Update config.json
Browse files- config.json +8 -14
config.json
CHANGED
|
@@ -3,7 +3,7 @@
|
|
| 3 |
"Mistral3ForConditionalGeneration"
|
| 4 |
],
|
| 5 |
"bos_token_id": 1,
|
| 6 |
-
"
|
| 7 |
"eos_token_id": 2,
|
| 8 |
"image_token_index": 10,
|
| 9 |
"model_type": "mistral3",
|
|
@@ -13,20 +13,17 @@
|
|
| 13 |
"spatial_merge_size": 2,
|
| 14 |
"text_config": {
|
| 15 |
"attention_dropout": 0.0,
|
| 16 |
-
"
|
| 17 |
-
"dtype": "bfloat16",
|
| 18 |
-
"eos_token_id": 2,
|
| 19 |
"head_dim": 128,
|
| 20 |
"hidden_act": "silu",
|
| 21 |
-
"hidden_size":
|
| 22 |
"initializer_range": 0.02,
|
| 23 |
-
"intermediate_size":
|
| 24 |
"max_position_embeddings": 262144,
|
| 25 |
"model_type": "ministral3",
|
| 26 |
"num_attention_heads": 32,
|
| 27 |
-
"num_hidden_layers":
|
| 28 |
"num_key_value_heads": 8,
|
| 29 |
-
"pad_token_id": 11,
|
| 30 |
"rms_norm_eps": 1e-05,
|
| 31 |
"rope_parameters": {
|
| 32 |
"beta_fast": 32.0,
|
|
@@ -41,18 +38,15 @@
|
|
| 41 |
"type": "yarn"
|
| 42 |
},
|
| 43 |
"sliding_window": null,
|
| 44 |
-
"tie_word_embeddings": true,
|
| 45 |
"use_cache": true,
|
| 46 |
"vocab_size": 131072
|
| 47 |
},
|
| 48 |
-
"tie_word_embeddings":
|
| 49 |
-
"transformers_version": "5.0.
|
| 50 |
"unsloth_fixed": true,
|
| 51 |
-
"unsloth_version": "2026.1.4",
|
| 52 |
-
"use_cache": false,
|
| 53 |
"vision_config": {
|
| 54 |
"attention_dropout": 0.0,
|
| 55 |
-
"
|
| 56 |
"head_dim": 64,
|
| 57 |
"hidden_act": "silu",
|
| 58 |
"hidden_size": 1024,
|
|
|
|
| 3 |
"Mistral3ForConditionalGeneration"
|
| 4 |
],
|
| 5 |
"bos_token_id": 1,
|
| 6 |
+
"torch_dtype": "bfloat16",
|
| 7 |
"eos_token_id": 2,
|
| 8 |
"image_token_index": 10,
|
| 9 |
"model_type": "mistral3",
|
|
|
|
| 13 |
"spatial_merge_size": 2,
|
| 14 |
"text_config": {
|
| 15 |
"attention_dropout": 0.0,
|
| 16 |
+
"torch_dtype": "bfloat16",
|
|
|
|
|
|
|
| 17 |
"head_dim": 128,
|
| 18 |
"hidden_act": "silu",
|
| 19 |
+
"hidden_size": 4096,
|
| 20 |
"initializer_range": 0.02,
|
| 21 |
+
"intermediate_size": 14336,
|
| 22 |
"max_position_embeddings": 262144,
|
| 23 |
"model_type": "ministral3",
|
| 24 |
"num_attention_heads": 32,
|
| 25 |
+
"num_hidden_layers": 34,
|
| 26 |
"num_key_value_heads": 8,
|
|
|
|
| 27 |
"rms_norm_eps": 1e-05,
|
| 28 |
"rope_parameters": {
|
| 29 |
"beta_fast": 32.0,
|
|
|
|
| 38 |
"type": "yarn"
|
| 39 |
},
|
| 40 |
"sliding_window": null,
|
|
|
|
| 41 |
"use_cache": true,
|
| 42 |
"vocab_size": 131072
|
| 43 |
},
|
| 44 |
+
"tie_word_embeddings": false,
|
| 45 |
+
"transformers_version": "5.0.0.dev0",
|
| 46 |
"unsloth_fixed": true,
|
|
|
|
|
|
|
| 47 |
"vision_config": {
|
| 48 |
"attention_dropout": 0.0,
|
| 49 |
+
"torch_dtype": "bfloat16",
|
| 50 |
"head_dim": 64,
|
| 51 |
"hidden_act": "silu",
|
| 52 |
"hidden_size": 1024,
|