Flax mistral (#26943)

* direct copy from llama work

* mistral modules forward pass working

* flax mistral forward pass with sliding window

* added tests

* added layer collection approach

* Revert "added layer collection approach"

This reverts commit 0e2905bf2236ec323163fc1a9f0c016b21aa8b8f.

* Revert "Revert "added layer collection approach""

This reverts commit fb17b6187ac5d16da7c461e1130514dc3d137a43.

* fixed attention outputs

* added mistral to init and auto

* fixed import name

* fixed layernorm weight dtype

* freeze initialized weights

* make sure conversion consideres bfloat16

* added backend

* added docstrings

* added cache

* fixed sliding window causal mask

* passes cache tests

* passed all tests

* applied make style

* removed commented out code

* applied fix-copies ignored other model changes

* applied make fix-copies

* removed unused functions

* passed generation integration test

* slow tests pass

* fixed slow tests

* changed default dtype from jax.numpy.float32 to float32 for docstring check

* skip cache test  for FlaxMistralForSequenceClassification since if pad_token_id in input_ids it doesn't score previous input_ids

* updated checkpoint since from_pt not included

* applied black style

* removed unused args

* Applied styling and fixup

* changed checkpoint for doc back

* fixed rf after adding it to hf hub

* Add dummy ckpt

* applied styling

* added tokenizer to new ckpt

* fixed slice format

* fix init and slice

* changed ref for placeholder TODO

* added copies from Llama

* applied styling

* applied fix-copies

* fixed docs

* update weight dtype reconversion for sharded weights

* removed Nullable input ids

* Removed unnecessary output attentions in Module

* added embedding weight initialziation

* removed unused past_key_values

* fixed deterministic

* Fixed RMS Norm and added copied from

* removed input_embeds

* applied make style

* removed nullable input ids from sequence classification model

* added copied from GPTJ

* added copied from Llama on FlaxMistralDecoderLayer

* added copied from to FlaxMistralPreTrainedModel methods

* fix test deprecation warning

* freeze gpt neox random_params and fix copies

* applied make style

* fixed doc issue

* skipped docstring test to allign # copied from

* applied make style

* removed FlaxMistralForSequenceClassification

* removed unused padding_idx

* removed more sequence classification

* removed sequence classification

* applied styling and consistency

* added copied from in tests

* removed sequence classification test logic

* applied styling

* applied make style

* removed freeze and fixed copies

* undo test change

* changed repeat_kv to tile

* fixed to key value groups

* updated copyright year

* split casual_mask

* empty to rerun failed pt_flax_equivalence test FlaxWav2Vec2ModelTest

* went back to 2023 for tests_pr_documentation_tests

* went back to 2024

* changed tile to repeat

* applied make style

* empty for retry on Wav2Vec2

docs/source/en/index.md

@@ -190,7 +190,7 @@ Flax), PyTorch, and/or TensorFlow.
 |                 [Megatron-BERT](model_doc/megatron-bert)                 |       ✅        |         ❌         |      ❌      |
 |                 [Megatron-GPT2](model_doc/megatron_gpt2)                 |       ✅        |         ✅         |      ✅      |
 |                       [MGP-STR](model_doc/mgp-str)                       |       ✅        |         ❌         |      ❌      |
-|                       [Mistral](model_doc/mistral)                       |       ✅        |         ❌         |      ❌      |
+|                       [Mistral](model_doc/mistral)                       |       ✅        |         ❌         |      ✅      |
 |                       [Mixtral](model_doc/mixtral)                       |       ✅        |         ❌         |      ❌      |
 |                         [mLUKE](model_doc/mluke)                         |       ✅        |         ❌         |      ❌      |
 |                           [MMS](model_doc/mms)                           |       ✅        |         ✅         |      ✅      |

docs/source/en/model_doc/mistral.md

@@ -149,3 +149,13 @@ Albert Jiang, Alexandre Sablayrolles, Arthur Mensch, Chris Bamford, Devendra Sin
 
 [[autodoc]] MistralForSequenceClassification
     - forward
+
+## FlaxMistralModel
+
+[[autodoc]] FlaxMistralModel
+    - __call__
+
+## FlaxMistralForCausalLM
+
+[[autodoc]] FlaxMistralForCausalLM
+    - __call__

src/transformers/__init__.py

@@ -4678,6 +4678,13 @@ else:
             "FlaxMBartPreTrainedModel",
         ]
     )
+    _import_structure["models.mistral"].extend(
+        [
+            "FlaxMistralForCausalLM",
+            "FlaxMistralModel",
+            "FlaxMistralPreTrainedModel",
+        ]
+    )
     _import_structure["models.mt5"].extend(["FlaxMT5EncoderModel", "FlaxMT5ForConditionalGeneration", "FlaxMT5Model"])
     _import_structure["models.opt"].extend(
         [
@@ -8830,6 +8837,11 @@ if TYPE_CHECKING:
             FlaxMBartModel,
             FlaxMBartPreTrainedModel,
         )
+        from .models.mistral import (
+            FlaxMistralForCausalLM,
+            FlaxMistralModel,
+            FlaxMistralPreTrainedModel,
+        )
         from .models.mt5 import (
             FlaxMT5EncoderModel,
             FlaxMT5ForConditionalGeneration,

src/transformers/modeling_flax_pytorch_utils.py

@@ -255,7 +255,10 @@ def convert_pytorch_sharded_state_dict_to_flax(shard_filenames, flax_model):
         # load using msgpack utils
         weights_only_kwarg = {"weights_only": True} if is_torch_greater_or_equal_than_1_13 else {}
         pt_state_dict = torch.load(shard_file, **weights_only_kwarg)
-        pt_state_dict = {k: v.numpy() for k, v in pt_state_dict.items()}
+        weight_dtypes = {k: v.dtype for k, v in pt_state_dict.items()}
+        pt_state_dict = {
+            k: v.numpy() if v.dtype != torch.bfloat16 else v.float().numpy() for k, v in pt_state_dict.items()
+        }
 
         model_prefix = flax_model.base_model_prefix
 
@@ -278,6 +281,7 @@ def convert_pytorch_sharded_state_dict_to_flax(shard_filenames, flax_model):
         # Need to change some parameters name to match Flax names
         for pt_key, pt_tensor in pt_state_dict.items():
             pt_tuple_key = tuple(pt_key.split("."))
+            is_bfloat_16 = weight_dtypes[pt_key] == torch.bfloat16
 
             # remove base model prefix if necessary
             has_base_model_prefix = pt_tuple_key[0] == model_prefix
@@ -314,11 +318,15 @@ def convert_pytorch_sharded_state_dict_to_flax(shard_filenames, flax_model):
                     continue
 
                 # also add unexpected weight so that warning is thrown
-                flax_state_dict[("params",) + flax_key] = jnp.asarray(flax_tensor)
+                flax_state_dict[("params",) + flax_key] = (
+                    jnp.asarray(flax_tensor) if not is_bfloat_16 else jnp.asarray(flax_tensor, dtype=jnp.bfloat16)
+                )
 
             else:
                 # also add unexpected weight so that warning is thrown
-                flax_state_dict[flax_key] = jnp.asarray(flax_tensor)
+                flax_state_dict[flax_key] = (
+                    jnp.asarray(flax_tensor) if not is_bfloat_16 else jnp.asarray(flax_tensor, dtype=jnp.bfloat16)
+                )
     return unflatten_dict(flax_state_dict)
 
 

src/transformers/models/auto/modeling_flax_auto.py

@@ -47,6 +47,7 @@ FLAX_MODEL_MAPPING_NAMES = OrderedDict(
         ("longt5", "FlaxLongT5Model"),
         ("marian", "FlaxMarianModel"),
         ("mbart", "FlaxMBartModel"),
+        ("mistral", "FlaxMistralModel"),
         ("mt5", "FlaxMT5Model"),
         ("opt", "FlaxOPTModel"),
         ("pegasus", "FlaxPegasusModel"),
@@ -148,6 +149,7 @@ FLAX_MODEL_FOR_CAUSAL_LM_MAPPING_NAMES = OrderedDict(
         ("gpt_neo", "FlaxGPTNeoForCausalLM"),
         ("gptj", "FlaxGPTJForCausalLM"),
         ("llama", "FlaxLlamaForCausalLM"),
+        ("mistral", "FlaxMistralForCausalLM"),
         ("opt", "FlaxOPTForCausalLM"),
         ("roberta", "FlaxRobertaForCausalLM"),
         ("roberta-prelayernorm", "FlaxRobertaPreLayerNormForCausalLM"),

src/transformers/models/mistral/__init__.py

@@ -13,11 +13,7 @@
 # limitations under the License.
 from typing import TYPE_CHECKING
 
-from ...utils import (
-    OptionalDependencyNotAvailable,
-    _LazyModule,
-    is_torch_available,
-)
+from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_flax_available, is_torch_available
 
 
 _import_structure = {
@@ -38,6 +34,18 @@ else:
         "MistralForSequenceClassification",
     ]
 
+try:
+    if not is_flax_available():
+        raise OptionalDependencyNotAvailable()
+except OptionalDependencyNotAvailable:
+    pass
+else:
+    _import_structure["modeling_flax_mistral"] = [
+        "FlaxMistralForCausalLM",
+        "FlaxMistralModel",
+        "FlaxMistralPreTrainedModel",
+    ]
+
 
 if TYPE_CHECKING:
     from .configuration_mistral import MISTRAL_PRETRAINED_CONFIG_ARCHIVE_MAP, MistralConfig
@@ -55,6 +63,18 @@ if TYPE_CHECKING:
             MistralPreTrainedModel,
         )
 
+    try:
+        if not is_flax_available():
+            raise OptionalDependencyNotAvailable()
+    except OptionalDependencyNotAvailable:
+        pass
+    else:
+        from .modeling_flax_mistral import (
+            FlaxMistralForCausalLM,
+            FlaxMistralModel,
+            FlaxMistralPreTrainedModel,
+        )
+
 
 else:
     import sys

src/transformers/utils/dummy_flax_objects.py

@@ -898,6 +898,27 @@ class FlaxMBartPreTrainedModel(metaclass=DummyObject):
         requires_backends(self, ["flax"])
 
 
+class FlaxMistralForCausalLM(metaclass=DummyObject):
+    _backends = ["flax"]
+
+    def __init__(self, *args, **kwargs):
+        requires_backends(self, ["flax"])
+
+
+class FlaxMistralModel(metaclass=DummyObject):
+    _backends = ["flax"]
+
+    def __init__(self, *args, **kwargs):
+        requires_backends(self, ["flax"])
+
+
+class FlaxMistralPreTrainedModel(metaclass=DummyObject):
+    _backends = ["flax"]
+
+    def __init__(self, *args, **kwargs):
+        requires_backends(self, ["flax"])
+
+
 class FlaxMT5EncoderModel(metaclass=DummyObject):
     _backends = ["flax"]
 

tests/models/mistral/test_modeling_flax_mistral.py

+# Copyright 2023 The HuggingFace 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.
+
+
+import unittest
+
+import numpy as np
+
+from transformers import MistralConfig, is_flax_available, is_tokenizers_available
+from transformers.testing_utils import require_flax, slow
+
+from ...generation.test_flax_utils import FlaxGenerationTesterMixin
+from ...test_modeling_flax_common import FlaxModelTesterMixin, ids_tensor
+
+
+if is_flax_available():
+    import jax.numpy as jnp
+
+    from transformers.models.mistral.modeling_flax_mistral import (
+        FlaxMistralForCausalLM,
+        FlaxMistralModel,
+    )
+
+
+if is_tokenizers_available():
+    from transformers import LlamaTokenizerFast
+
+
+class FlaxMistralModelTester:
+    def __init__(
+        self,
+        parent,
+        batch_size=2,
+        seq_length=7,
+        is_training=True,
+        use_input_mask=True,
+        use_token_type_ids=False,
+        use_labels=True,
+        vocab_size=99,
+        hidden_size=32,
+        num_hidden_layers=2,
+        num_attention_heads=4,
+        num_key_value_heads=2,
+        intermediate_size=37,
+        hidden_act="gelu",
+        hidden_dropout_prob=0.1,
+        attention_probs_dropout_prob=0.1,
+        max_position_embeddings=512,
+        window_size=7,
+        initializer_range=0.02,
+    ):
+        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.num_key_value_heads = num_key_value_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.window_size = window_size
+        self.initializer_range = initializer_range
+        self.scope = None
+        self.bos_token_id = vocab_size - 1
+        self.eos_token_id = vocab_size - 1
+        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 = np.tril(np.ones((self.batch_size, self.seq_length)))
+
+        config = MistralConfig(
+            vocab_size=self.vocab_size,
+            hidden_size=self.hidden_size,
+            num_hidden_layers=self.num_hidden_layers,
+            num_attention_heads=self.num_attention_heads,
+            num_key_value_heads=self.num_key_value_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,
+            use_cache=True,
+            is_decoder=False,
+            initializer_range=self.initializer_range,
+            sliding_window=self.window_size,
+        )
+        config.pad_token_id = config.eos_token_id
+
+        return (config, input_ids, input_mask)
+
+    # Copied from tests.models.gpt_neo.test_modeling_flax_gpt_neo.FlaxGPTNeoModelTester.prepare_config_and_inputs_for_common
+    def prepare_config_and_inputs_for_common(self):
+        config_and_inputs = self.prepare_config_and_inputs()
+        config, input_ids, attention_mask = config_and_inputs
+        inputs_dict = {"input_ids": input_ids, "attention_mask": attention_mask}
+        return config, inputs_dict
+
+    # Copied from tests.models.gpt_neo.test_modeling_flax_gpt_neo.FlaxGPTNeoModelTester.check_use_cache_forward
+    def check_use_cache_forward(self, model_class_name, config, input_ids, attention_mask):
+        max_decoder_length = 20
+        model = model_class_name(config)
+
+        past_key_values = model.init_cache(input_ids.shape[0], max_decoder_length)
+        attention_mask = jnp.ones((input_ids.shape[0], max_decoder_length), dtype="i4")
+
+        position_ids = jnp.broadcast_to(
+            jnp.arange(input_ids.shape[-1] - 1)[None, :], (input_ids.shape[0], input_ids.shape[-1] - 1)
+        )
+        outputs_cache = model(
+            input_ids[:, :-1],
+            attention_mask=attention_mask,
+            past_key_values=past_key_values,
+            position_ids=position_ids,
+        )
+
+        position_ids = jnp.array(input_ids.shape[0] * [[input_ids.shape[-1] - 1]], dtype="i4")
+        outputs_cache_next = model(
+            input_ids[:, -1:],
+            attention_mask=attention_mask,
+            past_key_values=outputs_cache.past_key_values,
+            position_ids=position_ids,
+        )
+
+        outputs = model(input_ids)
+
+        diff = np.max(np.abs((outputs_cache_next[0][:, -1, :5] - outputs[0][:, -1, :5])))
+        self.parent.assertTrue(diff < 1e-3, msg=f"Max diff is {diff}")
+
+    # Copied from tests.models.gpt_neo.test_modeling_flax_gpt_neo.FlaxGPTNeoModelTester.check_use_cache_forward_with_attn_mask
+    def check_use_cache_forward_with_attn_mask(self, model_class_name, config, input_ids, attention_mask):
+        max_decoder_length = 20
+        model = model_class_name(config)
+
+        attention_mask_cache = jnp.concatenate(
+            [attention_mask, jnp.zeros((attention_mask.shape[0], max_decoder_length - attention_mask.shape[1]))],
+            axis=-1,
+        )
+
+        past_key_values = model.init_cache(input_ids.shape[0], max_decoder_length)
+        position_ids = jnp.broadcast_to(
+            jnp.arange(input_ids.shape[-1] - 1)[None, :], (input_ids.shape[0], input_ids.shape[-1] - 1)
+        )
+
+        outputs_cache = model(
+            input_ids[:, :-1],
+            attention_mask=attention_mask_cache,
+            past_key_values=past_key_values,
+            position_ids=position_ids,
+        )
+        position_ids = jnp.array(input_ids.shape[0] * [[input_ids.shape[-1] - 1]], dtype="i4")
+        outputs_cache_next = model(
+            input_ids[:, -1:],
+            past_key_values=outputs_cache.past_key_values,
+            attention_mask=attention_mask_cache,
+            position_ids=position_ids,
+        )
+
+        outputs = model(input_ids, attention_mask=attention_mask)
+
+        diff = np.max(np.abs((outputs_cache_next[0][:, -1, :5] - outputs[0][:, -1, :5])))
+        self.parent.assertTrue(diff < 1e-3, msg=f"Max diff is {diff}")
+
+
+@require_flax
+class FlaxMistralModelTest(FlaxModelTesterMixin, FlaxGenerationTesterMixin, unittest.TestCase):
+    all_model_classes = (FlaxMistralModel, FlaxMistralForCausalLM) if is_flax_available() else ()
+    all_generative_model_classes = (FlaxMistralForCausalLM,) if is_flax_available() else ()
+
+    def setUp(self):
+        self.model_tester = FlaxMistralModelTester(self)
+
+    def test_use_cache_forward(self):
+        for model_class_name in self.all_model_classes:
+            config, input_ids, attention_mask = self.model_tester.prepare_config_and_inputs()
+            self.model_tester.check_use_cache_forward(model_class_name, config, input_ids, attention_mask)
+
+    def test_use_cache_forward_with_attn_mask(self):
+        for model_class_name in self.all_model_classes:
+            config, input_ids, attention_mask = self.model_tester.prepare_config_and_inputs()
+            self.model_tester.check_use_cache_forward_with_attn_mask(
+                model_class_name, config, input_ids, attention_mask
+            )
+
+    @slow
+    def test_model_from_pretrained(self):
+        for model_class_name in self.all_model_classes:
+            model = model_class_name.from_pretrained("mistralai/Mistral-7B-v0.1", from_pt=True)
+            outputs = model(np.ones((1, 1)))
+            self.assertIsNotNone(outputs)
+
+
+@slow
+@require_flax
+class FlaxMistralIntegrationTest(unittest.TestCase):
+    def setUp(self):
+        self.model_id = "mistralai/Mistral-7B-v0.1"
+        self.model = FlaxMistralForCausalLM.from_pretrained(self.model_id, from_pt=True)
+        self.test_batch = jnp.arange(32).reshape(4, 8) + 1911
+
+    def test_model_logits(self):
+        input_ids = jnp.array([[1, 306, 4658, 278, 6593, 310, 2834, 338]])
+        EXPECTED_MEAN = np.array([[-2.5548, -2.5737, -3.0600, -2.5906, -2.8478, -2.8118, -2.9325, -2.7694]])
+        EXPECTED_SLICE = np.array([-5.8781,-5.8616,-0.1052,-4.7200,-5.8781,-5.8774,-5.8773,-5.8777,-5.8781,-5.8780,-5.8781,-5.8779,-1.0787,1.7583,-5.8779,-5.8780,-5.8783,-5.8778,-5.8776,-5.8781,-5.8784,-5.8778,-5.8778,-5.8777,-5.8779,-5.8778,-5.8776,-5.8780,-5.8779,-5.8781])  # fmt: skip
+
+        flax_logits = self.model(input_ids).logits
+        diff_mean = jnp.abs(flax_logits.mean(-1) - EXPECTED_MEAN).max()
+        diff_slice = jnp.abs(flax_logits[0, 0, :30] - EXPECTED_SLICE).max()
+
+        self.assertAlmostEqual(diff_mean, 0, places=3)
+        self.assertAlmostEqual(diff_slice, 0, places=3)
+
+    def test_generated_text(self):
+        tokenizer = LlamaTokenizerFast.from_pretrained(self.model_id)
+        tokenizer.pad_token_id = 2
+        EXPECTED_TEXT_COMPLETION = """My favourite condiment is 100% ketchup. I love it on everything. I’m not a big"""
+        prompt = "My favourite condiment is "
+        inputs = tokenizer(prompt, return_tensors="np", truncation=True, padding=True)
+        generated_ids = self.model.generate(**inputs, max_new_tokens=20, temperature=0).sequences
+        generated_text = tokenizer.decode(generated_ids[0], skip_special_tokens=True)
+        self.assertEqual(generated_text, EXPECTED_TEXT_COMPLETION)

utils/check_docstrings.py

@@ -239,6 +239,8 @@ OBJECTS_TO_IGNORE = [
     "FlaxMBartModel",
     "FlaxMarianMTModel",
     "FlaxMarianModel",
+    "FlaxMistralForCausalLM",
+    "FlaxMistralModel",
     "FlaxOPTForCausalLM",
     "FlaxPegasusForConditionalGeneration",
     "FlaxPegasusModel",