Download testbed/huggingface__trl/docs/source/models.mdx from hc99/FEA-Bench: direct link, hf CLI and curl.
- Browser
- Download file 906 Bytes
-
https://huggingface.co/datasets/hc99/FEA-Bench/resolve/main/testbed/huggingface__trl/docs/source/models.mdx
- Command line
-
hf download hf://datasets/hc99/FEA-Bench/testbed/huggingface__trl/docs/source/models.mdx
-
curl -L -o models.mdx https://huggingface.co/datasets/hc99/FEA-Bench/resolve/main/testbed/huggingface__trl/docs/source/models.mdx
906 Bytes
| # Models | |
| With the `AutoModelForCausalLMWithValueHead` class TRL supports all decoder model architectures in transformers such as GPT-2, OPT, and GPT-Neo. In addition, with `AutoModelForSeq2SeqLMWithValueHead` you can use encoder-decoder architectures such as T5. TRL also requires reference models which are frozen copies of the model that is trained. With `create_reference_model` you can easily create a frozen copy and also share layers between the two models to save memory. | |
| ## PreTrainedModelWrapper | |
| [[autodoc]] PreTrainedModelWrapper | |
| ## AutoModelForCausalLMWithValueHead | |
| [[autodoc]] AutoModelForCausalLMWithValueHead | |
| - __init__ | |
| - forward | |
| - generate | |
| - _init_weights | |
| ## AutoModelForSeq2SeqLMWithValueHead | |
| [[autodoc]] AutoModelForSeq2SeqLMWithValueHead | |
| - __init__ | |
| - forward | |
| - generate | |
| - _init_weights | |
| ## create_reference_model | |
| [[autodoc]] create_reference_model |