waddledee's picture
Training in progress, step 100
86efcdf verified
{
"_name_or_path": "waddledee/myBit-Llama2-jp-127M-158b-hajime-0410",
"architectures": [
"LlamaForSequenceClassification"
],
"attention_bias": false,
"attention_dropout": 0.0,
"auto_map": {
"AutoConfig": "waddledee/myBit-Llama2-jp-127M-158b-hajime-0410--modeling_bit_llama.BitLlamaConfig",
"AutoModelForCausalLM": "waddledee/myBit-Llama2-jp-127M-158b-hajime-0410--modeling_bit_llama.BitLlamaForCausalLM"
},
"bitnet_type": "1.58b",
"bits": 8,
"bos_token_id": 1,
"eos_token_id": 2,
"hidden_act": "silu",
"hidden_size": 768,
"id2label": {
"0": "topic-news",
"1": "dokujo-tsushin",
"2": "livedoor-homme",
"3": "it-life-hack",
"4": "kaden-channel",
"5": "smax",
"6": "sports-watch",
"7": "peachy",
"8": "movie-enter"
},
"initializer_range": 0.02,
"intermediate_size": 1536,
"label2id": {
"dokujo-tsushin": 1,
"it-life-hack": 3,
"kaden-channel": 4,
"livedoor-homme": 2,
"movie-enter": 8,
"peachy": 7,
"smax": 5,
"sports-watch": 6,
"topic-news": 0
},
"max_position_embeddings": 1024,
"model_type": "llama",
"n_ctx": 128,
"num_attention_heads": 12,
"num_hidden_layers": 12,
"num_key_value_heads": 4,
"pad_token_id": 2,
"pretraining_tp": 1,
"problem_type": "single_label_classification",
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 10000.0,
"tie_word_embeddings": false,
"torch_dtype": "float32",
"transformers_version": "4.38.2",
"use_cache": true,
"vocab_size": 43176
}