Instructions to use aac6fef/laya-mlx with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use aac6fef/laya-mlx with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir laya-mlx aac6fef/laya-mlx
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
Download rl_agent_config.json from aac6fef/laya-mlx: direct link, hf CLI and curl.
- Browser
- Download file 746 Bytes
-
https://huggingface.co/aac6fef/laya-mlx/resolve/main/rl_agent_config.json
- Command line
-
hf download hf://aac6fef/laya-mlx/rl_agent_config.json
-
curl -L -o rl_agent_config.json https://huggingface.co/aac6fef/laya-mlx/resolve/main/rl_agent_config.json
746 Bytes
| { | |
| "encoder": "answerdotai/ModernBERT-large", | |
| "head_layers": 2, | |
| "max_len": 512, | |
| "head_max_len": 192, | |
| "max_prefixes": 6, | |
| "act_costs": { | |
| "escalate": 0.5 | |
| }, | |
| "cost_wrong_act": 3.0, | |
| "amp_dtype": "bf16", | |
| "model_name": "rl-agent", | |
| "temperature": [ | |
| 1.6369030475616455, | |
| 1.2514300346374512, | |
| 1.983399510383606 | |
| ], | |
| "temperature_by_options": { | |
| "choice:3-5": 1.7601518630981445, | |
| "choice:6-10": 1.0000158548355103, | |
| "score:3-5": 1.2514300346374512, | |
| "noul:2": 1.983399510383606, | |
| "choice:11+": 0.10058280825614929, | |
| "choice:2": 1.9063563346862793 | |
| }, | |
| "training": { | |
| "updates": 7313, | |
| "epochs_completed": 1, | |
| "hours": 1.96, | |
| "world_size": 1, | |
| "fine_tuned_from_checkpoint": true | |
| } | |
| } | |