We lifted 11 functions out of this paper's own repositories and ran 10 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.
| Repository | Role | Ran |
|---|---|---|
| flagai-open/flagai | canonical | 10 of 11 |
| Function | Status | Where it lives |
|---|---|---|
| accuracy_metric | Ran | flagai-open/flagai/flagai/metrics.py code served (permissive licence) · get_code("e923a6d5aaa37270") |
| change_json_to_cls | Ran | flagai-open/flagai/flagai/model/base_model.py code served (permissive licence) · get_code("9a58da55dcd68a4b") |
| cmd_load_hyperparam | Ran | flagai-open/flagai/flagai/launch.py code served (permissive licence) · get_code("329c97f25378e517") |
| create_custom_forward | Ran | flagai-open/flagai/flagai/model/aquila_model.py code served (permissive licence) · get_code("fd6777c43a4e9d66") |
| get_args_list | Ran | flagai-open/flagai/flagai/env_trainer.py code served (permissive licence) · get_code("2774d56117e0ff9f") |
| get_args_list | Ran | flagai-open/flagai/flagai/env_trainer_v1.py code served (permissive licence) · get_code("8977724aa4056e15") |
| save_best | Ran | flagai-open/flagai/flagai/env_args.py code served (permissive licence) · get_code("5011d61e17e5dcd5") |
| sigmoid | Ran | flagai-open/flagai/flagai/metrics.py code served (permissive licence) · get_code("98a2b2f174e040a3") |
| str2bool | Ran | flagai-open/flagai/flagai/env_args.py code served (permissive licence) · get_code("92a5f0d0912c3a91") |
| to_python_float | Ran | flagai-open/flagai/flagai/fp16/loss_scaler.py code served (permissive licence) · get_code("6b5d363ab4dbf012") |
| should_log_le | Not yet run | flagai-open/flagai/flagai/logger.py code served (permissive licence) · get_code("4e8fb0839eac7d8c") |
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
In this work, we present a conceptually simple and effective method to train a strong bilingual/multilingual multimodal representation model. Starting from the pre-trained multimodal representation model CLIP released by OpenAI, we altered its text encoder with a pre-trained multilingual text encoder XLM-R, and aligned both languages and image representations by a two-stage training schema consisting of teacher learning and contrastive learning. We validate our method through evaluations of a wide range of tasks. We set new state-of-the-art performances on a bunch of tasks including ImageNet-CN, Flicker30k-CN, COCO-CN and XTD. Further, we obtain very close performances with CLIP on almost all tasks, suggesting that one can simply alter the text encoder in CLIP for extended capabilities such as multilingual understanding. Our models and code are available at https://github.com/FlagAI-Open/FlagAI.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2211.06679")
get_code_for_paper("2211.06679")
have("2211.06679")
Connect an agent — have() is free.