Upload 2 files
Browse files- config.json +22 -11
- generation_config.json +6 -3
config.json
CHANGED
|
@@ -5,23 +5,34 @@
|
|
| 5 |
"attention_bias": false,
|
| 6 |
"attention_dropout": 0.0,
|
| 7 |
"bos_token_id": 128000,
|
| 8 |
-
"eos_token_id":
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
"hidden_act": "silu",
|
| 10 |
-
"hidden_size":
|
| 11 |
"initializer_range": 0.02,
|
| 12 |
-
"intermediate_size":
|
| 13 |
-
"max_position_embeddings":
|
|
|
|
| 14 |
"model_type": "llama",
|
| 15 |
-
"num_attention_heads":
|
| 16 |
-
"num_hidden_layers":
|
| 17 |
-
"num_key_value_heads":
|
| 18 |
"pretraining_tp": 1,
|
| 19 |
"rms_norm_eps": 1e-05,
|
| 20 |
-
"rope_scaling":
|
| 21 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 22 |
"tie_word_embeddings": false,
|
| 23 |
-
"torch_dtype": "
|
| 24 |
-
"transformers_version": "4.
|
| 25 |
"use_cache": true,
|
| 26 |
"vocab_size": 128256
|
| 27 |
}
|
|
|
|
| 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 |
"hidden_act": "silu",
|
| 14 |
+
"hidden_size": 4096,
|
| 15 |
"initializer_range": 0.02,
|
| 16 |
+
"intermediate_size": 14336,
|
| 17 |
+
"max_position_embeddings": 131072,
|
| 18 |
+
"mlp_bias": false,
|
| 19 |
"model_type": "llama",
|
| 20 |
+
"num_attention_heads": 32,
|
| 21 |
+
"num_hidden_layers": 32,
|
| 22 |
+
"num_key_value_heads": 8,
|
| 23 |
"pretraining_tp": 1,
|
| 24 |
"rms_norm_eps": 1e-05,
|
| 25 |
+
"rope_scaling": {
|
| 26 |
+
"factor": 8.0,
|
| 27 |
+
"low_freq_factor": 1.0,
|
| 28 |
+
"high_freq_factor": 4.0,
|
| 29 |
+
"original_max_position_embeddings": 8192,
|
| 30 |
+
"rope_type": "llama3"
|
| 31 |
+
},
|
| 32 |
+
"rope_theta": 500000.0,
|
| 33 |
"tie_word_embeddings": false,
|
| 34 |
+
"torch_dtype": "bfloat16",
|
| 35 |
+
"transformers_version": "4.42.3",
|
| 36 |
"use_cache": true,
|
| 37 |
"vocab_size": 128256
|
| 38 |
}
|
generation_config.json
CHANGED
|
@@ -1,9 +1,12 @@
|
|
| 1 |
{
|
| 2 |
"bos_token_id": 128000,
|
| 3 |
-
"eos_token_id": [128001, 128009],
|
| 4 |
"do_sample": true,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
"temperature": 0.6,
|
| 6 |
-
"max_length": 4096,
|
| 7 |
"top_p": 0.9,
|
| 8 |
-
"transformers_version": "4.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"bos_token_id": 128000,
|
|
|
|
| 3 |
"do_sample": true,
|
| 4 |
+
"eos_token_id": [
|
| 5 |
+
128001,
|
| 6 |
+
128008,
|
| 7 |
+
128009
|
| 8 |
+
],
|
| 9 |
"temperature": 0.6,
|
|
|
|
| 10 |
"top_p": 0.9,
|
| 11 |
+
"transformers_version": "4.42.3"
|
| 12 |
}
|