Add unpacked adapter files for direct from_pretrained loading
Browse files- config.json +36 -0
config.json
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"MixtralForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_dropout": 0.0,
|
| 6 |
+
"bos_token_id": 1,
|
| 7 |
+
"dtype": "bfloat16",
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"head_dim": null,
|
| 10 |
+
"hidden_act": "silu",
|
| 11 |
+
"hidden_size": 4096,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 14336,
|
| 14 |
+
"max_position_embeddings": 32768,
|
| 15 |
+
"model_type": "mixtral",
|
| 16 |
+
"num_attention_heads": 32,
|
| 17 |
+
"num_experts_per_tok": 2,
|
| 18 |
+
"num_hidden_layers": 32,
|
| 19 |
+
"num_key_value_heads": 8,
|
| 20 |
+
"num_local_experts": 8,
|
| 21 |
+
"output_router_logits": false,
|
| 22 |
+
"pad_token_id": 2,
|
| 23 |
+
"rms_norm_eps": 1e-05,
|
| 24 |
+
"rope_parameters": {
|
| 25 |
+
"rope_theta": 1000000.0,
|
| 26 |
+
"rope_type": "default"
|
| 27 |
+
},
|
| 28 |
+
"router_aux_loss_coef": 0.02,
|
| 29 |
+
"router_jitter_noise": 0.0,
|
| 30 |
+
"sliding_window": null,
|
| 31 |
+
"tie_word_embeddings": false,
|
| 32 |
+
"transformers_version": "5.10.1",
|
| 33 |
+
"use_cache": false,
|
| 34 |
+
"vocab_size": 32000,
|
| 35 |
+
"torch_dtype": "bfloat16"
|
| 36 |
+
}
|