Help to converto to gguf


i used SFTTrainer to fine tune a model (opt-350m) with a dataset, now i have a folder with several files and a model.tensorflow.

i used to convert but this generated a keyerror: intermediate_size

python3 []( ./models/salidamodelo/
Loading model file models/salidamodelo/model.safetensors
Traceback (most recent call last):
File “/home/con/llama.cpp/”, line 1471, in
File “/home/con/llama.cpp/”, line 1418, in main
params = Params.load(model_plus)
File “/home/con/llama.cpp/”, line 317, in load
params = Params.loadHFTransformerJson(model_plus.model, hf_config_path)
File “/home/con/llama.cpp/”, line 251, in loadHFTransformerJson
n_ff = config[“intermediate_size”],
KeyError: ‘intermediate_size’

how i can solve it?
thanks in advance

What’s in your config.json? Might be that key is missing?

“_name_or_path”: “facebook/opt-350m”,
“_remove_final_layer_norm”: false,
“activation_dropout”: 0.0,
“activation_function”: “relu”,
“architectures”: [
“attention_dropout”: 0.0,
“bos_token_id”: 2,
“do_layer_norm_before”: false,
“dropout”: 0.1,
“enable_bias”: true,
“eos_token_id”: 2,
“ffn_dim”: 4096,
“hidden_size”: 1024,
“init_std”: 0.02,
“layer_norm_elementwise_affine”: true,
“layerdrop”: 0.0,
“max_position_embeddings”: 2048,
“model_type”: “opt”,
“num_attention_heads”: 16,
“num_hidden_layers”: 24,
“pad_token_id”: 1,
“prefix”: “”,
“torch_dtype”: “float32”,
“transformers_version”: “4.37.0”,
“use_cache”: true,
“vocab_size”: 50272,
“word_embed_proj_dim”: 512