Image-to-Text
PEFT
Safetensors
Portuguese
English
vision-language
table-extraction
scientific-figures
markdown-table
qwen2.5-vl
lora
icdar-metric-loss
Instructions to use lucasoc/sci-image-models with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use lucasoc/sci-image-models with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen2.5-VL-3B-Instruct") model = PeftModel.from_pretrained(base_model, "lucasoc/sci-image-models") - Notebooks
- Google Colab
- Kaggle
File size: 1,138 Bytes
be90b31 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 | """
I/O utilities for reading and writing dataset files (JSONL, JSON, CSV).
"""
import json
import os
from typing import Any, Dict, List, Generator
def read_jsonl(file_path: str) -> List[Dict[str, Any]]:
"""Reads a JSONL file into a list of dictionaries."""
records = []
with open(file_path, "r", encoding="utf-8") as f:
for line in f:
line = line.strip()
if line:
records.append(json.loads(line))
return records
def iter_jsonl(file_path: str) -> Generator[Dict[str, Any], None, None]:
"""Yields records one by one from a JSONL file."""
with open(file_path, "r", encoding="utf-8") as f:
for line in f:
line = line.strip()
if line:
yield json.loads(line)
def write_jsonl(records: List[Dict[str, Any]], file_path: str) -> None:
"""Writes a list of dictionaries to a JSONL file."""
os.makedirs(os.path.dirname(os.path.abspath(file_path)), exist_ok=True)
with open(file_path, "w", encoding="utf-8") as f:
for record in records:
f.write(json.dumps(record, ensure_ascii=False) + "\n")
|