|
Download docs/PRETRAINED.md from devildasdf/devils-agent: direct link, hf CLI and curl.
- Browser
- Download file 1.92 kB
-
https://huggingface.co/devildasdf/devils-agent/resolve/main/docs/PRETRAINED.md
- Command line
-
hf download hf://devildasdf/devils-agent/docs/PRETRAINED.md
-
curl -L -o PRETRAINED.md https://huggingface.co/devildasdf/devils-agent/resolve/main/docs/PRETRAINED.md
1.92 kB
| # Pretrained diagnostic baselines | |
| Pinned local downloads for research only; no external inference endpoint is used. | |
| | Model | Revision | Hub license metadata | Purpose | | |
| |---|---|---|---| | |
| | Qwen/Qwen2.5-0.5B-Instruct | 7ae557604adf67be50417f59c2c2f167def9a775 | Apache-2.0 | Local autoregressive action baseline | | |
| | cross-encoder/ms-marco-MiniLM-L6-v2 | 233902d25c440f23af6f7d6e94d2946bac0bee0a | Apache-2.0 | Semantic candidate reranking | | |
| Models are loaded with local_files_only=True and trust_remote_code=False. | |
| Model weights use safetensors. Qwen's source LICENSE and both model cards were | |
| downloaded with the checkpoints under the intermediate work directory. | |
| Sources: [Qwen model card](https://huggingface.co/Qwen/Qwen2.5-0.5B-Instruct), | |
| [MiniLM model card](https://huggingface.co/cross-encoder/ms-marco-MiniLM-L6-v2). | |
| The metadata and revisions were checked through the authenticated HF CLI. | |
| Run from the project root after installing research extras: | |
| ```sh | |
| python -m baim.retrieval_audit --source ../../work/mind2web-source/data/train/train_10.json --model ../../work/pretrained/minilm-cross | |
| python -m baim.qwen_baseline --source ../../work/mind2web-source/data/train/train_10.json --model ../../work/pretrained/qwen-0.5b --limit 8 | |
| ``` | |
| History is teacher-forced: only prior human action descriptions are supplied. | |
| Neither evaluator receives the current labelled action or target as model input. | |
| This is not autonomous multi-step success. Only aggregate metrics and numeric | |
| per-step timing results are persisted, not raw page text, prompts or generations. | |
| The Qwen diagnostic scores action/target matching, not typed-value correctness. | |
| The broad BM25 candidate pool includes generic roles because source annotations | |
| already classify those records as candidates. This offline choice is not a live | |
| DOM interaction policy. Live generic elements still require interactability | |
| evidence and execution validation. | |