Using Reader's results in Answer Generator & use other models to generate results

[Asutos7]

Hello,

Currently, I am working with the Seq2SeqGenerator and RAGenerator models. It seems that these models only accept a retriever as an argument, rather than a reader. I am curious to know if it is possible to use the results from the reader instead of the retriever in these functions.

Furthermore, I attempted to use the following code :
"""from haystack.nodes import Seq2SeqGenerator
generator = Seq2SeqGenerator(model_name_or_path="dbddv01/gpt2-french-small")"""
Since I am working with French documents, I thought it would be best to use a model that is already trained on French documents. However, when I ran the code, I encountered the following error:
"""Unrecognized configuration class <class 'transformers.models.gpt2.configuration_gpt2.GPT2Config'> for
this kind of AutoModel: AutoModelForSeq2SeqLM.
Model type should be one of BartConfig, BigBirdPegasusConfig, BlenderbotConfig, BlenderbotSmallConfig,
EncoderDecoderConfig, FSMTConfig, GPTSanJapaneseConfig, LEDConfig, LongT5Config, M2M100Config, MarianConfig,
MBartConfig, MT5Config, MvpConfig, NllbMoeConfig, PegasusConfig, PegasusXConfig, PLBartConfig, ProphetNetConfig,
SwitchTransformersConfig, T5Config, XLMProphetNetConfig."""
This error suggests that I can only use certain models (listed above) instead of the one I attempted to use. I am curious to know if it is possible to use alternative models instead of the ones mentioned in the error message.

[ZanSara]

Hey @Asutos7, my first suggestion would be to move away from Seq2SeqGenerator and RAGenerator entirely: they're deprecated and scheduled for removal in the next release, at the end of this month: #4816

[Asutos7]

Hey @ZanSara ! Thank you for your answer