Spaces:
Sleeping
Sleeping
| import concurrent.futures | |
| from extractors.model import AnyParserModel, LlamaParseModel, UnstructuredModel, GPTModel, ClaudeModel | |
| ap_rt = AnyParserModel() | |
| lp = LlamaParseModel() | |
| un = UnstructuredModel() | |
| gpt = GPTModel() | |
| claude = ClaudeModel() | |
| model_function_map = { | |
| "AnyParser": ap_rt.extract, | |
| "LlamaParse": lp.extract, | |
| "Unstructured": un.extract, | |
| "GPT-4o-mini": gpt.extract, | |
| "Claude-3.5-Sonnet": claude.extract, | |
| } | |
| models = [key for key in model_function_map] | |
| def run_extract(model, file_path): | |
| print('Running extract: model', model, 'file_path', file_path) | |
| extractor = model_function_map[model] | |
| markdown = extractor(file_path) | |
| return markdown | |
| def run_extract_parallel(model_a, model_b, pdf): | |
| with concurrent.futures.ThreadPoolExecutor() as executor: | |
| # Submit tasks to the executor for parallel execution | |
| future_a = executor.submit(run_extract, model_a, pdf) | |
| future_b = executor.submit(run_extract, model_b, pdf) | |
| # Get the results as they complete | |
| result_a = future_a.result() | |
| result_b = future_b.result() | |
| return result_a, result_b | |