speech_recognition/FunASR
1
0
Fork 0
mirror of https://github.com/modelscope/FunASR synced 2025-09-15 14:48:36 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

fix bug

Browse Source
This commit is contained in:
游雁 2024-06-09 01:08:29 +08:00
parent b75d1e89bb
commit dcd40590df
1 changed files with 14 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

25
examples/industrial_data_pretraining/llm_asr/conf/whisper_qwen_linear2.yaml
View File

@ -59,17 +59,20 @@ scheduler_conf:
dataset: OpenAIDataset
dataset_conf:
index_ds: OpenAIIndexDSJsonl
batch_sampler: CustomDistributedBatchSampler
batch_type: example # example or length
batch_size: 4 # if batch_type is example, batch_size is the numbers of samples; if length, batch_size is source_token_len+target_token_len;
max_token_length: 3000 # filter samples if source_token_len+target_token_len > max_token_length,
shuffle: True
num_workers: 0
audio_adaptor_downsample_rate: ${audio_adaptor_conf.downsample_rate}
audio_encoder_downsample_rate: 2
# prompt: "<|startoftranscription|><|zh|><|transcribe|><|zh|><|notimestamps|><|wo_itn|>"
index_ds: OpenAIIndexDSJsonl
batch_sampler: BatchSampler
batch_type: token
batch_size: 900
max_token_length: 1024
shuffle: true
sort_size: 1024
batch_size_scale_ratio_max: 2
num_workers: 4
audio_adaptor_downsample_rate: ${audio_adaptor_conf.downsample_rate}
audio_encoder_downsample_rate: 2
data_split_num: 512
batch_size_sample_max: 15
retry: 20
tokenizer: HuggingfaceTokenizer