upload model_config.json
Browse files- model_config.json +58 -0
model_config.json
ADDED
|
@@ -0,0 +1,58 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model_name": "functiongemma-270m",
|
| 3 |
+
"architecture": "gemma3",
|
| 4 |
+
"hidden_size": 640,
|
| 5 |
+
"num_hidden_layers": 18,
|
| 6 |
+
"num_layers": 18,
|
| 7 |
+
"num_attention_heads": 4,
|
| 8 |
+
"num_key_value_heads": 1,
|
| 9 |
+
"head_dim": 256,
|
| 10 |
+
"intermediate_size": 2048,
|
| 11 |
+
"vocab_size": 262144,
|
| 12 |
+
"context_length": 2048,
|
| 13 |
+
"sliding_window": 512,
|
| 14 |
+
"sliding_window_pattern": 6,
|
| 15 |
+
"layer_types": [
|
| 16 |
+
"sliding_attention",
|
| 17 |
+
"sliding_attention",
|
| 18 |
+
"sliding_attention",
|
| 19 |
+
"sliding_attention",
|
| 20 |
+
"sliding_attention",
|
| 21 |
+
"full_attention",
|
| 22 |
+
"sliding_attention",
|
| 23 |
+
"sliding_attention",
|
| 24 |
+
"sliding_attention",
|
| 25 |
+
"sliding_attention",
|
| 26 |
+
"sliding_attention",
|
| 27 |
+
"full_attention",
|
| 28 |
+
"sliding_attention",
|
| 29 |
+
"sliding_attention",
|
| 30 |
+
"sliding_attention",
|
| 31 |
+
"sliding_attention",
|
| 32 |
+
"sliding_attention",
|
| 33 |
+
"full_attention"
|
| 34 |
+
],
|
| 35 |
+
"embed_scale": 25.298221281347036,
|
| 36 |
+
"rope_theta_global": 1000000.0,
|
| 37 |
+
"rope_theta_local": 10000.0,
|
| 38 |
+
"query_pre_attn_scalar": 256.0,
|
| 39 |
+
"rms_norm_eps": 1e-06,
|
| 40 |
+
"bos_token_id": 2,
|
| 41 |
+
"eos_token_id": [
|
| 42 |
+
1,
|
| 43 |
+
50
|
| 44 |
+
],
|
| 45 |
+
"tie_word_embeddings": true,
|
| 46 |
+
"final_logit_softcapping": 0.0,
|
| 47 |
+
"parts": {
|
| 48 |
+
"model": "model.mlpackage"
|
| 49 |
+
},
|
| 50 |
+
"quantization": "fp16",
|
| 51 |
+
"compute_units": "CPU_AND_NE",
|
| 52 |
+
"tokenizer_repo": "google/functiongemma-270m-it",
|
| 53 |
+
"chat_format": "functiongemma",
|
| 54 |
+
"function_call_markers": {
|
| 55 |
+
"start": "<start_function_call>",
|
| 56 |
+
"end": "<end_function_call>"
|
| 57 |
+
}
|
| 58 |
+
}
|