CodeBertForCodeTrans

This model is a fine-tuned version of microsoft/codebert-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0006

Model description

More information needed

Driectly uses

from transformers import AutoTokenizer, AutoModelForCausalLM
additional_special_tokens = {'additional_special_tokens':['<|begin_of_java_code|>','<|end_of_java_code|>'\
                                                           ,'<|begin_of_c-sharp_code|>','<|end_of_c-sharp_code|>',\
                                                            '<|translate|>']}
basemodel = "ljcnju/CodeBertForCodeTrans"
tokenizer = AutoTokenizer.from_pretrained(basemodel)
tokenizer.pad_token = tokenizer.eos_token
config = AutoConfig.from_pretrained(basemodel)
config.is_decoder = True
model = AutoModelForCausalLM.from_pretrained(basemodel,config=config)
device = torch.device("cuda") if torch.cuda.is_available() else torch.device('cpu') 
model.to(device)

code = "public void serialize(LittleEndianOutput out) {out.writeShort(field_1_vcenter);}\n"
prefix =  additional_special_tokens['additional_special_tokens'][0] 
input_str = prefix + code +additional_special_tokens['additional_special_tokens'][1] + additional_special_tokens['additional_special_tokens'][2]
input = tokenizer(input_str,return_tensors = "pt")
output = model.generate(**input, max_length = 256)
outputs_str = tokenizer.decode(output[0])
print(outputs_str)

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 12354.0
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
5.7169	1.0	644	4.5075
3.0571	2.0	1288	2.1423
0.7391	3.0	1932	0.2866
0.1028	4.0	2576	0.0219
0.0158	5.0	3220	0.0047
0.0065	6.0	3864	0.0024
0.0036	7.0	4508	0.0020
0.0028	8.0	5152	0.0014
0.0018	9.0	5796	0.0010
0.0023	10.0	6440	0.0017
0.002	11.0	7084	0.0009
0.002	12.0	7728	0.0012
0.0015	13.0	8372	0.0020
0.0028	14.0	9016	0.0010
0.0015	15.0	9660	0.0007
0.0027	16.0	10304	0.0015
0.002	17.0	10948	0.0007
0.0011	18.0	11592	0.0009
0.0019	19.0	12236	0.0007
0.0003	20.0	12880	0.0006

Framework versions

• Transformers 4.37.2
• Pytorch 2.1.2+cu121
• Datasets 2.15.0
• Tokenizers 0.15.0