Update config.json
Browse files- config.json +8 -11
config.json
CHANGED
|
@@ -1,15 +1,12 @@
|
|
| 1 |
{
|
|
|
|
| 2 |
"architectures": [
|
| 3 |
"LlamaForCausalLM"
|
| 4 |
],
|
| 5 |
"attention_bias": false,
|
| 6 |
"attention_dropout": 0.0,
|
| 7 |
"bos_token_id": 128000,
|
| 8 |
-
"eos_token_id":
|
| 9 |
-
128001,
|
| 10 |
-
128008,
|
| 11 |
-
128009
|
| 12 |
-
],
|
| 13 |
"head_dim": 64,
|
| 14 |
"hidden_act": "silu",
|
| 15 |
"hidden_size": 2048,
|
|
@@ -21,6 +18,7 @@
|
|
| 21 |
"num_attention_heads": 32,
|
| 22 |
"num_hidden_layers": 16,
|
| 23 |
"num_key_value_heads": 8,
|
|
|
|
| 24 |
"pretraining_tp": 1,
|
| 25 |
"rms_norm_eps": 1e-05,
|
| 26 |
"rope_scaling": {
|
|
@@ -32,12 +30,11 @@
|
|
| 32 |
},
|
| 33 |
"rope_theta": 500000.0,
|
| 34 |
"tie_word_embeddings": true,
|
| 35 |
-
"torch_dtype": "
|
| 36 |
-
"transformers_version": "4.
|
| 37 |
-
"
|
| 38 |
-
|
| 39 |
-
"kv_cache_dtype": "float16"
|
| 40 |
-
},
|
| 41 |
"use_cache": true,
|
| 42 |
"vocab_size": 128256
|
| 43 |
}
|
|
|
|
|
|
| 1 |
{
|
| 2 |
+
"_attn_implementation_autoset": true,
|
| 3 |
"architectures": [
|
| 4 |
"LlamaForCausalLM"
|
| 5 |
],
|
| 6 |
"attention_bias": false,
|
| 7 |
"attention_dropout": 0.0,
|
| 8 |
"bos_token_id": 128000,
|
| 9 |
+
"eos_token_id": 128009,
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
"head_dim": 64,
|
| 11 |
"hidden_act": "silu",
|
| 12 |
"hidden_size": 2048,
|
|
|
|
| 18 |
"num_attention_heads": 32,
|
| 19 |
"num_hidden_layers": 16,
|
| 20 |
"num_key_value_heads": 8,
|
| 21 |
+
"pad_token_id": 128004,
|
| 22 |
"pretraining_tp": 1,
|
| 23 |
"rms_norm_eps": 1e-05,
|
| 24 |
"rope_scaling": {
|
|
|
|
| 30 |
},
|
| 31 |
"rope_theta": 500000.0,
|
| 32 |
"tie_word_embeddings": true,
|
| 33 |
+
"torch_dtype": "float32",
|
| 34 |
+
"transformers_version": "4.51.3",
|
| 35 |
+
"unsloth_fixed": true,
|
| 36 |
+
"unsloth_version": "2025.4.7",
|
|
|
|
|
|
|
| 37 |
"use_cache": true,
|
| 38 |
"vocab_size": 128256
|
| 39 |
}
|
| 40 |
+
|