update model

README.md

@@ -18,7 +18,7 @@ All the models have a comparable model size between 90 MB and 150 MB, BPE tokeni
 ### Model Sources
 
 - **Repository:** [Plant DNA LLMs](https://github.com/zhangtaolab/plant_DNA_LLMs)
-- **Manuscript:** [Versatile applications of foundation DNA large language models in plant genomes]() 
+- **Manuscript:** [PDLLMs: A group of tailored DNA large language models for analyzing plant genomes]() 
 
 ### Architecture
 

config.json

@@ -1 +1,48 @@
-{"d_model": 768, "n_layer": 24, "vocab_size": 11, "ssm_cfg": {}, "rms_norm": true, "residual_in_fp32": true, "fused_add_norm": true, "pad_vocab_size_multiple": 1, "tie_embeddings": true}
+{
+  "_name_or_path": "Plant_DNAMamba_singlebase_promoter_strength_leaf",
+  "architectures": [
+    "MambaForSequenceClassification"
+  ],
+  "bos_token_id": 0,
+  "conv_kernel": 4,
+  "d_inner": 1536,
+  "d_model": 768,
+  "eos_token_id": 0,
+  "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",
+  "n_layer": 24,
+  "num_hidden_layers": 24,
+  "pad_token_id": 0,
+  "pad_vocab_size_multiple": 8,
+  "problem_type": "regression",
+  "rescale_prenorm_residual": false,
+  "residual_in_fp32": true,
+  "rms_norm": true,
+  "ssm_cfg": {},
+  "state_size": 16,
+  "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": "4.45.0.dev0",
+  "use_bias": false,
+  "use_cache": false,
+  "use_conv_bias": true,
+  "use_mambapy": false,
+  "vocab_size": 11
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d06dcf69e83a9bb086ef2b48d53bfccd37dd51e6c8b74e0bd91f51f9c6e70e4
+size 362144276

test_metrics.json

@@ -1 +0,0 @@
-{'test_loss': 0.6664415001869202, 'test_r2': 0.754, 'test_spearmanr': 0.8639747438872241, 'test_runtime': 22.3306, 'test_samples_per_second': 320.368, 'test_steps_per_second': 20.062}