Hugging Face's logo

tiiuae
/
Falcon3-7B-Base-1.58bit

Text Generation
Transformers
Safetensors
llama
bitnet
falcon3
text-generation-inference
Inference Endpoints
8-bit precision
Model card Files Community
Falcon3-7B-Base-1.58bit
File size: 1,532 Bytes 
d99c914
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
 
{
  "_name_or_path": "/home/ubuntu/qiyang/outputs/checkpoint/Falcon3-7B-BitNet-FT-FW-5B/hf_bitnet",
  "activation": "swiglu",
  "architectures": [
    "LlamaForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "auto_map": {
    "AutoConfig": "tiiuae/falcon3-7B--configuration_llama.LlamaConfig",
    "AutoModel": "tiiuae/falcon3-7B--modeling_llama.LlamaModel",
    "AutoModelForCausalLM": "tiiuae/falcon3-7B--modeling_llama.LlamaForCausalLM",
    "AutoModelForQuestionAnswering": "tiiuae/falcon3-7B--modeling_llama.LlamaForQuestionAnswering",
    "AutoModelForSequenceClassification": "tiiuae/falcon3-7B--modeling_llama.LlamaForSequenceClassification",
    "AutoModelForTokenClassification": "tiiuae/falcon3-7B--modeling_llama.LlamaForTokenClassification"
  },
  "bos_token_id": 11,
  "eos_token_id": 11,
  "head_dim": 256,
  "hidden_act": "silu",
  "hidden_size": 3072,
  "initializer_range": 0.02,
  "intermediate_size": 23040,
  "is_bitnet_config": true,
  "max_position_embeddings": 8192,
  "mlp_bias": false,
  "model_type": "llama",
  "num_attention_heads": 12,
  "num_hidden_layers": 28,
  "num_key_value_heads": 4,
  "parallel_attn": false,
  "pretraining_tp": 1,
  "quantization_config": {
    "modules_to_not_convert": [
      "lm_head"
    ],
    "quant_method": "bitnet"
  },
  "rms_norm_eps": 1e-06,
  "rope_scaling": null,
  "rope_theta": 1000042,
  "tie_word_embeddings": false,
  "torch_dtype": "bfloat16",
  "transformers_version": "4.46.0.dev0",
  "use_cache": true,
  "vocab_size": 131072
}