071.0000, 1071.0000, 1073.0000, 1070.0000, 1071.0000, 1075.0000, 1073.0000, 1075.0000, 1074.0000, 1069.0000, 1072.0000, 1071.0000, 1071.0000, 1071.0000, 1070.0000, 1069.0000, 1069.0000, 1069.0000, 1070.0000, 1075.0000, 1073.0000, 1074.0000])  # fmt: skip
+        input_ids = [29, 93, 303, 64, 5478, 49651, 10394, 187, 34, 12939, 875]
+        # alternative: tokenizer('<|im_start|>system\nA chat between')
+        input_ids = torch.as_tensor(input_ids)[None].to(torch_device)
+        outputs = model(input_ids)["logits"][:, -1][0, :30]
+        torch.testing.assert_close(EXPECTED_LOGITS, outputs, rtol=1e-5, atol=1e-5)