rename idefics3 HF repo

src/transformers/models/idefics3/convert_idefics3_weights_to_hf.py

@@ -31,7 +31,7 @@
 
 
 EPILOG_TXT = """Example:
-    python transformers/src/transformers/models/idefics3/convert_idefics3_weights_to_hf.py --original_model_id HuggingFaceM4/idefics3-8b --output_hub_path org/idefics3
+    python transformers/src/transformers/models/idefics3/convert_idefics3_weights_to_hf.py --original_model_id HuggingFaceM4/Idefics3-8B-Llama3 --output_hub_path org/idefics3
 """
 
 

src/transformers/models/idefics3/modeling_idefics3.py

@@ -1114,8 +1114,8 @@ def forward(
         >>> image2 = load_image("https://cdn.britannica.com/59/94459-050-DBA42467/Skyline-Chicago.jpg")
         >>> image3 = load_image("https://cdn.britannica.com/68/170868-050-8DDE8263/Golden-Gate-Bridge-San-Francisco.jpg")
 
-        >>> processor = AutoProcessor.from_pretrained("HuggingFaceM4/idefics3-8b")
-        >>> model = AutoModelForVision2Seq.from_pretrained("HuggingFaceM4/idefics3-8b", device_map="auto")
+        >>> processor = AutoProcessor.from_pretrained("HuggingFaceM4/Idefics3-8B-Llama3")
+        >>> model = AutoModelForVision2Seq.from_pretrained("HuggingFaceM4/Idefics3-8B-Llama3", device_map="auto")
 
         >>> BAD_WORDS_IDS = processor.tokenizer(["<image>", "<fake_token_around_image>"], add_special_tokens=False).input_ids
         >>> EOS_WORDS_IDS = [processor.tokenizer.eos_token_id]

src/transformers/models/idefics3/processing_idefics3.py

@@ -145,7 +145,7 @@ def __call__(
         >>> from transformers import Idefics3Processor
         >>> from transformers.image_utils import load_image
 
-        >>> processor = Idefics3Processor.from_pretrained("HuggingFaceM4/idefics3-8b", image_seq_len=2)
+        >>> processor = Idefics3Processor.from_pretrained("HuggingFaceM4/Idefics3-8B-Llama3", image_seq_len=2)
         >>> processor.image_processor.do_image_splitting = False  # Force as False to simplify the example
 
         >>> url1 = "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"

tests/models/idefics3/test_modeling_idefics3.py

@@ -467,7 +467,7 @@ def test_resize_embeddings_untied(self):
 @require_torch
 class Idefics3ForConditionalGenerationIntegrationTest(unittest.TestCase):
     def setUp(self):
-        self.processor = AutoProcessor.from_pretrained("HuggingFaceM4/idefics3-8b-base")
+        self.processor = AutoProcessor.from_pretrained("HuggingFaceM4/Idefics3-8B-Llama3-base")
         self.image1 = Image.open(
             BytesIO(
                 requests.get(
@@ -493,7 +493,7 @@ def tearDown(self):
     @slow
     def test_integration_test(self):
         model = Idefics3ForConditionalGeneration.from_pretrained(
-            "HuggingFaceM4/idefics3-8b-base",
+            "HuggingFaceM4/Idefics3-8B-Llama3-base",
             torch_dtype=torch.bfloat16,
             device_map="auto",
         )
@@ -517,7 +517,7 @@ def test_integration_test(self):
     def test_integration_test_4bit(self):
         # Let' s make sure we test the preprocessing to replace what is used
         model = Idefics3ForConditionalGeneration.from_pretrained(
-            "HuggingFaceM4/idefics3-8b-base", load_in_4bit=True, device_map="auto"
+            "HuggingFaceM4/Idefics3-8B-Llama3-base", load_in_4bit=True, device_map="auto"
         )
 
         # Create pixel inputs

tests/models/idefics3/test_processing_idefics3.py

@@ -31,7 +31,7 @@
 @require_vision
 class Idefics3ProcessorTest(unittest.TestCase):
     def setUp(self):
-        self.processor = Idefics3Processor.from_pretrained("HuggingFaceM4/idefics3-8b", image_seq_len=2)
+        self.processor = Idefics3Processor.from_pretrained("HuggingFaceM4/Idefics3-8B-Llama3", image_seq_len=2)
         self.image1 = Image.open(
             BytesIO(
                 requests.get(