Instructions to use JakeOh/LLaDA-70M-Base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use JakeOh/LLaDA-70M-Base with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("feature-extraction", model="JakeOh/LLaDA-70M-Base", trust_remote_code=True)# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("JakeOh/LLaDA-70M-Base", trust_remote_code=True, dtype="auto") - Notebooks
- Google Colab
- Kaggle
Upload config
Browse files- config.json +4 -4
config.json
CHANGED
|
@@ -10,8 +10,8 @@
|
|
| 10 |
"attention_layer_norm_with_affine": true,
|
| 11 |
"auto_map": {
|
| 12 |
"AutoConfig": "configuration_llada.LLaDAConfig",
|
| 13 |
-
"AutoModel": "
|
| 14 |
-
"AutoModelForCausalLM": "
|
| 15 |
},
|
| 16 |
"bias_for_layer_norm": false,
|
| 17 |
"block_group_size": 1,
|
|
@@ -19,7 +19,7 @@
|
|
| 19 |
"d_model": 768,
|
| 20 |
"dtype": "float32",
|
| 21 |
"embedding_dropout": 0.0,
|
| 22 |
-
"embedding_size":
|
| 23 |
"eos_token_id": 9,
|
| 24 |
"flash_attention": false,
|
| 25 |
"include_bias": false,
|
|
@@ -50,6 +50,6 @@
|
|
| 50 |
"scale_logits": false,
|
| 51 |
"transformers_version": "4.57.1",
|
| 52 |
"use_cache": false,
|
| 53 |
-
"vocab_size":
|
| 54 |
"weight_tying": false
|
| 55 |
}
|
|
|
|
| 10 |
"attention_layer_norm_with_affine": true,
|
| 11 |
"auto_map": {
|
| 12 |
"AutoConfig": "configuration_llada.LLaDAConfig",
|
| 13 |
+
"AutoModel": "modeling_llada.LLaDAModelLM",
|
| 14 |
+
"AutoModelForCausalLM": "modeling_llada.LLaDAModelLM"
|
| 15 |
},
|
| 16 |
"bias_for_layer_norm": false,
|
| 17 |
"block_group_size": 1,
|
|
|
|
| 19 |
"d_model": 768,
|
| 20 |
"dtype": "float32",
|
| 21 |
"embedding_dropout": 0.0,
|
| 22 |
+
"embedding_size": 50259,
|
| 23 |
"eos_token_id": 9,
|
| 24 |
"flash_attention": false,
|
| 25 |
"include_bias": false,
|
|
|
|
| 50 |
"scale_logits": false,
|
| 51 |
"transformers_version": "4.57.1",
|
| 52 |
"use_cache": false,
|
| 53 |
+
"vocab_size": 50259,
|
| 54 |
"weight_tying": false
|
| 55 |
}
|