Reinforcement Learning
Transformers
Safetensors
qwen2
text-generation
webshop
verl
text-generation-inference
Instructions to use Saulp/Webshop-7B-RL-vllm with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Saulp/Webshop-7B-RL-vllm with Transformers:
# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("Saulp/Webshop-7B-RL-vllm") model = AutoModelForCausalLM.from_pretrained("Saulp/Webshop-7B-RL-vllm", device_map="auto") - Notebooks
- Google Colab
- Kaggle
WebShop-7B-RL (merged)
This repository is a standard Transformers export of Jianwen/Webshop-7B-RL. The source checkpoint stored the post-RL policy in FSDP actor shards; those shards were reconstructed without numerically mixing weights.
Jianwen/Webshop-7B-SFT is the SFT predecessor and is used only as a metadata fallback if the RL checkpoint lacks configuration or tokenizer files.
- Downloads last month
- 37