config.json

{
  "_name_or_path": "distributed/llama-1b",
  "all_reduce_scores": {
    "0": "NOT_ALIVE",
    "1": "NON_PARTICIPATING",
    "10": "NOT_ALIVE",
    "100": "NOT_ALIVE",
    "101": "NON_PARTICIPATING",
    "102": "NON_PARTICIPATING",
    "103": "NON_PARTICIPATING",
    "104": "NON_PARTICIPATING",
    "105": "NON_PARTICIPATING",
    "106": "NON_PARTICIPATING",
    "107": "NOT_ALIVE",
    "108": "NOT_ALIVE",
    "109": "NOT_ALIVE",
    "11": "NON_PARTICIPATING",
    "110": "NON_PARTICIPATING",
    "111": "NON_PARTICIPATING",
    "112": "SUCCESS",
    "113": "NOT_ALIVE",
    "114": "NON_PARTICIPATING",
    "115": "NOT_ALIVE",
    "116": "NOT_ALIVE",
    "117": "SUCCESS",
    "118": "NOT_ALIVE",
    "119": "NON_PARTICIPATING",
    "12": "NON_PARTICIPATING",
    "120": "NON_PARTICIPATING",
    "121": "NOT_ALIVE",
    "122": "NON_PARTICIPATING",
    "123": "NOT_ALIVE",
    "124": "NON_PARTICIPATING",
    "125": "NON_PARTICIPATING",
    "126": "NOT_ALIVE",
    "127": "NOT_ALIVE",
    "128": "NOT_ALIVE",
    "129": "NON_PARTICIPATING",
    "13": "NOT_ALIVE",
    "130": "NOT_ALIVE",
    "131": "NOT_ALIVE",
    "132": "SUCCESS",
    "133": "SUCCESS",
    "134": "NON_PARTICIPATING",
    "135": "NOT_ALIVE",
    "136": "NON_PARTICIPATING",
    "137": "NON_PARTICIPATING",
    "138": "SUCCESS",
    "139": "NON_PARTICIPATING",
    "14": "NOT_ALIVE",
    "140": "NOT_ALIVE",
    "141": "NOT_ALIVE",
    "142": "NON_PARTICIPATING",
    "143": "NON_PARTICIPATING",
    "144": "NOT_ALIVE",
    "145": "NON_PARTICIPATING",
    "146": "NON_PARTICIPATING",
    "147": "NON_PARTICIPATING",
    "148": "NON_PARTICIPATING",
    "149": "NON_PARTICIPATING",
    "15": "NOT_ALIVE",
    "150": "NOT_ALIVE",
    "151": "NOT_ALIVE",
    "152": "NOT_ALIVE",
    "153": "NON_PARTICIPATING",
    "154": "NOT_ALIVE",
    "155": "NOT_ALIVE",
    "156": "NOT_ALIVE",
    "157": "NON_PARTICIPATING",
    "158": "NON_PARTICIPATING",
    "159": "SUCCESS",
    "16": "NOT_ALIVE",
    "160": "NON_PARTICIPATING",
    "161": "NOT_ALIVE",
    "162": "NON_PARTICIPATING",
    "163": "NOT_ALIVE",
    "164": "SUCCESS",
    "165": "NON_PARTICIPATING",
    "166": "NON_PARTICIPATING",
    "167": "NOT_ALIVE",
    "168": "SUCCESS",
    "169": "NOT_ALIVE",
    "17": "NOT_ALIVE",
    "170": "NON_PARTICIPATING",
    "171": "NOT_ALIVE",
    "172": "NON_PARTICIPATING",
    "173": "SUCCESS",
    "174": "NOT_ALIVE",
    "175": "NOT_ALIVE",
    "176": "NOT_ALIVE",
    "177": "NOT_ALIVE",
    "178": "NON_PARTICIPATING",
    "179": "NOT_ALIVE",
    "18": "NOT_ALIVE",
    "180": "NOT_ALIVE",
    "181": "NOT_ALIVE",
    "182": "NOT_ALIVE",
    "183": "NON_PARTICIPATING",
    "184": "NON_PARTICIPATING",
    "185": "NON_PARTICIPATING",
    "186": "NON_PARTICIPATING",
    "187": "NON_PARTICIPATING",
    "188": "NON_PARTICIPATING",
    "189": "NOT_ALIVE",
    "19": "NOT_ALIVE",
    "190": "NON_PARTICIPATING",
    "191": "NOT_ALIVE",
    "192": "NOT_ALIVE",
    "193": "NON_PARTICIPATING",
    "194": "NON_PARTICIPATING",
    "195": "NON_PARTICIPATING",
    "196": "NON_PARTICIPATING",
    "197": "NOT_ALIVE",
    "198": "NON_PARTICIPATING",
    "199": "NON_PARTICIPATING",
    "2": "NON_PARTICIPATING",
    "20": "NON_PARTICIPATING",
    "200": "NOT_ALIVE",
    "201": "NOT_ALIVE",
    "202": "NOT_ALIVE",
    "203": "NOT_ALIVE",
    "204": "NON_PARTICIPATING",
    "205": "NON_PARTICIPATING",
    "206": "NON_PARTICIPATING",
    "207": "NOT_ALIVE",
    "208": "NON_PARTICIPATING",
    "209": "NOT_ALIVE",
    "21": "NOT_ALIVE",
    "210": "SUCCESS",
    "211": "NON_PARTICIPATING",
    "212": "NON_PARTICIPATING",
    "213": "NON_PARTICIPATING",
    "214": "NON_PARTICIPATING",
    "215": "SUCCESS",
    "216": "NOT_ALIVE",
    "217": "NON_PARTICIPATING",
    "218": "NON_PARTICIPATING",
    "219": "NOT_ALIVE",
    "22": "NON_PARTICIPATING",
    "220": "NON_PARTICIPATING",
    "221": "NOT_ALIVE",
    "222": "NOT_ALIVE",
    "223": "NON_PARTICIPATING",
    "224": "NOT_ALIVE",
    "225": "NON_PARTICIPATING",
    "226": "NOT_ALIVE",
    "227": "NON_PARTICIPATING",
    "228": "NON_PARTICIPATING",
    "229": "NON_PARTICIPATING",
    "23": "NON_PARTICIPATING",
    "230": "NOT_ALIVE",
    "231": "NOT_ALIVE",
    "232": "NOT_ALIVE",
    "233": "NON_PARTICIPATING",
    "234": "NON_PARTICIPATING",
    "235": "NOT_ALIVE",
    "236": "NOT_ALIVE",
    "237": "NON_PARTICIPATING",
    "238": "SUCCESS",
    "239": "NOT_ALIVE",
    "24": "NOT_ALIVE",
    "240": "NOT_ALIVE",
    "241": "NOT_ALIVE",
    "242": "NON_PARTICIPATING",
    "243": "NOT_ALIVE",
    "244": "NON_PARTICIPATING",
    "245": "NON_PARTICIPATING",
    "246": "NON_PARTICIPATING",
    "247": "NON_PARTICIPATING",
    "248": "NOT_ALIVE",
    "249": "NOT_ALIVE",
    "25": "SUCCESS",
    "250": "NOT_ALIVE",
    "251": "NOT_ALIVE",
    "252": "NON_PARTICIPATING",
    "253": "NOT_ALIVE",
    "254": "NON_PARTICIPATING",
    "255": "NOT_ALIVE",
    "26": "NON_PARTICIPATING",
    "27": "NON_PARTICIPATING",
    "28": "NOT_ALIVE",
    "29": "NOT_ALIVE",
    "3": "NON_PARTICIPATING",
    "30": "NON_PARTICIPATING",
    "31": "NON_PARTICIPATING",
    "32": "NOT_ALIVE",
    "33": "NOT_ALIVE",
    "34": "NOT_ALIVE",
    "35": "SUCCESS",
    "36": "NOT_ALIVE",
    "37": "NOT_ALIVE",
    "38": "SUCCESS",
    "39": "SUCCESS",
    "4": "NOT_ALIVE",
    "40": "NON_PARTICIPATING",
    "41": "NON_PARTICIPATING",
    "42": "NON_PARTICIPATING",
    "43": "NOT_ALIVE",
    "44": "NON_PARTICIPATING",
    "45": "NOT_ALIVE",
    "46": "NOT_ALIVE",
    "47": "NON_PARTICIPATING",
    "48": "NOT_ALIVE",
    "49": "NON_PARTICIPATING",
    "5": "NOT_ALIVE",
    "50": "NON_PARTICIPATING",
    "51": "NON_PARTICIPATING",
    "52": "NON_PARTICIPATING",
    "53": "NON_PARTICIPATING",
    "54": "NON_PARTICIPATING",
    "55": "NOT_ALIVE",
    "56": "NOT_ALIVE",
    "57": "NOT_ALIVE",
    "58": "NON_PARTICIPATING",
    "59": "NON_PARTICIPATING",
    "6": "NOT_ALIVE",
    "60": "SUCCESS",
    "61": "NON_PARTICIPATING",
    "62": "NON_PARTICIPATING",
    "63": "SUCCESS",
    "64": "NOT_ALIVE",
    "65": "NON_PARTICIPATING",
    "66": "NOT_ALIVE",
    "67": "NOT_ALIVE",
    "68": "SUCCESS",
    "69": "NOT_ALIVE",
    "7": "SUCCESS",
    "70": "NON_PARTICIPATING",
    "71": "NON_PARTICIPATING",
    "72": "NON_PARTICIPATING",
    "73": "SUCCESS",
    "74": "NOT_ALIVE",
    "75": "NON_PARTICIPATING",
    "76": "NON_PARTICIPATING",
    "77": "NON_PARTICIPATING",
    "78": "NOT_ALIVE",
    "79": "NOT_ALIVE",
    "8": "NOT_ALIVE",
    "80": "NON_PARTICIPATING",
    "81": "NON_PARTICIPATING",
    "82": "NON_PARTICIPATING",
    "83": "NON_PARTICIPATING",
    "84": "NON_PARTICIPATING",
    "85": "NON_PARTICIPATING",
    "86": "NOT_ALIVE",
    "87": "NON_PARTICIPATING",
    "88": "NON_PARTICIPATING",
    "89": "NOT_ALIVE",
    "9": "NOT_ALIVE",
    "90": "NOT_ALIVE",
    "91": "NOT_ALIVE",
    "92": "NOT_ALIVE",
    "93": "NON_PARTICIPATING",
    "94": "NOT_ALIVE",
    "95": "NOT_ALIVE",
    "96": "SUCCESS",
    "97": "NON_PARTICIPATING",
    "98": "NOT_ALIVE",
    "99": "NOT_ALIVE"
  },
  "architectures": [
    "LlamaForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "block_list": [
    5961456,
    5961472
  ],
  "bos_token_id": 1,
  "eos_token_id": 2,
  "hidden_act": "silu",
  "hidden_size": 2048,
  "initializer_range": 0.02,
  "inner_step": 7,
  "intermediate_size": 5632,
  "last_allreduce_block": 5961214,
  "max_position_embeddings": 2048,
  "mlp_bias": false,
  "model_type": "llama",
  "num_attention_heads": 32,
  "num_hidden_layers": 22,
  "num_key_value_heads": 4,
  "pretraining_tp": 1,
  "rms_norm_eps": 1e-05,
  "rope_scaling": null,
  "rope_theta": 10000.0,
  "tie_word_embeddings": false,
  "torch_dtype": "float32",
  "transformers_version": "4.39.3",
  "use_cache": false,
  "vocab_size": 32000
}