# Wav2vec2 inference on my own model

**URL:** <https://discuss.huggingface.co/t/wav2vec2-inference-on-my-own-model/12028>\
**Category:** Beginners\
**Created:** [November 18, 2021, 3:43pm UTC](https://discuss.huggingface.co/t/wav2vec2-inference-on-my-own-model/12028 "2021-11-18T15:43:14Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![sanders](https://avatars.discourse-cdn.com/v4/letter/s/bc79bd/32.png) [@sanders](https://discuss.huggingface.co/u/sanders)\
**Post date:** [November 18, 2021, 3:43pm UTC](https://discuss.huggingface.co/t/wav2vec2-inference-on-my-own-model/12028/1 "2021-11-18T15:43:14Z")

</div>

Greetings, I have trained my own model using a fairseq setup. This process results in a `dict.ltr.txt`, `finutuned.pt`, `pretrainedd.pt`, `lexicon.txt` and `vocab.txt`. Also have a `kenlm.bin` but I understand that’s not yet supported.

My question is this - how do use these files with huggingface? It has it’s own setup with json files etc but I don’t know how to convert my set of files into a huggingface Transformers suitable set of files to do inference.

I haven’t been able to find anyone explaining via tutorial or otherwise if this is possible. Is it?