oamazonasgabriel/ minicpm-2b:mlx

490 1 month ago

A lightweight Ollama build of MiniCPM5-2B, the 2B-class open-source SOTA model from OpenBMB

tools thinking
... /
json
deb9ca33e863 ยท 886B
{
"architectures": [
"LlamaForCausalLM"
],
"bos_token_id": 0,
"eos_token_id": [
1,
130073
],
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 2048,
"initializer_range": 0.02,
"intermediate_size": 6144,
"max_position_embeddings": 131072,
"model_type": "llama",
"num_attention_heads": 16,
"num_hidden_layers": 42,
"num_key_value_heads": 2,
"pad_token_id": 1,
"quantization": {
"group_size": 64,
"bits": 4,
"mode": "affine"
},
"quantization_config": {
"group_size": 64,
"bits": 4,
"mode": "affine"
},
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 5000000,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "5.6.2",
"use_cache": true,
"vocab_size": 130560
}