Initial upload

README.md

@@ -0,0 +1,28 @@
+# Predicting music popularity using DNNs
+
+This is a pre-trained wav2vec2.0 model, trained on a fill Free Music Archive repository, created as part of DH-401: Digital Musicology class on EPFL
+
+## Team
+
+* Elisa (elisa.michelet@epfl.ch)
+* Michał (michal.bien@epfl.ch)
+* Noé (noe.durandard@epfl.ch)
+
+## Milestone 3
+
+Main notebook presenting out results is available [here](https://nbviewer.jupyter.org/github/Glorf/DH-401/blob/main/milestone3.ipynb)
+
+Notebook describing the details of Wav2Vec2.0 pre-training and fine-tuning for the task is available [here](https://nbviewer.jupyter.org/github/Glorf/DH-401/blob/main/milestone3-wav2vec2.ipynb)
+
+## Milestone 2
+
+Exploratory data analysis notebook is available [here](https://nbviewer.jupyter.org/github/Glorf/DH-401/blob/main/milestone2.ipynb)
+
+## Milestone 1
+
+Refined project proposal is available [here](https://github.com/Glorf/DH-401/blob/main/milestone0.md)
+
+
+## Milestone 0
+
+Original project proposal is available in git history [here](https://github.com/Glorf/DH-401/blob/bb14813ff2bbbd9cdc6b6eecf34c9e3c160598eb/milestone0.md)

config.json

@@ -0,0 +1,68 @@
+{
+  "activation_dropout": 0.1,
+  "apply_spec_augment": true,
+  "architectures": [
+    "Wav2Vec2Model"
+  ],
+  "attention_dropout": 0.1,
+  "bos_token_id": 1,
+  "conv_bias": false,
+  "conv_dim": [
+    512,
+    512,
+    512,
+    512,
+    512,
+    512,
+    512
+  ],
+  "conv_kernel": [
+    10,
+    3,
+    3,
+    3,
+    3,
+    2,
+    2
+  ],
+  "conv_stride": [
+    5,
+    2,
+    2,
+    2,
+    2,
+    2,
+    2
+  ],
+  "ctc_loss_reduction": "sum",
+  "ctc_zero_infinity": false,
+  "do_stable_layer_norm": false,
+  "eos_token_id": 2,
+  "feat_extract_activation": "gelu",
+  "feat_extract_dropout": 0.0,
+  "feat_extract_norm": "group",
+  "feat_proj_dropout": 0.1,
+  "final_dropout": 0.1,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout": 0.1,
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-05,
+  "layerdrop": 0.1,
+  "mask_feature_length": 10,
+  "mask_feature_prob": 0.0,
+  "mask_time_length": 10,
+  "mask_time_prob": 0.05,
+  "model_type": "wav2vec2",
+  "num_attention_heads": 12,
+  "num_conv_pos_embedding_groups": 16,
+  "num_conv_pos_embeddings": 128,
+  "num_feat_extract_layers": 7,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "transformers_version": "4.6.0.dev0",
+  "vocab_size": 32
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff5afc2d6459006682d6f3412946fd9b6554765084a9adb6c5b56b798e7ef2ac
+size 377569754