PatchTST trained on 1,600,000,000 samples

config.json

@@ -0,0 +1,50 @@
+{
+  "activation_function": "gelu",
+  "architectures": [
+    "PatchTSTForPretraining"
+  ],
+  "attention_dropout": 0.0,
+  "bias": true,
+  "channel_attention": true,
+  "channel_consistent_masking": false,
+  "context_length": 24,
+  "d_model": 512,
+  "distribution_output": "student_t",
+  "do_mask_input": true,
+  "dropout": 0.1,
+  "dtype": "float32",
+  "ff_dropout": 0.0,
+  "ffn_dim": 2048,
+  "head_dropout": 0.1,
+  "init_std": 0.02,
+  "loss": "mse",
+  "mask_type": "random",
+  "mask_value": 0,
+  "model_type": "patchtst",
+  "norm_eps": 1e-05,
+  "norm_type": "layernorm",
+  "num_attention_heads": 8,
+  "num_forecast_mask_patches": [
+    2
+  ],
+  "num_hidden_layers": 6,
+  "num_input_channels": 14,
+  "num_parallel_samples": 100,
+  "num_targets": 1,
+  "output_range": null,
+  "patch_length": 6,
+  "patch_stride": 6,
+  "path_dropout": 0.0,
+  "pooling_type": "mean",
+  "positional_dropout": 0.0,
+  "positional_encoding_type": "sincos",
+  "pre_norm": true,
+  "prediction_length": 24,
+  "random_mask_ratio": 0.4,
+  "scaling": "mean",
+  "share_embedding": true,
+  "share_projection": true,
+  "transformers_version": "4.57.3",
+  "unmasked_channel_indices": null,
+  "use_cls_token": true
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f330193a82abe7c7914e970d024f4eb58617f4c286173190d88ac7b5c54c86bc
+size 75733360