checkpoint-1500: config.json
Browse files- config.json +95 -0
config.json
ADDED
|
@@ -0,0 +1,95 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"MiniCPMSALAForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"attn_use_output_gate": true,
|
| 8 |
+
"attn_use_rope": false,
|
| 9 |
+
"auto_map": {
|
| 10 |
+
"AutoConfig": "configuration_minicpm_sala.MiniCPMSALAConfig",
|
| 11 |
+
"AutoModel": "modeling_minicpm_sala.MiniCPMSALAModel",
|
| 12 |
+
"AutoModelForCausalLM": "modeling_minicpm_sala.MiniCPMSALAForCausalLM",
|
| 13 |
+
"AutoModelForSeq2SeqLM": "modeling_minicpm_sala.MiniCPMSALAForCausalLM",
|
| 14 |
+
"AutoModelForSequenceClassification": "modeling_minicpm_sala.MiniCPMSALAForSequenceClassification"
|
| 15 |
+
},
|
| 16 |
+
"bos_token_id": 1,
|
| 17 |
+
"dim_model_base": 256,
|
| 18 |
+
"dtype": "bfloat16",
|
| 19 |
+
"eos_token_id": 73440,
|
| 20 |
+
"head_dim": 128,
|
| 21 |
+
"hidden_act": "silu",
|
| 22 |
+
"hidden_size": 4096,
|
| 23 |
+
"initializer_range": 0.1,
|
| 24 |
+
"intermediate_size": 16384,
|
| 25 |
+
"lightning_head_dim": 128,
|
| 26 |
+
"lightning_nh": 32,
|
| 27 |
+
"lightning_nkv": 32,
|
| 28 |
+
"lightning_scale": "1/sqrt(d)",
|
| 29 |
+
"lightning_use_rope": true,
|
| 30 |
+
"max_position_embeddings": 524288,
|
| 31 |
+
"mixer_types": [
|
| 32 |
+
"minicpm4",
|
| 33 |
+
"lightning-attn",
|
| 34 |
+
"lightning-attn",
|
| 35 |
+
"lightning-attn",
|
| 36 |
+
"lightning-attn",
|
| 37 |
+
"lightning-attn",
|
| 38 |
+
"lightning-attn",
|
| 39 |
+
"lightning-attn",
|
| 40 |
+
"lightning-attn",
|
| 41 |
+
"minicpm4",
|
| 42 |
+
"lightning-attn",
|
| 43 |
+
"lightning-attn",
|
| 44 |
+
"lightning-attn",
|
| 45 |
+
"lightning-attn",
|
| 46 |
+
"lightning-attn",
|
| 47 |
+
"lightning-attn",
|
| 48 |
+
"minicpm4",
|
| 49 |
+
"minicpm4",
|
| 50 |
+
"lightning-attn",
|
| 51 |
+
"lightning-attn",
|
| 52 |
+
"lightning-attn",
|
| 53 |
+
"lightning-attn",
|
| 54 |
+
"minicpm4",
|
| 55 |
+
"lightning-attn",
|
| 56 |
+
"lightning-attn",
|
| 57 |
+
"lightning-attn",
|
| 58 |
+
"lightning-attn",
|
| 59 |
+
"lightning-attn",
|
| 60 |
+
"lightning-attn",
|
| 61 |
+
"minicpm4",
|
| 62 |
+
"minicpm4",
|
| 63 |
+
"minicpm4"
|
| 64 |
+
],
|
| 65 |
+
"model_type": "minicpm_sala",
|
| 66 |
+
"mup_denominator": 32,
|
| 67 |
+
"num_attention_heads": 32,
|
| 68 |
+
"num_hidden_layers": 32,
|
| 69 |
+
"num_key_value_heads": 2,
|
| 70 |
+
"pad_token_id": 2,
|
| 71 |
+
"pretraining_tp": 1,
|
| 72 |
+
"qk_norm": true,
|
| 73 |
+
"rms_norm_eps": 1e-06,
|
| 74 |
+
"rope_scaling": null,
|
| 75 |
+
"rope_theta": 10000.0,
|
| 76 |
+
"scale_depth": 1.4,
|
| 77 |
+
"scale_emb": 12,
|
| 78 |
+
"shift_labels": true,
|
| 79 |
+
"sparse_config": {
|
| 80 |
+
"block_size": 64,
|
| 81 |
+
"dense_len": 8192,
|
| 82 |
+
"init_blocks": 1,
|
| 83 |
+
"kernel_size": 32,
|
| 84 |
+
"kernel_stride": 16,
|
| 85 |
+
"topk": 64,
|
| 86 |
+
"use_nope": false,
|
| 87 |
+
"window_size": 2048
|
| 88 |
+
},
|
| 89 |
+
"tie_word_embeddings": false,
|
| 90 |
+
"transformers_version": "4.57.6",
|
| 91 |
+
"use_cache": false,
|
| 92 |
+
"use_output_gate": true,
|
| 93 |
+
"use_output_norm": true,
|
| 94 |
+
"vocab_size": 73448
|
| 95 |
+
}
|