config.json

{
  "transforms": [
    "huggingface_tokenize"
  ],
  "src_vocab_size": 51865,
  "data": null,
  "tgt_vocab_size": 51865,
  "decoder_start_token": "<|startoftranscript|>",
  "src_vocab": null,
  "n_sample": 0,
  "pad_token": "<|endoftext|>",
  "vocab_size_multiple": 8,
  "save_data": null,
  "bos_token": "<|endoftext|>",
  "tgt_vocab": null,
  "skip_empty_level": "silent",
  "eos_token": "<|endoftext|>",
  "unk_token": "<|endoftext|>",
  "share_vocab": true,
  "training": {
    "normalization": "tokens",
    "batch_type": "tokens",
    "accum_steps": [
      0
    ],
    "valid_batch_size": 256,
    "quant_type": "",
    "batch_size_multiple": 1,
    "group_size": 0,
    "quant_layers": [],
    "w_bit": 0,
    "batch_size": 896,
    "compute_dtype": "torch.float32",
    "accum_count": [
      32
    ]
  },
  "model": {
    "sliding_window": 4096,
    "no_timestamps_token_id": 50363,
    "first_k_dense_replace": 0,
    "attn_scaling": null,
    "layer_norm": "standard",
    "layers": 12,
    "num_experts": 0,
    "adapter_bias": false,
    "num_shared_experts": 0,
    "add_final_linear_bias": true,
    "share_decoder_embeddings": true,
    "left_pad": false,
    "num_experts_per_tok": 0,
    "transformer_ff": 3072,
    "architecture": "whisper",
    "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
    ],
    "parallel_residual": false,
    "position_encoding_type": "Learned",
    "add_ffnbias": true,
    "shared_layer_norm": false,
    "norm_eps": 1e-05,
    "heads": 12,
    "heads_kv": 12,
    "huggingface_model": "openai/whisper-small",
    "mlp_activation_fn": "gelu",
    "hidden_size": 768,
    "add_key_bias": false,
    "word_timestamp_heads": [
      [
        5,
        3
      ],
      [
        5,
        9
      ],
      [
        8,
        0
      ],
      [
        8,
        4
      ],
      [
        8,
        7
      ],
      [
        8,
        8
      ],
      [
        9,
        0
      ],
      [
        9,
        7
      ],
      [
        9,
        9
      ],
      [
        10,
        5
      ]
    ],
    "add_qkvbias": true,
    "moe_transformer_ff": null,
    "head_dim": null,
    "begin_suppress_tokens": [
      220,
      50257
    ],
    "generator_bias": false,
    "encoder": {
      "src_word_vec_size": 768,
      "norm_eps": 1e-05,
      "sliding_window": 4096,
      "heads": 12,
      "heads_kv": 12,
      "max_source_positions": 1500,
      "encoder_type": "whisper",
      "first_k_dense_replace": 0,
      "mlp_activation_fn": "gelu",
      "hidden_size": 768,
      "attn_scaling": null,
      "add_key_bias": false,
      "add_qkvbias": true,
      "layer_norm": "standard",
      "layers": 12,
      "num_mel_bins": 80,
      "moe_transformer_ff": null,
      "head_dim": null,
      "num_experts": 0,
      "num_shared_experts": 0,
      "add_final_linear_bias": true,
      "num_experts_per_tok": 0,
      "transformer_ff": 3072,
      "position_encoding_type": null,
      "parallel_residual": false,
      "add_ffnbias": true,
      "shared_layer_norm": false
    },
    "embeddings": {
      "src_word_vec_size": 768,
      "tgt_word_vec_size": 768,
      "n_positions": 448,
      "position_encoding_type": "Learned"
    },
    "decoder": {
      "norm_eps": 1e-05,
      "n_positions": 448,
      "sliding_window": 4096,
      "heads": 12,
      "heads_kv": 12,
      "first_k_dense_replace": 0,
      "mlp_activation_fn": "gelu",
      "tgt_word_vec_size": 768,
      "hidden_size": 768,
      "attn_scaling": null,
      "decoder_type": "transformer",
      "add_key_bias": false,
      "add_qkvbias": true,
      "layers": 12,
      "layer_norm": "standard",
      "moe_transformer_ff": null,
      "head_dim": null,
      "num_experts": 0,
      "num_shared_experts": 0,
      "add_final_linear_bias": true,
      "num_experts_per_tok": 0,
      "transformer_ff": 3072,
      "position_encoding_type": "Learned",
      "parallel_residual": false,
      "add_ffnbias": true,
      "shared_layer_norm": false
    }
  },
  "transforms_configs": {
    "huggingface_tokenize": {
      "huggingface_model": "openai/whisper-small",
      "path": "/scratch/models/whisper-small-eole/tokenizer.json"
    }
  },
  "inference": {
    "optional_eos": [],
    "max_length": 448,
    "chat_template": null
  }
}