Training in progress, step 5

.ipynb_checkpoints/README-checkpoint.md

@@ -1,68 +0,0 @@
----
-language:
-- sv-SE
-license: apache-2.0
-tags:
-- automatic-speech-recognition
-- mozilla-foundation/common_voice_7_0
-- generated_from_trainer
-datasets:
-- common_voice
-model-index:
-- name: ''
-  results: []
----
-
-<!-- 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. -->
-
-# 
-
-This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the MOZILLA-FOUNDATION/COMMON_VOICE_7_0 - SV-SE dataset.
-It achieves the following results on the evaluation set:
-- Loss: 14.2136
-- Wer: 1.0
-
-## Model description
-
-More information needed
-
-## Intended uses & limitations
-
-More information needed
-
-## Training and evaluation data
-
-More information needed
-
-## Training procedure
-
-### Training hyperparameters
-
-The following hyperparameters were used during training:
-- learning_rate: 0.0075
-- train_batch_size: 8
-- eval_batch_size: 8
-- seed: 42
-- gradient_accumulation_steps: 4
-- total_train_batch_size: 32
-- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
-- lr_scheduler_type: linear
-- lr_scheduler_warmup_steps: 2000
-- training_steps: 10
-- mixed_precision_training: Native AMP
-
-### Training results
-
-| Training Loss | Epoch | Step | Validation Loss | Wer |
-|:-------------:|:-----:|:----:|:---------------:|:---:|
-| 10.3458       | 0.01  | 5    | 15.0991         | 1.0 |
-| 11.9029       | 0.03  | 10   | 14.2136         | 1.0 |
-
-
-### Framework versions
-
-- Transformers 4.16.0.dev0
-- Pytorch 1.10.1+cu102
-- Datasets 1.17.1.dev0
-- Tokenizers 0.11.0

.ipynb_checkpoints/added_tokens-checkpoint.json

@@ -1 +0,0 @@
-{"<s>": 35, "</s>": 36}

.ipynb_checkpoints/all_results-checkpoint.json

@@ -1,8 +0,0 @@
-{
-    "epoch": 0.03,
-    "train_loss": 12.162414741516113,
-    "train_runtime": 400.3916,
-    "train_samples": 11030,
-    "train_samples_per_second": 0.799,
-    "train_steps_per_second": 0.025
-}

.ipynb_checkpoints/config-checkpoint.json

@@ -1,107 +0,0 @@
-{
-  "_name_or_path": "facebook/wav2vec2-xls-r-300m",
-  "activation_dropout": 0.1,
-  "adapter_kernel_size": 3,
-  "adapter_stride": 2,
-  "add_adapter": false,
-  "apply_spec_augment": true,
-  "architectures": [
-    "Wav2Vec2ForCTC"
-  ],
-  "attention_dropout": 0.0,
-  "bos_token_id": 1,
-  "classifier_proj_size": 256,
-  "codevector_dim": 768,
-  "contrastive_logits_temperature": 0.1,
-  "conv_bias": true,
-  "conv_dim": [
-    512,
-    512,
-    512,
-    512,
-    512,
-    512,
-    512
-  ],
-  "conv_kernel": [
-    10,
-    3,
-    3,
-    3,
-    3,
-    2,
-    2
-  ],
-  "conv_stride": [
-    5,
-    2,
-    2,
-    2,
-    2,
-    2,
-    2
-  ],
-  "ctc_loss_reduction": "mean",
-  "ctc_zero_infinity": false,
-  "diversity_loss_weight": 0.1,
-  "do_stable_layer_norm": true,
-  "eos_token_id": 2,
-  "feat_extract_activation": "gelu",
-  "feat_extract_dropout": 0.0,
-  "feat_extract_norm": "layer",
-  "feat_proj_dropout": 0.0,
-  "feat_quantizer_dropout": 0.0,
-  "final_dropout": 0.0,
-  "hidden_act": "gelu",
-  "hidden_dropout": 0.0,
-  "hidden_size": 1024,
-  "initializer_range": 0.02,
-  "intermediate_size": 4096,
-  "layer_norm_eps": 1e-05,
-  "layerdrop": 0.0,
-  "mask_feature_length": 64,
-  "mask_feature_min_masks": 0,
-  "mask_feature_prob": 0.25,
-  "mask_time_length": 10,
-  "mask_time_min_masks": 2,
-  "mask_time_prob": 0.75,
-  "model_type": "wav2vec2",
-  "num_adapter_layers": 3,
-  "num_attention_heads": 16,
-  "num_codevector_groups": 2,
-  "num_codevectors_per_group": 320,
-  "num_conv_pos_embedding_groups": 16,
-  "num_conv_pos_embeddings": 128,
-  "num_feat_extract_layers": 7,
-  "num_hidden_layers": 24,
-  "num_negatives": 100,
-  "output_hidden_size": 1024,
-  "pad_token_id": 34,
-  "proj_codevector_dim": 768,
-  "tdnn_dilation": [
-    1,
-    2,
-    3,
-    1,
-    1
-  ],
-  "tdnn_dim": [
-    512,
-    512,
-    512,
-    512,
-    1500
-  ],
-  "tdnn_kernel": [
-    5,
-    3,
-    3,
-    1,
-    1
-  ],
-  "torch_dtype": "float32",
-  "transformers_version": "4.16.0.dev0",
-  "use_weighted_layer_sum": false,
-  "vocab_size": 37,
-  "xvector_output_dim": 512
-}

.ipynb_checkpoints/eval_results-checkpoint.json

@@ -1,9 +0,0 @@
-{
-    "epoch": 1.0,
-    "eval_loss": 3.82973051071167,
-    "eval_runtime": 132.808,
-    "eval_samples": 4620,
-    "eval_samples_per_second": 34.787,
-    "eval_steps_per_second": 4.352,
-    "eval_wer": 1.0
-}

.ipynb_checkpoints/preprocessor_config-checkpoint.json

@@ -1,9 +0,0 @@
-{
-  "do_normalize": true,
-  "feature_extractor_type": "Wav2Vec2FeatureExtractor",
-  "feature_size": 1,
-  "padding_side": "right",
-  "padding_value": 0,
-  "return_attention_mask": true,
-  "sampling_rate": 16000
-}

.ipynb_checkpoints/run-dummy-ab-gpu-checkpoint.sh

@@ -0,0 +1,22 @@
+python run_speech_recognition_ctc.py \
+--dataset_name="mozilla-foundation/common_voice_7_0" \
+--model_name_or_path="hf-test/xls-r-dummy" \
+--dataset_config_name="ab" \
+--output_dir="./" \
+--overwrite_output_dir \
+--max_steps="10" \
+--per_device_train_batch_size="2" \
+--learning_rate="3e-4" \
+--save_total_limit="1" \
+--evaluation_strategy="steps" \
+--text_column_name="sentence" \
+--length_column_name="input_length" \
+--save_steps="5" \
+--layerdrop="0.0" \
+--freeze_feature_encoder \
+--gradient_checkpointing \
+--fp16 \
+--group_by_length \
+--push_to_hub \
+--use_auth_token \
+--do_train --do_eval

.ipynb_checkpoints/run-dummy-sv-gpu-checkpoint.sh

@@ -1,7 +1,7 @@
 python run_speech_recognition_ctc.py \
 	--dataset_name="mozilla-foundation/common_voice_7_0" \
-	--model_name_or_path="facebook/wav2vec2-xls-r-300m" \
-	--dataset_config_name="sv-SE" \
+	--model_name_or_path="hf-test/xls-r-dummy" \
+	--dataset_config_name="ab" \
 	--output_dir="./" \
 	--overwrite_output_dir \
 	--max_steps="10" \

.ipynb_checkpoints/special_tokens_map-checkpoint.json

@@ -1 +0,0 @@
-{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "[UNK]", "pad_token": "[PAD]", "additional_special_tokens": [{"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}]}

.ipynb_checkpoints/tokenizer_config-checkpoint.json

@@ -1 +0,0 @@
-{"unk_token": "[UNK]", "bos_token": "<s>", "eos_token": "</s>", "pad_token": "[PAD]", "do_lower_case": false, "word_delimiter_token": "|", "special_tokens_map_file": null, "tokenizer_file": null, "name_or_path": "./", "tokenizer_class": "Wav2Vec2CTCTokenizer"}

.ipynb_checkpoints/train_results-checkpoint.json

@@ -1,8 +0,0 @@
-{
-    "epoch": 0.03,
-    "train_loss": 12.162414741516113,
-    "train_runtime": 400.3916,
-    "train_samples": 11030,
-    "train_samples_per_second": 0.799,
-    "train_steps_per_second": 0.025
-}

.ipynb_checkpoints/trainer_state-checkpoint.json

@@ -1,103 +0,0 @@
-{
-  "best_metric": null,
-  "best_model_checkpoint": null,
-  "epoch": 0.029006526468455404,
-  "global_step": 10,
-  "is_hyper_param_search": false,
-  "is_local_process_zero": true,
-  "is_world_process_zero": true,
-  "log_history": [
-    {
-      "epoch": 0.0,
-      "learning_rate": 3.75e-06,
-      "loss": 13.605,
-      "step": 1
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 7.5e-06,
-      "loss": 11.0063,
-      "step": 2
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.1249999999999999e-05,
-      "loss": 11.6693,
-      "step": 3
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.5e-05,
-      "loss": 13.432,
-      "step": 4
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.875e-05,
-      "loss": 10.3458,
-      "step": 5
-    },
-    {
-      "epoch": 0.01,
-      "eval_loss": 15.09913444519043,
-      "eval_runtime": 139.8551,
-      "eval_samples_per_second": 33.034,
-      "eval_steps_per_second": 4.133,
-      "eval_wer": 1.0,
-      "step": 5
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 2.2499999999999998e-05,
-      "loss": 15.2451,
-      "step": 6
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 2.625e-05,
-      "loss": 10.0481,
-      "step": 7
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 3e-05,
-      "loss": 12.3838,
-      "step": 8
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 3.3749999999999994e-05,
-      "loss": 11.9858,
-      "step": 9
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 3.75e-05,
-      "loss": 11.9029,
-      "step": 10
-    },
-    {
-      "epoch": 0.03,
-      "eval_loss": 14.213573455810547,
-      "eval_runtime": 160.2552,
-      "eval_samples_per_second": 28.829,
-      "eval_steps_per_second": 3.607,
-      "eval_wer": 1.0,
-      "step": 10
-    },
-    {
-      "epoch": 0.03,
-      "step": 10,
-      "total_flos": 4.405923604988928e+16,
-      "train_loss": 12.162414741516113,
-      "train_runtime": 400.3916,
-      "train_samples_per_second": 0.799,
-      "train_steps_per_second": 0.025
-    }
-  ],
-  "max_steps": 10,
-  "num_train_epochs": 1,
-  "total_flos": 4.405923604988928e+16,
-  "trial_name": null,
-  "trial_params": null
-}

.ipynb_checkpoints/vocab-checkpoint.json

@@ -1 +0,0 @@
-{"a": 1, "b": 2, "c": 3, "d": 4, "e": 5, "f": 6, "g": 7, "h": 8, "i": 9, "j": 10, "k": 11, "l": 12, "m": 13, "n": 14, "o": 15, "p": 16, "q": 17, "r": 18, "s": 19, "t": 20, "u": 21, "v": 22, "w": 23, "x": 24, "y": 25, "z": 26, "ä": 27, "å": 28, "é": 29, "ô": 30, "ö": 31, "ü": 32, "|": 0, "[UNK]": 33, "[PAD]": 34}

README.md

@@ -1,67 +0,0 @@
----
-language:
-- sv-SE
-license: apache-2.0
-tags:
-- automatic-speech-recognition
-- mozilla-foundation/common_voice_7_0
-- generated_from_trainer
-datasets:
-- common_voice
-model-index:
-- name: ''
-  results: []
----
-
-<!-- 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. -->
-
-# 
-
-This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the MOZILLA-FOUNDATION/COMMON_VOICE_7_0 - SV-SE dataset.
-It achieves the following results on the evaluation set:
-- Loss: 2.9141
-- Wer: 1.0
-
-## Model description
-
-More information needed
-
-## Intended uses & limitations
-
-More information needed
-
-## Training and evaluation data
-
-More information needed
-
-## Training procedure
-
-### Training hyperparameters
-
-The following hyperparameters were used during training:
-- learning_rate: 7.5e-05
-- train_batch_size: 8
-- eval_batch_size: 8
-- seed: 42
-- gradient_accumulation_steps: 4
-- total_train_batch_size: 32
-- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
-- lr_scheduler_type: linear
-- lr_scheduler_warmup_steps: 20
-- num_epochs: 2.0
-- mixed_precision_training: Native AMP
-
-### Training results
-
-| Training Loss | Epoch | Step | Validation Loss | Wer |
-|:-------------:|:-----:|:----:|:---------------:|:---:|
-| 2.9357        | 1.45  | 500  | 2.9459          | 1.0 |
-
-
-### Framework versions
-
-- Transformers 4.16.0.dev0
-- Pytorch 1.10.1+cu102
-- Datasets 1.17.1.dev0
-- Tokenizers 0.11.0

added_tokens.json

@@ -1 +1 @@
-{"<s>": 35, "</s>": 36}
+{"<s>": 51, "</s>": 52}

all_results.json

@@ -1,14 +0,0 @@
-{
-    "epoch": 2.0,
-    "eval_loss": 2.91414213180542,
-    "eval_runtime": 133.9783,
-    "eval_samples": 4620,
-    "eval_samples_per_second": 34.483,
-    "eval_steps_per_second": 4.314,
-    "eval_wer": 1.0,
-    "train_loss": 3.289040254992108,
-    "train_runtime": 1292.4856,
-    "train_samples": 11030,
-    "train_samples_per_second": 17.068,
-    "train_steps_per_second": 0.532
-}

checkpoint-400/config.json

@@ -1,107 +0,0 @@
-{
-  "_name_or_path": "facebook/wav2vec2-xls-r-300m",
-  "activation_dropout": 0.1,
-  "adapter_kernel_size": 3,
-  "adapter_stride": 2,
-  "add_adapter": false,
-  "apply_spec_augment": true,
-  "architectures": [
-    "Wav2Vec2ForCTC"
-  ],
-  "attention_dropout": 0.0,
-  "bos_token_id": 1,
-  "classifier_proj_size": 256,
-  "codevector_dim": 768,
-  "contrastive_logits_temperature": 0.1,
-  "conv_bias": true,
-  "conv_dim": [
-    512,
-    512,
-    512,
-    512,
-    512,
-    512,
-    512
-  ],
-  "conv_kernel": [
-    10,
-    3,
-    3,
-    3,
-    3,
-    2,
-    2
-  ],
-  "conv_stride": [
-    5,
-    2,
-    2,
-    2,
-    2,
-    2,
-    2
-  ],
-  "ctc_loss_reduction": "mean",
-  "ctc_zero_infinity": false,
-  "diversity_loss_weight": 0.1,
-  "do_stable_layer_norm": true,
-  "eos_token_id": 2,
-  "feat_extract_activation": "gelu",
-  "feat_extract_dropout": 0.0,
-  "feat_extract_norm": "layer",
-  "feat_proj_dropout": 0.0,
-  "feat_quantizer_dropout": 0.0,
-  "final_dropout": 0.0,
-  "hidden_act": "gelu",
-  "hidden_dropout": 0.0,
-  "hidden_size": 1024,
-  "initializer_range": 0.02,
-  "intermediate_size": 4096,
-  "layer_norm_eps": 1e-05,
-  "layerdrop": 0.0,
-  "mask_feature_length": 64,
-  "mask_feature_min_masks": 0,
-  "mask_feature_prob": 0.25,
-  "mask_time_length": 10,
-  "mask_time_min_masks": 2,
-  "mask_time_prob": 0.75,
-  "model_type": "wav2vec2",
-  "num_adapter_layers": 3,
-  "num_attention_heads": 16,
-  "num_codevector_groups": 2,
-  "num_codevectors_per_group": 320,
-  "num_conv_pos_embedding_groups": 16,
-  "num_conv_pos_embeddings": 128,
-  "num_feat_extract_layers": 7,
-  "num_hidden_layers": 24,
-  "num_negatives": 100,
-  "output_hidden_size": 1024,
-  "pad_token_id": 34,
-  "proj_codevector_dim": 768,
-  "tdnn_dilation": [
-    1,
-    2,
-    3,
-    1,
-    1
-  ],
-  "tdnn_dim": [
-    512,
-    512,
-    512,
-    512,
-    1500
-  ],
-  "tdnn_kernel": [
-    5,
-    3,
-    3,
-    1,
-    1
-  ],
-  "torch_dtype": "float32",
-  "transformers_version": "4.16.0.dev0",
-  "use_weighted_layer_sum": false,
-  "vocab_size": 37,
-  "xvector_output_dim": 512
-}

checkpoint-400/optimizer.pt

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:16fdeeb953ce80d82d1ba249d99ae68b45e80c943e36a80b8a6517c275a7b594
-size 2490362385

checkpoint-400/pytorch_model.bin

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:65431ecd0a10e000abb091aa8e3064f6e344c3e7771a07e792b6c0470ee8092c
-size 1262075377

checkpoint-400/trainer_state.json

@@ -1,328 +0,0 @@
-{
-  "best_metric": null,
-  "best_model_checkpoint": null,
-  "epoch": 1.1624365482233503,
-  "global_step": 400,
-  "is_hyper_param_search": false,
-  "is_local_process_zero": true,
-  "is_world_process_zero": true,
-  "log_history": [
-    {
-      "epoch": 0.03,
-      "learning_rate": 3.75e-05,
-      "loss": 12.1562,
-      "step": 10
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 7.125e-05,
-      "loss": 8.7679,
-      "step": 20
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 7.398952095808383e-05,
-      "loss": 5.3683,
-      "step": 30
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 7.286676646706586e-05,
-      "loss": 4.3219,
-      "step": 40
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 7.17440119760479e-05,
-      "loss": 3.7182,
-      "step": 50
-    },
-    {
-      "epoch": 0.15,
-      "eval_loss": 3.836604595184326,
-      "eval_runtime": 133.4846,
-      "eval_samples_per_second": 34.611,
-      "eval_steps_per_second": 4.33,
-      "eval_wer": 1.0,
-      "step": 50
-    },
-    {
-      "epoch": 0.17,
-      "learning_rate": 7.062125748502993e-05,
-      "loss": 3.478,
-      "step": 60
-    },
-    {
-      "epoch": 0.2,
-      "learning_rate": 6.949850299401197e-05,
-      "loss": 3.4492,
-      "step": 70
-    },
-    {
-      "epoch": 0.23,
-      "learning_rate": 6.837574850299401e-05,
-      "loss": 3.3928,
-      "step": 80
-    },
-    {
-      "epoch": 0.26,
-      "learning_rate": 6.725299401197604e-05,
-      "loss": 3.3183,
-      "step": 90
-    },
-    {
-      "epoch": 0.29,
-      "learning_rate": 6.613023952095809e-05,
-      "loss": 3.2075,
-      "step": 100
-    },
-    {
-      "epoch": 0.29,
-      "eval_loss": 3.258362293243408,
-      "eval_runtime": 126.6078,
-      "eval_samples_per_second": 36.491,
-      "eval_steps_per_second": 4.565,
-      "eval_wer": 1.0,
-      "step": 100
-    },
-    {
-      "epoch": 0.32,
-      "learning_rate": 6.500748502994012e-05,
-      "loss": 3.14,
-      "step": 110
-    },
-    {
-      "epoch": 0.35,
-      "learning_rate": 6.388473053892215e-05,
-      "loss": 3.1281,
-      "step": 120
-    },
-    {
-      "epoch": 0.38,
-      "learning_rate": 6.276197604790418e-05,
-      "loss": 3.0987,
-      "step": 130
-    },
-    {
-      "epoch": 0.41,
-      "learning_rate": 6.163922155688622e-05,
-      "loss": 3.1003,
-      "step": 140
-    },
-    {
-      "epoch": 0.44,
-      "learning_rate": 6.0516467065868256e-05,
-      "loss": 3.0922,
-      "step": 150
-    },
-    {
-      "epoch": 0.44,
-      "eval_loss": 3.127869129180908,
-      "eval_runtime": 126.3837,
-      "eval_samples_per_second": 36.555,
-      "eval_steps_per_second": 4.573,
-      "eval_wer": 1.0,
-      "step": 150
-    },
-    {
-      "epoch": 0.46,
-      "learning_rate": 5.9393712574850293e-05,
-      "loss": 3.0588,
-      "step": 160
-    },
-    {
-      "epoch": 0.49,
-      "learning_rate": 5.827095808383233e-05,
-      "loss": 3.0477,
-      "step": 170
-    },
-    {
-      "epoch": 0.52,
-      "learning_rate": 5.714820359281436e-05,
-      "loss": 3.045,
-      "step": 180
-    },
-    {
-      "epoch": 0.55,
-      "learning_rate": 5.602544910179641e-05,
-      "loss": 3.0439,
-      "step": 190
-    },
-    {
-      "epoch": 0.58,
-      "learning_rate": 5.490269461077844e-05,
-      "loss": 3.0846,
-      "step": 200
-    },
-    {
-      "epoch": 0.58,
-      "eval_loss": 3.079519271850586,
-      "eval_runtime": 125.7215,
-      "eval_samples_per_second": 36.748,
-      "eval_steps_per_second": 4.597,
-      "eval_wer": 1.0,
-      "step": 200
-    },
-    {
-      "epoch": 0.61,
-      "learning_rate": 5.3779940119760477e-05,
-      "loss": 3.0512,
-      "step": 210
-    },
-    {
-      "epoch": 0.64,
-      "learning_rate": 5.265718562874251e-05,
-      "loss": 3.0143,
-      "step": 220
-    },
-    {
-      "epoch": 0.67,
-      "learning_rate": 5.1534431137724546e-05,
-      "loss": 3.0387,
-      "step": 230
-    },
-    {
-      "epoch": 0.7,
-      "learning_rate": 5.0411676646706584e-05,
-      "loss": 3.0311,
-      "step": 240
-    },
-    {
-      "epoch": 0.73,
-      "learning_rate": 4.9288922155688615e-05,
-      "loss": 3.0417,
-      "step": 250
-    },
-    {
-      "epoch": 0.73,
-      "eval_loss": 3.069390058517456,
-      "eval_runtime": 125.7339,
-      "eval_samples_per_second": 36.744,
-      "eval_steps_per_second": 4.597,
-      "eval_wer": 1.0,
-      "step": 250
-    },
-    {
-      "epoch": 0.75,
-      "learning_rate": 4.816616766467066e-05,
-      "loss": 3.0219,
-      "step": 260
-    },
-    {
-      "epoch": 0.78,
-      "learning_rate": 4.704341317365269e-05,
-      "loss": 3.0194,
-      "step": 270
-    },
-    {
-      "epoch": 0.81,
-      "learning_rate": 4.592065868263473e-05,
-      "loss": 2.9974,
-      "step": 280
-    },
-    {
-      "epoch": 0.84,
-      "learning_rate": 4.479790419161676e-05,
-      "loss": 2.9996,
-      "step": 290
-    },
-    {
-      "epoch": 0.87,
-      "learning_rate": 4.36751497005988e-05,
-      "loss": 3.0016,
-      "step": 300
-    },
-    {
-      "epoch": 0.87,
-      "eval_loss": 3.0347490310668945,
-      "eval_runtime": 132.3039,
-      "eval_samples_per_second": 34.92,
-      "eval_steps_per_second": 4.369,
-      "eval_wer": 1.0,
-      "step": 300
-    },
-    {
-      "epoch": 0.9,
-      "learning_rate": 4.255239520958083e-05,
-      "loss": 3.0052,
-      "step": 310
-    },
-    {
-      "epoch": 0.93,
-      "learning_rate": 4.142964071856287e-05,
-      "loss": 2.9826,
-      "step": 320
-    },
-    {
-      "epoch": 0.96,
-      "learning_rate": 4.030688622754491e-05,
-      "loss": 2.9747,
-      "step": 330
-    },
-    {
-      "epoch": 0.99,
-      "learning_rate": 3.918413173652694e-05,
-      "loss": 2.9617,
-      "step": 340
-    },
-    {
-      "epoch": 1.02,
-      "learning_rate": 3.806137724550898e-05,
-      "loss": 3.2053,
-      "step": 350
-    },
-    {
-      "epoch": 1.02,
-      "eval_loss": 2.984886407852173,
-      "eval_runtime": 149.1508,
-      "eval_samples_per_second": 30.975,
-      "eval_steps_per_second": 3.875,
-      "eval_wer": 1.0,
-      "step": 350
-    },
-    {
-      "epoch": 1.05,
-      "learning_rate": 3.693862275449102e-05,
-      "loss": 2.9665,
-      "step": 360
-    },
-    {
-      "epoch": 1.08,
-      "learning_rate": 3.581586826347305e-05,
-      "loss": 2.9641,
-      "step": 370
-    },
-    {
-      "epoch": 1.1,
-      "learning_rate": 3.469311377245509e-05,
-      "loss": 2.9484,
-      "step": 380
-    },
-    {
-      "epoch": 1.13,
-      "learning_rate": 3.3570359281437126e-05,
-      "loss": 2.9494,
-      "step": 390
-    },
-    {
-      "epoch": 1.16,
-      "learning_rate": 3.244760479041916e-05,
-      "loss": 2.9698,
-      "step": 400
-    },
-    {
-      "epoch": 1.16,
-      "eval_loss": 2.989494562149048,
-      "eval_runtime": 150.6903,
-      "eval_samples_per_second": 30.659,
-      "eval_steps_per_second": 3.836,
-      "eval_wer": 1.0,
-      "step": 400
-    }
-  ],
-  "max_steps": 688,
-  "num_train_epochs": 2,
-  "total_flos": 1.1900340345446784e+18,
-  "trial_name": null,
-  "trial_params": null
-}

checkpoint-450/config.json

@@ -1,107 +0,0 @@
-{
-  "_name_or_path": "facebook/wav2vec2-xls-r-300m",
-  "activation_dropout": 0.1,
-  "adapter_kernel_size": 3,
-  "adapter_stride": 2,
-  "add_adapter": false,
-  "apply_spec_augment": true,
-  "architectures": [
-    "Wav2Vec2ForCTC"
-  ],
-  "attention_dropout": 0.0,
-  "bos_token_id": 1,
-  "classifier_proj_size": 256,
-  "codevector_dim": 768,
-  "contrastive_logits_temperature": 0.1,
-  "conv_bias": true,
-  "conv_dim": [
-    512,
-    512,
-    512,
-    512,
-    512,
-    512,
-    512
-  ],
-  "conv_kernel": [
-    10,
-    3,
-    3,
-    3,
-    3,
-    2,
-    2
-  ],
-  "conv_stride": [
-    5,
-    2,
-    2,
-    2,
-    2,
-    2,
-    2
-  ],
-  "ctc_loss_reduction": "mean",
-  "ctc_zero_infinity": false,
-  "diversity_loss_weight": 0.1,
-  "do_stable_layer_norm": true,
-  "eos_token_id": 2,
-  "feat_extract_activation": "gelu",
-  "feat_extract_dropout": 0.0,
-  "feat_extract_norm": "layer",
-  "feat_proj_dropout": 0.0,
-  "feat_quantizer_dropout": 0.0,
-  "final_dropout": 0.0,
-  "hidden_act": "gelu",
-  "hidden_dropout": 0.0,
-  "hidden_size": 1024,
-  "initializer_range": 0.02,
-  "intermediate_size": 4096,
-  "layer_norm_eps": 1e-05,
-  "layerdrop": 0.0,
-  "mask_feature_length": 64,
-  "mask_feature_min_masks": 0,
-  "mask_feature_prob": 0.25,
-  "mask_time_length": 10,
-  "mask_time_min_masks": 2,
-  "mask_time_prob": 0.75,
-  "model_type": "wav2vec2",
-  "num_adapter_layers": 3,
-  "num_attention_heads": 16,
-  "num_codevector_groups": 2,
-  "num_codevectors_per_group": 320,
-  "num_conv_pos_embedding_groups": 16,
-  "num_conv_pos_embeddings": 128,
-  "num_feat_extract_layers": 7,
-  "num_hidden_layers": 24,
-  "num_negatives": 100,
-  "output_hidden_size": 1024,
-  "pad_token_id": 34,
-  "proj_codevector_dim": 768,
-  "tdnn_dilation": [
-    1,
-    2,
-    3,
-    1,
-    1
-  ],
-  "tdnn_dim": [
-    512,
-    512,
-    512,
-    512,
-    1500
-  ],
-  "tdnn_kernel": [
-    5,
-    3,
-    3,
-    1,
-    1
-  ],
-  "torch_dtype": "float32",
-  "transformers_version": "4.16.0.dev0",
-  "use_weighted_layer_sum": false,
-  "vocab_size": 37,
-  "xvector_output_dim": 512
-}

checkpoint-450/optimizer.pt

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:8d20a2e7a06e4c6fe1ca1763518fe26e3bd509e54e5c6ce336b1cdd4ad352fc8
-size 2490362385

checkpoint-450/preprocessor_config.json

@@ -1,9 +0,0 @@
-{
-  "do_normalize": true,
-  "feature_extractor_type": "Wav2Vec2FeatureExtractor",
-  "feature_size": 1,
-  "padding_side": "right",
-  "padding_value": 0,
-  "return_attention_mask": true,
-  "sampling_rate": 16000
-}

checkpoint-450/pytorch_model.bin

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:64a67aef87c3664c6c90fef5a16264df1a1e9ec09017448f7dd3962f7e7ed3cd
-size 1262075377

checkpoint-450/rng_state.pth

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:171f5d4a1db63b5ecb87951d275cbcef882e72c7875f0439f386ec2a0cc474fc
-size 14503

checkpoint-450/scaler.pt

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:5b437a2168639a7226a6616cf299b21e2671732a153df5f941f17a2c38a20459
-size 559

checkpoint-450/scheduler.pt

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:89db00b13c6183184d8d5a3b77b2b0c26843c6c480825e0951a03deff712e541
-size 623

checkpoint-450/trainer_state.json

@@ -1,367 +0,0 @@
-{
-  "best_metric": null,
-  "best_model_checkpoint": null,
-  "epoch": 1.3074691805656273,
-  "global_step": 450,
-  "is_hyper_param_search": false,
-  "is_local_process_zero": true,
-  "is_world_process_zero": true,
-  "log_history": [
-    {
-      "epoch": 0.03,
-      "learning_rate": 3.75e-05,
-      "loss": 12.1562,
-      "step": 10
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 7.125e-05,
-      "loss": 8.7679,
-      "step": 20
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 7.398952095808383e-05,
-      "loss": 5.3683,
-      "step": 30
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 7.286676646706586e-05,
-      "loss": 4.3219,
-      "step": 40
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 7.17440119760479e-05,
-      "loss": 3.7182,
-      "step": 50
-    },
-    {
-      "epoch": 0.15,
-      "eval_loss": 3.836604595184326,
-      "eval_runtime": 133.4846,
-      "eval_samples_per_second": 34.611,
-      "eval_steps_per_second": 4.33,
-      "eval_wer": 1.0,
-      "step": 50
-    },
-    {
-      "epoch": 0.17,
-      "learning_rate": 7.062125748502993e-05,
-      "loss": 3.478,
-      "step": 60
-    },
-    {
-      "epoch": 0.2,
-      "learning_rate": 6.949850299401197e-05,
-      "loss": 3.4492,
-      "step": 70
-    },
-    {
-      "epoch": 0.23,
-      "learning_rate": 6.837574850299401e-05,
-      "loss": 3.3928,
-      "step": 80
-    },
-    {
-      "epoch": 0.26,
-      "learning_rate": 6.725299401197604e-05,
-      "loss": 3.3183,
-      "step": 90
-    },
-    {
-      "epoch": 0.29,
-      "learning_rate": 6.613023952095809e-05,
-      "loss": 3.2075,
-      "step": 100
-    },
-    {
-      "epoch": 0.29,
-      "eval_loss": 3.258362293243408,
-      "eval_runtime": 126.6078,
-      "eval_samples_per_second": 36.491,
-      "eval_steps_per_second": 4.565,
-      "eval_wer": 1.0,
-      "step": 100
-    },
-    {
-      "epoch": 0.32,
-      "learning_rate": 6.500748502994012e-05,
-      "loss": 3.14,
-      "step": 110
-    },
-    {
-      "epoch": 0.35,
-      "learning_rate": 6.388473053892215e-05,
-      "loss": 3.1281,
-      "step": 120
-    },
-    {
-      "epoch": 0.38,
-      "learning_rate": 6.276197604790418e-05,
-      "loss": 3.0987,
-      "step": 130
-    },
-    {
-      "epoch": 0.41,
-      "learning_rate": 6.163922155688622e-05,
-      "loss": 3.1003,
-      "step": 140
-    },
-    {
-      "epoch": 0.44,
-      "learning_rate": 6.0516467065868256e-05,
-      "loss": 3.0922,
-      "step": 150
-    },
-    {
-      "epoch": 0.44,
-      "eval_loss": 3.127869129180908,
-      "eval_runtime": 126.3837,
-      "eval_samples_per_second": 36.555,
-      "eval_steps_per_second": 4.573,
-      "eval_wer": 1.0,
-      "step": 150
-    },
-    {
-      "epoch": 0.46,
-      "learning_rate": 5.9393712574850293e-05,
-      "loss": 3.0588,
-      "step": 160
-    },
-    {
-      "epoch": 0.49,
-      "learning_rate": 5.827095808383233e-05,
-      "loss": 3.0477,
-      "step": 170
-    },
-    {
-      "epoch": 0.52,
-      "learning_rate": 5.714820359281436e-05,
-      "loss": 3.045,
-      "step": 180
-    },
-    {
-      "epoch": 0.55,
-      "learning_rate": 5.602544910179641e-05,
-      "loss": 3.0439,
-      "step": 190
-    },
-    {
-      "epoch": 0.58,
-      "learning_rate": 5.490269461077844e-05,
-      "loss": 3.0846,
-      "step": 200
-    },
-    {
-      "epoch": 0.58,
-      "eval_loss": 3.079519271850586,
-      "eval_runtime": 125.7215,
-      "eval_samples_per_second": 36.748,
-      "eval_steps_per_second": 4.597,
-      "eval_wer": 1.0,
-      "step": 200
-    },
-    {
-      "epoch": 0.61,
-      "learning_rate": 5.3779940119760477e-05,
-      "loss": 3.0512,
-      "step": 210
-    },
-    {
-      "epoch": 0.64,
-      "learning_rate": 5.265718562874251e-05,
-      "loss": 3.0143,
-      "step": 220
-    },
-    {
-      "epoch": 0.67,
-      "learning_rate": 5.1534431137724546e-05,
-      "loss": 3.0387,
-      "step": 230
-    },
-    {
-      "epoch": 0.7,
-      "learning_rate": 5.0411676646706584e-05,
-      "loss": 3.0311,
-      "step": 240
-    },
-    {
-      "epoch": 0.73,
-      "learning_rate": 4.9288922155688615e-05,
-      "loss": 3.0417,
-      "step": 250
-    },
-    {
-      "epoch": 0.73,
-      "eval_loss": 3.069390058517456,
-      "eval_runtime": 125.7339,
-      "eval_samples_per_second": 36.744,
-      "eval_steps_per_second": 4.597,
-      "eval_wer": 1.0,
-      "step": 250
-    },
-    {
-      "epoch": 0.75,
-      "learning_rate": 4.816616766467066e-05,
-      "loss": 3.0219,
-      "step": 260
-    },
-    {
-      "epoch": 0.78,
-      "learning_rate": 4.704341317365269e-05,
-      "loss": 3.0194,
-      "step": 270
-    },
-    {
-      "epoch": 0.81,
-      "learning_rate": 4.592065868263473e-05,
-      "loss": 2.9974,
-      "step": 280
-    },
-    {
-      "epoch": 0.84,
-      "learning_rate": 4.479790419161676e-05,
-      "loss": 2.9996,
-      "step": 290
-    },
-    {
-      "epoch": 0.87,
-      "learning_rate": 4.36751497005988e-05,
-      "loss": 3.0016,
-      "step": 300
-    },
-    {
-      "epoch": 0.87,
-      "eval_loss": 3.0347490310668945,
-      "eval_runtime": 132.3039,
-      "eval_samples_per_second": 34.92,
-      "eval_steps_per_second": 4.369,
-      "eval_wer": 1.0,
-      "step": 300
-    },
-    {
-      "epoch": 0.9,
-      "learning_rate": 4.255239520958083e-05,
-      "loss": 3.0052,
-      "step": 310
-    },
-    {
-      "epoch": 0.93,
-      "learning_rate": 4.142964071856287e-05,
-      "loss": 2.9826,
-      "step": 320
-    },
-    {
-      "epoch": 0.96,
-      "learning_rate": 4.030688622754491e-05,
-      "loss": 2.9747,
-      "step": 330
-    },
-    {
-      "epoch": 0.99,
-      "learning_rate": 3.918413173652694e-05,
-      "loss": 2.9617,
-      "step": 340
-    },
-    {
-      "epoch": 1.02,
-      "learning_rate": 3.806137724550898e-05,
-      "loss": 3.2053,
-      "step": 350
-    },
-    {
-      "epoch": 1.02,
-      "eval_loss": 2.984886407852173,
-      "eval_runtime": 149.1508,
-      "eval_samples_per_second": 30.975,
-      "eval_steps_per_second": 3.875,
-      "eval_wer": 1.0,
-      "step": 350
-    },
-    {
-      "epoch": 1.05,
-      "learning_rate": 3.693862275449102e-05,
-      "loss": 2.9665,
-      "step": 360
-    },
-    {
-      "epoch": 1.08,
-      "learning_rate": 3.581586826347305e-05,
-      "loss": 2.9641,
-      "step": 370
-    },
-    {
-      "epoch": 1.1,
-      "learning_rate": 3.469311377245509e-05,
-      "loss": 2.9484,
-      "step": 380
-    },
-    {
-      "epoch": 1.13,
-      "learning_rate": 3.3570359281437126e-05,
-      "loss": 2.9494,
-      "step": 390
-    },
-    {
-      "epoch": 1.16,
-      "learning_rate": 3.244760479041916e-05,
-      "loss": 2.9698,
-      "step": 400
-    },
-    {
-      "epoch": 1.16,
-      "eval_loss": 2.989494562149048,
-      "eval_runtime": 150.6903,
-      "eval_samples_per_second": 30.659,
-      "eval_steps_per_second": 3.836,
-      "eval_wer": 1.0,
-      "step": 400
-    },
-    {
-      "epoch": 1.19,
-      "learning_rate": 3.1324850299401195e-05,
-      "loss": 2.9664,
-      "step": 410
-    },
-    {
-      "epoch": 1.22,
-      "learning_rate": 3.020209580838323e-05,
-      "loss": 2.9494,
-      "step": 420
-    },
-    {
-      "epoch": 1.25,
-      "learning_rate": 2.9079341317365265e-05,
-      "loss": 2.935,
-      "step": 430
-    },
-    {
-      "epoch": 1.28,
-      "learning_rate": 2.7956586826347306e-05,
-      "loss": 2.9397,
-      "step": 440
-    },
-    {
-      "epoch": 1.31,
-      "learning_rate": 2.683383233532934e-05,
-      "loss": 2.9485,
-      "step": 450
-    },
-    {
-      "epoch": 1.31,
-      "eval_loss": 2.9584460258483887,
-      "eval_runtime": 140.6358,
-      "eval_samples_per_second": 32.851,
-      "eval_steps_per_second": 4.11,
-      "eval_wer": 1.0,
-      "step": 450
-    }
-  ],
-  "max_steps": 688,
-  "num_train_epochs": 2,
-  "total_flos": 1.336010418574825e+18,
-  "trial_name": null,
-  "trial_params": null
-}

{checkpoint-500 → checkpoint-5}/config.json

@@ -1,6 +1,6 @@
 {
-  "_name_or_path": "facebook/wav2vec2-xls-r-300m",
-  "activation_dropout": 0.1,
+  "_name_or_path": "hf-test/xls-r-dummy",
+  "activation_dropout": 0.0,
   "adapter_kernel_size": 3,
   "adapter_stride": 2,
   "add_adapter": false,
@@ -11,35 +11,23 @@
   "attention_dropout": 0.0,
   "bos_token_id": 1,
   "classifier_proj_size": 256,
-  "codevector_dim": 768,
+  "codevector_dim": 256,
   "contrastive_logits_temperature": 0.1,
-  "conv_bias": true,
+  "conv_bias": false,
   "conv_dim": [
-    512,
-    512,
-    512,
-    512,
-    512,
-    512,
-    512
+    32,
+    32,
+    32
   ],
   "conv_kernel": [
-    10,
-    3,
-    3,
-    3,
-    3,
-    2,
-    2
+    8,
+    8,
+    8
   ],
   "conv_stride": [
-    5,
-    2,
-    2,
-    2,
-    2,
-    2,
-    2
+    4,
+    4,
+    4
   ],
   "ctc_loss_reduction": "mean",
   "ctc_zero_infinity": false,
@@ -54,30 +42,31 @@
   "final_dropout": 0.0,
   "hidden_act": "gelu",
   "hidden_dropout": 0.0,
-  "hidden_size": 1024,
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 16,
   "initializer_range": 0.02,
-  "intermediate_size": 4096,
+  "intermediate_size": 20,
   "layer_norm_eps": 1e-05,
   "layerdrop": 0.0,
-  "mask_feature_length": 64,
+  "mask_feature_length": 10,
   "mask_feature_min_masks": 0,
-  "mask_feature_prob": 0.25,
+  "mask_feature_prob": 0.0,
   "mask_time_length": 10,
   "mask_time_min_masks": 2,
-  "mask_time_prob": 0.75,
+  "mask_time_prob": 0.05,
   "model_type": "wav2vec2",
   "num_adapter_layers": 3,
-  "num_attention_heads": 16,
+  "num_attention_heads": 2,
   "num_codevector_groups": 2,
   "num_codevectors_per_group": 320,
-  "num_conv_pos_embedding_groups": 16,
-  "num_conv_pos_embeddings": 128,
-  "num_feat_extract_layers": 7,
-  "num_hidden_layers": 24,
-  "num_negatives": 100,
-  "output_hidden_size": 1024,
-  "pad_token_id": 34,
-  "proj_codevector_dim": 768,
+  "num_conv_pos_embedding_groups": 2,
+  "num_conv_pos_embeddings": 16,
+  "num_feat_extract_layers": 3,
+  "num_hidden_layers": 4,
+  "num_negatives": 10,
+  "output_hidden_size": 16,
+  "pad_token_id": 50,
+  "proj_codevector_dim": 256,
   "tdnn_dilation": [
     1,
     2,
@@ -102,6 +91,6 @@
   "torch_dtype": "float32",
   "transformers_version": "4.16.0.dev0",
   "use_weighted_layer_sum": false,
-  "vocab_size": 37,
+  "vocab_size": 53,
   "xvector_output_dim": 512
 }

checkpoint-400/training_args.bin → checkpoint-5/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:77806a16ec6ef209c8a5c1c085159a0bc8e45c23f2d143f2c13e01527f13b5b2
-size 2991
+oid sha256:a245c50c71ce92148ecd921cac7789c454842de546c72d14ed7db3cfb2368f57
+size 130385

{checkpoint-400 → checkpoint-5}/preprocessor_config.json

@@ -3,7 +3,7 @@
   "feature_extractor_type": "Wav2Vec2FeatureExtractor",
   "feature_size": 1,
   "padding_side": "right",
-  "padding_value": 0,
-  "return_attention_mask": true,
+  "padding_value": 0.0,
+  "return_attention_mask": false,
   "sampling_rate": 16000
 }

checkpoint-400/scaler.pt → checkpoint-5/pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3828e6a0e34e076fa271e3dd1c08f47dc96711f380d9585d89de05befff54169
-size 559
+oid sha256:1f707ecc79ae941e56ffdb7641b649a684ab4c49c79e26a5c0e388853b9c3416
+size 143910

{checkpoint-400 → checkpoint-5}/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a426a909abceadbfb6682c295322fafea2b96fc77d190014718f2a843f386bc4
+oid sha256:ed2ba2d62f4307f8e3a73e07e52e7c639a10c1c25d1552bde4eee08c6914110c
 size 14567

{checkpoint-500 → checkpoint-5}/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0c8bceadebe118ff459b01a775a8e9b38a6b8302c162d022f78d3646163e6486
+oid sha256:9a648873cde1f456464b7741cbb832250d15acff5f65edbdbdecb0e392b63153
 size 559

{checkpoint-400 → checkpoint-5}/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b73abf21b39473fffe8d985c5918328870ddc7f68ec8e4b77b7e7ea44506a1b1
+oid sha256:1570f1540bbb5840ac5e01c84d0494b5fa441945dfd0c990d275c242d4173037
 size 623

checkpoint-5/trainer_state.json

@@ -0,0 +1,15 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.014204545454545454,
+  "global_step": 5,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [],
+  "max_steps": 10,
+  "num_train_epochs": 1,
+  "total_flos": 273921937920.0,
+  "trial_name": null,
+  "trial_params": null
+}

{checkpoint-450 → checkpoint-5}/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:77806a16ec6ef209c8a5c1c085159a0bc8e45c23f2d143f2c13e01527f13b5b2
+oid sha256:13f03d8c8ae9bd8496ddb1b6ef8d282d15b54e9aee2f515f000a57e9ac24d7c2
 size 2991

checkpoint-500/optimizer.pt

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:2dbd398c0993452243cb7df79d58793489c5b41d1a1000e547261b81959e2f45
-size 2490362385

checkpoint-500/preprocessor_config.json

@@ -1,9 +0,0 @@
-{
-  "do_normalize": true,
-  "feature_extractor_type": "Wav2Vec2FeatureExtractor",
-  "feature_size": 1,
-  "padding_side": "right",
-  "padding_value": 0,
-  "return_attention_mask": true,
-  "sampling_rate": 16000
-}

checkpoint-500/pytorch_model.bin

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ed159635bd947770842e25d915d8157fd198d2f06d5db2476c6663627e3beee7
-size 1262075377

checkpoint-500/rng_state.pth

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:32b82ac75b07e68dc3bc90e76f55a339f64dce724d87a9ae3c69ee46df441867
-size 14503

checkpoint-500/scheduler.pt

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:15019bcd1c8fc8b1bd39c46d9e1196c2fa76648918a0024eb84229f57debcf7e
-size 623

checkpoint-500/trainer_state.json

@@ -1,55 +0,0 @@
-{
-  "best_metric": null,
-  "best_model_checkpoint": null,
-  "epoch": 1.4525018129079044,
-  "global_step": 500,
-  "is_hyper_param_search": false,
-  "is_local_process_zero": true,
-  "is_world_process_zero": true,
-  "log_history": [
-    {
-      "epoch": 0.29,
-      "learning_rate": 6.613023952095809e-05,
-      "loss": 5.1206,
-      "step": 100
-    },
-    {
-      "epoch": 0.58,
-      "learning_rate": 5.490269461077844e-05,
-      "loss": 3.0901,
-      "step": 200
-    },
-    {
-      "epoch": 0.87,
-      "learning_rate": 4.36751497005988e-05,
-      "loss": 3.0224,
-      "step": 300
-    },
-    {
-      "epoch": 1.16,
-      "learning_rate": 3.244760479041916e-05,
-      "loss": 2.9922,
-      "step": 400
-    },
-    {
-      "epoch": 1.45,
-      "learning_rate": 2.1220059880239517e-05,
-      "loss": 2.9357,
-      "step": 500
-    },
-    {
-      "epoch": 1.45,
-      "eval_loss": 2.9458744525909424,
-      "eval_runtime": 138.8724,
-      "eval_samples_per_second": 33.268,
-      "eval_steps_per_second": 4.162,
-      "eval_wer": 1.0,
-      "step": 500
-    }
-  ],
-  "max_steps": 688,
-  "num_train_epochs": 2,
-  "total_flos": 1.4827194756605722e+18,
-  "trial_name": null,
-  "trial_params": null
-}

checkpoint-500/training_args.bin

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:07c8daeeea1ded5d5b75ab1c6033b9bed25c9ac1f192a365842399932683cfcc
-size 2991

config.json

@@ -1,6 +1,6 @@
 {
-  "_name_or_path": "facebook/wav2vec2-xls-r-300m",
-  "activation_dropout": 0.1,
+  "_name_or_path": "hf-test/xls-r-dummy",
+  "activation_dropout": 0.0,
   "adapter_kernel_size": 3,
   "adapter_stride": 2,
   "add_adapter": false,
@@ -11,35 +11,23 @@
   "attention_dropout": 0.0,
   "bos_token_id": 1,
   "classifier_proj_size": 256,
-  "codevector_dim": 768,
+  "codevector_dim": 256,
   "contrastive_logits_temperature": 0.1,
-  "conv_bias": true,
+  "conv_bias": false,
   "conv_dim": [
-    512,
-    512,
-    512,
-    512,
-    512,
-    512,
-    512
+    32,
+    32,
+    32
   ],
   "conv_kernel": [
-    10,
-    3,
-    3,
-    3,
-    3,
-    2,
-    2
+    8,
+    8,
+    8
   ],
   "conv_stride": [
-    5,
-    2,
-    2,
-    2,
-    2,
-    2,
-    2
+    4,
+    4,
+    4
   ],
   "ctc_loss_reduction": "mean",
   "ctc_zero_infinity": false,
@@ -54,30 +42,31 @@
   "final_dropout": 0.0,
   "hidden_act": "gelu",
   "hidden_dropout": 0.0,
-  "hidden_size": 1024,
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 16,
   "initializer_range": 0.02,
-  "intermediate_size": 4096,
+  "intermediate_size": 20,
   "layer_norm_eps": 1e-05,
   "layerdrop": 0.0,
-  "mask_feature_length": 64,
+  "mask_feature_length": 10,
   "mask_feature_min_masks": 0,
-  "mask_feature_prob": 0.25,
+  "mask_feature_prob": 0.0,
   "mask_time_length": 10,
   "mask_time_min_masks": 2,
-  "mask_time_prob": 0.75,
+  "mask_time_prob": 0.05,
   "model_type": "wav2vec2",
   "num_adapter_layers": 3,
-  "num_attention_heads": 16,
+  "num_attention_heads": 2,
   "num_codevector_groups": 2,
   "num_codevectors_per_group": 320,
-  "num_conv_pos_embedding_groups": 16,
-  "num_conv_pos_embeddings": 128,
-  "num_feat_extract_layers": 7,
-  "num_hidden_layers": 24,
-  "num_negatives": 100,
-  "output_hidden_size": 1024,
-  "pad_token_id": 34,
-  "proj_codevector_dim": 768,
+  "num_conv_pos_embedding_groups": 2,
+  "num_conv_pos_embeddings": 16,
+  "num_feat_extract_layers": 3,
+  "num_hidden_layers": 4,
+  "num_negatives": 10,
+  "output_hidden_size": 16,
+  "pad_token_id": 50,
+  "proj_codevector_dim": 256,
   "tdnn_dilation": [
     1,
     2,
@@ -102,6 +91,6 @@
   "torch_dtype": "float32",
   "transformers_version": "4.16.0.dev0",
   "use_weighted_layer_sum": false,
-  "vocab_size": 37,
+  "vocab_size": 53,
   "xvector_output_dim": 512
 }

eval_results.json

@@ -1,9 +0,0 @@
-{
-    "epoch": 2.0,
-    "eval_loss": 2.91414213180542,
-    "eval_runtime": 133.9783,
-    "eval_samples": 4620,
-    "eval_samples_per_second": 34.483,
-    "eval_steps_per_second": 4.314,
-    "eval_wer": 1.0
-}

preprocessor_config.json

@@ -3,7 +3,7 @@
   "feature_extractor_type": "Wav2Vec2FeatureExtractor",
   "feature_size": 1,
   "padding_side": "right",
-  "padding_value": 0,
-  "return_attention_mask": true,
+  "padding_value": 0.0,
+  "return_attention_mask": false,
   "sampling_rate": 16000
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5c01bc756f9d54db8c90d48e01252abf1115ffe23b3c7297aa9b48389b4e9132
-size 1262075377
+oid sha256:1f707ecc79ae941e56ffdb7641b649a684ab4c49c79e26a5c0e388853b9c3416
+size 143910

run-dummy-ab-gpu.sh

@@ -0,0 +1,22 @@
+python run_speech_recognition_ctc.py \
+--dataset_name="mozilla-foundation/common_voice_7_0" \
+--model_name_or_path="hf-test/xls-r-dummy" \
+--dataset_config_name="ab" \
+--output_dir="./" \
+--overwrite_output_dir \
+--max_steps="10" \
+--per_device_train_batch_size="2" \
+--learning_rate="3e-4" \
+--save_total_limit="1" \
+--evaluation_strategy="steps" \
+--text_column_name="sentence" \
+--length_column_name="input_length" \
+--save_steps="5" \
+--layerdrop="0.0" \
+--freeze_feature_encoder \
+--gradient_checkpointing \
+--fp16 \
+--group_by_length \
+--push_to_hub \
+--use_auth_token \
+--do_train --do_eval

run-dummy-sv-gpu.sh

@@ -1,7 +1,7 @@
 python run_speech_recognition_ctc.py \
 	--dataset_name="mozilla-foundation/common_voice_7_0" \
-	--model_name_or_path="facebook/wav2vec2-xls-r-300m" \
-	--dataset_config_name="sv-SE" \
+	--model_name_or_path="hf-test/xls-r-dummy" \
+	--dataset_config_name="ab" \
 	--output_dir="./" \
 	--overwrite_output_dir \
 	--max_steps="10" \