Upload config.json with huggingface_hub

config.json

@@ -1,58 +1,28 @@
 {
-  "_name_or_path": "./result_and_evaluation/SSM_20M_little_dropout",
-  "architectures": [
-    "MambaModel"
-  ],
-  "attn_cfg": {},
-  "attn_layer_idx": [],
-  "bos_token_id": 1,
-  "conv_kernel": 4,
-  "d_intermediate": 0,
-  "d_model": 512,
-  "dropout_rate": 0.1,
-  "eos_token_id": 2,
-  "expand": 2,
-  "fused_add_norm": true,
-  "hidden_act": "silu",
-  "hidden_size": 768,
-  "id2label": {
-    "0": "LABEL_0"
-  },
-  "initializer_range": 0.1,
-  "intermediate_size": 1536,
-  "label2id": {
-    "LABEL_0": 0
-  },
-  "layer_norm_epsilon": 1e-05,
   "model_type": "mamba",
+  "d_model": 512,
+  "d_intermediate": 0,
   "n_layer": 6,
-  "num_hidden_layers": 32,
-  "pad_token_id": 3,
-  "pad_vocab_size_multiple": 8,
-  "rescale_prenorm_residual": false,
-  "residual_in_fp32": true,
-  "rms_norm": true,
+  "vocab_size": 1880,
   "ssm_cfg": {
     "layer": "Mamba2"
   },
-  "state_size": 16,
-  "summary_activation": "tanh",
-  "summary_first_dropout": 0.4,
-  "summary_hidden_size": 128,
-  "summary_proj_to_labels": true,
+  "attn_layer_idx": [],
+  "attn_cfg": {},
+  "rms_norm": true,
+  "residual_in_fp32": true,
+  "fused_add_norm": true,
+  "pad_vocab_size_multiple": 8,
+  "tie_embeddings": true,
+  "dropout_rate": 0.1,
+  "eos_token_id": 2,
+  "bos_token_id": 1,
+  "pad_token_id": 3,
+  "num_labels": 1,
   "summary_type": "cls_index",
   "summary_use_proj": true,
-  "tie_embeddings": true,
-  "time_step_floor": 0.0001,
-  "time_step_init_scheme": "random",
-  "time_step_max": 0.1,
-  "time_step_min": 0.001,
-  "time_step_rank": 48,
-  "time_step_scale": 1.0,
-  "torch_dtype": "float32",
-  "transformers_version": "4.40.2",
-  "use_bias": false,
-  "use_cache": true,
-  "use_conv_bias": true,
-  "vocab_size": 1880
+  "summary_activation": "tanh",
+  "summary_proj_to_labels": true,
+  "summary_first_dropout": 0.4,
+  "summary_hidden_size": 128
 }