Training in progress, epoch 1

.gitignore

@@ -0,0 +1 @@
+checkpoint-*/

config.json

@@ -0,0 +1,353 @@
+{
+  "_name_or_path": "openai/whisper-medium",
+  "activation_dropout": 0.0,
+  "activation_function": "gelu",
+  "architectures": [
+    "WhisperForAudioClassification"
+  ],
+  "attention_dropout": 0.0,
+  "begin_suppress_tokens": [
+    220,
+    50257
+  ],
+  "bos_token_id": 50257,
+  "classifier_proj_size": 256,
+  "d_model": 1024,
+  "decoder_attention_heads": 16,
+  "decoder_ffn_dim": 4096,
+  "decoder_layerdrop": 0.0,
+  "decoder_layers": 24,
+  "decoder_start_token_id": 50258,
+  "dropout": 0.0,
+  "encoder_attention_heads": 16,
+  "encoder_ffn_dim": 4096,
+  "encoder_layerdrop": 0.0,
+  "encoder_layers": 24,
+  "eos_token_id": 50257,
+  "finetuning_task": "audio-classification",
+  "forced_decoder_ids": [
+    [
+      1,
+      50259
+    ],
+    [
+      2,
+      50359
+    ],
+    [
+      3,
+      50363
+    ]
+  ],
+  "id2label": {
+    "0": "af_za",
+    "1": "am_et",
+    "10": "ca_es",
+    "100": "yue_hant_hk",
+    "101": "zu_za",
+    "11": "ceb_ph",
+    "12": "ckb_iq",
+    "13": "cmn_hans_cn",
+    "14": "cs_cz",
+    "15": "cy_gb",
+    "16": "da_dk",
+    "17": "de_de",
+    "18": "el_gr",
+    "19": "en_us",
+    "2": "ar_eg",
+    "20": "es_419",
+    "21": "et_ee",
+    "22": "fa_ir",
+    "23": "ff_sn",
+    "24": "fi_fi",
+    "25": "fil_ph",
+    "26": "fr_fr",
+    "27": "ga_ie",
+    "28": "gl_es",
+    "29": "gu_in",
+    "3": "as_in",
+    "30": "ha_ng",
+    "31": "he_il",
+    "32": "hi_in",
+    "33": "hr_hr",
+    "34": "hu_hu",
+    "35": "hy_am",
+    "36": "id_id",
+    "37": "ig_ng",
+    "38": "is_is",
+    "39": "it_it",
+    "4": "ast_es",
+    "40": "ja_jp",
+    "41": "jv_id",
+    "42": "ka_ge",
+    "43": "kam_ke",
+    "44": "kea_cv",
+    "45": "kk_kz",
+    "46": "km_kh",
+    "47": "kn_in",
+    "48": "ko_kr",
+    "49": "ky_kg",
+    "5": "az_az",
+    "50": "lb_lu",
+    "51": "lg_ug",
+    "52": "ln_cd",
+    "53": "lo_la",
+    "54": "lt_lt",
+    "55": "luo_ke",
+    "56": "lv_lv",
+    "57": "mi_nz",
+    "58": "mk_mk",
+    "59": "ml_in",
+    "6": "be_by",
+    "60": "mn_mn",
+    "61": "mr_in",
+    "62": "ms_my",
+    "63": "mt_mt",
+    "64": "my_mm",
+    "65": "nb_no",
+    "66": "ne_np",
+    "67": "nl_nl",
+    "68": "nso_za",
+    "69": "ny_mw",
+    "7": "bg_bg",
+    "70": "oc_fr",
+    "71": "om_et",
+    "72": "or_in",
+    "73": "pa_in",
+    "74": "pl_pl",
+    "75": "ps_af",
+    "76": "pt_br",
+    "77": "ro_ro",
+    "78": "ru_ru",
+    "79": "sd_in",
+    "8": "bn_in",
+    "80": "sk_sk",
+    "81": "sl_si",
+    "82": "sn_zw",
+    "83": "so_so",
+    "84": "sr_rs",
+    "85": "sv_se",
+    "86": "sw_ke",
+    "87": "ta_in",
+    "88": "te_in",
+    "89": "tg_tj",
+    "9": "bs_ba",
+    "90": "th_th",
+    "91": "tr_tr",
+    "92": "uk_ua",
+    "93": "umb_ao",
+    "94": "ur_pk",
+    "95": "uz_uz",
+    "96": "vi_vn",
+    "97": "wo_sn",
+    "98": "xh_za",
+    "99": "yo_ng"
+  },
+  "init_std": 0.02,
+  "is_encoder_decoder": true,
+  "label2id": {
+    "af_za": "0",
+    "am_et": "1",
+    "ar_eg": "2",
+    "as_in": "3",
+    "ast_es": "4",
+    "az_az": "5",
+    "be_by": "6",
+    "bg_bg": "7",
+    "bn_in": "8",
+    "bs_ba": "9",
+    "ca_es": "10",
+    "ceb_ph": "11",
+    "ckb_iq": "12",
+    "cmn_hans_cn": "13",
+    "cs_cz": "14",
+    "cy_gb": "15",
+    "da_dk": "16",
+    "de_de": "17",
+    "el_gr": "18",
+    "en_us": "19",
+    "es_419": "20",
+    "et_ee": "21",
+    "fa_ir": "22",
+    "ff_sn": "23",
+    "fi_fi": "24",
+    "fil_ph": "25",
+    "fr_fr": "26",
+    "ga_ie": "27",
+    "gl_es": "28",
+    "gu_in": "29",
+    "ha_ng": "30",
+    "he_il": "31",
+    "hi_in": "32",
+    "hr_hr": "33",
+    "hu_hu": "34",
+    "hy_am": "35",
+    "id_id": "36",
+    "ig_ng": "37",
+    "is_is": "38",
+    "it_it": "39",
+    "ja_jp": "40",
+    "jv_id": "41",
+    "ka_ge": "42",
+    "kam_ke": "43",
+    "kea_cv": "44",
+    "kk_kz": "45",
+    "km_kh": "46",
+    "kn_in": "47",
+    "ko_kr": "48",
+    "ky_kg": "49",
+    "lb_lu": "50",
+    "lg_ug": "51",
+    "ln_cd": "52",
+    "lo_la": "53",
+    "lt_lt": "54",
+    "luo_ke": "55",
+    "lv_lv": "56",
+    "mi_nz": "57",
+    "mk_mk": "58",
+    "ml_in": "59",
+    "mn_mn": "60",
+    "mr_in": "61",
+    "ms_my": "62",
+    "mt_mt": "63",
+    "my_mm": "64",
+    "nb_no": "65",
+    "ne_np": "66",
+    "nl_nl": "67",
+    "nso_za": "68",
+    "ny_mw": "69",
+    "oc_fr": "70",
+    "om_et": "71",
+    "or_in": "72",
+    "pa_in": "73",
+    "pl_pl": "74",
+    "ps_af": "75",
+    "pt_br": "76",
+    "ro_ro": "77",
+    "ru_ru": "78",
+    "sd_in": "79",
+    "sk_sk": "80",
+    "sl_si": "81",
+    "sn_zw": "82",
+    "so_so": "83",
+    "sr_rs": "84",
+    "sv_se": "85",
+    "sw_ke": "86",
+    "ta_in": "87",
+    "te_in": "88",
+    "tg_tj": "89",
+    "th_th": "90",
+    "tr_tr": "91",
+    "uk_ua": "92",
+    "umb_ao": "93",
+    "ur_pk": "94",
+    "uz_uz": "95",
+    "vi_vn": "96",
+    "wo_sn": "97",
+    "xh_za": "98",
+    "yo_ng": "99",
+    "yue_hant_hk": "100",
+    "zu_za": "101"
+  },
+  "max_length": 448,
+  "max_source_positions": 1500,
+  "max_target_positions": 448,
+  "model_type": "whisper",
+  "num_hidden_layers": 24,
+  "num_mel_bins": 80,
+  "pad_token_id": 50257,
+  "scale_embedding": 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,
+    50360,
+    50361,
+    50362
+  ],
+  "torch_dtype": "float16",
+  "transformers_version": "4.27.0.dev0",
+  "use_cache": true,
+  "use_weighted_layer_sum": false,
+  "vocab_size": 51865
+}

preprocessor_config.json

@@ -0,0 +1,14 @@
+{
+  "chunk_length": 30,
+  "feature_extractor_type": "WhisperFeatureExtractor",
+  "feature_size": 80,
+  "hop_length": 160,
+  "n_fft": 400,
+  "n_samples": 480000,
+  "nb_max_frames": 3000,
+  "padding_side": "right",
+  "padding_value": 0.0,
+  "processor_class": "WhisperProcessor",
+  "return_attention_mask": false,
+  "sampling_rate": 16000
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3630bd8791baa8f7c8d725ebb8a257ba32420c7741a9fe6448826cb2effea411
+size 615058493

runs/Feb23_13-37-14_sanchit-t4-fleurs/1677159503.584576/events.out.tfevents.1677159503.sanchit-t4-fleurs.26232.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da751da5bf36be59098df00add531f0e8e31baf9fd0f8c91d17c5a67691b091f
+size 5840

runs/Feb23_13-37-14_sanchit-t4-fleurs/events.out.tfevents.1677159503.sanchit-t4-fleurs.26232.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ac8cb4dd3c6de5e0fece304cc822f5c19e5897e8bad8a23610fbb244d01f929
+size 62898

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:056cee6adb1b2f1fcd2f38aa61d20cb381ad85614636d6b75ea4483a58612531
+size 4731