Fix typos discovered by codespell

README.md

@@ -52,7 +52,7 @@ StarVector is a multimodal vision-language model for Scalable Vector Graphics (S
 
 ### Multimodal Architecture
 
-StarVector uses a multimodal architecture to process images and text. When performing Image-to-SVG (or image vectorization), the image is projected into visual tokens, and SVG code is generated. When performing Text-to-SVG, the model only recieves the text instruction (no image is provided), and a novel SVG is created. The LLM is based of StarCoder, which we leverage to transfer coding skills to SVG generation.
+StarVector uses a multimodal architecture to process images and text. When performing Image-to-SVG (or image vectorization), the image is projected into visual tokens, and SVG code is generated. When performing Text-to-SVG, the model only receives the text instruction (no image is provided), and a novel SVG is created. The LLM is based of StarCoder, which we leverage to transfer coding skills to SVG generation.
 
 <div align="center">
   <img src="assets/starvector-arch.png" alt="starvector" style="width: 700px; display: block; margin-left: auto; margin-right: auto;" />

starvector/metrics/inception.py

@@ -23,7 +23,7 @@ class InceptionV3(nn.Module):
     # Maps feature dimensionality to their output blocks indices
     BLOCK_INDEX_BY_DIM = {
         64: 0,   # First max pooling features
-        192: 1,  # Second max pooling featurs
+        192: 1,  # Second max pooling features
         768: 2,  # Pre-aux classifier features
         2048: 3  # Final average pooling features
     }
@@ -171,7 +171,7 @@ def _inception_v3(*args, **kwargs):
         # Just a caution against weird version strings
         version = (0,)
 
-    # Skips default weight inititialization if supported by torchvision
+    # Skips default weight initialization if supported by torchvision
     # version. See https://github.com/mseitzer/pytorch-fid/issues/28.
     if version >= (0, 6):
         kwargs['init_weights'] = False

starvector/model/gpt_bigcode/modeling_gpt_bigcode.py

@@ -297,7 +297,7 @@ def __init__(self, *args, **kwargs):
         super().__init__(*args, **kwargs)
 
         # TODO: Should be removed once Flash Attention for RoCm is bumped to 2.1.
-        # flash_attn<2.1 generates top-left aligned causal mask, while what is needed here is bottom-right alignement, that was made default for flash_attn>=2.1. This attribute is used to handle this difference. Reference: https://github.com/Dao-AILab/flash-attention/releases/tag/v2.1.0.
+        # flash_attn<2.1 generates top-left aligned causal mask, while what is needed here is bottom-right alignment, that was made default for flash_attn>=2.1. This attribute is used to handle this difference. Reference: https://github.com/Dao-AILab/flash-attention/releases/tag/v2.1.0.
         # Beware that with flash_attn<2.1, using q_seqlen != k_seqlen (except for the case q_seqlen == 1) produces a wrong mask (top-left).
         self._flash_attn_uses_top_left_mask = not is_flash_attn_greater_or_equal_2_10()
 

starvector/model/llm/starcoder.py

@@ -37,7 +37,7 @@ def __init__(self, config, **kwargs):
 
     def init_tokenizer(self, model_name):
         self.tokenizer = AutoTokenizer.from_pretrained(model_name, use_fast=False)
-        # Incude padding and eos tokens in the vocabulary
+        # Include padding and eos tokens in the vocabulary
         if self.tokenizer.eos_token_id is None:
             self.tokenizer.add_special_tokens({"eos_token": "[EOS]"})
         if self.tokenizer.pad_token_id is None:

starvector/model/llm/starcoder2.py

@@ -36,7 +36,7 @@ def __init__(self, config, **kwargs):
 
     def init_tokenizer(self, model_name):
         self.tokenizer = AutoTokenizer.from_pretrained(model_name, use_fast=False)
-        # Incude padding and eos tokens in the vocabulary
+        # Include padding and eos tokens in the vocabulary
         if self.tokenizer.eos_token_id is None:
             self.tokenizer.add_special_tokens({"eos_token": "[EOS]"})
         if self.tokenizer.pad_token_id is None:

starvector/train/util.py

@@ -214,7 +214,7 @@ def setup_train_env_variables(config):
 
 def load_fsdp_plugin(config, model):
     if config.fsdp.enable:
-        # get mixed precsion dtype
+        # get mixed precision dtype
         mixed_precision_dtype = {
             "fp16": torch.float16,
             "bf16": torch.bfloat16,