config.json

{
  "_name_or_path": "bigscience/bloom-3b",
  "apply_residual_connection_post_layernorm": false,
  "architectures": [
    "BloomForCausalLM"
  ],
  "attention_dropout": 0.0,
  "attention_softmax_in_fp32": true,
  "bias_dropout_fusion": true,
  "bos_token_id": 1,
  "eos_token_id": 2,
  "hidden_dropout": 0.0,
  "hidden_size": 2560,
  "initializer_range": 0.02,
  "layer_norm_epsilon": 1e-05,
  "masked_softmax_fusion": true,
  "model_type": "bloom",
  "n_head": 32,
  "n_inner": null,
  "n_layer": 30,
  "offset_alibi": 100,
  "pad_token_id": 3,
  "pretraining_tp": 4,
  "quantization_config": {
    "batch_size": 1,
    "bits": 2,
    "block_name_to_quantize": "transformer.h",
    "damp_percent": 0.1,
    "dataset": "wikitext2",
    "desc_act": false,
    "disable_exllama": false,
    "group_size": 128,
    "model_seqlen": 2048,
    "module_name_preceding_first_block": [
      "transformer.word_embeddings",
      "transformer.word_embeddings_layernorm"
    ],
    "pad_token_id": null,
    "quant_method": "gptq",
    "sym": true,
    "tokenizer": null,
    "true_sequential": true,
    "use_cuda_fp16": true
  },
  "skip_bias_add": true,
  "skip_bias_add_qkv": false,
  "slow_but_exact": false,
  "torch_dtype": "float16",
  "transformers_version": "4.32.0",
  "unk_token_id": 0,
  "use_cache": true,
  "vocab_size": 250880
}