susmitabhatt commited on
Commit
f93464c
1 Parent(s): cc55c72

Training in progress, epoch 1

Browse files
config.json ADDED
@@ -0,0 +1,48 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "openai/whisper-small",
3
+ "activation_dropout": 0.0,
4
+ "activation_function": "gelu",
5
+ "apply_spec_augment": false,
6
+ "architectures": [
7
+ "WhisperForConditionalGeneration"
8
+ ],
9
+ "attention_dropout": 0.0,
10
+ "begin_suppress_tokens": null,
11
+ "bos_token_id": 50257,
12
+ "classifier_proj_size": 256,
13
+ "d_model": 768,
14
+ "decoder_attention_heads": 12,
15
+ "decoder_ffn_dim": 3072,
16
+ "decoder_layerdrop": 0.0,
17
+ "decoder_layers": 12,
18
+ "decoder_start_token_id": 50258,
19
+ "dropout": 0.0,
20
+ "encoder_attention_heads": 12,
21
+ "encoder_ffn_dim": 3072,
22
+ "encoder_layerdrop": 0.0,
23
+ "encoder_layers": 12,
24
+ "eos_token_id": 50257,
25
+ "forced_decoder_ids": null,
26
+ "init_std": 0.02,
27
+ "is_encoder_decoder": true,
28
+ "mask_feature_length": 10,
29
+ "mask_feature_min_masks": 0,
30
+ "mask_feature_prob": 0.0,
31
+ "mask_time_length": 10,
32
+ "mask_time_min_masks": 2,
33
+ "mask_time_prob": 0.05,
34
+ "max_length": null,
35
+ "max_source_positions": 1500,
36
+ "max_target_positions": 448,
37
+ "median_filter_width": 7,
38
+ "model_type": "whisper",
39
+ "num_hidden_layers": 12,
40
+ "num_mel_bins": 80,
41
+ "pad_token_id": 50257,
42
+ "scale_embedding": false,
43
+ "torch_dtype": "float32",
44
+ "transformers_version": "4.47.0.dev0",
45
+ "use_cache": true,
46
+ "use_weighted_layer_sum": false,
47
+ "vocab_size": 51865
48
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:36caf590447e47d4e02bf968bf64564d0bcfa1f878dbfd855d1903f4bbcf948c
3
+ size 966995080
preprocessor_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "chunk_length": 30,
3
+ "feature_extractor_type": "WhisperFeatureExtractor",
4
+ "feature_size": 80,
5
+ "hop_length": 160,
6
+ "n_fft": 400,
7
+ "n_samples": 480000,
8
+ "nb_max_frames": 3000,
9
+ "padding_side": "right",
10
+ "padding_value": 0.0,
11
+ "processor_class": "WhisperProcessor",
12
+ "return_attention_mask": false,
13
+ "sampling_rate": 16000
14
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5862b9b10de800cb02ca56a5a378aab17a03b455288a33261b58d3270396780d
3
+ size 5496