Fine-tuned with CV17.0

README.md

@@ -1,43 +1,40 @@
 ---
-language:
-- eu
+library_name: transformers
 license: apache-2.0
+base_model: openai/whisper-small
 tags:
-- whisper-event
 - generated_from_trainer
 datasets:
-- mozilla-foundation/common_voice_16_0
+- common_voice_17_0
 metrics:
 - wer
 model-index:
-- name: Whisper Small Basque
+- name: openai/whisper-small
   results:
   - task:
       name: Automatic Speech Recognition
       type: automatic-speech-recognition
     dataset:
-      name: mozilla-foundation/common_voice_16_0 eu
-      type: mozilla-foundation/common_voice_16_0
+      name: common_voice_17_0
+      type: common_voice_17_0
       config: eu
       split: test
       args: eu
     metrics:
     - name: Wer
       type: wer
-      value: 12.012786552211754
+      value: 11.838103265051853
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 should probably proofread and complete it, then remove this comment. -->
 
-# Whisper Small Basque
+# openai/whisper-small
 
-This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the mozilla-foundation/common_voice_16_0 eu dataset.
+This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the common_voice_17_0 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.1996
-- Wer: 12.0128
-
-If you need to use this model with whisper.cpp, you can download the ggml file: [ggml-small-eu.bin](https://huggingface.co/xezpeleta/whisper-small-eu/resolve/main/ggml-small.eu.bin)
+- Loss: 0.1845
+- Wer: 11.8381
 
 ## Model description
 
@@ -60,7 +57,7 @@ The following hyperparameters were used during training:
 - train_batch_size: 32
 - eval_batch_size: 16
 - seed: 42
-- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_steps: 500
 - training_steps: 5000
@@ -68,18 +65,18 @@ The following hyperparameters were used during training:
 
 ### Training results
 
-| Training Loss | Epoch | Step | Validation Loss | Wer     |
-|:-------------:|:-----:|:----:|:---------------:|:-------:|
-| 0.2009        | 1.04  | 1000 | 0.2446          | 17.6881 |
-| 0.0759        | 2.09  | 2000 | 0.2102          | 14.2584 |
-| 0.0264        | 3.13  | 3000 | 0.2200          | 13.6898 |
-| 0.0633        | 5.02  | 4000 | 0.1955          | 12.5535 |
-| 0.0199        | 6.06  | 5000 | 0.1996          | 12.0128 |
+| Training Loss | Epoch  | Step | Validation Loss | Wer     |
+|:-------------:|:------:|:----:|:---------------:|:-------:|
+| 0.2458        | 1.0296 | 1000 | 0.2480          | 18.0723 |
+| 0.1099        | 2.0592 | 2000 | 0.1976          | 13.6355 |
+| 0.0403        | 3.0888 | 3000 | 0.2002          | 12.7249 |
+| 0.0211        | 4.1184 | 4000 | 0.2101          | 12.4327 |
+| 0.0609        | 5.148  | 5000 | 0.1845          | 11.8381 |
 
 
 ### Framework versions
 
-- Transformers 4.26.0.dev0
-- Pytorch 1.13.1+cu117
-- Datasets 2.8.1.dev0
-- Tokenizers 0.13.2
+- Transformers 4.46.0.dev0
+- Pytorch 2.4.1+cu121
+- Datasets 3.0.2.dev0
+- Tokenizers 0.20.0

config.json

@@ -2,15 +2,14 @@
   "_name_or_path": "openai/whisper-small",
   "activation_dropout": 0.0,
   "activation_function": "gelu",
+  "apply_spec_augment": false,
   "architectures": [
     "WhisperForConditionalGeneration"
   ],
   "attention_dropout": 0.0,
-  "begin_suppress_tokens": [
-    220,
-    50257
-  ],
+  "begin_suppress_tokens": null,
   "bos_token_id": 50257,
+  "classifier_proj_size": 256,
   "d_model": 768,
   "decoder_attention_heads": 12,
   "decoder_ffn_dim": 3072,
@@ -26,16 +25,24 @@
   "forced_decoder_ids": null,
   "init_std": 0.02,
   "is_encoder_decoder": true,
-  "max_length": 448,
+  "mask_feature_length": 10,
+  "mask_feature_min_masks": 0,
+  "mask_feature_prob": 0.0,
+  "mask_time_length": 10,
+  "mask_time_min_masks": 2,
+  "mask_time_prob": 0.05,
+  "max_length": null,
   "max_source_positions": 1500,
   "max_target_positions": 448,
+  "median_filter_width": 7,
   "model_type": "whisper",
   "num_hidden_layers": 12,
   "num_mel_bins": 80,
   "pad_token_id": 50257,
   "scale_embedding": false,
   "torch_dtype": "float32",
-  "transformers_version": "4.26.0.dev0",
+  "transformers_version": "4.46.0.dev0",
   "use_cache": false,
+  "use_weighted_layer_sum": false,
   "vocab_size": 51865
 }

generation_config.json

@@ -0,0 +1,264 @@
+{
+  "alignment_heads": [
+    [
+      5,
+      3
+    ],
+    [
+      5,
+      9
+    ],
+    [
+      8,
+      0
+    ],
+    [
+      8,
+      4
+    ],
+    [
+      8,
+      7
+    ],
+    [
+      8,
+      8
+    ],
+    [
+      9,
+      0
+    ],
+    [
+      9,
+      7
+    ],
+    [
+      9,
+      9
+    ],
+    [
+      10,
+      5
+    ]
+  ],
+  "begin_suppress_tokens": [
+    220,
+    50257
+  ],
+  "bos_token_id": 50257,
+  "decoder_start_token_id": 50258,
+  "eos_token_id": 50257,
+  "forced_decoder_ids": [
+    [
+      1,
+      null
+    ],
+    [
+      2,
+      50359
+    ]
+  ],
+  "is_multilingual": true,
+  "lang_to_id": {
+    "<|af|>": 50327,
+    "<|am|>": 50334,
+    "<|ar|>": 50272,
+    "<|as|>": 50350,
+    "<|az|>": 50304,
+    "<|ba|>": 50355,
+    "<|be|>": 50330,
+    "<|bg|>": 50292,
+    "<|bn|>": 50302,
+    "<|bo|>": 50347,
+    "<|br|>": 50309,
+    "<|bs|>": 50315,
+    "<|ca|>": 50270,
+    "<|cs|>": 50283,
+    "<|cy|>": 50297,
+    "<|da|>": 50285,
+    "<|de|>": 50261,
+    "<|el|>": 50281,
+    "<|en|>": 50259,
+    "<|es|>": 50262,
+    "<|et|>": 50307,
+    "<|eu|>": 50310,
+    "<|fa|>": 50300,
+    "<|fi|>": 50277,
+    "<|fo|>": 50338,
+    "<|fr|>": 50265,
+    "<|gl|>": 50319,
+    "<|gu|>": 50333,
+    "<|haw|>": 50352,
+    "<|ha|>": 50354,
+    "<|he|>": 50279,
+    "<|hi|>": 50276,
+    "<|hr|>": 50291,
+    "<|ht|>": 50339,
+    "<|hu|>": 50286,
+    "<|hy|>": 50312,
+    "<|id|>": 50275,
+    "<|is|>": 50311,
+    "<|it|>": 50274,
+    "<|ja|>": 50266,
+    "<|jw|>": 50356,
+    "<|ka|>": 50329,
+    "<|kk|>": 50316,
+    "<|km|>": 50323,
+    "<|kn|>": 50306,
+    "<|ko|>": 50264,
+    "<|la|>": 50294,
+    "<|lb|>": 50345,
+    "<|ln|>": 50353,
+    "<|lo|>": 50336,
+    "<|lt|>": 50293,
+    "<|lv|>": 50301,
+    "<|mg|>": 50349,
+    "<|mi|>": 50295,
+    "<|mk|>": 50308,
+    "<|ml|>": 50296,
+    "<|mn|>": 50314,
+    "<|mr|>": 50320,
+    "<|ms|>": 50282,
+    "<|mt|>": 50343,
+    "<|my|>": 50346,
+    "<|ne|>": 50313,
+    "<|nl|>": 50271,
+    "<|nn|>": 50342,
+    "<|no|>": 50288,
+    "<|oc|>": 50328,
+    "<|pa|>": 50321,
+    "<|pl|>": 50269,
+    "<|ps|>": 50340,
+    "<|pt|>": 50267,
+    "<|ro|>": 50284,
+    "<|ru|>": 50263,
+    "<|sa|>": 50344,
+    "<|sd|>": 50332,
+    "<|si|>": 50322,
+    "<|sk|>": 50298,
+    "<|sl|>": 50305,
+    "<|sn|>": 50324,
+    "<|so|>": 50326,
+    "<|sq|>": 50317,
+    "<|sr|>": 50303,
+    "<|su|>": 50357,
+    "<|sv|>": 50273,
+    "<|sw|>": 50318,
+    "<|ta|>": 50287,
+    "<|te|>": 50299,
+    "<|tg|>": 50331,
+    "<|th|>": 50289,
+    "<|tk|>": 50341,
+    "<|tl|>": 50348,
+    "<|tr|>": 50268,
+    "<|tt|>": 50351,
+    "<|uk|>": 50280,
+    "<|ur|>": 50290,
+    "<|uz|>": 50337,
+    "<|vi|>": 50278,
+    "<|yi|>": 50335,
+    "<|yo|>": 50325,
+    "<|zh|>": 50260
+  },
+  "max_initial_timestamp_index": 50,
+  "max_length": 448,
+  "no_timestamps_token_id": 50363,
+  "pad_token_id": 50257,
+  "prev_sot_token_id": 50361,
+  "return_timestamps": false,
+  "suppress_tokens": [
+    1,
+    2,
+    7,
+    8,
+    9,
+    10,
+    14,
+    25,
+    26,
+    27,
+    28,
+    29,
+    31,
+    58,
+    59,
+    60,
+    61,
+    62,
+    63,
+    90,
+    91,
+    92,
+    93,
+    359,
+    503,
+    522,
+    542,
+    873,
+    893,
+    902,
+    918,
+    922,
+    931,
+    1350,
+    1853,
+    1982,
+    2460,
+    2627,
+    3246,
+    3253,
+    3268,
+    3536,
+    3846,
+    3961,
+    4183,
+    4667,
+    6585,
+    6647,
+    7273,
+    9061,
+    9383,
+    10428,
+    10929,
+    11938,
+    12033,
+    12331,
+    12562,
+    13793,
+    14157,
+    14635,
+    15265,
+    15618,
+    16553,
+    16604,
+    18362,
+    18956,
+    20075,
+    21675,
+    22520,
+    26130,
+    26161,
+    26435,
+    28279,
+    29464,
+    31650,
+    32302,
+    32470,
+    36865,
+    42863,
+    47425,
+    49870,
+    50254,
+    50258,
+    50358,
+    50359,
+    50360,
+    50361,
+    50362
+  ],
+  "task_to_id": {
+    "transcribe": 50359,
+    "translate": 50358
+  },
+  "transformers_version": "4.46.0.dev0"
+}

merges.txt

@@ -1,4 +1,5 @@
 #version: 0.2
+Ġ t
 Ġ a
 Ġt h
 i n

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:51160e502a9753fabd3f154ffcebb40049941544eb344c4b8fc8243205072f73
+size 966995080

run.sh

@@ -1,6 +1,6 @@
 python run_speech_recognition_seq2seq_streaming.py \
 	--model_name_or_path="openai/whisper-small" \
-	--dataset_name="mozilla-foundation/common_voice_16_0" \
+	--dataset_name="mozilla-foundation/common_voice_17_0" \
 	--dataset_config_name="eu" \
 	--language="basque" \
 	--train_split_name="train+validation" \

run_speech_recognition_seq2seq_streaming.py

@@ -274,7 +274,7 @@ def load_maybe_streaming_dataset(dataset_name, dataset_config_name, split="train
     if "+" in split:
         # load multiple splits separated by the `+` symbol with streaming mode
         dataset_splits = [
-            load_dataset(dataset_name, dataset_config_name, split=split_name, streaming=streaming, **kwargs)
+            load_dataset(dataset_name, dataset_config_name, split=split_name, streaming=streaming, trust_remote_code=True, **kwargs)
             for split_name in split.split("+")
         ]
         # interleave multiple splits to form one dataset
@@ -282,7 +282,7 @@ def load_maybe_streaming_dataset(dataset_name, dataset_config_name, split="train
         return interleaved_dataset
     else:
         # load a single split *with* streaming mode
-        dataset = load_dataset(dataset_name, dataset_config_name, split=split, streaming=streaming, **kwargs)
+        dataset = load_dataset(dataset_name, dataset_config_name, split=split, streaming=streaming, trust_remote_code=True, **kwargs)
         return dataset
 
 
@@ -357,7 +357,8 @@ def main():
             data_args.dataset_name,
             data_args.dataset_config_name,
             split=data_args.train_split_name,
-            use_auth_token=True if model_args.use_auth_token else None,
+            # xezpeleta
+            #use_auth_token=True if model_args.use_auth_token else None,
             streaming=data_args.streaming,
         )
 
@@ -366,7 +367,7 @@ def main():
             data_args.dataset_name,
             data_args.dataset_config_name,
             split=data_args.eval_split_name,
-            use_auth_token=True if model_args.use_auth_token else None,
+            #use_auth_token=True if model_args.use_auth_token else None,
             streaming=data_args.streaming,
         )
 

special_tokens_map.json

@@ -111,22 +111,28 @@
   "bos_token": {
     "content": "<|endoftext|>",
     "lstrip": false,
-    "normalized": true,
+    "normalized": false,
     "rstrip": false,
     "single_word": false
   },
   "eos_token": {
     "content": "<|endoftext|>",
     "lstrip": false,
-    "normalized": true,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
     "rstrip": false,
     "single_word": false
   },
-  "pad_token": "<|endoftext|>",
   "unk_token": {
     "content": "<|endoftext|>",
     "lstrip": false,
-    "normalized": true,
+    "normalized": false,
     "rstrip": false,
     "single_word": false
   }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:aa82f3baa8211005bb4afa4b314455805f1fd0b5faaeebe6359417001abf7b6c
-size 3643
+oid sha256:1e5015f6997af49dd4702cbba394870a18c74f3b62b5a4ffcc8bf3aa71cc41ee
+size 5368