LLM/chandra
1
0
Fork 0
mirror of https://github.com/datalab-to/chandra.git synced 2026-01-20 05:50:42 +00:00
Code Issues Packages Projects Releases Wiki Activity

Code cleanup

Browse Source
This commit is contained in:
Vik Paruchuri
2025-10-21 12:11:37 -04:00
parent 2151833414
commit 47bd444f20
3 changed files with 10 additions and 25 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
README.md
View File

@@ -1,6 +1,6 @@
# Chandra
Chandra is an OCR model that converts images and PDFs into structured HTML/Markdown/JSON while preserving layout information.
Chandra is a highly accurate OCR model that converts images and PDFs into structured HTML/Markdown/JSON while preserving layout information.
## Features
@@ -154,7 +154,11 @@ VLLM_MODEL_NAME=chandra
VLLM_GPUS=0
```
## Benchmark table
# Commercial usage
This code is Apache 2.0, and our model weights use a modified OpenRAIL-M license (free for research, personal use, and startups under $2M funding/revenue, cannot be used competitively with our API). To remove the OpenRAIL license requirements, or for broader commercial licensing, visit our pricing page [here](https://www.datalab.to/pricing?utm_source=gh-chandra).
# Benchmark table
| **Model** | ArXiv | Old Scans Math | Tables | Old Scans | Headers and Footers | Multi column | Long tiny text | Base | Overall | Source |
|:--------------------------|:--------:|:--------------:|:--------:|:---------:|:-------------------:|:------------:|:--------------:|:----:|:--------------:|:------:|
@@ -168,9 +172,6 @@ VLLM_GPUS=0
| olmOCR v0.3.0 | 78.6 | 79.9 | 72.9 | 43.9 | **95.1** | 77.3 | 81.2 | 98.9 | 78.5 ± 1.1 | olmocr repo |
| dots.ocr | 82.1 | 64.2 | 88.3 | 40.9 | 94.1 | **82.4** | 81.2 | 99.5 | 79.1 ± 1.0 | dots.ocr repo |
# Commercial usage
This code is Apache 2.0, and our model weights use a modified OpenRAIL-M license (free for research, personal use, and startups under $2M funding/revenue, cannot be used competitively with our API). To remove the OpenRAIL license requirements, or for broader commercial licensing, visit our pricing page [here](https://www.datalab.to/pricing?utm_source=gh-chandra).
# Credits

9
chandra/model/util.py
View File

@@ -43,7 +43,10 @@ def scale_to_fit(
def detect_repeat_token(
predicted_tokens: str, max_repeats: int = 4, window_size: int = 500, cut_from_end: int = 0
predicted_tokens: str,
max_repeats: int = 4,
window_size: int = 500,
cut_from_end: int = 0,
):
try:
predicted_tokens = parse_markdown(predicted_tokens)
@@ -77,7 +80,3 @@ def detect_repeat_token(
return True
return False
def layout_failed(predicted_tokens: str, image: Image.Image):
pass

15
chandra/settings.py
View File

@@ -22,21 +22,6 @@ class Settings(BaseSettings):
VLLM_GPUS: str = "0"
MAX_VLLM_RETRIES: int = 6
# Transformers settings
@computed_field
@property
def TORCH_DEVICE_MODEL(self) -> str:
if self.TORCH_DEVICE is not None:
return self.TORCH_DEVICE
if torch.cuda.is_available():
return "cuda"
if torch.backends.mps.is_available():
return "mps"
return "cpu"
@computed_field
@property
def TORCH_DTYPE(self) -> torch.dtype: